预测银川市大气主要污染物浓度的一种动力统计模型

A dynamic- statistics model for predicting atmospheric main pollutant consistence in Yinchuan city

本文以箱模式为原理,从大气污染物平流扩散方程出发,经过简化推导,得出预测大气主要污染物浓度的动力统计模型。在模型中既考虑了气象条件的作用,又考虑了起报日的污染物浓度,与纯数学统计模型相比,有着更可信的物理基础;与数值模式相比,本模型不需要污染源排放清单,具有简便易行的优点。利用银川市2001-2004年大气主要污染物SO_2、NO_2、PM_(10)逐日平均浓度监测值和同期地面气象要素资料,通过回归分析确定了主要污染物浓度24 h变率预报方程,经2005年6～9月的预报检验,本模型对银川市大气主要污染物浓度的变化有一定的预报能力。

Environment and development, it is current societyg generally concerned problem. After experiencing all kinds of harm of the environmental pollution to the mankind, people begin to pay attention to environmental protection, set about carrying on the prediction research of the pollution accident, meanwhile predict the air pollution. With more and more aroused more interests of the public to air quality, many areas in China has all developed the air quality daily paper and the forecast work . The air pollution prediction is a cross subject, need to consider synthetically topographical ground form, meteorological condition, pollution sources discharging law, etc. At present, there are two kinds of prediction to be adopted that are divided into statistical prediction and numerical prediction. The weak point of statistical prediction lies in the change mechanism that be unable to announce the pollutant objectively, lack solid physics, chemical foundation; And the numerical one needs complete meteorological mode and atmospheric chemical reaction mode as its theoretical foundation, the present stage because initial field mode need terms difficult to provide, cause the output result of numeric prediction is not to be high . This paper regards boxmodel as the principle means, sets out from the spread equation of the laminar flow of air pollutant, through simplified ratiocination, and then obtain the dynamic-statistical model on predicting main atmospheric pollutant consistence. Box model is a fixed number of urban space as a fixed or a box fabric study, the average concentration of air pollutants and changes over time. Box model is the starting point pollution conserved hypothetical pill, the mixture of pollutants in the fabric. The model not only has considered the function of the meteorological condition promptly, but also has considered the consistence of beginning date to predict the pollutant. Compared with the way of pure mathematics-statistics, there are more believable physical foundations; Compared with numerical mode, this model period day by day of 2001- 2004 year in Yinchuan city, through analyzing that has found the main city atmospheric data of the same pollutant consistence predicting equation of 24 h becoming rate. After examining by prediction for June to September in 2005, the model has shown that it has certain prediction ability to Yinchuan city atmospheric main pollutant change of consistence.