Executive Summary : | The development of a flood forecasting system in the Alaknanda River basin will benefit various stakeholders, including water resource managers, hydropower plant operators, population downstream, and disaster management authorities. However, the complex topography, unavailability of dense hydro-meteorological measurement stations, presence of glaciers, frequent landslide and avalanche changes, and ongoing anthropogenic disturbances in the Himalayan region make it difficult to set up a hydrological model. The hydrological model calibration should include glacier mass balance estimates, snowmelt, glacier melt, rainfall, and baseflow to the total runoff. A fully distributed hydrological model capable of estimating different components of runoff at the outlet, including glacier melt and accumulation process, is highly desirable in the Himalayan region. This study aims to use all available meteorological and streamflow data, satellite precipitation estimates (SPEs), and reanalysis datasets to develop a two-stage blending approach to merge SPEs with observation data and produce better precipitation estimates with uncertainty measure. An optimization-based approach will be used to identify critical nodes in a river network, with discharge at critical nodes available either from observation or by routing surface runoff from reanalysis data. A fully-distributed hydrological model, WATFLOOD, will be calibrated for the Alaknanda River basin, and a multi-criteria calibration using observed discharge at critical hydrometric stations, surface runoff, snow-melt, glacier melt, and evapotranspiration data is proposed. A post-processor based on Bayesian Joint Probability approach will be used for streamflow/flood forecasting purposes. When successful, a flood forecasting system for the Alaknanda River basin can be operational with the help of concerned agencies. |