Changes in predictive skills through coupling a hydrological modelling framework with a glacier model

The Himalayan Mountain range has a substantial glacier cover, supplying melt water to local communities in the drier season for irrigation and human consumption. Effects of climate change on glacier retreat and therefore melt water availability are expected to be severe. An accurate representation of glacier processes is thus of great importance to predict water availability under future climate projections. Hydrological models focus mostly on processes occurring in non-glacierised areas with often overly simplified glacier parametrization. This can lead to uncertainties in streamflow predictions, especially in highly glacierized catchments. Coupling a glacier model to a hydrological model can resolve some of these uncertainties by a more accurate description of glacier-related processes including ice melt, and providing the extent of glacier retreat, which is essential to quantify changes under a transient climate. In this study, we test the hypothesis that coupling the glacier model GloGEM with the hydrological modelling framework Raven can lead to an increase in predictive skills through a better glacier parametrization. The chosen hydrological modelling framework Raven allows for testing multiple hydrological model structures, accounting for uncertainties along the full modelling chain. The relevance of coupling a glacier model with a hydrological model is analysed in a test basin with in-situ measurements of glacier mass balance and streamflow. Modelling results from coupled and non-coupled model runs are evaluated with the available streamflow data.