Optimizing Remote Sensing Data and Light Use Efficiency Model for Accurate Gross Primary Production Estimation in African Rangelands

This paper focuses on the meticulous selection of optimal remote sensing and climate datasets for Gross Primary Productivity (GPP) estimation in African rangelands. Utilizing Eddy Covariance Flux Tower data, we refine data selection and employ a Light Use Efficiency (LUE) model, with Sentinel 2 for photosynthetically active vegetation quantification, MODIS for Photosynthetically Active Radiation (PARin), and ERA5 Land reanalysis for climatic variables. The Eddy Covariance-based LUE-GPP model is identified as superior compare to other LUE based GPP models and further enhanced through fine-tuning LUE
max
and climate scalars. Footprint analysis determines a 500m footprint size, aligning with literature recommendations. Comparative analyses with various LUE models reveal ECLUE's superiority. Statistical validations affirm key parameter selections, leading to a reliable LUE-based GPP model tailored for African rangelands. The proposed model contributes to accurate GPP assessment, essential for informed environmental stewardship in these critical ecosystems.