Accurate forest biomass estimation is essential for quantifying carbon stocks, guiding sustainable forest management, and informing climate change mitigation strategies. Kenya’s forests are diverse, ranging from Afromontane and mangrove ecosystems to dryland woodlands and plantations, each presenting unique challenges for biomass measurement. This review synthesizes literature on field-based, remote sensing, and machine learning approaches applied in Kenya, highlighting their effectiveness, limitations, and integration potential. A systematic search across multiple databases identified peer-reviewed studies published in the last decade, screened against defined inclusion and exclusion criteria. The main findings are (1) Field-based techniques (e.g., allometric equations, quadrat sampling) provide reliable and site-specific estimates but are labor-intensive and limited in scalability. (2) Remote sensing methods (LiDAR, UAVs, multispectral and radar imagery) enable large-scale and repeat assessments, though they require extensive calibration and investment. (3) Machine learning and hybrid approaches enhance prediction accuracy by integrating multi-source data, but their success depends on data availability and methodological harmonization. This review identifies opportunities for integrating field and remote sensing data with machine learning to strengthen biomass monitoring. Establishing a national biomass inventory, supported by robust policy frameworks, is critical to align Kenya’s forest management with global climate and biodiversity goals.
Mkuzi et al. (Sat,) studied this question.