Simulating energy and carbon fluxes over winter wheat using coupled land surface and terrestrial ecosystem models

Coupled land surface and terrestrial ecosystem models are used to simulate energy and carbon fluxes over winter wheat at the Ponca City, Oklahoma, Ameriflux site. The terrestrial ecosystem model consists of photosynthesis and respiration (autotrophic and heterotrophic) sub-modules, which uses canopy temperature, soil moisture, and soil temperature simulated by the land surface scheme. The photosynthesis sub-module, which provides an estimate of canopy conductance to the land surface scheme, is designed to use both the big- and the two-leaf (sunlit/shaded) approaches, and canopy conductance formulations based on both relative humidity and vapor pressure deficit. This provides a tool to test the sensitivity of model results to these two different approaches of modeling photosynthesis and stomatal conductance. Model results for carbon and energy fluxes compare well with observations over the growing season of 1997, especially with the use of the two-leaf model and stomatal conductance formulation based on vapor pressure deficit. Averaged over the growing season, the model results suggest that for this particular site, the difference in simulated latent heat fluxes caused by the choice of the photosynthesis approach (big- or two-leaf) is smaller (∼1%) than the difference caused by the choice of the stomatal conductance formulation (∼10%). In regard to the carbon fluxes, averaged over the growing season and compared with the two-leaf model, the simulated net photosynthesis rate and net ecosystem exchange flux are about 5 and 18% higher, respectively, for the big-leaf model. It is shown that comparisons with both observed energy and carbon fluxes are necessary to constrain model behavior and test its performance adequately.