RESEARCH ARTICLE

Ensemble flood predictions for River Thames under climate change

Department of Civil and Environmental Engineering, Brunel University London, Middlesex UB8 3PH, UK

^* Corresponding author (email: yurui.fan@brunel.ac.uk)

Received: 14 April 2023
Revised: 9 November 2023
Accepted: 4 December 2023

Abstract

In this study, a Bayesian model averaging (BMA)-based ensemble modeling system is proposed to project future flood occurrences for the River Thames using downscaled high-resolution climate projections from the latest general circulation models (GCMs) in the Coupled Model Intercomparison Project Phase 6 (CMIP6). The BMA-based ensemble modeling system integrates multiple hydrological models into the BMA framework to enhance the accuracy of hydrological forecasting, which has shown good performance in validation with the NSE higher 0.91, KGE approaching 0.80, and correlation coefficient higher than 0.96. Daily projections of precipitation and temperature under all four shared socioeconomic pathways were obtained from three GCM models and were further employed to project future potential evaporation. The BMA-based ensemble modeling system was then used to forecast annual maximum flood rates and associated 3-day maximum flood volumes in the future. Our results show that the three GCM models exhibit considerable differences in terms of future flood projections, but all indicate a general increase in flood occurrence and magnitude under future climate change scenarios. The future daily flood events under different climate scenarios are likely to become more severe, as indicated by higher mean, maximum, and 90th quantile values of the AMAX flood series. Meanwhile, the corresponding 3-day flood volumes show varying patterns in terms of mean and extreme flood volumes under different scenarios, but we would have more chances to experience severe 3-day flood volumes in future. The results of our study can provide important information for flood risk management and adaptation planning in the River Thames basin.