Accuracy of the Bowen ratio-energy balance method for measuring latent heat flux in a semiarid advective environment

The Bowen Ratio-Energy Balance (BREB) is an accurate method often used to measure the latent heat flux (lambda E) due to its simplicity and portability. However, its performance in advective areas is less clear and its accuracy may depend on the equality of eddy transfer coefficients for heat and water vapor. In this work, hourly measured lambda E of a reference crop (Festuca arundinacea Schreb.) using a BREB system was compared with lysimeter-measured lambda E under moderate to severe advective conditions. The lysimeter resolution for hourly records was 22.6 W m(-2). The analysis of the eddy transfer coefficients was made using simultaneous measurements of fluxes and vertical gradients of temperature and humidity. To avoid computational problems when beta -> -1, some hourly periods were discarded in the analysis. Rejected data amounted to 37% of the total, although the cumulative evapotranspiration (ET) during these hours did not exceed 13% of the total ET. The BREB method overestimated daily ET by an average of 5.5% and by 5.7% when only daylight hours were considered. Under stable atmospheric conditions the method was less accurate, with relative errors of 21% vs. 11% under unstable conditions. For daylight hours, accuracy was higher under unstable conditions (RMSE = 36.15 W m(-2)) than under stable conditions (RMSE = 50.20 W m(-2)), which had larger overestimations of ET (6.3 vs. 5.1%). The main source of error appears to come from insufficient fetch resulting in local advective conditions. Nevertheless, and from a purely practical perspective, under the advective conditions of these measurements the BREB technique provides accurate ET fluxes with limited errors.