山东省PM2.5污染现状及预报效果对比分析
Overview of PM2.5 Pollution Status and Comparative Analysis on the Forecasting Effect of Air Quality in Shandong Province,China
王桂霞 1王妍然 1孟赫 2丁椿 3邱晓国 1张淼 1许杨 1解军1
作者信息
- 1. 山东省生态环境监测中心,山东 济南 250101
- 2. 山东省青岛生态环境监测中心,山东 青岛 266000
- 3. 中科三清科技有限公司,北京 100029
- 折叠
摘要
对 2019-2022 年山东省 16 个市的细颗粒物(PM2.5)污染特征进行了分析,并对 2021 和 2022 年的 4 个数值模式[社区多尺度空气质量模拟系统(CMAQ)、扩展综合空气质量模型(CAMx)、区域气象-大气化学在线耦合模式(WRF-Chem)、嵌套网格空气质量预报系统(NAQPMS)]及集合预报模式预测的效果进行评估.结果表明:2019-2022 年山东省PM2.5 年均质量浓度逐年降低,污染程度逐步减轻,但在 1-3,11-12 月,PM2.5 质量浓度超标现象较为普遍.2021 年底更换污染源清单后,2022 年 5 个模式的 24h级别准确率和相关系数(r)同比升高,均方根误差(RMSE)同比降低,模式预报准确率有所提升,但由于参数调整略大,CMAQ、CAMx、WRF-Chem、集合预报模式易漏报或偏轻预报PM2.5 的中度污染和重度污染天气.由于NAQPMS模式在更换污染源排放清单时,同时改进了非均相化学反应机制,因此对PM2.5 不同污染类别尤其是中度污染、重度污染的预报准确率明显提升.
Abstract
This paper analyzed the characteristics of PM2.5 pollution in 16 cities in Shandong Province from 2019 to 2022,and evaluated the forecasting effects in 2021 and 2022 of four numerical models(CMAQ,CAMx,WRF-Chem,NAQPMS)and ensemble forecasting model.The results showed as follows:The average annual concentration of PM2.5 in Shandong Province from 2019 to 2022 has decreased year by year,and the pollution degree has gradually decreased.However,the phenomenon of PM2.5 concentrations exceeding the national standard often occurred in the months of January-March and November-December.After updating of the pollutant emission inventory of models at the end of 2021,the 24-hour level accuracy and correlation coefficient(r)of the five models increased year on year in 2022,and the root mean square error(RMSE)decreased year on year.Although the model forecast accuracy was improved,CMAQ,CAMx,WRF-Chem,and ensemble forecasting model were easy to miss or underforecast the moderately and heavily polluted weather of PM2.5,due to that the parameter adjustment was slightly larger.Since the heterogeneous chemical reaction mechanism was improved when the pollution emission inventory of NAQPMS was updated,the forecast accuracy of different levels of PM2.5 days was significantly improved,especially on the moderately and heavily polluted days.
关键词
山东省/细颗粒物/预报模式/污染源清单/预报效果Key words
Shandong Province/PM2.5/Forecasting model/Pollution emission inventory/Forecasting effect引用本文复制引用
基金项目
山东省自然科学基金重大基础研究项目(ZR2020ZD21)
山东省自然科学基金面上项目(ZR2021MD013)
出版年
2024
NETL
NSTL
万方数据