Correlation and structure function analysis in the quality assessments of long-term eddy-covariance measurements over a deciduous forest

Quantifying the quality of long-term continuous eddy-covariance measurements of carbon and water fluxes is critical to properly derive annual net ecosystem production (NEP). Key theoretical assumptions of the eddy-covariance technique are the stationarity of well-developed turbulence and horizontal homogeneity. In this study, lagged correlation analysis is used to calculate turbulence integral time scales, a useful quantity in determining the length of time periods over which stationarity tests should be performed. It is also used with the structure function to identify periods of weak turbulence and non-stationarity. Differences of such quantities at two levels above two forests and its implication of horizontal homogeneity are discussed. The correlation and structure functions and integral scales of different scalar fluxes (carbon dioxide, water vapor and sensible heat) are also examined. Flow conditions under which lagged correlation function is sensitive to coordinate rotation are discussed. Results of the above analysis of annual eddy-covariance measurements over a deciduous forest and their implication in the estimate of NEP is presented.