Changes in future ice-jam flood severity of a regulated cascade reservoir-river system

Ice-jam flooding (IJF) is a complex and hazardous phenomenon that can result in devastating impacts on riverine communities and infrastructure. With the increasing concern about climate change and its potential implications on hydrological processes, there is a growing need to assess the future risks of IJF. In this study, we developed an integrated probabilistic modelling framework that combines a river ice model with bias-corrected CMIP6-GCMs data and a machine learning model to investigate the impacts of climate change and flow regulation on IJF backwater levels. The simulation of the IJF employs extensive Monte Carlo analysis (MOCA) simulations to capture a wide range of historical (1970–2019) and future (2020–2069) scenarios in the Sanhuhekou bend reach (SBR) of the Yellow River. Firstly, the results reveal that current flow regulation plays a significant role in increasing IJF backwater levels, outweighing the effects of 3 future climate scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5), which tended to reduce IJF backwater levels. Secondly, the findings suggest that any caps on regulated flows during the ice period may be eased in the future to partially meet growing demand for hydropower. Finally, the flow regulation strategy adapted from the MOCA simulation results can both increase hydropower generation and meet the high flood design criteria under SSP5-8.5 scenario.