Moses Isabirye
Research Interests
Moses Isabirye has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Daniela Requena Suarez
Section 1.4 Remote Sensing and Geoinformatics, GFZ Helmholtz Centre for Geosciences, Potsdam, Germany
1 shared publication Jorame Bahati
1 shared publication- Martin Herold
Helmholtz GFZ German Research Centre for Geosciences, Remote Sensing and Geoinformatics Section, Telegrafenberg, Potsdam, Germany
1 shared publication - Nandika Tsendbazar
Wageningen University & Research, Laboratory of Geo-Information Science and Remote Sensing, Droevendaalsesteeg 3, 6708 PB Wageningen, the Netherlands
1 shared publication - Olga Nepomshina1 shared publication
- Panpan Xu
Laboratory of Geo-Information Science and Remote Sensing, Department of Environmental Sciences, Wageningen University & Research, 6708PB Wageningen, The Netherlands
1 shared publication
Latest Publications
- Open Access
- 23 March 2026
- Published by Elsevier in SSRN Electronic Journal
High-resolution global land cover maps assessment for national-scale area change estimation and reporting: case study for Uganda
Accurate estimation of land cover (LC) change areas is essential for national-level monitoring and reporting of land use dynamics. Recently developed multi-year 10-meter global land cover (GLC) products require evaluation for their reliability for assessing land cover change and area estimate. This study evaluates the use of two high-resolution GLC products, Dynamic World (DW) and ESRI LULC for land cover change area estimation between 2019–2023 in Uganda. Eight change classes including deforestation, reforestation, urbanization, land abandonment, cropland expansion, wetland expansion, wetland degradation, and other transitions, alongside the no-change class were assessed. A design-based stratified random sampling was developed and applied to assess LC change accuracy and area estimation, including uncertainty quantification. Additionally, a model-assisted approach was evaluated to explore potential improvements in area estimates using land cover change probability data. For land cover change, the maps achieved an overall accuracy of 71.89% (±2.57%) for DW and 76.67% (±2.47%) for ESRI LULC. The no-change class is rather accurate, while change classes show low accuracy making the maps not suitable for direct change estimation. In particular, user’s accuracy for all change classes was low indicating overestimation of changes. The area estimation of LC change trends indicated that land degradation predominated over land improvement. The model-assisted approach provided modest additional enhancement, with limited gains in precision for national level reporting. The findings underscore a cautionary note on using current GLC maps for change assessments and area estimation, particularly those of small magnitude or high fragmentation, to inform policy and decision-making.
- 23 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessHigh-resolution global land cover maps assessment for national-scale area change estimation and reporting: case study for Uganda
Accurate estimation of land cover (LC) change areas is essential for national-level monitoring and reporting of land use dynamics. Recently developed multi-year 10-meter global land cover (GLC) products require evaluation for their reliability for assessing land cover change and area estimate. This study evaluates the use of two high-resolution GLC products, Dynamic World (DW) and ESRI LULC for land cover change area estimation between 2019–2023 in Uganda. Eight change classes including deforestation, reforestation, urbanization, land abandonment, cropland expansion, wetland expansion, wetland degradation, and other transitions, alongside the no-change class were assessed. A design-based stratified random sampling was developed and applied to assess LC change accuracy and area estimation, including uncertainty quantification. Additionally, a model-assisted approach was evaluated to explore potential improvements in area estimates using land cover change probability data. For land cover change, the maps achieved an overall accuracy of 71.89% (±2.57%) for DW and 76.67% (±2.47%) for ESRI LULC. The no-change class is rather accurate, while change classes show low accuracy making the maps not suitable for direct change estimation. In particular, user’s accuracy for all change classes was low indicating overestimation of changes. The area estimation of LC change trends indicated that land degradation predominated over land improvement. The model-assisted approach provided modest additional enhancement, with limited gains in precision for national level reporting. The findings underscore a cautionary note on using current GLC maps for change assessments and area estimation, particularly those of small magnitude or high fragmentation, to inform policy and decision-making.