| Sumario: | This study develops and validates a UAV-based approach for estimating dry-season herbaceous biomass in semi-arid rangelands at Kapiti Research Station, Kenya. High-resolution RGB and multispectral imagery, combined with extensive field measurements, were used to characterize vegetation structure, differentiate herbaceous cover from shrubs and bare soil, and quantify spectral–biophysical relationships across diverse soil and vegetation types. Orthomosaics, vegetation indices (NDVI, NDRE, GNDVI, SAVI), and digital surface models were derived to assess fine-scale variations in forage conditions. Preliminary results indicate limited NDVI variability during the dry season and weak correlations with fresh biomass, highlighting the need to integrate additional indices and structural metrics. The dataset establishes a foundation for seasonal biomass modeling and will support wet-season analysis to capture vegetation response dynamics. Ultimately, this work contributes to developing high-resolution, spatially explicit forage monitoring tools to improve grazing management, drought preparedness, and decision-making for pastoral and agropastoral systems in East Africa.
|
|---|