PROBABILISTIC RECONSTRUCTION OF A PEACH TREE CANOPY AND SIMULATION OF ITS PHOTOSYNTHETIC ACTIVITY

In the present study the probabilistic geometrical reconstruction of an adult peach tree crown (28.41 m² surface area; 14,260 leaves) was simulated from a leaf data sample identifying less then 1/20 of canopy leaves (i.e. less than 600 elements). More precisely, the data extraction was performed by applying a Monte Carlo method, whereas the foliage synthesis was accomplished by a non-parametric procedure, based on the adaptive Kernel method used in spatial statistics.
A bootstrap technique was used to generate (sub-)samples from the collected data making it possible to match the possibility to sample the canopy only once with the requirement of having a sufficient number of reconstruction to perform a statistical analysis.
The artificial tree crowns were evaluated from both structural and functional perspectives by comparing artificial geometrical traits and canopy carbon gain against the corresponding values computed on the real tree.
In general, a satisfactory agreement of all parameters was shown: the adaptive Kernel reconstruction procedure appears therefore a promising tool to be exploited in eco-physiological studies aimed to explain short- and long-term tree canopy responses.

VI International Peach Symposium

R. Giuliani, E. Muzzi, H. Sinoquet, F. Nerozzi, T.M. DeJong

adaptive Kernel, bootstrap, Monte Carlo, Prunus persica, intercepted radiation flux, gas exchange

https://doi.org/10.17660/ActaHortic.2006.713.28