Simulating the air quality impact of prescribed fires using graph neural network-based PM_2.5 forecasts

Environmental Data Science, Год журнала: 2025, Номер 4

Опубликована: Янв. 1, 2025

Abstract The increasing size and severity of wildfires across the western United States have generated dangerous levels PM 2.5 concentrations in recent years. In a changing climate, expanding use prescribed fires is widely considered to be most robust fire mitigation strategy. However, reliably forecasting potential air quality impact from fires, which critical planning fires’ location time, at hourly daily time scales remains challenging problem. this paper, we introduce spatio-temporal graph neural network (GNN)-based model for predictions California. Utilizing two-step approach, our predict net ambient concentrations, are used estimate wildfire contributions. Integrating GNN-based with simulations historically propose novel framework forecast their impact. This determines that March optimal month implementing California quantifies trade-offs involved conducting more outside peak season.

Язык: Английский

Prescribed burn related increases of population exposure to PM2.5 and O3 pollution in the southeastern US over 2013–2020

Environment International, Год журнала: 2024, Номер 193, С. 109101 - 109101

Опубликована: Окт. 28, 2024

Язык: Английский