Spatiotemporal Characteristics of Land Use Carbon Budget and Carbon Balance Capacity in Karst Mountainous Areas: A Case Study Using Social Network Analysis

Collaborative carbon regulation in Karst mountains critically reconciles socio-ecological conflicts. While intercity linkages drive spatial carbon heterogeneity, prior studies have focused on administrative-scale accounting, neglecting systematic spatial association network (SAN) analysis. Integrating SAN and geospatial detector models, we reveal county-level carbon balance dynamics in Guizhou, China (2000–2020). The key findings show the following: provincial carbon emissions rose 53% (0.96 to 1.47 × 108 t) against a 15% sequestration decline (0.67 to 0.57 × 108 t); emission networks shifted from single-core clustering to the axial Liupanshui–Guiyang–Tongren corridor, while sequestration networks retained peripheral ecological dominance; carbon balance capacity (CBC) exhibited an inverted C-shaped pattern (higher in the southeast, lower in the central–west) with westward centroid migration; and electricity consumption dominated spatial heterogeneity, with synergistic nighttime light–PM2.5 interactions showing strongest nonlinear enhancement. Notably, Jianhe County maintained peak CBC (16.5) via forest carbon sinks, whereas Shiqian County suffered the steepest decline due to industrial encroachment. This work pioneers dynamic carbon coupling analysis in fragile ecosystems, offering transdisciplinary tools for global “dual-carbon” governance.