USE OF STATISTICAL AND NEURAL-NETWORK TECHNIQUES TO DETECT HOW STOMATAL CONDUCTANCE RESPONDS TO CHANGES IN THE LOCAL ENVIRONMENT

The bulk stomatal conductance term, g(s), in the Penman-Monteith equat ion is modelled as a function of local environmental variables. The st omatal dependencies are frequently described as a product of individua l functions, f(j), each depending on only one variable X-j, where the functional forms have been estimated through laboratory experiments. O ther descriptions for g(s) are being developed (notably through photos ynthesis models) and so methods of comparing model performance are nee ded. A notion of the 'best possible fit' of g(s) to the X-j is require d, thereby providing a benchmark for any model. This paper introduces regression and neural network methods to analyze the stomatal conducta nce of pine forest, although the techniques are applicable to any vege tation type. In this paper the importance of a strong nonlinear depend ence of g(s) on the X-j is illustrated and further the frequently used 'Jarvis' type nonlinear functions, f(j), are shown to be nearly optim al. (C) 1997 Elsevier Science B.V.