This article conducts correlation analysis on the CUACE ground dust concentration and air quality AQI data developed by the China Meteorological Administration from June 2021 to July 2022.The K-means clustering algorithm is used to divide the CUACE dust concentration into five layers according to 300,1 000,3 000,and 5 000.A prediction model for air quality level is established based on a random forest model.The results show that the larger the CUACE,the greater the proportion of data with AQI ≥ 300;When CUACE≥5 000,all AQIs are greater than 300.When 5 000＞CUACE ≥ 3 000,AQI undergoes two-stage differentiation,and the data volume of AQI between 200-400 does not exceed 1％;In the data from the first to third layers,the proportion of AQI＜100 has significantly increased,accounting for 86.2％,68％,and 41.8％,respectively;The overall performance of the model is better than that of CUACE,with the fifth and fourth layers exhibiting better predictive performance.
万方数据
维普
