sing remote sensing information and simulation models to estimate pastures production at the national scale
Abstract
Pastures constitute an important terrestrial ecosystem. In France, pastures occupy 21% of the total area. There is a big effort to develop a real time systematic approach to estimate biomass production at the national level, focusing on spatial and seasonal variability in relation to drought. The non-existence of indirect methods of low cost that can be applied to large areas contributes to this situation. The advances registered in crop modeling and remote sensing offer new methodological and operative possibilities to solve this problem. In this work, thirteen Forage Regions (FR) of France were selected for their different geomorphologic, climatic and soil conditions with regard to pastoral productions. Images from the VEGETATION sensor on board SPOT4 satellite (1 km spatial resolution) were used to forecast some productive variables estimated by STICS Prairie simulation model. There was a general good agreement between the satellite and productive data. Particularly, the relationship between the middle infrared based Vegetation Index (SWVI) and the Leaf Area Index (LAI) demonstrated the best results regardless the Forage Region (FR). The obtained results confirm the capabilities of remote sensing data to be an accurate predictor of productive variables estimated from simulation models. Moreover, differences between satellite information and model estimations, especially during the harvesting periods of pasture systems, could be a good indicator to improve model estimations at the regional scale.
- Publication:
-
EGS - AGU - EUG Joint Assembly
- Pub Date:
- April 2003
- Bibcode:
- 2003EAEJA.....1855D