Taking seven national weather stations in Hebei Province as an example, the mixed layer height, ventilation capacity and atmospheric selfcleaning ability index are calculated using ERA5 cloud data and measured cloud data from 1979 to 2013, and a preliminarily assessment of ERA5 reanalysis cloud data is conducted through comparing the results by two sets of data calculated. ERA5 cloud cover is used to calculate the atmospheric self-cleaning capacity index from March 2013 to February 2021, and the relationship between it and PM_2.5 concentration is analyzed. The results are as follows: (1) The monthly average relative errors between the mixed layer height, ventilation capacity, and atmospheric self-cleaning ability index calculated by using two sets of cloud data approximately accord with the normal distribution, the highest frequency occurring in 0-5%; the monthly relative error of the mixed layer height > ventilation capacity > atmospheric self-cleaning ability index; and the correlation is opposite. The relative errors (correlation coefficients) of autumn and winter are mostly lower (higher) in spring and summer. The monthly mean relative errors of Xingtai, Shijiazhuang and Baoding are significantly smaller than those of other stations. (2) The annual average relative error of the atmospheric self-cleaning ability index is the smallest, followed by ventilation capacity. The fitting degree and deviation of the annual values in the south-central stations are also more ideal. (3) From March 2013 to February 2021, the monthly average atmospheric self-cleaning ability index and PM_2.5 concentration are inversely correlated at all stations. The correlativeness is more obvious in the central and southern stations.