Extreme values of snowpack water equivalent [SWE] are of interest for both flood management and engineering design. Direct analyses of historical SWE in the U.S. are inadequate for many purposes due to the fact that continuous observations are limited almost entirely to the rather sparse network of first-order National Weather Service stations. A more spatially dense network is available from the cooperative network stations, although these do not routinely measure SWE. Rather, analysis of the SWE climatology using these stations must be based on snow depths [SD], from which SWE may be estimated through a specified density for the snowpack.
Snowpack characteristics, including snowpack density and depth, vary significantly on the scale of a region such as the northeastern U.S. However, previous studies have assumed snowpack density to be spatially invariant for the purpose of estimating SWE from SD. Alternatively in this atlas, densities are allowed to take on different values at individual cooperative network stations. The specified densities are based on a regression analysis involving several climatic variables to improve the accuracy of the extreme SWE estimates derived from SD observations. In addition, the Wakeby distribution is used to extrapolate the estimated extreme SWE values beyond the length of the current observational record.
17 pp.