Work

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

Year: 2025

Type: article

Abstract: High Resolution Image Download MS PowerPoint Slide Several s tudies 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 km 2 ) 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 R 2 ). The range of the R 2 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 R 2 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.

Source: Environment & Health

DOI: 10.1021/envhealth.4c00201