Physically-based atmospheric models of evapotranspiration (ET) that consider the Penman combination energy balance and aerodynamic approach have achieved acceptance as useful tools for obtaining estimates of actual ET from land surfaces. These models have been made applicable to the case of non-saturated conditions through either surface resistance formulations (e.g., Penman-Monteith) or by application of the complementary evaporation theory of feedback between the atmosphere and surface moisture states (e.g., Granger-Gray). Their application becomes complicated under conditions of drought, when extremely low soil moisture availability severely restricts ET from the soil and plants. Under such severe conditions, consideration for the surface water balance and interactions with the balance of available energy and aerodynamic principles are Important for accurately estimating actual ET A modelling application is demonstrated using the Cold Regions Hydrological Model (CRHM) platform to examine the estimation of ET under drought conditions. CRHM allows users to assemble hydrological models by linking a suite of modular physically-based algorithms that describe the individual processes. In this case, the models assembled consider infiltration, evaporation, and sod moisture accounting and are applied to a mixed prairie located at Lethbridge, Alberta, Canada under drought conditions during the growing period in 2000 and 2001. Near surface meteorological and ecological observations used as model input and for evaluating model performance were obtained through the Ameriflux network and the Agriculture and Agri-Food Canada (AFC) Lethbridge Research Centre Results show that consideration for the effective rooting zone depth of the mixed-prairie at the site is important for estimating actual ET using the Penman-Montieth and Granger-Gray models during severe moisture stress. Relative differences in ET estimates provided by the models are discussed in the context of their contrasting theoretical approaches.
