Grassland ecological security assessment in Northwest Sichuan using the driver-pressure-state-impact-response...

Ecological security assessment is of fundamental importance for the conservation and sustainable management of fragile grassland ecosystems. Traditional assessment methods are frequently limited by subjective indicator weighting and inadequate consideration of spatial autocorrelation, which constrains their reliability in evaluating complex grassland systems. This study examined the spatiotemporal patterns of grassland ecological security in Northwest Sichuan, China, from 2000 to 2020, by integrating the Driver-Pressure-State-Impact-Response (DPSIR) framework, a Genetic Algorithm-optimized Projection Pursuit Model (GA-PPM) and spatial autocorrelation analysis. The GA-PPM was applied to achieve objective weighting of key indicators and identify spatial heterogeneity in ecological security. Results revealed that ecologically vulnerable zones were mainly distributed in the northeastern and southwestern parts of the study area. Slope and population density were identified as the dominant driving factors, illustrating the combined constraints of topographic characteristics and anthropogenic disturbances (e.g., grazing and tourism). Based on these ecological insights, study propose targeted strategies including differentiated zoning management and ecological restoration to mitigate grassland degradation. This integrated approach provides both a methodological basis and practical reference for ecological security assessment in other fragile grassland regions globally.