Refining erosion assessment with NDVI-based modeling: a case study in diverse climatic zones

Estimating soil erodibility (C-Factor) in data-scarce regions is complex due to the lack of appropriate and abundant data, as well as the variability in climatic and land cover conditions. This study develops and validates models to estimate C-Factor values based on Normalized Difference Vegetation Index (NDVI) data in Morocco's distinct settings. We delineated three climatic zones using the De Martonne Index and historical climate data: humid, transitional, and arid. The NDVI/C-Factor relationship was modeled using a five-parameter logistic (5PL) function, with each zone modeled separately. Validation using Root Mean Square Error (RMSE) showed promising performance, with RMSE values of 0.039, 0.047, and 0.053 for humid, transitional, and arid zones, respectively. In the humid zone, NDVI values ranged from 0.16 to 0.723, with corresponding C-Factor values from 0.0051 to 0.43, closely matching those in the literature (0.001–0.50). In the transitional zone, NDVI values ranged from 0.04 to 0.7, and C-Factor values from 0.06 to 0.70, aligning well with reported values (0.04–0.75). In the arid zone, NDVI values ranged from 0.04 to 0.45, with C-Factor values from 0.14 to 0.98, consistent with established values (0.10–1). These models provide valuable insights for erosion assessment, advancing methodologies for diverse climatic and land cover conditions. Our approach offers a precise, tailored assessment of soil erosion, supporting effective land management strategies.