Study On Intelligent Forecasting And Risk Controlling Of Flood And Storm Surge At Xiaoqing River Estuary

Due to the special geographical location and hydrometeorological conditions,the Xiaoqing River estuary has frequent flood tidal disasters and serious disaster losses,posing a huge threat to coastal economic development and the safety of people’s lives and properties.Scientific prevention and control of flood tidal disasters is essential.Therefore,in view of the practical problems such as the high disaster intensity,serious disaster losses and difficulty in risk control of the flood tide disaster at the Xiaoqing River estuary,this paper,on the basis of comprehensively considering the influencing factors of storm surge,carries out intelligent forecasting of storm surge,risk analysis of flood tide encounter,numerical simulation of flood and tide coupled dyke break flood and research on flood tide risk control.The specific research contents and results are as follows:(1)Considering the influence of typhoon factors,meteorological factors and astronomical tide factors on the storm surge comprehensively,combined with the respective advantages of the LSTM neural network and the GM model,an intelligent storm surge prediction model based on the LSTM-GM neural network model is established.The storm surge forecasting model is carried out at the Xiaoqing River estuary.The results show that the LSTM-GM neural network model has better prediction results,with NSE of 0.98,RMSE of 0.035,and MAPE of 0.025.All indicators are significantly better than the LSTM model and the BP neural network model.The model can realize accurate and stable forecast of storm surge at the Xiaoqing River estuary.(2)Based on the discussion on the relationship between the annual maximum24 h rainfall in the whole basin of Xiaoqing River and the highest high tide level in the flood season of Yangjiaogou station,the Copula function is used to analyze the combined risk of flood-tidal encounter with different design return period combinations in the Xiaoqing River estuary under flood-dominated and tidaldominated situations.The results show that when the return period of the maximum24-h rainfall in the year is constant,the risk of encountering the highest high tide level in the flood season becomes lower and lower with the improvement of the tide level standard.When the return period of the highest high tide level in the flood season is certain,the risk of its encounter with the annual maximum 24-h rainfall becomes lower and lower with the improvement of the rainstorm standard.When the design return period of flood tide encounter is certain,the risk of flood tide encounter in the case of flood tide is always greater than the flood tide encounter risk in the case of flood tide.(3)Zihe,Luqinggou and Xitaoyuan Village are selected as the study areas to construct a coupled numerical simulation model of dyke-break-flood in the tidesensing reaches of the Xiaoqing River considering the flood tide.Through model calibration and verification,the dynamic evolution process of dyke break inundation at the Xiaoqing River estuary under different design return periods of floods and tide levels are simulated.The results show that the model calibration results in the Zihe study area are consistent with the actual situation,and can meet the needs of floodtidal coupled breakwater flood evolution simulation and flood-tidal risk analysis.In the results,the inundation duration under different design floods and tidal return periods is more than 3 days,and the areas with submerged water depths greater than 1m are mainly concentrated in the areas near the breach and in the areas with lower terrain,which indicates that the terrain of the study area is relatively flat,the slope is gentle,and with the passage of the flood evolution time coupled with the flood tide,the evolution rate of the flood tide gradually decreases,and the flood tide submersion lasts too long.(4)Based on the calculation results of the flood tide encountering embankment breaching flood coupling model,the entropy weight-TOPSIS method is used to carry out a study on the risk assessment of the flood tide encountering embankment breaching in the Xiaoqing River estuary,and the spatial distribution characteristics of flood-tidal disaster risks in the tide-sensing reaches are analyzed.The target particle swarm optimization algorithm is used to solve the risk control model of dyke break and inundation,and the risk control effect is analyzed.The results show that the area proportion of extremely high risk areas in the study area of Xitaoyuan Village is higher than that of Luqinggou,and the proportion of extremely high risk areas is23.72% and 5.78% respectively.Under the control of the risk of dyke break and inundation,the proportion of extremely high-risk areas in the study area of Xitaoyuan Village and Luqinggou is reduced to 22.65% and 4.18%,respectively.