Based on the global radiation data, air temperature data, ground temperature data, relative humidity data and PM_2.5 concentration data from February 2013 to March 2014 in Hefei of Anhui province, the relationship among PM_2.5 concentration, global radiation, air temperature, ground temperature was analyzed. Results show that (1) The concentration of PM_2.5 is an important human factor to affect the global radiation in urban area. In the midday of cloud-free conditions, the global radiation and PM_2.5 concentration presented strong negative correlation, the correlation coefficient reaches -0.62. The correlation coefficient between normalized global radiation and PM_2.5 is-0.76, which is small in the morning and evening. On the average, when PM_2.5 concentration increases by 1 μg·m^-3, the global radiation declines by 0.92 W·m^-2 in the daytime of the cloud-free conditions. (2) In the daytime of the cloud-free conditions, air temperature, ground temperature and PM_2.5 concentration have obvious negative correlation. The effect of PM_2.5 concentration on the ground temperature is greater than that on air temperature. The effects in summer is greater than those in other seasons. The average slopes of linear fitting line between air temperature, ground temperature and PM_2.5 concentration are -0.12 and -0.022, respectively. It is also found that when PM_2.5 concentration increases by 10 μg·m^-3, ground temperature and air temperature on the average, fall 0.22℃ and 1.2℃, respectively. (3) On the weather scale, PM_2.5 concentration has obvious influences on global radiation, ground temperature and air temperature. In September and December,1 μg·m^-3 of PM_2.5 concentration increasing would cause the global radiation decline by 1.8 W·m^-2 and 0.5 W·m^-2, ground temperature drop 0.11℃ and 0.02℃, the air temperature drop 0.03℃ and 0.01℃. The air-pollution condition should be considered in weather forecasting.