Nationwide Machine Learning-Ensemble PM2.5 Mapping Prediction and Forecasting Models in South Korea with High Spatiotemporal Resolution and Health Risk Estimation-Based Evaluations

Several studies developed machine learning-based PM_2.5 prediction models; however, nationwide models addressing both mapping prediction and forecasting were limited. Further, although the prediction accuracy is different from PM_2.5-related health risk estimation, previous studies solely examined the prediction accuracy. This study suggests a method to assess the statistical properties of PM_2.5-health risk estimation, which also can be used as a model selection. We used three machine learning algorithms and an ensemble method to construct PM_2.5 mapping prediction (1 km²) and two-day forecasting models majorly using satellite-driven data in South Korea (2015–2022). We performed a simulation study to examine the statistical properties of short-term PM_2.5 risk estimation using prediction models. Our ensemble spatial prediction model showed better performance than single algorithms (0.956 test R²). The range of the R² values was 0.78–0.98 across the monitoring sites. The average % bias was from 1.403%–1.787% when our mapping models for PM_2.5-mortality risk estimation, compared to the estimates from monitored PM_2.5. The best R² of our forecasting models was 0.904. This study developed machine learning models for spatial PM_2.5 predictions and forecasting in Korea. This study also suggested a method to address risk estimation and model selection concurrently when multiple prediction models were used.