A study of the role of seven historically significant fast-response hygrometers and sensor calibration on eddy covariance H2O fluxes and surface energy balance closure

Abstract

Water vapor fluxes are a major component of the surface energy balance, hydrological cycle, and ecosystem processes. As newer sensors are developed for eddy covariance measurements, the role of different technologies and calibration uncertainties on H2O fluxes and surface energy balance closure should be assessed. Here, a field intercomparison is described testing historically significant fast-response hygrometers: Lyman-α, KH2O, LI-7000, LI-7500, LI-7200, EC155, and EC150. This study occurred in southeastern Wyoming in September 2015. Three methods of calibration are investigated: standard calibration with a dew point generator and zero-air gas, referencing to a first principles slow-response hygrometer, and piecewise calibration between fast-response hygrometers based on blocks of time with consistent relationships between 5-minute data. Differences in fluxes among the infrared gas analyzers were largest when using the standard calibration (maximum difference of 15.9%, standard deviations of differences between 4.1–6.0%) with differences reduced by 0.3–0.6 × when the calibrations were referenced to a first principles hygrometer. The smallest differences occurred using the piecewise calibration, which emphasizes the intrinsic differences among the sensor technologies, where differences were further reduced by 0.04–0.09 × (maximum difference of 8.2%, standard deviations of differences between 2.3–3.0%). The Lyman-α and KH2O experienced severe drifts, though their piecewise calibrations were generally within the range of the others. The two most influential factors beyond calibration uncertainty were the oxygen correction for the ultraviolet absorption sensors, especially the KH2O, and the tube attenuation spectral correction for closed-path analyzers, mostly with the lower flow EC155. The effect of hygrometers on the energy balance closure was minor with some uncertainty associated with the standard calibration, though this was less than the change in closure associated with the choice of sonic anemometer. These results focus attention on the need to further develop accurate calibration methodologies that are suitable for all flux tower sites.