On The Use Of Alternative Surface Atmospheric State VariablesFor Nighttime Flux Corrections

Eddy-covariance based observations of carbon fluxes across the globe have been shown to underestimate CO2 fluxes under stable atmospheric conditions. Net ecosystem exchange overestimations have been attributed to the selective systematic error caused by this problem, because stable conditions are frequently nocturnal, at a time that coincides with autotrophic respiration. The most general approach to correct observations during stable conditions is based on relationships between carbon or other fluxes and surface friction velocity (). This velocity scale variable represents the square root of the total vertical flux of horizontal momentum near the surface, which in turn is a function of the canopy density and roughness. Thus, under stable conditions, when near ground fluxes are decoupled from above canopy fluxes, we argue that might not be the most optimal criteria to correct nighttime fluxes. Alternative criteria to overcome the aforementioned issues were tested. We analyzed the use of the standard deviation of the vertical velocity fluctuations, σw, and a velocity scale relationship with the turbulent kinetic energy, , as surface atmospheric state variables to filter eddy covariance data under stable conditions. We found results that are similar to those reported by Acevedo et al. (2009) and Wharton et al. (2009), which confirm that in several cases the use of alternative surface atmospheric state variables for nighttime flux correction might be a more robust method in comparison to the traditional approach, and is theoretically more defensible.