Estimation of standing crop biomass in rangelands of the Middle Atlas mountains using remote sensing data

In the Middle Atlas rangelands, traditional methods for estimating standing crop biomass are labour-intensive and impractical. Remote sensing offers an initiative for standing crop biomass large-scale monitoring. The aim of this study was to estimate standing crop biomass, comprising annual and perennial forbs, grasses and perennial shrubs, using remote sensing data. The vegetation indices (NDVI, DVI, RVI, MSAVI and OSAVI) were derived from medium-resolution Landsat 8 and MODIS imagery. Sixty sampling sites were used for the biomass data collection. These sites were located across three grazing areas and data were collected in May and June 2016. Regression models were established between biomass field data and the five indices. Correlation analysis indicated that among the five vegetation indices, only DVI had the lowest value (r = 0.60). Linear models developed between the biomass field data and vegetation indices showed that NDVI, OSAVI and RVI explained a reasonable percentage of the variance in biomass. Values for R2 were 0.74, 0.77 and 0.71, respectively. Among these indices, the OSAVI performed better, with a high R2 and low error (MAPE = 11.03%). The established models represent a key tool for long-term monitoring of these rangelands.

This study was conducted with the supportof the Institut Agronomique et Vétérinaire (IAV) Hassan II, AfricanSustainable Agriculture Research Institute (ASARI-UM6P), InstitutNational de la Recherche Agronomique (INRA) Oujda in Morocco, andInstituto Pirenaico de Ecología (CSIC) in Spain. Special thanks to allthe participants for their time and willingness to share their knowledge.

Peer reviewed