Exploring Rangeland Dynamics in Punjab, Pakistan: Integrating LULC, LST, and Remote Sensing for Ecosystem Analysis (2000–2020)

Abstract

Rangelands, covering 40% of Earth's land surface, play a crucial role in global ecosystem services and local livelihoods. In Punjab, Pakistan, these ecosystems face significant pressures from land-use changes and climate variability. This study aims to assess rangeland dynamics in Punjab's Khushab and Bhakkar districts from 2000 to 2020 using multitemporal satellite imagery and geospatial analysis. We integrated Land Use Land Cover (LULC) classification, Land Surface Temperature (LST) analysis, and vegetation indices to provide a comprehensive assessment of rangeland conditions. Random Forest and Support Vector Machine classifiers were employed for LULC mapping, achieving overall accuracies above 85%. Results reveal a significant decline in rangeland areas, from 9% (1 338.1 km²) in 2000 to 7% (983.7 km²) in 2020, primarily due to conversion to cropland. LST analysis showed an increase in both minimum (from 26.47 to 28.19°C) and maximum (from 44.60 to 47.44°C) temperatures. Vegetation indices indicated complex dynamics, with improvements in vegetation cover but periods of stress observed. These findings highlight the vulnerability of Punjab's rangelands to land-use changes and climate variability, emphasizing the need for targeted conservation strategies and sustainable land management practices. The study demonstrates the effectiveness of integrated remote sensing approaches in monitoring rangeland ecosystems and provides a foundation for evidence-based policy decisions to balance agricultural development with rangeland conservation in semiarid regions.