Sensitivity of evapotranspiration in a wheat field, a forest, and a grassland to changes in climate and direct effects of carbon dioxide

Micrometeorological and physiological measurements were used to develop Penman-Monteith models of evapotranspiration for a wheat field in eastern Nebraska, a forest in Tennessee, and a grassland in east-central Kansas. The model fit the measurements well over the periods of observation. Model sensitivities to changes in climatic and physiological parameters were then analyzed. The range of changes considered was established from recent general circulation model output and from review of recent plant physiological research. Finally, climate change scenarios produced by general circulation models for the locations and seasons matching the observed data were applied to the micrometeorological models. Simulation studies show that when all climatic and plant factors are considered, evapotranspiration estimates can differ greatly from those that consider only temperature. Depending on ecosystem and on climate and plant input used, evapotranspiration can differ from the control (no climate or plant change) by about -20 to +40%.