A Framework for Integrating Heterogeneous Earth Observation Data for Comprehensive Land Degradation Neutrality Assessment in Nigeria

Achieving Land Degradation Neutrality, as committed to under the United Nations Convention to Combat Desertification, requires comprehensive spatial monitoring frameworks capable of integrating diverse Earth observation datasets across Nigeria's varied ecological zones. This study develops a theoretical and operational framework for integrating heterogeneous remote sensing datasets including MODIS, Landsat, Sentinel-2, and SAR data products for national-scale land degradation neutrality assessment. The framework addresses the fundamental challenges of sensor interoperability, spatial scale discordance, and temporal inconsistency that limit the utility of individual datasets for national assessments. A multi-scale spatiotemporal data fusion methodology is developed and applied across three pilot regions representing Nigeria's Sudan Savanna, Guinea Savanna, and Humid Forest zones. The framework incorporates the three biophysical sub-indicators prescribed by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, namely land cover change, land productivity dynamics, and carbon stocks. A Bayesian data assimilation approach integrates uncertainty from multiple input datasets into a probabilistic degradation assessment output. Validation using independent reference datasets across 450 field plots achieves a degradation classification accuracy of 85 percent. The framework provides Nigeria's Federal Ministry of Environment with an operational tool for biennial land degradation neutrality reporting. Keywords: land degradation neutrality, Earth observation, data integration, multi-scale, Nigeria