基于WRF和随机森林的臭氧短期预报模型研究——以青岛为例
Research on short-term ozone forecasting model based on WRF and random forest——A case study of Qingdao
卢维肖 1周易 2杨元建 1孙爱青 3方渊 4孟赫4
作者信息
- 1. 南京信息工程大学 大气物理学院,江苏 南京 210044
- 2. 南京信息工程大学 管理工程学院,江苏 南京 210044
- 3. 华润新能源(睢宁)有限公司,江苏 徐州 221200
- 4. 山东省青岛生态环境监测中心,山东 青岛 266003
- 折叠
摘要
本研究以青岛市为例,基于WRF模式预报与随机森林方法,利用 2022 年 9 月 9 个站点的预报与实况数据进行了气象要素预报的订正,回代气象要素订正数据并加入静态数据建立了O3 浓度的短期预报模型,实现了每个站点预报时效为 7d、时间分辨率为 1h的预报,对气象要素订正与O3 预报效果进行评估与分析.结果表明,气象要素 2m气温、地面气压、2m相对湿度、10 m风速均有不同程度的订正,尤其是地面气压与10m风速订正效果明显;O3 浓度预报模型有良好的预报效果,RMSE较WRF预报减少了 22.1%,站点预报误差均有改善且大部分效果明显.
Abstract
This study takes Qingdao as an example,based on the WRF model forecast and the random forest method,the forecast and the real data of 9 stations in September 2022 are used to revise the meteorological element forecast,and the short-term forecast model of O3 concentration is established by substituting the revised data of meteorological elements and adding the static data to realize the forecast with the time resolution of 1 hour for 7 days at each station,and the effect of the revision of the meteorological elements and the O3 prediction is evaluated and analyzed.The results show that the meteorological elements 2 m air temperature,surface pressure,2 m relative humidity,and 10 m wind speed are all revised to different degrees,especially the surface pressure and 10 m wind speed are revised with obvious effects;the O3 concentration prediction model has a good prediction effect,the RMSE is reduced by 22.1%compared with the WRF prediction,and the station prediction errors are improved and most of the effects are obvious.
关键词
臭氧/随机森林/预报订正/青岛市Key words
Ozone(O3)/random forest/revision of forecasts/Qingdao引用本文复制引用
基金项目
国家自然科学基金优秀青年科学基金(42222503)
出版年
2024
NETL
NSTL
万方数据