Hydrological Memory in the Himalayan Compound Floods: A Hydromet-to-Hydraulics Framework for Adaptive Flood Risk Management

Hydrological memory refers to previous wetness or dryness, which fundamentally alters the ecosystem's response to a disturbance. It describes how antecedent catchment wetness persists over time and amplifies subsequent flood response. This temporal compounding mechanism poses significant challenges for flood forecasting and reservoir operations in Himalayan basins where climate change is intensifying precipitation variability. The August 2023 Punjab floods, which impacted over 12,000 settlements and resulted in 65 fatalities, demonstrate the influence of hydrological memory in dam-regulated systems. This study employs a hydromet-to-hydraulics framework that begins with atmospheric analysis and goes all the way down to HEC-RAS hydrodynamic modelling and demographic impact assessment. This approach integrates hydrodynamic modelling with forecast aware dam operations to quantify flood exposure patterns.

A detailed spatiotemporal analysis shows that heavy rain in July raised the soil moisture levels in the Beas Basin significantly. The hydrological memory increased the likelihood of August floods, despite the rainfall received being less than the seasonal average. The study employed HEC-RAS hydrodynamic modelling integrated with 2011 census data to evaluate the effectiveness of dam operations. The results showed that controlled dam operations reduced the downstream population exposure by 80%. We proposed a Genetic Algorithm-based optimisation framework with piecewise penalty functions based on SWAT-generated inflow forecasts from GFS precipitation data. It showed that forecast-informed dam operations can more effectively balance flood mitigation with water conservation objectives than manual management. This integrated framework underscores the essential requirement for reservoir management systems that incorporate catchment memory states through continuous soil moisture monitoring and precipitation forecasting.