What question did this study set out to answer?

This study assesses the spatial distribution of soil organic carbon and evaluates machine learning models for prediction accuracy.

July 30, 2026Discover SustainabilityOpen Access

Spatial distribution and modeling of soil organic carbon using machine learning and a geostatistical approach

View Full Paper

Ask AI

Bookmark

This study assesses the spatial distribution of soil organic carbon and evaluates machine learning models for prediction accuracy.
Compared five modeling approaches: Linear Regression, Random Forest, Support Vector Machine, Artificial Neural Network, and Extreme Gradient Boosting.
Used environmental covariates such as temperature, soil texture, and drainage in the Fogera Plain.
Conducted correlation analysis and semivariogram analysis for evaluating spatial dependence.
SVM achieved the highest predictive accuracy (R² = 0.936) compared to other models (LR: R² = 0.925, RF: R² = 0.851, XGBoost: R² = 0.842, ANN: R² = 0.774).
Soil erosion was identified as the dominant negative driver of SOC (r = −0.756, p < 0.01) while soil texture had a positive relationship (r = 0.501, p < 0.01).
SOC concentrations were higher in the western and southwestern regions and lower in the northern and southeastern areas.

No takes yet. Share an insight, caveat, or question.

Randomized trial evaluated SOC modeling in Fogera Plain, indicating SVM's superiority for predictions.

Tiruneh et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark