Constraints on surface evapotranspiration: implications for modeling and observations

Abstract:

The continental hydrological cycle and especially evapotranspiration are constrained by additional factors such as the energy availability and the carbon cycle. As a results trying to understand and predict the surface hydrologic cycle in isolation might be highly unreliable. We present two examples were constraints induced by 1) radiation control through cloud albedo feedback and 2) carbon control on the surface water use efficiency are essential to correctly predict the seasonal hydrologic cycle.

In the first example we show that correctly modeling diurnal and seasonal convection and the associated cloud-albedo feedback (through land-atmosphere and convection-large-scale circulation feedbacks) is essential to correctly model the surface hydrologic cycle in the Amazon, and to correct biases observed in all general circulation models. This calls for improved modeling of convection to correctly predict the tropical continental hydrologic cycle.

In the second example we show that typical drought index based only on energy want water availability misses vegetation physiological and carbon feedback and cannot correctly represent the seasonal cycle of soil moisture stress. The typical Palmer Drought Stress Index is shown to be incapable of rejecting water stress in the future. This calls for new drought assessment metrics that may include vegetation and carbon feedback.