Contemporary variations of terrestrial net primary production: The use of satellite data in the light of an extremal principle

Combining ecological models and satellite data is likely to be the most feasible approach for resolving a current problem of carbon cycle modelling: evaluation of global NPP response to changes in climate and atmospheric CO2. The main difficulties encountered in implementing this approach are as follows. First, the dimensionality (the number of the parameters required to specify) of the comprehensive models of NPP is greater than the dimensionality (the number of independent measurements) of the available satellite data. Second, remotely sensed (macroscale) variables and modelled (microscale) variables are not equivalent in an ecological sense. In this paper we propose two means of coping with the difficulties: to introduce some ecological extremal principles for lowering the dimensionality of the NPP models, and to use satellite data for evaluating biogeographical (e.g., vegetation period) rather than biophysical (e.g., leaf area index) parameters of the NPP models.