气象条件和污染物排放对兰州市冬季空气质量的影响

发布时间：2018-07-23 09:37

【摘要】：利用人工神经网络(ANN)技术,基于气象条件、污染物排放变化和污染物浓度资料构建污染物浓度统计模型,在此基础上分析气象条件和污染物排放源排放变化对污染物浓度逐日变化和年际变化的影响。研究结果发现基于ANN建立的大气污染统计预报模型模拟NO_2浓度准确性较高,其次为SO2,PM10浓度准确性较低。ANN的输入参数更适合NO_2的模拟,SO2和PM10浓度的影响因子较为复杂。气象条件变化是NO_2浓度逐日变化的主要影响因子,污染物排放量变化是NO_2浓度年际变化的主要影响因子。因子分离法计算得到的气象条件、污染物排放及两者相互作用对NO_2浓度逐日变化的贡献率分别是57.9%、24.5%和17.6%,对NO_2浓度年际变化的贡献率分别是13.7%、73.3%和13%。
[Abstract]:Using artificial neural network (Ann) technology, based on meteorological conditions, pollutant emission change and pollutant concentration data, a statistical model of pollutant concentration is constructed. On this basis, the effects of meteorological conditions and pollutant emission sources on the diurnal and interannual variations of pollutant concentration were analyzed. The results show that the statistical prediction model of air pollution based on Ann has higher accuracy in simulating NO2 concentration, and the input parameters of Ann are more suitable for simulating SO2 and PM10 concentration of NO2 than those of so _ 2O _ 2PM10 concentration. The change of meteorological conditions is the main influence factor of the daily change of NO2 concentration, and the change of pollutant emission is the main factor of the interannual variation of NO2 concentration. The contribution rates of meteorological conditions, pollutant emissions and their interactions to the daily variation of NO2 concentration by factor separation method are 57.5% and 17.6%, respectively, and the contribution rates to the interannual variation of NO2 concentration are 13.73.3and 13.3%, respectively.
【作者单位】：中国科学院寒区旱区环境与工程研究所中国科学院寒旱区陆面过程与气候变化重点实验室;中国气象科学研究院;青海省气象局;济南市气象局;
【基金】：中国科学院百人计划项目(290827631) 兰州市科技攻关计划项目(2009KJLQ) 中国科学院寒旱区陆面过程与气候变化重点实验室开放基金项目(LPCC201405)
【分类号】：X51;X16

【相似文献】