A Study On The Method Of Air Pollution Forecast In Yinchuan City

A study on the method of air pollution forecast in Yinchuan CityThis dissertation analyses the air quality present situation, air pollution-meteorology condition and each kind of pollutant characteristic space-time distribution in Yinchuan city during 2001-2005, promulgating the quota or Yinchuan air pollutant change rule, and has established the air pollution model of numerical forecasting, dynamic-statistics and multiple linear regression analysis in this foundation. The results show that:1. The Yinchuan City year-leadership wind direction is the leaning north wind, summer primarily will lead the wind direction by the south wind, and it is leaning north wind at the end of the fall to the next spring;2. In a year, the Yinchuan city atmospheric stability by the neutrality and the stable kind primarily, the unstable kind of weather is easy to appear in the summer, but in the fall to winter season it is easy to appear the stable kind of weather;3. The high pollution time interval of main pollutant appears in the morning and night. PM₁₀ and SO₂ pollution morning is higher than the evening, the NO₂ pollution then just right is opposite. The PM₁₀ spring pollutes strongly, SO₂ and NO₂ winter pollution is strongest.4. The natural precipitation is extremely obvious to the atmospheric clean contribution function.In each kind of sand dust weather process, the floating dust weather causes PM₁₀ to pollute heavily, raises the sand weather to be next, the sand storm weather relative is lighter;5. The the forecast effect of dynamic-statistics pattern in the Yinchuan air pollution is most remarkable, has made up the numerical forecasting and the pure mathematical statistic forecast pattern insufficiency.Regarding the gas pollutant (SO₂, NO₂) the value pattern relative error is smaller than multiple linear regression analysis.But the pellet pollutant (PM₁₀), the numerical pattern forecast ability weaker than the multiple linear regression pattern;6. In the dynamic-statistics pattern predictor both has considered the meteorological condition function, and the pollutant density function.Compares with the pure statistical forecast method, it has the more credible physical foundation. And compares with the numerical pattern, the dynamic-statistics forecast pattern does not need the source of pollution strongly to discharge the detailed list, so it has the simpleness merit.