Seasonal Snow Accumulation, Melt and Water Input—A New England Model

Abstract

Two geographic parameters, elevation and latitude, account for much of the spatial variation of temperature and precipitation over New England during the winter and spring seasons. Since temperature controls the amount of precipitation which falls as snow and is highly related to snowmelt, a spatial seasonal climatology of snow accumulation, melt and water input can be derived by modeling the elevation and latitude effects on temperature and precipitation distributions. Such a model is developed primarily from Agricultural Research Service Sleepers River Watershed data in northern Vermont. Based only on predictor station temperature and precipitation observations and on elevation and latitude differences between predictor and predictand locations, snowpack accumulation, melt and water input are estimated for 10-day time steps through the snow accumulation and melt season (2 November through 30 May). Using a low-elevation station to estimate snowpack at higher elevations, model estimations of snowpack water equivalent represent a 46% improvement over climatological averages. Comparable skill is obtained in specifying water input. Skill diminishes with increasing predictor-predictand distance to near zero at about 100 mi (161 km). This model enables much of the climatic and 10-day synoptic information on snowpack accumulation, melt and water input over unobserved watersheds to be estimated from ordinary National Weather Service network temperature and precipitation observations.