Forecast Model of Xiangtan Air Quality Based on Meteorological Factors
In order to improve the forecast quality of AQI(air quality index)in Xiangtan,this analyzed the daily observations of AQI,the concentrations of three major pollutants(PM2.5,PM10,O3-8h),and six selected meteorological factors from January 1,2016 to December 31,2022.The temporal characteristics of AQI and three major pollutants were studied.The correlation and correlation coefficient between AQI,CPM2.5,CPM10 and CO3-8h and six selected meteorological factors were processed by SPSS software.The AQI forecasting model was constructed by multiple stepwise linear regression method,which was tested and evaluated.Finally,an AQI forecasting method based on meteorological factors,which was suitable for Xiangtan City and met the actual business needs,was established.The results show that:(1)The concentration of PM10 is the highest in spring,summer and autumn,followed by O3-8h.It is the highest concentration of PM2.5 in winter,followed by O3-8h.(2)The AQI is highly correlated with PM2.5 concentration and moderately correlated with O3-8h concentration and PM10 concentration.The PM2.5 concentration is moderately correlated with dai-ly maximum temperature,O3-8h concentration is moderately correlated with relative humidity,and PM10 concentration is moderately correlated with relative humidity and sunshine duration.(3)The accuracy of forecasting model without distinguishing between sea-sons is 55.62%.The accuracy of forecasting model in different seasons is 59.18%.(4)The AQI forecasting model in different seasons has been put into trial operation,and the classification forecasting accuracy rate reaches 68.68%.This model is of great application value to the air quality forecast business in Xiangtan.
Xiangtanmeteorological factorsAQIstepwise linear regression method
NETL
NSTL
万方数据
