The increasing problem of soil degradation, marked by the reduction of organic content and rising soil salinity is an urgent issue in many regions globally. Effective monitoring of soil properties, including soil organic carbon (SOC), pH, and electrical conductivity (EC), is crucial for sustainable land management to achieve high crop yields and mitigate degradation processes. This research focuses on mapping the spatial distribution of topsoil (0–15 cm) soil health properties, particularly SOC, pH, and EC, in the Washim District of Maharashtra, using advanced machine learning techniques. In the study area, measured SOC content (based on 1944 sampling locations) ranged from 0.06 to 1.23%, with an average value of 0.505%. Soil pH ranged from 6.2 to 8.82, with an average of 7.849. EC varied significantly from 0.02 to 8.7%, with an average value of 2.3%. In addition to soil health properties, key nutrient contents including Nitrogen (N), Phosphorus (P) and Potassium (K) were also mapped. Mapping these nutrients is crucial as they are essential for plant growth and development. Multiple statistical and machine learning models including Multiple Linear Regression (MLR), Random Forest (RF) and Extreme Gradient Boosting (XGBoost) were applied to model soil properties using environmental, topographic, and climate variables derived from remote sensing data. The performance of each model was evaluated using the coefficient of determination (R²) and root mean squared error (RMSE). The analysis of variable importance revealed that remote sensing variables, particularly soil moisture, precipitation, NDVI and elevation, were dominant in the spatial prediction of soil parameters. This research contributes to the broader field of digital soil mapping, highlighting the potential of integrating machine learning models with multi source remote sensing data to monitor and manage soil health and nutrient levels effectively enabling scalable assessments of soil properties.
Keywords: Digital Soil Mapping, Soil Properties, Remote Sensing, Machine Learning, Spatial Prediction.