Combining physical mechanisms and deep learning models for hourly surface ozone retrieval in China

As surface ozone (O₃) gains increasing attention, there is an urgent need for high temporal resolution and accurate O₃ monitoring. By taking advantage of the progress in artificial intelligence, deep learning models have been applied to satellite based O₃ retrieval. However, the underlying physical mechanisms that influence surface O₃ into model construction have rarely been considered. To overcome this issue, we considered the physical mechanisms influencing surface O₃ and used them to select relevant variable features for developing a novel deep learning model. We used a wide and deep model architecture to account for linear and non-linear relationships between the variables and surface O₃. Using the developed model, we performed hourly inversions of surface O₃ retrieval over China from 2017 to 2019 (9:00-17:00, local time). The validation results based on sample-based (site-based) methods yielded an R² of 0.94 (0.86) and an RMSE of 12.79 (19.13) μg/m³, indicating the accuracy of the models. The average surface O₃ concentrations in China in 2017, 2018, and 2019 were 82, 78, and 87 μg/m³, respectively. There was a diurnal pattern in surface O₃ in China, with levels rising significantly from 55 μg/m³ at 9:00 a.m. to 96 μg/m³ at 15:00. Between 15:00 and 16:00, the O₃ concentration remained stable at 95 μg/m³ and decreased slightly thereafter (16:00-17:00). The results of this study contribute to a deeper understanding of the physical mechanisms of ozone and facilitate further studies on ozone monitoring, thereby enhancing our understanding of the spatiotemporal characteristics of ozone.