基于XGBoost算法的近地面臭氧浓度遥感估算  

Remote-sensing estimation of near-surface ozone concentration based on XGBoost

作　　者：赵楠 卢毅敏[1,2,3] ZHAO Nan;LU Yimin(Key Lab of Spatial Data Mining and Information Sharing of Ministry of Education,Fuzhou University,Fuzhou 350108;Digital Region Engineering Technology Research Center in Fujian Province,Fuzhou University,Fuzhou 350108;Academy of Digital China(Fujian),Fuzhou 350003)

机构地区：[1]福州大学,空间数据挖掘与信息共享教育部重点实验室,福州350108 [2]福州大学,福建省数字区域工程技术研究中心,福州350108 [3]数字中国研究院(福建),福州350003

出　　处：《环境科学学报》2022年第5期95-108,共14页Acta Scientiae Circumstantiae

基　　金：国家重点研发科技专项(No.2017YFB0503500);福建省科技计划项目(No.2020L3005)。

摘　　要：本文采用XGBoost机器学习算法,融合臭氧浓度地面监测数据、欧洲中期天气预报中心的ERA5数据集、中国多尺度排放清单模型构建的排放清单数据集、高分辨率遥感影像(TROPOMI_NO_(2)、OMI_NO_(2))以及人口数据和DEM数据,构建训练估算数据集,开展近地面臭氧浓度估算研究.模型构建采用递归式特征消除法进行特征变量的选择,并对其进行十折交叉和自建模验证,R^(2)分别为0.871和0.955,RMSE分别为12.8μg·m^(-3)和7.514μg·m^(-3).同时进行了高分辨率遥感影像对估算结果的贡献分析,结果表明引入TROPOMI_NO_(2)因子参与建模可校正近地面臭氧浓度普遍被低估现象.模型模拟结果显示臭氧浓度回归估算结果层次更加分明、条带现象消失、连续性和平滑性明显改善.In this paper,the Extreme Gradient Boosting(XGBoost)machine learning algorithm was used to estimate near-surface ozone concentration.By integrating the ground monitoring data of ozone concentration,ERA5 dataset of European Centre for Medium-Range Weather Forecasts(ECMWF),emission inventory dataset of Multi-resolution Emission Inventory for China(MEIC),high-resolution remote sensing images (TROPOMI_NO_(2),OMI_NO_(2)),population data and DEM data,a training dataset was built to construct a remote sensing estimation model.A recursive feature elimination method was used to select features of the model,and the 10-fold cross-validation and self-modeling verification were performed.The R^(2)were respectively 0.871 and 0.955;the RMSE were 12.8μg·m^(-3)and 7.514μg·m^(-3)respectively.At the same time,the contribution of high-resolution remote sensing images to the estimation was analyzed.The results showed that the introduction of TROPOMI_NO_(2)factor to the modeling can improve the general underestimation of near-surface ozone concentration.The simulation results show that the ozone concentration regression estimation results are more distinct,the banding phenomenon disappears,and the continuity and smoothness are significantly improved.

关 键 词：近地面臭氧 XGBoost TROPOMI OMI 时空分布 

分 类 号：X51[环境科学与工程—环境工程]