Uncertainty analysis of computational methods for deriving sensible heat flux values from scintillometer measurements

The use of scintillometers to determine sensible heat fluxes is now common in studies of land-atmosphere interactions. The main interest in these instruments is due to their ability to quantify energy distributions at the landscape scale, as they can calculate sensible heat flux values over long distances, in contrast to Eddy Covariance systems. However, scintillometer data do not provide a direct measure of sensible heat flux, but require additional data, such as the Bowen ratio (β), to provide flux values. The Bowen ratio can either be measured using Eddy Covariance systems or derived from the energy balance closure. In this work, specific requirements for estimating energy fluxes using a scintillometer were analyzed, as well as the accuracy of two flux calculation methods. We first focused on the classical method (used in standard softwares) and we analysed the impact of the Bowen ratio on flux value and uncertainty. For instance, an averaged Bowen ratio (β) of less than 1 proved to be a significant source of measurement uncertainty. An alternative method, called the “β-closure method”, for which the Bowen ratio measurement is not necessary, was also tested. In this case, it was observed that even for low β values, flux uncertainties were reduced and scintillometer data were well correlated with the Eddy Covariance results. Besides, both methods should tend to the same results, but the second one slightly underestimates H while β decreases (<5%).