MULTIPLE LINEAR REGRESSION INVERSION MODELING OF WATER POLLUTION SOURCES BASED ON PCA

In order to accurately analyze the regional water pollution sources, the multiple linear regression inversion modeling method of water pollution sources based on PCA is studied. According to the relevant standards, water quality samples are collected in the study area, and a total of 13 indicators such as sodium, calcium, magnesium and iron are selected for multiple linear regression analysis to characterize the overall water quality of the study area. On this basis, the principal component analysis method is used to reduce the dimension of each indicator and determine the main factor classification of water quality indicators. Multiple linear regression method is used to build the inversion model, calculate the contribution of different main factors to water quality, extract the mathematical characteristics of water quality indicators, and determine the main factors affecting the groundwater quality in the study area and their corresponding pollution sources. The results show that the four main pollution factors in the study area are leaching and migration factor (contribution rate 35.32%), agricultural activity pollution factor (contribution rate 21.23%), geological environment background factor (contribution rate 14.49%) and industrial activity influence factor (contribution rate 9.07%); the linear fitting values between the analysis results of the proposed method and the actual values are all above 0.8, and the analysis results are consistent. The ratio between the predicted results and the actual values is close to 1, which indicates that the predicted results are in good agreement with the actual values, and the calculation and allocation of regional water pollution sources have good applicability.