A comparison of four snow models using observations from an alpine site

Abstract

 Results from four snow models-two used in climate models, one being developed for hydrological forecasting and one used for avalanche forecasting-are compared with observations made during two contrasting winters at a site in the French Alps. The models are all driven with hourly measurements of air temperature, windspeed, humidity, snowfall and downward longwave and shortwave radiation, but they differ greatly in complexity. Results from the models are compared with measurements of snowdepth, snow water equivalent, surface temperature, runoff and albedo. The models all represent the duration of snow cover well, but differ in their predictions of peak accumulation and timing of runoff. Experience gained in this study is used to make recommendations for a more ambitious intercomparison between a larger number of models for a wide range of environments.