Evaluation of a Surface/Vegetation Parameterization Using Satellite Measurements of Surface Temperature

Abstract

This paper compares surface sensible heat flux and soil moisture values derived by inverting two boundary layers models with a surface/vegetation formulation, using surface temperature measurements made from NOAA-7 satellite (the AVHRR) with measured values for a wheat-growing area of the Beauce in France. The vegetation parameterization enables the models to reproduce the dramatic increase in surface sensible heat flux and decrease in soil moisture which occurred over a 5-day period during the field experiment. A bare soil model proved incapable of capturing the increase of the sensible heat flux during the 5-day period even though it yielded similar values of root-zone moisture. The vegetation model responds sensitively to small changes in canopy temperature by producing large changes in surface sensible heat flux due to the parameterization of the foliage resistance and the fact that the foliage is considered a layer of zero thermal inertia. Both the vegetation and bare soil models showed a continuous moisture decrease to values near or below the wilting point in the upper part of the root zone. The sensitivity of the results to errors in the initial sounding values or measured surface temperature were tested by varying the initial sounding temperature, dewpoint and windspeed, and the measured surface temperature by amounts corresponding to typical measurement error. Accordingly, we found that an unlucky combination of such errors can totally mask even large variations in surface heat flux from day to day, such as was measured during the field experiment. The vegetation component, therefore, is apparently more sensitive to error than the bare soil model.