Integrating leaf-level combustion dynamics into wildfire risk assessment and carbon management: A decision-support framework for subtropical forests.

Forest litter strongly affects wildfire ignition, heat release, smoke generation, and post-fire carbon cycling, but species-specific leaf-level combustion data remain limited for subtropical green-space management. This study investigated the combustion behavior, carbon emission characteristics, and biochar properties of Liriodendron, Rhododendron, and Camphor litter collected from a managed subtropical green-space environment in East China. Leaf specimens were tested under external heat fluxes of 35, 55, and 75 kW/m2 using CONE tests, combined with proximate and ultimate analyses, carbon emission measurements, FESEM-EDS characterization, and Monte Carlo uncertainty analysis. The results showed clear interspecific differences. Rhododendron exhibited the highest PHRR, indicating strong short-term fire-growth potential, whereas Camphor showed the highest THR, ICF, and CO/CO2 ratio, suggesting greater sustained-burning, CO2-emission, and smoke-exposure risks. In contrast, Liriodendron generally showed lower heat-release and emission potential. Biochar analysis indicated that approximately 14.03% of the initial dry leaf mass was stored as CO2-equivalent carbon over a 100-year timescale, with a Monte Carlo 95% uncertainty interval of 11.2%-17.5%. These findings provide laboratory scale evidence for species-specific litter-risk classification and support differentiated litter management, vegetation planning, and carbon-accounting strategies in subtropical urban forests and wildland-urban interface environments.