Evaluation And Forecast Analysis Of Sanchuan River Water Quality In Luliang City

With the rapid development of China's economy,people's living standards are also improving steadily,but the problem of excessive development and utilization of water resources is very serious,which brings great challenges to the protection of water resources.Luliang city is rich in coal resources,but in the process of exploitation and use,it has brought a series of environmental pollution problems to the local area.As an important part of the surface water body of luliang city,sanchuan river has a particularly prominent water quality problem.Improving the water quality environment of sanchuan river and protecting sanchuan river are conducive to the sustainable development of social economy of luliang city.Based on the water quality monitoring data of the two state-controlled sections of the sanchuan river in luliang city,as well as the provincial and municipal water quality monitoring data of the last four years from 2016 to 2019,and combined with a variety of statistical software,this paper evaluated and predicted the water quality changes of the sanchuan river,providing scientific solutions for improving the water environment quality of the sanchuan river.This paper is divided into two parts:Evaluation: based on SPSS software,principal component analysis method was adopted to evaluate the water environment quality of sanchuan river from the perspectives of time and space,and the change of water quality of the section was analyzed according to the principal component scores of each section.Three principal components,namely three comprehensive indicators,can be extracted from the seven water quality indicators selected under the principal component analysis method to study the water quality change of sanchuan river,all of which can pass the KMO test and Bartlett spherical test.Through the analysis of principal component score table shows in three rivers of dangerous levels of ammonia nitrogen and total nitrogen is much higher than ? class water quality standard,the total phosphorus concentration rising sharply at the beginning of the 19 th,other indicators,especially in dissolved oxygen,COD levels continued to decline.The water pollution near the zhaidong bridge section in the middle reaches of the sanchuan river is the most serious,and the water pollution at the west cliff bottom section of the upper reaches has obvious seasonal differences.Prediction: on the basis of principal component analysis,ammonia nitrogen and COD indexes were selected to make short-term prediction on water quality of sanchuanhe zhai east bridge section in luliang city,and two methods of time series ARIMA model and BP neural network were adopted to predict water quality,and the accuracy of the two models was compared.Based on the water quality model of time series ARIMA(1,1,0),the seasonal model was also fitted for prediction.The average error of ammonia nitrogen and COD was about 1.58 and 2.16,respectively.In the BP neural network water quality model based on MATLAB software,the improved BP algorithm was adopted,and the average error of ammonia nitrogen and COD reached 0.898 and 1.33,respectively.The prediction results showed that the pollution level of ammonia nitrogen rose first and then slightly declined in a short period,and the level of COD continued to decrease.Two kinds of prediction model in the same conclusion,but improve the accuracy of the algorithm of BP neural network model to predict water quality is higher,the gradient descent algorithm of adaptive adjustment vector,vector algorithm with the adaptive adjustment of the change of the error,which makes the result more accurate,which has higher reference value.Sanchuan river waters the overall pollution is still severe,the water of water even worse ? class,should strengthen the water management work,strive to reduce sanchuan river pollution of ammonia nitrogen and total nitrogen level,effectively raise the quality of sanchuan river water environment.