Synthetic Year-Independent Spatio-Temporal Patterns of Snow Depletion

Snow cover extent is a key variable which is critically important for climate and hydrologic studies. In recent years, the MODerate Resolution Imaging Spectroradiometer (MODIS) satellite sensor has been used in a number of studies and gained widespread acceptance for estimation of snow cover extent due to its high spatial and temporal resolution. Several researchers have observed that snow accumulation and ablation occur in reasonably regular patterns from one year to the next. Thus, information from satellite imagery across different years ought to be able to inform the spatial distribution of snow in one year based on information in another year. This is a vital piece of information which has not yet been fully recognized or exploited. We have developed a method to synthesize the regular, year-independent, spatio-temporal pattern of snow depletion from the beginning to the end of melt seasons, using a series of snow cover maps produced from MODIS data across multiple years (2001 to 2011) coupled with the melt-out dates of a collection of SNOwpack TELemetry (SNOTEL) stations. The synthesized spatial time series have the capability of extrapolating snow covered area in space (e.g., during cloud obscuration) and time (e.g., for forecasting snow cover ablation). The accuracy of this method has been evaluated over the headwaters of the Upper Snake River in Western Wyoming with very good results. This method has many applications including cloud removal, within-season snow cover ablation forecasting, climate change impacts on snow and runoff, and modeling climatology of snow. These applications can be extended for use in water management and water supply forecasting.