THESE DE DOCTORAT Université Paris VI- Pierre

Transcription

THESE DE DOCTORAT
Présentée et soutenue publiquement pour obtention du titre de DOCTEUR des
Université Paris VI- Pierre et Marie Curie
&
Académie de la Science et de la Technologie du Vietnam
(Co-tutelle)
Spécialité: Biogéochimie des hydrosystèmes
Ecole Doctorale : Géoscience et Ressources Naturelles
LE Thi Phuong Quynh
FONCTIONNEMENT BIOGEOCHIMIQUE DU FLEUVE ROUGE
(NORD –VIETNAM) : BILANS ET MODELISATION
Soutenue le 7 Juillet 2005
Composition du jury:
M. Venu ITTEKKOT
Prof. Dr, CTME, Bremen, Allemagne
Rapporteur
M. Quang Cu BUI
Prof. Dr, VAST, HoChiMinh, Vietnam
Rapporteur
M. Georges VACHAUD
Prof. Dr, CNRS, Grenoble, France
Rapporteur
M. Ghislain DE MARSILY
Prof., Univ. Paris VI, Paris, France
Examinateur
M. Wolfgang LUDWIG
Dr, CEFREM, Perpignan, France
Examinateur
Mme. Josette GARNIER
Dr, CNRS- Univ. Paris VI, Paris, France Directrice de thèse
M. Gilles BILLEN
Dr, CNRS- Univ. Paris VI, Paris, France Directeur de thèse
M. Van Minh CHAU
Prof. Dr, VAST, Hanoi, Vietnam
Co-Directeur de thèse
Thèse préparée au sein des laboratoires
Sisyphe, UMR 7619, CNRS (France) – INPC, VAST (Vietnam)
THESIS
Written and defended for obtaining the doctorate degree of
Pierre et Marie Curie University (France)
&
Vietnamese Academy of Science and Technology (Vietnam)
(Co-supervision)
Speciality: Biogeochemistry of hydrosystems
PhD School: Geoscience and Natural Resources
LE Thi Phuong Quynh
BIOGEOCHEMICAL FUNCTIONING OF THE RED RIVER
(NORTH VIETNAM): BUDGETS AND MODELLING
Defended on July 7th 2005
Composition of the Committee:
M. Venu ITTEKKOT
Prof. Dr, CTME, Bremen, Germany
Reporter
M. Quang Cu BUI
Prof. Dr, VAST, HoChiMinh, Vietnam
Reporter
M. Georges VACHAUD
Prof. Dr, CNRS, Grenoble, France
Reporter
M. Ghislain DE MARSILY
Prof., Univ. Paris VI, Paris, France
Examinator
M. Wolfgang LUDWIG
Dr, CEFREM, Perpignan, France
Examinator
Mme. Josette GARNIER
Dr, CNRS- Univ. Paris VI, Paris, France
Advisor
M. Gilles BILLEN
Dr, CNRS- Univ. Paris VI, Paris, France
Advisor
M. Van Minh CHAU
Prof. Dr, VAST, Hanoi, Vietnam
Co- Advisor
This thesis is prepared at the laboratories
Sisyphe, UMR 7619, CNRS (France) – INPC, VAST (Vietnam)
Acknowledgements
Acknowledgements
First of all, I am extremely grateful to my advisors Dr. Josette Garnier and Dr. Gilles
Billen for accepting me as their PhD student and for their enormous assistance and helpful
discussions during my thesis. They have helped me in understanding concepts in a simple and
intuitive way, while introducing me to new ideas. They always know how to solve the
problems and always encourage me during the difficult periods. They have always offered me
special cares during my stays in France so that I could feel happy and comfortable. I would
express my particularly thanks to them.
I would like to thanks my Vietnamese co-advisor, Prof. Dr. Chau Van Minh who gives me the
opportunity to work in the ESPOIR project and to realize the cotutelle Ph.D. thesis. He always
provides me the favorable working conditions in the Institute of Natural Products Chemistry
(INPC). Without his helps in experiments, samplings and administrative papers in INPC in
Vietnam, the thesis would never be finished.
The PhD thesis was performed in the ESPOIR project, a French-Vietnamese program for
water quality and water treatment in the period from 2000 to 2004. I would like to thank Prof.
Georges Vachaud, Prof. Chau Van Minh and Prof. Nguyen The Dong to give me the chance
to pursue this thesis in the framework of the ESPOIR project.
I am also indebted to Prof. Ghislain de Marsily, the ex-director of the Ecole Doctorale
“Géosciences et Ressources Naturelles” for analysing my Vietnamese degree courses and
accepting my inscription. I must also thank Prof. Laurent Jolivet, the present director of the
Ecole Doctorale for his kindness with the administrative forms that permits my continuing
during the last period of this thesis.
Furthermore, I am deeply thankful to all the members of the jury: Prof. Venu Ittekkot, Prof.
Georges Vachaud, Prof. Bui Quang Cu, Prof. Ghislain De Marsily, Dr Wolfgang Ludwig, Dr.
Josette Garnier, Dr Gilles Billen and Prof. Chau Van Minh, who gave many interesting and
helpful comments and critics for my thesis manuscript and also for the enrichment of my
scientific knowledge.
During this work, I have been granted by the French Embassy in Vietnam at Hanoi. I would
like to express my thanks to the French Embassy in Vietnam and I especially thank Mr Bruno
Paing, attached to the cooperation of Science and Technology for his interest in this
programme and for always helping me with kindness in finding administrative solutions.
This work is a cotutelle thesis. I would also like to acknowledge the Leaders of Institute of
Natural Products Chemistry, the Leaders of University of Pierre and Marie Curie, who
i
Acknowledgements
permitted me to carry out this work. Helpful financial supports was provided by the Direction
of the International Cooperation of the Pierre and Marie Curie University.
I wish to extend a sincere gratitude to the director of Sisyphe laboratory, Prof. Alain Tabbagh,
to give me the warm welcome in this laboratory.
I express my sincere thanks to Sylvain Théry, a very humorous, friendly and hard working
person, for his huge helps, especially in the Red River data base elaboration, logical programs
creation and map drawing.
Moreover, I would like to thank the kind colleagues Nguyen Van Tuan, Tran Bich Nga and
Nguyen Van Tue in Meteorological and Hydrological Institute for their useful helps in
Vietnamese meteo-hydrological information. I would like to thank the sympathetic colleagues
in the Son Tay, Yen Bai, Hoa Binh, Vu Quang hydrological stations for their helps in water
samplings.
Among all the numerous people who have contributed to valuable ideas and experiments
related to this work, I would like to mention Dr Michel Meybeck and Dr Agnes Ducharne
(Sisyphe), Dr Pham Van Cu (Institute of Geography in Vietnam), Dr. Pham Huu Dien (Hanoi,
Pedagogic University I), Dr Nguyen Kien Cuong, Prof. Ngo Ngoc Cat and Dr Nguyen Thanh
Van (VAST). I express a deep gratitude to them.
I also wish to express my gratitude to my colleagues in the Institute of Natural Products
Chemistry: Luong, Thao, M. Ha; they have given so much help in the sampling campaigns
and sample analyses. My thanks are also due to Nicolas Prieur, who spent two years as a
CNRS Engineer making the link between French and Vietnamese team, and had a large
contribution in the organisation of the sampling campaigns for Nhue-Tolich urban rivers in
the framework of the ESPOIR project.
At Sisyphe, I truly thank Nadine and Valérie at the management and secretaryship, Maya
responsible for the informatics. I would like to sincerely acknowledge the generous assistance
provided by the following colleagues: Maïa, Séverine, Anun, Mohamed, Samia, Maïté. My
thanks are sent also to all these so kind friends: Agata, Véronique, Aurélie, Julien, Harouna,
Denis, Anne, Angelbert, Hans, Noémi, D. Thuy, Tam … for cheering me during the four halfyear stays in France.
At last but above all, I would like to extend my sentiments to my closest relatives. I am
greatly indebted to my parents and sisters for their morale and love supports as well as their
confidence in my scientific orientations and decisions. I am happy to get diploma, but my
parents are proud of that. I am deeply grateful to my husband who no only always
understands, believes and encourages me, but also helps me. My best Vietnamese friends
Binh, Trang, Long, Phong, Vu, Loi, Thuc are thanked for the wonderful days we spent
together at school and/or University and for their continuous encouragements.
ii
Résumé
Résumé
Le Fleuve Rouge (au Nord Vietnam et en Chine méridionale) couvre une surface de bassin
versant de 156 450 km2, avec une population de près de 30 millions d’habitants. L’axe
principal du Fleuve Rouge (aussi appelé Yuan, Thao ou Hong) reçoit deux affluents
principaux, le Da et le Lo, puis forme un large delta avant de se jeter dans le Golfe du Tonkin
(en Mer de Chine méridionale). Les trois sous-bassins supérieurs et le delta diffèrent
largement en terme de densité de population (de 101 hab.km-2 dans les bassins amont à plus
de 1000 hab.km-2 dans le delta), d’usage du sol et de pratiques agricoles.
Le but général du présent travail est de développer une compréhension d’ensemble du
fonctionnement biogéochimique de ce système sub-tropical de dimension régionale, et de
son contrôle par les processus naturels et anthropiques. L’épine dorsale du travail a consisté
dans l’implémentation du modèle RIVERSTRAHLER, développé antérieurement pour décrire le
lien entre la qualité de l’eau et les activités humaines dans le bassin de la Seine et d’autres
fleuves européens (Billen et al., 1994, 1997, 1999, 2005; Garnier et al., 1995, 1999, 2000,
2002), pour le cas particulier du système Fleuve Rouge.
La première étape dans cette étude a consisté dans la modélisation du régime hydrologique
et du transport solide du Fleuve Rouge (Le Thi Phuong Quynh et al., subm). Les
estimations antérieures de la charge solide du Fleuve Rouge variaient entre 100 et 170 106
t.an-1, c-à-d de 640 à 1060 t.km-².an-1. La forte dépendance du transport solide à l’hydrologie
est responsable d’une large variabilité inter-annuelle. Sur la base de données hydrologiques
relatives à la période 1997-2004, et d’un suivi journalier de la matière en suspension à
l’exutoire des 3 principaux tributaires du Fleuve Rouge en 2003, un modèle simplifié a été
établi pour estimer la charge solide moyenne interannuelle du Fleuve Rouge sous les
conditions actuelles. La valeur obtenue est de 40 106 t.an-1, correspondant à une charge
spécifique de 280 t.km-2.an-1. Elle reflète une réduction de 70% de la charge solide totale
suite à la mise en eau des réservoirs de Hoa Binh et de Thac Ba réservoirs dans les années
1980s. Le modèle prévoit une réduction supplémentaire de 20% de la charge en suspension
suite à la construction planifiée de deux grands réservoirs supplémentaires. Utilisant les
mesures de contenu en phosphore total dans la matière en suspension réalisées dans ce travail,
le flux de phosphore exporté par le Fleuve Rouge peut être estimé à 36 106 kgP an-1.
Les données de concentrations en nutriments dans le réseau hydrographique du Fleuve Rouge
étant assez rares, un suivi de la concentration des formes de l’azote, du phosphore, de la
silice, du carbone organique et de la chlorophylle à l’exutoire des principaux sous-bassins
amont, dans l’axe principal du Fleuve dans le delta et dans quelques rivières polluées de la
iii
Résumé
région d’Hanoï, a été réalisé à une fréquence mensuelle durant les années 2003 et 2004,
permettant de définir le niveau général de concentration en nutriments dans les eaux de
surface.
En vue d’examiner le degré de perturbation anthropique du cycle de l’azote et du phosphore
à l’échelle du bassin, des bilans de ces deux éléments ont été établis pour le système sol et
pour l’hydrosystème des 4 principaux sous-ensembles (Da, Lo, Thao et Delta) du bassin du
Fleuve Rouge (Le Thi Phuong Quynh et al., 2005). En terme de production agricole, d’une
part, de consommation de nourriture et de fourrage d’autre part, les sous-basins amont
apparaissent comme des systèmes autotrophes, exportant des produits agricoles, tandis que le
delta dépend d’importations de biens agricoles. Le bilan des sols agricoles révèle de fortes
pertes d’azote, principalement attribuables à la dénitrification dans les rizières, et de
phosphore, principalement dues à l’érosion. Le bilan du réseau hydrographique montre une
importante rétention/élimination d’azote (de 62 à 77 % dans les basins amont et de 59 % dans
le delta), et de phosphore, avec un taux de rétention de plus de 80 % dans le Da et le Lo, à
l’aval desquels sont localisés les grands réservoirs (Hoa Binh sur le Da et Thac Ba sur le Lo).
L’exportation spécifique estimée à l’exutoire du Fleuve Rouge est estimée à 855 kg.km-².an-1
d’azote total et 325 kg.km-².an-1 de phosphore total. L’azote plutôt que le phosphore semble
être l’élément limitant principal de la croissance algale dans les zones côtières influencées par
le Fleuve Rouge dans le Golfe du Tonkin.
Une base de données sous SIG a été assemblée à l’échelle du bassin du Fleuve Rouge, avec
des couches d’informations renseignant la géomorphologie du bassin, sa lithologie, la
météorologie, l’usage du sol et les pratiques agricoles, la population et les rejets d’eau usées
domestiques et industrielles. Cette base de données est conforme au format requis par le
logiciel SENEQUE/Riverstrahler (Ruelland et al, 2004), une version du modèle Riverstrahler
encapsulée dans une interface SIG constituant un outil de modélisation générique et
spatialement explicite de la qualité de l’eau à l’échelle des grands réseaux hydrographiques.
La première application de ce logiciel au système Fleuve Rouge est décrite et validée sur la
base des données acquises lors des suivis mensuels de qualité d’eau à l’exutoire des grands
sous-bassins et sur l’axe principal du Fleuve lors des années 2003 et 2004.
Enfin, le modèle a été utilisé pour explorer l’effet, en terme de qualité de l’eau et de
fonctionnement biogéochimique de divers scénarios décrivant de possibles changements
futurs du bassin du Fleuve Rouge concernant son aménagement hydraulique, l’usage de ses
sols et son agriculture, sa population et sa gestion des eaux usées.
Mots clés: Rivière Tropicale, Fleuve Rouge, Vietnam, modèle Riverstrahler/Sénèque,
nutriments, cycle de l’azote, du phosphore, de la silice, charge solide.
iv
Résumé
Summary
The Red River (in North Vietnam and South China) covers a watershed area of 156 450 km2
with a total population near 30 million inhabitants. The main branch of the Red River (also
called Yuan, Thao or Hong River) receives two major tributaries, the Da and Lo Rivers, then
forms a large delta before discharging into the Tonkin Bay (South China Sea). The 3 upstream
sub-basins and the Delta area differ widely in population density (from 101 inhab km-2 in the
upstream basins to more than 1000 inhab km-2 in the delta), land use and agricultural
practices.
The general goal of this work is to develop a comprehensive understanding of the
biogeochemical functioning of this sub-tropical regional system, and its control by natural
and anthropogenic processes. The backbone of the work consisted in implementing the
RIVERSTRAHLER Model, previously developed for describing the link between water quality
and human activities in the watershed in the Seine river and other European river systems
(Billen et al., 1994, 1997, 1999, 2005 ; Garnier et al, 1995, 1999,2000, 2002) to the special
case of the Red River system.
The first step of the study consisted in modeling the hydrological regime and the suspended
solid transport of the Red River (Le Thi Phuong Quynh et al., subm). Previous estimates of
its suspended matter loading range from 100 to 170 106 t.yr-1, i.e. from 640 to 1060 t.km-².yr-1.
The strong dependence of suspended solid transport on hydrology results in a large year-toyear variability. Based on available data on the hydrology over the period 1997-2004, and on
one -year survey of the daily suspended matter of the three main tributaries of the Red River
system in 2003, a simplified modeling approach is established to estimate the mean suspended
loading of the Red River under present conditions. The obtained value is 40 106 t.yr-1,
corresponding to a specific load of 280 t.km-2.yr-1. It reflects a 70% decrease of the total
suspended load since the impoundment of the Hoa Binh and Thac Ba reservoirs in the
1980’ies. The model predicts a further reduction by 20% of the suspended loading of the Red
River with the planned construction of two additional reservoirs. Using measurements of the
total phosphorus content of the suspended material in the different Red River tributaries, we
could estimate the present phosphorus delivery by the Red River as 36 106 kgP yr-1.
As data on nutrient concentration in the Red River drainage network are rather scarce, a
survey of nutrient concentration (N, P, Si, organic carbon and chlorophyll a) at the outlet of
the three main sub-basins, the main branch in the delta and some polluted rivers in the Hanoi
v
Résumé
region was carried on at monthly intervals in 2003 and 2004, allowing to define the general
levels of nutrient concentrations in surface water.
In order to examine the degree of human-induced alteration of the nitrogen and phosphorus
cycles at the scale of the watershed, budgets of these elements were established for the soil
and the drainage network of the 4 main sub-basins (Da, Lo, Thao and Delta) of the Red River
(Le Thi Phuong Quynh et al., 2005). In terms of agricultural production, on the one hand, and
consumption of food and feed on the other, the upstream sub-basins are autotrophic systems,
exporting agricultural goods, while the delta is a heterotrophic system, depending on
agricultural goods imports. The budget of the agricultural soils reveals great losses of
nitrogen, mostly attributable to denitrification in rice paddy fields and of phosphorus, mostly
caused by erosion. The budget of the drainage network shows high retention/elimination of
nitrogen (from 62 to 77 % in the upstream basins and 59 % in the delta), and of phosphorus,
with retention rates as high as 80 % in the Da and Lo sub-basins which have large reservoirs
in their downstream course (Hoa Binh on the Da and Thac Ba on the Lo). The total specific
delivery estimated at the outlet of the whole Red River System is 855 kg.km-².y-1 total N and
325 kg.km-².yr-1 total P. Nitrogen rather than phosphorus seems to be the potential limiting
factor of algal growth in the plume of the Red River in Tonkin Bay.
A GIS data base has been assembled at the scale of the whole Red River basin, with layers
documenting geomorphology, lithology, meteorology, land-use and agriculture, population,
domestic and industrial wastewater release, etc. This data base follows the format required
for running the SENEQUE/Riverstrahler software (Ruelland et al, 2004), a version of the
Riverstrahler model encapsulated into a GIS interface in order to build a generic and spatially
explicit water quality modelling tool. The first application of this model to the Red River
system is described and validated with the data acquired by the monthly surveys of water
quality at the outlet of the 3 sub-basins and in the main branch of the Red River during the
years 2003 and 2004.
Finally, the model is used to explore the effect in terms of water quality and biogeochemical
functioning of a variety of scenarios describing possible future changes in the Red River
basin concerning hydrological management, land use and agricultural practices, population
increase and wastewater treatment policy.
Key words: tropical river, Red River, Vietnam, Riverstrahler/Seneque model, nutrient
budgets, nitrogen, phosphorus, silica cycle, suspended solids.
vi
Résumé
Tãm t¾t
L−u vùc s«ng Hång n»m trªn ®Þa phËn miÒn B¾c ViÖt Nam vµ miÒn Nam Trung Quèc víi diÖn
tÝch toµn l−u vùc kho¶ng 156 450 km2 vµ d©n sè trong toµn l−u vùc ®¹t 30 triÖu ng−êi. Nh¸nh
chÝnh cña s«ng Hång (cßn gäi lµ s«ng Nguyªn, Thao, Çi, Hång) nhËn hai nh¸nh s«ng kh¸c lµ
s«ng §µ vµ s«ng L« t¹i ViÖt tr×, vµ b¾t ®Çu t¹o vïng ®ång b»ng ch©u thæ réng lín tr−íc khi ®æ
ra vÞnh B¾c Bé (biÓn §«ng). Ba tiÓu l−u vùc th−îng nguån vµ tiÓu l−u vùc ®ång b»ng hoµn
toµn kh¸c nhau vÒ mËt ®é d©n sè (tõ 101 ng−êi/km2 t¹i vïng th−îng nguån ®Õn h¬n 1000
ng−êi/km2 t¹i vïng ®ång b»ng ch©u thæ), vÒ t×nh h×nh sö dông ®Êt vµ çc ho¹t ®éng n«ng
nghiÖp trong tiÓu l−u vùc.
Môc tiªu chung cña luËn ¸n lµ ph¸t triÓn sù hiÓu biÕt vÒ çc ho¹t ®éng sinh th¸i ®Þa hãa cña
hÖ thèng b¸n nhiÖt ®íi chÞu t¸c ®éng cña çc qu¸ tr×nh tù nhiªn vµ cña con ng−êi. M« h×nh
RIVERSTRAHLER tr−íc ®©y ®· ®−îc x©y dùng ®Ó m« t¶ mèi quan hÖ gi÷a chÊt l−îng n−íc
vµ ho¹t ®éng cña con ng−êi trong l−u vùc s«ng Seine vµ mét sè l−u vùc s«ng lín ë Ch©u ¢u
(Billen et al., 1994, 1999, 2001; Garnier et al., 1995, 1999, 2002), lÇn ®Çu tiªn ®−îc ¸p dông
cho hÖ thèng s«ng nhiÖt ®íi, s«ng Hång.
B−íc ®Çu tiªn cña luËn ¸n lµ nghiªn cøu chÕ ®é thñy v¨n vµ chuyÓn t¶i hµm l−îng phï sa
trong hÖ thèng s«ng Hång. Çc nghiªn cøu tr−íc ®©y cho r»ng mçi n¨m s«ng Hång chuyÓn t¶i
ra biÓn kho¶ng 100-170tÊn, tøc lµ vµo kho¶ng 640-1060 tÊn/km2/n¨m. Sù phô thuéc m¹nh mÏ
cña hµm l−îng phï sa vµo chÕ ®é thñy v¨n ®· t¹o ra sù kh¸c biÖt râ rÖt vÒ tæng l−îng phï sa
chuyÓn t¶i ra biÓn hµng n¨m. Dùa vµo çc sè liÖu thu thËp ®−îc vÒ chÕ ®é thñy v¨n trong giai
®o¹n 1997-2004 vµ sè liÖu hµng ngµy vÒ hµm l−îng phï sa trong n¨m 2003 t¹i ba nh¸nh chÝnh
cña s«ng Hång, mét m« h×nh ®¬n gi¶n hãa ®· ®−îc thiÕt lËp ®Ó ®¸nh gi¸ t¶i l−îng trung b×nh
vÒ hµm l−îng phï sa víi çc ®iÒu kiÖn hiÖn t¹i. KÕt qu¶ cho thÊy h»ng n¨m s«ng Hång ®æ ra
biÓn kho¶ng 40.106tÊn/n¨m, tøc lµ kho¶ng 280 tÊn/km2/n¨m. §iÒu nµy ph¶n ¸nh 70% tæng
l−îng phï sa ®· bÞ gi¶m tõ khi cã sù vËn hµnh cña hå Hßa B×nh vµ hå Th¸c Bµ vµo nh÷ng n¨m
1980s. KÕt qu¶ dù b¸o cña m« h×nh cho thÊy sÏ cã kho¶ng thªm 20% tæng l−îng phï sa sÏ bÞ
gi¶m khi cã thªm 2 hå chøa n÷a ®i vµo ho¹t ®éng (S¬n La vµ §¹i ThÞ). Sö dông çc phÐp ®¸nh
gi¸ vÒ tæng l−îng phètpho trong phï sa t¹i çc nh¸nh chÝnh kh¸c nhau cña s«ng Hång, cho
thÊy, hiÖn nay mçi n¨m, s«ng Hång chuyÓn ra biÓn kho¶ng 36 106 kgP/n¨m.
Do thiÕu çc d÷ liÖu vÒ hµm l−îng chÊt dinh d−ìng trong m¹ng l−íi s«ng Hång nªn quan tr¾c
hµm l−îng çc chÊt dinh d−ìng (N, P, Si, Cacbon h÷u c¬ vµ chlorophyll a) t¹i çc h¹ nguån
cña ba nh¸nh s«ng chÝnh vµ trªn trôc chÝnh ë vïng ®ång b»ng vµ mét sè s«ng « nhiÔm t¹i Hµ
vii
Résumé
néi ®· ®−îc thùc hiÖn hµng th¸ng trong suèt hai n¨m 2003-2004, cho phÐp x¸c ®Þnh møc ®é
chung vÒ chÊt l−îng n−íc s«ng Hång.
Môc tiªu thø hai cña luËn ¸n lµ ®¸nh gi¸ møc ®é ¶nh h−ëng cña con ng−êi trong l−u vùc tíi
chu tr×nh nit¬ vµ phèpho. C©n b»ng dinh d−ìng cña hai nguyªn tè nµy ®−îc thiÕt lËp trong
bèn tiÓu l−u vùc §µ, L«, Thao vµ vïng ®ång b»ng cña hÖ thèng s«ng Hång. VÒ mÆt s¶n xuÊt
n«ng nghiÖp vµ tiªu thô l−¬ng thùc vµ thùc phÈm, çc tiÓu l−u vùc th−îng nguån ®−îc ®¸nh
gi¸ lµ çc hÖ thèng tù d−ìng, tøc lµ cã kh¶ n¨ng xuÊt khÈu hµng n«ng nghiÖp, trong khi vïng
®ång b»ng s«ng Hång l¹i ®−îc ®¸nh gi¸ lµ hÖ thèng dÞ d−ìng, phô thuéc vµo hµng n«ng
nghiÖp nhËp khÈu vµo l−u vùc. Nghiªn cøu vÒ c©n b»ng dinh d−ìng trong vïng ®Êt n«ng
nghiÖp cho thÊy nit¬ bÞ mÊt mét l−îng lín, hÇu hÕt lµ do qu¸ tr×nh khö nirat hãa trong vïng
®Êt trång lóa, trong khi l−îng phètpho mÊt chñ yÕu lµ do qu¸ tr×nh xãi mßn ®Êt. Nghiªn cøu vÒ
c©n b»ng dinh d−ìng trong hÖ thèng thñy v¨n cho thÊy qu¸ tr×nh l−u gi÷/lo¹i bá nit¬ diÔn ra
rÊt m¹nh (tõ 62-77% ë vïng th−îng nguån vµ 59% ë vïng ®ång b»ng) cßn phètpho th× ®−îc
l−u gi÷ rÊt nhiÒu trong çc hå chøa (Hßa B×nh, Th¸c Bµ) trong çc tiÓu l−u vùc s«ng §µ vµ
s«ng L«. T¶i l−îng tæng nit¬ vµ tæng phètpho chuyÓn t¶i ra biÓn cña toµn bé hÖ thèng s«ng
Hång ®−îc −íc tÝnh kho¶ng 855 kg/km²/n¨m vµ 325 kg/km²/n¨m. Nit¬ cã kh¶ n¨ng lµ yÕu tè
giíi h¹n sù ph¸t triÓn cña t¶o t¹i vÞnh B¾c Bé h¬n lµ phètpho.
HÖ d÷ liÖu GIS cña toµn bé l−u vùc s«ng víi çc líp vÒ ®Þa m¹o, ®Þa chÊt, thæ nh−ìng, khÝ
hËu, sö dông ®Êt- çc ho¹t ®éng n«ng nghiÖp, d©n sè, n−íc th¶i sinh ho¹t, c«ng nghiÖp … ®·
®−îc tËp hîp. HÖ d÷ liÖu nµy ®ßi hái d¹ng format ®Æc biÖt ®Ó cã thÓ ch¹y trong phÇn mÒm
SENEQUE/Riverstrahler (Ruelland, 2004), phiªn b¶n cña m« h×nh Riverstrahler) ®· ®−îc gãi
gän d−íi bÒ mÆt GIS ®Ó x©y dùng thµnh mét c«ng cô phÇn mÒm râ rµng thÓ hiÖn tÝnh tæng
qu¸t vµ tÝnh kh«ng gian cho phÐp x¸c ®Þnh chÊt l−îng n−íc. ¸p dông ®Çu tiªn cña m« h×nh
nµy ®èi víi s«ng Hång ®· ®−îc m« t¶ vµ ®¸nh gi¸ víi bé d÷ liÖu ®ßi hái chÊt l−îng n−íc cÇn
®−îc quan tr¾c hµng th¸ng trong suèt hai n¨m liªn tôc 2003-2004 t¹i h¹ l−u cña çc tiÓu l−u
vùc vµ trªn trôc chÝnh cña s«ng Hång.
Cuèi cïng, m« h×nh ®−îc sö dông ®Ó khai th¸c çc ¶nh h−ëng cña con ng−êi trong t−¬ng lai vÒ
çc mÆt qu¶n lý thñy v¨n, sö dông ®Êt vµ çc ho¹t ®éng n«ng nghiÖp, t¨ng d©n sè vµ çc chÝnh
s¸ch xö lý n−íc th¶i trong l−u vùc s«ng Hång ®Õn chÊt l−îng n−íc vµ çc ho¹t ®éng sinh th¸i
cña hÖ thèng s«ng Hång.
Tõ khãa: s«ng nhiÖt ®íi, s«ng Hång, ViÖt Nam, m« h×nh Riverstrahler/Seneque, c©n b»ng
dinh d−ìng, chu tr×nh nit¬, phètpho, silic, chÊt r¾n l¬ löng.
viii
Biogeochemical functioning of the Red River (North Vietnam): Budgets and Modelling
Main contents
Introduction
CHAPTER 1: Site description and major issues
1.1 Geographical presentation of the Red River basin
1.2 Geomorphology
1.3 Climate and hydrological regime
1.4 Hydrology
1.5 Social-economical context in the Red River basin and impacts
1.6 References
CHAPTER 2: General approach and methodology
2.1 Modelling the quality of the Red River hydrographic network
2.2 Experimental works
2.3 Nutrient budgets
2.4 References
CHAPTER 3: Hydrological regime and suspended load: observation and modelling
3.1 Introduction
3.2 General characteristics of the Red River basin
3.3 Hydrological regime of the Red River and its tributaries
3.4 Suspended solids loading of the Red River and its tributaries
3.5 Future scenarios of suspended solids loading
3.6 Conclusions
3.7 References
CHAPTER 4: Water quality
4.1 Discharge variations
4.1 Physical-chemical variables
4.3 General pattern of nutrients
4.4 Organic matter
4.5 Conclusions: water quality in the Red river
4.6 References
CHAPTER 5: Nutrient budgets (N, P)
5.1 Introduction
5.2 Description of the Red River Basin
5.3 The budget of the soil system
5.4 Domestic and industrial N, P loadings
5.5 The budget of the hydrographical network
5.6 Discussions
5.7 References
CHAPTER 6: Modelling the nutrient transfers in the river system
6.1 Introduction
6.2 The Riverstrahler model
6.3 Geomorphology
6.4 Hydrology
6.5 Role of reservoirs
6.6 Land use and non-point sources of nutrients
6.7 Wastewater point sources
6.8 Validation
6.9 References
CHAPTER 7: Exploring future trends of nutrient transfers
7.1 Impacts of new reservoirs constructed in the Red River basin
7.2 Fast increasing population and impact on water quality
7.3 Agricultural evolution and its impact on water quality
7.4 Prospective simulation at the 50 years horizon
7.5 References
General conclusions and perspectives
Contents
Annex
1
9
9
11
13
16
20
25
29
30
42
48
50
57
58
59
66
74
82
82
83
89
89
90
94
102
105
109
115
116
117
120
130
133
136
141
149
149
150
151
153
155
156
158
160
169
173
173
176
180
181
184
185
189
193
ix
x
Introduction
Introduction
Together with the Mekong, the Red River is the one of two largest rivers in Vietnam
(Figure 1). Both play an important role in the economic, cultural and political life of
Vietnamese people.
BangGiang-Kycung
Red-ThaiBinh
Ma-Chu
Ca
Huong (parfum)
ThuBon
Sesan
Sre pok
Ba
DongNai
Mekong
200km
Figure 1: Main river basins in Vietnam.
1
Introduction
The Red River brings many advantages with its abundant water resources. In general, the
water sources of the Red River in Vietnam are not only significantly used for irrigation but
also for domestic demand in country-village (Nguyen Ngoc Sinh et al., 1995). The river water
is also utilized for industries in the provinces of its upstream basin, Viet tri and Thai Nguyen
being typical examples of industrial zones. In addition, the water of the Red River is largely
exploited for power generation, since about 8.58 109 KW.h are provided each year by two
dams (the Hoa Binh and the Thac Ba dams) located in the Red River system (Trinh Quang
Hoa, 1998). Furthermore, the extensive network of wide and deep waterways in the Red River
basin represents an interesting potential for providing efficient means of transport, particularly
of heavy bulk cargo. Numerous inter-linked rivers, estuaries and coastal waters in the Red
River basin can be viewed as an excellent scope for the development of inland water-borne
transport facilities (Nguyen Ngoc Sinh et al., 1995). Whereas the benefits of the Red River are
clearly identified in Vietnam, its role in China has not been clearly recognised, perhaps less
important because of its morpho-geography unfavourable to human activities (94% of hills
and mountains in Yunnan province (Chinadata, 1998)).
The Red River has been strongly influenced by human activities in Vietnam. The
environmental pollution has regularly increased in the Red River basin, especially in its delta.
In the upstream of the Red River basin, deforestation (clear cutting or other harvesting
techniques) and land use changes are considered to cause a variety of environmental impacts
such as increased flooding and dramatically increased soil erosion from denuded watershed
exposed to high intensity tropical rainfall (Vo Tri Chung, 1998). In its downstream sector, the
high intensive farming areas attached to the use of nitrogen and phosphorus fertilizers, the
increase of population, the economic industrial development and urbanization as well as the
increased transportation network have strongly affected the water quality of the Red River
system and also influenced the coastal zone ecosystem (Ministry of Science and Technology MOSTE-: MOSTE 1998; MOSTE, 1999; Ministry of Environment and Natural Resources MONRE-, 2003).
The main objective of this Ph-D thesis, realized in a cooperative research program, was to
develop a comprehensive understanding of the linkage between land use and human activities
in the watershed in order to quantify the water quality and the transfer of nutrients (N, P, Si)
in the Red River drainage network (Vietnam and China). The mathematic model that has been
utilized for the Red River to establish this linkage is the RIVERSTRAHLER model. This has
been firstly developed for the Seine River (Billen et al., 1994; Garnier et al., 1995; Billen and
Garnier, 1999; Garnier et al., 1999), and then for several large European rivers (the Danube:
Garnier et al., 2002; the Mosel: Garnier et al., 1999; the Scheldt: Billen et al., 2005; the Rhine
2
Introduction
and the Loire: Garnier et al., 1997) to address the questions of organic pollution and oxygen
balance, nutrient contamination and related eutrophication, transfer and retention in the whole
basin. Moreover, this model would allow establishing the diagnostic of nutrient balance
(N:P:Si ratios), a key for controlling the eutrophication problem not only in the drainage
network but also at the coastal zone (Billen et al., 1985; Billen et al., 1997; Garnier and
Billen, 2002; Cugier et al., 2005). On the point of view of basic research, such an ecological
model has been applied to a sub-tropical river system for the first time, an approach devoted
to enlarge our knowledge on the ecological functioning of river ecosystem. Regarding the
management aspects, this study is also expected to serve as a guide for planning
environmental decisions at both regional and local scales. We implemented the
RIVERSTRAHLER model for the recent period of 8 years (from 1997 to 2004).
This work was undertaken in the framework of the ESPOIR on WATER project
aiming at identifying the water quality controls and at developing new processes for water
treatment. This three-year project (2001-2004) was supported by the activities of scientific
cooperation between different Vietnamese laboratories of the VAST (the Vietnamese
Academy of Sciences and Technology) and the French laboratories of CNRS (The French
National Centre for Scientific Research). Although this programme focused on the study on
water pollution and water treatment of urban rivers surrounding Hanoi, i.e. the Nhue-Tolich
river system located in the Red River delta, a special interest was given to the upstream
drainage network of the Red River, the Nhue river being one of diverted branched of the Red
River, upstream Hanoï (Figure 2). The Nhue receives directly the Tolich River draining Hanoï
(about 3.5 million inhabitants) therefore it is seriously polluted by the domestic and industrial
wastewater. It is important to note that Hanoi is equipped neither for domestic wastewater
collection and treatments nor for treatment systems of industrial wastewater; consequently the
Tolich River is extremely polluted and this pollution strongly impacts on water quality of the
Nhue River. Beside the Hanoï domestic and industrial pollution, the Nhue is also affected by
agricultural (irrigation in rice field and vegetation culture) and aquacultural (fish culture)
activities. The Nhue-Tolich hydrosystem is typically representative of the anthropogenic
rivers in the Red River Delta. As the Nhue River is supplied by the major branch of the Red
River through the Lien Mac dam, immediately upstream of Hanoi city (Figure 1), it was not
out of the scope of the programme to obtain a general knowledge of the quality of the Red
River, which constitutes the upstream limit condition of the Nhue River. A better regulation
of the inputs of water from the Red River to the Nhue River is indeed one of the possible
measures that can be proposed to improve the water quality of the Nhue River. Thus, although
the present study does not focuses on the small polluted urban rivers of the delta, a dialogue
3
Introduction
will now be possible between the model we developed for the Red River, and the one
developed in parallel in the framework of the ESPOIR programme on the special case of the
Nhue (Trinh Anh Duc, 2003).
Lo
.
R
Th
ao
R
.
Luc Nam R.
Da R
.
Red R.
Duong R.
Son Tay
Hanoï
ay
D
Tolich
R.
R.
ue
Nh
Hoa Binh
Haiphong
.
R
oî
B
.
R
Tra Ly
Ba
o
C
Day
R.
h
in
N
Tonkin Bay
La
t
Figure 2: Schematic representation of the Red River and its connections to the Nhue-Tolich
system.
This Ph-D thesis contains 7 chapters, several of them under the form of scientific papers
already published or submitted.
Chapter 1 is devoted to a general presentation of the Red River and its watershed, oriented
towards the construction of the model, the data required for the modelling approach being
physical constraints such as the geomorphology, geology and lithology and also of hydrometeorological nature, i.e. temperature, rainfall and hydrology.
Chapter 2 presents the general approaches and methodologies appropriate for the study of a
large regional system like the Red River basin. The general principles of the Riverstrahler
model, which has structured the whole study, are presented first. The experimental work that
was necessary to document the model regarding point and diffuse sources, as well as to
validate the modelling results, is then presented in this chapter. Indeed, whereas we have been
able to gather the data presented in chapter 1 from literature, or internet websites, water
4
Introduction
quality data in the Red River basin are scarce. Sampling campaigns were therefore realized
biological and chemical analyses were performed in the Vietnamese laboratory INPC
(VAST), after several trainings and inter-comparison have been organised with the French
Sisyphe laboratory (UMR 7619, CNRS and University Paris VI). The sampling strategies and
the methods used for these campaigns are described in this chapter. Lastly, the principles of
regional nutrient budget calculations, which offer a useful way of summarizing the overall
biogeochemical functioning of a regional system as well as of testing the coherency of the
data collected, are presented in this chapter.
Chapter 3 focuses on the modelling of the hydrology of the hydrographical network and on
the transport of suspended solid in the Red River basin. Daily meteorological and discharge
data have been analysed for a period of 8 years (1997-2004) with the simplified hydrological
model used as a part of the RIVERSTRAHLER MODEL. In addition, we have analysed the
behaviour of suspended solids in the drainage network in the context of the recent and future
large dam constructions. This chapter constitutes a scientific paper submitted in the Journal of
Hydrology.
The results of water quality observation in the rivers of the Red River drainage are reported in
Chapter 4. This chapter mentions the experimental results obtained in both INPC and
Sisyphe laboratories on water quality at the outlet of the three main sub-basins and in the
main branch of the Red River system in the period from 2002 to 2004. A comparison is made
with the data obtained in parallel on the much more polluted Nhue and ToLich rivers.
The establishment of nutrient budgets in the 4 sub-basins of the Red River is reported in
Chapter 5. In this part, nutrient budgets have been calculated using many statistical sources
within the Red River basin, and our own measurements in the hydrographic network. For the
first time, nutrient budgets were established for the agricultural soils using an agronomical
point of view and nutrient transfers calculated in the drainage network. This work is the
material of a paper published in the Journal of Global Biogeochemical Cycles.
The modelling of nutrient transport in the rivers of the Red River system is reported in
Chapter 6. This chapter describes how the Riverstrahler model takes into account the various
constraints to the drainage network functioning, and how the corresponding information has
been gathered for the special case of the Red River watershed. The results of the application
of the Seneque/Riverstrahler software to the Red River system are presented to validate the
model and illustrate its capabilities. This part will be submitted as a paper to the Journal of
Biogeochemistry.
5
Introduction
Lastly, in Chapter 7, we discuss the scenarios aiming to explore future conditions that could
be found in the Red River basin taking into account socio-economical trends observed and
new plans, for a rehabilitation of impacted systems of the urbanised areas, such as the delta,
but also to avoid ecosystem damage in zones of still good ecological status. A main objective
is to demonstrate that the tools implemented during this Ph-D thesis can be utilisable in
Vietnam to test scenarios for management purposes of human impacts in the watershed.
Explorations by the model such as rapid increase in population, reservoir construction in the
upstream basin of the Da and the Lo Rivers, are all subjects that are discussed in this chapter.
This part is also intended to form the basic material of a paper to be submitted to a scientific
journal.
The Conclusions stress the usefulness of our modelling approach as a framework to gather
pertinent information on a regional territory and to test the coherency of the data available at
this regional scale. We will defend the view that this approach, tested here on the Red River
system, can be extended for improving our knowledge on other poorly documented river
systems of the world.
References
Billen G., Somville M., DeBecker E. and Servais P., 1985. A nitrogen budget of the Scheldt
hydrographic basin. Neth J. Sea Res., 19: 223-230.
Billen G., Garnier J. and Hanset P., 1994. Modelling phytoplankton development in whole
drainage networks: The RIVERSTRAHLER model applied to the Seine river system.
Hydrobiologia, 289: 119-137.
Billen G. and Garnier J., 1997. The Phison River plume: coastal eutrophication in response to
change in land use and water management in the watershed, Aquat. Microb Ecol., 13:
3-17.
Billen G. and Garnier J., 1999. Nitrogen transfer through the Seine drainage network: a
budget based on the application of the RIVERSTRAHLER Model. Hydrobiologia,
410: 139-150.
Billen G., Garnier J. and Rousseau V., 2005. Nutrient fluxes and water quality in the drainage
network of the Scheldt basin over the last 50 years. Hydrobiologia (in press).
Chinadata 1998. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, Vol. 2000 China
Statistical Publishing House, (Basic Information of Yunnan, China).
(http://chinadatacenter.org)
Cugier Ph., Billen G., Guillaud J.F., Garnier J. and Ménesguen A., 2005. Modelling the
eutrophication of the Seine Bight (France) under historical, present and future riverine
nutrient loading. J. Hydrol. 304: 381-396
6
Introduction
Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and chlorophyecae
in the drainage network of the River Seine: Observations and modelling. Limnology.
and Oceanography, 40: 750-765.
Garnier J., Billen G. and Hannon E., 1997. Biogeochemical Nutrient Cycling in Large River
Systems (Binoculars). Final Technical Report. EC Environment Programme (ref
PL932037). 33 pp + Annexes.
Garnier J., Leporcq B., Sanchez N. and Philippon X., 1999. Biogeochemical budgets in three
large reservoirs of the Seine basin (Marne, Seine and Aube reservoirs).
Biogeochemistry, 47: 119-146.
Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget and related
ecological processes in the river Mosel: the Riverstrahler approach. Man and Rivers
System. J. G. J. M. M. Hydrobiologia. Netherland, 1999 Kluwer Academic Publishers.
410: 151-166.
Garnier J., Billen G., Hannon E., Fonbonne S., Videnina Y. and Soulie M., 2002. Modeling
transfer and retention of nutrients in the drainage network of the Danube River.
Estuarine, Coastal and Shelf Science, 54: 285-308.
MONRE, 2003. Report on water environment monitor in Vietnam in 2003. In “Studies on
Vietnam environmental statement in 2003”. Vietnam Ministry of Environment and
Natural Resources, 150pp., Hanoi.
MOSTE, 1998. Documentation on the Red River Delta (1997-1998), Ministry of Science,
Technology and Environment of Vietnam, Scientific and Technical Publisher, 214pp.,
Hanoi.
MOSTE, 1999. Environmental statement in Vietnam in the years 1990s. Ministry of Science,
Technology and Environment of Vietnam, Scientific and Technical Publisher, 219pp.,
Hanoi
Nguyen Ngoc Sinh, Hua Chien Thang, Nguyen Chu Hoi, Nguyen Van Tien, Lang Van Ken,
Pham Van Ninh and Nguyen Vu Trong., 1995. Case study report on Red River Delta
in Vietnam - Project on integrated management and conservation of near shore coastal
and marine areas in East Asia region (EAS-35) United Nations Environment program.
Regional coordinating for the East Seas (ESA/RCU), report, 78pp., U.N. Environ.
Programme, Nairobi.
Trinh Anh Duc, 2003. Etude de la qualité des eaux d’un hydrosystème fluvial urbain autour
de Hanoi (Vietnam); suivi expérimental et modélisation. Thèse de doctorat d'Etat de
l'Université Joseph Fourrier, Grenoble 1, France and Vietnam Academy of Science and
Technology (VAST). 265 p.
Trinh Quang Hoa, 1998. Water balance for purpose of socio-economic development in the
Red River delta. Proceedings of International Conference on Economic development
and environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar.
Vo Tri Chung, 1998. Forests on the Red River basin, Vietnam. Proceedings of International
Conference of Economic development and environmental protection in the Yuan-Red
River watershed, Hanoi 4th-5th March.
7
Introduction
Articles in press, submitted or to be submitted in the framework of this PhD thesis:
Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S., Fézard, C. and Chau, Van Minh
(2005). Nutrient (N, P) budgets for the Red River basin (Vietnam and China). Journal of
Global Biogeochemical cycles. Vol 19, GB2022, doi 10.1029/2004GB002405.
Le Thi Phuong Quynh, , Garnier J., Billen G., Thery S. and Chau V. M., 2005. Hydrological
regime and suspended matter flux of the Red River system (Vietnam): Observations
and modelling. Journal of Hydrology (submitted).
Le Thi Phuong Quynh, Billen G., Garnier J., Thery S., Ruelland D. and Chau V. M., 2005.
Nutrient transfers through the Red River basin (Vietnam): Observations and
modelling. Biogeochemistry (in prep.).
8
Site description and major issues
CHAPITRE 1
Site Description and Major Issues
The Red River basin (Figure 1.1) is located in the South-East Asia, from the latitude 20°00 to
25°30 North and from the longitude 100°00 to 107°10 East. The Red River is bordered by the
Truong Giang and the Chau Giang River basins (in China) in the North, by the Langcang
River (Mekong) basin in the West, by the Ma River basin (in Vietnam) in the South, and by
the Thai Binh River and the Tonkin Bay in the East (Nguyen Ngoc Sinh et al., 1995). In this
sub-tropical region, where chemical and mechanical erosion are among the highest of the
world (500 mm/1000 years), large rivers transport considerable amount of suspended solids
(Meybeck et al., 1989; Dupré et al., 2002). The climate is of monsoon type, with summer
dramatic inundations. The biggest floods in the Red River delta occurred in 1913, 1915, 1945
and 1971 when the serious dyke breakage happened in many places. The floods in 1971
submerged 250140 ha and affected about 2.71 million people, damaged 7 millions tons of
paddy (To Trung Nghia, 2000). In the Mekong delta, dramatic floods occurred in 2000 and
2001, affecting about 900 Vietnamese people. In the Red River delta, dikes dating back to the
early 1800s are maintained to protect the population in the delta area (To Trung Nghia, 2000).
Thac Ba reservoir
Hoa Binh reservoir
Figure 1.1: The Red River and its watershed
9
The Red River (or Thao River) originates in the mountainous region of South China
(100° 00’20’’ longitude, 25°30’10’ latitude), at the foot of the Himalaya mountains (Nguyen
Huu Khai and Nguyen Van Tuan 2001) in Dali city, in the Yunnan province, between the
Langcang and Jinsha river watersheds (Figure 1.2). The altitude of the source is about 3000
m. In the Chinese part, the Red River is named the Yuan River (the Yuanjiang or YuanjiangHong), located beside some other important rivers in Southeast Asia such as NujiangSalween, Nanpan, Jinsha, Lancang - Mekong, Dulong - Irrawaddy rivers. All of them are
originated from (eg. Yuanjiang and Nanpan) or go through the Yunnan province, and are
important pathways between China and Southeast and South Asia (Chinadata, 1998). In
average, the Red River has smaller discharge than other biggest rivers in South Asia (table
1.1).
Table 1.1: Characteristics of some largest rivers in South and Southeast Asia
River
Drainage area
km2
Water discharge
m3.s -1
Pearl (Zhujiang)
442585
10033
Zhang J., 1996
Yangtze (Changjiang)
1808500
24443
Zhang J., 1996
Mekong (Langcang)
803000
11000
Meybeck, 1989
Irrawaddy (Dulong)
430000
13600
Meybeck, 1989
Red River (1997-2004)
151448
3577
This study
References
The Red River is known as the “six-head
river” that enters into Vietnam at Lao Cai
province with its name of Thao (or Cai, or
Hong) River (Dang Anh Tuan, 2000). The
CHINA
name of the Red or Hong River originates
from its reddish-brown colour water, due to
the
transport
of
large
quantities
Mekong River
of
sediments, rich in iron dioxide. It runs
Dali
directly through Yunnan, Lao Cai, Yen Bai,
Phu Tho, Hanoi, Hung Yen and Thai Binh
provinces forming the Red River delta
before flowing into the China Sea (Gulf of
Yuan River
Yuanjiang
River
Red River
Tonkin) through four distributaries called,
Ba
Lat
(106° 32’10’’
longitude
and
20°20’00’ latitude), Lach Gia, Tra Ly, and
Day (Dang Anh Tuan, 2000).
10
Figure 1.2: The source of Red River in China
The Thao River receives two major tributaries: the Da (or the Black) River on the right bank
and the Lo (the Clear) River on the left bank.
The source of the Da is also located in the Yunnan province. It flows directly through
Yunnan, Lai Chau, Son La, Hoa Binh and Ha Tay provinces before reaching to the Thao
River at Ha Nong district, in Viet tri city (Figure 1.1). The Da River originates from a region
with a mean elevation of 2000m (Nguyen Huu Khai and Nguyen Van Tuan, 2001).
The Lo River also originates from in China and joins with the main branch at Viet Tri city.
The elevation of the source of the Lo River is 1100m (Nguyen Huu Khai and Nguyen Van
Tuan, 2001).
From the Viet tri confluence point to the estuary, the Thao River is named the Red (or Hong)
River.
1.2 Geomorphology
The area of the whole Red River basin takes different values depending on the authors,
because of the different ways of estimating, within the delta, the complex hydrographic
network of the Red-ThaiBinh River system, i.e., the Red River delta from the ThaiBinh river
network. In this study, the total area of the Red River catchment was first estimated to 156
451 km2. A subsequent analysis based on the treatment of the digital elevation model of the
NASA (global SRTM 3” resolution) lead to a slightly different watershed area of 142 950
km². Within the Red River watershed area, 47.9% is in Chinese (Chinadata, 1998), 51.2% in
Vietnamese (MOSTE, 1997) and 0.9% is Laotian territories.
In the Yunnan province (394000 km2, 4.1% of China), the Red River watershed occupies
about 20 % of the area of the province. It is important to note this proportion that will be used
below, to calculate figures related to the Red River basin, when we only obtained information
for the whole Yunnan.
The relief of the Red River basin that much varies from headwaters to the downstream areas
can be divided into three sections (figure 1.3).
i) In the Chinese part, mountainous landscapes dominate. Mean elevation of the
Yunnan province is at about 2000 m, but maximal elevation reaches 6740 m and the minimal
one is of 76.4 m (Chinadata, 2000). Within the total Yunnan province area, about 84% are
rugged mountains; 10% are highlands and hills; and only 6% are lowland and valleys
(Chinadata, 2000). Mountain areas are tectonically active and unstable, and this, combined
with intense rainfall, causes high erosion (Fullen et al., 1998). In Eastern Yunnan, the Red
11
River valley is surrounded by the Karst Plateau, composed of red stratum, called the Central
Yunnan Red Soil Plateau. Sandstones or mudstones of mixed colors including red, purple,
bluish gray, yellow and gray-white are widely exposed to erosion giving the red color water
of this river (Chinadata, 1998).
ii) In the Vietnamese part, about more than half of the Red River basin lies in the
mountainous region. The East-North Vietnam area is dominated by the Hoang Lien Son
Mountain with the highest pick as Phanxipan (3143m) in Sapa town, in LaoCai province.
Some other high mountains also locate in this area. In the North Vietnam, soils are mostly
(70%) grey and alluvial soils (MOSTE, 1997). Red soil occupies only 7% and rugged
mountains about 10%.
iii) The delta, the third section of the Red River basin, covers a very flat and low land,
elevation ranging from 0.4 to 12 m above sea level, with 36% lying below 2m (Dang Quang
Tinh, 2001). There are however higher areas in the delta which take the form of steep
limestone karsts, type formations which occur as isolated hills in Ninh Binh, Nam Ha, Ha
Tay, Ha Bac, Quang Ninh provinces including the famous Ha Long Bay (Nguyen Ngoc Sinh
et al., 1995).
Figure 1.3: False perspective view of the relief of the Red River basin (viewed from the
delta mouth), generated by treatment of a digital elevation model (global SRTM 3”
resolution, NASA, www:\\NASA.org)
12
Considering the 3 main watersheds of the whole Red River catchment, the mean elevations
are rather similar for the Da river basin (965 m), the Lo River region (884m) and the Thao
river watershed (647m) (Nguyen Viet Pho, 1984).
The total length of the Red River course is about of 1126 km from the source to the mouth, of
which 556 km is in the Vietnamese territory (To Trung Nghia, 2000). The mean slope of the
whole Red River basin is of 29.9% (Nguyen Huu Khai and Nguyen Van Tuan, 2001).
The Da and Lo rivers respectively have its length of 1010km (560 km in Vietnam) and of 470
km (275 km in Vietnam). Note that the Red River course can be split into the Thao (about 910
km) and the Hong River (delta, about 216 km, (Nguyen Viet Pho, 1984)).
The climate in the Red River basin, of sub-tropical East Asia monsoon type, is controlled by
the North East monsoon in winter and South West monsoon in summer. The climate is
characterized by two distinct seasons. The rainy season lasts from May to October and the dry
season covers the period from November to the next April.
During the study, we have gathered the meteorological data during the period from 1997 to
2004: daily rainfall, monthly temperature, monthly humidity, and monthly solar radiation,
obtained from 13 meteorological stations in the Red River basin (see Figure 1.4). The
evapotranspiration (ETP) data have been calculated by using Turc’s formula (Turc, 1961),
based on monthly temperature and sunshine duration data obtained from the respective
meteorological stations (see chapter 3).
The climate of the Red River basin is well described in the chapter 3. In the period from 1997
and 2004, the annual mean temperature, humidity, annual rainfall and ETP data in the
Vietnamese part are higher than values obtained in the China part.
The annual mean temperature varied from 14 to 27 °C in the whole Red River basin. The
monthly temperature varied from 14 to 25 °C in the upstream sub-basins and is higher in the
delta region (16 to 28°C) (IMH, 1997-2004).
As other tropical river basins, the humidity always remains in high level. In the whole Red
River basin, humidity averaged from 82 to 84% all over the year in the Vietnamese part of the
basin (IMH 1997-2004), while it was lower, about of 67÷70 %, in the Chinese part
(Chinadata, 1998; Chinadata 2000).
The rainy season cumulates 85 – 90% of the total annual rainfall in the Red River catchment.
It is also interesting to note that July and August are two months with the highest incidence of
13
typhoons in the Red River. The mean annual rainfall is 1587 mm in the whole Red River
basin.
Kunming
Ha Giang
Lao Cai
Yen Bai
Sa Pa
Tuyen Quang
Phu Tho
Lai Chau
Son Tay
Son La
Ha Noi
meteorological station
Thai Binh
Hoa Binh
hydrological station
0
20 50 70 100km
Nam Dinh
N
Figure 1.4: Meteorological and hydrological stations in the Red River basin
The climate of the Red River, characterized by a monsoon sub-tropical regime, confers the
typical hydrologic regime characterized by large runoff during summer and low runoff during
winter. Figure 1.5, constructed with data borrowed from Guilcher (1965) and other sources,
compares the climatic and hydrologic behaviour of the Red River with that of Arctic,
Mediterranean and Temperate Oceanic regions of the world. Both Mediterranean and
Temperate oceanic types of rivers have their maximum discharge during winter, because
evapotranspiration is the lowest in this season. Except for arctic rivers, which are
characterized by large discharge in spring due to snow melt at that time of the year (figure
1.5), the sub-tropical rivers are the only ones characterized by highest specific discharge
during the period of occurrence of highest radiative energy and temperature.
14
0
20
150
15
100
10
50
5
0
0
20
10
0
15
100
10
50
5
0
0
Seine R.
-
10
5
10
5
0
0
J F MAMJ J A S ON D
spec. disch., l.s .km ²
30
-1
spec. disch.
40
150
J F MAM J J AS ON D
25
200
20
150
15
100
10
50
5
0
J F MAM J J AS OND
0
J F MAMJ J A SOND
15
Ardèche R.
-
Kalix R.
20
30
250
J F MAM J J AS ON D
15
-1
-
50
60
-1
70
25
200
J F MAM J J AS ON D
J F MAM J J AS ON D
Oceanic Temperate
Monsoon tropical
300
temp., °C
200
250
temp., °C
rain or etr., mm/month
25
30
-²
50
250
300
-1
100
30
spec. disch., l.s .km
150
Mediterranean
rain. or etr., mm/month
200
300
temp., °C
20
15
10
5
0
-5
-10
-15
-20
Arctic
rain. or etr., mm/month
250
rain
etr
temp
temp., °C
rain or etr., mm/month
300
80
70
60
50
40
30
20
10
0
Red R., Vietnam
J F MAM J J A SON D
Figure 1.5: Climatic regime: rainfall (rain: mm/month); evapotranspiration (etr.: mm/month) and temperature (temp.: 0C) and specific discharge (spec.
disch.:L.s-1.km-2) of some rivers located in the different climatic regimes in the world. (Guilcher, 1965)
15
1.4. Hydrology
1.4.1 Hydrology in Vietnam
1.4.1.1 Surface water in Vietnam
Vietnam has an abundant water resource with a dense river network, of which 2360 rivers
have a length of more than 10 km (Nguyen Viet Pho, 1984). Within these rivers, eight have
large basins with a catchments area of 10000 km2 or more (table 1.2). The drainage density
varies from 0.25 to 1.94 km.km-2. Along the Vietnamese coastline (3260 km), about 20 km
separate the various river mouths. With an annual rainfall average in Vietnam of 1957 mm
and an annual evaporation of 983 mm, the total runoff of Vietnam is about 880.109 m3.y-1
(SEAMCAP, 2001).
Table 1.2: Major rivers and their watersheds in Vietnam (SEAMCAP, 2001)
River
Watershed area,
km2
Mean annual
discharge
Population in Vietnam
(in 1995)
Pop. Dens*,
inhab.km-2
total area
area in
Vietnam
total,
109m3
% of the total
Vietnam river
discharge
Inhabitants
(106)
Mekong
795000
72000
520.6
59.2
16.8
233
Red-ThaiBinh
169000
86660
137.0
15.6
24.2
279
DongNai
42655
36261
30.6
3.5
10.2
282
Ma
28490
17810
20.1
2.3
2.9
163
Ca
27200
17730
24.2
2.7
3.1
175
Ba
13900
13900
10.4
1.2
0.9
61
Bang Giang-KyCung
12880
11220
8.9
1.0
1.0
91
ThuBon
10496
10496
19.3
2.2
0.9
82
*:
population density (Pop. Dens*) in inhabitants.km-2
Note that the Red-Thai Binh and Mekong rivers carry 74.8 % of the total surface water
resource in Vietnam, while each of the other basins represents only 1÷3 % (table 1.2).
About two thirds of the water resources originate from catchment in neighbour countries.
Vietnam is the lower country for both the Mekong and the Red Rivers, and depends on the
water resource management and decisions taken in the upstream countries. This might
amplify the highly variable seasonal and geographical distribution of water (droughts in the
dry season and flood during in the monsoons) (MONRE, 2003).
16
Most dams and reservoirs in Vietnam have been constructed for multipurpose, including flood
control, irrigation, hydropower, water supply and other flow management. There are about
3600 reservoirs of various size of which less than 15% have a capacity above 1 million m3 or
a depth higher than 10 m). Some biggest reservoirs in Vietnam are presented in table 1.3.
Sedimentation from erosion within the watersheds leads to a decline in the reservoir capacity:
most reservoirs and dams were constructed since 20 - 30 years and some of them have lost up
to 70-30 % of their original capacities (MONRE, 2003).
Surface water is utilized for agricultural irrigation, aquaculture, domestic supply, livestock,
industry and service. In Vietnam, agriculture remains the largest consumer of water (about
82% of the total demand). Industry (6.5% of the total demand) and domestic use (about 2.5%
of the total demand) are however rising with population growth and economic development
(MONRE, 2003).
Table 1.3: Major reservoirs in Vietnam (MONRE, 2003).
Reservoir
Catchment
km2
Volume
km3
Hydropower
MW
*Hoa Binh
51700
9450
1920
* Thac Ba
6100
2940
108
Tri An
14600
2760
420
Dau Tieng
2700
1580
-
Thac Mo
2200
1370
150
Yaly
7455
1037
720
Phu Ninh
235
414
-
Song Hinh
772
357
66
Ke Go
223
345
-
* The reservoirs within the Red river basin
1.4.1.2 Groundwater in Vietnam
The groundwater resource in Vietnam is abundant, with a total potential exploitable reserve of
the aquifer with the whole country estimated at nearly 60 km3.y-1 (MONRE, 2003). Over
50 % of these resources are in the central part, about 40 % in the north and 10 % in the south
of Vietnam. A large amount of water is stored in unconsolidated alluvial sand and gravel
geological formations found in plains and valleys. A substantial part of these resources
(estimated at 35 km3.y-1) returns to the rivers as base flow, underground water being an
important river flow component in the dry season (MONRE, 2003). Groundwater is exploited
17
for irrigation of crash crops or for drinking water but less than 5% of the total underground
reserves is exploited for the whole country (MONRE, 2003).
1.4.2. Hydrology of the Red River
1.4.2.1 Drainage density
Within the Red River basin, the drainage density is quite high, in the range of 0.5 to 1.5
km.km-2 with about 500 streams and rivers (Le Bac Huynh, 1997).
In the upstream basin of the Red River, the Yunnan province territory is a vast land with
plentiful rivers: over 600 rivers and lakes (Chinadata, 1998).
The drainage density is much more complex in the delta areas, ranges from 0.7 to 1 km.km-2.
A dense system of irrigation channels for agricultural activities adds to the natural complexity
of the system. Trinh Quang Hoa (1998) reports that 30 main irrigation channels have been
constructed in the Red-Thai Binh river delta providing water for 735370 ha. Tran Duc Thanh
et al. (2004) mentions that the demand for irrigation water in dry season ranges from 25 to
50% of the river discharge in the Red River delta.
For this work, the hydrographic network of the Red River and its elementary watersheds,
constitute the first and basic layer of the GIS database. The details for the construction of the
hydrographic network representation are described in chapter 6. An important work has been
realized to geo-reference all the Vietnamese streams of the drainage network and to connect
them towards the direction of water flux. This network was then simplified, in order to adjust
the resolution to the one available for the Chinese part of the basin, finally producing the
simplified map of figure 1.6.
Figure 1.6: Drainage
network and elementary
watersheds of the Red
River basin, obtained by
treatment of the digital
elevation model of the
NASA (see chapter 6).
18
100 km
1.4.2.2. Water flows
The daily discharge data at the outlet of the 3 main branches and in the delta of the Red River
in the period from 1997 to 2004 were obtained at 6 hydrological stations from the Vietnamese
Ministry of Environment and Natural Resources (MONRE): the Hoa Binh station (in Hoa
Binh province) for the Da outlet; the Vu Quang station (in Phu Tho province) for the Lo
outlet; the Yen Bai station (in Yen Bai city) for the Thao outlet, and two stations along the
downstream course of the Hong river: Son Tay (in Ha Tay province) and Hanoi (in Hanoi
city) (Figure 1.4). In the period 1997-2003, the mean annual discharge of the main branch at
Son Tay station was of 3577 m3.s-1 (MONRE 1997-2004).
Whereas the flow of the Red river basin including the three main branches does not vary
greatly from year to year (within the period from 1997 to 2004), it largely varies seasonally.
The seasonal distribution of the water within the Red River basin depends on unevenly
distributed monsoon rainfalls. Such high variations combined with limited storage capacity
and insufficient flood control infrastructure result in devastating floods in the wet season and
damaging extreme low flows in the dry season.
According to long term hydrological data series, the annual discharge volume of the Red
River is around 130 109 m3 (a mean discharge of approximately 3600 m3.s-1 at Son Tay. This
accounts for about 15% of the total runoff for the whole Vietnam (Nguyen Ngoc Sinh et al.,
1995).
1.4.2.3. Reservoirs
The Hoa Binh and Thac Ba reservoirs are the two largest dam-reservoirs located in the Red
River basin (figure 1.1). Similarly to most reservoirs in Vietnam, they have been constructed
as multi-purpose reservoirs: for power generation, flood control, agricultural irrigation,
fishery and tourism. The Hoa Binh Reservoir, damming the Da River, is the largest reservoir
in North Vietnam (table 1.3). These two reservoirs on the Da and the Lo rivers represent a
storage capacity of nearly 7 km3, but only 6 percent of the mean annual flows of the Red
River (Vu Van Tuan, 2002). However, the influence of the Hoa Binh and Thac Ba reservoirs
on the flow and the suspended solid flux at Son Tay station (main branch of the Red River) is
not negligible. The detail about the hydrology and suspended solid transfers will be showed
below, in the chapter 3.
19
Table 1.4: Some major characteristics of the 3 main sub-basins (Da, Lo, Thao) of the Red
River system and its delta area.
Sub-basin
Da
Lo
Thao
Delta
51285
34559
61169
9435
1925
973
743
3290
(11100 ; 283
(8340 ; 165)
(6210 ; 146)
Reservoir, 109 m3
3.9-9.5
0.78-2.94
-
-
Population density
101
132
150
1173
Catchment area, km²
Average Discharge*,
(max ; min)
m3s-1
(20900 ; 555)
*average discharge for the period from the daily data from 1997 to 2004. Maximum and minimum values during
the same period between brackets.
Due to the high population density in the whole Red River (193 inhab.km-2) and mainly in the
delta, the impact by human activities is necessarily important. Contrarily to other densely
populated countries in Western Europe or North America, human influences on water quality
have not been well studied in South East Asia, including the Vietnam. In fact, until now the
major concerns to environmental problems are the damage caused by floods. In the Red River
delta, more attention has been paid for protecting population against flood during the rainy
season and water management to feed the population, than was devoted to water quality
issues.
1.5.1. General socio-economical context
Besides geomorphological and hydro-meteorological data which are major constraints to the
modelling approach, land use and fertilization, increasing population and domestic and
industrial pollution are also major constraints required to model water quality. Whereas these
constraints will be deeply analysed in chapter 6, general insights will be given here helping to
ask the appropriate questions.
1.5.1.1. Changes in land cover
Several changes in land cover of the Red River basin have been observed since the last 100
years.
Firstly, we have to mention about the deforestation and intensification of agriculture that have
largely occurred in both Vietnamese and Chinese parts.
20
In the Chinese part, the forest cover of Yunnan has declined from about 60% in the 1950s to
24.2% in 1990 (UNEP, 1990). About 10% of land in this province is categorized as severely
eroded in the 1980s. Only 7% of Yunnan land area is suitable for agricultural activities
(Fullen et al., 1998). Agriculture is restricted to a few of upland plains, open valley and
terraced hillsides. The main food crops such as maize, rice, wheat and potatoes and the main
cash crop such as tobacco, tea, sugarcane are grown in this area. The intensification of
agriculture has occurred thanks to deforestation, increasing cultivation of steep erodible
slopes, over cultivation and adoption of non-sustainable farming practices (Fullen et al.,
1998).
In Vietnam, land use and cover change is the most pervasive and immediately observable
component of the change. Deforestation, intensification of agriculture and urbanization
processes have occurred at variable and often rapid rates over the last couple of decades. It
was noted that warfare and deforestation associated with post-war development 1975 have left
the whole nation with only about 10% cover of closed tropical forests with less than 1% in
pristine state (Collins et al., 1995; Lebel, 1996).
In North Vietnam, deforestation processes was severe, especially in the northern mountains
and midlands. In this area, the forest which covered 95% in 1943 decreased to 17% in 1991
(World Bank, 1996, Nguyen Ngoc Sinh et al. 1995); a slight increased to 19% in the period
from 1995 to 1999 was observed due to the governmental policies of conservation and
development of cultivated forest (Pham Ngoc Dang et al., 2001).
Accounting for the forest area in the Red River basin in the Vietnamese territory, Vo Tri
Chung (1998) reported 3.6 million ha of forest, representing 31% in 1990 (58% for the barren
land). After carrying out the plan of 5 million ha of reforestation of which about 1.2 – 1.5
million ha should be given to the Red River basin, the forest area occupies about 45% of the
whole Red River watershed.
1.5.1.2. Increase of fertilizers utilisation
Fertiliser utilisation (as chemical fertilizer) has much increased in agricultural land in
Vietnam and in China for the recent 50 years. China is an agricultural country where
anthropogenic activity affects strongly surface and groundwater quality through chemical
fertilizer use (23.5 million tons in 1991) and irrigation. Weijin et al. (1999) mentioned that
China is the largest producer of nitrogen fertilizers and largest consumer of mineral fertilizers
in the world. In Vietnam, according to the FAO database (FAO 1990-1998), use of nitrogen
fertilizers has increased by 66 folds during a period from 1961 to 2000 (from 2.2 kgN.ha-1.y-1
in 1961 to 150 kgN.ha-1.y-1 in 2000). For phosphorus fertilizers, the amount used has been 5
21
folds multiplied during the same period. Application of chemical fertilizers may dramatically
increase nutrient concentrations in soils which may subsequently be removed by leaching and
transferred to the river water (figure 1.7). Further, serious erosion and soil loss in watersheds
accelerate the removal of nutrient elements.
Figure 1.7: pollution sources (non-point sources and point sources) in the Red River basin
1.5.1.3. Increase of the population and urbanisation
Increase in population and urbanization might also considerably impact the river system.
The total population of the Red River basin is estimated at 30 million inhabitants and is
growing at an annual rate of about 2.0 %. 65% of the Red River population is Vietnamese,
34% is Chinese and 1% is from Laos. Contrasted population density within the whole Red
River basin must be mentioned: averaging 195 inhabitants.km-2 for the whole basin;
101 inhabitants.km-2 are found in average in the northern mountainous region and
1174 inhabitants.km-2 in the Red River Delta region.
In the Chinese part, in the Yunnan province, where the inhabitants are living in 8 autonomous
prefectures and 11 cities (127 counties, towns), population of the Red River was estimated of
8.8 million inhabitants (about 20.9% of the total Yunnan population), (Chinadata, 1998), of
which 34% population belong to the ethnic minorities. In this area, the annual population
growth rate is of 1.29% (Chinadata, 1998).
In the Vietnamese part, the present annual population grow at a higher rate than in the
Chinese part (about 2.3%). The inhabitants are located in about 21 provinces and cities, and
comprise 18 different ethnic groups of minority people with typical traditions of culture
(MOSTE, 1997).
Parallel with the increasing population, the urbanisation in Vietnam has occurred at a high
rate in recent years. The number of agglomerations (city and town,) has increased from 500 in
of the early 1990’s to 623 in 2000. Whereas population living in agglomeration averaged to
19% in 1990 and increased to 23.5% in 1999, it should reach up to 30-33% in 2010 (Pham
Ngoc Dang et al., 2001). Such a continued rapid urban growth would be a big problem for the
22
future. However, Smith and Dixon (1997) reported that Vietnam has the lowest rate of urban
growth in the ASEAN group, except Singapore (the Association of Southeast Asian Nations
includes Vietnam, Laos, Cambodia, Thailand, Myanmar, Indonesia, Malaysia, Philippines,
Brunei, Singapore and East-Timor countries).
Domestic wastewater from cities and large agglomeration are mostly discharged directly into
rivers or lakes without treatment, leading to serious pollution of water environment in cities,
especially in Hanoi, Hai Phong, Viet tri (figure 1.8)…
Thai Nguyen
province
u
Ca
Vinh Phuc city
R.
Viet Tri city
Red R.
Son Tay
Hanoï
Quang Ninh
Duong R.
province
Ha Tay
province
Nhue R.
Thai Binh R.
Hai Duong
province
Haiphong city
y
Da
R.
Thai Binh
province
Tra Ly
Day
R.
Ba
La
t
n
Ni
h
Co
Figure 1.8: Cities and provinces with high population and industrial zones
1.5.1.4. Increase of industrial releases
Since the late 1980’s, Vietnam enters into the period of rapid economic growth that has been
closely associated with a re-engagement within the international and regional Pacific Asian
economies. The “Doi moi” programme introduced in 1986 has opened the economy to the
international monetary system and to a market economy, and reduces the central control
exercised by the State. Industrial activity in Vietnam has rapidly increased. According to the
(MOSTE, 2000), the number of industrial zones in Vietnam has increased from 16 in 1996 to
66 in 1999. In general, 90% of the industrial factories which were constructed before 1975
23
have no wastewater treatment systems. Since 1994, the factories have been located in these
new industrial zones where wastewater treatment systems have been constructed.
Industrial activities are mainly found in the delta of the Red River and Mekong River, and
have especially increased in some large cities. For example, industrial values of Hanoi
represent 8.2% of the whole Vietnamese industrial production in 1995 and increased to 9.4%
in 2002 (Le Qui An, 2003). With the rapid industrial increase and the absence of wastewater
purification infrastructures, the quality of surface and groundwater, at the local scale mainly,
but also at a larger scale, has strongly decreased.
In the Red River basin, there are several industrial zones which influence directly the water
quality of the Red River. In the middle of the basin, the Viet tri city is one of the most
important industrial zones in the North of Vietnam, where the food and drinks production,
paper, chemicals ... are concentrated. Almost all of the wastewaters related to these activities
are discharge directly into the Red River. Beside this zone, another industrial zone in the
mountainous region in the North Vietnam (Thai Nguyen, see figure 1.9) has less influence on
the Red River but strongly influences the Cau River (part of the Thai Binh river system).
Some other industrial zones such as Hai Duong, Hai Phong and Quang Ninh provinces in the
downstream delta of the Red River, and Hanoï, considered apart, have also their impacts on
aqua-ecological processes (see figure 1.8).
1.5.2. Impacts on water quality
1.5.2.1. Decline of surface water quality
In Vietnam, data on surface water quality is poor, and hardly exist in the upstream basin of
the Red River. However, the few existing researches have revealed that water quality of rivers
remains good in upstream rivers while downstream, domestic and industrial water releases
strongly pollute the river water especially in major cities. Urban rivers such as the To Lich,
Lu, Set, Kim Nguu Rivers in Hanoi are typical examples of open wastewater collectors, the
water quality of which being disastrous, especially in dry season. Suspended solid (SS) ranges
from 60 to 300 mg.L-1; dissolved oxygen (DO) within a range of 0.2 to 3 mgO2.L-1, biological
oxygen demand (BOD5) reaching values up to 180 mg.L-1 (MOSTE, 1998).
1.5.2.2. Increasing the water pollution in the delta and the coastal zone
The pollution brought by the Red River is a potential threat for coastal wetlands in the Red
River delta and coastal waters in the South China Sea. Any changes in human activities in the
basin will lead to a change in sediment discharge associated to nutrient loads at the coastal
zone of the Red river delta. For example, the deforestation in the upstream basin will lead to
an increase of floods in the delta, together with a sediment flow which will impact a coastal
24
marine zone larger than before. This problem can influence the coral reefs in the Southeast of
Cat Ba Island (Nguyen Ngoc Sinh et al., 1995). On the other hand, impoundment of large
reservoirs has decrease the sediment supply to the delta wetlands, increasing the salt intrusion
and reducing the production of wet-rice… Thus, changes of land use and hydrological
management can have contradictory or balancing effects able to temporary hide the problems.
Further, eutrophication in estuaries and coastal zones is another serious consequence of
human alteration of nutrient cycles. The increase of nutrient delivery to the coastal zones, and
the changes in their ratios (N:P:Si, Redfield et al., 1963) can decrease the diversity among
planktonic organisms, and modify the transfer of organic matter within the food web, leading
to phytoplankton accumulation, that paradoxically becomes an oxygen consumer. Moreover,
in eutrophied water, toxic algal blooms have been shown to dominate in many estuaries and
coastal zones in the recent decades (Vitousek et al., 1997), causing economical problems,
such as reduction of tourism activities, prohibition of selling fish and shellfish, etc. Excessive
development in tourism activity leading to pollution together with overfishing would already
have seriously contributed to reduce the productivity of the Red River Coastal zone. The
exceptional site of the Ha long Bay, although protected by UNESCO, is henceforth seriously
threatened.
1.6. References
Chinadata, 1998. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, China Statistical
Publishing House, (Basic Information of Yunnan, China) (http://chinadatacenter.org).
Chinadata, 2000. Statistical yearbook of Yunnan, Vol. 1999, Vol. 2000; China Statistical
Collins N.M., Sayer J.A. and Whitmore T.C., 1995. The conservation atlas of tropical forests.
Asia and the Pacific. World Conservation Monitoring Centre.
Dang Anh Tuan, 2000. The Red River Delta - The Cradle of the Nation (in Vietnamese), 53
pp., National University in Hanoi, Hanoi.
Dang Quang Tinh, 2001. Participatory planning and management for flood mitigation and
preparedness and trends in the Red River basin, Vietnam. Workshop international on
Strengthening capacity in participatory planning and management for flood mitigation and
preparedness in large river basin, Bangkok (Thailand) 20th-23rd Nov.
Dupre B., Gaillardet J. and Allegre C., 2002. A l’interface entre ciel et terre: les grands
fleuves d’Asie, Chapitre 2: 123-130. In: Himalaya-Tibet, le choc des continents. Ed.
CNRS and Museum National d’Histoire Naturelle, Paris. ISBN 2-271-05934-8, 191pp.
25
FAO, 1990-1998. Faostat statistics database (Fertilizer data used in China, Vietnam, Laos…),
Faostat database Results, copyright FAO 1990-1998, http://www.apps.fao.org/;
http://www.fertilizer.fao.org/. 2002.
Fullen M.A., Mitchell D.J., Barton A.P., Hocking T.J., Liu Liguang, Wu Bo Zhi, Zheng Yi
and Xia Zheng Yuan., 1998. Soil erosion and Conservation in the Headwaters of the
Yangtze River, Yunnan Province, China. In M.J. Haigh, J. Krecek, S. Rajwar and M.P.
Kilmartin (eds.), Headwaters: Water resources and Soil conservation. pp: 299-306.
Guilcher A., 1965. Prescis d’hydrologie marine et continentale, Masson, Paris. France,
389pp.
IMH, 1997-2004. Journal of Meteo-hydrology, Institute of Meteo-Hydrology in Vietnam,
Hanoi. (Monthly Journal during the periods of from 1997 to 2004).
Jhang Jing, 1996. Nutrient elements in large Chinese estuaries. Continental Shelf Research.
Vol 16 (8): 1023 -1045.
Le Bac Huynh, 1997. The especial flood in the Da River and downstream of the Red River in
August 1996: the role of Hoa Binh reservoir for flood control in the delta. Journal
Vietnamese of Meteo-Hydrology 4(439): 6-15.
Le Qui An, 2003. Environmental plan for the delta of the Red River and some environmental
problems in this area. Vietnamese Project National KC. 08.02. In “Proceedings of the first
conference on Science» of the Project KC.08: Environmental Protection and some
methods for natural disasters, Ministry of Science and Technology. 209 pp. Hanoi.
Lebel L., 1996. Change Challenges for Environment and Development in Vietnam. Seminar
on Environment and Development in Vietnam. Common Room, University House, and J
G Crawford Building, National Centre for Development Studies. Australian National
University, December 6-7. Funded by AusAID International seminar support scheme.
Available at http://coombs.anu.edu.au/~vern/env_dev/papers/pap05.html.
Meybeck M., Chapman D.V. and Helmer R., 1989. Global freshwater quality: a first
assessment. Cambridge, MA, World Health Organization/United Nations Environment
Programme Basil Blackwell, Inc. 306 pp.
MONRE, 2003. Report on water environment monitor in Vietnam in 2003. In “Studies on
Vietnam environmental statement in 2003”. Vietnam Ministry of Environment and
Natural Resources. 150 pp., Hanoi
MONRE, 1997-2004. Vietnamese Ministry of Environment and Natural Resources. Report
annual on hydrological observation in Vietnam, Hanoi.
MOSTE, 1997. Vietnamese general statistics officer, Ministry of Science, Technology and
Environment of Vietnam, general statistics editor, Hanoi. 550 pp.
26
MOSTE, 1998. Documentation on the Red River Delta (1997-1998), Ministry of Science,
Technology and Environment of Vietnam, Scientific and Technical Publisher, Hanoi. 214
pp.
MOSTE, 2000. Environmental statement in Vietnam. Ministry of Science, Technology and
Environment of Vietnam, rapport annual, Hanoi. 210 pp.
Nguyen Huu Khai and Nguyen Van Tuan, 2001. Geography and Hydrology in Vietnam.
Vietnam National University publisher, Hanoi, Vietnam.194pp.
Pham Van Ninh and Nguyen Vu Trong, 1995. Case study report on Red River Delta in
Vietnam - Project on integrated management and conservation of near shore coastal and
marine areas in East Asia region (EAS-35) United Nations Environment program.
Regional coordinating for the East Seas (ESA/RCU), 78pp, U.N. Environ. Programme,
Nairobi.
Nguyen Viet Pho, 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209pp.
Pham Ngoc Dang, Nguyen Quynh Huong, Nguyen Hoang Yen, Le Thai Bat, Vo Chi Trung,
Dang Huy Huynh, Ho Thanh Hai, Pham Van Ninh, Tran Hieu Nhue, Nguyen Chu Hoi,
Luu Van Dieu, Nguyen Kim Thai and Truong Manh Tien., 2001. Environmental
statement in Vietnam in the year of 2000. CEETIA. Report in the NORAD project
“Enforcing the environmental monitoring and assessment in the South Asia and the
modified Mekong basin”. Printed by UNEP RRC.AP. 170pp.
Redfield A.C., Ketchum B.A and Richards F.A., 1963. The influence of organisms on the
composition of sea-water. In: Hill MN (ed.) The Sea, Vol 2. Wiley, New York, p 26-77.
SEAMCAP, 2001. Report on the statement of environment in Vietnam in 2001. Strengthening
Environment assessment and monitoring capabilities in South Asia and the greater
Mekong sub region. SEAMCAP project, UNEP RRC.AP publisher, Klong Luang,
Pathumthani, Thailand. 200pp.
Smith D.D and Dixon C., 1997. Sustainable urbanisation in Vietnam. Geoforum, Vol 28(1):
21-38.
Turc L., 1961. Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle, Ann.
Agron., 12 (I), I 3-49.
Tran Duc Thanh, Yoshiki Saito, Dinh Van Huy, Nguyen Van Lap, Ta Thi Kim Oanh and
Masaaki Tateishi, 2004. Regimes of human and climate impacts on coastal changes in
Vietnam. Regional Environmental Change 4: 49-62
Trinh Quang Hoa, 1998. Water balance for purpose of socio-economic development in the
Red River delta. Proceedings of International Conference on Economic development and
27
environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar, 92pp.,
Hanoi.
To Trung Nghia, 2000. Flood control planning for the Red River Basin, International
European- Asian Workshop: Ecosystem & Flood 2000, June 27-29. 210pp Hanoi,
Vietnam.
UNEP., 1990. China Conservation Strategy, United Nations Environment Programme and
China Environmental Science Press. Beijing.
Vitousek P.M., Chair, Aber J., Howarth R.W., Likens G.E., Matson P.A., Schindler D.W.,
Schlesinger W.H. and Tilman G.D., 1997. Human alteration of the Global nitrogen cycle:
Causes and consequences. Issues in Ecology. Number 1, Published by the Ecological
Society
of
America,
Washington,
USA,
17pp.
Available
at
http://www.esa.org/science/issues/fileEnglish/issue1.pdf.
Vu Van Tuan, 2002. Program of Model construction for general gestation of Thac Ba
reservoirs basin served to the development socio-economy, General Centre for
Meteorology and Hydrology of Vietnam, 220pp. Hanoi.
Vo Tri Chung, 1998. Forests on the Red river Basin, Vietnam. Institute of forestry investment
and planning. Proceedings of International Conference on Economic development and
environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar. 92pp.
Weijin Yan, Chengging Yin and Shen Zhang, 1999. Nutrient budgets and biogeochemistry in
an experimental agricultural watershed in Southeastern China. Biogeochemistry 45: 1-19.
World Bank, 1996. Vietnam water resources sector review. Selected working papers of the
World Bank, ADB, FAO/UNP and NGO Water Resources Sectoral Group, Hanoi, 340 pp.
28
General approach and methodology
CHAPTER 2
General Approach and Methodology:
Modelling, experimental work and nutrient budgets
The modelling approach which forms the backbone of this thesis, aims at establishing the link
between the biogeochemical functioning of a large river system and the constraints set by the
meteorology, the morphology of the drainage network and the human activity in the
watershed. We have made use of the RIVERSTRAHLER model (Billen et al., 1994, 1999,
2001b; Garnier et al., 1995, Garnier et al., 2001), developed since 15 years to describe the
ecological functioning and nutrient transfers in the large regional river systems of Europe,
characterized by oceanic regime (the Seine River: Billen et al., 1994; Garnier et al., 1995;
Billen and Garnier, 1999; Billen et al., 2001b; the Mosel River: Garnier et al., 1999; the
Scheldt River: Billen et al., 2005). The Danube, of continental hydrological regime has also
been studied by the Riverstrahler approach (Trifu, 2002; Garnier et al., 2002b). This approach
had never been applied to a sub-tropical Asian system where the hydrological regime strongly
differs from occidental ones.
The implementation of the model offers a suitable framework to collect and synthesize data
on any river system. Beside the geomorphological, climatological and hydrological data, the
model requires gathering data on the point and diffuse sources and on the water quality,
leading to a deep understanding of the human activity in the watershed.
For trans-boundary watersheds, as already experienced on the Danube, the collection of the
data is difficult. Therefore the task undertaken for the Red River was a priori ambitious, as
the watershed of Red River is almost equally distributed within China and Vietnam.
Beside the difficulty to collect data in two different countries, the scarcity of data available,
particularly in the case of point and diffuse sources and water quality, is another problem.
Whereas routine survey of water quality have been organized in most of developed countries
(at variable temporal and spatial frequencies, however) by Institutions such as Water
Agencies or Ministries because severe pollution problems were encountered as soon as the
1960’s, such survey are not well organized within the drainage network of Vietnam. The
water quality survey presently organized by Vietnamese authorities is that of the Vietnamese
coastal line, comprising the outlet of the Red River at the delta. In view of the 3260 km of
29
coast line in Vietnam, this represents a considerable effort, but such data are not well adapted
to the requirement of a whole catchment modelling of the Red River which is the objective of
this thesis.
In order to fill this gap, a seasonal survey was undertaken in the scope of the programme
Espoir on Water, carried out by the research teams involved in the programme. A sampling
strategy was commonly decided to gather the classical water quality data taken into account in
ecological modelling (measurements on the field: oxygen concentration, temperature, pH,
conductivity; water samples for laboratory analysis: biological oxygen demand, suspended
solids, nutrients –the forms of nitrogen, phosphorus- and the silica, total carbon, chlorophyll a
as an estimator of phytoplankton biomass). For other research perspectives, metallic or
organic micropollutants were investigated in parallel as well as biological compartments, such
as bacteria, algae and zooplankton that were also studied on eco-toxicological point of view
(ESPOIR on Water, 2003).
Whereas a modelling approach was developed on the Urban System (Trinh Anh Duc, 2003),
we here constructed in complement the model for the whole upstream Red River basin. The
modelling of the whole delta rivers was not investigated here, although it should become a
research perspective in near future.
2.1 Modeling the quality of the Red River hydrographic network
2.1.1 What is a model?
First of all, a model is a tool for a synthesis and a creation of knowledge. It helps the
researcher to progress into the understanding of any complex systems. A model also allows
testing the general relevance of the conceptual schema adopted as well as the coherence of the
data gathered. According to Nordstrom (2003) “a scientific model is a testable idea,
hypothesis, theory or combination of theories that provide new insights or new interpretations
of an existing problem”. In any model, we only include properties and relationships needed to
understand those aspects of the real system in which we are interested (Nordstrom, 2003).
The process of constructing a model normally goes through the following steps. Firstly, we
try to identify the dominant elements of the system, which represents the state variables of the
model. For example, to study the water quality and ecological functioning of a system such
as a river, chemical species both under dissolved and particulate forms, the algae controlled
by light and nutrients, the bacteria and the zooplankton species have to be included as
30
variables. A second stage is to give a simplified representation of the complexity of the
interactions between the variables by formulating mathematically the kinetics of the
corresponding processes and determining the parameters of these kinetics (Chen, 1995). The
mathematical relations, algebraic or statistical, are set up under the form of a system of
(differential) equations which can be solved analytically in favorite cases but may require the
implementation of complex computational methods. The quantification of the stocks and
fluxes of all components taken into account within the system is then the major objective,
necessary to reach the knowledge of the functioning of the system in terms of the circulation
of material between its constituents. The system representation, either under conceptual
schemas or as mathematical model, contributes to a rational and scientific view helping to
understand the observed phenomena. To summarize, a model originates from a naturalist
description of the different elements of a system and attempts to apprehend its dynamical
functioning as the result of a series of cause-effects relationships. In the same time, the model
allows identifying the weakness of both of our knowledge and of the conceptual schemas
from which the model was constructed.
An ecological model of a river system should be able i) to reproduce spatial and temporal
variations of the concentrations of the various variables of the river system, ii) to allow
establishing budgets for any sub-systems within the system and iii) to calculate the flux
exchanged between these sub-systems (Sferratore et al., submitted).
Moreover, a model of knowledge able to reproduce some aspects of the behavior of an
ecosystem can be utilized as a predictive tool to simulate the behavior of the same ecosystem
in changing environmental conditions. In this way, although being first a research tool, the
model is possibly becoming a tool for environmental management and planning.
Finally, a model is a powerful tool of communication, particularly in the issue of water and
watershed management, as it makes available under an organized form a large amount of
knowledge.
2.1.2. Some definitions in the context of modeling
Spatial and temporal discretisation: An ecosystem may have a geographically variable
extension. A river system is limited by its watershed whereas a lake is contained in a
topographic depression and the position of its bank varies with the water level. The spatial
dimension of a model (0, 1, 2, 3 dimensions) characterizes the resolution of the studied
domain. The model for water quality of a lake can be a zero dimensional model, when
considered as a well mixed reactor, one dimensional if the depth distribution of variables is
31
considered, tri-dimensional if the heterogeneities in all directions are taken into account. A
model for a fluvial ecosystem in which only longitudinal evolutions are considered is as onedimensional or mono-dimensional. A model which describes the transversal and horizontal
heterogeneity of concentrations of any pollutant could be called bi-dimensional model. We
can also take into account the vertical heterogeneity in a tri-dimensional model (Poulin et al.,
1998).
Biological complexity: The level of complexity of a biological model characterizes the way in
which the variables and the biological processes are treated and taken into account in the
model (Chen, 1995).
State variables and forcing variables: A model allows simulating the spatial and temporal
variations of the physical, chemical and biological variables such as the water temperature,
the concentration in dissolved oxygen, in nutrients, in phytoplankton biomass, etc... The state
variables are calculated by the model. The forcing variables (or constraints) are the ones
provided to the model under the form of file of numeric values. The forcing variables control
the dynamic of the system but are not influenced by it. A state variable can be treated as a
forcing variable when it is easy to measure (the temperature for example) and when it is not
influenced by other variables (in the case of any influence between this variable and other
variables, it must be considered as state variable) (Tauson and Akimov, 1997).
Measured parameters and adjusted parameters: There are many coefficients named
parameters which used in the expression of the relationships describing the dynamics of the
system. An adjusted parameter is the one which can settle the results given by the model close
to the observations. We can be led to modify such a parameter for each river system studied
or even for each river section. Another approach is to experimentally determine the kinetics of
the processes and to determine the associated parameters. The value can then be adjusted
within its confidence interval range (Garnier et al. 2004; Sferratore et al., submitted). The
models based on this kind of approach are called deterministic or mechanistic: the
RIVERSTRAHLER model, used here, is of these two types: i) some parameters being
adjusted for the hydrological part and ii) some other are experimentally determined for the
ecological sub-model (see below).
Verification, setting, validation and control of quality: After that conceptual schema is chosen
and that the mathematical equations are established and analytically or numerically resolved,
the model has to be verified, adjusted and evaluated.
32
The verification controls the accuracy of the analytical or numerical solutions, the validity of
numeric approximation adopted. In this step, the comparison between the results by the model
are compared with the field observations.
This step of parameter adjusting is a procedure of sensibility analyses of the obtained results
with the variations of each and/or all parameter(s).
The validation of the model occurred when it is applied to a data set different from the one
used for verification and adjustment and simulations by the model compared to the
measurements. The validation can interfere or not with the adjustment step. The verification
and the validation represent a step of quality control for the model (Poulin et al., 1998).
2.1.3. The ecological functioning of hydrographic networks: RIVERSTRAHLER Model
2.1.3.1. General principles
The approach used in this study is based on the adaptation of the RIVERSTRAHLER model
which has been developed in the framework of the PIREN-Seine program and some other
international research programs (Billen et al., 1994; Garnier et al., 1995) to relate the
ecological and biogeochemical functioning of the whole drainage network of a large river
system to the constraints set by the climate, the morphology of the river system and the
human activities in its watershed (figure 2.1). It combines a simplified hydrological model
(HYDROSTRAHLER), relating meteorological constraints to hydrology, to an ecological
model (RIVE), describing in-stream ecological processes. Beside the Seine River, this model
has been successfully applied to several European rivers with differing population densities
(Billen et al., 1994; Billen and Garnier, 1999; Billen et al., 2001b; Billen et al., 2005; Garnier
et al., 1995; Garnier et al., 1999; Garnier et al., 2002b).
The RIVERSTRAHLER model takes into account the whole drainage network according to
the concept of stream orders (Strahler, 1957): the complex network of tributaries is
represented by a regular scheme of the confluence of rivers of increasing stream orders with
mean morphological characteristics. One obvious limitation of this approach is the fact that it
only provides simulations of the mean behaviour of tributaries of given orders, instead of
describing a real river with its own local characteristics. However, in order to improve the
geographical resolution, it is possible to apply the approach separately to several sub-basins
and connect the results to a model of the main branch of the drainage network.
33
Figure 2.1: the structure of the RIVERSTRAHLER model (ref: Billen and Garnier, 2000)
Within the whole drainage network, the model considers three kinds of interconnected
objects. The description of the sub-basin is typically based on the concept of stream order
(Strahler, 1957) with an idealized description (see 2.1.3.2). The main branch is represented
with a finer geographical resolution describing the longitudinal profile every kilometer. All
characteristics found on the main branch have to be described: depth, wetted section, length of
the river stretch canalized, geometry of navigation dam (location and water level), confluence
point of tributaries, location of reservoirs when existing, etc. When connected to the drainage
network, lakes and reservoirs (hydraulic annexes) constitutes the third kind of objects that are
taken into account by two ways: i) for large reservoirs, we consider them under the individual
form and describe their morphology and water inflow and outflow ii) lakes and ponds are
represented statistically by mean characteristics by hydrological order.
The first versions of the RIVERSTRAHLER model were developed under Quick-Basic
computing language. Recently, a new version of the model (Seneque 3-Riverstrahler), has
34
been developed (Ruelland, 2004, Ruelland et al., in prep.), Riverstrahler being embedded
within the SENEQUE GIS interface allowing the user to run the model with any structure of
basins and branches, selected on line according to the geographical resolution required for the
studied question (Figure 2.2).
Owing to this new software, the functionalities of the
RIVERSTRAHLER model are multiplied by those of a GIS, allowing an easy extraction of
the data required for separate runs of the Riverstrahler. The Seneque 3-Riverstrahler
developed under the Visual Basic in the Windows environment has been utilized here. This
version that comprises some software accessories is much friendlier for use. It requires
however to assemble a complete set of geo-referenced data on the different constraints under
the form of a GIS data base.
Figure 2.2: One of the working screens of the SENEQUE/Riverstrahler GIS software.
2.1.3.2. The hydrological model
The HYDROSTRAHLER model (Billen et al., 1994; Garnier et al., 2002a) allows simulating
the seasonal variations of the discharge at the outlet of each sub-basin with at a daily time
resolution. This model takes into account the rainfall, the potential evapotranspiration and the
geomorphological data which determine the flow rate. It is based on a simple representation
35
of the rainfall-discharge relationship considering the exchanges between 2 reservoirs: the soil,
contributing surface runoff, and the aquifer contributing base flow (Bultot and Dupriez,
1976). The model involves 4 parameters (soil saturation, infiltration rate, internal flow rate,
groundwater flow rate), and distinguishes between three components of the specific discharge
from the watershed: the base flow supplied by the water table, the internal (or hypodermic)
flow supplied by the soil reservoir, and the surface runoff supplied in periods of soil
saturation (Figure 2.3), (Billen et al., 1994; Garnier et al., 2004).
PLU
ETR
solsat
SW
superf.runoff
soil
surf.runoff
Infiltration
baseflow
GW
total spec discharge
groundwater
Figure 2.3: Representation of the rain-discharge relationship in the Hydrostrahler model
The discharge calculated for each order of any sub-basins and every km along the main
branch can be compared to the available observations. The simulations can be adjusted to the
data by considering an initial level of aquifer NAPo (mm) and by adjusting the 4 parameters
of the HYDROSTRAHLER model:
i) the level of soil saturation: SOLsat (mm)
36
ii) the infiltration rate: rinf (d-1)
iii) the internal flow rate: rssr (d-1)
iv) the aquifer flow rate: rgwr (d-1)
The daily variations of the soil water content (SW, mm) and of the groundwater stock (GW,
mm), as well as the specific discharge (mm.d-1) of any elementary watershed, are calculated
from rainfall (mm.d-1) and evapotranspiration (mm.d-1), as follows:
The evapotranspiration is taken equal to the potential evapotranspiration excepted when SW>
0.1 solsat, in which case evapotranspirtation is set to zero.
The total specific discharge is calculated as qspec tot = qbaseflow + qsurf.runoff
in which
The base flow supplied by the water table: qbaseflow = rgwr . GW
The infiltration from the soil water to the aquifer: infiltration = rinf. SW
The specific surface discharge is the sum of the superficial runoff and the (sub)surface runoff:
qsur.runoff = rssr. SW + qsup.runoff ,
the superficial runoff, qsup.runoff , is only supplied in periods of soil saturation:
If SW > solsat then = PLU-ETP else = 0
Within a sub-basin, the total discharge (Q, m3.s-1) in order n tributaries is calculated as the
sum of the discharges of their two n-1 order tributaries, the discharges of lateral tributaries of
order 1 to n-1, and the flow from its direct watershed, i.e. the part of the watershed which
does not belong to the catchments of the tributaries. In the main branch, the discharge is
calculated from the discharge of the tributaries and that of the direct watershed (Figure 2.4).
The main merit of this approach is that at any point in the drainage network, the baseflow and
the surface runoff component of the total discharge can be distinguished, which is the key for
taking into account the diffuse sources of material from the watershed (see below).
From the value of the discharge, calculated by stream order, width (w, m) and slope (s, m.m1
), mean depth (d, m) and flow velocity (v, m.s-1) are calculated by rearranging of the
Manning's empirical formula (Billen et al., 1994). The flow from the direct catchment area of
the river, or from its lateral tributaries of lower stream orders, 'dilutes' the water masses
flowing through the main channel. The corresponding dilution factor and its variations with
stream order and the season are very important for controlling the ecological functioning of
rivers.
37
In the main branch of the river, the calculation is similar, taking into account the contribution
to the flow of both the direct watershed and the considered sub-basins. In regulated sectors,
the values of the depth and the wetted section are taken into account.
n-2
Qn = 2.Q(n-1)
+ Q(lateral tributaries)
+ Q(direct watershed)
n-1
n
Figure 2.4: Calculation of water flow in the HYDROSTRAHLER module
In the case where reservoirs are present, their role in the hydrological regime must be taken
into account. This will be discussed in chapter 3.
2.1.3.3. The biogeochemical and ecological model: RIVE
The basic assumption in the RIVERSTRAHLER model is the unity of the microscopic
processes (biological and physical-chemical) involved in the functioning of river systems, i.e.
the kinetics of the processes are the same from headwaters to downstream sectors, whatever
the object considered (sub-basins, branches or stagnant annexes). On the contrary, the
hydrological constraints control their expression and differ widely along the upstreamdownstream gradient as do the constraints due to inputs from point and diffuse sources.
Therefore, the specificity of the ecological structure and function of the different sectors of
the river continuum depend on the constraints, rather than on the nature of the processes
involved.
A same model takes into account ecological processes (RIVE: see Garnier et al., 1999 where
developments taken into account in this version are included), and hence allows describing
the main variables of water quality.
Coupled to the HYDROSTRAHLER model, the RIVE model calculates the seasonal and
spatial variations of 22 variables characterizing the water quality and ecological functioning,
including nutrients (nitrate (NO3-) and ammonium (NH4+) dissolved phosphate (PO43-) and
particulate inorganic phosphorus – PIP- and dissolved silica - (SiO2) two taxonomic groups of
phytoplankton (diatoms and Chlorophyceae, Garnier et al., 1995), two kinds of zooplankton
(rotifers with a short generation time and microcrustaceae with a long generation time,
Garnier et al., 1999) and two compartments of bacteria (the small bacteria autochthonous and
38
the large bacteria allochtonous, Garnier et al., 1991), (Figure 2.5, Table 2.1). The description
of the phytoplankton dynamics is based on the Aquaphy module by Lancelot et al. (1991)
which distinguishes between photosynthesis -controlled by light intensity- and algal growth controlled by nutrient availability-. The module has been adapted to two groups of algae
(diatoms and non diatoms) and a formulation for loss processes by excretion and grazing has
been added (Garnier and Billen, 1993; Garnier et al., 1998). The degradation of organic
matter and heterotrophic bacterial dynamics are described according to the HSB module
(Billen and Servais, 1989) and split into two bacterioplankton compartments (the small
bacteria autochthonous and the large bacteria allochtonous, Garnier et al., 1992; Barillier and
Garnier, 1993) and also the nitrifying bacteria. The RIVE model also includes a calculation
of nutrient exchanges across the sediment-water interface (Venice) as a result of a given
sedimentation flux of organic material, taking into account organic matter degradation,
associated ammonium and phosphate release and oxygen consumption, nitrification and
denitrification, phosphate and ammonium adsorption onto inorganic material, mixing
processes in the interstitial and solid phases and accretion of the sedimentary column by
inorganic matter sedimentation (Billen et al., 1989 ; Sanchez, 1997; Billen et al., 1998).
Sedimented biogenic silica is re-dissolved (Garnier et al., 2004). Water column nitrification
(Brion and Billen, 1998; Brion et al. 2000) and phosphate adsorption on suspended inorganic
particles (and their subsequent sedimentation) are also taken into account in the model.
Table 2.1: Kinetic formulation of the processes taken into account in the RIVE model, and
values of the corresponding parameters (in Garnier et al., 1999)
Process
Kinetic expression
Parameters
meaning
Diatoms
Units
0.2
Chlorophyc.
0.5
kmax*
maximal rate of photosynth.
α
srmax*
initial slope of P/I curve
0.0012
0.0012
h-1/(µE.m-2 s -1)
max. rate of reserve synthesis
0.15
0.37
h-1
Ks
1/2 saturation cst
0.06
0.06
kcr.*
rate of R catabolism
0.2
0.2
Phytoplankton dynamics
Photosynthesis (phot)
kmax (1-exp-(α I/kmax)) PHY
reserves synthesis
srmax M(S/PHY,Ks) PHY
reserves catabolism
kcr R
h-1
h-1
growth (phygrwth)
mufmax M(S/PHY,Ks) lf PHY
mufmax
max. growth rate*
0.07
0.14
h-1
nutrient limitation factor
with lf = M(PO4,Kpp)
or M(NO3 +NH4, Kpn)
or M(Si02 , KpSi)
Kpp
Kpn
KpSi
1/2 sat. cst for P uptake
1/2 sat. cst for N uptake
1/2 sat. cst for Si uptake
15
70
0.42
46
70
-
µg P liter-1
µg N liter-1
mgSiO2 liter-1
respiration
maint PHY +ecbs phygrwth
maint*
ecbs
maintenance coefficient.
energetic cost of biosynthesis
0.002
0.5
0.002
0.5
h-1
-
excretion (phyex)
exp phot.+ exb PHY
exp
exb
"income tax" excretion
"property tax" excretion
0.0006
0.001
0.0006
0.001
h-1
h-1
lysis (phylys)
kdf + kdf (1+ vf)
kdf*
vf +
mortality rate
parasitic lysis factor
0.004
0 / 20
0.004
0 / 20
h-1
-
phyto sedimentation
(vsphy/depth).PHY
vsphy
sinking rate
.004
.0005
m/h
NH4 uptake
NO3 uptake
phygrwth /cn NH4/(NH4+NO3)
cn
algal C:N ratio
7
7
g C(g N)-1
phygrwth /cp
cp
algal C:P ratio
40
40
g C(g P)-1
phygrwth /cSi
cSi
algal C:Si ratio
2
-
g C(g Si02)-1
p(T) = p(Topt).exp(-(T-Topt)² /
dti²)
Topt
dti
optimal temperature
range of temperature
18
13
35
17
°C
°C
PO4 uptake
Si02 uptake
temperature dependency
phygrwth /cn NO3/(NH4+NO3)
39
Process
Kinetic expression
Parameters
Total zooplankton.
Zooplankton dynamics
ZOO growth
(zoogwth)
µzox.M(PHY-PHYo),KPHY).ZOO
µzox
KPHY
PHYo
max. growth rate
1/2 sat cst to PHY
threshold phyto conc.
0.02*
0.4
0.1
h-1
mgC/l
mgC/l
ZOO grazing
grmx.M((PHY-PHYo)
KPHY).ZOO
grmx
max grazing rate
0.035*
h-1
ZOO mortality
kdz.ZOO
kdz
mortality rate
0.001*
h-1
dti²)
Topt
dti
optimal temperature
22
12
small
bac
°C
°C
Bacterioplankton dynamics
large
bac
HPi production by lysis
εpi . (phylys+bactlys+zoomort)
εp1
εp2
εp3
HP1 fraction in lysis pducts
HP2 fraction in lysis pdcts
HP3 fraction in lysis pdcts
enzym. HPi hydrolysis
kib.HPi
k1b
k2b
HP1 lysis rate
HP2 lysis rate
HPi sedimentation
(vsm/depth).Hip
Vs
Hip sinking rate
0.05
m/h
Hid production by lysis
δe . (phylys+bactlys+zoomort)
εd1
εd2
εd3
HD1 fraction in lysis pdcts
0.2
0.2
0.1
-
enzym. HDi hydrolysis
eimax. M(HDi,KHi).BAC
e1max
e2max
KH1
KH2
max. rate of HD1 hydrolysis
max. rate of HD2 hydrolysis
1/2 sat cst for HD1 hydrol.
1/2 sat cst for HD1 hydrol.
0.75
0.25
0.25
2.5
0.75
0.25
0.25
2.5
h-1
h-1
mgC/l
mgC/l
direct substr. uptake
bmax. M(S,Ks).BAC
bmax
Ks
max. S uptake rate
1/2 sat cst for S uptake
0.2
0.1
0.8
0.1
h-1
mgC/l
bact. growth (bgwth)
Y. bmax. M(S,Ks).BAC
Y
growth yield
0.25
0.25
-
bact. mortality (bactlys)
kdb.BAC
kdb
bact. lysis rate
.01
0.1
h-1
bact. sedimentation
(vsb/depth).BAC
vsb
bacteria sinking rate
0
0.01
m/h
0.2
0.2
0.1
-
0.005
0.00025
h-1
h-1
ammonification
(1-Y)/Y.bgwth/cn
cn
bact. C:N ratio
7
gC/gN
PO4 production
(1-Y)/Y.bgwth/cp
cp
bact. C:P ratio
40
gC/gP
dti²)
Topt
dti
optimal temperature
meaning
25
25
15
15
nitrifying bacteria
°C
°C
Units
max growth rate of NIT
1/2 sat cst for NH4
1/2 sat cst for O2
0.05
7
0.6
h-1
mgN/l
mgO2/l
mgC/mg NH4
nitrification and phosphorus dynamics
NIT growth (nitgwth)
µnix.M(NH4,KNH4).M(O2,KO2).
NIT
µnix*
KNH4
KO2
NH4 oxidation
nitgwth/rdtnit
rdtnit
NIT growth yield NIT
0.1
NIT mortality
kdnit.NIT
kdnit*
NIT mortality rate
0.01
h-1
PO4 adsorpt/desorpt.
(planktonic phase)
Langmuir isotherm
Pac
KPads
SM max. adsorpt. capacity
1/2 saturation ads. cst.
0.0045
0.3
mgP/mgSM
mgP/l
dti²)
Topt
dti
optimal temperature
23
16
°C
°C
cm²/s
benthos remineralisation
susp. matter sedim.
(vsm/depth)*MES
vsm
sinking rate
Diffusion (interstitial ph.)
Fick law
Di
app. diffusion coefficient
Mixing (solid phase)
Fick law
Ds
mixing coefficient
2 10-5
2 10-6
orgN mineralis. (maorg)
kib.HPi/cn
orgP mineralis.
kip.HPi/cp
k1p*
k2p*
orgP hydrolysis rate of HP1
orgP hydrolysis rate of HP2
0.05*
0.0025*
h-1
h-1
benth. nitrification
kNi*NH4 (in oxic layer)
kNi
1st order nitrification cst
1
h-1
NH4 adsorpt/desorpt.
1st order equilibrium
Kam
1st order adsorpt. cst for NH4
30
-
PO4 adsorpt/desorpt.
(in benthos)
1st order equilibrium
Kpa
Kpe
PO4 adsorpt. (oxic layer)
PO4 adsorpt. (anoxic layer)
35
1.7
-
SiO2 redissolution
kdbSi.SIB
kdbSi
silica redissolution rate
0.01
h-1
Topt
dti
optimal temperature
25
20
°C
°C
p(T) = p(Topt).exp(-(T-Topt)² / dti²)
m/h
cm²/s
*These parameters depend on temperature according to the relation mentioned.
+ M(C,Kc) = C/(C+Kc) : hyperbolic Michaelis-Menten function .
+ vf: parasitic lysis amplification function. It is maintained at zero while algal density of each group remains lower than a
threshold value of 65 µg Chl a.L-1 and temperature is below 15°C.
40
THE RIVE MODEL
OXY
mineralization
Cyanobac
teria
Flagell
. Chloroph
.
PIP
photos
S
& resp.
PO
4
growth
SS
NH
R
HD
GRA
exoenz
.
1,2 hydrol
HD
Lysis&
excretion
OX
Y
3
Largeheterotr. bact
Smallheterotr.
bact
growth
& resp.
SM
4
nitrif.
NO
BAC
HP
1,2
3
NIT
Diatoms
CO
2 photos S
& resp.
R
HP
3
OXY
mortal
ity
growth
DSi
DIA
growth
& resp
.
grazing
dissol
.
microcrusteaceans
Rotifers
, Ciliates
ads
PO4
nitrif.
PO
4
NO
NH
3
4
OX
Y
organ.
matter
degrad
.
denit.
HP
1,2,3
AnoxicOxic
layer layer
sedim
.
BSi
ZOO
SO4
Figure 2.5: Processes taken into account in the RIVE module (from Garnier et al., 1999)
2.1.3.4. Point sources and non point sources
The point sources and non-point sources within the drainage basin are major constraints that
must be documented for modelling the water quality in any river system and are taken into
account in the RIVERSTRAHLER model. Starting from the level in the headwater streams,
whose water is a mixture of surface runoff and groundwater, the nutrient content evolves from
upstream to downstream of the hydrographic network both because of point discharges of
nutrients and because of the processes that transform, immobilise or eliminate them during
their downward transfer.
Diffuse sources are taken into account through mean nutrient concentrations in each of the
two components of runoff (surface- and groundwater flow) as calculated by the
HYDROSTRAHLER model. The documented variables are NO3, NH4, PO4, PIP, SiO2 and
suspended solids. Regarding nitrates in the surface water, the concentrations are calculated
from the land use in the watershed and from a coefficient of transfer through the riparian
zones (Billen and Garnier, 1999).
41
In the SENEQUE/Riverstrahler version of the model, suspended solids, organic carbon,
nitrogen and phosphorus composition of surface and groundwater flow are automatically
calculated, for each elementary watershed, from the GIS data base on land use, according to a
parameter file that should be documented for each new basin. Silicate content is similarly
calculated from the GIS layer on lithology. The details of the hypothesis used for calculating
the diffuse sources in the case of the Red River are presented in chapter 6.
Regarding the point sources, the variables taken into account in the domestic and industrial
wastewater are suspended mater, organic matter, and the various forms of nitrogen and
phosphorus. Note that SiO2- that typically originates from rock weathering is not considered
as a point source, although a recent work on the largest waste water treatment plant of the
Parisian region has allowed quantifying the amount of silica found in the raw and treated
wastewater (Garnier et al., 2002c). Organic matter in wastewater is an important constraint to
consider. A study carried out on the treated and untreated wastewater in the Paris urban area
(Servais et al., 1999; Garnier et al., submitted) made it possible to convert the variables
provided by sewage networks and treatment plants into state variables in the RIVE model;
biological oxygen demand (BOD) is for example converted into different fractions of organic
carbon. Bacteria brought by the effluents are also taken into account through a relationship
between BOD and heterotrophic bacteria.
Within a watershed, the distributions of all wastewater treatments are taken into account, as
well as the amount of treated or non treated effluents and the kind of treatment (through an
abatement percentage of the concerned variables).
However such kind of data, not necessarily available for European countries, hardly exist in
emerging countries, where wastewater treatment plants are rare, the polluted effluents being
brought directly to streams and rivers in the large cities. The hypothesis made to calculate the
point sources of wastewater in the case of the Red River basin are discussed in chapter 6.
To summarize, the RIVERSTRAHLER model is one of the few available means of modelling
nutrient cycling and ecological functioning of entire drainage networks as a function of the
distribution of natural constraints and human activities in the watershed.
2.2. Experimental work
2.2.1. Sampling campaigns
2.2.1.1. Monthly sampling in the sub-basin and the main branch of the Red River
42
Due to the lack of database on water quality of the Red River system, monthly sampling
campaigns were organized at the outlet of the three tributaries and in the main branch of the
Red River during the years of 2003 and 2004. At the beginning of investigation of water
quality in 2002, only two sampling campaigns were organized in dry season (in February) and
in rainy season (in August).
Thac Ba reservoir
Son Tay
Hoa Binh reservoir
Lien Mac
Figure 2.6: Sampling sites in the Red River system
The sampling sites chosen at the outlet of each of the three upstream sub-basins of the Red
River were those of the hydrological station of the Vietnam territory (see Figure 2.6). For the
Lo River, the samples were collected at the Vu Quang hydrological station, located in Vu
Quang city (Doan Hung district, Phu Tho province). For the Da River, the sampling site was
situated at the Pho Ngoc hydrological station, in Trung Minh city (Ky Son district, Hoa Binh
province). The sampling site of the Thao River was located at the Yen Bai hydrological
station, in Yen Bai city (Yen Bai province). In the delta of the Red River, due to the
complexity of the drainage network , we decided to limit our approach to the main branch of
the Red River at the Hanoi hydrological station, and chose three sampling stations located
between the confluence of the three main sub-basins (at Viet Tri city) and Hanoi city (see
Figure 2.6). In the main branch, from upstream to downstream, samples were taken at Son
43
Tay hydrological station, at Vien Son town (Son Tay district, Ha Tay province) where the
water quality of the Red River is a mixing of that of the three tributaries: Da, Lo and Thao. A
second sampling station is located just upstream the Lien Mac dam (Ha Tay province), which
is the source of the Nhue River (the urban river studied in the French-Vietnamese cooperation
program, ESPOIR on Water program).
Data obtained from this point were used for evaluating of the initial water quality of the Nhue,
seriously impacted by agricultural activities in the watershed and the Tolich River draining
the effluents of Hanoi, as already mentioned. The more downstream sampling station is
located at the Hanoi hydrological station (in Hanoi city), where the river is not completely
impacted by the wastewater of the agglomeration, as it is separated from the river by a huge
hydraulic works (dikes for protection against floods). Some images of sampling campaigns
are introduced in the figure 2.7a.
2.1.1.2. Sampling campaigns for non point source evaluations
Even fewer data are available for nutrient release from cultivated areas, especially in tropical
systems, including Vietnam. For this reason, some samples from agricultural channels in the
North of Vietnam were occasionally taken and analyzed to improve our knowledge on the
characteristics of the Red River basin, to compare the values with the ones scarcely found in
literature. The aim was to document as closely as possible the constraints of the model.
Samples were taken in 2002 and 2003 in regions of various agricultural activities, such as
vegetal culture (cabbage, salad greens…) and rice culture in the suburbs of Hanoi city, in Ha
Tay province (in the delta area) and in Viet Tri city (in the middle land) (figure 2.7b).
The main variables of interests are nutrients as nitrogen (nitrate, nitrite, ammonium) and
phosphorus (phosphate and phosphorus total). Organic carbon (dissolved and particulate
carbon) was also analyzed. Some other measurements such as pH, conductivity, dissolved
oxygen, water temperature were also realized.
2.1.1.3 Sampling campaigns for point source evaluation
Domestic wastewater
No wastewater treatment system for domestic wastewater exists in Vietnam. It can be noted
that wastewater from cities, towns or villages are occasionally diverted to canals, and then
brought to a lake, a small stream, a urban river but sometimes brought to the fields for
fertilizations (in the villages). Unfortunately, data on quantity and quality of domestic
wastewater reaching the surface water are still very poor. To fill this gap, we have therefore
44
collected samples from the various locations inside and around Hanoi city to estimate the
quality of the domestic wastewater in the whole basin of the Red River, and followed the
wastewater circulation in contrasted populated areas to roughly estimate a percentage of waste
really reaching the rivers.
a)
b)
Figure 2.7: a) Sampling campaigns in the upstream of the Red River; b) waste from the non
point-sources and point sources in the Red River basin
Industrial wastewater: sampling campaigns and data collection
Because of lack of a complete database of industrial wastewater as required by the modeling
approach, we have gathered the information of the representative enterprises within the Red
River basin as followings: daily production per enterprise, discharge of effluents, values of
variables such as pH, suspended solids (SS), dissolved oxygen, biological oxygen demand –
BOD-, chemical oxygen demand –COD-, nutrients (NO3, NO2, NH4, N total, PO4, P total).
45
Data were obtained by three ways in 2003: i) collection of available data; ii) elaboration of a
questionnaire; and iii) sampling followed by chemical analyses.
Reports of the Environmental State in Vietnam (reports of 1998, 1999, 2000, 2001) of the
MOSTE (Ministry of Science, Technology and Environment) and other reports issued from
research projects on wastewater in Hanoi and some Vietnamese provinces (JICA: a project of
Vietnam-Japan cooperation, 2000; projects of VAST (Vietnam Academy of Science and
Technology, 2000) were gathered to get the general database.
A questionnaire was elaborated and sent to a number of enterprises for which we got the
addresses by MOSTE (1998-2001). About 200 questionnaires were sent, with the duty to tick
an appropriate box; i) to document the size of the enterprises (range of wastewater effluent in
m3.s-1; number of workers; range of production in ton.day-1); ii) the quality of the wastewater
discharges (ranges of values for variables such as SS, BOD, N total and P total); and iii) the
ways of discharging the effluents (into the river, into a canal, into a lake or a pond, spread on
lands or stored in basin). Unexpectedly, we received about 20 answers.
In addition, we collected and analyzed samples taken from various industrial sectors in Viet
Tri and Hanoi cities. Several factories in Viet Tri as chemicals production, paper production
and textile plants were investigated. Around and inside the Hanoi city, we sampled the Duc
Giang district (West of Hanoi, representing chemicals, paper, wood, battery and electronics)
and the Dong Anh district (North of Hanoi representing electronics, battery, fertilizer, paper,
beer, milk and mechanics; samples were given by the Institute of Environmental Technology)
expected to lead to representative samples. It must be mentioned that it is not always easy to
enter the factories to collect samples so that the samples were mainly taken from wastewater
channels running outside the factories.
The results of water quality to evaluate the pollution sources are presented in the chapter 5.
2.2.2. In-situ measurements and samples analyses
2.2.2.1. In-situ measurements of physical-chemical variables and sampling
Water quality checker, model WQC-22A (TOA, Japan), was used in-situ to measure physicalchemical variables during the sampling campaigns. This instrument consists of the indicator
main body, the sensors and the standard accessories. By the built-in-one type sensor, five
variables such as temperature (0C), pH, conductivity (mS.cm-1) (or salinity, %0), turbidity
(NTU), and dissolved oxygen (DO, mgO2.L-1) were measured. Before each sampling
46
campaign, the instrument was calibrated using the pure water (for DO, turbidity sensors tests)
and using the standard solution (for pH sensor test).
At the sampling site of the hydrological station, surface samples were collected (30 cm below
the surface) at the middle of the river bed, in front of a boat by a sampled auto-collector.
The water samples were kept at 4 °C to 10 °C before treatment, during transportation to the
laboratory.
2.2.2.2. Filtration and preservation of samples in laboratory
Back to the laboratory, all samples were treated to avoid any changes (enrichment in
particulate and colloidal fractions due to a coagulation or transformation by biological and
chemical processes, (nitrification, denitrification, organic matter degradation, oxidation…)
during storage. The filtration was realized with a Gelman Science filter (Pall) equipped with a
high pressure and a high flow rate. Samples were sequentially filtered through:
i) Whatman GF/F paper-filter (glass micro-fiber filters 0.47µm) for dissolved nutrient
analyses as nitrogen (nitrite, nitrate and ammonia), phosphorus (phosphate), for dissolved
carbon (dissolved organic carbon DOC and dissolved inorganic carbon DIC). For SS
determination on the filter, GF/F filter-papers were pre-weighted.
ii) Whatman GF/C paper-filter for chlorophyll a determination.
iii) Whatman Cellulose nitrate membrane filters for silica.
After treatment, all samples were contained in disposable sterile polyethylene flasks (except
the dissolved organic carbon –DOC- samples stored in glass bottles). The samples were stored
frozen (except the silica samples stored at 4°C in the fridge) to minimize any possible
transformation (volatilization or biodegradation) between the sampling and the analyses.
2.2.2.3. Analyses of samples
Nutrient analyses: A Drell 2010 spectrophotometer (HACH, American) was used for all
nutrient analyses carried out at the Vietnamese INPC laboratory. This is a microprocessorcontrolled, single-beam instrument for colorimetric testing with wavelength range of 400900nm and silicon photodiode detector. It can be used both in the laboratory and in the field.
Most of the analyses were also realized at UMR Sisyphe laboratory using a double-beam UV
and visible spectrophotometer (UVIKON 922, KONTRON Instruments). The methods for
nutrient analyses in the laboratories are described in the chapter 5: phosphate, silica and
ammonium were spectro-photometrically determined on filtered water according to Eberlein
and Katter (1984), Rodier (1984) and Slawyck and MacIsaac (1972) respectively; total
47
phosphorus was evaluated on non-filtered samples after sodium persulfate digestion and
mineralization at 110°C in an acidic phase; nitrate was determined after reduction into nitrite
according to Jones (1984).
Chlorophyll determination was done at the Sisyphe laboratory. The chlorophyll was extracted
in 10ml of 90% acetone solution. The optical density of sample was spectrophotometrically
measured (using a 5 cm cell optical path) at 750nm and 650nm, before and after acidification
according to the Lorenzen’s method (1967).
Suspended solids were determined on a pre-weighed standard glass-fiber filter (GF/F) through
which a well-mixed sample was filtered. The material retained on the filter was dried for
about 1 hour at 1030C to 1050C. Taking into account the filtered volume, the increase in
weight of the filter represented the total suspended solids per unit volume (mgSS.L-1).
Dissolved organic carbon: The Total Organic Carbon Analyzer equipment, ANATOC Series
II, (SGE, Australia) was used to determine the dissolved organic carbon –DOC- and of
dissolved inorganic carbon –DIC- (SGE International Pty Ltd, 2002) of a water sample. A
same analyse on filtered water (0.22 µm cellulose acetate membrane filter) allows to
determine the dissolved fraction -DOC and DIC-. The principle is an UV oxidation. At room
temperature and UV light and oxygen, titanium dioxide catalyzes the oxidation of organic
compounds in an aqueous medium, generating carbon dioxide and water. Measurements were
triplicate.
The results of water quality at the outlet of upstream three sub-basins and in the stations in the
main branch of the Red River system are presented and discussed in chapter 4.
Particulate organic carbon analyses were performed on suspended matter harvested on a 12
mm diameter filter GF/F (ignited at 550°C) using a DC-180 Carbon Analyser (Dohrman).
Nutrient budgets (N, P) established at the basin or regional scale offer an insight into the
fluxes of biogenic material cycling in the various ecosystems constituting the terrestrial
regional system, or transferred into aquatic environments. The respective role of natural and
anthropogenic processes can be easily put in evidence. Regional systems differing in their
natural and anthropogenic characteristics can be compared in terms of their biogeochemical
functioning. Few such budgets have been established for Asian and West Pacific systems,
although rivers in this region of the world may supply about 30-40% of water and 60-70% of
sediment loads to the world’s ocean (Milliman and Meade, 1983). This is the first time
48
nutrient budgets are established for the Red River in Vietnam to evaluate the human impact to
natural nutrient cycling in this tropical region.
We present here the principle of the approach used for the establishment of the nutrient
budgets (N and P) for a given watershed. The details of the sources and hypothesis used to
establish the budgets for the Red River system are discussed in chapter 5.
2.3.1. Nutrients cycling in the soil system
The terrestrial soil sub-system of the considered watershed is divided into the forested area
(semi-natural) sub-system and the agricultural soil sub-system.
Both receives inputs from nitrogen atmospheric fixation, nitrogen and phosphorus in the wet
and dry atmospheric deposition, and are subject to losses through soil leaching and erosion.
Nutrient cycling in agricultural soils is described into more details, taking into account inputs
by chemical fertilizer and manure application, and by excretion by domestic animals, outputs
by export of agricultural goods, either consumed by human and animals in the watershed or
commercially exported outside the limits of the system. Commercial imports of food and feed
from outside the region should also be taken into account, which requires the complete
balance of food to be established for the system.
The principles of the soil budget are represented in figure 2.8.
agricultural goods
wood
exp.
N2fix
atm. fertilidepos. zers
Export Imp.
N2fix
cattle
farming
Forested
soils
dom.
act.
agricult
soils
Figure 2.8: Schematic representation of the nutrient budgets in the soil system
49
2.3.2. Nutrient budgets in the hydrosystem
The hydrosystem receives nutrients from the watershed as point and non-point sources. The
non-point sources comprise the amount of nutrient leached and eroded from forested or
agricultural land. The point sources include the domestic wastewater and industrial waste
discharges. The latter are extremely difficult to evaluate in the absence of a detailed census of
industrial water pollution. We have approached this question by estimating nutrient release
rate by ton of material produced from different industrial sectors, and using estimation of
industrial production in the system.
The outputs of nutrient from the hydrosystem represent the nutrient fluxes exported from
basin as calculated by the product of the annual discharge and nutrient concentration at the
outlet of the basin.
The difference between total inputs and total outputs from the hydrosystem allows putting in
evidence retention processes related either to elimination processes, like denitrification, or
retention processes, like sedimentation and storage in reservoir sediments.
The principles of the hydrosystem budget are represented in figure 2.9.
point sources
domestic & industrial activity
Forest soil
leaching
agricultural
soil leaching
river
export
denit
&
reton
Figure 2.9: Nutrient budget in the hydrosystem.
2.4. References
Barillier A. and Garnier J., 1993. Influence of temperature and substrate concentration on
bacterial growth yield in Seine River Water batch cultures. Appl. Environm. Microbiol.,
59: 1678-1682.
50
Billen G. and Servais P., 1989. Modélisation des processus de dégradation bactérienne de la
matière organique en milieu aquatique. In : Micro-organismes dans les écosystèmes
océaniques. Bianchi et coll. Masson. p. 219-245.
Billen G., Dessery S., Lancelot C. and Meybeck M., 1989. Seasonal and year-to-year
variations of nitrogen diagenesis in the sediments of a recently impounded basin,
Biogeochemistry, 8 : 73-100.
Billen G., Garnier J. and Hanset Ph., 1994. Modelling phytoplankton development in whole
drainage networks: the RIVERSTRAHLER model applied to the Seine River system.
Hydrobiologia, 289:119-137.
Billen G., Garnier J. and Meybeck M., 1998. Chapitre12 : Les sels nutritifs: l'ouverture des
cycles. La Seine en son bassin ; Fonctionnement écologique d'un système fluvial
anthropisé,. Meybeck M ., De Marsily G. and Fustec E. (eds). Elsevier. Paris: pp531-561.
Billen G. and Garnier J., 1999. Nitrogen transfers through the Seine drainage network: a
budget based on the application of the Riverstrahler model. Hydrobiologia, 410: 139-150.
Billen G., Garnier J., Deligne C., and Billen C., 1999. Estimates of early industrial inputs of
nutrients to river systems: implication for coastal eutrophication. The Sciences of the Total
Environment, 243/244: 43-52.
Billen G. and Garnier J., 2000. Nitrogen transfers though the Seine drainage network: a
budget based on the application of the 'Riverstrahler' model. Hydrobiologia, printed in
Netherland 410: 139-150.
Billen G., Garnier J. and LeGuern G., 2001a. SENEQUE 1.3 notice d’utilisation. Programme
PIREN-Seine. UMR CNRS 7619 Sysiphe, Paris.
Billen G., Garnier J., Ficht A. and Cun C., 2001b. Modeling response of water quality in the
Seine river estuary to human activity in its watershed over the last 50 years. Estuaries,
24(6B): 977-993.
network of the Scheldt basin over the last 50 years. Hydrobiologia. In press
Brion N. and Billen G., 1998. Une réévaluation de la méthode d’incorporation de 14HCO3pour mesurer la nitrification autotrophe et son application pour estimer des biomasses de
bactéries nitrifiantes. Rev. Sci. Eau, 11 : 283-302.
51
Brion N., Billen G.., Guezennec L. and Ficht A., 2000. Distribution of nitrifying activity in
the Seine River (France) and its estuary. Estuaries, 23: 669-682.
Bultot F. and Dupriez G., 1976. Conceptual hydrological model for an average-sized
catchment’s area. Journal of Hydrology 39:251-292.
Chen D., 1995. Basic concepts of the Design methodology for production management
systems: a state-of-the-art. International Transactions in Operational Ressearch. Vol
2(1): 137-146
Eberlein K and Katter G., 1984. Automatic method for the determination of orthophosphate
and dissolved phosphorus in the marine environment. Fresenius Z. Anal. Chem. 326: 354357.
ESPOIR on Water, 2003. Reports on the international conference in the framework of
ESPOIR project on water in urban river system in Hanoi, Vietnam, Feb. 2003, Hanoi.
Vietnam.
Garnier J., Servais P. and Billen G., 1991. Bacterioplankton in the Seine River: impact of the
parisian urban effluents. Can. J. Microbiol., 38: 56-64
Garnier J., Billen G. and Servais S., 1992. Physiological characteristics and ecological role of
small and large sized bacteria in a polluted river (Seine river, France). Archiv. Hydrobiol.
Ergebn. Limnol. 37:83-94.
Garnier J. and Billen G., 1993. Ecological interactions in a shallow sand-pit lake (Créteil
Lake, France). A modelling approach. In: Nutrient dynamics and biological structure
in shallow freshwater and brackish lakes. Hydrobiologia, 275/276:97-114.
Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and Chlorophyceae
in the drainage network of the river Seine: Observations and modelling. Limnol.
Oceanogr. 40: 750-765.
Garnier J., Billen G., Hanset Ph., Testard P. and Coste M., 1998. Développement algal et
eutrophisation. Chapitre 14. In: La Seine en son bassin Fonctionnement écologique d'un
système fluvial anthropisé, Meybeck M., De Marsily G. and Fustec E. (eds). Elsevier,
Paris.
Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget od the Mosel
drainage network with the concept of heterotrophic/autotrophic sequences: the
Riverstrahler approach. Hydrobiologia, 410: 151-166.
52
Garnier J., Dufayt O., Billen G. and Roulier M., 2001. Eutrophisation et gestion des apports
de phosphore dans le bassin de la Seine. Acte du Colloque, Lyon 2001. Scientifiques and
Décideurs : agir ensemble pour une gestion durable des systèmes fluviaux. 11 p. CD-rom
téléchargeable sur http://www.eaurmc.fr.
Garnier J. and Billen G., 2002a. The Riverstrahler modelling approach applied to a tropical
case study (The Red -Hong- River, Vietnam): nutrient transfer and impact on the Coastal.
SCOPE, Coll. Mar. Res. W., 12: 51-65.
Garnier J., Billen G., Hannon E., Fonbonne S., Videnina Y. and Soulie M., 2002b. Modelling
transfer and retention of nutrients in the drainage network of the Danube River. Estuar.
Coast. Shelf Sci., 54: 285-308.
Garnier J., d’Ayguesvives A., Billen G. and Sferratore A., 2002c. Silica dynamics in the
hydrographic network of the Seine River. Oceanis 29: 139-160.
Garnier J, Billen G. and Cugier Ph., 2004. Drainage basin use and nutrient supply by rivers to
the coastal zone. A modelling approach to the Seine River. p.60-87 In: Drainage basin
nutrient inputs and eutrophication: an integrated approach. P. Wassmann and K. Olli
(Eds). E-book in press, 309p. available at: www.ut.ee/~olli/eutr/
Garnier J., Billen G. and Cébron A., submitted. Modelling nitrogen transformations in the
lower Seine river and estuary (France): impact of wastewater release on oxygenation and
N2O emission. Hydrobiologia.
JICA, 2000. The Study on Environmental Improvement for Hanoi City in the Socialist
Republic of Vietnam, Final report, Data book., Nippon Koei Co. Ltd, Japan Internation
Cooperation Agency (JICA) and Hanoi People's Committee - Socialist Republic of
Vietnam., Hanoi, 350p.
Jones M.N., 1984. Nitrate reduction by shaking with cadmium, alternative to cadmium
columns. Water Res. 18:643-646.
Lancelot C., Veth C. and Mathot S., 1991. Modelling ice-edge phytoplankton bloom in the
Scotia-Weddel Sea sector of the Southern Ocean during spring 1988. J. Mar. Syst., 2 :
333-346.
Lorenzen C.J., 1967. Determination of chlorophyll and phaeopigments: spectrophotometric
equations. Limnol. Oceanogr., 12 : 343-346
Milliman J.D., and Mead R.H., 1983. World-wide delivery of river sediment to the oceans.
The Journal of Geology, 91(1), 1-21.
53
MOSTE., 1998-2001. Environmetal state in Vietnam. Ministry of Science, Technology and
Environment in Vietnam. Report annual, Hanoi. Vietnam.
Nordstrom D.K., 2003. Modelling low-temperature geochemical processes. In series books:
Treatise on Geochemistry, executive editors Holland H.D. and K.K. Turekian, Volume 5:
Surface and Ground Water, Weathering and Soils, Volume Editor: Drever J.I., Elsevier
Pergamon publisher, USA, pp: 37-72.
Poulin M., Even S., Billen G., Mouchel J.M., Garnier J., Levassor A. and Leviandier T., 1998.
Chapitre 6 Modèles: des processus au basin versant. Dans La Seine en son basin :
Fonctionnement écologique d’un system fluvial anthropisé. Meybeck M ., De Marsily G.
and Fustec E. (eds). Elsevier in Paris: 679-717.
Rodier J., 1984. L'analyse de l'eau. 7ème édition. Dunot (ed.), France, 1365 p.
Ruelland D., 2004. SENEQUE, logiciel SIG de modélisation prospective de la qualité de
l'eau. Revue Internationale de Géomatique, 14 (1), 97-117.
Ruelland D., Billen G., Brunstein D. and Garnier J., SENEQUE 3, in prep. A GIS interface to
the RIVERSTRAHLER model of the biogeochemical functioning of river systems. To be
submitted to Ecological Modelling.
Sanchez N., 1997. Le processus de dénitrification dans les sédiments du barrage-réservoir de
la Marne: étude de sa cinétique et modélisation. Thèse Univ. P and M. Curie. 140 pp.
Servais P., Garnier J., Demarteau N., Brion N. and Billen G., 1999. Supply of organic matter
and bacteria to aquatic ecosystems through waste water effluents. Water Research. 33:
3521-3531.
SGE International Pty Ltd., 2002. Guide manual ANATOC Series II TOC Analyzer.
“Installation and Operation Manual of Total Organic Carbon Analyser”, Publication
number: MN 0292E, Revision 2.0.0; 2.02.
Slavyck G. and McIsaac J.J., 1972. Comparison of two automated ammonium methods in a
region of coastal upwelling. Deep-Sea Res. 19:1-4.
Sferratore A., Billen G., Garnier J. and Théry S., Submitted. Modelling Nutrient (N, P, Si)
Budget in the Seine Watershed: Application of the Riverstrahler Model using Data from
Local to Global Scale Resolution. Global Biogeochemical Cycles.
Strahler A.H., 1957. Quantitative analysis of watershed geomorphology. Geophys. Union
Trans. 38: 913-920.
54
Tauson V.L. and Akimov V.V., 1997. Introduction to the theory of forced equilibria: general
principles, basic concepts and definitions. Geochimica et Cosmochimica Acta, 61(23):
4935-4943.
Trifu C. M., 2002. Transfert des nutrients dans le bassin du Danube et apports a la mer Noire:
modelisation et bilans. Thèse de doctorat d'Etat de l'Université Paris VI, 198pp.
VAST., 2000. Proceedings annual of Project Reports on Environments of Vietnam Academy
of Science and Technology in 2000, Hanoi. VAST publisher. 310p.
55
56
Hydrological regime and suspended load: observation and modelling
CHAPTER 3
Hydrological regime and suspended load of the Red River system
(Vietnam): observation and modelling
Abstract
Previous estimates of the suspended-matter loading of the Red River in Vietnam range from
100 to 170 106 t.yr-1, i.e. from 640 to 1060 t.km-².yr-1. The strong dependence on the
hydrology of the suspended-solid transport results in a large year-to-year variability. Based on
available hydrology data from the period 1997-2004, and on a one-year daily survey of
suspended-matter in the three main tributaries of the Red River system in 2003, a simplified
model was built to estimate the mean suspended load of the Red River in present conditions.
The obtained value is 40 106 t.yr-1, corresponding to a specific load of 280 t. km-2.yr-1. It
reflects a 70% decrease of the total suspended load since the impoundment of the Hoa Binh
and Thac Ba dams in the 1980s. The model predicts a further reduction by 20% of the
suspended load in the Red River with the planned construction of two additional dams. Using
measurements of the total phosphorus content of the suspended material in the different Red
River tributaries, we estimated the present phosphorus delivery by the Red River to be 36 106
kgP yr-1.
Keywords: the Vietnamese Red River, suspended-solids, particulate phosphorus, dams
This chapter is submitted as an article in the Journal of Hydrology under the reference:
Le Thi Phuong Quynh, , Josette Garnier, Gilles Billen, Sylvain Théry, and Chau Van Minh (2005, submitted).
57
3.1. Introduction
Sediments transported by rivers to coastal waters play an important role in the global
biogeochemical cycles of many elements. Martin and Meybeck (1979) estimated that over
90% of the global riverine delivery of some major biogenic elements such phosphorus or iron
are transported with suspended matter. Ludwig et al. (1996) showed that 45% of the total
organic carbon annually discharged globally from rivers into the ocean is in particulate form.
At the global and regional scale, the factors controlling riverine sediment fluxes have been
extensively studied (Milliman and Sywitski, 1992; Walling and Fang, 2003; Meybeck et al.,
2003; Syvitski, 2003; Syvitski et al. 2003). Non human factors linked to climate, topography
and lithology of the watershed, obviously play a major role. However, human actions may
also be important, e.g. deforestation, farming, surface mining, road construction and
urbanization among others have led to a 50% increase of global sediment discharge in the last
2,000 years (Milliman and Syvitski, 1992). More recently, on the contrary, dam construction
has caused a significant decrease of sediment loads globally (Milliman, 1997; Walling and
Fang, 2003). Vörösmarty et al. (1997, 2003) estimated that 30% of the world’s sediment flux
becomes trapped in large dams.
South and South-East Asia, due to their peculiar tectonic status linked to the Himalayan
formation, contribute a much larger share than other areas in the world to the global riverine
flux of suspended-solids to the ocean (Holeman, 1968, Miliman and Meade 1983, Milliman
and Syvitski, 1992, Ludwig and al., 1996). Milliman and Sywitski (1992) estimated that one
third of the present global suspended-matter delivered by rivers originates from Southern Asia
(about 20 109 t.yr-1). More recently, Meybeck et al. (2003) stressed that globally, most
sediment is carried to the oceans from a small proportion of the South East Asia and Pacific
Islands land masses. Milliman and Meade (1983) estimated the mean specific sediment yield
of Asian and South-Eastern Asian rivers to 380 t.km-².yr-1 and 600 t.km-².yr-1, compared to a
global mean of 116 t.km-².yr-1 (Milliman and Meade 1983).
The Red River in Vietnam and China is a good example of a South-East Asian river system
strongly affected by human activities. Its overall sediment load has been ranked 15th in the
world (Milliman and Syvitski, 1992). It is doubtful however whether these general estimates,
which have for decades been cross-cited by many authors (Holman, 1968; Meybeck, 1989;
Ludwig et al, 1996; Van Maren and Hoekstra, 2004), take into account the recent man-made
changes (e.g. deforestation of river systems).
In this study, we analyze a set of measured discharges and suspended-solid concentrations,
gathered from several sources in Vietnam, in order to assess the present suspended-matter
58
loading of the Red River system and the trends of its variations over the last 40 years.
Furthermore, we use a simplified modeling approach to predict possible future trends of
particle transport by this large sub-tropical river system. Another aim of the study is to
examine the link between the suspended-matter and the total phosphorus load in the Red
River. Recently, a study of the nutrient budget of the Red River and its major tributaries (Le
Thi Phuong Quynh et al., 2005) related the human activity in the watershed to nitrogen and
phosphorus delivery to the South China Sea. It showed that phosphorus deserved a more
detailed investigation, because of its close relationship with suspended-sediment transport.
3.2. General characteristics of the Red River basin
3.2.1. Geomorphology
The Red River basin (Figure 3.1) is located in South-East Asia (from 20°00 to 25°30 North;
from 100°00 to 107°10 East) and drains an area of 156 451km², of which 50.3% in Vietnam,
48.8% in China and 0.9% in Laos. The Red River, bordered by the Truong Giang and the
Chau Giang River basins of China, to the North, the Langcang River (Mekong) basin, to the
West, the Ma River basin (in Vietnam), to the South, and the Thai Binh River and the South
China Sea (Tonkin Bay), to the East, is the second largest river in Vietnam (Nguyen Ngoc
Sinh et al., 1995). The Red River gets its name from the reddish-brown colour caused by its
high load of iron-dioxide-rich sediments.
The Red River has its source in the mountainous Yunnan province, in southern China, at an
elevation of 2,000 m (Nguyen Huu Khai and Nguyen Van Tuan, 2001). It is named Yuan
River in China, and flows into Vietnam at Lao Cai where it is named Cai, Thao or Hong
River. The Red River then runs through 7 Vietnamese provinces before flowing into the
China Sea through 4 distributaries called, Ba Lat, Lach Gia, Tra Ly and Day (Figure 3.1),
(Dang Anh Tuan, 2000). The Red River receives two major tributaries: the Da and Lo rivers.
The Da River reaches the Thao River in Ha Nong district, at Viet tri city, position:
105° 20’50’’E and 21°15’00’N. The Lo River joins the main branch of the Red River at Viet
Tri city, slightly downstream, at position 105° 26’40’’E and 21°17’50’N. Some general
characteristics of the sub-basins (morphology, lithology, land use, population) are listed in
Table 3.1.
59
a)
100 km
Thac Ba
Reservoir
b)
Lo
R.
Th
ao
R.
Luc Nam R.
Red R.
Duong R.
Da R
.
Son Tay
Hanoï
D
ay
Hoa Binh Reservoir
Haiphong
R.
ue
Nh
Hoa Binh
R
.
B
tic
oi
h
Tra Ly
R
.
50 km
Ba
Tonkin Bay
La
t
Day
R.
nh
Ni
Co
Figure 3.1: Map of the Red River basin, a) its three major upstream tributaries and b) its delta
area
60
Table 3.1: Some characteristics of the Red River and its main tributaries (Nguyen Viet Pho
1984; Nguyen Viet Pho et al., 2003; MOSTE, 1997; Nguyen Huu Khai and Nguyen Van
Tuan, 2001; Tran Duc Thanh et al., 2004; Dürr, 2003).
Thao
Da
Lo
Hong
(Yen Bai) (Hoa Binh) (Viet Tri) Delta total
Red River
(total)
Topography
Basin area, km²
Length main branch, km
Maximum elevation, m
Slope, ‰
57 150
51 285
34 559
9435
156 451
902
1010
470
236.5
1138.5
6740
3143
3076
10
6740
33.2
37
20
-
29.9
9.0
0.0
0.1
0
-
0.0
0.0
0.5
0
-
55.5
85.3
72.7
0
-
18.0
0.0
21.5
0
-
16.7
14.7
5.2
0
-
0.9
0.0
0.0
100
-
18.7
12.5
8.1
63
17
14.4
3.0
58.6
3.9
19.8
7.2
3.6
3.9
2.6
5.0
54.2
74.4
22.4
17.8
51.6
4.1
6.2
6.4
5.9
5.4
1.4
0.3
0.6
6.8
1.2
150
101
132
1174
192
Lithology
Plutonic acid rocks, %
Basic volcanic rocks, %
paleozoic sedimentary rocks,
%
Mesozoic silicic rocks, %
mesozoic carbonated rocks, %
alluvial deposits, %
Land use (in 1997)
Rice, %
Industrial and other cultures,
%
Grassland, %
Forest, %
Rocky areas, %
Urban area, %
Population (in 1997)
population density, inhab/km²
The mountain areas that form a large part of the upstream basin of the Red River are
tectonically very active and show high erosion rates (Fullen et al., 1998). The geologic
substratum of the upper basin is dominated by consolidated paleozoic sedimentary rocks of
complex lithology with variable contributions of mesozoic silicic or carbonated rocks.
Naturally alluvial deposits dominate the delta area. Soils in the upper basins are typically
Ultisols (by U.S. classification) or “red soil” (by Chinese soil classification), while in the
delta area grey soil and alluvial soil dominate (MOSTE, 1997).
61
3.2.2. Meteorology
The climate of the Red River basin is of sub-tropical monsoon type, characterised by the
alternation of a dry and a rainy season, the latter lasting from May to October and receiving
85 – 95% of the total yearly rainfall. Meteorological data (temperature, humidity, rainfall,
solar radiation) were obtained from 12 Vietnamese meteorological stations in the Red River
basin (IMH 1997-2004), as well as from the Kunming station in China (Chinadata, 19982001) for the period 1997 to 2004. The 13 meteorological stations were distributed within the
different sub-basins on the basis of Thiessen polygons (Figure 3.2), and the integrated mean
values by sub-basin were calculated.
Kunming
1228
Annual rainfall values
in period of 1997- 2003
Ha Giang
2332
Lao Cai
1771
Yen Bai
1738
Sa Pa
2578
Tuyen Quang
1635
Phu Tho
1242
Ha Noi
1600
Lai Chau
2413
Son La
1341
≥ 2000 mm
1600 ÷ 2000 mm
Hoa Binh
1930
≤ 1600 mm
0
20 50 70 100km
N
Nam Dinh
1602
Thai Binh
1577
Figure 3.2: Rainfall distribution within the Red River basin: annual values for the period
1997-2003, at the different stations
In the period from 1997 to 2004, the 10-day mean temperature in the three upstream subbasins varied from 14-16°C in winter to 26-27°C in summer. In the Delta area, temperatures
were higher, varying from 17 to 30°C. Humidity averaged from 82 to 84% throughout the
year in the Vietnamese part of the basin (IMH 1997-2004), while it was lower, about 67÷70
%, in the Chinese part (Chinadata, 1998-2001).
62
The mean annual rainfall is 1,590 mm for the whole Red River basin. The rainfall in the
Chinese territory (annual average of 1,230 mm) is much lower than in the Vietnamese part
(1,810 mm). It varies greatly in space (Figure 3.2), with the highest values (2,000-2,600
mm.yr-1) in the upstream area of Lai Chau, Sapa, Ha Giang, smaller values in the upper
Chinese basin (above 1,230 mm.yr-1) and in the intermediate zone of Tuyen Quang, Yen Bai
and Lao Cai (from 1630 to 1770 mm.yr-1), and shows its lowest values in the median zone of
Son La, Phu Tho (1250 to 1350 mm.yr-1). In the Delta, the values are around 1,600 mm.yr-1
(Figure 3.2). Considering the 4 sub-basins, the mean annual values are 1,904 mm.yr-1 for the
Thao, 1,889 mm.yr-1 for the Da, 1,874 mm.yr-1 for the Lo and 1,677 mm.yr-1 in the Delta. In
the period from 1997 to 2004, the lowest rainfall amount was observed in 1999 and the
highest one in 1997 and 2001 in all three upstream sub-basins (Figure 3.3).
PLU
15
20
ETP
-1
Thao sub-basin
PLU-ETP, mm.d
PLU-ETP, mm.d
-1
20
10
5
Da sub-basin
5
20
15
10
5
0
1997 1998 1999 2000 2001 2002 2003
-1
Delta sub-basin
PLU-ETP, mm.d
-1
10
1997 1998 1999 2000 2001 2002 2003
1997 1998 1999 2000 2001 2002 2003
PLU-ETP, mm.d
15
0
0
20
Lo sub-basin
15
10
5
0
1997 1998 1999 2000 2001 2002 2003
Figure 3.3: Evapotranspiration (ETP) and rainfall (PLU) variations (mm d-1) in the upstream
sub-basins of the Red River from 1997 to 2003 (Thao, Da, Lo) and in the Delta sub-basin
(Delta).
Evapotranspiration (ETP, mm) was calculated by Turc’s formula (Turc, 1961), based on
63
monthly temperature (T°C) and sunshine duration (Sdur, h). These data were obtained from the
respective meteorological stations:
ETPmm/month = 0.4 T°C (Ig+50)/(T°C+15)
where
T°C is the atmospheric temperature in 0C during the considered period
Ig is the total solar radiation expressed in cal.cm-2.d-1 during the period, which can be
calculated by:
Ig = IgA (0.18+ 0.62 h/H)
where
IgA is the energy of solar radiation in the absence of atmospheric attenuation,
expressed in cal.cm-2.d-1.
h/H is the relative duration of sunshine, H is the duration of the astronomic day and h,
the duration of the sunshine period per day.
IgA and H values, which only depend on the latitude and the period of the year, are
provided by Turc (1961).
The mean annual evapotranspiration (ETP) (from 1997 to 2004) is rather homogeneously
distributed over the whole basin area, varying within quite a narrow range, from 880 to 1,150
mm.yr-1. Annual ETP values are: 1,040 mm.yr-1 in the Thao basin, 1,040 mm.yr-1 in the Da,
1,000 mm.yr-1 for the Lo and 1,080 mm.yr-1 in the Delta (Figure 3.3).
3.2.3. Population and land use
The population in the Red River basin was estimated at 30 million inhabitants in 1997, of
which 34 % in China, 65 % in Vietnam and less than 1% in Laos (Chinadata, 1998, MOSTE,
1997). The population density in the different sub-basins varies significantly from 101, 132
and 150 inhab.km-2 in the Da, Lo and Thao sub-basins respectively, to 1,174 inhab.km-2 in the
Delta area (Table 3.1).
Land use is very diverse from one upstream sub-basin to another, as well as between the
basins and the Delta area (Le Thi Phuong et al., 2005). Industrial crops (mainly coffee,
rubber, cotton, sugar, tobacco, etc.) dominate (58.1%) in the Lo basin, forests (74.4%) in the
Da basin, paddy rice fields (66.3%) in the Delta area, while the Thao basin is characterized by
a larger diversity of land use including forest (54.2), paddy rice fields (18.7) and industrial
crops (12.8).
64
The forest cover of the Chinese Yunnan province, in the upper part of the Red River basin,
declined from about 60% in the 1950s to 24.2% in 1990 (UNEP, 1990). About 10% of the
land in this province was categorized as severely eroded in the 1980s. Moreover, Fullen et al.
(1998) reported that, over the last 250-500 years, erosion rates in the Yunnan province have
increased fifteen folds as a result of poor land management, cultivation on steep slopes,
deforestation and lack of conservation. Similarly, deforestation has been intense in North
Vietnam, especially in the northern mountains and the center, where the percentage of forest
cover decreased from 95% in 1943 to 17% in 1991 (World Bank, 1996).
3.2.4. Dams and discharge regulation
Table 3.2: Some characteristics of the large dams already impounded (Hoa Binh and Thac
Ba) or planned, in italics (Son La and Dai Thi) in the Red River basin (data gathered from
Trinh Quang Hoa, 1998; Vu Van Tuan, 2002a, 2002b; Nguyen Huu Khai and Nguyen Van
Tuan 2001; To Trung Nghia, 2000).
Name of the dams
Characteristics
Hoa Binh
Thac Ba
Son La
Dai Thi
Da
Chay (Lo)
Da
Gam (Lo)
1985
1972
2010-2015
2010
3.9 – 9.5
0.78-2.94
9.3-25.5
0.5-3.0
1750-1500
200-190
850-750
Surface area, km²
208
235
440
42
Length, m
210
60
-
-
Mean depth, m
50
42
60
70
Water level (normal), m
115
58
215-265
115
Upstream watershed, km2
57285
6170
26000
9700
Electricity production, 106 KWh.yr-1
1920
386
2400-3600
313
River (sub-basin)
Date of impoundment
*Volume (min-max), 109 m3
*critical upstream discharge, m3.s-1
•
parameters used for the hydrological simulations (see text for explanation)
There are two large dams in the Red River watershed: Hoa Binh and Thac Ba (Figure 3.1,
Table 3.2). Hoa Binh, damming the Da River, is the largest one in Vietnam. It was
constructed in 1985, has a surface area of 208 km² and an effective storage capacity of 9.5
km3 (Vu Van Tuan, 2002b; Ngo Trong Thuan and Tran Bich Nga, 1998). Besides protecting
the city of Hanoi from exceptional floods such as the one in 1971, and providing water for
irrigation at low river flow, it serves to generate electric power and provides 40% of
Vietnam’s electricity (7.8 billion KWh). The Thac Ba dam, impounded in 1972 on the Chay
65
River (a tributary to the Lo River) is the second largest in Vietnam, with a surface area of 235
km² and a storage volume of 2.94 km3 (Vu Van Tuan, 2002.a). It provides 0.4 billion KWh
(Dang Quang Tinh, 2001; Vu Van Tuan, 2002.a).
Another large dam, the Son La, upstream of the Hoa Binh on the Da river, is planned to start
operating in 2010-2015. It will have a surface area of 440 km², a total storage capacity of 25.5
km3 (an effective storage of 16.2 km3), and a water level 265 m above sea level (a.s.l.). The
Dai Thi dam is already in construction on the Lo River and is planned to be operational in
2010. It will have a surface area of 42 km², a total storage of 3.0 km3 and a water level a.s.l. of
115 m (Dang Anh Tuan, 2000).
3.3. Hydrological regime of the Red River and its tributaries
3.3.1. Total and specific discharge of the sub-basins
The average annual discharge at Son Tay station (downstream from the confluence of the
three main tributaries) reported by To Trung Nghia (2000) for the period 1902-1990 was
3,740 m3.s-1, corresponding to a specific discharge of 26.1 L.s-1.km-². For the period 19972004, for which we obtained the daily values from MONRE (1997-2004), we calculated a
similar value of 3,389 m3.s-1 (23 L.s-1.km-²).
The discharge measured at Son Tay station is equal to the sum of the discharges of the three
major tributaries Da, Thao and Lo, except in the dry season, when some water is diverted
through irrigation channels (Figure 3.4). The discharge at Hanoi, however, is lower by about
20% in all seasons. Downstream of the Son Tay station, and upstream from Hanoi, 4 main
distributaries divert water from the main branch: the Day River and the Nhue River on the
south-east side and the Ca Lo River and the Duong River on the north-eastern side (Figure
3.1). Since the implementation of the Day River Flood Diversion Scheme in 1937, the Day
River draws water from the Red River main branch through the Day Dam, located 35 km
upstream from Hanoi, open during the flood season. The Red River has an irregular flow and
is subject to flooding. In order to protect Hanoi city, the Day River system has been designed
to be the first and largest flood diversion route in case of emergency. The Nhue River receives
water from the Lien Mac dam with an annual discharge of 24.5 m3.s-1 and serves both to
evacuate wastewater from Hanoï city and irrigate the Delta. The Duong River on the left bank
of the Red River, initially a tributary 5 km upstream from Hanoi, is presently a distributary
that with a mean annual discharge as high as 1,060 m3.s-1 diverted from the Red River to the
Thai Binh River (MONRE, 1997-2004). The Ca Lo river mouth is now almost filled up with
sediment and no longer plays an important hydrologic role. As the complexity of the
66
hydrology of the delta system, comprising the multiple distributaries of the Red River, would
require a separate study, we mainly analysed the outputs of water and suspended-matter from
measurements at the outlets of the three main tributaries and/or at the upstream Son Tay
station on the main branch.
10000
a)
Thao
Da
Lo
discharge, m3.s-1
8000
6000
4000
2000
0
2001
2001
2002
2002
20000
discharge, m3.s-1
b)
2003
2003
Son Tay
Hanoi
Total
15000
10000
5000
0
2001
2001
2002
2002
2003 2003
Figure 3.4: Discharge variations in 2001, 2002, 2003, a) at the outlets of the sub-basins
Thao, Da, Lo, and b) in the main branch of the Red River (at the Son Tay, Hanoi stations).
The sum of the discharge (Total) at the outlet of three upstream sub-basins is shown in
comparison (b).
Over the last 100 years, the maximum daily value at Son Tay station, 37 800 m3.s-1 was
observed in August 1971, while the minimum, 368 m3.s-1 was observed in May 1960. High
floods are always a threat to the highly populated delta area. In the recent history of Vietnam,
67
serious floods causing dykes to break were noted in 1913, 1915, 1945 and 1971 when the
water level in Hanoi reached respectively 11.35 m, 11.2 m, 11.45 m and 13.3 m (the highest
known), (To Trung Nghia, 2000). In fact, the 4 major floods within the return period of 100
years were caused by simultaneous strong floods in the Lo, Thao and Da rivers. The Da River
generally plays the major role, representing 53-57% of total discharge. Since the Hoa Binh
dam was constructed on the Da River (1985), the floods in Hanoi have been fairly well
controlled (Le Bac Huynh, 1997).
The seasonal variations of specific discharge at the outlets of the three upstream sub-basins
and the main branch during the period 1997-2004 (MONRE, 1997-2004) are shown in Figure
3.5. The Da and Lo basins have much higher specific discharges (34 and 25 L.s-1.km-2 as an
annual mean in 2003), than the Thao river (9.6 L.s-1.km-2), which has a large part of its basin
-1
Thao River
250
200
Spec. disch., L.s .km ²
-
200
Spec. disch., L.s .km ²
250
-1
in the drier Chinese territory.
150
100
50
0
150
100
50
0
1997 1998 1999 2000 2001 2002 2003
150
100
50
0
-²
-1
Da River
250
200
Spec. disch., L.s .km
-²
200
-1
1997 1998 1999 2000 2001 2002 2003
250
Lo River
Hong River
150
100
1997 1998 1999 2000 2001 2002 2003
50
0
1997 1998 1999 2000 2001 2002 2003
Figure 3.5: Seasonal specific discharge (Spec. disch.: L.s-1 km-2) at the outlet of the three
sub-basins (Thao, Da, Lo) and in the main branch of the Red River (at Son Tay station,
Hong) from 1997 to 2003.
3.3.2. Modelling the rain-discharge relationship
68
PLU
ETR
If SW > 0.1 solsat then ETR=ETP
else ETR=0
a)
solsat
SW
soil
superf.runoff
If SW > solsat then =PLU-ETP
else =0
surf.runoff
= rssr . SW + sup.runoff
Infiltration
= rinf . SW
baseflow
= rgwr . GW
GW
total spec discharge
= baseflow + surf. runoff
groundwater
b)
Qin
RBA
= qspec. RBA
volmax
Qfill
If Vol<volmax and Qin>qcritf then Qfill = Qin
else Qfill = 0
Vol
Qempt
volmin
If Vol>volmin and Qin<qcrite then Qempt = 0.5 Qin
else Qempt = 0
Qout = Qin – Qfill + Qempt
Figure 3.6: Principles of the hydrological model. a) The Hydrostrahler model (Billen et al,
1994). SW: soil water (mm); GW: groundwater (mm). ETR: real evapotranspiration (mm day1
); ETP: potential evapotranspiration (mm day-1); Solsat: soil saturation content (above which
all excess rainfall is evacuated as surface runoff), infr: infiltration rate, srr : surface runoff rate
and gwrr: groundwater runoff rate. b) Representation of the hydrology of the large dams in
the model. RBA: watershed area upstream from the dam. Qin, Qout: inflowing and
outflowing discharge (m3 s-1); Qfill, Qempt: discharge of filling or emptying of the dam;
volmax, volmin (m3) : minimum and maximum volume of the dam; qcritf, qcrite (m3 s-1):
critical discharge above which the dam is allowed to be filled or below which it is allowed to
be emptied.
69
In order to further explain the differences in mean specific discharge between the sub-basins
as well as their seasonal variations, we tried to relate the specific discharge to rainfall. In
view of the small number of meteorological stations where rainfall data are available, only a
simplified approach was possible. We chose to use the Hydrostrahler model, as described by
Billen et al. (1994). This simple and non distributed model of the rainfall-discharge
relationships considers two water reservoirs in the watershed (Figure 3.6a), i) a superficial (or
soil) reservoir, with short residence time, supplied by rainfall and feeding evapotranspiration,
infiltration and surface/sub-surface runoff, ii) a groundwater reservoir, with longer residence
time, fed by infiltration and at the origin of the base flow. The model involves 4 parameters:
surface runoff rate, soil saturation content (above which all excess rainfall is evacuated as
surface runoff), infiltration rate and groundwater discharge rate. A calculation procedure was
developed to optimize the values of these 4 parameters, based on the Nash criterion (Nash and
Sutcliffe, 1970) calculated with the observed (obsQ) and calculated (calcQ) values of daily
discharge:
Nash = 1 – [Σ (obsQ-calcQ)² / Σ (obsQ-meanQ)² ]
In order to avoid systematic bias related to poor knowledge of the total rainfall over the whole
basin, we adjusted the mean daily rainfall data of the 4-5 meteorological stations available for
each sub-basin by multiplying them with the factor required to equilibrate the annual balance
between observed cumulated discharge at the outlet of the basin and cumulated rainfall minus
potential evapotranspiration. The assumption behind this procedure is that the available
rainfall data provide a correct picture of the temporal distribution of precipitation but only a
poor estimate of its absolute value. The approach was applied to the series of daily rainfall
and discharge data available over the period 1997-2004 for the Thao basin, providing the
simulation of discharge over 3 years shown in Figure 3.7. The calibrated values of the
parameters are listed in Table 3.3. The value of the required rainfall correcting coefficient,
always below 1 for the Thao River, indicates that, in general, the rainfall data gathered from
the Chinese territory, spatially under-represented, overestimated the water balance in the Thao
basin. With corrected rainfall data, a discharge simulation with a Nash criterion above 0.7
could be obtained (Table 3.3). The main advantage of this procedure is that, in the total
discharge of the river, a component corresponding to (sub)-surface runoff can be
distinguished from another corresponding to base flow (Figure 3.7). This is the basis for a
suspended-load model (see below).
70
discharge, m3.s-1
4000
obs.
sim.
base flow
3000
2002
Thao R.
2000
1000
0
0
J 30F 60M 90A120M150J 180J 210
A 240
S
discharge, m3.s-1
4000
270
O 300
N 330D360
2003
3000
2000
1000
0
J j F f Mm A a Mm J j J j A a S s O o N n D d
j
discharge, m3.s-1
4000
2004
3000
2000
1000
0
j
J
j
F
f
m
M A
a
m
j
M J J
j
A
a
S
s
o
O N
n
D
d
Figure 3.7: Simulations and observations of the discharge at the outlet of the Thao sub-basins
from 2002 to 2004 (obs: discharge observations; sim.: discharge simulations; base flow).
71
Table 3.3: Adjusted hydrological parameters of the hydrological model for the three subbasins of the Red River (Thao, Da, Lo). Solsat: soil saturation content (above which all
excess rainfall is evacuated as surface runoff), infr: infiltration rate, srr : surface runoff
rate and gwrr: groundwater runoff rate. Factor PLU.: factor used in the hydrological
model to correct rainfall data and Nash: Nash criterion based on observed
and calculated values by ten-day periods, see text for explanations).
Parameters
Thao
Da
Lo
solsat, mm
110
165
210
0.0619
0.0375
0.05
0.0384
0.0745
0.0675
0.0132
0.0026
0.0010
-1
infr, d
srr, d
-1
-1
gwrr, d
Year
factor PLU.
Nash
factor PLU.
Nash
factor PLU.
Nash
1997
0.65
0.83
0.87
0.69
1.10
0.77
1998
0.68
0.91
1.07
0.75
1.20
0.65
1999
0.76
0.81
1.05
0.80
1.20
0.77
2000
0.61
0.73
1.01
0.81
1.00
0.68
2001
0.81
0.66
1.09
0.90
1.05
0.87
2002
0.73
0.79
1.00
0.86
1.10
0.93
2003
0.58
0.83
0.90
0.86
0.80
0.79
2004
0.51
0.73
0.95
0.51
-
-
The model was adapted to take into account the filling and emptying of a dam, if it is present
in the watershed (case of the Da and Lo rivers, Figure 3.8). Four additional parameters are
taken into account to describe, in a simplified way, the management rules of each dam (Table
3.2): the minimum and maximum volume of the dam and two critical values of the river
discharge above which water is stored (Qcritf), or below which the dam is emptied (Qcrite) to
sustain the downstream flow, provided the volume of the dam has not yet reached its
maximum or minimum value, respectively (Figure 3.6b). The procedure first calculates the
daily specific discharge for the whole watershed area, then the absolute discharge entering the
dam (Qin), considering the watershed area upstream from the dam. It is allowed to store water
when its volume is below the maximum value and Qin is above Qcritf. It is emptied if its
volume is above the minimum value and if Qin is below Qcrite. In all other situations, the
discharge downstream of the dam is equal to the one entering it. The results of the discharge
simulation of the Da and Lo rivers, show major differences in May and June if their dams are
taken into account or ignored (Figure 3.8).
72
10000
3
discharge, m .s
-1
obs.
sim.
sim. without dam
Base flow.
7500
5000
2500
0
0J 30F 60 M90 A120M
150J180J210A
240S270O300N
330 D
360
5000
Lo R., 2003
-1
3
discharge, m .s
Da R., 2003
4000
3000
2000
1000
0
0
J 30 F 60 M90 A120 M150 J180J210A240S270O300N330D360
Figure 3.8: Simulations and observations of the discharge at the outlet of the Da and Lo subbasins in 2003 (obs: discharge observations; sim.: discharge simulations; sim. Without res.:
simulation without the presence of any dam; base flow).
The optimized hydrological parameters obtained on the basis of the observed discharge and
rainfall data during the period 1997-2004 for the Lo river and, to a less extent for the Da
(Table 3.3) show higher infiltration rates and lower groundwater runoff rates than for the
Thao basin, indicating a more stable contribution of base flow in their hydrological regimes.
73
3.4. Suspended-matter loading of the Red River and its tributaries
The results of a detailed survey of daily suspended-material concentration in the Red River
and its three main tributaries in 2003 were made available to us by MONRE. This data base
is the only one available at this frequency for recent years. It concerns the stations Yen Bai,
Son Tay and Hanoï on the Thao River, Hoa Binh on the Da River, and Vu Quang on the Lo
River. We compared these data with earlier results published by several authors in the
Vietnamese and international literature, often providing only monthly or annual means.
3.4.1 Total and specific suspended load
Table 3.4: Sediment load (106 tons.y-1) transported by the Red River gathered from several
studies.
Authors
Total suspended
load, 106 tons.y-1
Remarks
Lisitzin, 1972; Holman, 1968; UNESCO,
1991; Milliman and Meade, 1983; Milliman
and Syvitski, 1992
130
Before Hoa Binh dam
impoundment
Nguyen Viet Pho, 1984, World Bank, 1996
116
Before Hoa Binh dam
impoundment
Meybeck et al., 1989;
160
Before Hoa Binh dam
impoundment
140-150
Before Hoa Binh dam
impoundment
Ludwig et al., 1996
166
Before Hoa Binh dam
impoundment
Van Maren and Hoekstra, 2004
100
Period not stated
This paper, observations in 2003
41
Measurements at Hanoi
station in 2003
38.8
Calculation with the
model for the period
1997-2003
Nguyen Ngoc Sinh et al. 1995
This paper, calculations for the period 19972003
Several authors have reported figures for the total annual loading of the Red River system at
its outlet (Table 3.4). Their estimates range from 100 to 166 106 t.yr-1, i.e. from 640 to 1,060
t.km-².yr-1. As mentioned above, many of these figures are cross-cited from one author to
another, and it is rather difficult to determine to which period they refer. Year-to-year
variations of the hydrology introduce a large variability. Thus, Nguyen Viet Pho (1984)
pointed out that, although the mean annual sediment load of the Red River in the period 1958
to 1971 was about 111 106 tons, it varied from a minimum of 56 106 t.yr-1 in 1963, a rather dry
year, to a maximum of 202 106 t.yr-1 in 1971, when a disastrous flood occurred. Nguyen Viet
Pho et al. (2003) reported that the suspended-solid load at Son Tay station decreased from 114
106 t.yr-1 in the period 1958-1985 to 73 106 t.yr-1 in the period 1986-1997, after the Hoa Binh
74
dam on the Da river has come into operation. The detailed data obtained from MONRE for
the year 2003 at Son Tay and Hanoï stations lead to much lower values, respectively 26 and
41 106 t.yr-1, i.e 178 - 274 t.km-².yr-1 (Table 3.4).
Data provided by Pham Quang Son (1998) and Tran Thanh Xuan and Pham Hong Phuong
(1998) show large differences between the sub-basins with the specific suspended load
varying between 262 - 417 t.km-².yr-1 for the Lo River, 228 – 1,193 t.km-².yr-1 for the Da
River and 551- 1,060 t.km-².yr-1 for the Thao River in the period from 1958 to 1995 (Figure
3.9). These differences are confirmed by our data from 2003, which however show lower
values due to rather dry hydrological conditions (Figure 3.9).
SS, tons.km-2.y-1
1500
Thao R.
1000
500
0
1958-1985 1976-1985 1986-1995
SS, tons.km-2.y-1
1500
2003
Da R.
1000
500
0
1958-1985 1976-1985 1986-1995
Lo R.
1000
500
0
1958-1985 1976-1985 1986-1995
2003
1500
SS, tons.km-2.y-1
Figure 3.9: Distribution of the specific
suspended-solid load (SS, tons.km-2.y1
) between 4 time periods for the three
sub-basins (Thao River at Yen Bai
station, Da River at Hoa Binh station,
Lo River at Vu Quang station) and the
main branch (Hong River at Son Tay
station), (Pham Quang Son, 1998;
Tran Thanh Xuan and Pham Hong
Phuong, 1998; Trinh Dinh Lu and
Doan Chi Dung, 1998 and MONRE,
1997-2004).
SS, tons.km-2.y-1
1500
2003
Hong R.
1000
500
0
1958-1985 1976-1985 1986-1995
2003
75
3.4.2. Seasonal and long-term variations of suspended load
It is a general characteristic of tropical river systems, and of the Red River as well, that most
of the suspended load is transported during the rainy season and at high discharge (Nguyen
Viet Pho, 1984, Pham Quang Son, 1998, Van Maren and Hoekstra, 2004). This is because
during the rainy summer months, both the discharge and the suspended-matter concentration
are high. During a flood, suspended-matter concentrations often increase from 1,000 to 5,000
mg.L-1 in the Thao River, from 500 to 2,500 mg.L-1 in the Da River and from 150 to 500
mg.L-1 in the Lo River (Nguyen Viet Pho et al., 2003). For a given sub-basin, there is a
significant linear relationship between suspended-matter concentration and specific discharge
and the different behaviors of the three sub-basins described above are clearly visible (Figure
3.10). The rather low suspended-matter concentrations in the Lo River, compared to those of
the Thao and the Da rivers before damming are striking. Note that the name of the “Lo” river
means “clear” in Vietnamese, indicating that its low suspended-matter content is an ancient
characteristic. Nguyen Viet Pho et al. (2003) reported that year to year mean suspendedmatter concentrations over the period 1961-1990 in the Lo river, were 710 mg.L-1 at Dao Duc
station (in Ha Giang province), 410 mg.L-1 at Chiem Hoa station (in Tuyen Quang province)
and 290 mg.L-1 at Vu Quang station (in Phu Tho province). These low values might be
surprising when it is remembered that the Lo watershed is the one with the smallest share of
forest and the largest share of industrial crops as compared with the other two sub-basins
(Table 3.2). The lower general slope and different geology of the Lo basin (Table 3.2)
probably explain this paradox.
The changing suspended-matter - discharge relationship observed in the Da River over the
long term (Figure 3.10), with a strong reduction after the filling of the Hoa Binh dam in 1985,
illustrates the prominent role of large dams in the trapping of suspended material (Vorösmarty
et al., 2003). Pham Quang Son (1998) estimated that in the first years of operation of the Hoa
Binh dam, about 50 106 tons.yr-1 of suspended-solids were deposited in the dam
(corresponding to more than 80 % of the total SS flux transported by the upstream part of the
river). Furthermore, Nguyen Viet Pho et al. (2003) reported an interannual mean of
suspended-solid concentrations over the period 1961-1989 decreasing from 1,600 mg.L-1 at
Lai Chau station and 1,430 mg.L-1 at the Ta Bu station (just upstream of the Hoa Binh dam) to
209 mg.L-1 at the Hoa Binh station (downstream of the HoaBinh dam).
The data concerning the Thao river basin shows higher suspended loading at Lao Cai, on the
Chinese border, than at the Yen Bai station, indicating that the upstream part of the basin is
subjected to greater erosion than the lower part (Figure 3.10). Nguyen Viet Pho et al. (2003)
reported mean a suspended-matter concentration of 2,730 mg.L-1 at Lao Cai against 1,760
76
mg.L-1 at the Yen Bai station during the same period, from 1958-1990 (Figure 3.10). No
significant long-term trends appear in the data from the Yen Bai station over the period 1956-
Suspended solids, mg.L -1
2003.
Thao R.
5000
1956-1978
1956-1990
2003
4000
3000
2000
1000
0
Suspended Solids, mg.L -1
0
20
5000
40
60
80
100
120
1964-1968
1981-1984
1990-1993
1986-1997
2003
Da R.
4000
3000
2000
1000
0
Suspended solid, mg.L -1
0
20
5000
40
60
80
100
Lo R.
120
1961-1970
1961-1990
1960-1990
2003
4000
3000
2000
1000
0
0
20
40
60
80
-1
100
120
-2
Figure 3.10: Relationships between suspended-solid concentrations (mg.L-1) as functions of
the specific discharge (Spec. disch.: L.s-1.km-2) at the outlet of the three sub-basins: for the
Thao river at the Lao Cai station in 1956-1978 and at Yen Bai station in 1956-1990 and 2003;
for the Da river at Hoa Binh station for all periods ; for the Lo river (or its tributaries, i.e. LoGam-Chay river), at Thac Ba station in the Chay river in 1961-1970, at the Ghenh Ga station
in the Lo in 1961-1990, at Chiem Hoa station in the Gam river in 1960-1990, at Vu Quang
station in the Lo river in 2003. Linear trends of the relationship are indicated.
77
3.4.3. Modelling the suspended load
On the basis of the simple hydrological model discussed above, providing a daily estimate of
the base-flow and the surface runoff, we proposed a simple model for calculating the
suspended load of each of the three sub-basins. We assumed that the former component of the
discharge is characterized by a constant and low base-line suspended-matter concentration,
while the latter comprises higher suspended concentrations, resulting from erosion processes
and depending on topography, lithology and land use in the watershed. These two values can
be calibrated by optimizing the reconstructed daily variations of the suspended-matter load
with respect to the observed ones, using the Nash criterion as explained above for the
discharge modeling.
Where there is a dam, a simple, steady state, model is taken into account, relating suspended
mater concentration at the outlet of the dam (SMout) to the concentration at the inlet (SMin)
and the hydraulic residence time (τ) in the dam:
SMout = SMin . [1 / (1 + ksed .τ)]
A reasonable value for ksed (day-1) is 0.5 day-1, representing the ratio of the particle setting
rate (about 1 m.h-1) to the mean depth of the dam (about 50 m).
Results of these calculations are compared with the observations in the Thao, Da and Lo
rivers for the year 2003 for which daily suspended-matter concentration data are available
(Figure 3.11). We also applied the model to a set of monthly suspended load data from the
Da River before the Hoa Binh dam was impounded using the same calibrated parameters
(Table 3.5). The Nash criterion, calculated on mean values by decades, ranges between 0.50.76. Note that an alternative model in which the suspended concentration of the surface
runoff component was considered as a linear function of the specific surface runoff, instead of
as a constant, did not provide better results. The model calculations of annual loading for the
Thao, Da and Lo rivers in 2003, respectively 17.8 106, 6.2 106 and 8.8 106 t.yr-1 were very
close to the values calculated from measured daily discharge and suspended-matter
concentrations, respectively 20.0 106, 5.5 106 and 7.9 106 t.yr-1. This model, although
admittedly rather simplistic, is capable of estimating the suspended-solid loading of the three
sub-basins from hydrological data. The retention of suspended-matter by the dam is fairly
well evaluated during the high water period. It is overestimated during low flow (Figure
3.11), possibly because it neglects the role of non- or slowly settling particles, and possibly
that of algal biomass produced in the lakes. Nevertheless, this does not severely affect the
capability of the model to correctly assess the total annual sediment load.
78
4000
simulation
sim.
obs.
observation
3000
2000
1000
0
Thao R.
0
J
30
F
800
60
90
12 0
15 0
18 0
2 10
240
270
300
330
360
M A M J J A S O N D
Da R.
600
400
200
0
0
J
30
F 60 M90 A120 M150J180J210A240S270O300 N330D360
800
Lo R.
600
400
200
0
J
0
30
F 60 M90 A120M150J 180J 210A 240S 270O300N330D 360
Figure 3.11: Seasonal simulations (sim.) and observations (obs.) of suspended-solid
concentrations (mg.L-1) at the outlets of the Thao, Da and Lo Rivers for the year 2003.
Using the same model, and the calibrated parameters of Table 3.5, we calculated the
suspended load over the period 1997-2004, for which the discharge values were modeled and
validated (see above), but daily suspended-matter data are not available. The results provide a
79
mean year-to-year suspended-matter load of 22.5 106, 6.5 106 and 9.8 106 t.yr-1 respectively
for the Thao, Da and Lo sub-basins, and a total suspended-matter loading at Son Tay of 38.8
106 t.yr-1 (Table 3.6).
Table 3.5: Suspended-solid concentrations (SS, mg l-1) in the base flow and surface runoff
determined for the three sub-basins (Thao, Da, Lo) of the Red River to calculate the
suspended-matter load.
SS concentrations, mg l-1
Thao
Da
Lo
Surface runoff concentration
2400
2100
520
Base flow concentration
160
30
30
Table 3.6: Calculations by the model of the mean annual sediment load (106 tons.y-1)
transported by the three main tributaries of the Red River for the year 2003
at the various stations (st.) of the different rivers (R.).
Mean annual sediment load
(106 tons.y-1)
Yen Bai st.
Thao R.
Hoa Binh
st. Da R.
Vu Quang
st. Lo R.
Son Tay st.
Hong R.
with the presence of the two dams
(Hoa Binh and Thac Ba)
22.5
6.5
9.8
38.8
without the presence of the two
dams (Hoa Binh and Thac Ba)
22.5
86.8
12.5
121.8
with the presence of two
additional dams (Son La and Dai
Thi)
with climate change
22.5
3.3
6.5
32.3
27.9
8.5
11.5
47.9
The model was also used to calculate the suspended-matter load in the Da and Lo basins that
would have been observed in the absence of any dams in. The values of 87 106 and 12 106 t.yr1
were obtained for the Da and Lo respectively, resulting a value of 122 106 t.yr-1 for the total
loading at Son Tay (Table 3.6).
To what extent our estimates of the suspended load at the mouth of the three main tributaries
reflect the suspended-matter delivery of the Red River to the sea is difficult to assess. The
Delta area, where the slope of the rivers is very weak and the flow diverted into many natural
and man-made distributaries might be the site of a net sediment deposition. However,
sedimentation processes are observed mostly in areas along the coast or influenced by sea
80
water like Haiphong harbour, within Ba Lat, Ninh Co and Day estuaries, as well as on the
tidal flats of the neighbouring seashore (Nguyen Ngoc Sinh et al., 1995, Van Maren and
Hoekstra, 2004). In 2003, the daily suspended-matter and discharge data at the Hanoi station
showed a suspended-matter flow of 40 106 t.yr-1, close to our loading estimate for the three
main upstream tributaries for the same year. This indicates that no significant net deposition
of solid material occurs in the upper portion of the Delta, upstream from Hanoi.
3.4.4. Relationship between suspended-solid and phosphorus transport
The median phosphorus concentration of suspended-solids in world rivers was estimated by
Meybeck (1982) to 1.15 mgP.g-1. The phosphorus content of suspended-matter was measured
on samples taken monthly at the stations of Son Tay, Hoa Binh, Vu Quang and Hanoï in 2003
with the method described by Rodier (1984). The results show significantly different mean
values in the sub-basins, ranging from 0.42 to 0.85 mgP.g-1 in the upper tributaries (Table
3.7). The values found at the Hanoi station are consistently higher (0.93 mgP. g-1), probably
reflecting the adsorption on the suspended-matter of some phosphorus of human origin in the
lower course of the Red River basin (Table 3.7). As these values show no clear seasonal
variations, we calculated the mean annual particulate phosphorus loading by simply
multiplying the P contents by the respective suspended-solid loads estimated above. The
values obtained compare well with the phosphorus export previously calculated by another
method (Le Thi Phuong Quynh et al., 2005; Table 3.7).
Table 3.7: Phosphorus content (P content, mgP.g-1) of the suspended load (SS load, 106
tons.yr-1) measured at the outlet of the three main tributaries (Thao, Da, Lo) of the Red River
system and in the main branch at Hanoi. Calculation of the corresponding mean annual
phosphorus load (106 kgP yr-1) and comparison with the budget estimates reported by Le Thi
Phuong Quynh et al. (2005).
Thao
P content, mgP.g-1
Da
Lo
Main branch
0.43 ± 0.09 0.68 ± 0.17 0.85 ± 0.28 0.93 ± 0.14
SS load, 106 tons.yr-1
22.5
6.5
9.8
38.8
P load, 106 kgP yr-1 (this study)
9.7
4.4
8.3
36.1
8.3
3.5
5.1
51
P load, 106 kgP yr-1
(see Le Thi Phuong Quynh et al., 2005)
81
3.5. Future scenarios of suspended-matter loading
3.5.1. Effects of planned dams
As mentioned above, the construction of additional large dams is planned for the next 10
years in the Red River basin (Table 3.2). Assuming a constant rainfall regime over a period
of 8 years, the model described above can be used to calculate the effect of these new dams on
the suspended load of the Red river system (Table 3.6). The Son La dam, on the Da River
upstream from the Hoa Binh one will further decrease the suspended load by about 50%. The
Dai Thi dam, in the upper basin of the Lo River, will double the effect of the existing Thac Ba
dam on the total suspended-matter delivery by the Lo river, reducing its load by about 30 %.
As a whole, the total loading of the Red River will decrease from 40 to 32 106 t.yr-1 basing on
the hydrology of the period 1997-2004.
3.5.2. Effect of Climate change
As a result of increased greenhouse gas concentrations in the atmosphere and consecutive
global warming, the hydrological cycle is setting to be amplified. In Asia, in particular, the
Intergovernmental Panel on Climate Change (IPCC, 2001) predicts an annual mean rainfall
increase of about 3 ± 1% in the 2020s and 11 ± 3% in the 2080s, along with a 2-5°C mean
temperature increase. More local or detailed predictions are extremely uncertain because of
the large inter-model variations. Moreover, no models are presently able to predict the effect
of climate change on the frequency of paroxysmal events. In order to gain some insights into
the order of magnitude of the suspended load variations in the Red River induced by climate
change, we ran the described model for the last 8 years, with a 10% rainfall increase, and an
increase of evapotranspiration corresponding to a 3°C temperature rise (which leads to a
roughly 5% increase of ETP) (Table 3.6). With respect to the conditions of the period 19972004, the model predicts an increase of about 20% in the suspended-matter loading, i.e. from
40 to 48 106 t.yr-1.
3.6. Conclusions
With the data presented in this article, a good estimate can be obtained of the total suspendedmatter load presently carried by the Red River main branch and its two major tributaries.
Our estimate is around 40 106 t.yr-1 over the period 1997-2004, corresponding to a specific
load of 280 t. km-2.yr-1. The figures show both the variability linked to year to year variations
in the hydrology and sub-basin to sub-basin differences related to their lithology and
82
morphology. The specific loads of the Thao, the Da and the Lo rivers during the same period
are respectively 394, 127 and 282 t. km-2.yr-1. Our results are lower than the range of previous
estimates (Table 3.4), which, however, in many cases are based on measurements made
several decades ago.
Many authors have discussed the increase in erosion and riverine suspended-solid transport
resulting from forest clearance in South-East Asian countries. The Philippines is a well
documented example where sediment load has increased from 1,100 t.km-2.yr-1 to 4,500 t.km2
.y-1 during the last three decades (Dudgeon et al., 2000). On the other hand, the construction
of large dams has resulted in a decrease of suspended-solid transport by many Asian rivers.
Walling and Fang (2003) thus report the case of the Yellow River in China whose the
sediment flux declined by 50% between the 1950s and the 1980s.
Over the period investigated, dating back to about 40-50 years, the data we collected for the
Red River show no evidence of an increase in suspended loading, even at the scale of
individual sub-basins, despite the well-documented reduction of forested areas and increase of
bare land in the watershed. The suspended-solid loads observed at the Son Tay station in the
1930s by Pouyanne (1931), prior to the period of intense deforestation, were already as high
as 500 mg.l-1 by low flow and 3,500 mg.l-1 by high flow. It seems that the material eroded
from the upstream basin does not reach the downstream course of the Red River and its main
tributaries.
On the other hand, the impoundment of two large dams in the Da and the Lo watersheds has
resulted in a considerable reduction of the total suspended load carried to the sea by the Red
River. Using a simplified modeling approach, we estimated this reduction to about 70%, on
the basis of a calculation carried out with the hydrological data of the last 8 years. The
planned construction of two additional dams would further reduce the total suspended load by
20%, according to our model. This is also the order of magnitude of the expected increase in
suspended loading due to higher rainfall rates induced by climate change (Table 3.6), so that
over the long term, one effect should compensate for the other.
3.7. References
Billen, G., Garnier, J. and Hanset, P., 1994. Modelling phytoplankton development in whole
drainage networks: the RIVERSTRALER model applied to the Seine river system. In: J.
P. Descy, C. S. Reynolds. and. J. Padisak (eds.). Phytoplankton in Turbid Environments:
Rivers
and
Shallow
lakes.
Belgium,
Kluwer
academic
Publishers,
Belgium.
Hydrobiologia 289: 119-137.
83
Chinadata, 1998-2001. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, Vol. 1999, Vol.
2000. China Statistical Publishing House, (Basic Information of Yunnan, China).
(http://chinadatacenter.org).
Dang, Anh Tuan, 2000. The Red River Delta - The Cradle of the Nation. Hanoi national
university publisher, Hanoi.
Dang, Quang Tinh, 2001. Participatory planning and management for flood mitigation and
preparedness and trends in the Red River basin, Vietnam. Workshop international on
Strengthening capacity in participatory planning and management for flood mitigation
and preparedness in large river basin, Bangkok (Thailand) 20th-23rd Nov.
Dudgeon, D., Choowaew, S. and Ho, S.C., 2000. River conservation in South-East Asia. In
Boon P.J, Davies B.R and G.E Petts (eds.). Global perspectives on river conservation:
Science, Policy and Practice. John Wiley& Sons Ltd., Chichester, England. 564 pp.
Dürr, H.H., 2003. Towards a typology of global river systems : some concepts and examples
at medium resolution. Ph-D, University Paris VI - Pierre and Marie Curie. 432 pp.
Fullen, M.A., Mitchel, D.J., Barton, A.P., Hocking, T.J., Liu, Liguang, Wu, Bo Zhi, Zheng,
Yi and Xia, Zheng Yuan., 1998. Soil erosion and Conservation in the Headwaters of the
Yangtze River, Yunnan Province, China. In M.J. Haigh, J. Krecek, S. Rajwar and M.P.
Kilmartin (eds.), Headwaters: Water resources and Soil conservation. pp: 299-306.
Holeman, J.N., 1968. Sediment yield of major rivers of the world: Water Resources Res., 4:
737-747
IMH, 1997-2004. Vietnamese Journal of Meteo-hydrology, Institute of Meteo-Hydrology in
Vietnam, Hanoi. (Journal monthly during the periods of from 1997 to 2004)
IPCC, 2001. Climate change 2001. Intergovernmental Panel on climate change. Third
Assessement Report. http://www.grida.no/climate/ipcc_tar/
Le, Bac Huynh, 1997. The especial flood in the Da River and downstream of the Red River in
August 1996: the role of Hoa Binh dam for flood control in the delta. Vietnamese Journal
of Meteo-Hydrology 4 (439): 6-15
Global Biogeochemical cycles.Vol 19, GB2022, doi 10.1029/2004GB002405.
Lisitzin, A.P., 1972. Sedimentation in the world ocean: Soc.Econ.Paleont.Mineral. Spec.Pub.
17, 218 pp.
84
Ludwig, W., Probst, J.L., and Kempe, S., 1996. Predicting the oceanic input of organic carbon
by continental erosion. Global Biogeochemical Cycles, 10(1): 23-41.
Martin, J.M and Meybeck, M., 1979. Elemental mass-balance of material carried by major
world rivers. Marine Chemistry 7: 173-206.
Meybeck, M., 1982. Carbon, nitrogen and phosphorus transport by world rivers. Am.J.Sci.
282: 401-450
Meybeck, M., Chapman, D. and Helmer, R., 1989. Global freshwater quality: a first
Programme Basil Blackwell, Inc.
Meybeck, M., Laroche, L., Durr, H.H. and Syvitski, J.P.M. 2003. Global variability of daily
total suspended-solids and their fluxes in rivers. Global and Planetary Change 39: 65-93.
Milliman, J. D., and Mead, R. H., 1983. World-wide delivery of river sediment to the oceans.
The journal of Geology, 91(1): 1-21.
Milliman, J. D., and Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment
discharge to the ocean: the importance of small mountainous rivers. The Journal of
Geology 100: 525-544.
Milliman, J.D. 1997. Blessed dams or damned dams? Nature 386: 325-327
MONRE., 1997-2004. Vietnamese Ministry of Environment and Natural Resources. Report
annual on hydrological observation in Vietnam.
MOSTE., 1997. Vietnamese general statistics officer, Ministry of Science, Technology and
Environment of Vietnam, general statistics editor, Hanoi.
Nash, J.E. and Sutcliffe, J.V., 1970. River flow forecasting through conceptual models.
Journal of Hydrology 10: 282–290.
Ngo, Trong Thuan and Tran, Bich Nga, 1998. Erosion of Da and Hong rivers caused by the
Hoa Binh dam operation. Proceedings of International Conference on Economic
development and environmental protection of the Yuan-Red River watershed, Hanoi 4th5th Mar.
Nguyen, Huu Khai and Nguyen, Van Tuan, 2001. Geography and Hydrology in Vietnam.
Nguyen, Ngoc Sinh, Hua, Chien Thang, Nguyen, Chu Hoi, Nguyen, Van Tien, Lang, Van
Ken, Pham, Van Ninh and Nguyen, Vu Trong, 1995. Case study report on Red River
Delta in Vietnam - Project on integrated management and conservation of near shore
85
coastal and marine areas in East Asia region (EAS-35) United Nations Environment
program. Regional coordinating for the East Seas (ESA/RCU).
Nguyen, Viet Pho, 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209 pp.
Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan, 2003. Water resources in Vietnam.
Vietnamese Institute of Meteo-hydrologie. Agicultural Editor (in Vietnamese).
Pham, Quang Son, 1998. Fundamental characteristics of the Red River bed evolution.
Proceedings of International Conference on Economic development and environmental
protection of the Yuan-Red River watershed, Hanoi 4-5 March.
Pouyanne, A.A., 1931. L’hydraulique agricole au Tonkin. Bulletin économique de
l’Indochine, Section A, 34e année, No2, 126 pp.
Rodier, J., 1984. L'analyse de l'eau. 7ème édition. Dunot (ed.), France, 1365 pp.
Syvitski, J.P.M., 2003. Supply and flux of sediment along hydrological pathways: research for
the 21st century. Global and Planetary Change 39: 1-11.
Syvitsk, J.P.M., Peckham, S.D., Hilberman, R. and Mulder, T., 2003. Predicting the terrestrial
flux of sediment to the global ocean: a planetary perspective. Sedimentary Geology 162:
5-24.
To, Trung Nghia, 2000. Flood control planning for the Red River Basin, International
European- Asian Workshop: Ecosystem & Flood 2000, Hanoi, Vietnam, June 27-29.
Tran, Duc Thanh, Yoshiki, Saito, Dinh, Van Huy, Nguyen, Van Lap, Ta, Thi Kim Oanh and
Masaaki, Tateishi, 2004. Regimes of human and climate impacts on coastal changes in
Vietnam. Regional Environmental Change 4: 49-62
Tran, Thanh Xuan and Pham, Hong Phuong, 1998. Impact of Hoa Binh dam on sediments
flux to the downstream of the Red River. Vietnamese Journal of Meteo-Hydrology 4(448):
7-12
Trinh, Dinh Lu and Doan, Chi Dung, 1998. Impact of Hoa Binh dam operation on
hydrological factors at the downstream of the Red River. Vietnamese Journal of MeteoHydrology 4(448): 34-40.
Trinh, Quang Hoa, 1998. Water balance for purpose of socio-economic development in the
Red River delta. Proceedings of International Conference on Economic development
environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar.
Turc, L., 1961. Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle.
Ann. Agron., 12 (I): I 3-49.
86
UNESCO., 1991. Water related issues and problems of the humid tropics and other warm
humid regions. International hydrological Programme. UNESCO, Paris.
UNEP, 1990. United Nations Environment Programme. China Conservation strategy, UNEP
and China Environemental science press, Beijing. China.
Van Maren, D.S. and Hoekstra, P., 2004. Seasonal variation of hydrodynamics and sediment
dynamics in a shallows subtropical estuary: the Ba Lat River, Vietnam. Estuarine, Coastal
and Shelf Science 60: 529-540.
Vörösmarty, C.J, Meybeck, M., Fekete, B. and Sharma, K., 1997. The potential impact of
neo-Castorization on sediment transport by the global network of rivers. In: Walling D.E.
Probst J.L. (Eds.), Human Impact on Erosion and Sedimentation. (Proc. Rabat
symposium, April 1997), IAHS Publication No. 245. IAHS Press, Wallingford, UK, pp
261-273.
Vörösmarty, C.J, Meybeck, M., Fekete, B., Sharma, K., Green, P. and Syvitski, J.PM., 2003.
Anthropogenic sediment retention: major global impact from registered river
impoundments. Global and Planetary Change 39: 169 -190.
Vu, Van Tuan, 2002.a. Report of Program of Model construction for general gestation of
Thac Ba dams basin served to the development socio-economy, General Centre for
Meteorology and Hydrology of Vietnam, Hanoi, 220pp.
Vu, Van Tuan, 2002.b. The Hoa Binh dam and the general gestation of its basin. Report of
General Centre for Meteorology and Hydrology of Vietnam, Hanoi, 250pp.
Walling, D.E and Fang, D., 2003. Recent trends in suspended sediment loads of the world’s
rivers. Global and Planetary Change 39: 111 -126.
World Bank, 1996. Vietnam water resources sector review. Selected working papers of the
World Bank, ADB, FAO/UNP and NGO Water Resources Sectoral Group, Hanoi, 340 pp.
Acknowledgements
This study was realized in the framework of a French-Vietnamese co-operation. Thanks are
due to Georges Vachaud, Research Director at the CNRS, for the coordination of the
programme ESPOIR (CNRS-CNSTV). Le Thi Phuong Quynh’s Ph-D thesis is supported by
the French Ambassy and by the Pierre and Marie Curie University (Paris 6).
87
88
Water quality in the Red River system
CHAPTER 4
Water quality in the Red River System
Few data are available on water quality in the Red River system, both in China and in
Vietnam, excepted those collected at the outlet of the rivers in the delta area by the
Oceanographic Institute of Nha Trang (Dr. Tac An, pers. comm.). For filling this gap of
knowledge, we decided to organize monthly sampling campaigns at the outlet of each three
sub-basins Da, Lo and Thao and in the main branch of the Red River over two annual cycles
(2003 and 2004). In addition, the water quality of the Nhue-To Lich urban system draining
the large city of Hanoi and its densely populated and industrialized surroundings were
analysed, so that it can be compared with the water quality of the contrasted upstream sectors.
The methods are described in Chapter 2. The results of the analyses will be used for both
establishing nutrients fluxes (Chapter 5) and validating the model (Chapter 6).
Beside nutrients (nitrogen, phosphorus, silica), other informative variables (conductivity,
dissolved oxygen, chlorophyll a, etc.) will help to better characterise the water quality and
nutrient status of the Red River.
4.1. Discharge variations
The discharge values in the years do not show much difference in 2003 compared to 2004
(Figure 4.1). During the recent period analysed (1997-2004), the year 2002 was the wettest
one, mainly due to the contribution of the Da R., (see chapter 3, Le Thi Phuong Quynh et al.
submitted).
4000
2000
Thao R.
Lo R.
Da R.
12000
Discharge, m3 s-1
Discharge, m3 s-1
6000
Son Tay
Hanoi
8000
4000
0
0
0.J1.F2.M3.A4.M5.J6.J7.A8.S9.O10N11D
12J
13F
14M15A
16M
17J18J19A20S21O
22N23D
24
00 00 0000 00
00 00 0000 00 .0 .0 .0 .0 .0 .0 .0 .0 2004
.0 .0 .0 .0 .0 .0 .0
2003
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
A20S21O22N23D24
0.J1.F2.M3.A4.M5.J6.J 7.A8.S9.O10N11D12J13F14M15A
16M17J18J19
2003
00 00 00 0000
00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .02004
.0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 4.1: Interannual variations of the discharge in 2003 and 2004, in the three main
tributaries of the Red River (Thao, Da, Lo) and in the main branch at the stations Son Tay and
Hanoi).
89
To summarize, the hydrological regime of the Red River is of monsoon type, with higher
discharges in summer. The contribution of the Da River (on right bank) to the discharge of the
Red River is the highest, the levels being similar for the Thao River (the upstream Red River)
and the Lo River. The discharge of the main branch follows the same trend, but due to
distributaries in the delta, the discharge at the Hanoi station is lower than at the Son Tay
station, located immediately downstream from the confluence of the three main sub-basins
(Figure 4.1).
4.2. Physical-chemical variables
4.2.1. Temperature and conductivity
During the two years 2003 and 2004, the water temperature in the Red River system was in
the range from 17.2 to 30.4 °C, mean value averaging 250 C at the outlets of the three
tributaries (sub-basin) and at the three stations of the main branch (Figure 4.2). The similar
temperature was also found in the urban To Lich and Nhue rivers (Figure 4.2).
Regarding conductivity, whereas values varied in a narrow range around 20 µS m-1 in the Red
River tributaries and its main branch, much higher values (70 µS m-1) were found in the To
Lich, typically indicating the importance of the pollution. This small To Lich River (discharge
around 5 m3 s-1, from 1.5 to 15 m3 s-1 in extreme values) can be considered as a waste water
collector draining Hanoi city (Figure 4.2). The To Lich River represents a significant source
of pollution for the Nhue River (average discharge at 35 m3 s-1, from 8 to 50 m3 s-1 in extreme
values).
The pH values did not vary much in the Red River, as well as in the urban rivers (around 7.4,
extreme values from 6.8 to 8).
4.2.2. Suspended matter and dissolved oxygen
In addition to its role in the equilibrium of oxygen, i.e. aquatic life (production
vs.
respiration, suspended solids (SS) and light climate (Garnier and Benest, 1991; Ryding and
Thornton, 1999; Garnier et al., 2001) are currently monitored in rivers because it is a major
carrier of inorganic and organic pollutants, as well as nutrients (Meybeck et al. 1989). Most
toxic heavy metals, organic pollutants, pathogens, and nutrients such as phosphorus and
appreciable amount of biodegradable organic material are associated to suspended material.
Measurements of suspended solids are also relevant to other environmental issues such as soil
conservation, land denudation, rocks weathering, inputs of elements to the ocean,
sedimentation rate in reservoirs, river bed erosion, etc. (Meybeck et al., 1989).
90
In order to evaluate the representation of our monthly sampling survey, we have plotted the
daily values available for the year 2003 only, with those we gathered during the study (Figure
4.3).
Thao R.
Lo R.
Da R.
30
20
10
0
80
60
40
20
0
M15A16
M17
J 18
J 19
A 20
S 21
O 22
N 23
D 24
0 J 1F 2M3A 4M5J 6J 7A 8S 9O10N11D12J13F14
Temp, °C
30
20
10
0J 1F 2M 3A 4M5J 6J 7A 8S 9O10N11D12J13F14M15A16M17J18J19A20S21O22N23D24
0
80
60
40
20
0
0 J1 F2 M3 A4 M5 J6 J7 A8 S9O10N11 D
12 J13 F14M15A
16M17J18J19A20S21O22N23D24
2004
Nhue R.
To Lich R.
40
30
20
10
0
J
N11D12J13F14
M15A16
M17
J 18
J 19
A20S21O22N23D24
0 F
1 M
2 A
3 4M 5J 6J A
7 8S O
9 10
2003
2004
Nhue R.
To Lich R.
100
Conduct, µS m-1
2003
Son Tay
Lien Mac
Hanoi
100
Conduct, µS m-1
Son Tay
Lien Mac
Hanoi
40
Temp, °C
Thao R.
Lo R.
Da R.
100
Conduct, µS m-1
Temp, °C
40
80
60
40
20
0
0.J1.F2.M3.A4.M5.J6.J7.A8.S9.O10N11D12J13F14M15A16M17J18J19A20S21O22N23D24
0 0 0 0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2002
2003
O10N11D12J13F14
M15A16M17J 18
J 19
A20S21O22N23D2
0.J 1.F 2.M3.A4.M5.J 6.J 7.A 8.S9.
0 0 0 0 0 2002
0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0
.0 .0 .0 .0 .0
2003
Figure 4.2: Seasonal variations, during the years 2003 and 2004, of water temperature and
conductivity (conduct.) in the three main tributaries of the Red River (Thao, Da, Lo) and in
the main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue and To
Lich are shown for comparison for the years 2002 and 2003.
It has been long recognized that low frequency of sampling programs lead to a severe
underestimation of the mean annual suspended solid concentration. This is indeed the case
when we compare the mean of our monthly measurements with those found at the same
station from the daily values: 533 against 698 mg l-1 respectively for the Thao R., 35 against
74 mg l-1 for the Da R. and 92 against 170 mg l-1 for the Lo R..
91
Thao R.
SS, mg l-1
5000
4000
3000
2000
1000
0
0.00 J1.00 F2.00 M3.00 A4.00M5.00J 6.00J 7.00A8.00 S9.00O10.0 N11.0 D12.0
0
0
0
SS, mg l-1
2000
Lo R.
1500
1000
500
0
J 6.00
J 7.00A 8.00S 9.00O 10.0N11.0D 12.0
0.00J 1.00F 2.00M3.00A4.00M5.00
0
0
0
SS, mg l-1
1000
Da R.
750
500
250
0
0.00J1.00F2.00 M3.00A4.00M5.00J 6.00J 7.00A8.00 S9.00O10.0 N11.0D12.0
0
0
0
Figure 4.3: Seasonal variations, during the year 2003, of suspended solids (SS) from daily
(continuous lines) and monthly sampling (open circles) in the three main tributaries of the
Red River (Thao, Da, Lo).
Mean suspended solid concentrations appeared higher in 2004 than in 2003 (Figure 4.4), by a
factor of 5. This difference is significant with respect to the one caused by sampling
frequency. It cannot be explained by the hydrology that was comparable for the two years.
Note here again, as shown in chapter 3, the SS concentrations in the Thao River were much
higher than the ones in the Lo and Da Rivers (Figure 4.4). The SS concentrations in the main
branch typically represented a mixing of the three upstream water masses.
92
Thao R.
Lo R.
Da R.
5000
8
4000
SS, mg l-1
Oxygen, mg l-1
10
6
4
2
3000
2000
1000
0
0
N11D12J13F14
M 15
A16
M 17
J 18
J 19
A 20
S 21
O 22
N 23
D 24
0 J 1F 2M3A 4M5J 6J 7A 8S 9O 10
Son Tay
Lien Mac
Hanoi
10
8
5000
4000
SS, mg l-1
Oxygen, mg l-1
0J 1F 2M3A 4M5J 6J 7A 8S9O10N11D12J13F
14M15A
16M
17J18J19A20S21O22N23D24
6
4
3000
2000
1000
2
0
0
0 J1 F2 M3 A4 M5 J6 J7 A8 S9O10N11D
12 J
13 F
14M15A
16M
17J18J19A20S21O
22N23D24
2003
2003
Nhue R.
To Lich R.
10
8
6
4
2004
500
SS, mg l-1
Oxygen, mg l-1
J0 F1 M
N D J F14151617
M A M J 18192021
J A S O222324
N D
2A
3 M
4 J
5 J6 A
7 8S O
9 10111213
2004
400
300
200
100
2
0
0
A20S21O22
N23D24
0.J1.F2.M3.A4.M5.J6.J7.A8.S9.O10N11D12J13F14M15A16M17J18J 19
0 0 0 0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2002
2003
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A8.S9.O10N11D12J13F14M15A
16M17J18J19A20S21O
22N23D24
0 0 0 0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2002
2003
Figure 4.4: Seasonal variations, during the years 2003 and 2004, of dissolved oxygen and
suspended solids (SS) in the three main tributaries of the Red River (Thao, Da, Lo) and in the
main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue and
To Lich are shown for comparison for the years 2002 and 2003.
Dissolved oxygen concentrations appeared in average lower in 2004 than in 2003 in the
tributaries and also in the main branch of the Red River (Figure 4.4), simultaneously with
lower concentration in suspended solids (SS).
From April to September, i.e. during the rainy season, oxygen concentrations averaged 5.5 mg
O2 l-1 in 2003 against 4.5 mg O2 l-1 in 2004 for the upstream tributaries, and 6.5 mg O2 l-1 in
2003 against 4.5 mg O2 l-1 in 2004 for the main branch. These differences might be explained
by the difference in SS that besides limiting photosynthesis and algal growth are known to be
a support for heterotrophic bacteria which consume oxygen.
93
In the urban river, concomitantly to much lower suspended solids than in the Red River (by
a factor of 100), oxygen concentration was lower due to water organic (domestic) pollution,
and much variable, the water becoming occasionally anoxic.
4.3. General pattern of nutrients
4.3.1. Inter-comparison of nutrient analyses by two laboratories
Before the beginning of this study, the Vietnamese laboratory (INPC, Institute of Natural
Products) did not currently measured nutrients with the standard methods used here.
Therefore a transfer of methods was realized and many samples have been analyzed in
duplicate with the same methodologies in the two laboratories involved in the study (Figure
4.5). The results are compared in x-y graphs (Figure 4.5).
-1
N-NO3, mgN l
2.0
1.5
Sisyphe
1.5
Sisyphe
-1
N-NH4, mgN l
2.0
1.0
1.0
0.5
0.5
0.0
0.0
0.0
0.5
1.0
1.5
0.0
2.0
0.5
1.0
2.0
INPC
INPC
-1
1.5
-1
10
1.5
DSi, mgSi l
P-PO4, mgP l
Sisyphe
Sisyphe
8
6
4
1.0
0.5
2
0
0.0
0
2
4
6
INPC
8
10
0.0
0.5
INPC
1.0
1.5
Figure 4.5: Relationship between the results of analyses carried out in parallel by the two
concerned laboratories (Sisyphe, Paris and INPC, Hanoi) for the stations of the Red River in
the years 2003 and 2004 and one of the Nhue for the years 2002 and 2003.
94
Although the results are in the same range, a considerable variability is observed, due to
several factors. Besides the quality of the spectrophotometer (see method, Chapter 2), the
quality of the chemical product used and the dilution water, the analyzes of the Paris
laboratory were realized on samples that were transported frozen but sometimes thawed at the
arrival (24 hours later) and frozen again until analysis. According to these results, although
interpreting seasonal variations could be speculative when of less than a factor of 2, we can
however state that the general levels of nutrient concentrations are correctly estimated.
4.3.2. Nutrient variations
Since the industrial revolution, human activities have caused strong impact on structure and
function of their environment, including the aquatic environment. In recent years, human
perturbations of agricultural, domestic and industrial origins have largely impacted on water
quality. Some influences, like deforestation, agricultural fertilizers, fossil fuel combustion and
urbanization, result in increasing contamination (N, P, heavy metals) in rivers, while others,
like reservoir construction, soil conservation, result in decreasing concentration of silica and
of suspended solid associated nutrients.
The enrichment of riverine water in nitrogen and phosphorus, together with decreasing
suspended solids and silica, often result in eutrophication of coastal marine (Conley et al.
1993; Billen and Garnier, 1997; Cugier et al., 2005), characterized by non-diatoms harmful
algal blooms.
Nitrate and ammonium
Nitrate content in water river originates mainly from leaching of agricultural lands (Billen et
al., 1998; Billen and Garnier, 1999), but in river sectors impacted by domestic wastewater, a
significant contribution originates from the nitrification of ammonia (Chestérikoff et al.,
1992; Brion et al., 2000, Garnier et al., 2001). In river catchment influenced by agricultural
activities, nitrate contamination increased in parallel with the quantity of fertilizer used.
In the Western European rivers, e.g. the Seine River upstream from Paris, nitrate
concentration has increased by a factor of about 5 from the 1950’s to 2000 (from 1.5 mgN.L-1
up to 8 mgN.L-1) while nitrogen fertilizers application increased from 13 kgN.ha-1.y-1 to 150
kgN.ha-1.y-1 during the same period of time. In Vietnam, according to the FAO database
(FAO, 1990-1998), the use of nitrogen fertilizers has increased 66 folds during the period
from 1961 to 2000 (from 2.2 kgN.ha-1.y-1 to 150 kgN.ha-1.y-1) but the concentrations in the
Red River system are still low (Figure 4.6), compared to those found in Western Europe.
95
2.5
2.0
Thao R.
Lo R.
Da R.
1.5
1.0
0.5
2.5
NH4, mgN l-1
NO3, mg N l-1
0.0
Thao R
Lo R.
Da R.
2.0
1.5
1.0
0.5
0.0
0.
9. 10
J 1.
F 2.
M 3.
A 4.
M 5.
J 6.
J 7.
A 8.
SO
N 11
D 12J 13F 14
M 15A16
M 17
J 18
J 19
A 20
S 21
O 22
N 23
D 24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2.5
2.0
Son Tay
Lien Mac
Hanoi
1.5
1.0
0.5
0.0
NH4, mgN l-1
NO3, mgN l-1
0.J1.F2.M3.A4.M5.J 6.J 7.A 8.S9.O 10N11D12J13F14
M15A16
M17
J 18
J 19
A 20
S 21
O 22
N 23
D 24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2.5
2.0
1.5
1.0
0.5
0.0
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A 8.S 9.
O10N11D12J13F14M15A16M17J 18
J 19
A 20S21O22N23D2
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .
2003
0 0 0 0 0 0 0 0 2004
0 0 0 0 0 0 0
O 10N11D12J13F14
M15A16M17
J 18
J 19
A 20
S 21O22
N 23D24
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A 8.S 9.
00 00 00 00 00
00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .02004
.0 .0 .0 .0 .0 .0 .0
2003
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
25
2.0
Nhue R.
To Lich R.
1.5
1.0
0.5
NH4, mgN l-1
2.5
NO3, mgN l-1
Son Tay
Lien Ma
Hanoi
15
10
5
0
0.0
A20S21O22N23D24
0. J1. F2. M
3. A4. M
5. J6.J7.A8. S9.O10N11D12 J13F14M15A16M17J 18J 19
0 0 0 0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2002
2003
Nhue R.
To Lich R
20
A 4.M5.J 6.J 7.
A 8.S 9.
O 10
N 11D12J 13F 14
M 15
A 16
M 17
J 18
J 19
A 20
S 21
O 22
N 23
D 24
0.J 1.F 2.M 3.
0 0 0 0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2002
2003
Figure 4.6: Seasonal variations, during the years 2003 and 2004, of nitrate (NO3, mgN l-1)
and ammonium (NH4, mgN l-1) in the three main tributaries of the Red River (Thao, Da, Lo)
and in the main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue
and To Lich are shown for comparison for the years 2002 and 2003.
Taken into account the variability of the analyses, it is difficult to put in evidence any
significant seasonal variations in the nitrate concentrations. The highest values which are
logically found in the rainy season in 2003 (nitrate is of diffuse origin), are found in April,
before the rainy season (Figure 4.6). Similarly, the nitrate concentrations in the three main
tributaries might not be significantly different, although the highest values are observed in the
Lo and Thao Rivers. Land use in the Lo basin is dominated by agriculture, while the Thao
basin is the more populated compared to the Da, less impacted. Nitrate concentrations
averaged 0.5 mg N-NO3 l-1 for the Lo and Thao R. and 0.18 mg N-NO3 l-1 for the Da R.
respectively. In the main branch, the nitrate concentrations have similar levels, reflecting the
mixing of the waters (0.31 N-NO3 l-1 at the Son Tay upstream station, 0.36 N-NO3 l-1 at the
downstream Hanoi station, in average). In the urban rivers, nitrate concentrations are much
96
more fluctuating, with mean values much higher than in the Red River: 3 mg N-NO3 l-1 in the
Nhue and 2 mg N-NO3 in the To Lich.
Regarding ammonium, concentrations are usually low in natural waters, as it is taken up very
quickly by microorganisms including autotrophic algae, heterotrophic bacteria, and
autotrophic nitrifying bacteria. High ammonia level in water is typically a sign of domestic
wastewater pollution. The average ammonium concentrations at the outlet of the rivers Thao,
Lo and Da (Yen Bai, Vu Quang and Hoa Binh stations respectively) are very low: 0.1, 0.06
and 0.03 mgN-NH4.L-1, respectively (Figure 4.6).
The mean values increased considerably in the main branch, from upstream (Son Tay station)
to downstream (Hanoi station) i.e., from 0.1 to 0.85 mgN-NH4.L-1, and much more in the
urban rivers, the Nhue (2.7 mgN-NH4.L-1) and the To Lich river (9.5 mgN-NH4.L-1).
Contrarily to the nitrate concentrations that tend to increase during rainy seasons under
leaching from the agricultural lands, ammonium concentrations in the To Lich tended to show
a dilution (Figure 4.6).
Whereas the values found for nitrate are still far below the Vietnamese Standards (15 mgNNO3.L-1), it was not the case for ammonium in urban rivers, which were clearly above the
standards of 1.0 mgN-NH4.L-1).
Considering the proportion of nitrate, nitrite and ammonium in total inorganic nitrogen, it
appeared that nitrate was, in proportion, the dominant form (around 80 %) in the upstream
basins, decreasing in the main branch (from 69 to 25 %) at the benefit of ammonium. The
proportion in nitrite remained low (< 2 %).
In the Nhue and To Lich rivers, the proportion in ammonium reached up to 98 %. Note that,
downstream of the city of Paris, after the treated domestic effluents of the 6.5 million
inhabitants discharging their waste waters to the purification plant of Achères and then being
driven to the lower Seine River, ammonium and nitrate are in a 50 %-50 % proportion (5 mg
N-NO3 l-1 and 5 mgN-NH4.L-1), (Garnier et al., 2001).
97
Table 4.1: Proportion (%) of nitrate (N-NO3), nitrite (N-NO2) and ammonium (N-NH4)
compared to the total inorganic nitrogen at the different stations in the sub-basins (Thao, Lo
and Da), in the main branch (Son Tay and Hanoi), and in the urban river system (To Lich and
Nhue).
Location
N-NO3, %
N-NO2, %
N-NH4, %
Thao (Yen Bai)
76
2
22
Lo (VuQuang)
84
1
15
Da (Hoa Binh)
79
1
20
Hong (Son Tay)
69
2
29
Hong (Hanoï)
25
1
74
To Lich river
2
0
98
Nhue river
9
1
90
Phosphate and total phosphorus
In freshwater, phosphorus is often the main factor that limits the production of plant biomass.
Phosphate can be dissolved or adsorbed to particles, remaining available by desorption
(Némery, 2003; Némery et al., 2005). Phosphorus was limiting in the range from 0.01 to 0.04
mgP-PO4 l-1, which corresponded to the value of the half-saturation constant for phosphate
uptake by algae (Garnier et al., 1995; Garnier et al., 1998; Garnier et al., 2005).
In the upstream tributaries, average concentrations of 0.03, 0.03 and 0.02 mgP-PO4 l-1 were
found in the Thao, Lo and Da Rivers respectively. Such low levels might be limiting for algal
growth at least at certain periods, depending on the seasonal variations (Figure 4.7).
However, total phosphorus concentrations were much higher, 0.29, 0.18 and 0.16 mg P l-1 in
the Thao, Lo and Da respectively (Figure 4.7), a significant proportion being probably
exchangeable. Taking into account the Redfield ratio (Redfield et al., 1963), and the amount
of TOC (see below), at least 20 to 40 % of the total phosphorus can be estimated to be under
mineral form. Differences in the total phosphorus levels between the sub-basins reflect mainly
their difference in suspended matter concentration, although the higher concentrations in the
Thao River are also due to its higher population density.
In the main branch, phosphates and total phosphorus, increased from 0.03 and 0.23 mg P l-1
to 0.11 and 0.27 mg P l-1 respectively, from upstream (Son Tay station) to downstream (Hanoi
station), due to the increase of population density along the river bank in the delta area
98
(Figure 4.7). The total phosphorus concentration in the downstream sector of the Red River is
PO4, mg P l-1
1.0
Thao R.
Lo R.
Da R.
0.5
Tot P, mgP l-1
very close to the one observed in the Amazon (0.24 mgP l-1) (Meybeck and Ragu, 1996).
0.0
2.0
1.0
0.5
0.0
J 1.
F 2.
M 3.
A 4.M 5.J 6.J 7.
A 8.S 9.
O 10N11D12J13F14
M15A16M17J 18J 19A20S21O22N23D24
0.
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Son Tay
Lien Mac
Hanoi
0.5
0.
J 1.
F 2.M 3.
A 4.M 5.J 6.J 7.A 8.S 9.
O 10N 11D12J13F14M15A16M17J 18J 19A20S21O22N23D24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Tot P, mgP l-1
PO4, mgP l
-1
1.0
0.0
2.0
1.0
0.5
0.0
O10N11D12J13F14M15A16M17J18J 19A20S21O22N23D24
0.J 1.F 2.M3.A4.M5.J 6.J 7.A 8.S9.
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2003
2004
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3.0
2.0
1.0
0.0
Nhue R.
To Lich R.
5.0
Nhue R.
To Lich R.
J 1.
F 2.
M 3.
A 4.M 5.J 6.
J 7.
A 8.S 9.
O 10
N 11D12J13F14
M15A16M17
J 18
J 19
A 20S21O22N23D24
0.
0 0 0 2002
0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .02003
.0 .0 .0 .0 .0 .0 .0 .0
Tot P, mgP l-1
PO4, mgP l-1
4.0
Son Tay
Lien Mac
Hanoi
1.5
0.J 1.F 2.M3.A4.M5.J 6.J 7.A 8.S9.O10N11D12J13F14
M15A16M17J 18
J 19
A 20S 21O22
N 23D24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
2003
2004
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
5.0
Thao R.
Lo R.
Da R.
1.5
4.0
3.0
2.0
1.0
0.0
0.J 1.F 2.M3.A4.M5.J 6.J 7.A 8.S9.O10N11D12J13F14M15A16M17J 18J 19
A20S21O22N23D24
0 0 0 0 0 2002
0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 .0 .02003
.0 .0 .0 .0 .0 .0
Figure 4.7: Seasonal variations, during the yeas 2003 and 2004, of dissolved phosphates
(PO4, mgP l-1) and total phosphorus (Tot P, mgP l-1) in the three main tributaries of the Red
River (Thao, Da, Lo) and in the main branch at the stations Son Tay, Lien Mac and Hanoi.
The urban rivers Nhue and To Lich are shown for comparison for the years 2002 and 2003.
According to Nguyen Viet Pho (1984), in the estuarine water of the Red River, the mean total
phosphorus concentrations varied from 0.21 to 0.56 mgP l-1 between the end of the dry season
and the flood period. This statement would imply that phosphorus is more issued from
diffuse than from point sources. Our results do not show any clear seasonal variation.
In urban rivers, phosphorus levels drastically increased, largely fluctuating up to 3 mg P l-1 for
phosphates (1.8 mg P l-1 in average) and 5 mg P l-1 (2.8 mg P l-1 in average) for total
phosphorus in the To Lich, sensibly diluted in the Nhue (in average, phosphates equal 0.5 mg
P l-1 and total P, 0.7 mg P l-1). The fraction of dissolved phosphate in the total phosphorus
concentration, represents less than 15 % in the upstream rivers where it comes mainly from
99
diffuse sources, but reaches 40 % in the Red River (at Hanoi), and up to 70 % in the urban
rivers where domestic pollution becomes a major source.
Phosphorus content of suspended solid (Tot P – P-PO4/ SS, in mgP gSS-1) in the 3 major
tributaries varied from 0.43 mgP gSS-1 in the Thao River, the most turbid of the tributaries, to
0.85 in the Lo river (Table 4.2). In the main branch, the values were slightly higher, from 0.7
to 1.2 mgP gSS-1, and increased to 18 mgP gSS-1 in the To Lich River (Table 4.2). A similar
upstream downstream gradient has been observed in the drainage network of the Seine River
by Némery (2003), with phosphorus content of suspended matter ranging from 1 mgP gSS-1 in
small streams draining agricultural soils, to values as high as 6 mgP gSS-1 in the Seine
downstream from Paris agglomeration.
Table 4.2: Phosphorus content (mgP gSS-1) of suspended solids (SS, mg L-1) at the different
stations in the sub-basins (Thao, Lo and Da), in the main branch (Son Tay and Hanoi), and in
the urban river system (To Lich and Nhue). Average values for the two study-years.
TP, mgP gSS-1
SS, mg L-1
Thao (Yen Bai)
0.43
1550
Lo (VuQuang)
0.85
460
Da (Hoa Binh)
0.65
110
Hong (Son Tay)
0.7
640
Hong (Hanoï)
1.2
600
To Lich river
18.0
70
Nhue river
11.0
50
Location
Dissolved silica and algal pigments
Dissolved silica concentrations in rivers mainly originate from rock weathering, and therefore
depend on the lithology (Meybeck, 1986). The lithological composition of the Red River
watershed is dominated by sedimentary rocks, with about half of carbonated rocks (Dürr,
2003; H. Durr and M. Meybeck, pers. comm., based on data from the UNESCO World
Geological Map). Meybeck (1986) assigned a silica concentration between 2 and 5 mgSi.L-1
to these lithological types. In addition, it was shown that, for a given rock composition, the
silica concentration in drainage water is much higher under warm and wet climate than under
colder climatic conditions (Meybeck, 1986; Garnier et al., in press). Dissolved silica
100
concentrations averaged 4 mgSi l-1 in the Lo and the Da Rivers (Figure 4.8); values were
notably higher in the Thao River (5.4 mgSi l-1), probably explaining by its lithology
characterized by a greater proportion of basic volcanic rocks and silico-clastic sedimentary
consolidated rocks. In the main branch, the DSi concentrations are typically intermediate (4.5
mgSi l-1).
5
5
15
20
0 J 1F 2M3A4 M5 J 6J 7A8 S9O10N11D
12 J13 F14M15A
16M
17J18J19A20S21O
22N23D2
100
Son Tay
Lien Mac
Hanoi
80
60
40
20
0
0.J1.F2.M3.A4.M5.J6.J7.A8.S9.O10N11D12J13F14M15A16M17J18J19A20S21O22N23D24
00 00 00 00 002003
00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 2004
.0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Nhue R.
To Lich R.
10
5
M1516
A M17
J 18
J 19
A 2021
S O22
N23D2
0J 1F 2M 3A 4M5J 6J 7A 8S 9O10N11D12J13F14
2003
2004
Nhue R.
To Lich R
100
T Chla , µg l-1
DSi, mg l-1
10
0
DSi, mg l-1
Son Tay
Lien Mac
Hanoi
40
0
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A8.S9.O10N11D12J13F14M15A
16M
17J18J19A20S
21O
22N
23 D
24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
15
60
-1
0
80
T Chla, µg l
DSi, mg l-1
10
Thao R
Lo R.
Da R.
100
T Chla , µg l-1
Thao R.
Lo R.
Da R.
15
80
60
40
20
0
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A 8.S9.O10N11D12J13F14M15A16M17J18J 19A20S21O22N23D24
0 0 0 02002
0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .0 2003
.0 .0 .0 .0 .0 .0 .0 .0
0
O 10
N11D12J13F14
M15A16M17
J 18
J 19
A 20S21O22
N23D2
0J 1F 2M 3A 4M 5J 6J 7A 8S 9
2002
2003
Figure 4.8: Seasonal variations, during the years 2003 and 2004, of dissolved silica (DSi, mg
Si L-1) and Chlorophyll a + Pheopigments (T Chl a, µg L-1) in the three main tributaries of the
Red River (Thao, Da, Lo) and in the main branch at the stations Son Tay, Lien Mac and
Hanoi. The urban rivers Nhue and To Lich are shown for comparison for the years 2002 and
2003.
Higher dissolved silica concentrations were found in the urban rivers (7.9 and 5.9 mgSi l-1 in
the To Lich and Nhue, respectively). This would tend to show that effluent of the Hanoi city
could be a source for silica, the silica being then diluted in the Nhue (Figure 4.8). The origin
of these high silica concentrations in wastewater is not clear. It was already noted that
domestic wastewater in Europe contains significant dissolved silica concentrations related to
the use of sodium metasilicates as a corrosion inhibitor in modern washing powders (Billen et
101
al, 2001; Garnier et al., 2002). On the other hand, industrial effluents may also be a significant
source: analyzing the effluents of a number of industries we found particularly high
concentrations (more than 30 mgSi l-1) in some of them.
Except in the urban system where the large fluctuations of concentrations were observed, the
silica concentrations in the Red River tributaries and main branch, were rather stable (Figure
4.8), showing that biological consumption of silica is low. Diatoms use silica to elaborate
their frustules and in eutrophicated rivers where nitrogen and phosphorus are not limiting,
silica can be seriously depleted during algal blooms (Garnier et al., 1995; Garnier et al., 1998;
Billen et al., 2005). When the Redfield ratios in the water (Si:N, Si:P: Redfield et al., 1963)
are too low compared to the algal requirement, silica becomes limiting for the diatoms, which
are then replaced at the coastal zones mostly, by other non-siliceous algae, sometimes
producing toxins, a phenomenon known as harmful algal bloom (HAB). Such situations are
currently encountered in North Western Europe, in the Manche Channel and North Sea
(Lancelot, 1995; Cugier et al., 2005), Black Sea (Humborg et al., 1997) or in the Gulf of
Mississippi (Rabalais and Turner, 2001). There is, of course, no Vietnamese standard level
for silica concentrations, but to avoid silica depletion and harmful non-diatom blooms, it is
necessary to control the N and P inputs to the rivers.
Levels of phytoplankton biomass, as expressed by the sum of chlorophyll a and pheopigments
(T Chl a, µg L-1), were relatively low. The values were the highest in the Thao river, despite
its higher suspended solid concentration compared to the other two tributaries (Figure 4.8,
Table 4.3). As the Thao was also the richest in phosphorus, this would suggest that algal
growth in the Red River tributaries is more limited by phosphorus than by available light. A
further increase in phytoplankton biomass occurs in the main branch, where nutrients
concentrations increase too. In the To Lich, phytoplankton biomass was rather high, but this
biomass can originate from the several fish ponds, in communication along its course.
Phytoplankton production can occur despite suspended matter concentrations as high as 60
mg l-1, in the absence of nutrient limitation (cf. Garnier et al., 2001). In the Nhue river, the
variations of phytoplankton concentrations is closely parallel to those of the To Lich river
(Figure 4.8).
4.4. Organic matter
Organic carbon is found under dissolved or particulate form and is either autochthonous
(produced in situ by algal biomass production and subsequently released by lysis or
102
excretion) or allochtonous (brought to the river from soil leaching, or domestic and industrial
effluents).
Leaching of the organic layers of soils is the primary source of dissolved organic matter
(DOC) in rivers. The level of DOC resulting from this process is strongly influenced by the
regional vegetation, climate and hydrology (Sempéré et al., 2002; Lilienfein et al., 2001). In
Nordic countries, high DOC values (up to 15 mgC.L-1) were found in rivers draining forested
and peatland area, with considerable seasonal variations linked to variations of temperature
and hydrology (Bishop and Pettersson, 1996). Lobbes et al. (2000) estimated that TOC
concentrations (total = dissolved + particulate) of 12 Russian rivers which enter into the Artic
Ocean ranged from 2.8 to 12.1 mgC.L-1. Meybeck and Ragu (1996) reported the mean value
of DOC and the total organic carbon TOC of rivers in the Amazon zone was about 4.0 and 6.6
mgC.L-1 respectively. Meybeck (1988) concluded that dissolved organic carbon concentration
for the wet tropics were higher than those in dry tropical regions and also higher than those in
temperate zones and proposed a mean value of 8 mgC L-1 for dissolved organic carbon in the
wet tropical regions. In the head waters of temperate climates, DOC originating from soil
leaching is mostly refractory (Servais et al., 1998).
When brought by domestic effluent, a large part is biodegradable (> 50 %, Servais et al.,
1995), these inputs possibly leading to oxygen depletion (or even anoxia), due to the
respiration of heterotrophic bacteria (Servais and Garnier, 1993; Garnier et al., 2001; Garnier
et al., 2004). Similarly, when autochthonous primary production is high, due to ample nutrient
concentrations, the organic biomass of the organisms can represent a large stock of
biodegradable organic matter, the heterotrophic degradation of which can also lead to oxygen
depletion (Garnier et al., 1999; Garnier et al., 2001; Garnier et al., 2004). These two types of
organic pollution, can lead to reduce the oxygen level down to values inappropriate for
aquatic life, fish in particular.
The mean DOC values at Hoa Binh, Vu Quang and Yen Bai, Son Tay and Hanoi during the
years 2003 and 2004 were 2.5, 2.6, 2.6, 2.8, and 3.6 mgC L-1 respectively. These values might
seem low compared with the above figures proposed by Meybeck (1988), but probably
reflects the absence of alluvial forests in the Red River basin.
The results of POC occasional analyses are shown in table 4.3. These values are lower than
those proposed by Ittekkot and Laane (2002) for the different ranges of river suspended solid
concentration but the DOC/POC ratios giving a range from 0.9 to 12.8 (highest in the Da
River and lowest in the To Lich and Thao River) are very close to the data reported by the
same authors.
103
Phytoplankton biomass figures, which represent a biodegradable fraction of the total organic
carbon have been converted in carbon unit using a C / T Chl a ratio of 24 (Servais and
Garnier, submitted) (Table 4.3). The results indicate that phytoplankton biomass represent
only a small fraction of the particulate organic carbon, and, a fortiori, a small fraction, from 2
to 6 % of total organic carbon (Table 4.3).
DOC, mgC l-1
15
Thao R.
Lo R.
Da R.
10
5
0
0.J 1.F 2.M3.A 4.M5.J 6.J 7.A8.S9.O10N11D
12 J
13 F
14M
15 A
16M
17J18J19A20 S
21 O
22 N
23 D
24
00 00 00 00 00 00 00 00 00 00 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DOC, mgC l-1
15
10
5
0
15
DOC, mgC l-1
Son Tay
Lien Mac
Hanoi
0.J1.F2.M3.A4.M5.J6.J7.A8.S9.O10N11D
12J13F14M15A
16M17J18J19A20S21O22N23D24
00 00 00 00 00 00
00
00
00
00
.0
.0
.0
.0
.0
.0
.0 .0 .02004
.0 .0 .0 .0 .0 .0
2003
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Nhue R.
To Lich R.
10
5
0
J 1.F 2.M 3.A 4.M5.J 6.J 7.A 8.S 9.
O 10N11D12J13F14M15A16M17J 18J 19
A 20S21O22N23D24
0.
0 0 0 2002
0 0 0 0 0 0 0 .0 .0 .0 .0 .0 .0 .02003
.0 .0 .0 .0 .0 .0 .0 .0
Figure 4.9: Seasonal variations, during the years 2003 and 2004, of dissolved organic carbon
(DOC, mgC L-1) in the three main tributaries of the Red River (Thao, Da, Lo) and in the main
branch at the stations Son Tay, Lien Mac and Hanoi. The urban rivers Nhue and To Lich are
shown for comparison for the years 2002 and 2003.
104
Table 4.3: Chlorophyll a + Pheo-pigments (T Chl a, µg L-1), Phytoplankton biomass (Phy,
mgC L-1) and particulate and dissolved organic carbon (POC, DOC, mgC L-1) at the different
stations in the sub-basins (Thao, Lo and Da), in the main branch (Son Tay and Hanoi), and in
the urban river system (To Lich and Nhue). Average values for the two study-years.
Location
Tchla, µg.l-1 Phy (*), mgC l-1 POC, mgC l-1 DOC, mgC l-1 DOC/POC
Thao (Yen Bai)
11.0
0.3
2.3
2.6
1.1
Lo (VuQuang)
3.2
0.1
0.9
2.6
2.9
Da (Hoa Binh)
3.4
0.1
0.2
2.5
12.5(**)
Hong (Son Tay)
5.6
0.1
Hong (Hanoï)
6.9
0.2
1.6
3.6
2.0
To Lich river
37.2
0.9
12.0
10.3
0.9
Nhue river
19.1
0.5
2.8
5.4
1.9
(*)
2.8
: Phy, represent the fraction of carbon content due to algal biomass, taking into account a ratio of C: T Chla equalling 24
(Servais and Garnier, submitted).
(**) Although the low SS values that can explain the high DOC/POC ratio (Ittekko and Laane, 2002), this ratio was
calculated with a low number of POC data.
4.5. Conclusions: water quality in the Red River
4.5.1. General levels of nutrients in the Red River drainage network
Our two-year surveys of water quality in the Red River tributaries allow for the first time to
assess the general level of nutrient concentrations in this sub-tropical river system (Table 4.4
in the part 4.5.3). As discussed above, these levels are low compared to river systems in the
temperate region of the world with similar population densities. Note that the values found in
head water (see Table 4.4) are close to the ones found at the outlet of the sub-basins, showing
that the low anthropogenic impact within the upstream basin of the Red River. Only in small
urban river systems of the delta region, extreme signs of domestic pollution are present.
Another general conclusion drawn from our survey is the absence of important seasonal
variations of nutrient concentrations, excepted those directly related to suspended solid.
105
Table 4.4: Water quality for the three main tributaries of the Red River (Thao, Da, Lo) and in
the main branch at the Hanoi station. The urban rivers Nhue and To Lich are shown for
comparison for the years 2002 and 2003. Values from head water are also given for 5 stations
sampled in October 2004 in the surroundings of Lao Cai, close to the Chinese border. DO:
minimum value of oxygen concentration observed; NH4: ammonium; Tot N: sum of inorganic
nitrogen; Tot P: total phosphorus; DSi: dissolved silica; T Chla: sum of chlorophyll a and
pheopigments.
Locations
DO
mg L-1
Head waters
NH4
Tot N,
Tot P,
mg.N L-1 mgN.L-1 mgP.L-1
DSi
T Chl a,
mgSi.L-1
μg.L-1
classification
0.00
0.4
0.12
4.7
1.5
Oligotrophic
Da
4.9
0.03
0.2
0.16
4.3
3.4
OligotrophicMesotrophic
Lo
5.4
0.06
0.6
0.18
4.2
3.2
Mesotrophic
Thao
5.4
0.10
0.6
0.29
5.4
11.0
Mesotrophic
Red-HongHanoi
5.8
0.85
1.2
0.27
4.5
6.9
Mesotrophic
To Lich
0.9
9.5
9.7
2.80
7.9
37.2
Eutrophic
Nhue
2.9
2.7
3.0
0.70
5.9
19.1
Organically
polluted
4.5.2. Behaviour of nutrients with increasing specific discharges in the Red River System
In order to analyse the general trends of variation of nutrients with respect to discharge for all
tributaries, we plotted the measured concentration against specific discharge.
Nitrate shows an increase with specific discharge, supporting its predominantly diffuse origin,
from soil leaching (Figure 4.10). Similarly, total phosphorus and suspended solid
concentrations originate from erosion of soil material to which adsorbed phosphorus is
associated. This trend also points out the higher concentrations of these elements in the
superficial water rather than in ground waters.
Regarding ammonium and ortho-phosphates, their dilution with increasing specific discharge
reveals their point source origin, the dilution being particularly evidenced for the downstream
Hanoi station (Figure 4.10). The low phosphate values at high discharge, also results from an
efficient adsorption of ortho-phosphates on the high concentrations of suspended solids.
Silica concentrations, although showing a large dispersion of values at low specific discharge
values, are rather stable within the range of specific discharges observed during the study
(Figure 4.10).
106
N-NO3, mg l-1
2.0
1.5
1.0
0.5
0.0
1.5
1.0
0.5
100
0.5
0
0.0
Thao R.
Lo R.
Da R.
Hanoi
5000
4000
3000
2000
1000
0
100
Thao R.
Lo R.
Da R.
Hanoi
0.5
0.0
25
50
75
100
Spec. disch., L. km-2.s-1
0
25
50
75
100
Thao R.
Lo R.
Da R.
Hanoi
15
DSi, mg Si l -1
0
25
50
75
1.0
P-PO4, mg l-1
25
50
75
Thao R.
Lo R.
Da R.
Hanoi
1.0
Tot P, mgP l-1
2.0
0.0
0
SS, mg l-1
Thao R.
Lo R.
Da R.
Hanoi
2.5
N-NH4, mg l-1
Thao R.
Lo R.
Da R.
Hanoi
2.5
10
5
0
0
25
50
75
100
-2 -1
Spec. disch., L. km .s
0
25
50
75
100
Figure 4.10: Relationship between the concentrations of nitrate (NO3), ammonium (NH4),
total phosphorus (Tot P), phosphates (PO4), suspended solids (SS) and dissolved silica (DSi)
and the specific discharge (Spec. Disch), in the three main tributaries of the Red River (Thao,
Da, Lo) and in the main branch at the station Hanoi for the years 2003 and 2004.
4.5.3. Classification of pollution level
Dodds et al. (1998) and Dodds and Welch (2000) proposed a general typology of rivers
according to their level of nutrient pollution (Table 4.5). On the other hand, Tran Hieu Nhue
et al. (1994) proposed a classification of nutrient pollution level specially adapted for tropical
climatic region like Vietnam (Table 4.6).
107
Table 4.5: Classification of trophic levels of rivers according to Dodds et al., 1998; Dodds
and Welch, 2000.
Trophic level
Total N
Total P
Suspended Chl a,
Benthic Chl a,
mgN.L-1
mgP.L-1
µg.L-1
mg.m-2
Eutrophic
> 1.5
> 0.075
> 30
> 60
Mesotrophic
0.7 – 1.5
0.025 - 0.075
10 - 30
20 - 70
Oligotrophic
< 0.7
< 0.025
< 10
< 20
Table 4.6: Classification of pollution level on the basis of diverse variables of water quality
(DO: dissolved oxygen; BOD5: Biological oxygen demand -5 days-, Tran Hieu Nhue
and al., 1994).
Pollution
level
Eutrophic
DO
BOD5
Organic
degradation
0 ÷1
> 40
Anaerobic
Water
statement
Rich in nutrients
1÷3
20÷40
Mesotrophic
Aerobic
degradation
[NH4+] > 10mg.l-1;
Microbial
contents
Strong
trace of CH4 and H2S development of
degradation
α-
Nutrient contents
Rich in nutrients,
occurrence of
algal blooms
in sediment layer
microbes
[NH4+]: 8÷10mg.l-1
Hundreds to
occurrence of
NO2-
thousands
microbes per
liter
β-
3÷5
10÷20
Mesotrophic
Aerobic
degradation
Rich in nutrients,
frequent
occurrence of
Nitrate and nitrite
content: several
-1
mg.l
algal blooms
Oligotrophic
>5
< 10
Stable levels of
organic matter
no algal blooms
Several
thousands
microbes per
liter
Nitrate and nitrite
occurrence of
content low and
macrophyte and
stable
pink agar
On the basis of these two references, we tried to classify the nutrient pollution level of the
different sectors of the Red River drainage network as indicated in Table 4.4. The upstream of
the Red River may be classified as Oligotrophic- Mesotrophic (β), the main branch in the
delta area is Mesotrophic (β), while the urban rivers are clearly organically polluted (Table
4.4).
As a whole, the water of the Red River is oligotrophic before entering the urbanized region of
delta, as expected by the origin of the nutrients, essentially of diffuse type, with limited
anthropogenic impact.
108
Among the major types of degradation of surface water that have occurred in the recent times,
the Red River does not seem to be touched in its sub-basins; eutrophication seems to be
limited by nutrients more than by light, at least during the dry season, from September to June
whereas siltation, particularly from agriculture or deforestation, would have not changed
much, and/or counterbalanced by the role of the two reservoirs, on the Lo and the Da rivers
(cf. Chapter 3). However, i) future nutrient enrichment due to increasing population, in urban
areas mainly, and ii) the future impoundment of two additional reservoirs as planned at the
horizon 2010-2015 which will further reduce the suspended solid concentrations, could
together quickly lead to major disruptions in term of river eutrophication.
In the delta, aquatic ecosystems are seriously damaged in a number of classical ways (Wetzel,
2001). We have clearly observed the most common type of degradation through the
contamination by inorganic (NH4) and organic (DOC) pollutants. Other types of degradation
come from irrigation, channelization that modify the aquatic habitats, toxic material, etc.
Presently, the To Lich river has reached a domestic and industrial pollution level close to the
one mentioned at the end of the XIX century in Western Europe, e.g. for the Bièvre urban
tributary of the Seine in Paris intra muros (Billen et al., 1999) or to that of the Senne crossing
Brussels (Garnier et al., 1992). Note that to face such pollution, these two rivers were
covered! It is interesting to note that, after more than 50 years of wastewater treatment effort,
the re-opening of these rivers is presently under debate. The treatment of urban wastewater
appears therefore a priority for Hanoi.
4.6. References
Billen G. and Garnier J., 1997. The Phison River Plume: coastal eutrophication in response to
changes in land use and water management in the watershed. Aqu. Microb. Ecol., 13: 317.
Billen G., Garnier J. and Meybeck M., 1998. Les sels nutritifs: l’ouverture des cycles.
Chapitre 12. In: La Seine en son bassin Fonctionnement écologique d’un système fluvial
anthropisé (Meybeck M., De Marsily G. and Fustec E. eds). Elsevier, Paris. p. 531-565.
Billen G. and Garnier J., 1999. Nitrogen transformation through the Seine drainage network: a
budget based on the application of the 'Riverstrahler' model, Hydrobiologia 410: 139-150.
Billen G., Garnier J., Deligne C., and Billen C., 1999. Estimates of early industrial inputs of
nutrients to river systems: implication for coastal eutrophication. The Sciences of the Total
Environment, 243/244: 43-52.
109
Billen, G., Garnier J., Ficht A. and Cun C., 2001. Modelling the response of water quality in
the Seine estuary to human activity in its watershed. Estuaries, 24(6): 977-993.
network of the Scheldt basin over the last 50 years. Hydrobiologia. 540: 47-67
Bishop, K. H and Pettersson C., 1996. Organic carbon in the boreal spring flood from
adjacent sub-catchments. Environment International, Vol 22 (5): 535-540.
Brion N., Billen G.., Guezennec L. and Ficht A., 2000. Distribution of nitrifying activity in
the Seine River (France) and its estuary. Estuaries, 23: 669-682.
Chesterikoff A., Garban B., Billen G. and Poulin M., 1992. Inorganic nitrogen dynamics in
the River Seine downstream from Paris (France), Biogeochemical mass-balances (C, N, P,
Si) in three large reservoirs of the Seine Catchment (France), Biogeochemistry 17: 147164.
Conley D.J., Claire L. S. and Stoermer E. F., 1993. Modification of the biogeochemical cycle
of silica with eutrophication. Marine Ecology Progress Series, 101: 179-192.
Cugier J., Billen G., Guillaud J.F., Garnier J., and Menesguen A., 2005. Modelling the
nutrient loading. Journal of Hydrology. Volume 304 (1-4): 381-396.
Dodds W.K., Jones J.R., and Welch E.B., 1998. Suggested classification of stream trophic
state: Distributions of temperate stream types by chlorophyll, total nitrogen, and
phosphorus. Water Res. 32:1455-1462.
Dodds W.K. and Welch E.B., 2000. Establishing nutrient criteria in streams. J. N. Am.
Benthol. Soc. 19: 186- 196.
Dürr H.H. 2003. Towards a typology of global river systems : some concepts and examples at
medium resolution. Thèse de doctorat, Université Paris VI - Pierre et Marie Curie. 432pp.
FAO, (1990-1998), Faostat statistics database (Fertilizer data used for China, Vietnam, Laos),
Faostat
database
Results,
copyright
FAO
1990-1998, http://www.apps.fao.org/;
Garnier J. and Benest D., 1991. Influence of light attenuation on primary production in a
shallow sand-pit lake. Verh. Internat. Verein. Limnol. 24: 1536-1542.
Garnier J., Billen G. and Billen C., 1992. Modélisation des modifications du paysage
écologique des rivières. Application à la Senne. Rapport de synthèse, Action "Research in
Brussels, Secrétariat d'Etat à la Région. 69 pp.
110
Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and Chlorophyceae
in the drainage network of the river Seine: Observations and modeling, Limnol. Oceanogr.
40: 750-765.
Garnier J., Billen G., Hanset Ph., Testard P. and Coste M,. 1998. Développement algal et
eutrophisation. Chapitre 14. In: La Seine en son bassin Fonctionnement écologique d’un
système fluvial anthropisé (Meybeck M., De Marsily G. and Fustec E. eds). Elsevier,
Paris. p. 593-626.
Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget od the Mosel
drainage network with the concept of heterotrophic/autotrophic sequences: the
Riverstrahler approach. Hydrobiologia, 410: 151-166.
Garnier J., Servais P., Billen G., Akopian M. and Brion N., 2001. Lower Seine river and
estuary (France): carbon and oxygen budgets during low flow. Estuaries, 24: 964-976.
Garnier J., d’Ayguesvives A., Billen G., Conley D. and Sferratore A., 2002. Silica dynamics
in the hydrographic network of the Seine River. Oceanis, 28 : 487-508
Garnier J, Billen G. and Cugier Ph., 2004. Drainage basin use and nutrient supply by rivers to
the coastal zone. A modelling approach to the Seine River. p 60-87 In: Drainage
basin nutrient inputs and eutrophication: an integrated approach. P. Wassmann and
K. Olli (Eds). E-book in press, 309pp. available at: www.ut.ee/~olli/eutr/
Garnier J., Nemery J., Billen G. and Théry S., 2005. Nutrient dynamics and control of
eutrophication in the Marne River system: modelling the role of exchangeable
phosphorus. Journal of hydrology, Volume 304(1-4): 397-412.
Garnier J, Sferratore A., Meybeck M., Billen G. and Dürr H. (in press). Modelling silica
transfer processes in river catchments. Chapter in SCOPE book series.
Humborg C., Ittekot V., Cosiascu A., and Bodungen B. V., 1997. Effect of Danube River dam
on Black sea biogeochemistry and ecosystem structure. Nature, 386: 385-388.
Ittekkot V. and Laane R. W. P. M., 2002. Chapter 10: Fate of Riverine Particulate Organic
Matter, SCOPE 42 - Biogeochemistry of Major World Rivers, available at
http://www.icsu-scope.org/downloadpubs/scope42/chapter10.html
Lancelot C., 1995. The mucilage phenomenon in the continental coastal waters of the North
Sea. The Science of The Total Environment, 165: 83-102.
111
Le Thi Phuong Q., Garnier J., Billen G., Théry S. and Chau Van Minh (submitted).
Hydrological regime and suspended matter flux of the Red River System (Vietnam):
Observations and modelling. To be submitted to Journal of Hydrology.
Lilienfein J., Wolfgang W., Richard T., Lourival V., Do Carmo Lima S. and Wolfgang Z.,
2001. The effects of Pinus caribaea forests on the C, N, P and S status of Brazilian
savanna Oxisols, Forest Ecology and Management 147: 171- 182.
Lobbes J. M., Fiznar H.P. and Kattner G., 2000. Biogeochemical characteristics of dissolved
and particulate organic matter in Russian rivers entering Artic Ocean, Geochimica et
Cosmochimica Acta, Vol. 64: 2973-2983.
Meybeck M., 1986. Composition chimique des ruisseaux non pollués de France. Sci. Géol.,
Bull. (Strasbourg) 39: 3-77.
Meybeck M., 1988. How to establish and use world budgets of river material, in Physical and
chemical weathering in Geochemical Cycles, edited by A. Lerman and M. Meybeck,
p. 247-272, Kluwer, Norwell, Mass.
Meybeck M., Chapman D.V. and Helmer R., 1989. “Global environment monitoring: Global
freshwater quality, a first assessment”, Chapter 7: Suspended matter in rivers and
lakes , p. 93-104, published by WHO and UNEP.
Meybeck M. and Ragu A., 1996. River discharges to the oceans: An assessment of suspended
solids, major ions and nutrients. UNEP Env. Information and Assessment Report.
Nairobi, 245pp.
Némery J., 2003. Origine et devenir du phosphore dans le continuum aquatique de la Seine,
des petits basins à l'estuaire. Rôle du phosphore échangeable sur l'eutrophisation.
Thèse Univ. Paris 6, 258pp.
Némery J., Garnier J. and Morel C., 2005. Phosphorus budget in the Marne watershed
(France): urban vs. diffuse sources, dissolved vs. particulate forms. Biogeochemistry
(in press).
Nguyen Viet Pho., 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209pp.
Rabalais N.N. and Turner R.E. (eds), 2001. Coastal Hypoxia. Coastal and Estuarine Studies
Studies, 52, American Geophysical Union, Washington, D.C.
Redfield A.C., Ketchum B.H. and Richards F.A., 1963. The influence of organisms on the
composition of sea-water.In M. N. Hill (ed.), The Sea, John Wiley and Sons, New
York, p. 12-37.
112
Ryding S.O. and Thornton J.A., 1999. Chapter 5: Types of aquatic pollutants, impacts on
water quality and determination of critical levels. In Assessment and control of nonpoint source pollution of aquatic ecosystem”- Man and the biosphere series, Unesco
Paris and the Parthenon Publishing group, Vol. 23, p. 73-105.
Sempéré R., Panagiotopoulos C., Lafont R., Marroni B. and Van Wambeke F., 2002. Total
organic carbon dynamics in the Aegean Sea, Journal of Marine Systems, Vol 33-34:
355-364
Servais P. and Garnier J., 1993. Contribution of heterotrophic bacterial production to the
carbon budget of the River Seine (France). Microb. Ecol. 25:19-33.
Servais P., Barillier A. and Garnier J., 1995. Determination of the biodegradable fraction of
dissolved and particulate organic carbon in waters. Int. J. Limnol. 31: 75-80
Servais P., Billen G., Garnier J., Idlafkih Z., Mouchel J.M., Seidl M. and Meybeck M.., 1998.
Le carbone organique : origines et biodégradabilité. In: La Seine en son bassin :
Fonctionnement écologique d’un système fluvial anthropisé. (Meybeck M., De
Marsily G. and Fustec E. eds). Elsevier, Paris p. 483-529.
Servais P. and Garnier J., submitted. Organic carbon and bacterial heterotrophic activity in the
Seine estuary maximum turbidity zone France). Aquatic Sci.
Tran Hieu Nhue, Phan Ngoc Dang, Hua Chien Quang and Lam Duc Huynh. 1994. Polluted
rivers in Vietnam. Construction editor, Hanoi, 65pp.
Wetzel R.G., 2001. Limnology. Lake and river ecosystems. Elsevier Ed., Third Edition.
Academic Press. London. 1006pp.
113
114
Nutrient budgets (N, P)
CHAPTER 5
Nutrient budgets (N, P) for the Red River Basin
Abstract
In order to examine the degree of human-induced alteration of the nitrogen and phosphorus
cycles at the scale of a tropical watershed of regional dimension, the budget of these two
elements were estimated in the 4 main sub-basins (Da, Lo, Thao and Delta) of the Red River
system (156 448 km², Vietnam and China). The 4 sub-basins differ widely in population
density (from 101 inhab km-2 in the upstream basins to more than 1000 inhab km-2 in the
delta), land use and agricultural practices. In terms of agricultural production, on the one
hand, and consumption of food and feed on the other, the upstream sub-basins are autotrophic
systems, exporting agricultural goods, while the delta is a heterotrophic system, depending on
agricultural goods imports. The budget of the agricultural soils reveals great losses of
nitrogen, mostly attributable to denitrification in rice paddy fields and of phosphorus, mostly
caused by erosion. The budget of the drainage network shows high retention/elimination of
nitrogen (from 62 to 77 % in the upstream basins and 59 % in the delta), and of phosphorus,
with retention rates as high as 80 % in the Da and Lo sub-basins which have large reservoirs
in their downstream course (Hoa Binh on the Da and Thac Ba on the Lo). The total specific
delivery estimated at the outlet of the whole Red River System is 855 kg.km-².y-1 total N and
325 kg.km-².yr-1 total P. Nitrogen rather than phosphorus seems to be the potential limiting
factor of algal growth in the plume of the Red River in Tonkin Bay.
Key-words: Nitrogen cycle, phosphorus cycle, Red River, autotrophy/heterotrophy of
regional systems, nutrient retention
This chapter is published as an article in the Journal Global Biogeochemical Cycles under the reference:
Le Thi Phuong Quynh, Gilles Billen, Josette Garnier, Sylvain Théry, Cédric Fézard, Chau Van Minh (2005, in
press). Received 8 November 2004; revised 20 March 2005; accepted 11 Avril 2005
115
5.1 Introduction
Today, human perturbation of the global N and P biogeochemical cycles is a matter of great
concern [Galloway et al., 1995; Howarth et al., 1996; Smil, 1999; Galloway and Cowling,
2002]. At the global scale, anthropogenic nitrogen fixation, either deliberate through
cultivation of nitrogen fixing crops and production of industrial fertilizer, or unintentional
through high temperature combustion, presently equals the natural rates. The resulting
increased nitrate contamination enhances the global denitrification rate and N2O emissions,
which contribute to the green-house effect and the destruction of the stratospheric ozone
layer.
Similary, world-wide mining and processing of phosphorus minerals, mainly for
fertilizers production, reach a level of the same order of magnitude as natural weathering and
erosion processes [Weijin et al., 1999]. The riverine transfer of nitrogen and phosphorus to the
coastal waters has therefore increased considerably in many areas of the world, making
marine eutrophication a symptom of global change [Green et al., 2004].
To obtain a good understanding (and possibly control) of these global phenomena, they have
to be examined at a regional scale, where the diversity of climatic and socio-economical
constraints can be taken into account. The regional scale is that of the human perception of
the environment, at which management decisions are taken. Moreover, a description of the
cycling of nutrients within a given territory offers an insight into how humans have managed
their environment and, to some extent, how they live.
Numerous studies have been devoted to the calculation of the nitrogen or phosphorus budget
of regional systems in Europe and Northern America [Billen et al., 1985; Howarth et al.,
1996; Boyer et al.; 2002; Van Breemen et al., 2002]. There are few similar attempts in other
regions of the world, despite the early work by Robertson and Rosswall [1986] in the Niger
basin, and some recent studies of nitrogen budgets in Asian countries [Bashkin et al., 2002;
Xing et al., 2002].
Here the analysis concerns the nitrogen and phosphorus cycling in the terrestrial and aquatic
components of the Red River watershed, a tropical river system which has been profoundly
modified by human intervention for two millennia, as it was the cradle of an ancient
civilization. This region is now the place of an original economic development scheme where
rural population remains dominant [Pham Xuan Nam, 2001]: since 1986, the introduction of a
market-oriented socialist economy (“Doi Moi”) has resulted in the rapid growth of
agricultural and industrial production, but has avoided explosive urban growth and
uncontrolled rural exodus.
116
5.2 Description of the Red River Basin
5.2.1 Geomorphology
The Red River basin (Figure 5.1) is located in South East Asia (from 20°00 to 25°30 North;
from 100°00 to 107°10 East) and its watershed covers 156 448 km². It is bordered by the
Truong Giang and Chau Giang River basins in China to the North, the Langcang River
(Mekong) basin to the West, the Ma River basin (in Vietnam) to the South. The Red River
flows eastwards into the Tonkin Bay (South China Sea) [Nguyen Ngoc Sinh et al., 1995] and
rises in a mountainous region of South-eastern China, in theYunnan province, where its name
is Yuan River; it crosses into Vietnam near Lao Cai where it is named Cai, Thao or Hong
River. The main branch is about 1140 km long [Dang Anh Tuan, 2000], and passes through 8
Chinese and Vietnamese provinces before flowing into the China Sea through 4 defluent
branches named Day, Lach Gia, Ba Lat, and Tra Ly. The Thao River has two major
tributaries, the Da and Lo rivers, downstream of which the main branch is named Hong (Red)
River. The drainage density in the Red River basin is rather high, in the range of 0.5 to 1
km.km-2.
Figure 5.1: map of the Red River basin, its 3 upstreams sub-basins (Da, Lo and Thao) and its
delta area. Circles indicate the gauging stations.
For budget calculations in this study, we divided the total basin area into 4 sub-basins
corresponding to the drainage area of the three main branches (Thao, Da and Lo rivers) and
the delta (Figure 5.1). Regular measurement of the discharge and water quality were carried
117
out at the outlet of the three upper sub-basins. Because of the difficulty to monitoring the
numerous diverging outlets of the Hong River in its delta area, a gauging station at Hanoï was
included, covering only about 20% of the delta, from which the output fluxes of the whole
delta were extrapolated (see below). Four sub-basins are therefore considered, i.e. those of the
three main tributaries and a portion of the delta.
5.2.2 Administrative divisions
50.3% of the Red River basin is located in Vietnam, 48.8% in China and 0.9% in Laos. Some
of the data used for budget calculations, including land use, fertilizer application, agricultural
production, livestock and industrial activity were taken from recent (1997) official provincial
statistics, i.e from 21 provinces in Vietnam [MOSTE, 1997] and one province in China
[Chinadata, 1998]. In this case, the data by province were reaffected to the 4 sub-basins on
the basis of the percentage of province surface area located inside each sub-basin, as shown in
Figure 5.1.
Table 5.1. Distribution of the surface area of Vietnamese and Chinese provinces (in %) within
the sub-basins of the Red River system.
Sub-basins
Provinces
Bac Kan
Cao Bang
Ha Giang
Ha Nam
Ha Tay
Hai Duong
Hung Yen
Hoa Binh
Lao Cai
Lai Chau
Nam Dinh
Ninh Binh
Phu Tho
Son La
Thai Binh
Thai Nguyen
Thanh Hoa
TP. Ha Noi
Tuyen Quang
Vinh Phuc
Yen Bai
Yunnan (China)
118
Area, km²
4 796
6 387
7 831
823
2 148
1 661
895
4 612
8 050
17 133
1 669
1 387
3 465
14 210
1 509
3 769
11 106
921
5 801
1 371
6 808
394100
Lo
37.57
30.24
88.40
Da
Thao
Total Hong
delta
11.60
0.28
7.40
24.75
5.69
33.25
21.63
77.99
1.53
72.69
0.06
100
67.57
2.49
64.71
54.65
73.57
89.97
14.08
10.96
62.80
74.96
1.97
46.14
0.41
3.21
94.02
87.72
22.42
3.85
0.29
63.88
13.22
5.89
3.86
64.34
11.16
8.08
5.2.3 Meteorological and hydrological characteristics
The climate in the Red River basin is quite homogeneous across the 4 sub-basins and of subtropical character. The average annual temperature is 19°C and the average annual rainfall
was 1470 mm in the whole basin in 1997 [IMH of Vietnam, 1997-2003], [Chinadata, 1998].
The rainy season lasting from May to October, represents 85 to 90% of the total annual
rainfall, and the dry season from November to April only 10 to 15%.
The mean annual discharge of the main branch (at Son Tay station, just downstream of the
outlets of the three main tributaries) is 3577 m3.s-1 [IMH of Vietnam, 1997-2003]. In the last
100 years, the highest daily discharge, 37 800 m3.s-1 was observed in August 1971, and the
lowest, 368 m3.s-1 in May 1960. Figure 5.2 shows the seasonal variations of discharge at the
outlets of the 4 sub-basins in 2003. The Da and Lo basins have higher specific discharges
(respectively an annual mean of 34 and 25 L.s-1.km-2 in 2003), while the Thao river, with a
large part of its basin in the drier Chinese territory, has by far the lowest specific discharge
(9.6 L.s-1.km-2 ).
Yen Bai (Thao)
6000
Vu Quang (Lo)
discharge, m3 s-1
Hoa Binh (Da)
Ha noi (Hong)
4000
2000
0
J
F
M
A M
J
J
A
S
O
N
D
Figure 5.2: Discharges at the outlet of the 3 upstream sub-basins of the Red River system,
and at the Hanoï station in the delta area, in 2003
5.2.4 Land use and population
As shown in Table 5.2 [MOSTE, 1997], land use differs markedly between the 3 upstream
sub-basins and the delta of the Red River. Overall, forest occupies the largest part of the
upstream Red River sub-basins (54 %), while cultivated land represents 33% (12 % for the
rice culture, 20 % for industrial crops). The Lo sub-basin differs from the other two upstream
sub-basins by a greater acreage of industrial crops (58.1 %) than the Da (2.6 %) and Thao
119
(12.8 %) sub-basins. Forest dominates the Da sub-basin (74.4 %). Urban areas represent a
very small proportion (1 %) of the upstream Red River basin. In the delta however, cultivated
land (mainly rice fields) holds the largest share of the land use (63 %), far above forest (18%);
urbanized areas represents a much larger surface (6.8 %) than in the upstream basins.
Table 5.2: Land use in the upstream sub-basins and in the main branch of the Red River
(delta) in 1997 (in % area)
Sub-basin
Rice
Industr.
cult.
Dry
cereals
Grassland Fruits Forest Rocks Urban
areas
Da
12.5
2.6
0.4
3.6
0.0
74.4
6.2
0.3
Thao
18.7
12.8
0.7
7.2
0.9
54.2
4.1
1.4
Lo
8.1
58.1
0.4
3.9
0.1
22.4
6.4
0.6
Hong delta at Hanoî
66.3
7.6
0.7
2.2
0.6
14.9
1.0
6.7
Total Hong Delta
63.0
3.7
0.0
2.6
0.2
17.8
5.9
6.8
In the whole basin, the population was estimated at 30 million inhabitants in 1997, of which
34 % in China [Chinadata, 1998] and 65 % in Vietnam [MOSTE, 1997]. The proportion in
Laos is low, less than 1%. These values were obtained from 5235 villages and towns, all of
which were geo-referenced in the Red River basin with the help of a GIS (Arc Info). The
population density differs greatly among the sub-basins, from 101, 132 and 150 inhab.km-2 in
the Da, Lo and Thao sub-basins respectively, to 1173 inhab.km-2 in the delta area (Table 5.3).
Table 5.3: Population and population density (inhab.km-2) in the sub-basins (Lo, Thao and
Da) and in the basin of the main branch (Delta) of the Red River in 1997.
Sub-basins
Surface
km²
Population
million inhab.
Population density
Inhab.km-²
Da
51 285
5.19
101
Thao
61 169
9.17
150
Lo
34 559
4.56
132
Hong Delta at Hanoï
1578
2.47
1565
whole Hong Delta
9 435
11.08
1173
Total
156 448
30.00
192
5.3 The budget of the soil system
In the nutrient budget for the soil subsystem of each sub-basin, we take into account the
following inputs and outputs, considering agricultural and forested areas separately: i) input
by atmospheric deposition, atmospheric nitrogen fixation, fertilizer application and excretion
by domestic animals, ii) output through harvested crops and grazing by domestic animals
120
(Figure 5.3). The nutrient losses through leaching or erosion into surface- or groundwater
will be discussed below in connection with the hydrosystem budget (Figure 5.3).
agricultural goods
530
20
wood
exp.
N2fix
370
500
60
510
200
20
1
Export Imp.
N2fix
800
640 100
130 10
1790
180
cattle
farming
dom.
act.
390
170
-45
Forested
soils
370
40
660
30
agricult
soils
640
220
130
20
390
335
ind.
act.
830
250
290
130
6
10
100
40
soil
denit
90
river
export
740
70
denit
&
reton
416
345
Figure 5.3 (a): Da River sub-basin (51 285 km²)
agricultural goods
wood
exp.
N2fix
110
410
25
490
60
1040
410
30
1
Export
N2fix
Imp.
900
1450
280
2140
240
cattle
farming
dom.
act.
-25
Forested
soils
150
7
110
15
380
45
agricult
soils
1420
410
1610
415
210
30
ind.
act.
500
220
15
15
640
610
soil
denit
130
55
930
150
1160
river
export
denit
&
reton
5
537
Figure 5.3 (b): Lo River sub-basin (34 560 km²)
121
agricultural goods
310
20
wood
exp.
N2fix
280
500
70
510
200
280
20
Export
N2fix
Imp.
1140
650
130
1340
130
100
10
cattle
farming
dom.
act.
245
Forested
soils
270
30
230
30
70
10
agricult
soils
740
230
190
170
ind.
act.
570
255
970
315
420
195
20
20
150
60
soil
denit
370
river 140
export
1320
denit
&
reton
60
120
Figure 5.3(c): Thao River sub-basin (61 170km²)
agricultural goods
wood
exp.
N2fix
85
640
30
Exp.
atm.
depos.
500
60
fertilizers
N2fix
8270
3230
3290
3390
130
Import
5720
1200
8370
1210
cattle
farming
6360
2110
ind.
act.
domestic
activity
2240
Forested
soils
100
50
agricult
soils
8680
3280
90
10
410
50
1060
910
4870
1080
850
120
460
120?
6360
2110
soil
denit
7410
river
export
14190
5390
9880
1790
denit
&
retention
3670
-410 ??
Figure 5.3 (d): the Delta sub-basin (9 435 km²)
Figure 5.3 (a,b,c,d): Nitrogen and phosphorus budgets in the 4 sub-basins of the Red River
system, expressed per km² of catchment area (nitrogen, in bold: kgN.km-2.year-1; phosphorus,
in italics: kgP.km-2.year-1).
122
5.3.1 Atmospheric deposition
Due to the increase in nitrous oxide emissions by automobile engines and thermal power
production in industrialized countries, wet and dry nutrient atmospheric deposition has
become a significant term in the nitrogen budget of terrestrial systems [Galloway, 2002;
Sheldrick et al., 2003]. Van Drecht et al. [2003] mentioned a global mean N deposition rate of
450 kg.km-².y-1. A compilation of available data from non-industrial countries shows values
of atmospheric N deposition rates in the range of 100-800 kg.km-².y-1 (230 kg.km-².y-1 in a
rural area of south-eastern China [Weijin et al., 1999], 760 kg.km-².y-1 in the Middle Hills of
Nepal [Collins and Jenkins, 1996], 500 kg.km-².y-1 in a hillslope forest in Puerto-Rico
[Chestnut et al., 1999], 580 kg.km-².y-1 in Ecuador [Wilcke et al., 2001]), while global models
provide values of between 500 and 1000 kg.km-².y-1 in the Red River basin area [Lelieveld
and Dentener, 2000] and Pham Hung Viet et al. [1998] report the value of 2000 kg.km-².y-1
in a suburb of Haiphong. These figures often represents the sum of nitrate and ammonium
deposition rates, in which the proportion of both forms are generally similar. However, as
discussed by Howarth et al (1996), among other authors, only nitrate deposition should be
considered as a ‘new’ nitrogen input when examining large watershed nitrogen budget,
because most ammonium deposition depends on short distance transfer of locally volatilized
nitrogen, thus representing internal cycling within the watershed.
Phosphorus deposition, although much lower, is still significant. Values of between 7 and 156
kg.km-².y-1 are cited in the literature for regions with climate and economic conditions similar
to those of the Red River: 45 kg.km-².y-1 in agricultural areas in south-eastern China [Weijin et
al., 1999], 7-28 kg.km-².y-1 in a Ivory Coast rain forest [Stoorvogel et al., 1997], 60 kg.km-².y1
in Ecuador, 156 kg.km-².y-1 in a dry forest in Mexico [Campo et al., 2001].
Because of the lack of direct measurements in North Vietnam, the above data were used as a
basis for the chosen values of 500 kg.km-2.y-1 for nitrate-N deposition, and 60 kg.km-2.y-1 for
phosphorus deposition rates, considered as representative for the ‘new’ atmospheric
deposition of nutrients in the whole Red River basin (Table 5.4 a,b).
123
Table 5.4a: Nitrogen budgets of the sub-basins of the Red River (106 kg N.yr-1).
Da
Lo
Thao
Whole
Hong
delta
forest
19
3.9
16.6
0.8
40.4
agriculture+grass
6.6
13
14
3.9
37.5
forest
19
3.9
17
0.8
40.4
grass- and cropland
41
31
70
31
173
Fertiliser application
26
36
31
78
171
Human manure application
15
12.9
26.1
-
-
meat and dairy production
4.6
7.3
5.7
8.1
25.7
excretion
28
43
34
46
151
grazing and feed consumption
33
50
40
54
179
agricultural production
92
74
82
79
327
commercial import
1
1
17
32
-
commercial export
27
14
19
6
-
human consumption
20
17.3
34.8
60
132
Domestic wastewater release
5
4.3
8.7
60
77.9
Industrial wastewater release
0.3
0.5
1
1.9
3.8
Leaching from forest soil
17.5
5.3
4.3
0.7
27.7
Leaching from agricultural soil
20
22
12
10
64
Input from upstream tributaries
-
-
-
93.2
-
Riverine delivery at basin outlet
38
32
22.5
117
169
106 kg.y-1 as N
Total Red
R. basin
Soil system
Atmospheric deposition
Nitrogen fixation
Cattle farming
Agriculture and food balance
Hydrosystem
124
Table 5.4b. Phosphorus budget in the sub-basins of the Red River (106 kg P.yr-1)
Da
Lo
Thao
Total
Hong
delta
forest
2.0
0.5
2.0
0.1
4.6
agriculture+grass
1.0
1.5
2.0
0.5
5.0
Fertiliser application
10.3
14.1
12.2
30.5
67.1
Human manure application
6.6
5.8
11.7
-
-
meat and dairy production
0.6
0.9
0.7
1.1
3.3
excretion
6.0
8.8
7.3
9.9
32
grazing and feed consumption
6.6
9.8
8.0
11.3
35.7
vegetal production
9.3
8.2
8.2
11.4
36.0
commercial importation
0.03
0.04
1.4
1.2
-
commercial exportation
1.1
0.85
1.3
0.32
-
human consumption
8.8
7.7
15.6
19.9
52
Domestic wastewater release
2.2
1.9
3.9
19.9
27.9
Industrial wastewater release
0.65
0.51
1.3
1.1
3.6
Leaching from forest soil
1.4
0.2
0.4
0.5
2.5
Leaching and erosion from agr. soil
17.0
21.0
10.0
8.6
57.0
Input from upstream tributaries
-
-
-
16.9
-
Riverine delivery at basin outlet
3.5
5.1
8.3
51.0
51.0
106 kg.y-1 as P
Total
basin
Soil sub-systems
Atmospheric deposition
Cattle farming
Agriculture
Hydrosystem
5.3.2 Atmospheric nitrogen fixation
Atmospheric nitrogen fixation can represent high inputs of reactive nitrogen in tropical
systems. A compilation of specific fixation rates corresponding to the main land use classes in
the Red River basin cited in the literature [Chestnut et al, 1999; Smil 1999; Weijin et al., 1999;
Boyer et al., 2002, Vitousek et al., 2002, Xing et al., 2002; Basking et al, 2002] led to the
125
following values: 105 kg.ha-1.y-1 for nitrogen fixation by soybean and peanut crops, 50 kg.ha1
.y-1 for paddy rice, 5 kg.ha-1.y-1 for other cultures, 5 kgN.ha-1.y-1 for forest and 15 kg.ha-1.y-1
for grassland. On the basis of the distribution of these land use classes in each sub-basin
(Table 5.2), the total nitrogen fixation was calculated (Tables 5.4 and 5.5). Due to its large
share in all basins, rice cultivation always represents the major part of the total nitrogen
fixation. Note however that the value of 50 kg.ha-1.y-1, often cited for paddy rice fields, might
be over-estimated in the case of intensive chemical ferilization [Roger and Ladha, 1992].
Table 5.5: Nitrogen fixation (106 kg N.yr-1) for the largest land use classes in the sub-basins
of the Red River basin [MOSTE 1997; Chinadata, 1998] and total per sub-basins.
Sub-basins
Rice
Soybean
& peanut
Other
cultures
Grassland
Forest
106 kg.yr-1
Da
Lo
Thao
Hong-delta
Total Red River
32.1
14.0
57.2
29.7
133.0
5.0
4.6
7.6
0.3
17.5
Total
N fixation
per unit
watershed
surface area
kg.km-2.y-1
0.8
10.1
4.4
0.2
15.5
2.8
2.0
1.1
0.4
6.3
19.1
3.9
16.6
0.8
40.3
59.7
34.6
86.8
31.5
212.6
1165
1000
1420
3334
1359
As far as phosphorus is concerned, the process of ‘new’ phosphorus input to soils, i.e.
phosphorus mobilization from bed rock weathering, cannot be easily estimated, and is
neglected in the budgets.
5.3.3 Chemical fertilizers
The use of chemical fertilizers in agriculture has increased significantly in Vietnam and
China in the last 50 years. According to Weijin et al. [1999] China is presently the largest
producer of nitrogen fertilizers and the greatest consumer of mineral fertilizers in the world.
In Vietnam, according to the FAO database, the use of nitrogen fertilizers has increased 66
fold during the period from 1961 to 2000 (from 2.2 kg.ha-1.y-1 in 1961 to 150 kg.ha-1.y-1 in
2000). For phosphorus fertilizers, the increase was 5 fold during the same period. For the late
1990’s (e.g. 1997), the average application rate was 115 kg.ha-1.y-1 of N fertilizers and 45
kg.ha-1.y-1 of P fertilizers on cropland in Vietnam, on the basis of the FAO data [FAO 19901998]. The annual fertilizer inputs in the Red River sub-basins were calculated from these
rates and the agricultural surface area in each sub-basin (Table 5.4).
5.3.4 Feed consumption, food production and excretion by domestic animals
The excretion by domestic animals, either directly on grazed land or through spreading of
manure on cropland, must be considered as an input into the agricultural soil system, while
126
grazing and feed consumption constitute an output. These terms of the budget (Table 5.4)
were estimated on the basis of a livestock census in each sub-basin. Five livestock categories
were taken into account: pig, bovine, horse, sheep/goat and poultry, and the corresponding
data were taken from Vietnamese and Chinese statistics by province in 1997 (Table 5.6). Per
capita production rates of manure, as well as of meat and dairy products, compiled from the
literature (Table 5.7) were used to calculate the budget of animal farming in each sub-basin
(Table 5.4 and 5.8). The sum of excretion and food production was used to estimate total
feedstuff consumption by livestock. Note that pigs and bovines are responsible for more than
80% of the total fluxes in all sub-basins.
Table 5.6: Livestock census (in 103 capita) in the sub-basins of the Red River in 1997
[MOSTE 1997, Chinadata 1998]
Pigs
Bovines
Horses
Sheep
Poultry
x 103
x 103
x 103
x 103
x 103
Da
688
389
48
52
4148
Lo
1033
594
33
34
13095
Thao
980
436
35
85
9911
Hong-delta at Hanoi
433
91
2
1
3843
Total Red River
5130
1802
131
173
51367
Sub-basin
Table 5.7: Per capita excretion and animal food (meat, eggs and/or milk) production for the
main livestock categories in Vietnam.
Category
Excretion
nitrogen
Meat (and dairy) production
Phosphorus
kg capita-1 y-1 kg capita-1 y-1
Nitrogen
Phosphorus
kg capita-1 y-1
kg capita-1 y-1
Pig
7.7
2.25
1.5
0.18
Bovine
50
9.6
8.5
1
Horse
43
9.6
-
-
Goat and sheep
5.8
1.9
0.9
0.11
Poultry
0.3
0.04
0.05
0.01
(data compiled from different sources including Soltner 1979; SCS, 1992; Smil, 1999; ITP,
2000; Boyer et al., 2002; Van der Hoek, 1999; Bleken and Bakken, 1997; Thomas and
Gilliam, 1977; Weijin et al., 1999, Hedlund et al., 2003).
127
5.3.5 Nutrient export by crop harvesting and grazing
Nutrient outputs from the soil by harvested crops were determined from the figures of
agricultural production in each sub-basin combined with the N and P content of harvested
products (Table 5.8). The main crops considered are rice, wheat, maize, starchy roots (tubers
and potatoes), vegetables, soybean, peanut, fruit, sugarcane, tobacco, tea, coffee, rubber and
cotton. The production of forage was also considered, including that directly grazed by cattle.
Because data on grass production are not available, an overall yield value was used, i.e. 8000
kg.ha-1.y-1, proposed by Stevenson and Cole [1999], with a nutrient content of 2 % N and
0.25% P. Table 5.8 shows these estimates.
In order to estimate the fate of nutrient fluxes exported from agricultural or grassland soils
with the crops and grass (either consumed locally or exported from the sub-basins), the crop
production figures were compared (Table 5.8) with the food requirements of the local
population and the feed requirement of the cattle.
To estimate the human consumption, the average per capita diet of Vietnamese people
provided by the FAO (Table 5.9), was combined with the population figures of each subbasin. To take into account the interregional differences in living standards, the overall FAO
figure was corrected by a factor of 0.7 for the upland regions while the delta area was
considered representative of the national mean [Liu, 2001]. The FAO figures for fish and
seafood consumption were also corrected by region with data from MOSTE [1997].
The total animal feed requirements are estimated in Table 5.4. In order to meet these
requirements, grass production, as well as residues of cereals, starchy roots and sugar cane
were considered as fodder. When necessary, a part of the cereal production not included in the
local human nutrition was allotted to the livestock diet. Finally, the feed budget was balanced
by introducing an ‘other feed’ source represented by fodder that did not figure in the available
statistics, e.g. grazing on rangeland or in forests, aquatic plants used as fodder, etc… (Table
5.8). This additional term also accounts for inaccuracies or gaps in the statistical data. For
instance, official statistics do not correctly take into account the production of home gardens
and backyard plots, which can make significant nutritional contributions. However, the size of
this ‘other feed’ category, is reasonable, which demonstrates the reliability of the overall
budget; it was therefore not included in the calculation of nutrient export from the agricultural
soil.
Besides dividing the agricultural production between local human and animal consumption,
the study estimated commercial import and export to and from the sub-basins (Table 5.8).
Agricultural products other than foodstuff, e.g. cotton or rubber, were considered to be
entirely exported.
128
Table 5.8. Agricultural production and its destination (human and livestock consumption, or exportation) in the sub-basin of the Red River in
1997. (Figures in 106 kg harvested products yr-1 unless stated)
1
2820
51
6
export/import
1122
23
2
309
16
0
0
0
1087
200
0
0
animal
consumption.
2725
4228
54 19/-17 74
5
1/-1
8
592
0
0
0
0
592
500
0
0
Human
consumption
214
16
84
38
132
0
114
0
0
Production,
kt/yr
238
381
273
73
34
33
41
77
61
9
22
523
0
18
0
19
3
112
0
223
80
41
4
2
0
0
0
0
80
16
Production,
kt/yr
2256
47
3
-246
-75
592
8
124
2
18
0
761
381
291
73
53
36
153
77
284
37
659
20
62
1
24
1087
Whole delta
export/import
27/-1
1/0
4492
82
8
0
0
0
Production,
kt/yr
169
54
0
0
animal
consumption.
2099
34
6
-556
-556
0
44
36
1053
0
36
0
39
6
225
0
449
160
83
9
3
0
0
0
0
160
32
0
249
424
115
156
-37
767
7
91
2
34
0
497
249
460
115
153
90
263
132
203
85
1267
17
127
2
18
592
Human
consumption
1277
27
2
export/import
91
18
0
0
0
0
0
1489
-350
0
0
export/import
animal
consumption.
9
303
265
71
59
50
93
110
Lo sub-basin
animal
consumption.
4826
92
9
596
0
20
0
22
4
127
0
254
91
47
5
2
Thao sub-basin
Human
consumption
total kt/yr
total ktN/yr
total ktP/yr
%P1
0.22 605
0.22 303
0.35 285
0.35 71
0.48 81
0.46 54
0.12 220
0.22 110
0.06 410
0.09 54
0.08 814
0.23 12
0.15 93
0.43
2
0.43 34
0.26 1489
0.06
0.3 135
0.3
54
Human
consumption
%N1
1.1
rice
leaves
1.1
1.2
maïze
leaves
1.2
1.8
wheat
2.2
soja
2.4
starchy roots
leaves
2.4
3.7
vegetables
2.4
fruit
2.1
sugar cane
1.3
peanuts
tea,coffee,tobacco 2.9
2.2
cotton
2.9
rubber
2
grass'
other feed
2.9
3.4
animal pdcts
3.4
fish & sea food
Production,
kt/yr
Da sub-basin
1817
952
1385
116
45
0
0
0
-66
-8
-262
0
-43
309
0
60
1
24
0
2769
1385
178
45
0
3
67
34
150
505
255
41
13
0
12
198
134
0
239
54
277
188
111
0
0
0
0
198
650
0
0
14/-1
1/0
5947
79
11
3969
85
6
3490
54
7
0
0
0
0
0
62
66
11
329
34
776
277
144
16
6
0
0
-626
228
0
25
7
0
12
0
-38
-134
6/-32
0/-1
Data compiled from several sources including Weijin et al, 1999; Martin-Prével et al., 1984; Stevenson and Cole, 1999; Pilbeam et al., 2000, Morel, 1996; Beaton et al, 1995; Smil, 1999.
129
5.4 Domestic and industrial P, N loadings
When concentrated to urban areas, domestic and industrial activities represent a major source
of nitrogen and phosphorus transfer from the agricultural soil system to the hydrosystem, by
direct point discharge. An analysis was made of the data from which it is possible to estimate
the release of nutrients by domestic and industrial wastewater in the sub-basins of the Red
River. It is worth mentioning that presently, wastewater treatment is practically nonexistent in
the domestic and most of the industrial sectors in Vietnam.
5.4.1 Domestic wastewater in cities and villages
With the human per capita food consumption data (Table 5.9), and the N and P content
discussed above, one can calculate a yearly per capita nutrient loading in the range of 3.8 - 5.4
kg.cap-1.y-1 for nitrogen and 0.5 - 0.65 kg.cap-1.y-1 for phosphorus in Vietnam, the lower
values characterizing the poorest population in the upland areas, while the higher represent
the Vietnamese mean and delta population (see above). The P loading corresponding to the P
content of builders and sequestering agents accompanying the 4.8 kg of active detergents used
annually per capita in washing powders and personal care products in Vietnam [Vietnam
Parorama,
2004,
http://www.vietnampanorama.com;
Vietparners,
2004,
http://www.
vietpartners.com ] should be added to the figures for phosphorus. Considering a mean active
detergent content of 20% [Madsen et al., 2001], and a mean P content of 5 % in cleaning
products, the additional P loading from these products is about 1.2 kg.cap-1.y-1 as P, making a
total annual phosphorus loading of 1.7 –1.8 kg.cap-1.y-1. Although these values are very close
to European standards [see eg. Billen et al., 1999; Servais et al., 1999], and within the range
found by McKee et al. [2000] in a sub-tropical catchment in Australia (2.2 to 6.2 kg.capita-1.y1
for N, and 0.66 to 1.8 kg.cap-1.y-1 for P), they are high compared to those found in the
literature for Asian countries. Cao Van Sung [1995] estimated the specific per capita load of
the Vietnamese at 3.65 kg.y-1 for N and 0.62 kg.y-1 for P, in good agreement with the figures
proposed by Meybeck et al. [1989] , i.e. 3.3 kg.capita-1.y-1 for N and 0.4 kg.capita-1.y-1 for P.
Bashking et al. [2002] for Korea, and Weijin et al. [1999] and Sheldrik et al. [2003] used
much lower values, 0.18-0.7 kg.capita-1.y-1 for N and 0.09-0.25 kg.capita-1.y-1 for P,
respectively. Although this is not entirely clear from their paper, these authors probably took
into account the fact that only a part of the produced human wastes is discharged into surface
water, another part is spread on agricultural soils.
130
Table 5.9: a) Average human diet per capita per year (kg capita-1 yr-1) in Vietnam (1997)
(FAO), and N and P content (%N or %P); b) Average human diet per capita per year
expressed in nitrogen and phosphorus.
a)
Products
kg /capita/yr
%N1
%P1
Rice
164
1.1
0.22
Maize
5.6
1.2
0.35
Wheat
6
1.8
0.38
Starchy roots
35
0.9
0.12
Soybean
1
2.2
0.46
Vegetables
70
1.1
0.06
Fruits
25
2.4
0.09
Sugar cane
13
2.1
0.08
Peanut
1.4
1.3
0.23
Tea and coffee
0.5
2.9
0.15
Meat
24
3.4
0.3
Dairy products
1
2.1
0.35
Fish and seafood
17
3.4
0.3
Total
364
b)
kgN capita-1 yr-1
-1
kgP capita yr
-1
5.4
0.65
1
Data compiled from several sources including Weijin et al., 1999; Martin-Prével et al., 1984; Stevenson and
Cole, 1999; Pilbeam et al., 2000, Morel, 1996; Beaton et al, 1995; Smil, 1999, Boyer et al., 2002, Xing et al.,
2002, Vitousek et al., 2002, Bashkin et al., 2002
For the Red River delta region, where most of the population is agglomerated, and where
running water is available everywhere, we considered that all domestic waste is discharged
into the hydrosystem. However, in the upstream watersheds, where only 25% of the
population live in urban areas [Cao Van Sung, 1995], it was estimated that only 25% of the
domestic wastewater reaches surface waters and the rest is recycled in agriculture (Table 5.4).
5.4.2. Industrial activity
Several large industrial sites exist in the Red River basin, namely those of Viet Tri, Thai
Nguyen and Ha Bac (chemistry, textile and paper). Moreover, smaller cottage industries
(textile, food processing, …) are found everywhere in traditional villages and cause
131
significant pollution of surface waters. The contribution by these industrial activities to
nitrogen and phosphorus loading of the hydrosystem is extremely difficult to evaluate.
Vietnamese and Chinese economical statistics by province were initially used to estimate the
industrial production in each sub-basin (expressed in tons of finished products) in the
following branches, thought to be the most significant ones in terms of aquatic N and P
pollution: cement production, wood processing, the production of paper industry, chemicals,
food and drink and textiles (Table 5.10). A large amount of data were then gathered to
characterise the wastewater discharged by specific factories, in order to estimate a general
specific N and P loading value for each one of these industrial branches in the present
Vietnamese conditions. Sectorial studies carried out for, or by, the Vietnamese Ministry of
Science, Technology and Environment [MOSTE, 1999, 2003; Le Xuan Tu and Huynh Phu,
1998, VAST, 2000…] or by International Cooperation Agencies [Japan International
Cooperation Agency, 2000] were examined. This compilation was augmented by enquiries,
collection and analysis of effluents in samples from about 20 factories in the Hanoï district.
Overall, 20-30 factories were adequately investigated in each industrial branch. The median
value of the N and P release rate by ton of material produced by the different industrial
branches (Table 5.11) together with a production census (Table 5.10), led to a calculation of
the overall N and P discharge from industrial activities for each sub-basin (see Table 5.4).
These estimates indicate that industrial activities generate nutrient fluxes amounting to less
than 10 % of those from domestic activity. The textile and chemical industries (fertilizers and
detergents) dominates nutrient point sources to surface water. Because it is difficult to obtain
reliable data on pollution fluxes generated by industries and handicraft activities in villages,
the estimates of the contribution to nutrient water contamination by industries might be
severely underestimated.
Table 5.10: Industrial production (in 106 kg yr-1 final product) of the most polluting sectors
in the sub-basins in 1997 (sources: Vietnam General Statistic Office 1997, Chinadata 1998)
in 106 kg y-1
Cement
Paper industry
Wood industry
Chemical industry
Textile industry
Food industry
drinks
milled food
sugar
132
Da
1002.5
21.3
827.8
176.7
2.9
Lo
704.3
19.5
268.4
168.1
2.5
Thao
1843.2
75.2
2688.0
690.4
7.6
Hong delta
381.2
9.0
5.1
167.5
20.5
total basin
3931.1
125.0
3789.2
1202.7
33.6
16.4
20.3
97.3
11.2
87.6
66.9
33.2
84.6
186.2
69.8
1168.6
2.5
130.6
1361.2
353.0
Table 5.11: Specific N & P loading for the most polluting industrial activities in Vietnam
estimated from a sample of investigated factories in North Vietnam (see text for details).
Industrial sector
Concrete
Wood
Paper industry
Chemical industry
Textile industry
Food industry
drinks
milled food
sugar
Wastewater produced
m3. 10-3 kg of product
300
1
100
100
200
Specific N loading
kg. 10-3 kg product
0.002
1.000
0.700
30.000
Specific P loading
kg. 10-3 kg product
0.600
0.0003
0.200
0.050
4.000
10
50
15
0.450
3.000
0.300
0.064
0.700
0.045
(sources: MOSTE 1999, 2003, projects on environments of VAST 1997-2003, projects JICA
2000; this study: chemical analysis results and questionary)
5.5 The budget of the hydrographical network
During their downstream transfer through the aquatic continuum, from the headwaters to
large river branches and reservoirs, nutrients from watershed-based sources undergo several
biogeochemical processes with the result that a fraction of their load is immobilized or
eliminated before it reaches the outlet of the basin. A comparison of the estimates of the total
diffuse fluxes (from agricultural and forested soils) and point inputs (from domestic and
industrial activities) in the watershed with the calculation of the N and P fluxes discharged at
its outlet (our measurements) gives an insight into these “retention” processes.
5.5.1 Diffuse nutrient loss from forested soils to the hydrosystem
Nitrogen concentrations in the surface water draining tropical forests are fairly well
documented [Forti and Neal, 1992; MacDowell and Asbury, 1994; Stoorvogel et al., 1997;
Roldan and Ruiz, 2001; Colins and Jenkins, 1996], and range between 0.05 and 0.5 (median
0.4) mg.L-1 for nitrate-N and between 0.01 and 0.07 (median 0.03) mg.L-1 for ammonium-N,
while dissolved organic nitrogen often represents a large fraction (50-60%) of total nitrogen,
averaging a concentration of 0.4 mg.L-1. These figures are significantly higher than those
found for forested ecosystems at temperate latitudes [Howarth et al. 1996]. The total
phosphate-P content in headwaters of tropical forested watersheds amounts to around 0.0150.05 (median 0.04) mg.L-1; the strongest concentrations were associated with periods of high
133
runoff, and the weakest ones with low flow conditions [Forti and Neal, 1992; Stoorvogel et
al., 1997; Roldan and Ruiz, 2001; Colins and Jenkins, 1996].
These values were used to calculate the contribution to total N and P diffuse loading by
forested soils in the 4 sub-basins (see Table 5.4), taking into account the discharge measured
at the outlet in 2003, and the forested area of each sub-basin.
5.5.2 Diffuse nutrient loss from agricultural soils
Much fewer data are available for cultivated areas in tropical systems. Kao et al. [in press]
report nitrate-N concentrations of between 0.42 at low runoff and 3.5 mg.L-1 at high runoff in
streams draining vegetables cultures in mountain areas in Taiwan. Roldan and Ruiz [2001]
measured inorganic nutrient concentrations of 0.67 mg.L-1 for N and 0.55 mg.L-1 for P in
rivers draining industrial plantations in Columbia. Taking into account these ranges, and
considering ammonium and organic nitrogen releases similar to those from forested soils, a
value of 2.5 mgN.L-1 was used to estimate total dissolved nitrogen leaching from cultivated
soils in the upstream sub-basins of the Red River. However, due to the anaerobic nature of
waterlogged paddy-field soils, no nitrate nitrogen is exported from wet rice fields [Reddy and
Patrick, 1986]. This is confirmed by our measurements of water draining paddy rice fields in
the Hanoï area, where the nitrate concentrations were below 0.05 mgN.L-1, while the
ammonium-N concentrations amounted to 2 mg.L-1. Total phosphorus exportation was
considered to occur mainly in particulate form and to depend on soil erosion, particularly
under high flow conditions. The suspended sediment concentration, typical of the Red River
tributaries upstream from the large reservoirs, is greater than 5 g.L-1, with a phosphorus
content of 0.42 mg.g-1 and an nitrogen content of 1.3 mg.g-1 (see below). On this basis, we
attributed a mean P concentration of 2 mg.L-1 to headwaters draining cultivated soils in the
Red River basin.
At the outlet of each sub-basin diffuse nutrient fluxes from cultivated areas were estimated
from these concentrations and the specific discharge value measured in 2003 (see Table 5.4).
5.5.3 Nutrient output at the outlet of the sub-basins
Monthly sampling and nutrient analysis were carried out at the outlet of each upstream subbasin and at the Hanoï station during 2003. All samples were stored frozen in disposable
sterile polyethylene flasks. Phosphate, silica and ammonium were determined spectrophotometrically on water filtered through glass-fiber filters, according to respectively
Eberlein and Katter [1984], Rodier [1984] and Slawyck and MacIsaac [1972]. Nitrate was
determined after reduction into nitrite according to Jones [1984].
Total nitrogen and
phosphorus were determined on non-filtered water after sodium persulfate digestion and
134
mineralization at 110°C in an acidic phase. Total organic nitrogen concentration (TON,
mgN.L-1), only determined on 10 occasions, obeys the following relationship with suspended
solids (SS, mg.L-1): TON = 0.4 + 0.0013. SS (r²=0.91), from which it was extrapolated to the
other samples. Higher nutrient concentrations were observed during the rainy season than
during the dry one (Figure 5.4). N and P concentrations were almost always higher in the
Thao River, both at the upstream station of Yen Bai and upstream of Hanoï, than in the outlet
of the Da and Lo rivers.
4
Ptot, mgP.L-1
3
a)
Ntot, mgN.L-1
Yen Bai (Thao)
Vu Quang (Lo)
Hoa Binh (Da)
Ha noi (Hong)
1,0
Yen Bai (Thao)
Vu Quang (Lo)
Hoa Binh (Da)
Ha noi (Hong)
2
b)
0,5
1
0,0
0
J
F
M
A
M
J
J
A
S
O
N
D
J
F
M
A
M
J
J
A
S
O
N
D
Figure 5.4: Measured concentration of a) total phosphorus (Ptot, mgP.L-1) and b) total
nitrogen (Ntot, mgN.L-1) at the outlet of the 3 upstream sub-basins of the Red River system,
and at the Hanoï station in the delta area, in 2003. Ntot represents the sum of inorganic and
organic nitrogen.
The output of nitrogen and phosphorus at the outlet of each sub-basin (Table 5.4) was
estimated from the monthly data combined with daily measurements of discharge at the same
stations (Figure 5.2).
The complexity of the hydrological network in the Delta area and the lack of regular analyses
at the outlet of each of the numerous branches of the Red River discharging into the Tonkin
Bay, prevent accurate estimates of the total riverine nutrient delivery. Nguyen Ngoc Sinh et
al. [1995] estimated the flux of total nitrogen and phosphorus at 220-250 106 kg.y-1 and 61130 106 kg.y-1 respectively for the entire basin of the Red-Thai Binh river system, but this
represents a larger river basin than the one considered here. For this reason, the nutrient fluxes
at the outlet of the whole delta area, were calculated by extrapolating the flux measured at the
Hanoï gauging station taking into account the respective delta area in the watershed,
according to the following formula:
Flux at delta outlet =
(Flux at Hanoï – Σ flux upstr. tribut.) x tot.delta area / delta area at Hanoï +Σ flux upstr. tribut.
135
5.6 Discussion
All the nitrogen and phosphorus flux estimates discussed above are expressed in watershed
area specific fluxes to facilitate comparisons between the sub-basins (Figure 5.3 a, b, c, d).
5.6.1 Balancing the budgets
The flux estimates (Table 5.4 and Figure 5.3) rely on a wide variety of sources of differing
reliability, as well as on several debatable hypotheses. Values deduced from official statistics
may be inaccurate because all activities, particularly agricultural and cottage industry
production are not always correctly registered. Population data, at least for Vietnam, were
available as a geographically referenced data base, so that they could be allocated fairly
accurately to each sub-basin. This was not the case with many data (e.g. agricultural and
industrial productions) only available at the province level, that were reallocated to the subbasins according to the fractions of the province surface area where they belong (Figure 5.1),
with the implicit assumption that these activities have homogeneous spatial distribution.
Moreover, for lack of direct measurements in the studied area, many fluxes, such as
atmospheric deposition or nitrogen fixation, were estimated from data in the literature
concerning similar regions. For these reasons, we a priori estimate in the order of 25-50 %
the confidence level of our figures, which must be taken with caution. Nevertheless, the
resulting budgets appear quite consistent. As mentioned above the food and feed budget
(Table 5.8) can be balanced provided an ‘unregistered’ feed source is taken into account.
However, the magnitude of this ‘missing feed’ (which might represent grazing on rangeland
or in non agricultural areas) nowhere exceeds 20% of the calculated livestock requirements.
The nutrient budgets of agricultural soils are also coherent. Their nitrogen budget regularly
shows an excess of inputs over the outputs, which might be explained, either by an
overestimation of fertilizer or of other inputs, or by underestimation of loss processes. The
fact that cultivated plant uptake is lower than N inputs from fertilizer is not particularly
surprising: it is generally so excepted in countries with very low inputs. (Krupnik et al.,
2004). On the other hand, gaseous losses from soils have not been taken into account in our
agricultural budget, and probably explain the apparent surplus. Both denitrification and
ammonia volatilization and denitrification are known to be quite significant in paddy rice
soils (Bouwman et al., 2002). If the former process should not be taken into account in our
‘new’ nitrogen budget (as ammonium deposition has not been considered neither), the latter
might in itself explain the gap of the budget. Indeed, the shortfall in the balance of the
different agricultural soil budgets in the sub-basins is related to the relative acreage of paddy
136
rice fields (Figure 5.5). Extrapolating the observed trend to 100% of paddy-field surface area
would provide a denitrification rate of about 100 kgN.ha-1.y-1, which is in the range of
reported values for denitrifcation in fertilized paddy fields or nitrate contaminated wetlands
(50-120 kg ha-1.y-1 as N) [Reddy and Patrick, 1986; Rolston et al., 1978]. As far as
phosphorus is concerned, its budget in agricultural soils shows either excess or default inputs
compared to the outputs according to the sub-basins. The mountainous Da River basin, and
the Lo River basin, characterized by a dominance of industrial crops both appear to show
large erosion losses of phosphorus, while the phosphorus budget of the Thao basin and the
Delta area show a phosphorus accumulation in the agricultural soils.
excess N, kgN/km²/yr
10000
range of
denitrification
rate in paddy
rice fields
7500
5000
2500
0
0
20
40
60
80
100
% paddy rice
Figure 5.5. Estimated balance default of the agricultural soil budget in the sub-basins of the
Red River, plotted against the percentage area occupied by paddy rice fields. The trend
extrapolates to plausible denitrification rates in fertilized paddy rice plots (i.e. 100 kgN. ha1 -1
.y at 100%).
The budget of forested soil was not fully established because of a lack of reliable estimates of
forest primary production. Our estimates of nitrogen fixation and deposition on forested soils
lead to a total input of 1000 kg.km-².y-1 (expressed per surface of forested areas), far in excess
of the nitrogen output by forested soil leaching and erosion (312 kg.km-².y-1). Wood export
can be estimated from the total wood production of the basin, i.e. 3700 103 kg.y-1 (Table
5.10). Considering a mean N content of 0.2% in wood, this represents a nitrogen export of
only 92 kg.km-².y-1 from the 80700 km² of forested area in the Red River basin as a whole.
137
As far as the budget of the hydrosystem is concerned, the nitrogen export calculated at the
outlet of the sub-basins is clearly smaller than the sum of the inputs. The corresponding
nitrogen retention within the hydrographic network represents respectively 36%, 0.5 %, 14 %
and 20% of all the inputs to the river network for the Da, Lo, Thao and Delta sub-basins.
Similarly, the phosphorus budget shows the retention of 83 %, 78 % and 46 % for the Da, Lo
and Thao sub-basins respectively. The much higher retention in the Da and Lo sub-basins
commonly reported in the literature, is obviously related to the presence of large reservoirs
(Hoa Binh on the Da, 208 km² and Thac Ba on the Lo, 235 km²) in their downstream course,
which trap a great deal of suspended matter and associated phosphorus [Vorösmarty et al.,
1997; Garnier et al., 1999). The phosphorus budget of the delta, however, shows a deficit of
about 8%. Although this only represents a minor imbalance, it might reflect an
underestimation of the industrial contribution of phosphorus to the river system.
5.6.2 Biogeochemical functioning of the sub-basins
The four sub-basins in this study have quite different land use patterns. The population
densities of the three upstream river basins are similar (101-150 inhab.km-², Table 5.3), but
their agricultural activities differ greatly (Table 5.2): i) the Da river basin is mostly forested,
ii) the Lo river basin is predominantly devoted to industrial crops, mainly sugar cane, tea and
rubber and iii) the Thao river basin also has large areas of industrial crops but a greater
fraction of its surface area is devoted to rice production. The population is concentrated to the
delta, (population density greater than 1000 inhab.km-²) where rice production and livestock
farming are the most important agricultural activities.
The differing land use patterns result in a varied biogeochemical functioning of the systems.
When the total agricultural production in each sub-basin is plotted against the total
consumption by humans and cattle (both expressed in terms of kg.km-².y-1 of nitrogen), the
resulting diagram, similar to the classical P/R diagram of functional ecosystem analysis,
makes it possible to define the degree of autotrophy (P) or heterotrophy (R) of a regional
human system (Figure 5.6). An ecosystem is said autotrophic when its net primary production
(integrated over a certain time period) surpasses its respiration: it then accumulates or exports
biomass and represents a sink for nutrients and carbon dioxide. When respiration (either
supported by external inputs or by consumption of internal stocks of organic matter) is larger
than primary production, the system is said heterotrophic, and exports nutrient and CO2
(Odum, 1959). By analogy, a regional watershed can be said autotrophic (with respect to
human economy) when its agricultural production is greater than the consumption of
agricultural products by human and cattle. Whereas the three upstream Red River sub-basins
(particularly the Da basin), are slightly autotrophic, the delta system is clearly heterotrophic,
138
which is in agreement with the fact that the former three basins export agricultural products,
while the latter imports them (Figure 5.3). Similar data derived from analyses of a few other
regional budgets published in the literature for other regions of the world are included in
figure 5.6 for comparison. In the Republic of Korea [Bashkin et al., 2002], with a population
density of 395 inhab.km-², the situation is rather similar to that of the upstream catchment of
the Red River system, but with a slightly greater autotrophy. The Scheldt basin [De Becker et
al, 1988] as well as those of the east coast of the United States [Boyer et al, 2002], both area
with high human population densities and intensive cattle farming, are examples of
heterotrophic systems, depending however much more on imports of food and feed than the
Red River delta. The upstream Seine river basins [Billen et al., 2001], as well as the
Mississippi basin [Howarth et al, 1996] are examples of autotrophic systems with moderate
population densities, that export large amounts of agricultural products.
agricultural production, kgN/km²/yr
15000
Autotrophy
10000
Scheldt
Seine
Delta
5000
Heterotrophy
Mississippi
Korea
Da
Thao
Lo
US East coast
0
0
5000
10000
15000
human and animal consumption, kgN/km²/yr
Figure 5.6: Characterisation of the degree of auto- or heterotrophy of regional human
exploited systems: plot of agricultural production against total food and feed consumption by
humans and domestic animals. The data from the Red River are compared with literature data
from other river systems. (See text for explanation).
5.6.3 Riverine nutrient export
The specific riverine export of nutrients from the three sub-basins is quite low (respectively
740, 930 and 370 kg.km-².y-1 for nitrogen and 70, 150 and 140 kg.km-².yr-1 for phosphorus
from the Da, Lo, Thao sub-basins). The estimated total delivery at the outlet of the delta is
139
855 kg.km-².y-1 for nitrogen and 325 kg.km-².y-1 for phosphorus when expressed with respect
to the total Red River basin area, slightly lower than the specific fluxes cited by Nguyen Ngoc
Sinh et al., [1995] for the outlet of the Red-Thai Binh river system (1180-1480 kg.km-².y-1 for
nitrogen and 350-700 kg.km-².y-1 for phosphorus). According to our estimates, the
contribution by the delta area alone represents 4310 kg.km-².y-1 of nitrogen and 3600 kg.km².y-1 of phosphorus .
These results are in agreement with the view, expressed by Howarth et al., (1996) and Boyer
et al., [2002], that nitrogen export in streamflow is strongly related to total new inputs of
nitrogen to the catchment (Figure 5.7), although only 20-25 % of these new inputs of nitrogen
are exported by the river system. The Red River delta area appears to be one of the most
heavily loaded systems documented in the literature.
5000
riverine export, kgN/km²/yr
Delta
4000
3000
2000
whole Red R
Lo
Da
1000
Thao
0
0
5000
10000
15000
20000
total new inputs, kgN/km²/yr
Figure 5.7: Riverine export of nitrogen plotted against total inputs of new (see text for
definition) fixed nitrogen to the watershed. The data from the Red River are compared with
literature data from other river systems. (See text for explanation).
The molar N/P ratio of riverine delivery at the outlet of the Red River basin is 5.8. This value
is much lower than the Redfield ratio (16) of marine phytoplanktonic algae, indicating that
nitrogen rather than phosphorus is the potentially limiting factor of algal growth in the plume
of the Red River in the Tonkin Bay. This conclusion is particularly important in view of the
recent work by Wu et al [2003] demonstrating that nitrogen also limits net phytoplankton
growth in the offshore waters of the South China Sea, where nitrogen fixation remains at very
low levels.
140
Our measurements of silica flux at the Hanoï station show a silica delivery of 2920 kg.km-².y-1
as Si, indicating a molar Si/N of 1.7 in the nutrient fluxes carried by the river, in excess of the
requirements of marine diatom growth (Si/N generally close to 1, [Conley et al., 1993; Billen
and Garnier, 1997]). The increased human activity in the Red River watershed, particularly in
its delta, may further enrich the system in nitrogen and phosphorus along its aquatic
continuum. However, the Tonkin Bay does not, at the moment, seem threatened by harmful
marine eutrophication processes characterized by depletion of silica in relation to nitrogen and
phosphorus as well as by a proliferation of undesirable non diatom algae [Officer and Ryther,
1980; Billen and Garnier, 1997; Conley et al., 1993; Garnier and Billen, 2002; Garnier et al.,
in press].
5.7 References
Bashkin, V. N., S.U. Park, M. S. Choi and C.B.Lee (2002), Nitrogen budgets for the Republic
of Korea and the Yellow Sea region. Biogeochemistry 57/58: 387-403.
Beaton, J.D., T. L. Roberts, E.D.H. Halstead and L.E. Cowell (1995), Global transfers of P in
fertilizer materials and agricultural commodities. In Phosphorus in the global
environment: Transfer, cycles and management. H. Tiessen eds. Chichester, England,
John Wiley and Sons Ltd. Scope 54: 462.
Billen, G., M. Somville, E. DeBecker and P. Servais (1985), A nitrogen budget of the Scheldt
hydrographic basin. Neth J. Sea Res. 19: 223-230
Billen, G. and J. Garnier (1997), The Phison River plume: coastal eutrophication in response
to change in land use and water management in the watershed, Aquat. Microb Ecol 13:
3-17.
Billen, G., J. Garnier, C. Deligne, C. Billen (1999), Estimates of early industrial inputs of
nutrients to river systems : implication for coastal eutrophication. The Sciences of the
Total Environment, 243/244: 43-52.
Billen, G., J.Garnier, A. Ficht and C. Cun (2001), Modelling the response of water quality in
the Seine estuary to human activity in its watershed. Estuaries, 24(6): 977-993.
Bleken, M.A. and L. Bakken (1997), The nitrogen cost of food production: Norwegian society
26. Ambio 26(3): 134-142.
Boyer, E.W., C.L. Goodale, N.A. Jaworski and R.W. Howarth (2002), Anthropogenic
nitrogen sources and relationships to riverine nitrogen export in the northeastern
U.S.A. Biogeochemistry 57/58: 137-169.
141
Bouwman A.F, J.M Boumans and N.H.Batjes (2002) Emissions of N2O and NO from
fertilied fields: Summary of available measurement data. Global Biogeochem. Cycles,
16: 6/1-6/12
Campo, J., M. Maass, V.J. Jaramillo, A. Martinez-Yrizar and J. Sarukhan (2001), Phosphorus
cycling in a Mexican tropical dry forest ecosystem. Biogeochemistry 53: 161 - 179.
Cao Van Sung (1995), Environment and bioresources of Vietnam: present situation and
solutions. Thê Gioi Publishers, Hanoi, Vietnam: 235 pp.
Collins, R and A. Jenkins (1996), The impact of agricultural land use on stream chemistry in
the Middle Hills of the Himalayas, Nepal. Journal of Hydrology 185: 71-86
Conley, D.J., L. S Claire, E. F. Stoermer (1993), Modification of the biogeochemical cycle of
silica with eutrophication. Marine Ecology progress series, Published November 4,
101: 179-192.
Chestnut, T.J., D. J. Zarin, W.H. McDowell and M. Keller (1999), A nitrogen budget for latesuccessional hillslope tabonuco forest, P. Biogeochemistry 46: 85-108.
Chinadata, (1998), Statistical yearbook of Yunnan, Vol 1997, Vol 1998, Vol 2000 China
Statistical
Publishing
House,
(Basic
Information
of
Yunnan,
China),
(http://chinadatacenter.org), China.
Dang Anh Tuan, (2000), The Red River Delta - The Cradle of the Nation, published in Hanoi,
Vietnam, 53 pp.
De Becker, E., G. Billen and P. Servais (1988), Nitrogen budget at the scale of the Belgian
hydrographical network, Water Supply 6: 327-333.
Eberlein, K and G. Katter (1984), Automatic method for the determination of orthophosphate
and dissolved phosphorus in the marine environment. Fresenius Z. Anal. Chem. 326:
354-357
FAO, (1990-1998), Faostat statistics database (Fertilizer data used for China, Vietnam, Laos),
Faostat database Results, copyright FAO 1990-1998, http://www.apps.fao.org/;
Forti, M.C. and C. Neal (1992), Hydrochemical cycles in tropical rain forests: an overview
with emphasis on central Amazonia. Hydrology 134: 103-115.
Galloway, J. N., W. H. Schlesinger, H.L. III, A. Michaels and J.L. Schnoor (1995), Nitrogen
fixation: anthropogenic enhancement-environmental response. Global Biogeochemical
cycles 9: 235-252.
Galloway, J.N, E.B Cowling, S.P Seitzinger and R.H. Socolow (2002), Reactive Nitrogen: too
much of a good thing? Ambio 31(2): 60-63
142
Galloway, J.N and E.B. Cowling (2002), Reactive Nitrogen and the world: 200 years of
change Ambio 31(2): 64-71
Garnier, J., Leporcq B. Sanchez N. and Philippon (1999)- Biogeochemical budgets in three
large reservoirs of the Seine basin (Marne, Seine and Aube reservoirs).
Biogeochemistry 47: 119-146
Garnier, J. and Billen G. (2002). The Riverstrahler modelling approach applied to a tropical
case study (The Red –Hong- River, Vietnam): nutrient transfer and impact on the
Coastal Zone. SCOPE, Coll. Mar. Res. W. 12: 51-65
Garnier, J, Sferratore A., Meybeck M., Billen G. and Dürr H. (in press). Modelling silica
transfer processes in river catchments. SCOPE book series.
Green, P. A., C.J. Vorosmarty, M. Meybeck, J.M. Galloway, B.J. Peterson and E.W. Boyer
(2004), Pre-industrial and contemporary fluxes of nitrogen through rivers: a global
assessment based on typology. Biogeochemistry 68: 71-105
Hedlund, A., E. Witter and Bui Xuan An. (2003), Assessment of N, P and K management by
nutrient balances and flows on peri-urban smallholder farms in southern Vietnam,
European Journal of Agronomy, Volume 20, Issues 1-2: 71-87
Howarth, R. W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A.Downing,
R. Elmgren, N. Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch
and Zhu Zhao-Liang. (1996), Regional nitrogen budgets and riverine N &P fluxes for
the drainnages to the North Atlantic Ocean: Natural and human influences.
Biogeochemistry 35: 75-139.
IMH, (1997-2003), Journal of Meteo-hydrology, Insitute of Meteo-Hydrology in Vietnam,
Hanoi. (Journal monthly during the periods of from 1997 to 2003)
ITP, Institute Technique du Porc. (2000), Phosphore et Environnment, ITP editor, Paris,
France.
JICA, (2000), The Study on Environmental Improvement for Hanoi City in the Socialist
Republic of Vietnam, Final report, Data book., Nippon Koei Co. Ltd, Japan Internation
Cooperation Agency (JICA) and Hanoi People's Committee - Socialist Republic of
Vietnam., Hanoi, 350p.
Jones, M.N. (1984). Nitrate reduction by shaking with cadmium, alternative to cadmium
columns. Water Res. 18:643-646.
Kao, S.J., F.K. Shiah, K.K. Liu and J. Owen (submitted), Export of dissolved inorganic
nitrogen from a partially cultivated mountainous watershed (Lanyang-Hsi) in Taiwan.
Biogeochemistry.
143
Krupnik et al., 2004. in Agriculture and the nitrogen cycle, edited by A.R. Mosier, J.K. Syers,
and J.R. Freney, pp. 193-207, Island Press, Washington, D.C., 2004.
Le Xuan Tu and Huynh Phu, (1998), Studies on treatment of waste water from paper
production factories. Procedings of National Conference on Environment, Hanoi,
Vietnam.
Lelieveld, J. and F. Dentener, (2000), What controls tropospheric ozones? J. Geophys. Res.
105: 3531-3551.
Liu, A.Y.C. (2001), Markets, Inequality and Poverty in Vietnam. Asian Economic Journal 15(
21): 217-235.
MacDowell, W.H. and C. E. Asbury (1994), Export of carbon, nitrogen and major ions from
three tropical mountain watersheds. Limnol. Oceanogr. 39: 111-125.
Madsen, T, H.B. Boyd, D. Nylén, A.R Pedersen, G. Petersenand and F. Simonsen (2001),
Environmental and Health Assessment of Substances in Household Detergents and
Cosmetic Detergent Products. Environmental Project No. 615 Miljøprojekt. Danish
Environmental Protection Agency.
Martin-Prével, P., J. Gagnard and P. Gautier (1984), L'Analyse végétale dans le contrôle de
l'alimentation des plantes tempérées et tropicales. Paris, Technique et Documentation
(Lavoisier),
McKee, L.J. and B.D. Eyre (2000), Nitrogen and phosphorus budgets for the sub-tropical
Richmond River catchment, Australia. Biogeochemistry 50: 207-239.
Meybeck, M., D. Chapman and R. Helmer (1989), Global freshwater quality: a first
Programme Basil Blackwell, Inc.
Morel, R. (1996), Les sols cultivés. Technique and documentation, 2ème édition. Paris,
France.
MOSTE, (1997), Vietnamese general statistics officer, Ministry of Science, Technology and
Environment of Vietnam, General statistics editor, Hanoi.
MOSTE, (1999), Wastewater quality of some industrial branches in Viet Nam in Resource
and present condition of environment in Vietnam in years of 1990s, Hanoi, Vietnam,
219pp.
Pham Van Ninh and Nguyen Vu Trong. (1995), Case study report on Red River Delta
in Vietnam - Project on integrated management and conservation of nearshore coastal
144
and marine areas in East Asia region (EAS-35) United Nations Environment program.
Regional coordinating for the east Seas (ESA/RCU), 78 pp.
Odum, E.P. 1959. Fundamentals of Ecology. Saunders, ed. 546 pp.
Officer, C.B. and J.H. Ryther (1980), The possible importance of silicon in marine
eutrophication. Mar Ecol Prog Ser 3: 83-91
Pham Hung Viet, Pham Manh Hoai, Nguyen Xuan Trung and Vu Duc Nam. (1998),
Chemical composition and acidity of atmospheric deposition in the northeastern part
of Vietnam. In: Rhode, Kuylenstierna, Boonjawat, Ayers (Eds.), Proceedings of the
Fourth Meeting of the IGAC/DEBITS: Composition and Acidity of Asian
Precipitation (CAAP), 9/12 November, Chulalongkorn University, Bangkok.
Pham Xuan Nam (2001). Rural development in Vietnam, the search for ustainable
development. Social Scineces publishing House, Hanoï.
Pilbeam, C. J., B.P. Tripathi, D.P. Sherchan, P.J. Gregory and J. Gaunt (2000), Nitrogen
balances for households in the mid-hills of Nepal. Agriculture, Ecosystem and
Environnment 79: 61-72.
Reddy, K. R and W. H. Patrick (1986), Denitrification losses from flooded rice fields. In: S.K.
De Datta and W.H. Patrick (editors), Nitrogen Economy of Flooded Rice Soils.
Martinus Nijhoff, Dordrecht, 15-39p
Robertson, G.P and T. Rosswall (1986), Nitrogen in West Africa: The regional cycle. Eco.
Monogr. 56: 43-72
Rodier, J (1984), L'analyse de l'eau. 7ème édition. Dunot (ed), France, 1365 p.
Roger, P.A. and J.K. Ladha (1992), Biological nitrogen fixation in wetland rice fields:
estimation and contribution to nitrogen balance. Int. Rice Res. Inst.,
Philippines.
Manila,
Plant and Soil 141(1-2): 41-55.
Roldan, G. and E. Ruiz (2001), Development of Limnology in Colombia. In Limnology in
Developing Countries, edited by R. G. Gopal.B and Wetzel. Volume 3: 69-119.
Rolston, D.E., D.L. Hoffman and D.T. Toy (1978), Field measurements of denitrification, I.
Flux of N2 and N2O. Soil Sci. Soc. Am. J. 41: 863-869.
SCS, (1992), Soil Conservation service, Agricultural waste management field handbook,
chapter 4. US Gouvernment priting office, Washington, DC.
Servais, P., J. Garnier, N. Demarteau, N. Brion and G. Billen (1999), Supply of organic
matter and bacteria to aquatic ecosystems through waste water effluents.
Water
Research. 33: 3521-3531.
145
Sheldrick, W., J.K. Syers and J. Lingard (2003), Soil nutrient audits for China to estimate
nutrient balances and output/input relationships. Agricultural, Ecosystems and
Environment 94: 341-354.
Slavyck, G. and J.J. McIsaac (1972), Comparison of two automated ammonium methods in a
region of coastal upwelling. Deep-Sea Res. 19: 1-4
Smil, V. (1999), Nitrogen in crop production: an account of global flows. Global
Biogeochem. Cycles 13(2): 647-662.
Soltner, D. (1979), Alimentation des animaux domestiques, 13éme édition, France.
Stevenson, F.J. and M.A. Cole (1999), Cycles of soil, Carbon, Nitrogen, Phosphorus, Sulfur,
Micronutrients. John Wiley and Sons publisher, Canada, volume 1, 427 p.
Stoorvogel, J.J, B.H. Janssen and N. Van Breemen (1997), The nutrient budgets of a
watershed and its forest ecosystem in Taï National Park in Côte d'Ivoire, Africa.
Biogeochemistry 37: 159-172.
Thomas, G.W. and J.W. Gilliam (1977), Agro-ecosystems in the USA. Agroecosystem 4: 182239.
Van Breemen, N., E.W. Boyer, C.L. Goodale, N.A. Jaworski, K. Paustian, S.P. Seitzinger, K.
Lajtha, B. Mayer, D. Vandam, R.W. Howarth, K.J. Nadelhoffer, M. Eve and G. Billen
(2002), Where did all the nitrogen go? Fate of nitrogen inputs to large watersheds in
the northeastern U.S.A. Biogeochemistry 57/58: 267-293.
Van der Hoek, K.W. and A.F. Bouwman (1999), Upscaling of nutrient budgets from
agroecological niche to global scale, Book: Nutrient disequilibri in agroecosystems
Smaling EMA, O. Ocenema and LO Fresco eds., CAB International publisher, 57-73p.
Van Drecht, G., A.F Bouwman, J.M. Knoop, A.H.W. Beusen and C. R. Meinardi (2003)
Global modelling of the fate of nitrogen from point and non-point sources in soils,
groundwater and surface water, Global Biogechem. Cycles, 17(4) 26.1-26.20, doi:
10.1029/2003GB002060,2003.
VAST., 2000. Proceedings annual of Project Reports on Environments of Vietnam Academy
of Science and Technology in 2000, Hanoi. VAST publisher. 310p.
Vitousek, P.M., K. Cassman, C. Cleveland, T. Crews, C.B. Field, N.B. Grimm, R.W.
Howarth, R. Marino, L. Martinelli, E. B. Rastetter and J. I. Sprent (2002), Towards an
ecological understanding of biological nitrogen fixation. Biogeochemistry 57/58: 1-45.
Vörösmarty C.J., M. Meybeck, B. Fekene and K. Sharma (1997). The potential impact of neoCastorization on sediment transport by the global network of rivers. In: Human impact
146
on erosion and sedimentation (Proceedings of the Rabat Symposium, April 1997),
IAHS Publ, n° 245.
Weijin, Yan, C. Yin and Shen Zhang. (1999), Nutrient budgets and biogeochemistry in an
experimental agricultural watershed in Southeastern China. Biogeochemistry 45: 1-19.
Wilcke, W., S. Yasin, C. Valarezo and W. Zech (2001), Change in water quality during the
passage through a tropical montane rain forest in Ecuador. Biogeochemistry 55: 45-72.
Wu Jingfeng, Shi-Wei Chung, Liang-Saw Wen, Kon-Kee Liu, Yuh-ling Lee Chen, HoungYung Chen and David M. Karl. (2003), Dissolved inorganic phosphorus, dissolved
iron and Trichodesmium in the oligotrophic South China Sea. Global Biogeochem.
Cycles, 17(1), 1008, doi 10.1029/2002GB001924
Xing, G. X. and L. Zhu (2002), Regional nitrogen budgets for China and its major watershed.
Biogeochemistry 57/58: 405-427.
Acknowledgements
This study was realized in the framework of a French-Vietnamese co-operation. Thanks are
due to Georges Vachaud, Research Director at the CNRS, for the coordination of the
programme ESPOIR (CNRS-CNSTV). Le Thi Phuong Quynh’s Ph-D thesis is supported by
the French Ambassy and by the Pierre and Marie Curie University (Paris 6).
147
148
Modelling nutrient transfer in the river system
CHAPTER 6
Modelling nutrient transfer in the river system:
implementation of the Seneque/Riverstrahler software
6.1. Introduction
The biogeochemical functioning of a river system is characterized by the fluxes of transfer,
transformation and retention of biogenic elements during their downwards travel from the
terrestrial watershed to the sea through the whole drainage network.
Both the importance of the inputs of these elements and the intensity of the processes they
undergo within the system depend on the complex interplay of climatic, geomorphologic and
anthropogenic factors. Understanding the role of all these factors, and assessing the
relationship between human activity in the watershed and ‘water quality’ of the river system,
given the variability of natural factors, is a prerequisite for a rational management of water
resources. The Riverstrahler model (Billen et al., 1994; Garnier et al., 1995; Billen and
Garnier, 1999) has been established for this purpose.
Recently, the model has been
encapsulated into a GIS interface in order to build a generic and spatially explicit software
(Seneque/Riverstrahler) which can be implemented to any watershed, provided a suitable data
base is assembled under a specific format (Ruelland, 2004; Ruelland et al. in prep). This
approach, which was first developed for the well documented Seine river system, has proven
particularly fruitful for addressing different water management issues in temperate regions,
including the questions of oxygen deficits in regulated rivers (Garnier et al., 1999), of nitrate
contamination from diffuse sources (Billen and Garnier, 1989), of excessive algal
development either in the river drainage or in the adjacent coastal zones (Garnier et al., 2005;
Cugier et al., 2005).
Although it is conceived as a generic tool, implementing this model for the case of less
documented river systems where no Water Agencies or similar organisms are established,
might be much more difficult, by lack of access to the required data bases.
This chapter is prepared as an article to be submitted under the title: Modelling the biogeochemical
functioning of the Red River system: implementation of the Seneque/Riverstrahler software.
149
In this paper we describe how the Seneque/Riverstrahler has been successfully applied to the
case of a tropical river: the Red River, in Vietnam and China. We will describe the minimum
information required by the software to take into account the various constraints to the river
drainage functioning, and how the corresponding information has been gathered for the
special case of the Red River. Confrontation of the model results with observations on water
quality gathered at the outlet of the major tributaries of the Red River system in the scope of a
survey programme at monthly intervals throughout two years, will allow validating the
modelling approach. Some general conclusions on the biogeochemical functioning of the Red
River system and its nutrient export will be presented.
6.2. The Riverstrahler model
Riverstrahler describes the drainage network of any river system as a combination of basins,
idealized as a regular scheme of confluence of tributaries of increasing stream order, each
characterized by mean morphologic properties, connected to branches, represented more
realistically, with a higher spatial resolution. The advantage of this representation of the
drainage network is that it allows, with reasonable calculation time, to take into account both
the processes occurring in small first orders, headwater streams and those occurring in large
tributaries. The water flows in the hydrographical network are calculated from the specific
discharges generated within the watershed of the different sub-basins and branches
considered. These are calculated from rainfall and potential evapotranspiration by a simple
rainfall-discharge model with two compartments (Hydrostrahler module). The discharge in
any stream order river or branch is the sum of two components, one corresponding to surface,
or sub-root (hypodermic) runoff, the other to groundwater, base flow.
The essence of the model is to couple these water flows routed through the defined structure
of basins and branches, with a model describing biological, microbiological and
physicochemical processes occurring within the water masses. The module representing the
kinetics of the processes is known as the Rive model. The state variables comprise nutrients,
oxygen, suspended matter, dissolved and particulate non living organic carbon, as well as
algal, bacterial and zooplanktonic biomasses. Most processes important in the transformation,
elimination and/or immobilization of nutrients during their transfer within the network of
rivers and streams are explicitly calculated by the Rive, including algal primary production,
aerobic and anaerobic organic matter degradation by planktonic as well as benthic bacteria
150
with coupled oxidant consumption and nutrient remineralization, nitrification and
denitrification, phosphate reversible adsorption onto suspended matter and subsequent
sedimentation, etc. A detailed description of the Rive Model and of the physiological
parameters used can be found in Garnier et al. (2002a).
Besides morphological and climatic constraints, the Riverstrahler takes into account diffuse
and point sources of nutrient from land based anthropogenic sources. Diffuse sources of
nutrients are taken into account by assigning a constant concentration for all nutrients to
surface and groundwater flow respectively. Point sources, typically wastewater discharges,
must be specified by stream-order for the basins, and at their exact location for the branches.
The newly developed Seneque interface allows to derive from a general GIS data base
covering the watershed, all the input files required for running the Riverstrahler model, and
this for any portion of the drainage area, represented as a particular structure of basins and
branches defined by the user according to the spatial resolution required for his application
(Ruelland, 2004 ; Ruelland et al., in prep). Assembling a suitable GIS database is thus the
key for running this generic software.
6.3. Geomorphology
The whole GIS data base is structured according to the representation of the drainage
network, as a system of connected directional arcs, with each confluence marking the
beginning of a new arc. Each arc is described by data such as Strahler stream order, length,
width, slope... To each arc corresponds an elementary watershed, representing the area
directly drained to this arc, which thus constitutes the elementary spatial grid of the model.
The best way to obtain the drainage network and the elementary watershed coverage is to
derive them from a Digital Elevation Model (DEM) (Tarboton et al., 1991). For the Red
River, we made use of the digital elevation model SRTM at 3” resolution made available at a
global scale by NASA (www:\\NASA.org), (Figure 6.1a). The Standard ArcInfo Workstation
software has been used to generate the watersheds and the network coverage. First, SRTM
data sets have been converted into Arcinfo Grid, and then sinks have been filled. A minimum
watershed size of 25000 cells (approximatively 200 km²) was imposed in the process, limiting
the upstream drainage network to stream orders 2 or 3. The obtained drainage network was
compared to available topographic maps and the few required corrections were made
manually.
151
100 km
Figure 6.1: Relief of the Red River basin deduced from the STRM Digital elevation model
(www:\\NASA.org) and the structure of the drainage network and elementary watershed
derived from it, using the procedure described by Tarboton et al., 1991.
152
6.4. Hydrology
The Hydrostrahler model included in the Seneque/Riverstrahler software calculates the
seasonal variations of specific base flow and superficial runoff by periods of ten days for each
sub-basin. This result is used for reconstructing the discharge anywhere in the drainage
network, but also to calculate the diffuse sources of nutrient from the watershed. The
Hydrostrahler model requires daily rainfall and potential evapotranspiration data at a number
of stations in the watershed. We could obtain daily rainfall data for 13 meteorological stations
in the Vietnamese part of the basin (IMH, 1997-2004) and at the station Kunming in China,
for the period from 1997 to 2003. Evapotranspiration was calculated using Turc’s formula
(Turc, 1961), based on monthly temperature (T°C) and sunshine duration (Sdur, h) data
obtained from the respective meteorological stations:
ETPmm/month = 0.4 T°C (Ig+50)/(T°C+15)
where
T°C is the atmospheric temperature in °C in the period considering
Ig is the total solar radiation expressed in cal.cm-2.d-1 in the period considered, which
can be calculated by the relation:
Ig = IgA (0.18+ 0.62 h/H)
in which
IgA is the energy of solar radiation in the absence of atmospheric attenuation,
expressed in cal.cm-2.d-1.
h/H is the relative duration of sunshine, H being the duration of the astronomic
day and h, the duration of sunshine period per day.
IgA and H values, which only depend on the latitude and the period of the year,
are provided by Turc (1961).
The Hydrostrahler model involves four empirical hydrological parameters (soil saturation
level (solsat), superficial runoff rate (srr), infiltration rate (infr), groundwater runoff rate
(gwrr), the value of which is calculated by the software for any sub-basin from its lithological
characteristics, provided suitable parameters are defined for each lithological class of the
watershed.
Independently of the software, we developed an automated procedure (Le Thi Phuong Quynh
et al., subm.) allowing to calibrate the values of these parameters for the three upstream sub153
basins of the Red River, on the basis of observed discharge values at their outlets,
communicated by the Ministry of Natural Resources and Environment (MONRE, 2004),
(Table 6.1).
Table 6.1: Hydrological parameters derived by calibration of the Hydrostrahler model on
observed discharge values at the outlet of the three main sub-basins of the Red River for the
period 1997-2003 (Le Thi Phuong Quynh, subm).
Parameter
Thao
Da
Lo
solsat, mm
110
165
210
0.062
0.038
0.05
0.038
0.075
0.068
0.013
0.0026
0.001
-1
infr, d
srr, d
-1
gwrr, d
-1
A detailed lithological map of the Red River basin is not available. We used the information
provided by the global geological/lithological database of Dürr (2003) (Figure 6.2).
Combining the distribution of the lithology in the three sub-basins, with the calibrated values
of Table 6.1, we assigned a value of the hydrological parameters to each lithological class in
order to reproduce as well as possible the combined values for the three sub-basins (Table
6.2).
Figure 6.2: Lithological map of the Red River basin derived from the global
geological/lithological map of Dürr (2003).
154
Table 6.2: Hydrological parameters for each lithological class
Parameter
solsat, mm
-1
infr, d
srr, d-1
gwrr, d-1
plutonic
basic
paleozoic
acid
volcanic sedimentary
mesozoic
silicic
mesozoic
carbonated
alluvial
deposits
150
100
180
30
250
400
0.050
0.040
0.001
0.050
0.070
0.020
0.060
0.050
0.001
0.020
0.080
0.025
0.040
0.040
0.001
0.050
0.020
0.001
6.5. Role of dams
Two major dams are in operation in the upstream sub-basins of the Red River. Two others are
planned to be impounded in the next decade (Table 6.3). Le Thi Phuong, et al., (subm) have
proposed to represent the hydraulic behavior of these dams by simple management rules,
based on the value of their maximum and minimum volume, as well as two critical values of
the input discharge above which the dam is allowed to fill or below which it is emptied.
Based on these rules, an algorithm determining their period of filling and emptying in
function of upstream river discharge was constructed and corresponding parameters
determined (Table 6.3). Based on these parameters and on the pre-calculated water quality of
the inflowing water, a software associated to the Seneque/Riverstrahler model and very
similar to the Barman model described by Garnier et al. (2000), calculates the hydrological
and biogeochemical functioning of the dam, using exactly the same kinetic formulation of the
ecological processes. This model provides the files required for allowing Seneque to fully
take into account the role of the dams: a file providing daily values of inflowing and
outflowing discharge, and a file providing the quality of dam water.
Table 6.3: Some characteristics of the large dams impounded (Hoa Binh and Thac Ba) or planned
(Son La and Dai Thi) in the Red River basin (from Le Thi Phuong Quynh et al, submitted)
Name of dam
Hoa Binh
Thac Ba
Son La
River (sub-basin)
Date of impoundment
*Volume (min-max), 109 m3
*Critical upstream discharge, m3/s
Surface area, km²
Mean depth, m
Upstream watershed, km2
Da
Chay (Lo)
1972
0.78-2.94
200-190
235
42
6170
Da
Gam (Lo)
2010-2015
2010
9.3-25.5
0.5-3.0
850-750
?
440
42
60
70
26000
9700
1985
3.9 – 9.5
1750-1500
208
50
57285
Dai Thi
155
6.6. Land use and non-point sources of nutrients
To calculate diffuse sources of nutrients in each sub-basin, the Riverstrahler model assigns a
yearly constant mean composition to surface- and base flow runoff respectively, according to
land use of the watershed.
a.
b.
Figure 6.3: a. Land use map for the Vietnamese part of the Red River basin, based on the
data from MONRE (2004); b. Land use map for the whole Red River basin based on the
Global 1° land cover map of DeFries et al., 1998 and Hansen, et al., 2000.
156
A GIS land use coverage of the Vietnamese territory was made available from the Ministry of
Science, Technology and Environment (MOSTE, 1997). We considered the following 6
classes as the most relevant for our purpose: forest, grassland, paddy rice fields, other (dry)
cultures including industrial cultures, rocks and bareland, urban areas (Figure 6.3a). Similar
information was not available for the Chinese part of the basin. However, a global GIS
covering (DeFries et al., 1998; Hansen, et al., 2000) at a 1° resolution provide information on
the spatial repartition of forest (with a lot of details on the kind of forested formations),
grassland, cropland, bare ground, and urban area (Figure 3b). The legends of these two data
bases differ in the fact that the latter does not distinguish paddy rice fields from other
croplands. We arbitrarily assigned a constant proportion of 33% of rice fields in total cropland
to all elementary watersheds in China, a figure obtained from the general statistics of land use
of the Yunnan province as a whole (Chinadata, 2000).
The Seneque associates the spatial distribution of these land use classes with a parameter file
providing the corresponding concentrations of all variables in superficial and base flow issued
from these classes. The water composition (organic matter, nitrate, ammonium, total
phosphorus) assigned to each of the land use classes considered in the data base should be
calculated from data of empirical surveys of surface water composition of small streams
draining homogeneous basins with given land use. For the Red River basin, we relied on an
extensive survey of literature (see Le Thi Phuong Quynh et al., 2005, chapter 5) and on our
own unpublished measurements (see chapter 4), (Table 6.4).
As far as suspended matter is concerned, the rather low concentration observed in the Lo subbasin compared with the two other basins is rather paradoxical in view of the fact that this
basin has the greatest proportion of industrial crops in its watershed (Le Thi Phuong Quynh et
al., 2005). This difference, which probably results from a different geomorphological and/or
geological context in the Lo basin, led us to define two different values for the suspended
matter concentration associated with industrial crops in the Lo and the other sub-basins (Table
6.4). The total inorganic phosphate concentration is calculated from the suspended matter
concentrations using the measured total phosphorus content of suspended matter at the outlet
of the sub-basins, namely 0.43 mgP.g-1 for the Thao and Da Rivers, and 0.85 mgP.g-1 for the
Lo River (Le Thi Phuong Quynh et al., 2005).
157
Table 6.4: Composition of surface runoff according to land use in the Red River basin, as
taken into account by the Seneque/Riverstrahler model.
Land use class
mgN/l
NH4+
mgN/l
Ptot
mgP/l
5000
0.4
0.015
3
1000
3000
8000 (Lo 150)
3000
8000
0.4
1.4
2.8
0.02
2.8
0.015
0.015
0.015
2
5
0.6
0.43
4.8 (Lo 0.13)
1.8
4.8
Susp. matter.
mg/l
Rocks
Forest
Grassland
Industrial (dry) cultures
Paddy rice field
Urban areas
NO3-
Regarding dissolved silica concentration, the concentration is similarly related to the
distribution of lithological classes, obtained from the global lithological world map of Dürr
(2003) (Figure 6.2), using the corresponding mean SiO2 proposed by Meybeck (1986, 1987),
taking into account the mean temperature of the Red River basin (Table 6.5).
Table 6.5. Dissolved silica concentration associated with each lithological class (according to
Meybeck, 1986, 1987.
lithological class
plutonic acid
basic volcanic
paleozoic sedimentary
mesozoic silicic
mesozoic carbonated
alluvial deposits
Dissolved silica , mgSi.L-1
4.9
7.7
4.4
5.5
3.2
3.8
6.7. Wastewater point sources
The population in the whole basin was estimated to 30.02 million inhabitants for the year
1997, of which 65 % in Vietnam (MOSTE 1997), 34 % in China (Statistical Yearbook of
China data, 1998) and 1% in Laos. For Vietnam, the population of 5235 villages and towns
could be geo-referenced in the Red River basin GIS. For China, only data on the urban
population and rural population of the Yunnan province as a whole were available. In fact, the
Red River basin in Yunnan drains none of the major cities of the province (Kunming,
Dali,…). We therefore uniformly distributed the rural population density of Yunnan (81
inhab.km-2) within the Chinese part of the upstream Red River watershed. The population
158
density of the whole basin thus varies from 81-150 inhab.km-2 in the upstream watershed to
over 1000 inhab.km-2 in the delta area (Figure 6.4).
From the analysis of the domestic consumption budget of food and washing powders in
Vietnam, Le Thi Phuong Quynh et al. (2005) estimated the human per capita release of
nutrients to 0.010 – 0.015 kgN.cap.d-1 and 0.0046 – 0.005 kgP.cap.d-1, the lower values
corresponding to the poor rural region of the upstream basin, while the highest hold for the
urban area in the delta. Moreover, in small rural villages (<10 000 inhab.) in the upstream
part of the basin, we estimated that only 25% of domestic wastewater reaches the surface
waters, the remaining part being recycled in agriculture, while in large villages, and in the
delta region, where the population is mostly agglomerated and running water is present
everywhere, we considered that all domestic wastewater is discharged to the hydrosystem.
On this basis, a database of all domestic inputs of wastewater was established for the
Seneque/Riverstrahler model. A first census of industrial wastewater discharges has been
carried on (Le Thi Phuong Quynh et al., 2005) but it remains very partial and has not been
included in the data base.
1 -50
50-100
100-200
200-500
500-1000
1000-2000
2000-5000
5000-10000
Figure 6.4: Distribution of the population density in the elementary watersheds of the Red
River basin, as used for calculating the point sources of wastewater in the
Seneque/Riverstrahler model.
159
6.8. Validation and flux calculation
The Seneque/Riverstrahler software can be used to calculate the spatial and seasonal
variations of water quality at the scale of the whole drainage network. The interface allows
the user to immediately visualize the results under three formats:
- seasonal variations of discharge or concentrations of any variable at one station (either at the
outlet of the tributary of a specified stream order for a sub-basin, or at a specified kilometric
position for a river branch);
- longitudinal variations of discharge or concentrations along a river branch at a specified
time;
- cartographic representation of the variables (with an adjustable color code) over all basins
and branches of the simulations at a specified time period.
For the two former representations, the possibility exists to automatically compare the
calculation results with measured data when these are stored in the database. This comparison
allows the validation of the modelling procedure on recent well documented situations.
A two-year survey of water quality has been carried out at monthly intervals during the year
2003 and 2004 at the stations Yen Bai (Thao River), Vu Quang (Lo River) and Hoa Binh (Da
River), Son Tay and Hanoi (Hong river) (see chapter 4). In order to adapt the resolution of
the model to the requirement of the validation with respect to these sampling stations, as well
as in order to take into account the presence of the two dams in the upper drainage network, a
suitable spatial representation of the river system in the model should be defined. The chosen
representation is shown in Figure 6.5. It involves 7 basins, 5 branches and the two presently
operating dams. It treats the upper half of the whole Red River basins as “Strahler-idealized”
basins, while the lower half courses of all three major tributaries are treated as river branches
with a kilometric resolution.
As an example, the data obtained during the field campaigns in 2003 at the 4 stations cited
above are compared with the results of the model (Figure 6.6). On the other hand, we
compare the observations and the simulations at Yen Bai, on the Thao River, for both years
2003 and 2004 (Figure 6.7). At this stage, the agreement between observed and calculated
concentrations, although far from being perfect, is in general not bad: the model is able to
reproduce the observed general levels of nutrient concentration, which is not a priori obvious.
160
Figure 6.5: Spatial representation of the Red River drainage network by the
Seneque/Riverstrahler software for the validation of the results.
In fact, very few calibration procedures have been applied in the construction of the model,
excepted concerning the hydrological sub-model (see chapter 3). The agreement between the
model and the observations thus represents the exact measure of the correctness of our
representation of the system functioning and/or of our knowledge of the constraints
controlling it. The discrepancies between simulations and observations thus deserve close
examination, because they inform us on the weak aspects of our approach.
Concerning our knowledge of the constraints several weak points can be mentioned which
could be responsible for a part of the discrepancies between simulation results and
observation. The first lies in the meteorological data, especially in the Chinese part, which
covers a haft area of the Red River basin where we have only one station (Kunming) with
mean monthly meteorological data in the period from 1997 to 2001. We applied the mean
monthly values of this series to the period 2003 and 2004. This leads to errors both on
hydrology and water quality simulations. The same limitation of our knowledge of the
conditions in the Chinese part of the Red River basin concerns the distribution of the
population and industrial activity, that we know only as an overall figure for the whole
Yunnan province, of which the Red River basin makes 20.9%. Even for the Vietnamese part
of the basin, on the other hand, our knowledge of industrial wastewater releases is quite
insufficient and the figures we used are probably largely underestimated. We hope that a
161
better knowledge of all these constraints and their distribution at the scale of the whole Red
River basin would improve the simulation by the model.
As far as suspended matter is concerned, the level in the Lo River is strongly overestimated
by the model, despite the lower suspended solids delivery considered for industrial cultures in
the Lo basin with respect to the other sub-basins. On the contrary, the model underestimates
the suspended matter of the Thao River (Fig. 6.6 and 6.7). Clearly, a general assumption
concerning the relation of suspended matter yield with land use is not able to reproduce the
differences in suspended loading among the sub-basins. The process of erosion, which
generates the suspended solid load of the river network, should be described in a much more
refined way, even at the regional scale at which we are working. This problem is also
apparent on the simulations of total phosphorus, a large part of which is linked to suspended
matter: the model also overestimates total phosphorus in the Lo basin. However, as far as
dissolved phosphate concentrations are concerned, taking into account the variability of their
measurements at the low levels occurring in most of the stations, the model predictions are
rather satisfactory, which indicates that the kinetic formulation of the adsorption-desorption
equilibrium used in the Rive model (Garnier et al., 2005) is valid for the Red River.
Chlorophyll a concentrations predicted by the model, although they fluctuate may be too
much, reproduce some important trends revealed by the measurements: as observed, the Thao
river is the only one among the three large tributaries to develop a significant planktonic algal
biomass. The phytoplankton development occurs in spring, by low discharge, and to a lower
extend in autumn, after the flood.
The model does not capture all the observed variability of nitrate concentrations. Although the
general level is correctly reproduced for most stations, the model underestimates nitrate
concentration in the Thao River. On the other hand, the very low ammonium levels are
satisfactorily reproduced by the model. These low levels, in spite of significant inputs of
ammonium by point sources of waste water and diffuse sources from paddy rice soils, could
be the result of an active in-stream nitrifying activity. Setting the nitrifying activity to zero in
the model, results in much higher calculated ammonium concentration at Hanoi station.
Finally, the agreement between the model simulations and the observations of dissolved silica
concentration is so good. An important conclusion from the model simulation is that there is
no significant retention process of silica along the river system. No period of silica depletion
occurs, even during the limited planktonic blooms in the Thao River. The large dams on the
Da and Lo Rivers are not responsible for any significant silica retention.
162
0
0
J F MAM J J A SO ND
10
J F M AM J J A S O N D
J F MAM J J A SO ND
SiO2, mg.L-1
P-PO4, mg.L-1
SiO2, mg.L-1
Ptot, mg.L-1
Ptot, mg.L-1
P-PO4, mg.L-1
0
20
0.2
0.6
0.0
J F M AM J J A S O ND
SiO2, mg.L-1
P-PO4, mg.L-1
Ptot, mg.L
-1
N-NH4, mg.L-1
N-NH4, mg.L-1
N-NH4, mg.L-1
1
0.1
10
0
0.0
10
J F M AM J J AS O ND
1.2
0
0
0
J F MAM J J A S O ND
2
20
0.0
2
SiO2, mg.L-1
2000
4
0
0.1
P-PO4, mg.L-1
4000
N-NO3, mg.L-1
SS, mg.L
-1
Chla, µg.L-1
20
10
0.2
0.0
Ptot, mg.L-1
4000
Hong R.
0.0
0.6
J F MAM J J A S O N D
20
0.1
J FM AM J J A S OND
1.2
0
0
0.0
1
2
0
0.2
0.6
J FM AM J J A S OND
2
-1
0
J F MA M J J A S O N D
1
10
0.0
1.2
0
4
10
2
J F MAM J J A SO ND
N-NO3, mg.L
Chla, µg.L-1
SS, mg.L-1
Discharge, m3.s-1
0
0
0
20
2000
2
4000
Thao R.
J F MAM J J AS O ND
Discharge, m3.s-1
0
4000
10
4
0.1
J F M AM J J A S ON D
20
0.2
0.0
0
J F MAM J J A SO ND
N-NO3, mg.L-1
2000
0
0
8000
20
Chla, µg.L-1
SS, mg.L-1
Discharge, m3.s-1
4000
0.6
1
0
1.2
2
2
0
Lo R.
4000
8000
10
0
4
N-NO3, mg.L-1
2000
0
8000
Chla, µg.L-1
4000
20
N-NH4, mg.L-1
4000
Da R.
SS, mg.L-1
Discharge, m3.s-1
8000
Figure 6.6: Seasonal variations in 2003 of observed (open circles) and calculated (solid curve) variables (from left to right): discharge, suspended
solids (SS), chlorophyll a (Chla: µg.L-1), nutrient concentrations (nitrate and ammonium: mg N.L-1; total phosphorus and phosphates mg P.L-1;
dissolved silica: mgSiO2.L-1). From top to bottom are figured the four sampled stations: Hoa Binh (Da R.), Vu Quang (Lo R.), Yen Bai (Thao R.) and
Hanoi (in the delta, Red R.)
163
164
4000
Discharge, m .s
3
2000
2000
0
0
J F MAM J J A S OND
4000
SS, mg.L
SS, mg.L
-1
-1
4000
2000
2000
0
0
J F M A M J J A S ON D
20
2004
Chla, µg.L
Chla, µg.L
-1
-1
20
10
10
0
0
J F M A M J J A S O N D
4
N-NO3, mg.L
-1
N-NO3, mg.L
-1
4
2
0
2
0
2
N-NH4, mg.L
N-NH4, mg.L
-1
-1
2
1
0
1
0
1.2
Ptot, mg.L
-1
Ptot, mg.L
-1
1.2
0.6
0.0
0.6
0.0
0.2
P-PO4, mg.L
P-PO4, mg.L-1
-1
0.2
0.1
0.0
0.1
0.0
20
20
SiO2, mg.L
-1
-1
SiO2, mg.L
Figure 6.7: Station Yen Bai
(upper Thao R.) in 2003 (left) and
in
2004
(right):
seasonal
variations of observed (open
circles) and calculated (solid
curve) variables. From top to
bottom
are
represented:
discharge, suspended solids (SS),
chlorophyll a (Chla: µg.L-1),
nutrient concentrations (nitrate
and ammonium: mgN.L-1; total
phosphorus
and
phosphates
-1
mgP.L ;
dissolved
silica:
-1
mgSiO2.L ).
2004
-1
2003
3
Discharge, m .s
-1
4000
10
0
10
0
165
Figure 6.8 illustrates how the Seneque/Riverstrahler software can take into account the effect
of a large dam on the concentration of particulate material in river water. Two calculated
longitudinal profiles of suspended solid in the Da river are shown, respectively in March (dry
season) and July (wet season), two periods for which the level in suspended solids strongly
contrasts (250 mg.L-1 against 1150 mg.L-1 respectively). Downstream from the dam, a huge
abatement of suspended solids concentrations is shown, up to 85 % during high water flows
but hardly lower in low waters. This pattern is confirmed by the seasonal variations of
suspended matter and total phosphorus at two stations upstream and downstream the dam
(Figure 6.8); the comparison with the available observed data shows that the model is able to
correctly predict the effect of the presence of a dam in the drainage network.
Two examples of a cartographic representation of the calculated results are presented here
(Figure 6.9). Figure 6.9.a illustrates the spring algal development in the drainage network as
calculated by the model. The stronger algal growth in the Thao River is well illustrated.
Clearly, this algal development is already initiated in the upstream sector of the river. Figure
6.9 b shows the distribution of total phosphorus and the role of the two dams (Hoa Binh on
the Da River and Thac Ba on the Lo river) in the abatement of total phosphorus concentration.
The model can also be used to estimate the total flux of nutrient delivery at the outlet of the
different sub-basins (Table 6.7). As far as suspended matter is concerned, the agreement with
our previous estimate (see chapter 3) is acceptable for the Da, but the models, as already
discussed, severely overestimates the suspended solid load of the Lo while it underestimates
that of the Thao River. Regarding nitrogen and phosphorus delivery, the model estimates
differ by less than a factor 2 from those resulting from our budgeting approach (Le Thi
Phuong Quynh et al., 2005), (see chapter 5).
The model provides also an estimate of the silica delivery, which can be used to calculate the
molar Si/N, Si/P and N/P ratios at the outlet of the systems.
166
a)
Suspended solid, mg.L
-1
1500
Mar sim.
Jul. sim.
1000
500
0
0
50
100
150
200
km
250
-1
1500
upstream the dam
1000
500
0
-1
b)
1500
downstream the dam
1000
sim.
Obs.
500
0
J F MAM J J AS ON D
1.2
Ptot, mg.L
Ptot, mg.L
-1
-1
1.2
0.8
0.4
0.0
0.8
0.4
0.0
J F MAM J J ASOND
Figure 6.8: a) Longitudinal variations of suspended solid concentrations along the Da river
branch in March and in July in 2003; b) Seasonal variations, in 2003, of the simulations (sim.)
and observations (obs.) of suspended solid (mgSS.L-1) and total phosphorus (Ptot, mgP.L-1)
concentrations obtained at a station upstream (left) and downstream (right) the Hoa Binh dam
on the Da River. (The observed suspended solid concentration upstream from the dam are
mean monthly values reported by Nguyen Viet Pho et al. (2003) for the Lai Chau station in
the period 1961-1989).
167
Early April 2003
<2 µgChla/l
>2 µgChla/l
>4 µgChla/l
>6 µgChla/l
<0.2 mgP/l
>0.2 mgP/l
>0.4 mgP/l
>0.6 mgP/l
Chlorophyll a
Total Phosphorus
Figure 6.9: Cartographic representation of the geographical distribution of chlorophyll a and
total phosphorus concentration in the drainage network of the Red River system at the
beginning of April 2003, as provided by the Seneque/Riverstrahler model. The higher
(although limited) algal development, as well as the higher concentration of total phosphorus
in the Thao river than in the other tributaries is quite apparent, as well as the role of the dams
(open circle) in reducing the particulate phosphorus concentration.
Table 6.7: Nutrient flux (N: nitrogen; P: phosphorus; Si: silica) calculated by the model (Sim)
at the outlet of the Da, Lo, Thao and Hong Rivers at the stati
ons Hoa Binh, Vu Quang, Yen Bai and Hanoi respectively, in 2003.
Total P
Total N
3
10 tons N.y
-1
3
Dissolved Si
10 tons P.y
-1
3
10 tons Si.y
-1
Suspended solid
106 tons SS.y -1
Sim.
Ref.*
Sim.
Ref.*
Sim
Sim.
Ref.*
Da
61
38
9.3
3.5
274
7.4
5.5
Lo
55
32
24.0**
5.1
128
35.0**
7.9
Thao
15
22.5
7.9
8.3
57
8.4
20
Hong Hanoi
146
-
48.2
-
475
58.5
40
Ref.*: data provided in Le Thi Phuong Quynh et al., 2005; Silica fluxes not calculated in Le Thi Phuong Quynh
et al., 2005
** note the overestimations by the model (cf. text) of total phosphorus and suspended solids fluxes for the Lo
River.
168
The molar N/P ratios at the outlet of the three sub-basins Thao, Lo, Da and in the main branch
of the Red River at Hanoi are respectively 1.9, 2.3, 6.6 and 3.0. They are much lower than
the Redfield ratio of 16 characterizing the requirement of algal growth (Redfield et al., 1963).
The molar Si/N ratios at the same stations, 3.7, 2.3, 4.5 and 3.2 respectively, are much higher
than 1 (Conley et al., 1993; Billen and Garnier, 1997), indicating that silica is largely in
excess over the requirements of diatoms. We already arrived to this conclusion of great
importance (Garnier and Billen, 2002; Le Thi Phuong Quynh et al., 2005), as silica limitation
is often at the origin of harmful algal blooms at the coastal zone (see Cugier et al., 2005 and
included references).
Finally, it must be stressed that the model, in its present development stage is not able to
focus on the small urban rivers in the Delta area where water environment is the most
seriously polluted, as mentioned in the chapter 4. In fact, the Seneque/Riverstrahler model
does not here consider distributaries; it is only here able to model one main branch of the Red
River in the Delta area, which in fact limits seriously its ability to represent the real situation
of water pollution in this, very populated area. One of our aims in the next future, is to be able
to adapt the Seneque/Riverstrahler model to pursue the modelling work down to the coastal
zone through the complex drainage network of the Red River delta.
6.9. References
Billen, G., Garnier, J. and Hanset, P. (1994). Modelling phytoplankton development in whole
drainage networks: The RIVERSTRAHLER model applied to the Seine river system.
Hydrobiologia, 289: 119-137.
Billen, G. and Garnier, J. (1997). The Phison River plume: coastal eutrophication in response
to change in land use and water management in the watershed, Aquat. Microb Ecol 13:
3-17.
Billen, G. and Garnier, J. (1999). Nitrogen transfer through the Seine drainage network: a
budget based on the application of the RIVERSTRAHLER Model. Hydrobiologia, 410:
139-150.
Chinadata, (1998). Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, China Statistical
Chinadata, (2000). Statistical yearbook of Yunnan, Vol. 1999, Vol. 2000; China Statistical
Conley, D.J., Claire, L. S and Stoermer, E. F. (1993). Modification of the biogeochemical
cycle of silica with eutrophication. Marine Ecology progress series, Published
November 4, 101: 179-192.
169
Cugier, J., Billen, G., Guillaud, J.F., Garnier, J., and Memesguen, A. (2005). Modelling the
nutrient loading. Journal of Hydrology. Volume 304, issues 1-4: 381-396.
DeFries, R., Hansen, M., Townshend, J.R.G. and Sohlberg, R. (1998). Global land cover
classifications at 8 km spatial resolution: The use of training data derived from Landsat
imagery in decision tree classifiers, International Journal of Remote Sensing; 19(16):
3141-3168.
Dürr, H.H. (2003). Towards a typology of global river systems: some concepts and examples
at medium resolution. Thèse de doctorat, Université Paris VI - Pierre et Marie Curie.
732pp.
Garnier, J., Billen, G. and Coste, M. (1995). Seasonal succession of diatoms and
chlorophyecae in the drainage network of the River Seine : Observations and modelling.
Limnology. and Oceanography, 40: 750-765.
Garnier, J., Billen, G. and Palfner, L. (1999). Understanding the oxygen budget and related
ecological processes in the river Mosel: the RIVERSTRAHLER approach. Hydrobiologia
410: 151-166.
Garnier, J., Billen, G., Sanchez, N. and Leporcq, B. (2000). "Ecological functioning of the
Marne dam (upper Seine basin, France)." Regulated rivers: Research & Management,
Regul. Rivers: Res. Mgmt. 16: 51-71.
Garnier, J., Billen, G., Hannon, E., Fonbonne, S., Videnina, Y. and Soulie, M. (2002a).
Modeling transfer and retention of nutrients in the drainage network of the Danube River.
Estuarine, Coastal and. Shelf Science, 54: 285-308.
Garnier, J. and Billen, G. (2002b). The Riverstrahler modelling approach applied to a tropical
case study (The Red –Hong- River, Vietnam): nutrient transfer and impact on the
Coastal Zone. SCOPE, Coll. Mar. Res. W., 12: 51-65
Garnier, J., Nemery, J., Billen, G., and Thery, S. (2005). Nutrient dynamics and control of
eutrophication in the Marne River system: modelling the role of exchangeable
phosphorus. Journal of Hydrology, Volume 304, issues 1-4: 397-412.
Greenlee, D.D. (1987). Raster and Vector Processing for Scanned
Photogrammetric Engineering and Remote Sensing. Vol. 53(10): 1383-1387.
Linework,
Hansen, M., DeFries, R., Townshend, J. R. G. and Sohlberg, R. (2000). Global land cover
classification at 1km resolution using a decision tree classifier, International Journal of
Remote Sensing. 21: 1331-1365.
IMH, (1997-2004). Journal of Meteo-hydrology, Institute of Meteo-Hydrology in Vietnam,
Hanoi. (Monthly Journal during the periods of from 1997 to 2004).
Jenson, S.K. and Domingue, J.O. (1988). Extracting Topographic Structure from Digital
Elevation Data for Geographic Information System Analysis, Photogrammetric
Engineering and Remote Sensing. Vol. 4(11): 1593-1600.
170
Global Biogeochemical cycles.Vol 19, GB2022, doi 10.1029/2004GB002405.
Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S. and Chau, Van Minh (submitted).
Hydrological regime and suspended matter flux of the Red River System (Vietnam):
Observations and modelling. Journal of Hydrology.
Meybeck, M. (1986). Composition chimique naturelle des ruisseaux non pollués en France Sci. Geol.Bull. , 39, 3-77.
Meybeck, M. (1987). Global chemical weathering of superficial rocks estimated from river
dissolved loads. American Journal of Science, 287: 401-428.
MONRE, (1997-2004). Vietnamese Ministry of Environment and Natural Resources. Report
annual on hydrological observation in Vietnam, Hanoi.
MOSTE, (1997). Vietnamese general statistics officer, Ministry of Science, Technology and
Environment of Vietnam, general statistics editor, Hanoi. 550 pp.
Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan (2003). Water resources in Vietnam.
Institute of Meteo-Hydrology. Agricultural Editor. Hanoi
Redfield, A. C., Ketchum, B.H. and Richards, F.A. (1963). The influence of organisms on the
composition of sea-water.In M. N. Hill (ed.), The Sea, John Wiley and Sons, New
York, p. 12-37.
Ruelland, D. (2004). SENEQUE, a GIS software to evaluate water quality. Hermès-Lavoisier
Ed., Revue Internationale de Géomatique, vol. 14(1): 97-117.
Ruelland, D., Billen, G., Brunstein, D. and Garnier, J. (in prep). SENEQUE 3A GIS interface
to the RIVERSTRAHLER model of the biogeochemical functioning of river systems. To
be submitted to Ecological Modelling
Tarboton, D.G., Bras, R.L. and Rodriguez-Iturbe, I. (1991). On the Extraction of Channel
Networks from Digital Elevation Data, Hydrological Processes. Vol. 5 : 81-100.
Turc, L. (1961). Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle,
Ann. Agron., 12 (I), I 3-49
171
172
Exploring future trends of nutrient transfers
CHAPTER 7
The material presented above, as well as the different modelling tools developed in the scope
of this work, now offer the possibility of exploring a number of prospective scenarios
concerning the future biogeochemical functioning of the Red river system.
We have
restricted ourselves to two three important aspects: (1) the impoundment of new large dams
on the drainage network of the Red River; (2) the increase of the population and its degree of
urbanization; (3) the changes in land use and the intensification of agricultural practices. The
Seneque/Riverstrahler model implemented on the Red River basin will allow us to predict the
results of possible future changes in these three aspects on the overall water quality and
nutrient delivery of the river system.
7.1. Impacts of new dams constructed in the Red River basin
As mentioned in chapter 3, the construction of two large dams, in addition to the already
existing Hoa Binh and Thac Ba dams, is planned for the next decade. The Son La dam, with a
volume of 9.3-25.5 109 m3 will be constructed on the upper course of the Da river, upstream
from the Hoa Binh reservoir. The Dai Thi dam, with a volume of 0.5-3 109 m3 will be
constructed on the Gam river, a tributary of the Lo river. We already calculated, with the
simplified approach described in chapter 3 that these dams would reduce by about 20% the
solid load of the Red River. As a consequence, the phosphorus loading should also be
significantly reduced.
We have run the Seneque/Riverstrahler model for two scenarios differing from the standard
validated scenario of the year 2003 in that (i) no dams at all are considered (scenario called
“1970”), or (ii) the four large dams are considered operating (scenario called”2050”).
Excepted for this aspect, all the other constraints (hydrology, land use, point sources of waste
water) were taken identical with those of the reference “2003” scenario. Figure 7.1 shows the
results of these simulations for suspended solid and total phosphorus concentrations at the
173
outlet of the Da and Lo rivers and in the main branch at Hanoï station. The effect of the Son
La dam is less apparent than that of the Dai Thi dam (compare sc 2003 and sc 2050), because
the Hoa Binh dam already reduced severely the suspended solid load of the Da river at its
outlet (compare sc 1970 and 2003).
The annual flux of suspended solid and total phosphorus delivery were calculated and given
in Table 7.1. The results show a clear decrease of both suspended solid and total phosphorus
fluxes at the outlet of the Da and Lo rivers, as well, as in the main branch, at Hanoï. In
particular the impoundment of the Dai Thi dam on the Lo river will result in 50% reduction of
the suspended solids flux. These conclusions are in agreement with those of other authors
(Nguyen Huu Khai and Nguyen Van Tuan, 2001; Pham Quang Son, 1998; Nguyen Viet Pho,
2003).
Table 7.1: Simulated fluxes of suspended solid and total phosphorus delivery at the outlet of
the Da and Lo rivers and at Hanoi station, calculated for the conditions of the year 2003,
without any dam (“1970”), with the two presently existing dams (“2003”) and with two
additional dams (“2050”).
174
Suspended solid
Total phosphorus
106 ton SS.y -1
103 ton P.y -1
1970
2003
2050
1970
2003
2050
Da
58.1
7.4
6.7
39.6
9.3
8.8
Lo
41.5
35
21.5
27.9
24
11.6
Hong Hanoi
113.8
58.5
44.5
81.6
48.2
39.7
Da R.
1.2
-1
1500
in 1970
in 2003
in 2050
TotP, mg.L
SS, mg.L
-1
2000
0.8
1000
0.4
500
0
0.0
1.2
2000
-1
1500
TotP, mg.L
SS, mg.L
-1
Lo R.
1000
500
0
0.8
0.4
0.0
1.2
2000
-1
TotP, mg.L
SS, mg.L
-1
Hong R.
1500
1000
500
0
0.8
0.4
0.0
Figure 7.1: simulation results of suspended solid (SS, mg.L-1) and total phosphorus (Ptot,
mgP.L-1) at the stations Hoa Binh (in the Da River), Vu Quang (in the Lo River) and Hanoi
(in the Hong River) in the scenario ‘1970’ (no dam at all), in 2003 (presence of Hoa Binh and
Thac Ba dams) and in the scenario ‘2050’ (presence of two more new dams) in the Red River
system.
175
Several studies have been carried out concerning the factors controlling the long term
demographic evolution of Vietnam. Hoang Xuyen (2000), on the basis of a detailed analysis
of the present demographic structure of the Vietnamese population and of the evolution of
birth, mortality and migration rates since the last 50 years; they conclude that the process of
demographic transition, characterized by a reduction of the mortality rate, followed by a
reduction of the birth rate, has been initiated in Vietnam, particularly in the North of the
country, since the mid 1950’ies, and entered its final stage after the end of the war in 1975.
This means that the population of the country should stabilize within about one generation.
The same author evaluates the population of Vietnam in 2020 to 100 millions inhabitants
(between 98 and 103 millions inhabitants). According to FAO statistics (FAO, 2004), the total
population in Vietnam increased from 27.4 106 in 1950 to 83.6 106 in 2005, and will reach
117.7 in 2050 (Figure 7.2). The Vietnamese population should thus increase by a factor 1.4
by 2050.
6
10 inhabitants
140
120
Urban
100
Rural
80
60
40
20
2050
2005
2000
1995
1990
1985
1980
1975
1970
1965
1960
1955
1950
0
Figure 7.2: Evolution of the total population in Vietnam (FAO database, 2004), as well as the
urban and rural components.
The rural population always occupies the largest proportion in the whole country (about 80%
in 1990s) (Figure 7.2), and this represents a typical characteristics of the social organisation in
Vietnam. The analysis of the population data base by villages in the Vietnamese Red River
176
basin shows indeed that 80% of the population lives in agglomerations of less than 10 000
inhabitants (Figure 7.3).
Urban population, however, has increased at a higher rate in the last decades than rural
population. The FAO figures for the whole country show an increase of rural population at a
rate of 1.1%.yr-1 over the last 5 years, while urban population has raised at the rate of 2.2%.yr1
in the same period (Fig. 7.2). Based on these figures and on the total population increase
forecasted in 2050 by FAO, we estimate that urban population in Vietnam will represent
about 40% and the fraction of rural population should be 60% by the year 2050.
cumulated popul., million inhab
12
Red River basin in Vietnam
10
8
6
Rural population
4
2
0
0
50000
100000
150000
200000
size of agglomerations, nb inhab
Figure 7.3: Cumulated population in the Vietnamese part of the Red River basin as a function
of agglomeration size in 1999. 80% of the population live in villages of less than 10 000
inhabitants.
Analysing the population figures for the provinces of the Red River Delta area and of the
Vietnamese North Mountains regions (roughly, the upstream Vietnamese basin of the Red
River), we arrived at rather similar results concerning the rate of increase of urban versus
rural population (Table 7.2).
Obviously, these rates of increase cannot be extrapolated over the next 50 years, as we know
that the total population will stabilize. Instead we made the hypothesis that the ratio between
177
the urban and the rural population growth rate will remain constant in the two regions
considered.
Table 7.2: a) Population census in North Vietnam (106) inhabitants in the period from 1997 to
2002; b) annual rate of population increase (%.yr-1) in North Vietnam in the period 19972002.
Data from MOSTE 1998-2003, ‘Delta’ gathers the data from the provinces Hanoi, Hai Phong, Ha Tay, Hai
Duong, Hung Yen, Nam Dinh, Bac Ninh, Thai Binh, Quang Ninh, Ninh Binh
The Mountainous region comprises the provinces of Ha Giang, Cao Bang, Lao Cai, Bac Can, Lang Son, Tuyen
Quang, Yen Bai, Thai Nguyen, Phu Tho, Vinh Phuc, Bac Giang, Lai Chau, Son La and Hoa Binh.
a)
Region
Red River
Delta
Mountainous
region
1997
rural urban
1998
rural urban
1999
rural urban
2000
rural urban
2001
rural urban
2002
rural urban
12.39
3.43
12.47
3.58
12.53
3.69
12.59
3.79
12.67
3.92
12.75
4.06
9.47
1.43
9.57
1.47
9.69
1.49
9.76
1.55
9.84
1.60
9.94
1.57
b)
Rural
0.5
0.8
Regions
Delta
Mountains
Urban
3.0
2.4
Assuming in addition that the total population of both region will be multiplied by a factor of
1.4 by 2050 (similarly with the factor of increase of the total Vietnamese population), we
were able to calculate, for the two regions considered, the factor of increase of rural and urban
population (urbf and rurf). The reasoning is as follows:
For each region (Mountains and Delta respectively):
totP2050 = 1.4 totP2003
hence
urbP2003. urbf + rurP2003. rurf
= 1.4 (urbP2003 + rurP2003)
if urbf and rurf are respectively the increase factors of rural and urban population by 2050.
Considering that urbf/rurf = ur (where ur is the observed ratio between the urban and rural
population growth rate in the region considered, see table 7.2), the following relationships can
be established:
urbf = 1.4 [totP2003 / (urbP2003 + rurP2003/ur) ]
178
rurf = urbf / ur
The application of the above relationships to the two regions of the Red River Basin yields the
figures of table 7.3.
Table 7.3: factor of increase of urban (Urbf) and rural (Rurf) population of the Red River
basin by 2050
Regions
Delta
Mountains
Urbf
3.9
3.2
Rurf
0.6
1.1
These figures allowed us to construct the spatial distribution of the population of the
Vietnamese Red River basin by 2050. For the Chinese part of the basin, that we considered
mostly rural, we applied the same figure as for the mountainous Vietnamese part of the basin.
According to these calculations, the total population in the Red River basin considered in the
Seneque data base will increase from 16 106 inhabitants to 23 106 inhabitants by 2050. Figure
7.4 compares the spatial distribution of the present and the future population.
> 100 000 inhab.
< 10 000 inhab.
2003
2050
179
Figure 7.4: Distribution of the individual agglomerations of the Red River basin in the 2003
and 2050 scenarios. An ellipse indicate the zone of major changes in the upper delta.
Essentially, the 2050 scenario tend to reinforce the major trends of the spatial distribution,
already apparent in the present situation, namely an accentuated concentration of the
population in the delta area, with a few centres of population agglomeration in the upper Lo
and Da river sub-basins and all along the Thao River (figure 7.4).
The increase of the population requires a parallel evolution of agricultural production. The
recent trends of land use evolution in North Vietnam are the stabilization of forest cover and
the increase in urban area. The latter is probably mostly increasing at the expense of
agricultural land, particularly paddy rice fields, the total surface of which is decreasing in the
recent years. This evolution is made possible owing to the spectacular and continuing
increase in the productivity of Vietnamese agriculture observed since the last decades in terms
of yield per unit cultivated surface. This increase is largely due to the use of increasing
amounts of chemical fertilizers (Figure 7. 5).
300
kg fertiliser.ha
-1
in Vietnam
262
200
100
0
1950
14
1960
1970
1980
1990
2000
2010
Figure 7.5: Evolution of fertilizer application (N and P) in Vietnam (FAO database, 2002)
Accordingly, we established a “2050” land use GIS file, by increasing the urban area, for each
elementary watershed unit of the ‘2003’ land use file, at the expense of paddy rice fields or
other agricultural surface if necessary.
180
In order to account for the intensification of agricultural practices, we also assumed that the
nitrate concentration resulting from soil leaching of ‘dry’ agricultural soils (all crops excepted
paddy rice) would have reached the levels typically observed in West European countries (10
mgN/l), and we considered a 25% increase of the export of total phosphorus from agricultural
soils.
7.4.
Prospective simulation at the 50 year horizon
The Seneque/Riverstrahler model has been run for an hypothetical “2050” scenario
characterized by the hydrological conditions of the year 2003, the presence of 4 large dams, as
discussed in §7.1, a 40% increase of population distributed between rural and urban centres as
discussed in §7.2, land use and agricultural practices as discussed in §7.3. The results show
the trends of the changes to be expected from this “business as usual” scenario of the future
evolution of the human activity in the Red River basin, compared to the present “2003”
situation (Figure 7.5). Note that we have not considered any difference in wastewater
treatment practices in 2050 compared to 2003, i.e. the same hypothesis concerning the release
of wastewater in urban (no treatment and total release to surface waters) and rural areas (75%
recycling in agriculture) has been made.
The results show a very important increase in nitrate and ammonium contamination of the
Red River, while the level of phosphorus contamination remains nearly the same.
Apparently, the retention of phosphorus by the two additional dams counterbalance to a large
extend the increased release of phosphorus by agricultural soils and human population.
The calculated fluxes of nutrient delivery by the Red River and its main tributaries show the
same trends (Table 7.4). Nitrogen fluxes will be considerably increased at the outlet of the
Thao, Da, Lo rivers and in the main branch of the Hong River, while phosphorus flux at the
outlet of Da and Lo rivers tends to decrease. Only at the outlet of the Thao basin, the most
populated and free of dams basin, phosphorus flux increased in 2050 with respect to 2003.
The resulting total phosphorus flux at Hanoi station is nearly unchanged. Silica fluxes are also
predicted to remain essentially unchanged in response to the 2050 scenario.
The resulting nutrient ratios obviously reflect these trends (Table 7.5). A clear increase of the
N/P ratios is predicted for the 2050 scenario with respect to the 2003 situation, along with a
clear decrease of the Si/N ratios at the outlet of the three rivers Thao, Lo and Da and in the
main branch Hong River.
181
4
in 2050
in 2003
2
2
0
0
J F M A M J J A S ON D
1.0
N-NH4, mg.L
N-NH4, mg.L
-1
-1
1.0
0.5
0.5
0.0
0.0
J F MAM J J A S ON D
1.6
-1
1.2
1.2
Ptot, mg.L
-1
1.6
Ptot, mg.L
Hong R.
-1
Thao R.
N-NO3, mg.L
N-NO3, mg.L-1
4
0.8
0.8
0.4
0.4
0.0
0.0
J F M AMJ J A SON D
0.3
P-PO4, mg.L
P-PO4, mg.L
-1
-1
0.3
0.2
0.2
0.1
0.1
0.0
0.0
J F M AM J J A SON D
Figure 7.5: Simulation results obtained at Yen Bai station in the Thao River and at Hanoi
station for the ‘2050’ and the ‘reference 2003’ scenarios.
182
Table 7.4: Calculated nutrient fluxes (N, P and Si) at the stations Yen Bai (Thao river), Vu
Quang (Lo River), Hoa Binh (Da River) and Hanoi (Hong River) in the year 2003 and in the
‘2050’ scenario with the same hydrological conditions
Sub-basin
Total N
3
Total P
10 tons N.y
-1
3
Dissolved Si
10 tons P.y
-1
3
-1
10 tons Si.y
in 2003
in 2050
in 2003
in 2050
in 2003
in 2050
Da
61
70
9.3
9
274
279
Lo
55
104
24
18
128
128
Thao
15
26
7.9
9.5
57
57
Hong Hanoi
146
234
48
49
475
479
Table 7.5: Molar nutrient ratio of fluxes delivered at the outlet of the sub-basins and the main
branch of the Red River as calculated by the model for the reference year 2003 and in the
‘2050’ scenario with the same hydrological conditions
Thao
Lo
Da
Hong Hanoi
in 2003
1.9
2.3
6.6
3.0
N/P
in 2050
2.8
5.8
7.8
4.8
Si/N
in 2003
in 2050
3.7
2.2
2.3
1.2
4.5
4.0
3.3
2.1
The Seneque/Riverstrahler model implemented on the Red River basin enables the diagnostic
of the N:P:Si nutrient balance, which is the key for the control of freshwater and coastal
marine eutrophication problems. The implementation of such a water quality model at the
regional scale can offer an excellent framework for initiating and developing the dialogue,
both between scientists of different disciplines but also between scientists, decision-makers
and the public. It will also enable to very clearly point out the gaps subsisting in our
understanding of the system, thus indicating the need for further research.
Because of the role of water in the history of human development, the scale of the catchment
of large rivers often represents a pertinent scale regarding major environmental land planning
issues. On the other hand, downscaling should be possible with water quality models, in order
to use them operationally for local management of water quality. A coupling of this model to
other models of similar conceptual approach at more local scale, as the one describing the
functioning of polluted urban rivers in the delta (Trinh Anh Duc, 2003), could be currently
used as a support for Water Policy in the future .
183
7.5. References
FAO database, 2005. FAO Statistical Databases, available at http://faostat.fao.org/faostat.
Hoang Xuyen, 2000. La Transition démographique. In Population et Développement au
Vietnam. Patrick Gubry, ed. Karthala/Ceped Paris. p 61-82.
MOSTE, 1998 – 2003. Vietnamese general statistics officer, Annual rapport of Ministry of
Science, Technology and Environment of Vietnam, general statistics editor, Hanoi.
Nguyen, Huu Khai and Nguyen, Van Tuan, 2001. Geography and Hydrology in Vietnam.
Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan, 2003. Water resources in Vietnam.
Vietnamese Institute of Meteo-hydrologie. Agicultural Editor (in Vietnamese).
Pham, Quang Son, 1998. Fundamental characteristics of the Red River bed evolution.
Proceedings of International Conference on Economic development and environmental
protection of the Yuan-Red River watershed, Hanoi 4-5 March.
184
General Conclusions and Perspectives
The Red River system (a tropical river with its watershed of 156 448 km², mainly laid in
Vietnam and China) has been considerably influenced by human activities. While previous
studies gave mainly emphasis on the delta area of the Red River, addressing issues like flood
disasters, irrigation systems, etc., or on the upstream sectors, addressing the questions of
forest management, erosion, reservoir construction, etc., our work, for the first time, addresses
other important aspects at the scale of the whole basin.
1. The hydrological regime, characterized by irregular flow including floods in July and
August, has been studied in the period from 1997 to 2004. The Hydrostrahler model, based on
a simple description of the rain-discharge relationship, has been utilized for modelling the
discharge at the outlet of the three sub-basins Da, Lo, Thao of the Red River. The results
show that discharge simulation with a Nash criterion around 0.7 can be obtained using only
the limited number of meteorological stations available in the basin. Closely related with
hydrological regime, the suspended load carried by the Red River has also been investigated.
The results reveal that the Hoa Binh and the Thac Ba reservoirs have a major influence on
suspended solid concentrations in the Red River system. The suspended load decreased from
100-170 106 t.yr-1 to 40 106 t.yr-1 since the impoundment of the two reservoirs. With the
planned construction of two additional reservoirs (Son La and Dai Thi) in 2010, the
suspended load of the Red River in future will be further reduced by 20%.
2. A two-year field survey, at monthly frequency, has been organised in order to collect
data on water quality at the outlet of the major tributaries and in the main branch of the Red
River, as well as in some polluted rivers of the Hanoi region. This allowed defining the
general level of nutrient (N, P, Si) concentrations in the drainage network of this large river
system. It also allowed demonstrating the low level of algal growth in the major rivers as
well as in the large reservoirs. To our knowledge, water quality data did not exist up to now in
the upstream basin of the Red River, and could be used as reference in the future.
3. The degree of human-induced alteration of the nitrogen and phosphorus cycles at the
scale of a sub-tropical watershed was investigated by budgeting N, P, within the 4 main subbasins (Da, Lo, Thao and Delta) of the Red River system, differing in population density (by a
185
factor of 10), land use and agricultural practices. In terms of agricultural production, on the
one hand and consumption of food and feed on the other, the upstream sub-basins are
autotrophic systems (they produced more than they consumed), while the delta is, at the
opposite, a heterotrophic one. Great losses of nitrogen are attributable to denitrification in rice
paddy fields, those of phosphorus being caused by erosion. In stream elimination of nitrogen
and retention of phosphorus are the highest in the Da and Lo sub-basins which have large
reservoirs in their downstream course. The total specific delivery estimated at the outlet of
the whole Red River System is 855 kgN.km-².yr-1 and 325 kgP.km-².yr-1. Nitrogen rather than
phosphorus seems to be the potential limiting factor of algal growth in the plume of the Red
River, and further in the South Chinese Sea (Tonkin Bay).
4.
For assessing the link between human activity in the watershed and water quality of
the river system, the Seneque/Riverstrahler model was successfully applied for the first time
to a tropical river system. A GIS data base has been assembled at the scale of the whole basin,
with layers documenting geomorphology, lithology, meteorology, land-use and agriculture,
population, industrial wastewater release, etc. for estimating the role of natural and
anthropogenic factors in the watershed on the water quality and biogeochemical functioning
of the whole river system. The results could be validated, by a rather good agreement
between the modeling results and the observed data of water quality, at the outlet of the three
sub-basins and in the main branch of the Red River during the years 2003 and 2004. Beside
the modeling tool, the GIS data base has allowed to check the coherence of existing data and
the synthesize them. This data base can be permanently enriched with new data, allowing a
better forcing of the model and/or a stronger validation.
5. The model has allowed to explore a variety of scenarios describing potential changes
in the watershed (climate, hydrology, land use and agricultural practices, population increase,
wastewater treatment policy), in terms of river water quality and overall export of nutrients.
As an example, with the hypothesis of a climate change in the 2080s (increase of 10% of
rainfall data and 3°C increase of temperature), the model predicts an increase of about 20% of
the suspended matter loading of the Red River (from 40 to 48 106 t.yr-1) with respect to the
conditions of the period 1997-2004. Some other changes in the Red River basin in 2050 such
as increasing population, planning of new large reservoirs and changes in land use in the next
50 years were tested to obtain the future nutrient levels of the Red River. These results are
186
expected to serve as a guide for planning environmental decisions at both regional and local
scales.
Beside all interesting results mentioned above, this work has also its restrictions. Regarding
to the results obtained, it is possible to note that the water quality of the Red River system is
not seriously polluted, especially in the upstream of the Red river basin. The most polluted
rivers are the ones mainly located in the Delta area, especially in the Hanoi city area where
rivers are considered as waste water collectors. This work has not focused on modeling these
small rivers, although we analyzed their quality level, but was complementary to a similar
modelling approach on one of the urban river (The Nhue River). Therefore, several
perspectives are fully open for the nearby future:
1. We can remind here, that this work was undertaken in the framework of the ESPOIR
project aiming at identifying the water quality controls and at developing new processes for
water treatment. As already mentioned in the Introduction, although this programme focused
on the study on water pollution and water treatment of urban rivers surrounding Hanoi, i.e. the
Nhue-Tolich river system located in the Red River delta, a special interest was given to the
upstream drainage network of the Red River, the Nhue river being one of diverted branched
of the Red River, upstream Hanoï. The Nhue receives directly the Tolich River draining
Hanoi (about 3.5 million inhabitants), being therefore seriously polluted by the domestic and
industrial wastewater from Hanoi city and also by agricultural activities (irrigation in rice
field and vegetation culture) and aquaculture (fish and crustaceans production). As the Nhue
River is supplied by the major branch of the Red River through the Lien Mac dam,
immediately upstream of Hanoi city, it could now become possible to explore how the
management of the discharge of water from the Red River to the Nhue River could be used to
improve the water quality of the Nhue River. A dialog between the Red River model
developed in this work, with the Nhue/Tolich Rivers model developed by Duc’s thesis (2003,
see reference in chapter 7) could be useful in this context.
2. In the framework of this thesis, the complex drainage network characterizing the Red
River delta area, with a lot of distributaries and irrigation channels, connected with the Thai
Binh River system, has not been examined in details. In the next step, we wish to be able to
adapt the model in terms of hydrology and water quality to be able to pursue the modelling
work down to the coastal zone. An application has just been submitted for further
187
cooperation between France and Vietnam to investigate a major branch of the delta, also
largely polluted, the Day River.
3. The Seneque/Riverstrahler model, which has now shown its capability to represent
water quality of a tropical river system, although further refinements are always necessary,
should well be utilized for addressing some other important environmental issues in other
region of Vietnam. The management of water quality in the lagoon of the Huong (Parfum)
River in Hue city is a possible example.
188
Table of contents
Introduction
1
CHAPTER 1: Site description and major issues
9
1.2 Geomorphology
1.4 Hydrology
1.4.1 Hydrology in Vietnam
1.4.1.1Surface water
1.4.1.2 Ground water
1.4.2 Hydrology of the Red River
1.4.1.1 Drainage density
1.4.1.2 Water flow
1.4.1.3 Reservoirs
1.5.1 General social-economical context
1.5.1.1 Change in land covers
1.5.1.2 Increase of fertilizers utilization
1.5.1.3 Increase of population and of urbanisation process
1.5.1.4 Increase of industrial releases
1.5.2 Impacts on the water quality
1.5.2.1 Decline of surface water quality
1.5.2.2 Increasing the water pollution in the delta and the coastal zone
1.6 References
9
11
13
16
16
16
17
18
18
19
19
20
20
20
21
22
23
24
24
24
25
CHAPTER 2: General approach and methodology
29
2.1 Modelling the quality of the Red River hydrographic network
2.1.1 What is a model?
2.1.2 Some model definitions in the context of modelling
2.1.3 The ecological functioning of hydrographic networks: RIVERSTRAHLER
model
2.1.3.1 General Principles
2.1.3.2 Hydrological model
2.1.3.3 Biogeochemical and ecological model: RIVE
2.1.3.4 Point sources and non-point sources
2.2 Experimental works
2.2.1 Sampling campaigns
2.2.1.1 Monthly sampling in the sub-basin and in the main branch of the Red
30
30
31
33
33
35
38
41
42
42
42
189
River
2.2.1.2 Sampling campaigns for non point source evaluation
2.2.1.3 Sampling campaigns for point source evaluation
2.2.2. In-situ measurements and sample analyses
2.2.2.1 Measurements of physical-chemical variables
2.2.2.2 Filtration and preservation of samples in the laboratory
2.2.2.3 Analyses of samples
2.3.1 Nutrient cycling in the soils system
2.3.2 Nutrient cycling in the hydrosystem
2.4 References
44
44
46
46
47
47
48
49
50
50
CHAPTER 3: Hydrological regime and suspended load:
observation and modelling
57
3.1 Introduction
3.2 General characteristics of the Red River basin
3.2.1 Geomorphology
3.2.2 Meteorology
3.2.3 Population and land use
3.2.4 Dams and discharge regulation
3.3 Hydrological regime of the Red River and its affluents
3.3.1 Total and specific discharge of the different sub-basins
3.3.2 Modelling the rain-discharge relationship
3.4 Suspended matter loading of the Red River and its tributaries
3.4.1. Total and specific suspended load
3.4.2. Seasonal and long term variations of suspended load
3.4.3. Modelling the suspended load
3.4.4. Relationship between suspended solid and phosphorus transport
3.5 Future scenarios of suspended matter loading
3.5.1. Effects of planned dams
3.5.2. Effects of climate change
3.6 Conclusions
3.7 References
58
59
59
62
64
65
66
66
68
74
74
76
78
81
82
82
82
82
83
CHAPTER 4: Water quality of the Red River
89
4.1 Discharge variations
4.2 Physical-chemical variables
4.2.1 Temperature and conductivity
4.2.2 Suspended matter and dissolved oxygen
4.3 General pattern of nutrients
4.3.1 Inter-comparison of nutrient analyses by two laboratories
89
90
90
90
94
94
190
6.6 Land use and non-point sources of nutrients
6.7 Wastewater point sources
6.8 Validation
6.9 References
CHAPTER 7: Exploring future trends of nutrient transfers
7.1 Impacts of new dams constructed in the Red River basin
7.4 Prospective simulation at the 50 years horizon
7.5 References
156
158
160
169
173
173
176
180
181
184
General conclusions and perspectives
185
Contents
Annex
189
193
192
Annex
Annex
Table A1: Water quality and discharge observations of the Thao River at Yen Bai station
Year
Date
NO3-N
mg.L-1
PO4-P
mg.L-1
DSi
mgSi.L-1
SS
mg.L-1
Discharge
m3.s-1
2003
20/01/03
0.27
0.000
5.44
109
378
2003
16/02/03
0.79
0.004
5.47
34
364
2003
15/03/03
0.26
0.003
4.81
55
204
2003
15/04/03
0.005
4.88
39
171
2003
15/05/03
0.04
0.66
0.004
6.00
257
188
2003
15/06/03
0.94
0.003
4.61
1620
1380
2003
15/07/03
0.36
0.007
3.64
2177
826
2003
15/08/03
0.63
0.009
5.25
1363
986
2003
15/09/03
0.42
0.018
5.25
315
1090
2003
16/10/03
0.49
0.014
6.43
283
586
2003
15/11/03
0.20
0.027
7.25
73
284
2003
15/12/03
0.54
0.009
7.20
71
291
2004
16/01/04
0.36
0.104
4.86
194
264
2004
15/02/04
0.25
0.078
5.23
528
176
2004
15/03//04
0.54
0.000
5.45
-
148
2004
14/04/04
0.54
0.026
4.30
1640
260
2004
10/05/04
0.42
0.055
4.40
2870
276
2004
23/06/04
0.47
0.023
5.13
3990
1130
2004
20/07/04
0.60
0.011
5.72
6000
1620
2004
18/08/04
0.64
0.011
6.17
1650
1920
2004
18/09/04
0.64
0.013
6.17
10000
999
2004
20/10/04
0.50
0.007
6.28
385
502
2004
17/11/04
0.36
0.141
7.28
1110
368
2004
05/12//04
0.64
0.137
6.03
756
275
DSi: dissolved silica concentration
SS: suspended solid concentration
193
Annex
Table A2: Water quality and discharge observations of the Da River at Hoa Binh station
Year
Date
NO3-N
mg.L-1
PO4-P
mg.L-1
DSi
mgSi.L-1
SS
mg.L-1
Discharge
m3.s-1
2003
20/01/03
0.18
0.001
2.69
2
899
2003
16/02/03
0.16
0.004
2.51
1
929
2003
15/03/03
0.25
0.001
3.51
1
766
2003
15/04/03
0.004
3.67
1
1030
2003
15/05/03
0.18
0.02
0.002
4.64
3
1310
2003
15/06/03
0.01
0.009
5.02
38
1630
2003
15/07/03
0.01
0.012
2.64
75
2660
2003
15/08/03
0.01
0.008
4.30
56
2610
2003
15/09/03
0.01
0.011
5.56
91
2160
2003
16/10/03
0.05
0.015
5.53
9
1270
2003
15/11/03
0.04
0.018
5.47
4
699
2003
15/12/03
0.43
0.001
5.68
3
757
2004
16/01/04
0.19
0.003
5.28
18
723
2004
15/02/04
0.06
0.016
4.16
8
667
2004
15/03//04
0.11
0.017
3.89
68
644
2004
14/04/04
0.13
0.003
4.28
46
898
2004
10/05/04
0.19
0.020
3.70
75
1330
2004
23/06/04
0.46
0.001
3.84
280
2400
2004
20/07/04
0.27
0.003
3.56
450
4150
2004
18/08/04
0.28
0.003
3.80
605
2010
2004
18/09/04
0.32
0.005
3.46
463
2110
2004
20/10/04
0.20
0.004
3.75
60
1590
2004
17/11/04
0.33
0.073
3.42
43
945
2004
05/12//04
0.44
0.073
2.90
210
824
194
Annex
Table A3: Water quality and discharge observation s of the Lo River at Vu Quang station
Year
Date
NO3-N
mg.L-1
PO4-P
mg.L-1
DSi
mgSi.L-1
SS
mg.L-1
Discharge
m3.s-1
2003
20/01/03
0.44
0.000
3.30
15
368
2003
16/02/03
0.14
0.007
4.25
6
379
2003
15/03/03
0.10
0.002
4.30
7
442
2003
15/04/03
0.005
4.69
5
408
2003
15/05/03
0.09
0.09
0.002
3.64
242
1190
2003
15/06/03
1.20
0.008
4.20
107
1080
2003
15/07/03
1.52
0.004
4.20
72
1400
2003
15/08/03
0.66
0.001
3.52
487
2000
2003
15/09/03
0.64
0.008
4.03
76
1100
2003
16/10/03
0.67
0.006
3.65
152
873
2003
15/11/03
0.19
0.016
4.76
16
297
2003
15/12/03
0.11
0.002
4.88
6
282
2004
16/01/04
0.20
0.006
3.91
66
257
2004
15/02/04
0.61
0.026
3.71
42
165
2004
15/03//04
0.48
0.078
4.31
40
217
2004
14/04/04
1.00
0.016
3.75
80
241
2004
10/05/04
0.66
0.055
4.27
4270
391
2004
23/06/04
0.64
0.003
4.23
337
1390
2004
20/07/04
0.14
0.003
2.82
2215
1320
2004
18/08/04
0.29
0.010
3.28
2215
1420
2004
18/09/04
0.41
0.006
3.62
177
754
2004
20/10/04
0.51
0.006
4.29
30
356
2004
17/11/04
0.61
0.150
4.60
90
292
2004
05/12//04
0.21
0.228
4.13
190
283
195
Annex
Table A4: Water quality and discharge observations of the Hong River at Son Tay station
Year
Date
NO3-N
mg.L-1
PO4-P
mg.L-1
DSi
mgSi.L-1
SS
mg.L-1
Discharge
m3.s-1
2003
20/01/03
0.17
0.003
2.42
93
1220
2003
16/02/03
0.08
0.004
2.44
33
1400
2003
15/03/03
0.12
0.010
2.71
60
1140
2003
15/04/03
0.018
4.39
35
1270
2003
15/05/03
0.07
0.05
0.003
4.68
52
1520
2003
15/06/03
0.38
0.002
5.20
307
3600
2003
15/07/03
0.44
0.007
5.22
142
6750
2003
15/08/03
0.23
0.007
5.35
145
5800
2003
15/09/03
0.52
0.020
5.18
278
5000
2003
16/10/03
0.53
0.015
5.99
141
2920
2003
15/11/03
0.19
0.030
6.20
41
1120
2003
15/12/03
0.07
0.003
4.27
31
1080
2004
16/01/04
0.21
0.068
4.15
204
1120
2004
15/02/04
0.20
0.026
3.21
251
994
2004
15/03//04
0.17
0.001
4.16
210
987
2004
14/04/04
0.28
0.019
4.01
520
1110
2004
10/05/04
0.29
0.039
4.78
1705
2020
2004
23/06/04
0.58
0.002
4.81
1925
4370
2004
20/07/04
0.29
0.002
5.10
2660
6670
2004
18/08/04
0.41
0.001
4.22
1855
5330
2004
18/09/04
0.29
0.001
4.72
2340
4390
2004
20/10/04
0.68
0.021
6.03
900
2520
2004
17/11/04
0.47
0.175
2.60
690
1780
2004
05/12//04
0.47
0.156
1.96
675
1790
196

THESE DE DOCTORAT Université Paris VI- Pierre

Transcription

Similar documents

Le Hong Phong Forest Vicinity

Please contact our Global Event Associate, Thu Thao Nguyen: E

INVITATION

In Vietnam, the lame walk and more.

Vietnam - IndoChina1911

Minutes of 2013 AGM

Page 1 TECTONOSTRATIGRAPHY AND NEOTECT ONIC

Returning to Vietnam

ANTI-VIETNAM WAR IMAGERY - Copy / Paste by Peter Pappas