M a n a g e m e n t

a n d

A n a l y s i s

o f

S n o w,

A v a l a n c h e

a n d

C l i m a t e

D a t a

RESULTS

Our findings indicate that the overall amount of SWE has
not significantlychanged overthe past 65 years through-
out the Sierra Nevada (figure 1).However,the snow ac-
cumulation patterns have narrowed, with less snow both
earlyandlateintheseason.January1snowfractions
seem stable (figure 6), which suggeststhat average Decem-
bersnow levels remain unchanged.February and March
(figures 7,8) also show stable SWE fractions below about
2000m,butathigherelevationstheslopesareconsist-
entlyestimatinglesssnow,withsignificantdeclining
trends at 30% of the elevation zones.
Due to increasing precipitation in March, the April 1
levels indicate that the snow has "caught up," and looks
likepastAprilamounts,butwithanelevationaldis-
tinction (figures 3,9).The lowerelevations retain less
snow,whilehigherareasreceivemore.Thisismost
likelydue torain on snow events,which meltsnow in
the lowerelevations.

reduced March 1 values.

Average meltmonthsand maximumyearlySWE values
were computedfor each station.The month of maximum
SWE was calculated foreach station-year,as wellas the
differencefromtheaverage meltmonth.Monthlyfrac-
tions of the station-year's maximum SWE were calculated,
yielding 9000 station-years with at least three months of
data each and spanning up to 78 years.
We computed linear regressions of monthlyfractions,
maximum SWEvalues, anddifferent melt month data over
an average of 65 years and grouped theminto20eleva-
tion zones, each extending 100 meters.These regressions
werechecked witharobustkernelestimator,whichis
usefulfornonparametricregressionswithoneexplana-
tory variable.This technique uses generalized cross-vali-
dation,where points are left outone at a time while the
regressionisestimatedontheremainingobservations,
thereby minimizing the mean square errorand selecting
that fit.

4