A test of the potential accuracy of the water-budget approach to estimating evapotranspiration

Abstract

The water-budget approach to estimating evapotranspiration was tested against energy-budget calculations to determine its potential accuracy. Measurements of soil moisture change were made using both neutron attenuation and gravimetric techniques for one-day and four-day periods. Hydraulic head gradients were measured routinely in the field and moisture characteristics and capillary conductivities were calculated from both field and laboratory data. The energy-budget calculation of evapotranspiration was derived from measurements of net radiation, soil heat flow and Bowen ratios. The measurements were made in a corn field and the experimental period spanned 25 days in July, 1969. An error analysis of the water-budget approach and the experimental results lead to the same conclusion. The water-budget cannot accurately estimate evapotranspiration on a daily basis due to unavoidable large potential errors. Accuracy arises with an increasing time span between soil moisture measurements. If actual evapotranspiration is high, there is no precipitation and six sites are averaged a reasonable accuracy can be achieved for four-day periods. If only one measurement site is used or if evapotranspiration is low or if precipitation occurs, the necessary time lapse between measurements increases substantially. Moisture characteristics and capillary conductivities varied greatly from site to site and from soil horizon to soil horizon. For the entire measurement period the water lost by deep seepage was less than 10% of that lost by evapotranspiration.