Articles | Volume 16, issue 16
Development and technical paper
29 Aug 2023
Development and technical paper |  | 29 Aug 2023

A gridded air quality forecast through fusing site-available machine learning predictions from RFSML v1.0 and chemical transport model results from GEOS-Chem v13.1.0 using the ensemble Kalman filter

Li Fang, Jianbing Jin, Arjo Segers, Hong Liao, Ke Li, Bufan Xu, Wei Han, Mijie Pang, and Hai Xiang Lin

Data sets

The PM2.5 data from observations and model outputs for fused prediction Li Fang

Model code and software

geoschem/GCClassic: GEOS-Chem 13.1.0 (13.1.0) The International GEOS-Chem User Community

Python source code of EnKF-based prediction fusion Li Fang

Short summary
Machine learning models have gained great popularity in air quality prediction. However, they are only available at air quality monitoring stations. In contrast, chemical transport models (CTM) provide predictions that are continuous in the 3D field. Owing to complex error sources, they are typically biased. In this study, we proposed a gridded prediction with high accuracy by fusing predictions from our regional feature selection machine learning prediction (RFSML v1.0) and a CTM prediction.