Drivers of short-term recovery in vegetation greenness and canopy height in burned areas of Southwest China

Forest fires are major disturbances to forest ecosystem structure and function. Understanding post-fire vegetation recovery and its drivers is crucial for forest restoration. This study investigates 20 forest sites burned in Southwest China in 2020. Post-fire vegetation recovery was evaluated in two dimensions using multi-source remote sensing: vegetation greenness represented by Enhanced Vegetation Index (EVI) and canopy height. Recovery was quantified with Relative Recovery Index (RRI) and annual growth rate. A Generalized Additive Model (GAM) was employed to explore the driving effects of multiple factors,including topography, climate, fire severity, and pre-fire vegetation conditions, on post-fire vegetation recovery. It was found that there was an asynchrony between the recovery of EVI and canopy height after fire, EVI recovered faster than canopy height after fire (RRI≤0, 38.27 % vs 6.37 %; RRI 0.5–1.0, 53.8 % vs 17.4 %), thus potential overestimation for forest recovery if assessed using optical indices solely. GAM results indicated that EVI recovery was primarily driven by precipitation, temperature, fire severity, and pre-fire EVI; canopy height recovery was mainly influenced by slope, fire severity, and pre-fire canopy height, with elevation and precipitation influenced recovery through interactions with other factors. High fire severity enhanced EVI recovery but suppressed canopy height recovery, while pre-fire vegetation conditions negatively affected the short-term recovery of both metrics. For both single and interactive driving factors, the effects on the recovery of EVI and canopy height were predominantly nonlinear rather than purely linear. The results advance knowledge of post-fire vegetation recovery mechanisms and support informed evaluation and management of affected ecosystems. • Recovery of vegetation greenness and canopy height after the fire is asynchronous. • Vegetation greenness increases with fire intensity, but canopy height shows decline. • Most drivers show obvious nonlinear relationships with vegetation restoration. • Topographic–meteorological interactions significantly affect vegetation recovery.