Source Identification Of River Sudden Water Pollution Based On CA-MCMC Method

Now our country is in an important period of industrial structure transformation and upgrading.The rapid development of industrialization and urbanization not only brings huge economic and social benefits,but also aggravates the plight of environmental pollution in our country.In the process of industrial production,the water pollution,air pollution and solid waste pollution caused by illegal and excessive discharge are becoming increasingly prominent.According to relevant statistics in recent years,water pollution accidents account for more than half of the total number of environmental pollution accidents in China,especially the sudden surface water pollution accounts for the highest proportion and the greatest impact.Therefore,we should focus on strengthening the source control and rapid response to the sudden water pollution events,improving the emergency monitoring ability of the emergency events,so as to timely and accurately lock the pollution source information,evaluate the level and impact scope of pollution accidents,and provide strong support for scientific decision-making.In this paper,aiming at the problem of Pollution Tracing of river burst water,a model of pollutant migration and diffusion based on cellular automata(CA)is constructed,and a new method and framework of CA-MCMC tracing is proposed.The main work is as follows:(1)The diffusion and migration mechanism of pollutants in different emission types and diffusion stages are discussed.On this basis,one-dimensional and two-dimensional models of instantaneous and continuous discharge of sudden point source pollutants are established.At the same time,the principle of pollutant traceability is described,and the specific solution methods of different parameters to be traced are given.(2)In order to reduce the uncertainty of the traceability solution,this paper studies the traceability algorithm of river sudden water pollution based on Markov Monte Carlo method from the perspective of Bayesian reasoning.In this paper,the problem of pollutant traceability is transformed into Bayesian estimation problem,and the estimation value of traceability solution is described by probability distribution through historical prior information and observation value sequence of monitoring section,combined with two-dimensional mathematical model of pollutant migration and diffusion.(3)In order to improve the sampling efficiency of Markovian Monte Carlo method,the commonly used standard M-H sampling method is improved.When generating sample values according to the recommended distribution,a condition is added to judge whether the sample values meet the posterior probability density function,so that the sample interval can approach the real value faster and more accurately.(4)The model of pollutant migration and diffusion is integrated with the model of pollutant traceability.In order to improve the accuracy of traceability and reduce the uncertainty of pollutant diffusion model,a CA model was constructed to simulate the process of pollutant migration and diffusion.The results show that the CA model has good applicability for different emission scenarios,can reproduce the process of pollutant transport,and improve the accuracy and efficiency of the traceability solution.(5)In order to verify the applicability and reliability of CA-MCMC traceability method,two kinds of model experiments of continuous discharge and instantaneous discharge of point source pollutants are designed and carried out.The traceability results of CA-MCMC method are compared with the measured values to verify the applicability of the traceability method.