Ground-level ozone estimation based on geo-intelligent machine learning by fusing in-situ observations, remote sensing data, and model simulation data

Abstract

In recent years, near-surface ozone (O3) pollution has been increasing, seriously endangering both the ecological environment and human health. Accurately monitoring spatially continuous surface O3 is still difficult with only remote sensing observations. In this paper, to address this issue, we propose a method for estimating surface O3 by fusing multi-source data, including in-situ observations, O3 precursors obtained by remote sensing, and model simulation data, including O3 profile data and reanalysis products of meteorological and radiative elements. The estimation method is geo-intelligent light gradient boosting (Geoi-LGB) which takes into account both the spatial and temporal geographical correlation based on the standard LGB model. The spatio-temporal autocorrelation factors of the site observations are also constructed and added into the input variables. In a case study of China, centered on North China in 2019, the Geoi-LGB method obtained a root-mean-square error of 10.25 μg/m3, a mean absolute error of 7.30 μg/m3, and a coefficient of determination of 0.912 under the site-based cross-validation strategy. The proposed method has the advantages of being able to obtain a higher accuracy than some of the popular O3 estimation models. Furthermore, the excellent spatial mapping ability of the Geoi-LGB method was demonstrated, in that about 85 % of the sites had an annual average absolute error of less than 10 μg/m3. We believe that this study could provide some important reference information for the accurate estimation of ground-level O3.

Keywords

• Near-surface ozone estimation
• Light gradient boosting machine
• Spatio-temporal correlation
• Ozone profile of model simulation
• S5P-TROPOMI