Two vegetation dynamic models (VDMs) were coupled to a soil–vegetation–atmosphere transfer scheme aimed at simulating water, heat, and CO2 fluxes and to a three-dimensional unsaturated flow and heat diffusion model allowing the study of heterogeneous soils and vegetation. The VDMs differ in reproducing gross photosynthesis. One was adapted from a daily model, while the other is a more complex model, needing the continuous monitoring of CO2 concentration and a larger number of parameters, although it is useful for assessing feedback due to an increase in CO2 concentration. The two coupled models were validated on a half-hour time step through leaf area index (LAI), ground energy, and water flux measurements performed during the summer of 2007 and 2009 in an alfalfa (Medicago sativa L.) field in the Mediterranean (southern Italy) and during the spring of 2001 in a predominantly C3 grass covered field (California). Furthermore, the models were used at the field scale in numerical experiments for assessing the effects of model dimensionality on evapotranspiration, CO2 flux, and LAI in the presence of flat heterogeneous or sloping homogeneous terrains using both the fully three-dimensional soil water flow model and a simpler model allowing only vertical flow. For the flat terrain experiment, one-dimensional modeling generally provided significant differences in the simulated quantities with respect to the three-dimensional model, even though in some cases similar results were found. For the sloping terrain, one-dimensional modeling proved quite unsuitable for analyzing slope effects, while three-dimensional modeling allowed detailed description of different degrees of slope interactions between soil moisture dynamics and surface fluxes for different slope angles.