An integrated assessment of extreme hydrometeorological events in Bangladesh

Climate and extreme hydrometeorological studies are required to reduce risk and vulnerabilities. This study uses different cumulus and microphysics schemes in Weather Research and Forecasting (WRF) model to simulate heavy rainfall events and Cyclone Sidr in Bangladesh, where many extreme events occur. Results show that WRF can capture the cyclone track, intensity, and landfall position. In addition, regionalization and an ensemble method through Bayesian regression model (BRM) are used to improve WRF rainfall simulations. Although regionalization can improve results of the experiments with different schemes, BRM leads to the best performance. To consider uncertainties and evaluate hazards, a probabilistic framework and proper indices based on distributions are used. Spatiotemporal precipitation distributions are used to develop the flash flood index (FFI) to compare the intensity of heavy rainfall events. Results show a large FFI over the central and southeast divisions that indicates the areas prone to the flash flood hazards and high-level risk. We use standardized precipitation index (SPI) in 6- and 12-month time scales based on monthly precipitation of ACCESS-CM2 in 2015–2100 under two scenarios (SSP-126 and 585) to evaluate long-term precipitation changes and drought/flood tendency. In addition, the probability distributions of the precipitation and wind speed are used in reliability analysis of the infrastructure. Target reliability indices determine the proper design rainfall (105 mm d −1 ) and wind speed (60 m s −1 ) that leads to a safe design of infrastructure. This study provides an integrated analysis of extreme hydrometeorological events, which is crucial for sustainable adaptation and mitigation plans.