Investigation of the thermodynamic component of Penman's method for estimating evaporation

The study presented here was initially commissioned by the Sydney Catchment Authority (SCA) in 2004, in response to the independent recommendations for the improvement of evaporation estimates being used in SCA reservoir water balance models. A study by Sinclair Knight Merz (SKM) independently reviewed the WATHNET model used by the SCA, and suggested improvement of the basic “Pan Conversion” evaporation estimates used in the model. Best estimates suggest that evaporation can cause as much as 14 billion litres per month (14Ml/month) to be lost from water supply reservoirs in Sydney in the worst of conditions. This accounts for as much as 20% for worst months of all water lost from the main water supply reservoirs in Sydney, such as Warragamba Dam (Lake Burragorang) and Woronora Dam. These estimates are based on observations of historical Pan Evaporation data. Evaporation is the largest loss component of water balance in open water reservoirs. However, it is difficult to estimate evaporation with any reasonable certainty. The Penman Method is universally acknowledged as one of the best physically based evaporation estimation methods to date (Penman, 1948, 1956). Use of this method on a continual basis is a difficult task due to limited data availability of a fairly large number of variables. These variables are a crude representation of the thermodynamic and aerodynamic properties of an open water surface. Focusing on the thermodynamic component of the Penman Method, the study sought to reduce errors associated with the estimation of solar energy flux into a water surface, particularly in Sydney. In most applications of the Penman Method, a series of semi-physical and semi-empirical equations are used to derive net energy flux based on a solar radiation constant, air and water temperature readings, and a cloudiness coefficient, as the input variables. This paper presents modifications to the Penman Method which make it suitable for use with high quality solar radiation measurements. These modifications involved exclusion of some empirical equations reported to have errors of up to ±20%, and resulted in a statistical improvement of general evaporation estimation. The study also produces a set of solar radiation constants calibrated based on the results of the best evaporation estimates. These constants are useful substitutes for the frequent gaps in measured solar radiation data, and allow consistency for long periods of estimation. Surprisingly, it was discovered that the semi-empirical approximations of solar radiation terms cause an average annual error of only 23mm, compared to the estimates derived from real measured solar radiation data. This level of improvement may be useful where accuracy of evaporation estimates is an absolute priority, or where vast water surfaces mean that evaporation rate equates to the loss of a very large volume of water. The findings can also benefit cases where the resolution of evaporation estimates is critical, by improving accuracy for shorter time steps. Two conclusions were drawn from this study which can benefit the use and understanding of the Penman Method in general. The first was that the semi-empirical equations suggested by some texts such as Thompson (1999), which are commonly used to approximate many of the inputs into the Penman Method, though seemingly far fetched and based on very limited data, do not result in any acute errors in the annual total evaporation estimate. The second conclusion drawn from this study was that the Pan Conversion Method of estimating evaporation has a strong tendency to over estimate evaporation by as much as 7% if compared to the best Penman Method estimates made in this study.