Development of a model for estimating the biomass of pastures using normalized difference vegetation index (NDVI): the case of the Niassa pastoral zone in Burkina Faso

<div>In Burkina Faso, livestock feeding is based on pasture productivity, which is a key factor in managing livestock productivity. As such, this management requires a good knowledge of pastures. However, the high cost of producing pasture data and the inaccessibility of pastures due to insecurity mean that there is a low availability or lack of data on pasture. This makes it difficult for stakeholders to take decisions to improve livestock feed in the event of a local food crisis. This is why this study aims to facilitate data production by developing a model for estimating herbaceous biomass in pastures using land observation tools. To this end, the study correlated biomass data collected on the ground with satellite imagery. A total of 24 the wooded, shrubby, and grassy savannahs and gallery forests vegetation units were used as ground observation stations. Simple multilinear regression was chosen for the correlation test. The model parameters gave an R² of 0.70, an RMSE of 236.7 kg DM/ha with a p-value of less than 0.0001 and an average production of 1,423 kg DM/ha. Based on these parameters, there is a strong relationship between vegetation indices and pasture biomass. The equation Y=45*Vav+9*Vmx-19*Rrg-2,256 derived from the model can be used to estimate pasture herbaceous biomass. This will improve the availability of data on pasture forage resources for better action planning in a context of insecurity.</div>