Quantifying the "Urban Carbon Penalty" in the Nigerian Sudan Savanna: A Multi-Temporal Remote Sensing and Sensitivity Analysis (2005–2025)

Abstract

Land use and land cover change is a primary driver of terrestrial carbon emissions, yet the differential carbon costs of specific transition pathways in dryland ecosystems remain poorly quantified. This study analyzes the spatiotemporal dynamics of LUCC and associated carbon fluxes in Kebbi State, Nigeria, over twenty years (2005–2025). Using multi-temporal Landsat and Sentinel-2 imagery, we achieved classification accuracies exceeding 87% (Kappa > 0.84) to map six land cover classes. Results indicate a profound landscape transformation: built-up areas tripled (+6.3%), farmland expanded by 8,330 km2, while dense vegetation declined by 9.1%. Integrating these maps with IPCC Tier 2 carbon density parameters validated against local field data, we estimated a net carbon loss of 33.7 MtC (69.25 MtCO2e). Crucially, sensitivity analysis revealed that the conversion of dense vegetation to built-up areas carries a disproportionately high "urban carbon penalty" (Sensitivity Index = 1.76), contributing 17.97% of total emissions from only 10.21% of converted area. In contrast, agricultural extensification onto sparse vegetation exhibited low sensitivity (SI = 0.40). Ordinary Least Squares regression confirmed built-up expansion (β = 0.68, p < 0.001) as the strongest predictor of carbon loss. These findings challenge the assumption that agricultural expansion is the sole dominant driver of dryland emissions, highlighting the critical need for spatially explicit urban containment policies to mitigate high- intensity carbon losses in rapidly urbanizing Sahelian regions.

Keywords

References