High and dry: model development to improve simulations of large-scale hydrological droughts in the Alps  

Hydrological droughts are often not limited to a single river basin but affect several basins simultaneously. The co-occurrence of droughts in different basins is influenced by meteorological processes, catchment characteristics and surface processes such as soil moisture and snow cover, the latter of which is particularly important in mountain regions.Large-scale hydrological models are a useful tool to study the drivers of large-scale droughts and to understand their evolution in a changing climate. In recent years, these models have moved towards increasingly high spatial resolutions, making them more applicable in regions with complex heterogenous mountain topography. However, large-scale hydrological models often have simplified representations of snow and glacier processes.Here, we set up the PCR-GLOBWB 2.0 global hydrological model, which represents snow cover by using a temperature index model with a constant degree-day factor and which has no explicit representation of glaciers, at a resolution of 30 arcseconds (approximately 1 km) over the Alps. We adapted the model to make it more suitable for mountain regions. Specifically, we (1)  improved the snow module and compare different implementations of temperature-index models and (2) implemented a new dynamic glacier module. In the new model set-up, we calibrated snow water equivalent and glacier elevation changes against observations and reanalysis products and evaluated the model over the Alps, with specific emphasis on the representation of spatial patterns in hydrological extremes. With this new set-up, we are able to tackle model issues related to excess snow accumulation and improve discharge simulations, particularly in glacierized catchments. We apply this new model set-up to the larger Alpine region to study the drivers of large-scale hydrological droughts.