Overview A map of smallholder-dominated landscapes covering the provinces Niassa, Zambezia, Cabo Delgado, and Nampula in Northern Mozambique. The map includes active cropland and short-term fallows as separate classes, as well as five land cover classes (herbaceous vegetation, open woodlands, closed woodlands, non-vegetated land, water). The map is based on PlanetScope mosaics and consequently comes at 4.77m spatial resolution. The download contains the following files: ps_lc_nmoz.tif / .qml: land cover map and associated QGIS style file ps_lc_nmoz_probmargins.tif / .qml: probability margins and associated QGIS style file training.gpkg: training samples with class labels LICENSE.pdf: NICFI data program user license Map accuracy We conducted an area-adjusted accuracy assessment based on a stratified random sample, which yielded important insights regarding accuracies and error types. The area-adjusted overall accuracy of the map is 88.9%, but users should be aware of the most important error types: Active cropland were overestimated, whereas local topographical depressions with moist soils, and regions with exposed soils/rocks and sparse vegetation cover were found to be falsely classified. Short-term fallows were underestimated, particularly in regions with high growth rates and extensive land management, such as parts of the northern and north-eastern study region. Further resources The production of this map was made possible through the NICFI data program, providing the PlanetScope mosaics and the Google Earth Engine cloud computing platform for preprocessing of the satellite data and classification. As such, the use of the map falls under the NICFI data program license agreement included in the download. The code for preprocessing the PlanetScope mosaics is based on the Google Earth Engine Python API and made available at https://github.com/philipperufin/eepypr/. We advise map users to read the preprint or the open access paper for detailed insights. In case of questions please consult these resources or contact the lead author of the work.

This work was supported by the FRS-FNRS, grant no. T.0154.21 and the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (Grant agreement No 677140 MIDLAND). This research contributes to the Global Land Program. The authors would like to thank Yara Ubisse, Sá Nogueira Lisboa, and Dr. Almeida Sitoe for their invaluable help in preparing and conducting the fieldwork in Northern Mozambique in 2021.