The Assimilation Simulation Study Of The Total Suspended Solids Concentration In Taihu Lake Based On The Initial Disturbance Of ETKF

Under rapid urbanization,the water environment in Taihu Lake has deteriorated and the total amount of pollutants has remained high.Sediment,one of the main influencing factors of Taihu Lake water environment,can be used as a carrier of water pollutants.The mechanism of sediment formation and its dynamic evolution process has been a hot topic in related fields.Among the current research methods of sediment,the dynamic model simulation can effectively describe the spatiotemporal dynamic evolution process of sediment.However,due to the large number of water environment parameters and the complex mechanism in long-term sequential simulation,the dynamic model is often generalized by simplified algorithms,and the accuracy of the simulation results still needs to be improved.The completing Taihu Lake water quality monitoring network can obtain monitoring data from buoys in real time at high frequency,thus provide the input data for designing a Taihu Lake sediment data assimilation method.In the existing data assimilation method,the widely-adopted Ensemble Kalman Filter(EnKF)avoids the complicated calculations accompanying the pattern.The EnKF assimilation method uses the Monte Carlo(MC)method to generate the initial perturbation field.The bottleneck problems of this method include the excessively low error growth rate,and the unclear meaning for dynamics.These problems restrict the quality of assimilation simulation.The Ensemble Transform Kalman Filter(ETKF),as the initial perturbation method,can introduce a transformation matrix to adjust the set perturbation,which can reduce the influence of the initial perturbation on the simulation results,and thus become one of the popular initial perturbation methods in recent years.This study takes the total sediment of Taihu Lake as the research object,studies and designs an initial perturbation field generation method of sediment simulation in Taihu Lake,and constructs a coupled simulation model of sediment assimilation in Taihu Lake based on EnKF.A case-based analysis on different initial perturbation methods based on high-frequency buoy monitoring data of the sediment of Taihu Lake has been carried out.The main research contents and results of this paper are as follows:(1)Method for generating initial perturbation field in assimilation simulation of Taihu Lake water body:Focusing on the low error growth rate problem of EnKF while generating the initial perturbation field using the MC method,the concentration distribution and evolution of the sediment in the Taihu Lake are studied.Based on the ETKF initial perturbation method,the concept of perturbation expansion factor is introduced to construct the covariance transformation matrix.The initial perturbation field generation method of sediment assimilation simulation in Taihu Lake is studied to provides a theoretical basis for the simulation analysis of assimilation coupling model.(2)Study on simulation mechanism of coupled model of sediment assimilation in Taihu Lake:In this study,the data organization scheme and implementation flow of the coupled model of sediment assimilation in Taihu Lake are designed.The parameterization mechanism and model-driven coupling method are studied,and the hydrodynamic sediment model and the Ensemble Kalman Filter assimilation based on ETKF initial perturbation method are realized.The coupled simulation scheme can reproduce the dynamic evolution process of sediment in Taihu Lake and provide decision support for the treatment of water pollution in Taihu Lake.(3)Case analysis of sediment assimilation of Taihu Lake:The case analysis is carried out on the buoy monitoring data collected on Taihu Lake,using the assimilation coupling simulation model constructed in this study based on EnKF assimilation method and three-dimensional sediment hydrodynamic model.A comparative case analysis based on the dynamic BGM and the ETKF initial perturbation methods is completed with the EnKF assimilation model simulation.Results show that the three methods improve the simulation accuracy of FVCOM model to different extents,with the most precise method being the assimilation simulation results based on ETKF initial perturbation method.