The data file defines a global seasonal snow classification system. Based on the physical properties of the snow (depth, density, thermal conductivity, number of layers, degree of wetting, etc.), the world's seasonal snow covers were divided into six classes, plus classes for water and ice fields. Each class was defined by its physical properties, then empirically related to climate using three variables (precipitation, wind, and air temperature). A vegetation proxy was used for wind data: tall vegetation equals low wind, short vegetation equals high wind.

The snow classification data are formatted into an array of integers, each value representing a 0.5-degree latitude by 0.5-degree longitude cell. The data set was developed and tested for the Northern Hemisphere. Results for the Southern Hemisphere are untested and have not been evaluated. The Southern Hemisphere suffers from a reduced density of meteorological data which may well have influenced the quality of the snow classification in that area. Also note that much of Antarctica is inappropriately defined as water (the vegetation data set did not extend that far south). A 0.5- x 0.5-degree land-sea mask could be used to set the Antarctica region to the ice classification if this region is of interest.

National Snow and Ice Data Center (NSIDC) has created a land/ocean mask from the original (updated) Global Ecosystem map (see Related links) used in the derivation of the snow classification. The mask may be used to set the land mass of Antarctica to ice, as described above, altering the data set as submitted.