Long-term effects of post-earthquake landslides on vegetation ecosystem net carbon

Landslides significantly impact carbon dynamics across global regions, yet the long-term vegetation ecosystem net carbon effect from post-earthquake landslides remains unclear. To address this gap, this study focuses on the 2008 Wenchuan earthquake in Sichuan, China. Using landslide inventories and satellite data, we applied the Seasonal Autoregressive Integrated Moving Average Model (SARIMA) model and Bayesian change-point detection to examine long-term effects of post-earthquake landslides on vegetation ecosystem net carbon uptake (ENCU). For the restored areas, the multi-year total vegetation ecosystem net carbon loss (ENCL) was calculated for different landslide-prone areas. Subsequently, the Random Forest Regression Model and Structural Equation Model were used to explore the environmental drivers of ENCL. A substantial decline in ENCU was observed in the post-earthquake, with a 17.40 % reduction noted in May 2008. By 2019, only 45 out of 86 affected areas recovered to pre-earthquake ENCU levels, with 18 areas surpassing their initial levels. Recovery times for most landslides ranged from 8 to 11 years. From 2008 to 2019, the total ENCL across various landslide-affected areas primarily ranged from 5000 to 20,000 gC m⁻². Precipitation emerged as a key driver of ENCL, while landscape fragmentation hindered recovery. These findings underscore the need for sustainable post-earthquake land management to maintain carbon balance, offering new insights into how seismic events influence global carbon cycles.