Snow residence time (the average length of time that snow will remain on the ground, in days, for the contiguous United States) was modeled using the spatial analog models of Luce et al., [2014] (see also Lute and Luce, in review). These models are built on precipitation and snow data from Snowpack Telemetry (SNOTEL) stations across the western United States and temperature data from the TopoWx dataset [Oyler et al., 2014]. To simulate historical and future gridded snow, the models ingest gridded winter cumulative precipitation and winter average temperature (from the MACAv2-Metdata dataset described above). While the snow models have been rigorously validated [Lute and Luce, in review], their application to gridded climate model data has not been evaluated. Note that this models of Snow Water Equivalent (the amount of water contained in the snowpack on April 1, in millimeters, for the contiguous United States) is available from the data provider.

Spatial Extent: CONUS and Alaska* (*comparable SWE data, also available in the EDW, is available for Alaska)

Units: Percent change in Snow Residence Time, current vs. future