Spatiotemporal patterns and influencing factors of conflict risk in Pakistan: A multi-scalar analysis based on GDELT data

Against the backdrop of an increasingly complex geopolitical environment, identifying the spatiotemporal patterns of conflict risk and its influencing factors is of importance for regional security governance. Taking Pakistan as the study area, this study measures conflict risk from 2015 to 2024 based on GDELT data, examines its spatiotemporal patterns, and differences across conflict types, and employs spatial econometric models to investigate the multidimensional influencing factors and their spatial heterogeneity at national and provincial scales. Results show: (1) Conflict risk exhibits significant spatial inequality, with high-risk areas primarily concentrated in border regions, major urban centers, and political cores, while showing notable temporal fluctuations; (2) conflict risk is shaped by the joint effects of multiple dimensions, with border location playing a fundamental role and exhibiting a nonlinear “border–transition zone–inland” gradient. population agglomeration and transportation conditions significantly intensify conflict risk, while the effects of social structural factors are more context-dependent; and (3) the mechanisms influencing conflict risk exhibit significant spatial heterogeneity. Exposure to conflict risk in neighboring countries is significantly and positively associated with local conflict risk in Khyber Pakhtunkhwa and Balochistan. Meanwhile, the effects of transportation and social structural factors vary along the border-inland gradient, indicating that spatial heterogeneity in the mechanisms influencing conflict risk exists not only across provinces but also within provinces. This study reveals the spatial differentiation mechanisms of conflict risk from a multi-scalar perspective and provides implications for conflict risk identification and regional security governance in Pakistan.