Evaluation of the Maximum Evaporation and the Priestley-Taylor Models for Inland Waterbodies

Abstract

The Priestley-Taylor (PT) model is a classic model of potential evaporation of terrestrial and marine surfaces. It is now recognized that the Bowen ratio (β) implicit in the PT model is too sensitive to temperature. The model also requires the surface net radiation (R_n) as input even though R_n is not an independent external forcing. The maximum evaporation model (MEM) proposed by Yang and Roderick (2019, https://doi.org/10.1002/qj.3481) is a potential candidate for replacing the PT model. Past studies have evaluated the MEM for land ecosystems and for the global ocean. This study represents the first evaluation of the MEM for inland waterbodies. Results are based on eddy-covariance observation at a large lake and at a small fishpond. Although there were complex error structures and error compensation patterns among its intermediate variables, the MEM was able to reproduce the observed monthly (R² > 0.95) and interannual variability (R² > 0.78) in the lake latent heat flux. A key assumption of the MEM, that the incoming and outgoing longwave radiation is coupled, was a reasonable approximation at both the monthly and the annual time scale for the large lake and at the monthly time scale for the small fishpond. This assumption allows the MEM to treat R_n as an intermediate variable instead of a prescribed forcing. These results support the MEM as an alternative to the PT model at locations where R_n measurements are not available. In situations where R_n data is available, a revised PT model with reduced β temperature sensitivity is recommended.