A simple model for the temporal variations of NDVI at regional scale over agricultural countries. Validation with ground radiometric measurements

Abstract

The NDVI profile computed from coarse resolution satellite records over temperate agricultural countries, is an average profile which results from the combination of radiative responses of the different species which are found within a pixel. In this paper we are interested in the extraction of the NOVI time profile of some specific crops using a semi-empirical model to test for the temporal variations of the NDVI. First, a simple model (a double logistic function with five parameters) is used to describe the temporal profile of the NOVI over an homogeneous canopy. The two parameters which locate the inflexion points, are particularly interesting to define the growing period. In this paper, we assume that, from a radiative point of view, the Beauce agricultural country (France) can be assimilated to an association of two homogeneous canopies: winter crops group and summer crops group. The double logistic functions which describe the NOVI time profile for these two groups are: NDVI_w(t) and NDVI_s,(t). Second, in such a case, an approximate model based on the sum of NDVI_w(t) and NDVI_s,(t) weighted by the areas, can describe the NOVI temporal data at a regional scale. A sensitive study is done, in order to examine how the application of this model to Regional Data permits the retrieval of parameters values close to that of NDVI_w(t) and NDVI_s,(t). The results show that with an a priori knowledge on the asymptotical value of each double logistic function, the positions of the inflexion points are retrieved with a maximum error of 6 per cent when the relative contributions of winter crops and summer crops are nearly the same, when an error of 11 per cent is reached for the location of the growing period of the summer crops and when their contribution to the regional response is reduced.