On Scalar Transport in Plant Canopies

Abstract

We examine the notion that scalars are transported along their mean concentration gradients in the air space of the canopy. Recent observations and theory indicate that this concept is both inappropriate and misleading. Independent measurements of the fluxes and gradients of heat, water vapour and CO2 in a forest canopy show that counter-gradient fluxes are common. The intermittency of the transport processes and their large scale are seen as important reasons for this. Eddy-correlation and/or ecological techniques seem to be the only viable alternatives for measuring flux densities and source-sink strengths at present, but the logistical problems are formidable. For modelling exchange processes at leaf surfaces, hence source-sink distributions, analyses based on the gradient-diffusion concept may not be too much in error in as much as they employ essentially correct descriptions of transfer across leaf boundary-layers, if not in the canopy air space. An empirical description of transport in the latter may suffice. The utility of alternative models of scalar transport based on the nature of canopy turbulence is examined. Second-order closure models appear to have great pedagogic value in identifying the existence and relative importance of mechanisms for the production, transport and dissipation of scalar fluxes, but they are of limited use for prediction. Lagrangian models, though, appear to predict dispersion and profile development very well, provided the source distribution is known. However, the inverse problem of inferring source distributions from the concentration profiles remains a challenge.