Long-Term Trends in the Probability of Different Grades of Haze Days in China from 1960 to 2013

The number of haze days and daily visibility data for 543 stations in China were used to define the probabilities of four grades of haze days:slight haze(SLH)days;light haze(LIH)days;moderate haze(MOH)days;and severe haze(SEH)days.The change trends of the four grades of haze were investigated and the following results were obtained.The highest probability was obtained for SLH days(95.138%),which showed a decreasing trend over the last54 years with the fastest rate of decrease of-0.903%·(10 years)-1 and a trend coefficient of-0.699,passing the 99.9%confidence level.The probabilities of LIH and MOH days increased steadily,whereas the probability of SEH days showed a slight downward trend during that period.The increasing probability of SLH days was mainly distributed to the east of 105°E and the south of 42°N and the highest value of the trend coefficient was located in the Pearl River Delta and Yangtze River Delta regions.The increasing probability of LIH days was mainly distributed in eastern China and the southeastern coastal region.The probabilities of MOH and SEH days was similar to the probability of LIH days.An analysis of the four grades of haze days in cities with different sizes suggested that the probability of SLH days in large cities and medium cities clearly decreased during the last 54 years.However,the probabilities of LIH days was<10%and increased steadily.The probability of MOH days showed a clear interdecadal fluctuation and the probability of SEH days showed a weak upward trend.The probability of SLH days in small cities within 0.8°of large or medium cities decreased steadily,but the probability of LIH and MOH days clearly increased,which might be attributed to the impact of large and medium cities.The probability of SLH days in small cities>1.5°from a large or medium city showed an increasing trend and reached 100%after 1990;the probability of the other three grades was small and decreased significantly.