Monitoring Afforestation Areas with Earth Observation: A Case Study in Togo

Abstract

In Agou, Togo, a strategic afforestation initiative is actively restoring degraded landscapes by planting tree species. This study assesses the utility of the Normalized Difference Vegetation Index (NDVI) to monitor vegetation changes using Sentinel 2 satellite imagery from 2016 to 2024. Leveraging the capabilities of the Google Earth Engine (GEE) platform, NDVI values were processed to identify the top 95th percentile annually. The methodology extended to determining spatial correlations among vegetative covers, establishing that these were significantly clustered. A multidimensional cluster validation in R determined an optimal count of three clusters. This facilitated the implementation of a K-means unsupervised classifier that integrates NDVI-derived metrics, including trend analysis, cumulative values, and slope calculations. Additionally, assessment of regional precipitation and temperature data were incorporated into the analysis, which showed no statistically significant impacts. Our findings underscore the effectiveness of integrating remote sensing analysis and its convenience in monitoring changes in vegetation land cover.