Snow cover sensitivity to horizontal resolution, parameterizations, and atmospheric forcing in a land surface model

This paper assesses the impacts of horizontal resolution, snow physics, and atmospheric forcing in snow cover simulations by the European Centre for Medium-Range Weather Forecasts (ECWMF) land surface model Hydrology Tiled ECMWF Scheme for Surface Exchanges (HTESSEL). Off-line simulations are carried out forced by the ECMWF deterministic short-term weather forecasts (WFC) with a resolution of 25 km from March 2006 to June 2010. The horizontal-resolution impact on snow cover is addressed by performing simulations at 25, 80, and 200 km forced by WFC. The impact of atmospheric forcing on snow cover is assessed by forcing the model additionally with the ECMWF Era-Interim (ERAI) reanalysis, at 80 km resolution. Snow physics effects are analyzed by performing an extra simulation forced by WFC using a different snow scheme. The simulations are validated against four independent observational data sets: (1) snow water equivalent (SWE) over Switzerland; (2) snow cover duration in Europe (SNOWCLIM); (3) interactive multisensor snow and ice-mapping system (IMS) snow cover; and (4) Moderate Resolution Imaging Spectroradiometer (MODIS) surface albedo. ERAI forced simulations show a systematic underestimation of SWE and snow cover fraction, which is due to an underprediction of snowfall by ERAI. The snow physics experiment highlights the sensitivity of the model to the partitioning between rainfall and snowfall when rainfall interception in the snowpack is neglected. The horizontal resolution has a crucial role in characterizing the snow cover over complex terrain (e.g., orographic areas, coastal and lakes regions). However, improved snow physical parameterizations and meteorological forcing are shown to be the key elements to achieve more accurate simulations of the snowpack and of the snow-atmosphere interactions, also over flat terrain.