河北省城市PM_(2.5)浓度模型构建研究

Study on PM_(2.5) Concentration Model of the Cities in Hebei Province

多年PM_(2.5)浓度序列可用于区域大气污染的时空演变特征分析,现有PM_(2.5)浓度监测数据积累时间短,有必要构建PM_(2.5)浓度模型。以河北省为例,选择空气污染程度不同的6个城市(邢台、保定、承德、张家口、唐山、秦皇岛),开展河北省城市PM_(2.5)浓度模型构建研究。利用2013-2016年的PM_(2.5)浓度与PM_(10)、大气污染物(SO_2、NO_2、CO和O_3)观测数据,分析PM_(2.5)与PM_(10)、大气污染物(SO_2、NO_2、CO和O_3)相关性,PM_(2.5)浓度与PM_(10)、NO_2、SO_2和CO呈显著正相关,与O_3呈负相关。利用逐步多元回归方法分析构建基于PM_(10)与气态污染物的PM_(2.5)浓度模型,模型预测PM_(2.5)浓度与实测PM_(2.5)浓度序列存在较高的相关性,平均偏差优于3.5μg/m^3。河北省城市PM_(2.5)浓度模型预测值与PM_(2.5)浓度观测值基本吻合。

PM2. 5 concentrations sequence for many years can be used to temporal and spatial evolution charac- teristic of regional air pollution. Due to the short data of the existing PM2. 5 concentrations, it is necessary to con- struct a PM2.5 concentration model. Taking the example of Hebei province, it was selected six cities with different air pollution levels （ Xingtai, Baoding, Chengde, Zhangjiakou, Tangshan and Qin hangdao） to construct the mod- els of PMz.5 concentration. Using the observation data of PM2.5, PM10 and gaseous pollutants from 2013 to 2016, it was analyzed the correlation between PM2.5, PM10 and gaseous pollutants. There was notable positive correlation ex- isting in PMI0, NO2, SO2 and CO, and negative correlation in 03. It was constructed the models of PM2.5 concen- trations in different cities based on PM10 and gaseous pollutants by using the stepwise regression analysis method. There was a high correlation of PM2. 5 concentration between the predicted value and the measured value which the average deviation was better than 3. 5μg/mIn3. Indeed, the PM2.5 concentration model is in agreement with its observed values.