Abstract: Based on the visibility data in 2009 and relative humidity (RH), PM₁₀ and PM_2.5 data in Tianjin, their relationships were analyzed. The results indicate that the linear correlation coefficient between RH and visibility is the highest and that between PM_2.5 and visibility ranks second. The visibility decreases obviously with the increase of RH. When RH is less than 60%, there exists a significant nonlinear correlation between visibility and PM_2.5 concentration and a weak correlation between visibility and PM₁₀ concentration, while the correlation between the visibility and the concentration difference between PM₁₀ and PM_2.5 is not significant. When RH is greater than 60%, the nonlinear correlation between visibility and PM₁₀ concentration is more significant than that between the visibility and PM_2.5 concentration. The nonlinear correlation efficient between the visibility and RH is higher than their corresponding linear correlation efficient. Finally, a nonlinear fitting formula of visibility in Tianjin is developed using RH, PM₁₀ and PM_2.5.  It has been shown through examination against the observation that this formula is able to simulate the visibility in Tianjin with a relatively good accuracy.

Key words: Atmospheric visibility, PM₁₀, PM_2.5, Relative humidity, Nonlinear correlation