Statistical Analysis Of Haze In Shanghai

With the development of the society, environmental pollution becomes more serious. And haze, as one of the air pollution, has had a bad impact on air quality. This article studies the characteristics of the haze and the reason why haze happens by the meteorological data and particle data in Shanghai from Dec.2009to Nov.2010. And the regression equation of PM2.5is established on the basis of the study of haze. At last, the conceptual discriminant model is built. The main work and results are as follows:(1) The statistics of the meteorological factors in haze days. The statistical characteristics of relative humidity, wind direction, wind speed and weather situations in haze days are as follows:The range of relative humidity in haze days is always from60%to90%. The wind speed is from0to2m/s in about75%haze days, and is from2m/s to4m/s in about25%haze days. The wind direction is always from180皌o360°n haze days and sometimes from0°o90°. And it can be found that haze is easy to happen in2kinds of weather situations. One is saddle field. Another is weak cold high pressure field.(2) The establishment of the regression equation of PM2.5. It can be found through the correlation analysis that PM2.5has a strong relationship to the meteorological factors of temperature, dew point, wind speed, wind direction, visibility, pressure and relative humidity. Thus, the monthly regression equation of PM2.5can be established whose independent variables are meteorological factors. The significant test of regression equation passes, which means the regression equation of PM2.5can be used.(3) The building of the conceptual discriminant model of haze. The conceptual discriminant model of haze is made up in3parts. The first part is the probability of the happing of haze. It's measured by relative humidity, wind speed and wind direction. The second part is the rank of the haze. Learning from the establishment of API, the haze index is built by visibility and the concentration of PM2.5. According to the value of the haze index, the degree of the haze can be decided. The third part is the type of the haze. OC/EC, NO3-/SO42-are used to distinguish the haze type.Through the research, the characteristics of haze and why haze happens can be learned. And the establishment of regression of PM2.5and the building of haze discriminant model make haze forecast possible.