A Relaxed Eddy Accumulator for Surface Flux Measurements on Ground-Based Platforms and Aboard Research Vessels

The relaxed eddy accumulation (REA) method has become a standard tool for scalar flux measurements near the surface during field experiments. With this technique, turbulent fluxes of relatively slowly reacting species are measured using the difference in the mean concentration between air samples accumulated during updrafts and downdrafts. A prototype of a relaxed eddy accumulator adapted to air–sea flux measurements aboard a research vessel was developed at the Centre National de Recherches Meteorologiques (CNRM) and was implemented on board the research vessel LʹAtalante during the Programme Ocean Multidisciplinaire Meso-Echelle (POMME) experiment (March–April 2001). When this technique is used on a mobile platform, a major difficulty is the real-time calculation of the true vertical wind corrected for platform motions. The different stages of this computation are described in this paper. In addition, a ground-based version of the system was tested over a maize field during the Experience sur Sites pour Contraindre les Modèles de Pollution Atmospherique et de Transport dʹEmission (ESCOMPTE) experiment. With this simplified prototype in which the motion corrections were not needed, the effect of filtering the vertical velocity signal was evaluated and the accuracy of the gas analysis system was estimated. Water vapor, carbon dioxide (CO2), and nitrogen oxides flux over the maize field was measured by this technique. The validation of the system was carried out through comparisons between fluxes obtained by other methods (eddy covariance for CO2 and latent heat flux during ESCOMPTE, and bulk method for latent heat flux during POMME). Good correlations were observed between latent heat and CO2 fluxes measured by eddy covariance and REA techniques for the ESCOMPTE dataset. During the POMME experiment, the correlation between REA and the bulk method is reasonably good (r = 0.83) for latent heat flux. However, since the bulk method is not a direct method, this correlation is slightly lower than the comparisons during the ESCOMPTE experiment. Eventually, the scatter is of the same order of magnitude as that observed between various air–sea flux techniques (for instance, between eddy-covariance and bulk air–sea fluxes). This result leads the authors to consider that their system can be used aboard ships in the future to measure fluxes of trace species at the ocean–atmosphere interface, provided that the analyzer used is accurate enough to detect and measure the concentration difference between the two reservoirs.