Abstract
The aim of this study was to evaluate the performance of different variants of the parsimonious large-scale GSM-WASMOD glacio-hydrological model in quantifying the water resources of the Upper Beas basin under past and changed climate conditions. The study area is a meso-scale glacierized Himalayan basin with limited long-term ground-based hydro-climate and glaciological data. 10 km- and 50 km-resolution gridded GSM-WASMOD models were setup with the limited ground based data and calibrated with daily discharge data from Thalout station. Both resolutions were tried with linear and exponential forms of slow and fast flow formulations. The model variants with linear formulations performed poorer as compared to their counterpart with exponential formulations, and were eliminated from further consideration. At monthly and yearly time step, both the 10 km and 50 km resolution models performed well (monthly NSE > 0.80) considering the poor quality input data. At daily time step, the 10 km resolution performed much better (NSE=0.72) than the 50 km resolution model, and its performance was found to be comparable to that of the 1km resolution HBV and ENKI models. The GSM-WASMOD, however, performances were not satisfactory in their monsoon runoff predictions. The cross validation of model parameters between the 10 km and 50 km resolution models suggest that the optimal parameters are scale dependent. Thus, experiences with lumped models might not be transferable to fine resolution gridded models. The calibrated 10 km-resolution model was used to quantify the total runoff, snowmelt runoff, glacier runoff, and other components of the water balance under past (1991 to 2001) and future hypothetical climate scenarios. 21 synthetic scenarios were constructed using a combination of 1 to 3 oC temperature changes and -15 to +15% precipitation changes. Under the reference scenario snow melt from glacier-free areas contributes about 42 % of the total runoff. Snow and glacier melt from glacier covered areas, however, were estimated just less than 2% which is much less than previously reported by Li et al. (2014) and Singh and Kumar (1997). Hence, the GSM module did not perform quite well for the study basin and needs a revisit. The largest effect of warmer climate in the study basin is on glacier runoff and total runoff while the biggest impact of wetter climate is on snow melt, total runoff and land moisture. Among the seasons, pre-monsoon and monsoon seasons are predicted to be affected most by the climate change scenarios studied. But, no impact is noticed on the peak timing of monthly runoff despite the predicted earlier snowmelt peak.