Effect of Spatial Resolution on Land Cover Mapping in an Agropastoral Area of Niger (Aguié and Mayahi) Using Sentinel-2 and Landsat 8 Imagery Within a Random Forest Regression Framework

Monitoring environmental changes over time requires images with extensive historical depth. However, high spatial resolution images often lack such depth. This study investigates the impact of spatial resolution on image classification. Thus, Landsat 8 and Sentinel-2 images acquired between October and December 2020 were processed and classified using Random Forest regression on Google Earth Engine (GEE). This method allows for continuous land cover maps, required for robust assessment of land cover dynamics in patchy landscapes. A total of 1719 training samples were collected from the Collect Earth Online (CEO) platform to train the model. In addition to the spectral bands, vegetation indices were considered to optimize classification results. The study revealed statistical differences in land cover areas estimated by the two sensors. These differences are statistically significant at p < 0.001, although they are small. Validation results showed that the RMSE from Sentinel-2 is slightly lower than that from Landsat 8, with this difference significant at p < 0.05. Therefore, spatial resolution influences the accuracy of image classification. Nevertheless, given the observed differences between the two sensors, which ranged from 0.03% to 3.94% across land covers, Landsat imagery remains suitable for producing reliable land cover maps in heterogeneous landscapes.