Vapor pressure deficit is used in iLand as proxy for atmospheric humidity. It serves as the mediator between atmospheric conditions and transpiration via stomatal conductance, and acts as a direct modifier on primary production. We follow Landsberg and Waring (1997) in applying the same response function for both production and stomatal conductance (Eq. 1)

\[\begin{aligned} f_{D,d}=e^{-k\cdot D} \end{aligned} \]

Eq. 1

with d a index for day of the year, k an empirical constant, and D the atmospheric vapor pressure deficit in kPa. The model is sensitive to this modifier, both with regard to water cycle feedbacks and direct effects on NPP. It is thus noteworthy that Mäkelä et al. (2008), conducting a empirical parametrization of Eq. 1 over sites in Europe, obtained strongly divergent parameters for k compared to Landsberg and Waring (1997). A possible explanation lies in differences in temporal resolution, with daily (Mäkelä et al. 2008) vs. monthly (Landsberg and Waring 1997) responses.

In iLand Eq. 1 is calculated on a daily basis. It is used as daily modifier on stomatal conductance in the water cycle simulations and averaged to obtain the f_vpd modifier for the monthly calculation of utilizable radiation and production.