The Effect of Relative Humidity on Eddy Covariance Latent Heat Flux Measurements and its Implication for Partitioning into Transpiration and Evaporation

• DOI: 10.2139/ssrn.4106267
• Publication Date: 2022-05-28T12:00:57Z
• Authors: Weijie Zhang; Martin Jung; Mirco Migliavacca; Rafael Poyatos; Diego Miralles; Tarek S. El-Madany; Marta Galvagno; Arnaud Carrara; Nicola Arriga; Andreas Ibrom; Ivan Mammarella; Dario Papale; Jamie Cleverly; Michael J. Liddell; Georg Wohlfahrt; Christian Markwitz; Matthias Mauder; Eugenie Paul-Limoges; Marius Schmidt; Sebastian Wolf; Christian Brümmer; M. Altaf Arain; Silvano Fares; Tomomichi Kato; Jonas Ardö; Walter Oechel; Chad Hanson; Mika Korkiakoski; Sébastien Biraud; Rainer Steinbrecher; Dave Billesbach; Leonardo Montagnani; William Woodgate; Changliang Shao; Nuno Carvalhais; Markus Reichstein; Jacob A. Nelson
• Keywords: Civil and Environmental Engineering; Agriculture and Food Sciences; Environmental Engineering; 550; FLUXNET; Other Civil and Environmental Engineering; 0207 environmental engineering; Eddy covariance; 02 engineering and technology; Atmospheric Sciences; CARBON; Engineering; ENERGY-BALANCE CLOSURE; ATTENUATION; AGRICULTURAL MANAGEMENT; veterinary and food sciences; Meteorology & Atmospheric Sciences; EXCHANGE; Agricultural; info:eu-repo/classification/ddc/550; Bioresource and Agricultural Engineering; Agricultural and Veterinary Sciences; Evapotranspiration; ddc:550; Forestry; Energy balance closure; Biological Sciences; FOREST; Latent energy; Eddy; Biological sciences; Earth sciences; MODIS; WATER-VAPOR; covariance; Earth Sciences; EVAPOTRANSPIRATION ALGORITHM; SAPFLUXNET

While the eddy covariance (EC) technique is a well-established method for measuring water fluxes (i.e., evaporation or 'evapotranspiration’, ET), susceptible to many uncertainties. One such issue potential underestimation of ET when relative humidity (RH) high (>70%), due low-pass filtering with some EC systems. The influence these errors different types systems (e.g. open- closed-path sensors) has not been characterized synthesis datasets as widely used FLUXNET2015 dataset. Here, we assess RH-associated from 163 sites in We found that RH were most apparent during hours was approximately 70% higher. This predominantly observed at using systems, underestimations estimated 58% higher than 90%. To impact on and EC-based transpiration (T) estimates, corrected machine learning based empirical approach then partitioned into T E two data-driven methods. Results showed opposite (increasing vs decreasing T-to-ET ratios, T/ET) responses owing contrasting ways methods model WUE. Analysis estimates three correction approaches no clear improvement correlation independent sap flow T. Overall, our results demonstrate existence RH-related bias dataset suggest significant source uncertainty must be reckoned estimating ecosystem T/ET work further demonstrates need high-quality, flux measurements understand uncertainties estimates.