Implications of Snowpack Data Uncertainties for Hydrological Modelling of Climate Change Impacts in Northern India

Abstract:

Glaciers and snowpacks influence streamflow by altering the volume and timing of discharge. With climate change, an initial increase in total runoff is expected followed by an eventual decrease, as the glaciers or permanent snowpacks retreat. Without reliable data on snow and ice volumes and properties, initialising hydrological models for climate impact assessment is challenging.

In this paper we explore the implications of the model-setup assumptions regarding snow/ice reserves for future water resources simulations in the Beas catchment in northern India. Two contrasting HySIM model builds, in which snowcover in the high-elevation subcatchment is either seasonal (snow melt is snowpack limited) or permanent (snow melt is energy-limited) were calibrated and validated against observed discharge data (2000-2008). We then applied both models within a scenario-neutral framework to develop Impact Response Surface of hydrological response to future changes in annual temperature and precipitation for the region from AR5.

Both models had similar baseline model performance (NSE of 0.68-0.70 in calibration and 0.66 in validation), but the impact response surfaces differ in the magnitude and (for some combinations) direction of model response at low (Q10) and high (Q90) daily flows. For example, for +1^oC / +10% change in annual temperature / precipitation, corresponding to around the 75^th percentile changes for RCP 8.5 for the region in 2065, future Q10 changes by between -5% (assuming seasonal snow cover) and +45% (permanent snowcover) and Q90 by between +10 (seasonal snow cover) and +55% (permanent snowcover). This paper will consider the implications of data inadequacies in snowpack characterization for assessing the impacts of climate change and the associated timing of hydrological tipping points.