Uncertainty Analysis of Evapotranspiration Estimates in Ecosystems

Abstract

Many hydrologic models and agricultural management applications require evapotranspiration estimates. The intensity of evapotranspiration is mainly determined by mathematical models rather than by direct measurement. In addition to its own estimate of evapotranspiration it is necessary to determine the uncertainty of this estimate. This uncertainty is not usually mentioned. In this paper these formulas are derived for the uncertainty estimate of evapotranspiration under simplifying assumptions. These assumptions enabled one to derive an expression of evapotranspiration estimation uncertainty suitable for practical applications. The paper focuses on both the absolute and the relative uncertainty of evapotranspiration estimation. The derived formulas can be used for determining the uncertainty in evapotranspiration estimation, but as well as for the accuracy estimate which is necessary for the measuring of input variables. The derived relationship shows that the net radiation should be more accurately measured than the other energy fluxes that have an influence on evapotranspiration. It follows that the relative uncertainty of evapotranspiration is primarily influenced by the relative uncertainty of net radiation. The uncertainty in the measurement of net radiation was derived from data obtained by using a radiometer which was equipped with a pair of pyranometers and with a pair of pyrgeometers. Planck’s Law was used for spectral analysis. The possible presence of systematic errors in the measuring of net radiation was evaluated for its potential impact on the errors of the evapotranspiration estimate. This paper is accompanied by measurement records and graphs documenting the achieved results.