Realistic representation of stomatal conductance (gst) is a key component of land surface modeling for simulation of evapotranspiration as well as dry deposition. It is well known, through laboratory and field experiments, that stomatal conductance decreases with exposure to dryer air. However, there is, as yet, no strong mechanistic explanation for this relationship. Therefore, there is considerable disagreement on how to best represent the stomatal dependence on humidity. Until recently, most stomatal parameterizations were simple empirical functions (usually linear) of vapor pressure deficit (D). Ideally D should be the difference between the air immediately outside the leaf surface and the air inside the leaf. However, the air humidity at the leaf surface is difficult to measure or to model. Therefore, the ambient air humidity above the canopy is usually used along with the saturation humidity at the leaf temperature. Clearly, this form of D is somewhat removed from the direct experience of the leaf and, therefore, not the most relevant parameter. More recently, a new generation of models compute stomatal conductance based on rates of photosynthesis. Many of these models have stomatal conductance a linear function of relative humidity at the surface of the leaf (RHs) which typically requires an iterative solution. The reason that these models tend to use a measure of humidity at the leaf surface while the earlier models mostly use the ambient humidity above the canopy seems to be that the photosynthesis models require numerical iterative solutions anyway to solve for CO2 concentration at the leaf surface. However, for applications in complex Eulerian grid models, particularly for mesoscale meteorology models where CO2 variations are not important, it is often desirable to avoid iterative solutions.
For application to mesoscale air quality issues we have developed a land surface model coupled to the Penn State/NCAR Mesoscale Meteorology model (MM5) and the Community Model for Air Quality (CMAQ) for both evapotranspiration and chemical dry deposition. The land surface model is mostly based on the earlier empirical schemes but with the linear dependence on leaf surface relative humidity, characteristic of the newer photosynthesis based models. With a small approximation (that stomatal conductance goes to zero as RHs goes to zero) RHs is solved for, analytically. The solution involves ambient humidity, leaf temperature, aerodynamic and laminar boundary layer conductances, and stomatal conductance without closure due to humidity. Therefore, stomatal response to humidity becomes a complex function of many environmental quantities. The aim of this study is to investigate the behavior of this model through numerical experimentation in comparison to laboratory data. For example, the model should respond realistically to relative humidity changes due to either changes in ambient absolute humidity or changes in ambient temperatures. Also, recent studies suggest that stomata respond more directly to rates of evapotranspiration (E) rather than humidity. Therefore, it is important to investigate relationships between gst, D, and E under a variety of conditions.