Research On Monitoring And Predicting Methods Of Land Subsidence In Coastal Zone

With developed economy and large population,the land-sea transition zone is an extremely important area for human production and life from a global perspective.Due to the natural processes of river sand transport,natural sedimentation,weathering and stripping,as well as the long-term effects of human activities such as land reclamation,coastal transformation and offshore engineering,the topography and geomorphology of the land-sea transition zone are complex and varied,and disasters such as land subsidence,seawater intrusion and storm surge occur frequently,which have seriously threatened the safety of life and property and sustainable social development in the region.Rapid land subsidence is a typical geohazard in the land-sea transition zone,but how to monitor and warn the land subsidence in the coastal zone area has been a hot and difficult problem for research.To this end,this paper takes Hangzhou Bay Bridge area as an example and carries out the research on the land subsidence pattern and prediction in the sea-land transition zone area based on In SAR technology.The main research contents and results of the paper are as follows:(1)The synthetic aperture interferometric radar(In SAR)technology approach for land subsidence monitoring is systematically discussed.For the processing and analysis of long time series of In SAR data,timely,continuous and long-term deformation monitoring of the land surface can be carried out,and permanent scatterer points can be extracted as a reference datum,and on this basis,the deformation time series data can be modeled,and then potential land subsidence problems can be analyzed,providing a reliable technical means for the monitoring and prediction of land subsidence.(2)A time series prediction technique method HLA(Hybird LSTM and ARIMA)with hybrid LSTM and ARIMA is established.To address the difficult problem that land subsidence cannot be accurately predicted in the complex and variable environment of the land-sea transition zone,a time series prediction method HLA combining the Auto Regressive Integrated Moving Average(ARIMA)model with the Long Short-Term Memory(LSTM)model in deep learning is proposed for the first time,i.e.,the time series of form variables obtained by using the time series In SAR technique and the time series predicted by the ARIMA model are The difference is then trained on this difference time series using LSTM,and finally the prediction of land subsidence is performed.(3)An analysis and prediction study of land subsidence in the Hangzhou Bay area was conducted.Taking the In SAR monitoring data of Hangzhou Bay from 2017-2019 as an example,the HLA prediction method established in this paper was used for land subsidence analysis and prediction research,whose results showed that compared with a single prediction algorithm,the root mean square error(RMSE)of the method proposed in this paper was reduced by at least 28.51%,the mean absolute error(MAE)was reduced by at least 19.89%,and the average prediction accuracy is improved by at least 15.19%,which verifies the feasibility of the method in this paper and shows a good application prospect.