Evaluation of Evapotranspiration from Eddy Covariance and Surface Layer Scintillometry.

Evapotranspiration (ET) is an important component in the water budget and used extensively in water resources management, such as water planning and irrigation scheduling. In semi-arid regions, irrigation is used to supplement limited and erratic growing season rainfall to meet crop water demand. Although lysimetry is considered the most accurate method for crop water use measurements, large precision weighing lysimeters are expensive to build and operate. Alternatively, other measurement systems such as eddy covariance (EC) and scintillometry are being used to estimate crop water use. However, due to numerous explicit and implicit assumptions in the EC method, an energy balance closure problem is widely acknowledged. Path integrating capabilities of scintillometers over several kilometers have the potential to bridge the gap between primary, more local observations (lysimeters, Bowen ratio, or eddy covariance) and the demand for large-scale spatially averaged surface heat fluxes. However, inherent errors associated with scintillation measurements by widely available visible and infrared scintillometers may propagate errors in estimating surface energy fluxes. Numerous studies have evaluated the measurement accuracy of scintillometers using eddy covariance systems; however, the energy balance closure problems with EC measurements may not allow for a thorough evaluation. In this study, an EC and a visible surface layer scintillometer (SLS) were deployed in a lysimeter field at the USDA-ARS-Conservation and Production Research Laboratory in Bushland, TX for the 2015 and 2016 growing seasons. Results indicated that the EC system achieved energy balance closure of 75%, which is similar to literature report values. The EC system and the SLS produced similar results for ET with RMSEs of approximately 60% and R² values of 0.81 and 0.75, respectively. Both instruments were found to have poor correlation for sensible heat flux, but good correlation with latent heat flux and ET. Additional analysis should provide indications of potential areas to reduce ET errors.