Flood modelling of a partly glacierized catchment in the Himalayas in a context of climate change

The Himalayan region is severely exposed to the flood risk due to the heavy rainfall during the summer monsoon. The dynamics of the hydrological response during extreme events is relatively less understood, because of several complex and interactive processes. In this scenario, the use of rainfall-runoff models capable of adequately taking these processes into account could be crucial for reliable flood forecasting. However, in areas with such complex topography, accurately characterizing meteorological forcing and streamflow dynamics remains a challenging task due to the lack of ground measurements. Furthermore, in highly glacierized Himalayan basins, the significant contribution to streamflow by snow and ice melting has been shown to be progressively increasing due to its sensitivity to climate change, in parallel with the loss in glacier mass. In this study, a conceptual and parsimonious hydrological model was implemented in semi-distributed mode and calibrated against streamflow and glacier loss volume data simultaneously. The MISDc-2L model was modified to simulate not only the snow accumulation and melt, but also the glacier melting in the ice-covered fraction of sub-basin area, assumed to occur once the seasonal snowpack is locally depleted. The Alaknanda River (one of the two headstreams of the Ganges) was chosen as a case study because it experiences several disastrous flood events in recent years. The basin upstream the Rudraprayag gauge was considered (8600 km2), for the period 2000-2020. The Randolph Glacier Inventory v7.0 was employed to locate glacierized areas, while glacier storage change data were extracted from available literature studies. Elevation data from NASADEM and hourly variables from ERA5-Land reanalysis dataset were used. A joint objective function was considered for calibration, including the Kling-Gupta efficiency, a high-flows hydrological signature and the error in glacier stored water loss. The model, constrained with glacier storage change data, was found to be able to provide good hydrological performances, both in calibration and validation, also with specific reference to annual flood peaks. Despite the simplicity and the flood-oriented approach, the proposed modelling procedure simulated the dominant hydrological processes in a physically plausible way, in a basin with high-altitude glacierized areas in a context of climate change. The goal of adequately characterizing the contribution of glacier melt to total streamflow was pursued by aiming for consistency with additional data sources. This work was funded by: - the Next Generation EU - Italian NRRP, Mission 4, Component 2, Investment 1.5, call for the creation and strengthening of 'Innovation Ecosystems', building 'Territorial RD Leaders' (Directorial Decree n. 2021/3277) - project Tech4You - Technologies for climate change adaptation and quality of life improvement, n. ECS0000009. This work reflects only the authors views and opinions, neither the Ministry for University and Research nor the European Commission can be considered responsible for them. - the FLOSET Project 'Probabilistic floods and sediment transport forecasting in the Himalayas during the extreme events, funded in the context of the 'ITALY-INDIA JOINT SCIENCE AND TECHNOLOGY COOPERATION CALL FOR JOINT PROJECT PROPOSALS FOR THE YEARS 2021 2023'.