Daily snow water equivalent (SWE), gridded at 500-m spatial resolution, was reconstructed for 2001-2009 using canopy-corrected fractional snow covered area (fSCA) from MODIS and a temperature-index snowmelt calculation. The MODIS fractional SCA was based on the MODSCAG (MODIS Snow Covered Area and Grain size/albedo) model, and provides a daily estimate of SCA across complex terrain that avoids some of the inherent biases in a binary SCA product. fSCA corrections of up to 40% were made in dense forests of the western and 30% in the eastern Sierra Nevada to account for the satellite only detecting snow in viewable gaps in the canopy. Degree days for the temperature-index model were calculated using monthly average lapse rates, with values varying by ±25% over the seasonal snowmelt period. Elevation-dependent degree-day coefficients were estimated from snowmelt rates observed at snow-pillow sites distributed over the region, with values generally increasing over the snowmelt period. As there are no spatial ground-truth measurements for validation of the amount and spatial variability of the resulting product, SWE amounts were compared with point precipitation measurements, point SWE measurements and an interpolated/gridded precipitation product (PRISM). Changes in gridded SWE, representing snowmelt rates, were compared with melt recorded at snow pillows and with discharge recorded at stream gauges. At peak accumulation the gridded SWE values were generally within 25% of SWE measured at snow pillows in the same 500-m grid cell, but SWE at each snow pillow melted out 2-6 weeks before snow disappeared from the corresponding grid cell. Comparison of the gridded SWE with precipitation required partitioning the precipitation into rain versus snow, which was based on a combination of ground temperature and available dew-point temperature values. At higher elevations both station measurements and interpolated /gridded precipitation were generally within 50% of SWE reconstructed for the same location, with greater uncertainty near the rain-snow transition elevation. Results highlight how gridded, reconstructed SWE estimates can improve basin-scale spatial estimates of precipitation and snowmelt, estimates of the high-elevation rain-shadow effect, and basin response to climate warming. They also highlight the severe shortcomings of current measurements to accurately estimate these quantities.
Bales, R.C., Rice, R., Meng, X. (2010): Spatial distribution of snow water equivalent across the central and southern Sierra Nevada. Fall meeting, American Geophysical Union, December 2010. Abstract C33E-0592..