Item

Abstract

In the Alps, the seasonal snow cover controls on freshwater availability and hydropower production. Distributed snow cover mapping is essential to monitor snow amounts and anticipate risks. The ESA Alpine Regional Initiative AlpSnow EXPRO+ proposes near-real time remotely sensed snow cover observations that make the best of ESA’s Sentinel 1-3 suite. These products offer daily to bi-weekly, high resolution (20m-200m) observations of snow cover fraction, snow albedo, and snow wetness. Data assimilation of these observations into energy-balance snowpack models could provide us with the best, multi-variable estimate of the current snow cover and its associated uncertainties. Such analyses would serve as an initial state for sub-weekly snow hydrology operational forecasts, which is a key demand from end-users. In this work, we bring together ensemble simulations of the snowpack, accounting for meteorological and snowpack modelling uncertainties, with AlpSnow observations in view of assessing their compatibility for data assimilation in a pixel-by-pixel approach. In particular, the observed snow wetness requires the design of a dedicated transfer function from modeled liquid water content to a variable that represents the observations. We focus on a 1000km2 subdomain of the operational snow hydrology service for Switzerland, exhibiting a variety of snow conditions and landcovers throughout the melting season. We show that multi-parameter observations have the potential to enhance model simulations, by better constraining the onset of melt as well as unresolved processes such as fractional snow cover.