Spatial assessment of soil erosion variability in Kwara State, Nigeria using RUSLE and GIS

Abstract

Soil erosion poses a significant challenge to agricultural productivity and land sustainability in Kwara State, Nigeria. This study assesses the spatial variability of soil erosion risk across the state using the Revised Universal Soil Loss Equation (RUSLE) integrated with Geographic Information Systems (GIS) and remote sensing data. Rainfall erosivity (R-factor) was estimated from 30 years (1992 – 2022) of mean annual precipitation using an empirical relationship suitable for data-limited environments, while other RUSLE factors (K, LS, C, and P) were derived from soil data, digital elevation models, and satellite-based vegetation indices. The model outputs were integrated to generate a spatial map of average annual soil loss, which was classified into low, medium, and high erosion risk categories. Results indicate that approximately 73.15% of the study area falls within the low-risk class, 23.04% within the medium-risk class, and 3.81% within the high-risk class, with higher erosion susceptibility observed in areas associated with steeper slopes, sparse vegetation, and intensive land use. Model validation was conducted using a limited number of ground-referenced observations through a confusion matrix approach, yielding an overall accuracy of 93% and a kappa coefficient of 0.87. These metrics indicate good agreement but should be interpreted with consideration of the sample size and spatial coverage of validation data. The findings provide a regional-scale indication of erosion risk patterns and highlight areas where soil conservation measures may be prioritized. However, the results represent model-based estimates of relative erosion susceptibility and should be interpreted cautiously. The study offers a useful baseline for supporting land management planning while underscoring the need for further validation and uncertainty analysis.