Synergistic Effects of Drought and Fire Intensity Drive Carbon Source Shifts in Chinese Forests

Forest fires profoundly affect ecosystem carbon sink functions through direct carbon emissions and ecological process disturbances. However, the mechanisms by which fire carbon emissions (FCE) influence net ecosystem productivity (NEP) remain poorly understood. Here, we analyzed the nonlinear impacts of FCE on NEP using long-term remote sensing data (2000~2018) across China’s forest ecosystems, employing linear mixed-effects models and logistic generalized linear mixed-effects models. Our results revealed significant spatiotemporal heterogeneity in the FCE-NEP relationship, with a conditional coefficient of determination (R2 = 0.785) substantially higher than the marginal R2 (0.332), indicating that the effect of fire on carbon sinks depends on where and when it occurs, with contrasting responses between evergreen and deciduous forests and between early and late growing seasons. Different vegetation functional types exhibited contrasting responses to fire. In temperate evergreen broadleaf forests, FCE showed a significant positive correlation with NEP, consistent with the mechanism that fire promotes carbon uptake through nutrient release. In contrast, a significant negative correlation was observed in temperate deciduous broadleaf forests, reflecting their vulnerability associated with shallow root systems and high leaf turnover rates. Drought and fire intensity exerted a strong synergistic amplification effect. The three-way interaction among FCE, vegetation type, and water deficit greatly improved model fit relative to models without this interaction (ΔAIC > 2500). Among multiple fire characteristic metrics, fire radiative brightness (BRIGHTN), representing fire intensity, was the most critical factor influencing carbon sinks and driving carbon decline. This study elucidates the nonlinear, heterogeneous, and synergistic characteristics of fire impacts on forest carbon sinks. The results highlight that fire effects on carbon sinks are not uniform but depend critically on vegetation type and the compounded effects of drought and fire intensity. These findings provide a scientific basis for vegetation-specific assessments and dynamic monitoring of forest carbon sink vulnerability under changing fire and drought regimes.