河南省供暖期大气污染特征及预测分析——以郑州市为例  

The Characteristics and Prediction of Air Pollution during the Heating Period in Henan Province:A Case Study of Zhengzhou City

作　　者：韩艳[1] 李东璞 李思杰 刘俊杰 张宁丹 HAN Yan;LI Dongpu;LI Sijie;LIU Junjie;ZHANG Ningdan(College of Environment and Planning,Henan University,Henan Kaifeng 475004,China)

机构地区：[1]河南大学环境与规划学院,河南开封475004

出　　处：《河南大学学报（自然科学版）》2018年第6期639-649,共11页Journal of Henan University:Natural Science

基　　金：中央引导地方科技发展项目(HN2016-149);河南省大气污染综合防治与生态安全重点实验室开放基金项目(PAP 201703);河南省高校科技创新团队支持计划资助(16IRTSTHN012)

摘　　要：近年来河南省大气污染问题引起社会的广泛关注,但有关供暖期间大气污染方面的研究相对较少.以郑州市为例,分析郑州市供暖期间大气颗粒物的浓度变化并进行预测,对提高当地空气质量具有重要意义.基于2014-2016年郑州市空气质量监测数据和同期气象数据,利用SPSS相关分析和BP神经网络模型,分析郑州市供暖期间PM2.5、PM10的超标情况、日变化特征,探究气象要素对PM2.5和PM10的影响,最后预测AQI指数的变化.结果表明:2014年供暖期郑州市空气质量相对较差,PM2.5和PM10平均质量浓度超标率最高;2015年供暖期郑州市空气质量相对较好,PM2.5和PM10平均浓度变化幅度较大;2014-2016年供暖期间郑州市PM2.5和PM10浓度具有明显的日变化特征,呈现双峰型变化;2014-2016年供暖期郑州市PM2.5、PM10与日均气温相关性不显著,与日均风速呈显著负相关,与日均相对湿度呈显著正相关;当供暖期郑州市主导风向为正西风时,污染天气出现频率较低;利用BP神经网络预测2016年AQI的精度较高,预测值与实测值相关系数为0.85.In recent years, atmospheric pollution of Henan Province has caused widespread concern in society. However, there were relatively few studies about atmospheric pollution during the heating period. Taking Zhengzhou City as the research object, in order to analyze the characteristics of atmospheric pollution and forecasting pollution, it is of significance to improve air quality of Zhengzhou. Based on the air quality monitoring data and the meteorological data from 2014 to 2016, correlation analysis and BP neural network model were used to analyze the over-limit rate and diurnal variation characteristics of PM2.5 and PM10 during the heating period of Zhengzhou. The paper investigated the impact among meteorological elements, PM2.5 and PM10, and finally forecasted AQI. The results show： In 2014, the air quality of Zhengzhou was relatively poor, in which the average level of exceeded PM2.5 and PM10 was the highest; the air quality during heating period was relatively good in 2015, however, the average level of PM2.5 and PM10 had changed significantly; during the heating period from 2014 to 2016, the level of PM2.5 and PM10 in Zhengzhou had obvious diurnal variations, which was double-peak; there was no significant correlation among PM2.5, PM10 and the average daily air temperature during the heating period of Zhengzhou from 2014 to 2016; PM2.5 and PM10 were significantly negatively correlated with the average daily wind speed and positively correlated with the daily average relative humidity; when the dominant wind direction of Zhengzhou during the heating Period was West Wind, the frequency of polluted weather was low; the accuracy of predicting AQI in 2016 by the BP neural network was high, and the correlation coefficient between the predicted value and the measured value was 0.85.

关 键 词：供暖期 超标率 气象因素 预测分析 

分 类 号：X513[环境科学与工程—环境工程]