Spatiotemporal Assessment of Forest Dynamics to Decipher Landscape Ecological Insights on Forest Fragmentation in Kangra region (2019-2025) using FRAGSTATS and Geospatial AI Techniques

: The dynamics of vegetation and forest changes are crucial for sustainable development in the era of increasing anthropogenic pressures and climate change. Here, we introduce a novel combination of cloud-based geospatial analysis and landscape ecology metrics to quantify forest dynamics, disturbance, and fragmentation in the Kangra region from 2019 to 2025. In this study we used multi-temporal classification with landscape metrics (e.g., forest cover fraction, connectivity indices) to measure temporal changes in the structure and fragmentation of forests over space. The use of GIS-based biophysical datasets and landscape metrics, including Class Area(CA), Patch Density(PD), Largest Patch Index(LPI), Aggregation Index(AI), and CLUMPY, revealed notable ecological and fragmentation-related changes throughout the study period. Temporal trends analysis highlighted shifts in forest structure, fragmentation, and connectivity. Dense forest and low forest classes increased by 2.02% and 3,19 %, respectively, whereas moderate forest and grassland decreased by -1.47% and -1.38% during 2019-2025. Classification accuracy, assessed using kappa coefficients, ranged from 0.84 (2019) to 0.93 (2025), indicating reliable performance. Significant re-vegetation was verified by regression analysis, where r 2 values ranged from 0.78 to 0.95. The findings further indicate substantial disturbances in forested areas, including fragmentation, deforestation, and changes in vegetation corridors in response to urban growth, land conversion, and climatic variability. The proposed integrated framework advances existing studies in the western Himalaya by enabling scalable, multi-year monitoring of fragmentation dynamics using cloud computing and landscape metrics, thereby offering improved capability for identifying disturbance hotspots and supporting sustainable forest management.