大气污染控制有效性的评估模型及应用

基于小波分析理论及BP神经网络,建立了大气污染控制有效性的评估模型.利用该模型解析天津市污染源排放及气象因素对大气污染物环境浓度的影响.结果表明,SO2浓度波动主要由污染源的季节变化引起,与天气过程有关的短期波动也不容忽视;而PM10浓度波动主要由污染源排放及气象条件的短期变化引起.2002~2003年的气象条件总体上不利于大气扩散,使SO2日均浓度长期分量增加约为3μg/m3;而2004年的气象条件有利于大气扩散,使SO2日均浓度长期分量减小约为5μg/m3.PM10日均浓度长期分量也有类似的规律,2002~2003年的不利气象条件使PM10日均浓度长期分量增高更显著,约为10μg/m3.

Evaluation model for the effectiveness of air pollution control and its application

A model for evaluating the effectiveness of air pollution control was presented based on wavelet analysis theory and backpropagation neural network. The influence of source emissions and meteorological factors on ambient air pollutant concentration in Tianjin was analyzed. Fluctuations of 24h average SO2 concentration were mainly attributed to seasonal changes in source emissions, and weather-related short-term variations could not be neglected. Whereas variations of 24h average PM10 concentration were mainly due to short-term changes in source emissions and weather conditions. Meteorological conditions had adverse effects on atmospheric diffusion in 2002~2003, which led to an increase of 3μg/m3 in long-term SO2 concentration. While beneficial meteorological conditions in the year of 2004 accounted for a reduction of 5μg/m3. Meteorological conditions had similar effects on long-term PM10 concentrations, as the concentration of PM10 had a more significant increase of about 10μg/m3 in 2002~2003.