- 期刊
Based Mixed Mathematical Model of Contaminant Related Meteorological Factors Used In Indicator System
摘要
This paper based on the pollutant concentration and meteorological data of the central and southern cities of Hebei Province, this paper analyzes the main meteorological indicators affecting the concentration of various pollutants, and proposes a multi-index screening method for the influence factors of pollutant concentration. First, combined with a variety of data visualization methods, it is found that the relationship between the same pollutant and its meteorological impact factors is different in different cities or different seasons. Subsequently, by plotting the linear correlation matrix heat maps of six pollutants and 13 meteorological indicators, it is found that the same meteorological impact factor has different effects on different pollutant concentrations. Huge difference. Moreover, the intrinsic function relationship between different pollutant concentrations and various meteorological factors is also complex and variable. Based on the above analysis, considering the comprehensive use of Pearson correlation coefficient, Kendall correlation coefficient and maximum information coefficient MIC, which have different correlations between measurement variables, an automatic and efficient multiindicator screening method for pollutant concentration meteorological impact factors is proposed. Finally, taking Shijiazhuang and Xingtai as examples, the screening results of major meteorological impact factors corresponding to 48 combinations of two different cities, four different seasons and six different pollutants were given.