Temporal Analysis of Landuse and Land Cover Changes in Siliguri City (1991- 2021), West Bengal, India : A Hybrid Machine Learning and GIS Approach

Land Use and Land Cover (LULC) dynamics are critical for understanding urban growth and associated environmental pressures. This study evaluates the spatio-temporal LULC changes in Siliguri City from 1991 to 2021 using a hybrid approach integrating Google Earth Engine (GEE), Random Forest (RF) classification, and GIS-based spatial analysis. Results show a persistent expansion of built-up areas across three decades. Urban built-up land increased from 12.42 sq km in 1991 to 23.74 sq km in 2021, while Rural/slum built-up areas expanded from 1.92 sq km to 15.09 sq km, reflecting rapid peripheral urbanization. Natural land classes consistently declined: vegetation reduced from 7.56 sq km to 2.72 sq km, agricultural land from 6.83 sq km to 1.13 sq km, and scrub/shrub land nearly disappeared (from 9.87 sq km to 0.0045 sq km), indicating extensive land conversion. Water bodies and riverine areas showed fluctuations and an overall reduction, underscoring growing hydrological stress. Newly developed parks in 2021 (1.26 sq km) represent limited attempts at restoring green space. Overall, the findings reveal aggressive urban expansion at the cost of natural and agricultural land, emphasizing the need for stricter land-use regulations, ecological conservation, and sustainable urban planning to guide Siliguri's future