On the variability of the Priestley-Taylor coefficient over water bodies

Deviations in the Priestley-Taylor (PT) coefficient α_PT from its accepted 1.26 value are analyzed over large lakes, reservoirs, and wetlands where stomatal or soil controls are minimal or absent. The data sets feature wide variations in water body sizes and climatic conditions. Neither surface temperature nor sensible heat flux variations alone, which proved successful in characterizing α_PT variations over some crops, explain measured deviations in α_PT over water. It is shown that the relative transport efficiency of turbulent heat and water vapor is key to explaining variations in α_PT over water surfaces, thereby offering a new perspective over the concept of minimal advection or entrainment introduced by PT. Methods that allow the determination of α_PT based on low-frequency sampling (i.e., 0.1 Hz) are then developed and tested, which are usable with standard meteorological sensors that filter some but not all turbulent fluctuations. Using approximations to the Gram determinant inequality, the relative transport efficiency is derived as a function of the correlation coefficient between temperature and water vapor concentration fluctuations (R_Tq). The proposed approach reasonably explains the measured deviations from the conventional α_PT = 1.26 value even when R_Tq is determined from air temperature and water vapor concentration time series that are Gaussian-filtered and subsampled to a cutoff frequency of 0.1 Hz. Because over water bodies, R_Tq deviations from unity are often associated with advection and/or entrainment, linkages between α_PT and R_Tq offer both a diagnostic approach to assess their significance and a prognostic approach to correct the 1.26 value when using routine meteorological measurements of temperature and humidity.