Practical and Theoretical Benefits of an Alternative to the Penman-Monteith Evapotranspiration Equation

The Penman-Monteith equation is used widely to estimate evapotranspiration (E) and to understand its governing physics. I present an alternative to the Penman-Monteith equation that has both practical and theoretical advantages, at no appreciable cost. In particular, the new equation requires no additional assumptions, empiricism, or computational cost compared with the Penman-Monteith equation. Practically, the new equation is consistently more accurate over a wide range of conditions when compared with eddy covariance observations: The new equation has lower errors compared with Penman-Monteith estimates of ET at all of the 79 eddy covariance sites available for the analysis. Using the new equation reduces errors, on average, by 67%, from 8.55 to 2.81 [W m(-2)]. At night, the improvement is even greater (92% reduction in error; from 1.26 to 0.097 [W m(-2)]). This improvement is achieved without calibration. Theoretically, the new equation corrects a conceptual error in the Penman-Monteith equation, in which the Penman-Monteith equation incorrectly implies that E from a saturated surface into a saturated, turbulent atmosphere (equilibrium E) is exactly equivalent to E from an unsaturated surface into an unsaturated, laminar atmosphere. The conceptual error is traced back to the failure of the Penman-Monteith equation in important limiting cases; these errors are eliminated by the new equation. I use the new equation to revise an existing theory of land-atmosphere coupling affected by the conceptual error in the Penman-Monteith equation and to reassess several common but incorrect definitions of equilibrium E.