The present study integrates GIS and remote sensing technology to do a morphometric analysis of Karanthaimalai Hill in Tamil Nadu's Dindigul district, India. Morphometric analysis is the measurement and quantitative study of many Earth surface features, including shape, length, height, and slope. Karanthaimalai Hill in the Natham region is mostly composed of granite and gneiss. The primary purpose of this study is to identify a wide range of morphometric parameters, including slope, contour, aspect, curvature, drainage, elevation, flow accumulation, flow direction, drainage density, dissection index, relative relief, roughness, and hill stream order. The data used in this study include Survey of India (SOI) toposheet maps and Shuttle Radar Topography Mission (SRTM) digital elevation model (DEM) satellite photos with a resolution of 30 metres. The study found that Karanthaimalai Hill has steep and highly dissected topography, with heights ranging from 300 m to 900 m. Notably, the hill provides large relative relief, particularly at the summit and middle, with moderate relief throughout. The dissection index reveals extensive erosion and stream incision, notably on the hill's southern side. The topography roughness varies greatly across the area, indicating a diverse terrain. The slope aspect is primarily orientated west and east, while the hill's curvature displays steep slopes and dips. Furthermore, the hill's drainage density represents a complex network of streams and drainage patterns. These findings have significant implications for regional land-use planning, conservation, and management strategies. However, additional research is required to understand the fundamental principles that affect the landscape and to promote sustainable land management approaches.
Unregulated and aggressive Land Use Land Cover (LULC) dynamics such as urban sprawl and informal deforestation require a monitoring system that is close to real time, accurate, and readily available. Currently available commercial and cloud-based systems often demand high levels of technical knowledge in geospatial and programming, posing a stumbling block to local practitioners and NGOs. We propose Dhristhi, an automated, end-to-end Web-GIS platform designed to overcome this expertise bottleneck. Dhristhi integrates a sophisticated hybrid methodology: it employs a pre-trained U-Net Deep Learning model for high-precision, pixel-wise semantic segmentation of multispectral imagery, followed by an Object-Based Post-Classification Comparison (OBC) approach to aggregate and validate changes into meaningful geographic regions. An important component of the framework is a Random Forest (RF) classifier with user defined dynamic thresholding mechanism & seasonal variation in values of spectral indices (NDVI, NDBI) to prevent false positive and improve change validation. This platform works on a fully automated, plug-and-play workflow, it automatically handles everything from data collection to pre-processing, so anyone can use the platform easily without having technical expertise. After any successful change detection system automatically sends alert and generates report in the user’s dashboard. Dhristhi successfully demonstrates advanced geospatial intelligence, providing a robust, noise-resilient tool for environmental and civil governance. On a validation dataset covering forest-sprawled and urban regions, the proposed system achieves an overall change detection accuracy of 92.1%, with a Kappa coefficient of about 0.88 and mean IoU of about 0.89 for binary anthropogenic change detection.
Accurate temperature prediction is essential for climate adaptation, environmental monitoring, and sustainable urban planning. This study evaluates the performance of two machine learning techniques Random Forest (RF) and Gradient Boosting Regression (GBR) for predicting near-surface air temperature in the Greater Accra Region of Ghana. Daily temperature observations obtained from the Ghana Meteorological Agency covering the period 1960–2018 were used for model development. The average daily air temperature was computed from minimum and maximum temperature observations. The predictive performance of the models was compared with a classical statistical time-series model, Autoregressive Integrated Moving Average (ARIMA). Model evaluation was performed using five-fold cross-validation to improve the robustness of the results. Performance metrics included Mean Squared Error (MSE) and the coefficient of determination (R²). The results show that the Random Forest model achieved the highest predictive accuracy with MSE = 0.0010 °C and R² = 0.9996, while the Gradient Boosting Regression model produced MSE = 0.0015 °C and R² = 0.9994. The ARIMA model showed significantly lower performance with MSE ≈ 0.598 °C and R² ≈ 0.30. The high predictive performance of the machine learning models is partly attributed to the deterministic relationship between the input variables and the computed target temperature. The study demonstrates the potential of machine learning approaches for climate-related prediction tasks and provides insights for environmental planning and climate resilience strategies in rapidly urbanizing regions such as Greater Accra.
A study was conducted in the Tirupathur district of Tamil Nadu, India to identify suitable sites for groundwater recharge and to suggest appropriate site specific recharge mechanisms. The potential of groundwater depends on topography, lithology, geological structure, depth of weathering, slope, drainage pattern, landuse land cover, soil, rainfall, lineament density, drainage density, magnetic breaks and topographic wetness index. All thematic layers were prepared and assigned comparative weights using Saaty's 9-point scale and then normalized using the Analytical Hierarchy Process. According to the investigation, groundwater recharge zones are categorised into five classes; very low, low, moderate, high, and very high. The study found the region of Vaniyambadi and Natrampalli had very high and high potential zones, respectively, covering 6.22% (128.75km2) and 15.2% (312.79km2) area. Conversely, the region of south Natrampalli, Tirupathur, and eastern Ambur had moderate, low, and very low potentials, covering 29.31% (607.06km2), 24.35% (518.30km2), and 25.02% (504.41km2) area. The study mainly focused on moderate to very low potential zones for artificial recharge. High and very high zones were not considered as priority due to their high infiltration rates. This approach helped to identify 46 potential sites for artificial recharge based on the best execution of AHP to boost groundwater conditions and meet the shortage of water resources in agriculture and domestic use. This study reveals that Remote Sensing and GIS with AHP provide an efficient and effective platform for convergent analysis of various data for groundwater management and planning.
The Hindu Kush Himalaya (HKH) is a globally significant biodiversity hotspot, with extensive forests, protected areas (PAs) and substantial carbon reserves. However, increasing frequency and intensity of forest fires threaten its ecological integrity. Despite these concerns, there is a lack of comprehensive spatial assessment of forest fire susceptibility, necessitating a data-driven approach to evaluate environmental risks. This study evaluates forest fire susceptibility and its impact on biodiversity and carbon stocks across the HKH region using remote sensing and machine learning models. The models include Analytic Hierarchy Process, Certainty Factor, Maximum Entropy, and Random Forest (RF), based on thirteen ignition factors, representing environmental, meteorological, edaphic, socio-economic, and topographic factors. Active fire data from MODIS and VIIRS was used for training and testing of models. Model performance, evaluated using Area Under Curve (AUC) of Receiver Operating Characteristic curve, showed that RF (AUC = 0.95) outperformed other models. Results indicate that about 13.54–20.47% of HKH forested region is highly susceptible to forest fires, with higher risk in Himalayan belt, Bangladesh, and Myanmar. Key factors influencing fire risk include wind speed, solar radiation, elevation, and precipitation. Forest fires threaten biodiversity, with around 25,878.66 sq. km of PAs identified as highly vulnerable. Additionally, fire-induced carbon emissions from aboveground biomass, estimated at 32.22 million Mg, jeopardize carbon stocks by depleting stored carbon and increase atmospheric CO₂ levels. Forest fire susceptibility maps and risk assessments provide essential spatial insights for policymakers, supporting proactive fire mitigation, biodiversity conservation efforts, and carbon management.
