Spatio-temporal dynamics of forest restoration in the karst region: integrating trend analysis and the multi-order adjacency index with Landsat imagery

Monitoring forest restoration in fragile landscapes like karst demands frameworks that disentangle coupled spatial-temporal dynamics and address spectral and spatial pattern gaps. TFPW-MK trend testing and the Multi-order Adjacency Index (MAI) were used to analyze the temporal trends and spatial expansion patterns of forest restoration, based on Landsat time-series data (1986–2024). Our findings demonstrated that after controlling for precipitation effects, 68.74% of the study area exhibited significant NDMI increases, which was negatively correlated with precipitation trends, confirming that forest restoration is the dominant driver of NDMI increases. Forest restoration was primarily characterized by continuous increase, with accelerated restoration rates after 2000 and 2006. MAI analysis revealed that edge and proximity expansion types dominated the restoration (>60% coverage), forming spatially coherent buffers around core forests. However, a distinct phase of spread-type accounted for 20% of new growth during 2010–2015, closely aligned with policy interventions. The northern Mulun Nature Reserve exhibited sustained contiguous restoration, while the central and southern regions displayed more complex recovery patterns influenced by topography and human activities. The integrated framework provides a novel perspective for understanding the multi-scale mechanisms in karst regions and offers a scientific basis for formulating differentiated ecological restoration strategies.