Investigating the characteristics of PM2.5 emissions from coniferous trees during indoor simulated combustion utilizing a random forest model

Wildfire smoke is a significant pollutant and severe threat to climate, air quality, and public health within forest ecosystems. This study focused on quantifying PM 2.5 emission characteristics from the combustion of major coniferous species in China. Indoor ignition experiments were conducted on needles, branches, and bark of six coniferous species—— Pinus koraiensis (HS), Larix gmelinii (LYS), Pinus sylvestris var. Mongolica (ZZS), Abies fabri (LS), Picea jezoensis (YLYS), and Picea koraiensis (HPYS)—from the Liangshui National Nature Reserve. The experiments systematically varied fuel moisture content, fuel load, and wind speed to assess their effects on PM 2.5 emissions. Substantial disparities in PM 2.5 emission concentrations were observed among different tree species and their organs. Both individual and interactive effects of fuel moisture content, fuel load, and wind speed significantly impacted PM 2.5 emissions. Elevated wind speed and fuel load were identified as predominant factors influencing PM 2.5 concentrations, whereas the impact of high fuel moisture content was more complex. The random forest model trained on these data effectively predicted PM 2.5 emissions at the laboratory scale. This study provides a crucial reference for estimating wildfire smoke emissions, evaluating their atmospheric impact, and informing refined forest fuel management strategies.