Surrounding el< 
Projection 
on an hemisphere 
on a plane 
1 reference plane 
Figure 2: The fisheye method is a double projection that allows to integrate easily the solid angles involved, 
paying attention to overlaying. 
22 The crop model AFRCWHEAT2 
The original model AFRCWHEAT for winter wheat is described in Weir et al. (1984) for the explanation of all 
the processes related to the phonological development, the li ght interception, die photosynthesis, the respiration, 
the dry-matter partitioning, the root growth and the grain growth. This model is distinguishable from other crop 
production models by a complete and discrete description of the development of the canopy, \fery often, the 
canopy is just represented by a “Big-Leaf’, characterized by the value of the LAI, and eventually, die cover 
¿action. When the photosynthesis is computed at various levels within the canopy, a parameter describing the 
rate of attenuation of the radiation wi thin the cover is used to allow the computation of the intercepted 
photosynthetically active radiation at each level. This is the case of the SUCROS crop models (Spiders et al., 
1989). Of course, when studying the radiative transfer within the canopy, it appears that this parameter is related 
to the feature of the distribution of the leaves angles, and more completely, to the architecture and the structure of 
the canopy. But most of the crop production models can simulate growth and give correct yield estimations 
without any description of the architecture of the canopy. 
On the opposite, in AFRCWHEAT, the stems and leaves are countable. Simulation of emergence, 
growth and Miy OTno-. is done for individual leaves, in a quantitative way: lentgh and width are computed, and 
the maximum dimension of each kind of leaf is known. The model explains the production of the tillers, followed 
by their death or their survival. Details of the simulati on of tiller and leaf growth are given in Porter (1984). The 
first version of the model was usable for a crop without water or nutrients limitations, which is generally the case 
fa most of the winter wheat fields in the temperate regions, at least in Europe. However, the new version 
AFRCWHEAT2 (Porter, 1993) now incorporates responses to water and nitrogen, taking into account the soil 
characteristics. An important improvment in the AFRCWHEAT2 version is the stronger linking than before 
between the canopy development and the photosynthesis assimilates sent to leaves. The canopy development 
submodel computes the potential daily increase in the leaf area resulting from the growth of the active leaves. 
The photosynthesis «nd partitionning submodels compute the «ggimiUtad dry matter potentially sent to the 
leaves. Interaction is done in order to reduce the increase in the LAI if the dry matter sent to the leaves is limited. 
The daily meteorological records required to run the model are solar radiation, minimum and maximum 
temperature, dry bulb and wet bulb temperature at 09:00 GMT, precipitation and wind. 
Apart from the satisfaction to be able to reproduce the architecture of the canopy, the interest to 
use such a rather complex description in a crop model for which the main objective is to estimate correctly the 
yield, and not the number and the area of the leaves, is real and must be detailed a little bit The crop production 
is linked to the grain weights and to the number of grains per square meter. The number of grains per square 
meter is related to the number of shoots able to carry ears: main shoots, plus the surviving tillers. To reproduce 
the population behavior, three initial parameters due to the farming practices are very important: the day of 
sowing, the sowing density, and the chosen variety. The variety is essentially characterized by its sensitivity to 
the photoperiod. The well managed fields reach generally a number of shoots at anthesis which could be guessed 
at die first order more simply However, depending on the values of the three parameters discussed above, the 
behavior of the canopy development in the first phenological stages, in terms of shoot population time profile and