African rangelands are ecologically and socio-economically significant ecosystems that require effective monitoring to support sustainable management. Remote sensing (RS) technologies, ranging from RGB, multispectral, and hyperspectral sensors to unmanned aerial vehicles (UAVs) and satellites, offer valuable tools for observing rangeland dynamics. Various data analysis approaches, including traditional regression techniques and emerging machine learning (ML) methods, have been used to monitor rangelands, such as classifying vegetation, estimating biomass, assessing rangeland condition, and identifying grazing patterns. This review systematically explores existing studies that apply RS for monitoring African rangelands, organizing the literature based on monitoring objectives and highlighting commonly used sensors, platforms, and methods. While ML approaches, such as convolutional neural networks (CNNs), show potential, their adoption remains limited due to data scarcity and computational constraints. In addition, hyperspectral data, UAV-based monitoring, and high-resolution commercial imagery are also underutilized, largely due to cost and accessibility challenges. The review refers to emerging opportunities such as integrating Synthetic Aperture Radar (SAR) and optical data, leveraging community-driven data collection, and combining UAV and satellite imagery. The paper concludes by outlining current challenges and pointing to potential directions for further research. • Review RS platforms, sensors, and methods for African rangeland monitoring. • Discuss RS, ML, and other approaches applied in African rangeland studies. • Highlight challenges and research gaps in RS and ML use across African rangelands.
Hosseini et al. (Sun,) studied this question.