Scientific and technical changes to flood forecasting models are implemented to improve forecasts. However, responses to such changes are complex, particularly in global models, and evaluation of improvements remains focussed on generalised skill assessments and not on the most relevant outcomes for those taking decisions. Recently, the Global Flood Awareness System (GloFAS) flood forecasting model has been upgraded from version 2.1 to 3.1 with a significant change to its hydrological model structure. In the updated version 3.1, a single fully configured hydrological model (LISFLOOD) has been adopted, including ground water and river routing processes, instead of two coupled models, a land surface and a simplified hydrological model, of the previous version 2.1. This study aims to evaluate changes in the simulated behaviour of floods and the forecast skill of the two GloFAS versions based on different decision criteria for early action. We evaluate GloFAS reforecasts for the Brahmaputra and the Ganges Rivers in Bangladesh for the period 1999–2018. For the Brahmaputra River, the old GloFAS 2.1 version performs better than the 3.1 version, especially in predicting low‐ (90th percentile) and medium‐level (95th percentile) floods. For the Ganges, GloFAS 3.1 shows improved probability of detection of low‐ to medium‐level floods compared to version 2.1, especially for lead times longer than 10 days. Both versions show limited skill for more extreme floods (99th percentile) but results are less robust for these less frequent floods given the lower number of events. Using lead‐time dependent thresholds improves the false alarm ratio while reducing the probability of detection. The changes in model structures influence the model performance in a complex and varied way and forecast skill needs further investigation across regions and decision‐making criteria. Understanding the skill changes between different model versions is important for decision‐makers; however, focused case studies such as this should also be used by model developers to guide future changes to the system to ensure that they lead to improvements in decision‐making ability.

In Bangladesh, where floods frequently occur, there is a severe annual risk to community displacement, agriculture, fisheries, livestock, public health, and food security. Extreme flooding events are becoming more common due to a combination of human‐induced climate change, irregular upstream river water flows, increased proportion of sediment distribution on the riverbed, institutional fragility, lack of planning regulations, and changing rainfall patterns. Effective flood management requires precise and timely flood mapping methodologies to adopt disaster risk reduction strategies and enable efficient response efforts. This study presents an approach that facilitates timely flood identification, improving emergency response, evacuation initiatives, relief distribution, and disaster risk reduction. This research introduces a novel methodology for expedited flood inundation mapping, using the August 2022, 2023, and especially the 2024 flood events in the Feni District as the primary case study. The study employs Google Earth Engine (GEE) and Sentinel‐1 synthetic aperture radar (SAR) data to accurately delineate flood inundation regions by utilizing vertical transmit and vertical receive (VV), vertical transmit and horizontal receive (VH), and VV/VH polarization bands. Water bodies characterized by lower backscatter values in VH polarization ranging from −41.15 to −24.06 dB and VV polarization from −31.66 to −15.94 dB were identified as suitable thresholds for flood inundation area delineation. To assess the accuracy of flood map, this study focuses on pixel‐based digital classification and machine learning (ML) techniques separately for flood inundation mapping. The classification accuracy values of 95.60% for the pixel‐based method and 94.40% for the random forest ML model specifically correspond to the 2024 flood event. This study developed a GEE‐based operational methodology by evaluating two innovative techniques designed for rapid flood inundation mapping to support effective flood management and disaster risk reduction efforts.

Dhaka, one of the world’s fastest-growing megacities, faces severe urban flood risks driven by rapid urbanization, inadequate drainage infrastructure, and climate change–induced extreme weather. This study develops an integrated urban flood forecasting system for a densely built-up area of Dhaka using the MIKE + hydrodynamic model. The system links hydrological and hydraulic modules to simulate the complex interactions among surface runoff (pluvial), river overflow (fluvial), drainage networks, and overland flow. Real-time data from rainfall, river gauge, and pump stations are incorporated to improve forecasting precision. Sensitivity analysis identified ten influential parameters, optimized during model calibration. Historical flood events from 2019–2022 were used for calibration and validation, yielding high performance with Nash–Sutcliffe Efficiency (NSE) above 0.70, R² above 0.80 and Mean Absolute Error (MAE) below 0.1 m. The model’s robustness was further verified using major flood events in 2020 2021, 2022, 2024 and 2025 validated against flood marks and community reports and observed data. This advanced forecasting framework offers reliable early warnings and scenario simulations, strengthening decision-making and emergency response for flood-prone urban areas in Dhaka.

While flooding is an annual occurrence in the Brahmaputra basin during the South Asian summer monsoon, there is large variability in the flood characteristics that drive risk: flood duration, rate of water level rise and peak water level. The aim of this study is to understand the key hydrometeorological drivers influencing these flood characteristics. We analyse hydrometeorological time series of the last 33 years to understand flood dynamics focusing on three extraordinary floods in 1998 (long duration), 2017 (rapid rise) and 2019 (high water level). We find that long duration floods in the basin have been driven by basin-wide seasonal rainfall extremes associated with the development phase of strong La Niña events, whereas floods with a rapid rate of rise have been driven by more localized rainfall falling in a hydrological ‘sweet spot’ that leads to a concurrent contribution from the tributaries into the main stem of the river. We find that recent record high water levels are not coincident with extreme river flows, hinting that sedimentation and morphological changes are also important drivers of flood risk that should be further investigated. Understanding these drivers is essential for flood forecasting and early warning and also to study the impact of future climate change on flood.