Comment on “An Alternative Approach for CO2 Flux Correction Caused by Heat and Water Vapour Transfer” by Liu

The need and methodology to correct open-path eddy covariance estimates and determine surface exchange were documented in a landmark article a quarter of a century ago (Webb et al., 1980). Interest in these ‘Webb corrections’ is increasing, as seen in both their widespread application and recent articles re-examining their use (Paw U et al., 2000; Fuehrer and Friehe, 2002; Liebethal and Foken, 2003). Indeed, hundreds of eddy covariance tower sites now form a global network (Baldocchi et al., 2001), many employing open-path sensors to measure CO2 and water vapour densities. Regrettably, the recent article of Liu (2005) represents a flawed interpretation, of both the basic equation of state and the Webb corrections, whose publication obscures the open-path correction issue. The essential first step in deriving the Webb corrections is to define an equation of state for turbulent perturbations. This can take various forms including those shown by Equation (9b) of Webb et al., (1980), and Equations (4) and (6) of Liu (2005). To demonstrate clearly Liu’s error in interpretation, the perturbation equation of state is reproduced here in the form most relevant to the correction of open-path eddy covariance. Open-path gas analysers detect extinction targets in a constant sensing volume, and thus respond to molar density (n). Partitioning the molar density of air (n= nv + na) yields components representing water vapour (nv) and all other gases (‘dry air’, na) including trace amounts of CO2 (nc). In molar density terms, Equation (4) of Liu (2005) may be simplified to an equally valid expression of the perturbation equation of state