A Multi-Horizon Morphodynamic Forecasting and Near Real-Time Scour Monitoring Framework for the Jamuna River

Large braided rivers pose persistent challenges for the protection of major river training works, as rapid channel migration, evolving char formations, and highly localized scour can compromise structures within short timescales. The Jamuna River in Bangladesh is one of the most morphodynamically active systems of this kind, making it an ideal but demanding environment for operational forecasting and adaptive management.

Since 2014, a morphological model of the Jamuna River has been progressively developed using detailed survey data and MIKE 21C simulations. This model underpins annual monsoon-season morphological forecasts, predicting planform adjustments and potential scour depths for the upcoming monsoon. The 2025 forecast report which was submitted on 30 April, identified elevated scour risk between CH 1300–2500, with maximum predicted depths ranging from -31.51 mPWD under a 1 in 100 year flood to -37.16 mPWD under a 1-in-2.33-year flood. These predictions guided initial preparedness and monitoring plans for the monsoon season.

In recent years, the framework has been extended to provide near-real-time scour forecasts for all major river training works, integrating short-term hydrological forecasts with high-frequency bathymetric observations. During the 2025 monsoon, the near-real-time hydro-morphodynamic modelling system was continuously updated using the latest 5-day water level forecasts from the Flood Forecasting and Warning Centre (FFWC) and validated through frequent single-beam and multibeam bathymetry surveys. This approach enabled timely detection of rapid scour intensification near CH 2600-2700. Based on combined survey and model results for August, a targeted dumping plan along CH 2470-2780 was formulated and executed. This represents an adaptive intervention strategy, where protective measures are triggered in response to evolving river dynamics indicated by both predictive simulations and real-time observations. By late August, measured scour reached -33.34 mPWD exceeding the design threshold by more than 6 m. Yet timely adaptive interventions maintained apron stability and prevented wider structural exposure.

This study demonstrates that operational morphodynamic forecasting integrating annual monsoon-season predictions with near real time model updates and survey observations, can significantly enhance the resilience of major river training works in highly dynamic sand bed rivers. It represents one of the first operational applications of an integrated multi-horizon and near real time morphodynamic forecasting framework for guiding adaptive river training interventions providing a practical and scalable blueprint for infrastructure risk management under increasing hydrological variability and climate-driven extremes.