1 août 2020
Arsène Adou Kouassi et al., « Land use land cover ultra-high-spatial-resolution digital maps (5-cm) for 4 urban districts of the city of Bouaké (Air France, Belleville, Koko, Sossoribougou), center Côte d'Ivoire, 2020 », DataSuds, ID : 10.23708/PUYNSG
This dataset provides ultra-high-spatial-resolution Land Use / Land Cover (LULC) digital maps of 4 urban districts of Bouaké, a city located in center Côte d’Ivoire (Ivory Coast). They were produced using images acquired by Unmanned Aerial Vehicles (UAVs) in 2020 at a spatial resolution of 5 cm. They were originally created to study the landscape determinants of the abundance of the main malaria mosquito species in these districts, and are made available here for a wide range of uses. These LULC products contain 8 land cover classes: bare soil, scrub, rice-growing surfaces, market gardening, dense deciduous trees, palm trees, completed buildings, unfinished buildings. The method used to generate the maps involved a supervised object-based image classification using aerial images acquired by an UAV, a ground-truth dataset acquired by photo-interpretation, and a random forest classifier. The classification accuracy varies from 91% to 94% depending on the district. In addition to the LULC georeferenced raster datasets of the four districts, we deliver the following files in this release: the raster attribute table, including definitions of the land cover classes in English and French ; a layer style (*.qml) to visualize the LULC georeferenced raster datasets in QGIS ; a map of the LULC products as a .jpeg image (for visualization purposes only) ; a map of the original orthophotos acquired by the UAV as a .jpeg image (for visualization purposes only) ; representative pictures of the land cover classes ; the detailed methodology used to generate the data, with the resulting confusion matrices (in English and French) ; the ground-truth georeferenced datasets used for training and test in the classification (including the predictive variables).