Increasing severity of large-scale fires prolongs recovery time of forests globally since 2001.

Ongoing and sharply increased global forest fires, especially extreme large-scale fires (LFs) with their greater destructiveness, have significantly altered forest structures and functions. However, long-term variations in the severity of LFs and corresponding effects on the natural post-LF recovery time of global forests remain unclear. Here, we rigorously identified 3,281 global large-scale (>10 km) single-time fire events (LSFs) from 2001 to 2021, and used multiple indicators to understand the post-LSF recovery dynamics from different perspectives and comprehensively reveal major driving factors across regions and forests types based on multiple models. Compared with pre-2010, LSFs after 2010 caused greater forest damage, with the fire severity expanding further from low to high latitudes and from humid to arid regions, particularly affecting evergreen needleleaf forests. Fewer than one-third of the forests recovered successfully within 7 years, and most of these were tropical, moisture-rich broadleaf forests. The average time required for three indicators to recover to pre-fire conditions increased by 7.5% (vegetation density), 11.1% (canopy structure) and 27.3% (gross primary productivity). Moreover, the positive sensitivity of recovery time to increased fire severity was significantly intensified. Notably, more forests experienced recovery stagnation with increased severity, especially in boreal forests, further extending recovery time. The negative impact of the severity of LSFs on forest recovery was much stronger than that of post-LSF climate conditions. Soil moisture after LSFs was identified as the primary facilitating factor. Temperature generally had a positive role before 2010, but a strong negative influence on post-LSF forest recovery after 2010. These findings provide a useful reference for better understanding global forest recovery mechanisms, estimating forest carbon sinks and implementing post-LSF management accordingly.