Seasonally snow covered mid-latitude forests may be highly sensitive to climate change as they often overlap or reside near the present-day synoptic mean rain-snow transition zone. Limited capabilities of satellite remote sensing in forested, steep terrain combined with sparse in-situ observations emphasize the need for improved numerical simulations of the distribution of snow water equivalent in these regions. The land surface / snowmelt model Alpine3D was used to simulate snow accumulation and melt in the 7.22 km2 sub-alpine Wolverton basin in the southern Sierra Nevada, California. The basin is part of the Southern Sierra Nevada Critical Zone Observatory. Results from three snow seasons were evaluated against data from a distributed network of automated snow depth sensors, repeated catchment-wide snow survey measurements conducted in 2008 and 2009, and LiDAR data from 2010. Compared to the local 86-year historical record, the three years of observation accumulated average (2008), 48% below average (2009) and 43% above average (2010) maximum annual SWE. A mid-winter rain-on-snow event occurred in both 2008 and 2009. The inter-annual variability in maximum SWE combined with inter-annual differences in the timing and type of precipitation events, the timing of seasonal melt onset, and differences in the persistence of spring cloud cover caused significant inter-annual variability in areal snow cover depletion rates. In 2009, the year with the least precipitation, the most spring cloud cover, and a basin-wide late-January rain event, SWE patterns exhibited the least spatial variability and areal snow cover depletion was rapid. Conversely, the greatest spatial variability in SWE was simulated in 2010, the year with the most precipitation, no rain events, and a melt season that extended into early summer. The areal snow cover depletion curve for this year exhibited a rapid exponential phase as in 2009, but a distinctly different transitional phase as deep snow cover persisted at forested upper elevations (confirmed by automated depth sensors) long after >90% of the basin was snow-free. Simulations of the average snow season (2008) predicted an areal snow cover depletion curve that exhibited dampened characteristics of both the wet and dry years. Additionally, significant inter-annual differences in accumulation and melt patterns between open and sub-canopy environments were simulated. The results illustrate the utility of physically based models to simulate highly heterogeneous seasonal snowpack dynamics. Furthermore, results support the use of such a land surface model to predict potential impacts of climate change on the future of the region’s hydrologic regime and the role of snow in these sub-alpine systems.
Musselman, K.N., Molotch, N.P., Margulis, S.A., Kirchner, P.B., Bales, R.C. (2011): Inter-annual snow accumulation and melt patterns in a sub-alpine mixed conifer forest: results from a distributed physically based snow model. Fall meeting, American Geophysical Union, December 2011, Halls A-C, 1:40 PM. Abstract C33D-0671..