Ground Tilt Measurements During the 1973 Heimaey Eruption
Transcription
Ground Tilt Measurements During the 1973 Heimaey Eruption
vol. 80, No. JOURNAL OF GEOPHYSICAL RESEARCH 20 JULY IO, I9?5 Ground Tilt Measurements During the 1973 Heimaey Eruption LEo KRrsrJANssoN S.iqce Instittte, AniaeBitt o[ I.eldd, Relkjaoik, lceland IvaN SrMoN AND MARTTN L, CodEN A.thu. D. Lrtl., In -, Conbridse, M6sach6etts 02140 SvBINBJdRN BJdRNssoN Science Innituk. Uni'.rsitr oJ lc.land, Rerkjaaik, Iceland A conrinuous re@rd ofground tilt vs obrained at a site Ll5 kn ollhe active dater during the 'est l9?3 *uprion in Heimaey, Wstman lslands. An uplift mri€spondi.s lo a tilt valueotapproximalely 200 rrad pcrsist.d through lhe active period of the eruplion dnd was followed by rapid subsidene nedr the end ofvohanic acrivily. krge spasmodic lill surges sup.rimpos.d on the gen.ral uplift s€emed lo b. poorly co.r.laled with volcanic, or seismic, aclivity. They may have been due to localiz€d inliltrarion of shallow fractures by lava or sas€s, whereas tbe mor€ Bradual and moderatechang6 in lill levelr€sultcd fron the rise and decay ol pr.ssure in a dap{ea1ed source. A fissure e.uption began in Heimaey, Wesha, lslands (20'l5'w, 63'26'N), on January 23, l9?3, contracling in a few Tilt changes relative to the inilid level set on Februar/ 18 areshown in Fisure2. Teo half-day gaps in the r€.ord in mid- days to a single crater. The initial ourpur of lava and tephra May have be€n nll€d in as a best estinate. aided by Iosbook records. Po$tive values of tih in the figure indicate an uplift ol m'ls lEinaa.ior, 1974], and thetotal volum€ of €xtrusiv€s has been estimated as 250.10 m'. The intensity of the eruption decayed i. an app.oximately exhas been estimared at 100 ponential fashion, antl it end€d on Jun€ 26. In order to study ground movements during the eruption and. if possible to forecast chang€s in its progress a onecomponent tiltmete. was inslallen in the viciniiy ofth€ nssure on F€bruary 4, 1971, b@omins fully op€rational on February 18. Figure I shows the localion andorienlation ofthetillmeter with respecl to the fissure. The inslrument was set up on the concrete floor of a temporarily abandon€d farn building at Dalir, where it rcmain€d through February 1974. The build- the ground loward the €raler. It is seen thal in laleFebruary, March. and April there was a general uplift amounling 10 a total of 150-200 rrad, interrupled only by a brief downtilt lrom February 23 to 2? and two sharp uplift peaks on April4 and 8. The firsl hallofMay was marked by several larse tilt surses. The peak value of 550 prad was.eached on May l3 and followed by adrop in 3 days ing is on tho weslorn flank ofthe extinct volcano Helgafell and Sroximately 900 m west of the fissu.q the distance liom the c.ater was app.oximalely l.15 km. '-,uve I Tho liltmoter was of the diamagnetic suspension type desctibed by Sinon et al. [968], a model III'I-S-14 made by Arthur D. L;ttle, lnc. Its sensitivity was l-15 mvarad, and its output wasconlinually r€corded by n battery-ope.ated inpact recorder. The chart spe€d \vas nominally 2.5 cm/h, bu1 ir varied somewhst dep€nding on th€ battery voltag€. The l% linear range (50 !rad) ofthe tiltmeter was exterded by nanual r€lev€ling during the pe.iods of larS€ and rapid lilt changes. Whenever tho of-scale condition app€ared to beimminent, the tiltmeter l'as reset by a leveling microscrew on its l0-cm base. The ourput could also be read from a4-tim€s more sen' sitive pa.allel microemmere. on .he control panel. The tiltnetcr was calibrat€d before shipment and installed by one of us (M.L.C.): its calibradon was rech€ck€d before removal from Heima€y. Aftcr F€bruary l8 th€ liltmcter was att€nded to once or lwice a day. Due to nechanical problems the .ecorder was out ofopera.ioo fton February 26 to March 22 and from May L3 to 19;howev€r, wi.h lh€ daily meter readinss the record was esen.ially complet€. Fig. CopJrighr @ I975 by rhe Anerian Gmphysical Union. 2951 L Mdp of Hejmaey indi€ting th€ location of tilh.lcr (center of arow). volcanic fi$urc aid qare., ad de.p drill hole. KRIsTJ^NssoN ET AL,: GROUND TIL1 ME^SUREMENTS 2952 .> *.E :E 60 5g 5E ta E: --5o / ah t^ 6 E F5 KRBTJANSSoN Er GRoUND ^1.: TrLr ME^suReMENrs 2951 til litr 1r-il,: 6 APR. FiE r. 21 06 o4 02 F\rnple o. r lrne.er re.ord lo the ll0'prdd --l it 1 7 GMT APR. 1973 o lheperiodi.rori,tiol:ofC_oLndno'ene.rr,.c.le_ar'orJdrnngrr.e'Lar'or level. Another series of spasmodic surSes oc cured later in May and in earlyJune, thelast peak o1450!rad occurring on June I5. followed by a .apid drop lo 170 prad. Afl€r June 19 the till showed only minor fluctuat'ons atd stayed al a mcan value around 130 l.rad through Februa.y 1974. For the purpos€ ofcompari.S the lih changes *ith the intensily ot eruptive activity we ircluded in Figure 2 a record oiground movement tor the period frorn February lT lo June 13. This record was oblained from the peak'to'peak amplilude of the hofizortal acceleraliors recorded by the tilh.ter and measured by the widlh ofthe record trace; tho tikmele. had a nalural frequency of0.9 Hz and was nearly crilically damped. Simila. measurewas oblained lrom obse.veB'noles on !h€ o$ cillations of the panel meler; bolh eslimates compared quile *ell, and observers' noles were used to compleb rhe record when thecharlrecorderwasinoperative.Thestrongestground novemenl persisted during lale February 1o early April, a low possibly occuriing ir mid-Fcbrua.y. A peculiar lealure oflhe ground novement observed in the tilheter records is the occur.once of periodic varialions in amplilude, changing from unusually high to low values and back in I lr, hou6 (Figute 3). Similar periodjc variarions were observed in the seismic records fiom local Seophones at the corelpo,dirs dme periods (S. Bjdrnsson and P. Einarsson, unpublished manuscript. 1974). These periodic variations were obseNed occlsionall) in Aprilend May, when the erup' lion was slill in progress, bul the lava ourpur was on the decline. They were much less distincl during the later intervals of spasmodic rilt acriviry in lat€ May and June. The main feature oflhe e'rlire lilt record of Figu.e 2 can be seen as a general uplilt lrom lhe beginning until abo ur June 15, followed by a .apjd relufn to qhat was presumably the , ro-anys P s 3 u J I uly 3 E E m Frg.4. Rateolflowand levcl of wargin a l560'd d€ep well l? eruprior lftan Bjbmten a km w€st ofthe crutefur Heinre! durine thc l97l al. t915],. 2954 KRrsrrANssoN Er AL.: CRoUND TrLr MEASURtrMlNrs original level. As all output of lava aho slopped in lale June, we conclude thal at this point rhe drivinS pressure in a deepseated source responsible both for tho uptill and lor theextrusion of lava relurned to its normal (lithoslatic) level. This downtill may aho be corclaled with a drop in pressure in the l56o'm-d@p w€ll in Heimaey. as shown in Fisurc 4 [rrom Ejbmsson et al., t9l5l. The large tilt surses (and drops, €.g., those on Februa.y 23) superimposd on thh genor. sequence are more problemalic. They do not seem 1o conelate rilh details in the intensity of the errprion as ibdicaled by the number ol aclive venls, th€ height of eruption columns, and the produclion of etrusives. However, these indicators ol inlensity may depend on condi tions at very shallow leveh. e.g.,1he access ofseawater 1o lhe Aralysis ol elaslic ground deformation in various model spu.ious fluctuations peculiat ro rhe sire lEato, and Muruta, t960; Fiske aad Kidothila, 19691, and data from several (wocomponent lihm€lers as aell as ho.izontal distance merers mighl havo been needed to describe the ground defo.mation Acknowle.tgnenn- Th. aulhoc skh lo dck.owledge suppo.l by the Sciene Inslitule ofrhe UniveBiry ol lceland, which provid.d for the lognlics and op€ration ol th. installalion ar Heimaey, and rheloan of thc tiltmeter equipnenl by Arthur D. Linl€, Inc., ol Cambiidgc, Massachusetts, Our pcrsonal tha.ks ate aho exlended to lhe many obs€rvetr, especially H. Johnscn, who gave their tioe and besl erorts to keep the exp.rin€nl Soins under tbe nost adv.rse condrtions. REFERENcES Bjorn$on, A.. L. Kristjdn$od, lnd H. Johnsen,Som.observationsof the Heimaet d@p drill holc doting the er.qpt ion ol 1913, Jbkull, 25, in pr€ss, 19?5. (Absrlact appeared in Eos 1'taks. AeU, 55, 488, 1914.) geomelries irdicares thatthe maximumlilt occurs ala distance roDghly equal to rhe deplh oi the source lMogi 1958i l/atu and DeckeL l91ll. Since rhe tihmerer at Dalir *as al a dislance of Ll5 km liom rhe crater, it was mosl responsive to sources a1 a depth ol approximately l-2 km from th€ crater. Brander, The observed tih changes may lherefore siSnily changos in presure a1 this deprh o. even minor intrusive activity inlo lhe helerogeneous volcanic and sedimenlary lbrmalions undeF lying the islard. The qu€stion of a possible value ol tilt obsrvat'ons lbr pr€dicling the course of a volcanic €ruption cannot besalisfactorily answered on the bash of th€ presenl dala. First we did not have time to install the tillmeter al Heimaey beiore the eruption, ard so it remainsunknown at whal tim€and rate lhe till b€gan to changej d^ra by Bnhdet ond wadse lt974l ;ndicale large horizontal strains in the firsl 2 weeks. Seco.d th€ tilt data froln a single observalion point are likely b include Fiske, R. S., ard J., and G. Wadge, Dhtance measurenents afioss the Heimaey erupliv€ fisure,,{"ture, 244, 496498, t914. Eaton, J. P., and K, J. Murata, How volcadoes erow, S.,?r.e, /J2, 925-938, 1960. Eina$san, L. Aosid a Heihdry, Heimskringla, Reykjavik, lceland, 1914. w- T. Kinoshila, Inflalion oaKihuea vol€no prior to i1s 1967 1968 e.uplion, ,t ience. 165,341-149,1969, Mosi, K.. Relations of the €ruptions of various volcanoes and !t'e deloimaLion oi Lhc ground around then, Inst. Tokto Uiit,, J6,99 \J4. t954. ,ll Ea hq@ke Re!. Simon,l., A.C. Emsli€, P. F. Suong. and R. K, Mcconnell. Sensilire tilheler ulilizing a diamagn€dc supedsion. lq?o. S.i, Iror4z., J9, 1666-r671, 1968 Wdhh, J.8., and R. W. Deck€r, Slrfae delornarion associat.d wilh eolcanisn,.1, G.aphls. Res., 76, 3291-3302, 1911{Receiv€d Ociober 29, 1974i Elised March 5, 1975i a@epted March 24, l9?5.)