Assessment and Prediction of Jacking Effects On Coastal Cities Under Compound Flooding: A Physics-based Modeling Approach
摘要
As climate change intensifies, the jacking effect caused by rising tides and river levels increasingly impedes coastal drainage systems, potentially exacerbates compound flood risks in coastal urban areas, but the mechanism of jacking effects on drainage networks remains incompletely understood. In this study, we take the area of Meishe River Basin in Haikou City as an example, a coupled compound flooding physical model with particle tracking methodology is constructed to analyze the characteristics of jacking effect under high tide and river levels, and an LSTM model is developed to rapidly predict the jacking extent of the pipe network and the jacking length of vulnerable pipelines. The results show that the jacking effect is significant in the right-bank drainage network of the Meishe River, while the flow inter-feedback between tide-river-drainage is weaker on the left bank due to topographic features, underlying surface conditions, and boundary constraints. The jacking effect is negatively correlated with rainfall intensity. When river level rises, the jacking effect intensifies, with the maximum jacking extent under the 20-year rainfall and 50-year rainfall being 82.58% and 81.69%. As tide level increases, the duration and extent of jacking effect decrease. The LSTM-based prediction model for jacking effect provides reliable predictions and outperforms, for the jacking extent of drainage networks, with the coefficient of determination (R2) of 0.943, and the prediction time for 1-h test scenarios is within 10 s, significantly improving efficiency. This study provides a theoretical basis for quantifying the jacking characteristics of high tides and river levels in drainage networks, and predicts the extent of jacking effect and vulnerable pipeline locations, offering important references for urban flood disaster warning and mitigation applications.