Improvement of the representation of canopy architecture

Starting fromWOFOST-GT, a further version of the model (named WOFOST-GT2) was developed, aimed at explicitly considering the vertical canopy profile, via the implementation of a model for plant height (Confalonieri et al., 2011) coupled to a function for deriving the number of canopy layers from DVS (Equations (12) and (13)).

where the second equation is the floor function of

This dynamic simulation of the emission of canopy layers assumes that during early stages (DVS < 0.2) the canopy can be adequately described as composed by two layers; as long as the crop is growing, the number of photosynthetic layers increases according to a logistic function, with the maximum rate of emission of new layers set during tillering (DVS = 0.25-0.35). The emission rate of new layers decreases during stem elongation, and ends with the emission of the flag leaf (DVS = 0.9). Leaf senescence is then computed allocating dead LAI units starting from the lowest canopy layer until the dead LAI of the layer is equal to its total LAI. Then, this layer is considered no longer photosynthetically active, and dead LAI units start to be allocated to the layer above. Simulating leaf senescence with such a bottom-up dynamic is coherent with both the drivers for leaves senescence reproduced byWOFOST: the ageing of leaves (the bottom layers contains the first emitted leaves) and selfshading (the bottom layers are those which are shaded). The maximum number of canopy layers is set to 20 and is reached at anthesis. This value represents a compromise between the need to increase the vertical resolution to allow a fine description of leaf senescence dynamics and the need to limit the increase in the computational cost of the simulation, given the frequent adoption of WOFOST in projects requiring simulations against large-area databases. The adoption of two canopy layers at emergence is justified by (i) the pronounced worsening of the performance of the original WOFOST version run with a single canopy layer (data not shown) and (ii) by simulation experiments that revealed that the original WOFOST version markedly changes its behaviour only while increasing the number of canopy layers from one to two. Moreover, coupling the approach for dynamic emission of canopy layers and bottom-up leaves senescence with a model for estimating plant height allows to assign an explicit thickness to each canopy layer and to identify the height above which LAI is photosynthetically

active and below which it is represented by senescent tissues. This gives the opportunity to improve the simulation of micrometeorological aspects within the canopy (e.g., dead LAI units do not transpire, thus their temperature is higher), which, in turns, could lead to a more realistic simulation of biophysical processes involved with biotic (e.g., fungal pathogens) and abiotic stressors affecting the crop.

Created with the Personal Edition of HelpNDoc: Single source CHM, PDF, DOC and HTML Help creation