Remote sensing capabilities to estimate pasture production in France
Pastures constitute an important terrestrial ecosystem. In France, pastures occupy 21% of the total area. A big effort is being made to develop a real-time systematic approach to estimate biomass production at a national level, focusing on spatial and seasonal variability in relation to drought. The absence of indirect low-cost methods that could be applied to large areas contributes to this situation. Advances in remote sensing and crop models offer new methodological and operative possibilities to solve this problem. In this paper, 13 forage regions (FR) in France were selected on the basis of their different geomorphologic, climatic and soil conditions with regard to pastoral production. Images from Système Probatoire de l'Observation de la Terre (SPOT) 4-VEGETATION were used to forecast productive variables estimated by the STICS-Prairie simulation model. In general terms, both satellite and productive data agreed properly. Particularly, the relationship between the middle infrared based vegetation index (SWVI) and the Leaf Area Index (LAI) demonstrated the best results whatever the FR. Results obtained confirm the capabilities of remote sensing data as an accurate predictor of productive variables estimated as from simulation models. Differences between satellite information and model estimations of pasture systems, especially during the harvesting periods, could be good indicators to improve model estimations at a regional scale as well.