Since about 1950 there has been a trend in the California Sierra Nevad
a toward a decreasing portion of the total annual streamflow occurring
during April through July, while the streamflow during autumn and win
ter has increased. This trend not only has important ramifications wit
h regard to water management, it also brings up the question of whethe
r this represents a shift toward earlier release of the snowpack resul
ting from greenhouse warming. Therefore, the observed record has been
examined in terms of relative influences of temperature and precipitat
ion anomalies on the timing of streamflow in this region. To carry out
this study, the fraction of annual streamflow (called the fractional
streamflow) occurring in November-January (NDJ), February-April (FMA),
and May-July (MJJ) at low, medium, and high elevation basins in Calif
ornia and Oregon was examined. Linear regression models were used to r
elate precipitation and temperature to the fractional streamflow at th
e three elevations for each season. Composites of monthly temperature
and precipitation were employed to further examine the fractional stre
amflow in its high and low tercile extremes. Long time series of clima
tic and hydrologic data were also looked at to infer the causes in the
trend toward earlier runoff. For the low-elevation basins, there is a
dominant influence of precipitation on seasonal fractional streamflow
. Middle-elevation basins exhibit a longer memory of precipitation and
temperature in relation to their fractional streamflow. In-season pre
cipitation is still the most important influence upon NDJ and FMA frac
tional streamflow; however, the influence of temperature in melting th
e snowpack is seen on MJJ fractional streamflow, whose strongest influ
ence is FMA temperature. At higher elevations, prior-season precipitat
ion exerts a greater influence than at low and middle elevations, and
seasonal temperature anomalies have an effect on all seasonal streamfl
ow fractions. There are several causes for the trend toward decreasing
fractional streamflow in the spring and summer. Concomitant with the
trend in the timing of streamflow was an increase in NDJ (most notably
November) precipitation. There also has been a trend toward higher sp
ring temperatures over most of the western United States, but since th
ere has also been a trend toward decreasing temperatures in the southe
ast, we do not interpret this as a signal of anthropogenic warming. Ot
her factors in the trend toward earlier streamflow may include a decre
ase in MJJ precipitation and an increase in August-October precipitati
on.