Technical Abstract: Evapotranspiration (ET) or latent heat flux (LE) can be spatially estimated as an energy balance (EB) residual for land surfaces using remote sensing inputs. The EB equation requires the estimation of net radiation (Rn), soil heat flux (G), and sensible heat flux (H). Rn and G can be estimated with an acceptable accuracy. In computing H, radiometric surface temperature (Ts) is often used instead of surface aerodynamic temperature (To), as To is neither measured nor easily estimated. This may cause an underestimation of ET because H will be overestimated as Ts is typically larger than To for unstable atmospheric conditions. The objectives of this study were to (1) model To to improve the estimation of H and consequently ET for advective environments in the semi-arid Texas High Plains, and (2) assess the accuracy of the To model using three different methods (aerodynamic profile, lysimeter, and eddy covariance). A 6.5 m tower platform was used to measure profiles of wind speed, air temperature, and relative humidity in and above cotton canopy near a large weighing lysimeter managed under rainfed conditions at the USDA-ARS Conservation and Production Research Laboratory, Bushland, Texas. The To was modeled using H as a residual from the EB at the lysimeter location. Results indicated that To was better modeled as a linear function of Ts, air temperature, and surface aerodynamic resistance. Modeled To showed a very small estimation error (0.1% mean bias error and 3.8% root mean square error) when compared to To values measured using the aerodynamic profile data. Even though excellent results were found in this study, the model is only valid for dryland cotton with a leaf area index ranging from 0.2 to 1.3 m**2 m**-2. Furthermore, more research is needed to expand the To model to cover cotton grown under irrigated conditions and showing larger crop percent cover and leaf area index values, and under different environmental and atmospheric conditions.