Anusha Boya Nakkala

Landslides are among the most destructive natural hazards in the Western Ghats region of Kerala, driven by complex interactions between geological, hydrological, and anthropogenic factors. This study aims to generate a high-resolution Landslide Susceptibility Level Map (LSLM) using a machine learning (ML)-based Random Forest (RF) model integrated with Geographic Information Systems (GIS). A total of 231 historical landslide locations obtained from the Bhukosh portal were used as reference data. Eight predictive factors—Stream Order, Drainage Density, Slope, Aspect, Geology, Land Use/Land Cover (LULC), Normalized Difference Vegetation Index (NDVI), and Moisture Stress Index (MSI)—were derived from remote sensing and ancillary datasets, preprocessed, and reclassified for model input. The RF model was trained and validated using a 50:50 split of landslide and non-landslide points, with variable importance values derived to weight each predictive factor of the raster layer in ArcGIS. The resulting Landslide Susceptibility Index (LSI) was reclassified into five susceptibility zones: Very Low, Low, Moderate, High, and Very High. Results indicate that approximately 17.82% of the study area falls under high to very high susceptibility, predominantly in the steep, weathered, and high rainfall zones of the Western Ghats. Validation using Area Under the Curve–Receiver Operating Characteristic (AUC-ROC) analysis yielded an accuracy of 0.890, demonstrating excellent model performance. The output LSM provides valuable spatial insights for planners, disaster managers, and policymakers, enabling targeted mitigation strategies and sustainable land-use planning in landslide-prone regions.

The Pandameru River Basin, South India, is affected by high levels of contamination from human activities and the over-exploitation of groundwater for agriculture, both of which pose significant threats to water quality and its availability for drinking and irrigation. To explore sustainable groundwater management, this study presents a machine learning-driven approach to basin-scale groundwater potential zone (GWPZ) mapping by integrating remote sensing (RS), a geographic information system (GIS), and the random forest (RF) algorithm. The research leverages ten thematic layers—including lithology, geomorphology, soil type, lineament density, slope, drainage density, land use/land cover (LULC), NDVI, SAVI, and rainfall—to assess groundwater availability. The RF model, trained with well-distributed groundwater data, provides an optimized classification of GWPZs into five categories: very good (5.84%), good (15.21%), moderate (27.25%), poor (27.22%), and very poor (24.47%). The results indicate that excellent groundwater zones are predominantly located along highly permeable alluvial deposits, whereas low-potential zones coincide with impermeable geological formations and steep terrains. Field validation using piezometric readings and well data confirmed significant variations in water table depths, ranging from 5 m to over 150 m. The groundwater potential map achieved an accuracy of 86%, underscoring the effectiveness of the RF model in predicting groundwater availability. This high-precision mapping technique enhances decision-making for sustainable groundwater management, supporting long-term water conservation, equitable resource allocation, and climate-resilient water strategies. By providing reliable insights into groundwater distribution, this study contributes to the sustainable utilization of groundwater resources in semiarid regions, aiding policymakers and planners in mitigating water scarcity challenges and ensuring water security for future generations.

Land degradation poses significant challenges to sustainable development, particularly in semi-arid regions where ecosystems are highly vulnerable. This study employs a cutting-edge geospatial and multi-criteria decision-making (MCDM) approach to delineate land degradation severity zones (LDSZs) in Anantapur district, India—a region characterized by persistent environmental stress. Utilizing thematic layers such as geology, geomorphology, soil properties, slope, and remote sensing indices (NDVI, MNDWI, NDSI, and LST), the study integrates high-resolution Landsat 8 OLI/TIRS (2023) and DEM datasets with local meteorological data for precise spatial analysis. The LDSZ classification identified critical degradation patterns, with river/stream/waterbody areas occupying 3.06% of the landscape and varying severity zones covering the remaining areas: very low (4.58%), low (20.56%), moderate (31.09%), high (27.62%), and very high (13.08%). Validation using the receiver operating characteristic (ROC) curve resulted in an area under the curve (AUC) value of 0.825, demonstrating the model’s reliability. By synthesizing geospatial data and MCDM, this research offers a dynamic framework for mapping and quantifying land degradation. It underscores the pressing need for context-specific land management practices to mitigate severe degradation while paving the way for broader applications in other semi-arid regions. This approach represents a significant leap in assessing and addressing land degradation, providing a robust scientific basis for future interventions and policy development.