Data Assimilation Of Cellular Automata Model For River Water Quality

Rivers contain nearly 75%of the water on the earth’s surface.As an important flow channel of surface water,rivers undertake the task of transporting water to various places and play an irreplaceable important role in the water cycle.River water quality determines the quality of people’s production and life to a great extent.However,due to the global climate change and the increase of human activities,the prospect of river water quality is not optimistic,and river water quality management is still an urgent problem in the field of water environment.Observation and simulation are two indispensable methods in river water quality management at present,but they have their own advantages and disadvantages.The water quality data obtained from observation is highly accurate,but it is difficult to achieve spatiotemporal continuity,which does not meet the requirements of river water quality dynamic management.Simulation can obtain spatiotemporal continuity data,but the accuracy is low.Therefore,how to integrate the advantages of the two aspects to obtain more accurate spatiotemporal continuous water quality data is of great significance for river water environment management.In this paper,the construction of water quality model and the data assimilation of the model are studied.The research contents include:(1)Cellular automata model(CA)was used to simulate water quality.The transport and diffusion process of pollutants in rivers was expressed by CA model,and the conversion rules were divided into flow movement rules and transport and diffusion rules.The flow movement rules were established based on the hydraulic equation and logic rules.The transport and diffusion rules of pollutants are established based on the diffusion equation under the premise of considering the influence of water movement.Finally,the compilation and integration of the CA model of river water quality is realized on the Matlab platform to realize the model input,the calculation of pollutant transport and diffusion and the output of the results.(2)Combination of data assimilation algorithm and CA model of river water quality.Starting from the selection of parameters of ensemble Kalman filter,the steps of data assimilation for CA model are introduced.With pollutant concentration as the state variable and measured pollutant concentration data from river water quality monitoring stations as the observation data,the set size was set as 30,the CA model error was set as 20%,and the observation error was set as 30%.The degree to which data assimilation technology could improve the CA model accuracy of river water quality was initially explored.(3)Research on model application.The water quality CA model and its corresponding data assimilation scheme were applied in the section from Muli Zha to Huangtaiqiao of the main stream of Xiaoqing River to simulate the concentration of COD.According to the observation of the overall COD concentration distribution map of the river and the concentration curve of individual monitoring stations,it is found that the water quality simulation results accord with the actual diffusion law,and after comparison of the measured COD concentration value,it is found that the simulation error is within a reasonable range.After combining ensemble Kalman filter with CA model,it is found that when observed data appear,the corrected simulation results jump,making the simulation trajectory as a whole close to the real situation,and the COD concentration value closer to the measured data,indicating that data assimilation is an effective method to improve the accuracy of CA model of river water quality.In general,the CA Model avoids the complicated solution process of the traditional water quality model based on physical equations,has high calculation efficiency,and can easily adjust the conversion rules and parameters of each part according to the actual river conditions,and has practical value.In addition,the use of data assimilation technology can improve the accuracy of water quality simulation,which has a good research prospect in the future water quality precise management.