Abstract: Background and Objective: Land use information is increasingly essential for knowledge of the environment mainly in the context of global change. The objective of this study was to investigate the benefits of using field-collected ecological data and high-resolution images to map different types of land cover and vegetation from a Landsat OLI 8 image. Materials and Methods: The selection of training areas was made without and with the use of GPS data collected in the field and the google Earth tool. The supervised classification was applied using the maximum likelihood algorithm Spectral Angle Maping. In addition, data collected in the field made it possible to estimate above-ground biomass. Results: The results obtained showed that land use maps made without input from both ecological field data and Google Earth images have classes that are highly contaminated with pixels from other classes. The overall accuracy was estimated by applying the Intergovernmental Panel on Climate Change (IPCC) good practices It goes from O = 92.7% for the first map produced to O = 89.8% for the second map. In addition, the average carbon stock in our study area is estimated at 112 MgC ha–1 while it varies from 84 -213 MgC ha–1 in areas of land forests to and in areas of seasonally flooded forests. Conclusion: This small-scale study highlights several challenges to be addressed in reducing emissions from deforestation and forest degradation (REDD+) before arriving at a very accurate final map of land use in the intertropical zone.