1112
tree covers influence the total measured reflectance of the scene modifying the relationship ENDVI /
herbaceous production ; secondly, the production model used assumes a direct relationship between p rimar y
production and Absorbed PAR (APAR) neglecting the effects of other limiting factors on the vegetation growth.
Recently, to overcome this last limitation, Prince (1991b) proposed a modified Monteith's production model
which includes different coefficients taking into account important limiting factors like water availability and
nutrient shortage. However, as such, the representation of vegetation processes remains essentially empirical
and important climatic and physiological factors affecting vegetation activity cannot be satisfactorily analysed.
An alternative approach relies on the combined use of remote sensing spectral measurements with an ecosystem
model which describes the relevant processes of the soil-vegetation-atmosphcre system to simulate vegetation
growth and soil water dynamics. In this approach, the ecosystem process model is not bound by the sole use of
satellite data, as for the Monteith's production model, but integrates the latter in an explicit formulation of the
main processes. In this way, a close analysis of the relationships between different processes (photosynthesis,
transpiration,...) described by the model and satellite data can be carried out and the capability of the model to
be driven by satellite data can also be investigated. This alternative approach appears the most appropriate to
analyse and understand the underlying complexities in the interactions between vegetation
parameters/processes and radiation.
The objective of this study is twofold. Firstly, an ecosystem process model for sahelian herbaceous vegetation is
developped. This model is defined to be used at a regional scale in order to simulate the temporal evolution of
the main parameters and processes associated with vegetation fiinctionning in arid countries. Secondly, the
coupling with satellite derived parameters is examined. The present paper is structured as follows : in the first
section, the ecosystem process model is described and its validation using biomass measurements acquired
during the period 1976-1992 in two different regions of the Sahel, is shown. The second section presents the
coupling of the ecosystem model with a simple reflectance model. A comparison with NOAA/AVHRR data is
presented. The third section deals with the perspectives and future evolution of the model.
ECOSYSTEM MODEL DEFINITION AND VALIDATION
As a key component of the overall approach, the grassland model has to satisfy a certain number of
requirements :
1) it must give a realistic description of the relevant processes, particularly net photosynthesis, water fluxes
and primary productivity at a scale compatible with satellite observations i.e. at a regional scale ;
therefore, it has to be validated with appropriate field measurements at the scale considered.
2) it must be able to simulate the temporal evolution of the main structural parameters like LAI, vegetation
cover fraction,...that may be directly incorporated into physical models of reflectivity in the visible and
microwave domains.
3) it must have a relatively simple formulation in order to incorporate vegetation parameters that may be
retrieved from satellite measurements. The use of data from different parts of the electromagnetic spectrum
should provide access to a larger number of biophysical parameters.
4) it must be driven by standard meteorological data obtained from synoptic network and site specific data
that can be readily obtained.
5) finally, it must be open with a certain capability of generalization. Future improvements should also be
easily incorporated in the model.
Following these requirements, an ecosystem model for sahelian grasslands, named STEP (Sahelian
Transpiration - Evaporation and Productivity model), was developped (Mougin et al, 1993). This model is
based on an existing model originally developped for arid and semi-arid rangelands (Rambal, 1980 ; Rambal
and Comet, 1982) and was adapted to the specific case of regional application in a sahelian environment At
present, the modelling of the aboveground primary production only deals with the herbaceous layer, the major
component of the sahelian vegetation, for which growth and mortality arc mostly controlled by water
availability. The processes considered relevant in the soil-plant-atmosphere system are modeled to simulate the
aboveground vegetation growth and soil water dynamics. In the present version, the following processes are
modeled : water fluxes in the soil, evaporation from bare soil, transpiration of the vegetation, photosynthesis,
respiration, senescence, litter production and litter decomposition at the soil surface. Moreover, structural
