Energy and water flow through the soil-vegetation-atmosphere system: the fiction of measurements and the reality of models

Abstract

This paper summarizes basic concepts and definitions in models of the SVA system and then emphasizes inconsistencies between model variables and observations for the soil, vegetation and atmosphere elements. This is done first in a qualitative sense, then analytically for the observations of the radiometric temperature of vegetation canopies. Notwithstanding the significant efforts dedicated to systematic comparison of models and to performing complex experiments to construct a rich data base for model validation, the inter-model variability of predicted fluxes and state variables remains large and it is hard to pin-point specific causes. The argument developed in this paper is that a different avenue should be explored in search of a solution, namely an in-depth analysis of the nature of feasible observations of the SVA system, in order to detect and understand inconsistencies in model variables and parameterizations. Model equations define state variables and parameters rather precisely, while parameterizations are often established from experimental data, assuming that observed and model variables are consistent. The latter cannot be taken for granted, however. There are now tools of investigation that were not available in the early years of land-surface science. We are now able to construct detailed and realistic 3D models of elements of the SVA system, e.g. a soil-foliage system, and to model radiative and convective processes in such 3D systems. The latter gives us the opportunity of modeling observations of elements of the SVA system, such as the soil matrix or a vegetation canopy, in a rather realistic way. When dealing with highly heterogeneous systems, this capability provides ways and means to understand how the integral magnitudes we measure, such as soil electrical resistivity or radiance emitted by a canopy, relate to the object properties we seek to determine