Navigation mit Access Keys

The SLF’s operational snow-hydrological models

Main menu

 

By conducting ongoing analyses of snow water resources, the OSHD contributes to hydrological forecasting in Switzerland. The analyses are based on a modelling framework that combines data stemming from snow measurements, several snow melt models and data assimilation techniques.

 

The SLF's operational snow-hydrological service (OSHD) has developed a modelling system which encompasses both meteorological data and information captured by Switzerland's snow measurement networks. The first step entails applying an enhanced temperature index model (Slater et al., 2006) to estimate snow quantities and their melting rates from precipitation and temperature data which has been obtained. With the aid of data assimilation techniques, such as the Ensemble Kalman filter method (Evensen, 2009), these estimates are continuously compared with snow observation data and thereby improved. This modelling framework provides an accurate picture of the spatial distribution of snow quantities and the regional snow melt pattern (Figs. 1 and 2; Magnusson et al., 2015). The framework can be combined with weather forecasting models to enable a prediction of the amount of snow melt over the following days. The only inputs required by the temperature index method are precipitation and temperature data, which are acquired in Switzerland by dense measuring networks. Temperature index models are capable of delivering sound and accurate results in numerous situations. In the context of complex meteorological conditions, however, they are subject to limitations. In these circumstances, models that give consideration to detailed snowpack processes are superior to such conceptual approaches.

 

Energy balance model for snow

For this reason we also use a snow melt model based on energy and mass balancing (Essery, 2013). This energy balance model calculates various thermal energy fluxes that influence snow melt (e.g. shortwave radiation and heat conducted from the soil). Especially in the context of complex meteorological conditions, this model type predicts snowpack runoff more accurately than a conceptual temperature index model – when rainfall is deposited on an existing snowpack, for example (Fig. 3). On the other hand, energy balance models depend on additional meteorological inputs that are often difficult to obtain with the required accuracy. We therefore use several models alongside each other in order to benefit from their individual strengths.