Spatiotemporal dynamics of carbon emissions induced by land-use change and their implications for climate resilience in West African drylands

Abstract

Understanding the effects of land-use and land-cover change (LULC) on carbon emissions is crucial for ensuring the sustainable management of dryland ecosystems. This study investigates spatial and temporal changes in carbon stock associated with LULC in one of the key biodiversity hotspots of the West African drylands. The study uses a Landsat imagery dataset spanning 2000, 2010, and 2022, processed in Google Earth Engine using a random forest algorithm. LULC-induced changes in carbon stock are estimated using ecosystem-specific emission factors for the designated study area. Additionally, integrating NDVI and climate stressors enables the projection of vegetation productivity and future carbon trajectories. The findings indicate a substantial decline in wooded savanna (~ 21%), tree savanna (7%), and shrub savanna (5%), alongside a ~ 35% increase in cropland. These transitions generated an estimated 300,000 MgC of carbon stock changes between 2000 and 2022. Climate projections under SSP1-2.6 and SSP5-8.5 scenarios indicate continued warming and reductions in annual rainfall of 5.414 mm and 9.359 mm, respectively, leading to additional carbon increases of ~ 1.3 MgC/ha and ~ 2 MgC/ha by 2070. These combined pressures accentuate land degradation and climate feedbacks. It emphasizes the need for integrated strategies to enhance climate resilience in West African dryland ecosystems in line with Sustainable Development Goal (SDG 15).