North-east region of Bangladesh is comprised of hilly transboundary rivers in the upstream and complex river network along with floodplain in the downstream. In this study, a satellite observation and numerical modeling based near real-time flash flood forecasting framework is developed for the north-east part of Bangladesh. For nowcast information, Global Precipitation Measurement Mission (GPM) developed latest near real-time IMERG precipitation product is chosen. To produce improved quality precipitation forecasts, Global Forecasting System (GFS) based precipitation predictions are corrected by applying climatology based bias adjustment technique. Flow from upstream river channels into the complex river network are calculated by using the nowcast and forecast precipitation products forced into Soil and Water Assessment Tool hydrological Model. The inflow discharges from the hydrological model then used in HecRAS hydrodynamic model to produce forecasted water height/discharge, flood depth and inundation in prominent locations in the study region. Both of the hydrological and hydrodynamic models are simulated sequentially in sub-daily scale to produce 120 hours lead time forecasts. SWAT Model was calibrated and validated in six inflow stations using rated discharge. HecRAS Model was calibrated and validated using measured water level at different locations.

Dam construction is on the rise in developing nations. Monitoring of these dams is necessary to understand downstream hydrologic impacts and for better planning and management of water resources. Satellite observations and advancements in information technology now present a unique opportunity to overcome the traditional limitations of reservoir monitoring. In this study, a global reservoir monitoring framework is developed as an online tool for near-realtime monitoring and impact analysis of existing and planned reservoirs based on publicly available and global satellite observations. The framework uses a mass balance approach to monitor 1598 reservoirs in South America, Africa, and South-East Asia region. The framework simulated streamflow was validated in 18 river basins and the storage change was validated against in-situ data of 77 reservoirs. The framework was able to capture reservoir state realistically for more than 75% of the reservoirs. The tool can now be used to study existing and planned reservoirs for hydrologic impact and operating pattern for short and long term decision making and policy analysis.