Estimation of Crop Evapotranspiration by Means of the Penman-Monteith Equation

This study evaluated the potential of using simple, semi-empirical resistance models for the direct estimation of crop evapotranspiration as an alternative to the traditional approach involving a reference equation and a crop coefficient. It consisted of three major parts. The first part originally aimed at the development of simple expressions for the aerodynamic and surface resistance terms in the Penman-Monteith equation. This goal could not be achieved because of two reasons: First, the determination of aerodynamic resistance was rendered impossible because of problems with the measurements, and, secondly, the values of surface resistance, back-calculated from the Penman-Monteith equation, turned out to be very sensitive to the estimates of aerodynamic resistance. In the second part, two forms of the Penman-Monteith equation and one form of the Shuttleworth-Wallace model were compared to each other and to the traditional KimberlyPenman approach. The analysis showed that (1) it was possible to fit a simple form of the Penman-Monteith equation to measured data, (2) one form of the Penman-Monteith equation allowed a better fit than the KimberlyPenman approach and ( 3) the Shuttleworth-Wallace model provided a slightly better fit to the data than the Penman-Monteith equation. In the third part, the possibility of estimating net radiation and soil heat flux was investigated and a comparison was made between meteorological observations obtained at a grassed weather station and those obtained above an agricultural crop. The results indicated that (1) it was difficult to obtain an accurate estimate of net radiation or soil heat flux for a partial canopy and (2) major differences existed between measurements of vapor pressure deficit and wind speed obtained above an agricultural crop and at a grassed weather station. From the analysis, it is concluded that the benefit of using a semi-empirical form of the Penman-Monteith equation instead of the traditional approach is limited, in particular for the prediction of crop evapotranspiration from limited historical measurements executed at a grassed weather station. As an alternative, the use of an elaborate multi-layer model for the determination of more versatile crop coefficients is suggested.