Estimation of FAO Evapotranspiration Coefficients

Abstract

The estimation, by computers, of correction coefficients for the United Nations Food and Agriculture Organization (FAO) evapotranspiration computation method had required a ``table lookup'' process involving arrays of up to four dimensions. Since such arrays cannot be stored on many computer systems, an alternative approach to estimating the correction coefficients is desirable. This note presents one such approach which utilizes regression of climatic parameters to predict correction coefficients. Predictions obtained by this approach are compared to original coefficient values by use of a normalized error term. While problems have been experienced with certain intermediate values of the independent variables, the results obtained have been generally satisfactory.