Surface layer characteristics and SVAT modelling of a fetch-limited forest

The work in this thesis concerns data evaluation and soil-vegetation-atmospheretransfer (SVAT) modelling of measurements taken in and above a temperate beech forest canopy. The SVAT modelling framework used here has been a mixture of an electrical network analogy (single-layer model) and an analytical type of scheme where vertical gradients within the canopy are captured with continuous functions, which can be integrated to the ecosystem level. The beech forest site is fetch limited and influence on forest mast measurements from upwind farmland was investigated. The influence was studied (1) by analyzing ratios of friction velocity taken at different heights in the mast as a function of fetch length and (2) by studying the flux-profile relationship of wind speed and momentum flux, and temperature and sensible heat flux, respectively. The flux-profile relationships also yielded information on the nature of the roughness sublayer which is the lowermost part of the surface layer. The results indicated that the forest site is influenced by the upwind conditions via internal boundary layers. The forest flux-profile relationships were interpreted in terms of aerodynamic resistances. Additional measurements yielded information on the viscous sublayer resistance for heat. Total atmospheric resistances for sensible heat and momentum were approximately equal. Regarding the modelling of soil-vegetation-atmosphere interaction, the focus was on carbon dioxide exchange. A simple model was developed where leaf measurements taken at ambient conditions were used to construct a mean canopy light response curve, which was integrated to yield the carbon dioxide uptake into the tree crowns. In order to compare with the eddy-correlation measurements of net ecosystem exchange, models and measurements of the ecosystem respiration were included. Agreement between the eddy-correlation based estimate and the leaf-measurement based scheme was generally good. The radiation scheme for the integration (upscaling) does not include a distinction between direct and diffuse light. Together with a simple model for the atmospheric resistances, temperature measurements and water vapor fluxes were analyzed to give estimates of mean canopy resistance. The derived estimates were compared to a canopy scale version of the optimal stomatal conductance hypothesis, which regards water lost via transpiration as a cost when assimilating carbon. An approach which calculates canopy photosynthesis analytically is also presented. The work in this thesis represents an attempt to derive simple but accurate formulations for single-layer models of trace gas and heat exchange in forested areas.