Development of a recurrent spatiotemporal deep-learning method coupled with data fusion for correction of hourly ozone forecasts.

Ambient ozone (O) predictions can be very challenging mainly due to the highly nonlinear photochemistry among its precursors, and meteorological conditions and regional transport can further complicate the O formation processes. The emission-based chemical transport models (CTM) are broadly used to predict O formation, but they may deviate from observations due to input uncertainties such as emissions and meteorological data, in addition to the treatment of O nonlinear chemistry. In this study, an innovative recurrent spatiotemporal deep-learning (RSDL) method with model-monitor coupled convolutional recurrent neural networks (ConvRNN) has been developed to improve O predictions of CTM. The RSDL method was first used to build the ConvRNN within a 24-h scale to characterize the spatiotemporal relationships between the monitored O data and CTM simulations, and then incorporated the recurrent pattern to achieve 72-h multi-site forecasts based on a pilot study over the Pearl River Delta (PRD) region of China. The results showed that the RSDL method predicted O with high accuracy over this case study, with an increase of 27.54% in the correlation coefficient (R) average for all sites as well as an increase in R of 0.14-0.21 for all cities compared to CTM. Moreover, the regional distribution of CTM was further improved by the RSDL predictions with the data fusion technique, which greatly reduced the underpredictions of O concentrations, particularly in high O-level areas (concentrations >160 μg/m), with a 33.55% reduction in the mean absolute error (MAE).