Variability of the Atuel River streamflow annual cycle and its

Transcription

Variability of the Atuel River streamflow annual cycle and its
Variability of the Atuel River streamflow annual cycle and
its relationship with the tropospheric circulation
Diego Araneo1,2 and Eduardo Agosta3
1. IANIGLA, CONICET, Mendoza, Mendoza, Argentina.
2. ICB, UNCuyo, Mendoza, Mendoza, Argentina.
3. PEPACG , CONICET/UCA, Buenos Aires, Argentina.
OBJECTIVES
Río Atuel, draining from the eastern side of the Argentinean Central Andes, is the
most important irrigation source for the southern oases in the Province of Mendoza.
The present study determines the main atmospheric circulation patterns associated
with long-term variability in the annual streamflow cycle of Río Atuel.
METHODS
Principal Component Analysis (PCA) was applied to the annual hydrological cycle to
extract the main features in the annual regime.
Correlation fields between the PC loading timeseries and atmospheric-oceanic
variables were also examined to establish potential links between the variability in the
annual streamflow regime and tropospheric circulation.
1
DATA
Atmospheric-oceanic Variables
*Geopotential heights
*Wind vector
*Streamfunction
*Velocity potential
*Precipitable Water
*Precipitation rates
*Surface and air temperature
*Sea Surface Temperature
Data sources
*NCEP Reanalysis 1 dataset and the Extended Reconstructed Sea Surface
Temperature (ERSST) of NOAA’s Climate Diagnostics Center
*University of Delaware (last version)
*Streamflows Provided by the Argentinean Water Resource Administration.
Period
1958-2004
First PC loading timeserie (F1)
Streamflow annual cycles averaged for periods
of predominantly positive or negative PC1
loading values.
Av. Ruiz Leal s/n., Parque Gral. San Martín, (5500)
Mendoza, Argentina
C. Correo 330 – Tel-fax: 54-261-4286010
e-mail: daraneo@mendoza-conicet.gov.ar
The first (PC1) and second (PC2) Principal Components explain 48% and 22%
of the total variance in the streamflow annual cycle, respectively.
The direct (indirect)
mode of PC1
corresponds to annual
cycles characterized by
streamflows above
(below) the mean.
The PC2 represents
seasonal lags of the
streamflow peak shifted
to late-spring/earlysummer (direct mode) or
late-summer/earlyautumn (indirect mode).
Second PC loading timeserie (F2)
Streamflow annual cycles averaged for periods
of predominantly positive or negative PC2
loading values.
Positive (negative) values correspond to direct (indirect) modes.
PC1 CORRELATION FIELDS
The PC1 streamflow pattern is significantly related to the winter-spring (Jun-Nov) atmospheric-oceanic conditions. Shaded areas from light to dark colors correspond to significant values at 90, 95, 99 and 99.5 % levels respectively.
A
A
A
A
B
B
B
C
Correlation fields between F1 and Jun-Nov geopotential height and wind vector
at 500 hPa (A) and 850 hPa (B).
Correlation fields between F1 and Jun-Nov stream function at 250 hPa (A),
and Jun-Nov stream function and precipitable water at 1000 hPa (B). Vectors
represent non-divergent circulations related to the correlations.
Correlation fields between F1 and Jun-Nov potential velocity at 250
hPa (A), and Jun-Nov SST (B). Vectors represent non-rotational
circulations related to the correlations.
PC2 CORRELATION FIELDS
The PC2 streamflow pattern is significantly correlated with late-spring (November-December) and middle-to-late summer (January) atmospheric-oceanic conditions. Signals are different in both
seasons. Shaded areas from light to dark colors correspond to significant values at 90, 95, 99 and 99.5 % levels respectively.
LATE SPRING
B
A
A
A
B
B
C
Correlation fields between F2 and Nov-Dec geopotential height and
wind vector at 500 hPa (A) and 850 hPa (B).
Correlation fields between F2 and December stream function at 250 hPa (A),
and December stream function and precipitable water at 1000 hPa (B). Vectors
represent non-divergent circulations related to the correlations.
Correlation fields between F2 and December surface temperature (A),
SST (B) and 700 hPa air temperature (C).
MIDDLE-TO-LATE SUMMER
A
A
A
B
B
B
B
Correlation fields between F1 and Jun-Nov surface temperature (A),
precipitation rate (B) and 700 hPa air temperature (C).
CONCLUSIONS
The first (PC1) and second (PC2) Principal Components
explain 48% and 22% of the total variance in the annual
cycle, respectively. The direct (indirect) mode of PC1
corresponds to annual cycles characterized by streamflows
above (below) the mean. The PC2 represents seasonal
lags of the streamflow peak shifted to late spring (direct
mode) or early autumn (indirect mode).
The temporal evolution of the PC factor loadings reveals
that the PC1 pattern was predominantly indirect during the
periods 1917-1976 and 1988-1999, suggesting a tendency
to undergo extended periods of streamflow deficit with
relatively shorter (1906-1916 and 1977-1987) periods of
above-average runoff. In turn, the PC2 pattern is
predominantly indirect from 1906 to 1926 but direct from
1948 to 2002, suggesting a tendency to a dominance of
streamflow peaks during late spring during most of the
second half of the 20th century. Relationships between the
temporal evolution of the annual cycle of Río Atuel and
large-scale circulation indexes are discussed.
The PC1 streamflow pattern is significantly related to the
winter-spring tropospheric circulation conditions. Direct
(indirect) PC1 conditions are associated with significant
positive (negative) pressure anomalies over the
southeastern Pacific Ocean concurrent with warmer
(cooler) sea surface temperature (SST) anomalies in the
central equatorial Pacific. A northward (southward) shift of
the storm tracks is induced by the anomalies in
tropospheric circulation over the southeastern Pacific in
response to SST anomalies in the equatorial Pacific (El
Niño/La Niña events).
The PC2 streamflow pattern is significantly correlated with
late-spring and middle-to-late summer tropospheric
circulation. Streamflow peaks in the late spring (middle-tolate summer) are associated with anomalous warming
(cooling) over the subtropical Andes due to the
strengthening (weakening) of the South Atlantic anticyclone
and the enhanced (reduced) meridional circulation over
subtropical latitudes in November-December.
In addition, years with a tendency to experience streamflow
peaks in the late spring (the middle-to-late summer) also
occurred associated with anomalous cooling (warming) at
high levels in the atmosphere over the Rio Atuel basin
during January, related to a low pressure anomaly center
linked to a stationary quasi-zonal wave train from Australia
across the Pacific Ocean.
Acknowledgments
Correlation fields between F2 and January geopotential height and wind
vector at 500 hPa (A) and 850 hPa (B).
Correlation fields between F2 and January stream function at 250 hPa (A), and January stream
function and precipitable water at 1000 hPa (B). Vectors represent non-divergent circulations
associated to the correlations.
Correlation fields between F2 and January surface
temperature (A) and 700 hPa air temperature (B).
To the Subsecretaría de Recursos Hídricos de la Nación.
This work was funded by CONICET (PIP2010-439) and IAI (CRN2-047 ).