WMOD00275 - Laserfiche WebLink

Home Browse Search

706 . JOURNAL OF IIYI>RO~ETEOROLOOY VOI,UMF 7 were plilCcllUlllT ncar snow courses thai exhihited high C()rTclalinn~ with streamflow (Schaefer and Johnson 1t)lJ2), IIntl locations with uniform snow cover thai were thought 10 represent snowpadr. conditions specific 10 the elevation and aspect of thai panicular suhhllsin Walmer 19M). SNOTEL and snow course networks were installed to hoth provido.:: and improve upon prediclions of sea. sonal runoH through point estimates of SWE. The sul>- sequent sCllwna] streamflow forecasts typic.llly have er- ron strongly rdaled 10 the liming and character of pre- cipitation (Pagano cl al. 201M). When climatic. and therefore, slrcamnow. conditions deviate from the long-term mean. forecast errors increase. An alternative 10 stllli!>licnlly hascd forcCHsls is mudel-hased forecasting. Phr;icHlly based hydrologic models (e.g.. Leavesley el al. 20(2) use ;n1erpolalcd prl..-cipitalion and temperature valucs 10 estimate basin conditions (e.g., snowpad., soil moislure, slorage. ele). I>islrihuled eslimates from poinl measuremenls of ground-hased SWE have the pOlentialto improve upon these simulations of basin conditions by adjusting model mass and energy states (Dres....ler et al. 2IX16). Ullimately, this improves prediclions of snowmelt run. off liming and seasonal streamflow volumes. It should he noted, however,thal snow course nelworks were not designed as measures of the abwlute alllounl of SWE in a basin, or 10 yield SWE estimates acros.'. a hasin. However, for Ihe prescnt, these arc the best dala avail- ahle on historical SWE ncross the western United St.ltes. This paper examines spali.ll different."Cs in basin-wide SWE estim,lted using SNOTEL versus snow course data for the Colorado River basin. We firsl compare differences in poinl values al collocated SNOTEL .!Od snow course sites over represenlalivc dry. average, and wet years. To cvalullte differences between the dalasets in eSlimating lolal water resour~"Cs from snow, hasin. scale SWE was estimated using hypsomelric inlerpola- tion of poinl SNOTEL versus snow course measure- ments, and combinations of Ihe two dalasels. Knuwl- edge of spalial diffen..nces between SNOTEI. amI !>now course dala at Ihe poinl and basin !>Cllle will heller in. form appliclItions such as water resource management regarding is..sues of using eilher dal,l !>ource for opera- tions or hydrolugie mudding. 2, Stud) un'u The Colorado River h:lsin (Fig. I) is over IJ(X) km long and up to 8(X) km wide. The Upper Colorado hasin (Serrele et al. 1<)99) has a dminage area of 277 (XXI kml, an elevation range of 975-4260 m, and an llvenlge e1~ , +? -'"..kloI.rn .Ol.SOO .I.soo-2.500 11:12.soo.),ooo DJ.OOO.uoo D),5OO.4.5OO + ~El 16J.5 ns , . -- 1'10. I. A C.,lnrlkln River t..,in tuclIli,," map. evation of 2150 m, while the Lower Colorado has a draimlgc area of 346 (J()O kml. with an clc~'ation range of 0--3771 m, and an average elevation of 1310 m. Al- most 60% of the Upper Colorado, but only 16% of Ihe Lower Colorado. is aoove 2{J()1) m. More of thl.' annual precipitation in the Upper Colo!iu.lu falls ;l-'i !inllW and is slored throughout the wi Iller seilson than in the Lower Colorado (Scrrele et al. 19'},)). The focus of this paper is the entire Colorado River (623 (KXl kml) IInd three of its suhhasins: Gunnison (21) 5m kml). San Juan (63 7lXl km2l, ,md Sail-Verde (35 IlXl km2) (Fig. I). These ba- sins were chosen to rl.'present large pori ions of snow area for Ihe Upper (Gunnison and S,m Juan) and Lower (Sail-Verde) Colorado, and to represent a range of terrain complexity and forest density. .\' ()uhl ulld methods The SNOTEL SWE data arc available for more Ih;m (,(Xl sites lllld !>now course d,lta for nearly 2(XIlI sites in the western United Slales from lhe NRCS (see infor- mation online al hllp:f1www.nrcs.usda.gov). Snow courses provide sn.lp!>huts of SWE lIear the 1st of each month, hased on lhe average of 10 manual lIIe;lsure- men Is al each site. SWE is measured by in'oCrling a hollow aluminum or sleel tube in III the snnw, elllfacting a core. and weighing il. Silme manual snuw ~lurses have been eliminaled as more aultJmaled !>tations have been e!>lah1ished lllld Iheir U'oC in wllter foreeasls ac- cepled. At Ihe SNOTEL sites, snow accumuhlling over II nuid.fiIled pillow forces the nuid into a manomelric column and height changes arc recorded.