Soil erosion risk prediction: an assessment of qualitative and quantitative modeling approaches

Identifying and assessing erosion-prone areas in Iran's watersheds are crucial due to the annual loss of thousands of tons of fertile soil. This comparative study examined the qualitative Fargas and Institute for the Conservation of the Nature (ICONA) models against the quantitative Erosion Potential Method (EPM) model for assessing soil erosion risk in the Chehelgazi watershed, a sub-basin of Qeshlaq Dam in Sanandaj. The study extracted 380 random points from model predictions and used statistical methods to evaluate qualitative models against EPM as the reference. Results showed significant convergence in identifying high-risk erosion areas despite model structural differences, with the Fargas and ICONA models demonstrating 77.26% and 82.15% agreement with EPM, respectively, in the very high-risk category, confirming their reliability in assessing severe erosion risks. Analysis of the erosion risk maps produced by the models revealed that, based on the predicted areas by each model, 63–87% of the study area faces a high-to-very high-erosion risk, emphasizing the importance of multiple approaches in soil erosion management. These results underscore the importance of adopting multiple approaches to manage and control soil erosion. The recommendations of this study included developing artificial intelligence-enhanced integrated models, conducting comparative studies in watersheds with diverse geomorphological features, and examining the model’s sensitivity to environmental variables to enhance the identification and evaluation of erosion-prone areas. This study, emphasizing comparative analysis, provides a framework for the scientific application of qualitative models and represents an effective step toward comprehensive soil erosion management. The findings can inform soil conservation strategies and management programs by providing accurate assessments for researchers and land managers.