Towards an enhanced metric for detecting vertical flow decoupling in eddy covariance flux observations

Abstract

The eddy covariance (EC) technique has emerged as the method of choice for observing ecosystem–atmosphere interactions across biomes and climate zones. However, EC measurements are biased when the turbulent flow is decoupled from the underlying surface, severely limiting the applicability of the technique in observing surface–atmosphere fluxes. Friction velocity ($u_{\ast }$) is typically used to detect and filter these periods from EC flux time series. The processes that control decoupling are understood qualitatively, including the strength of vertical turbulent mixing, stable stratification and canopy drag. However, the standard practice utilising $u_{\ast }$ misses most of these processes, resulting in a significant uncertainty in detecting decoupling. Consequently, a quantitative metric, $\Omega$, which encapsulates all these processes in a unified framework, was recently proposed. However, it has not yet been systematically tested over a range of ecosystems and site characteristics. The objectives of this study were therefore to test the efficacy of $\Omega$ at a diverse range of EC sites, to quantify the processes controlling decoupling across sites, and to compare $\Omega$ against other decoupling metrics, such as $u_{\ast }$. A similar $\Omega$ threshold value for coupling was observed at all the 45 tested EC sites, with a value of 0.59 (median; 0.5...0.63 interquartile range). This indicates that the $\Omega$ metric captured the essential features of decoupling across sites, thereby enabling deeper analyzes of the causes of decoupling. For example, $\Omega$ indicates that (1) flows above dense forest canopies can be decoupled from the forest floor also during the daytime due to canopy drag and that (2) during stable stratification decoupling is more likely with tall towers. These findings significantly enhance our scientific understanding of the underlying causes of decoupling, will inform improved analyzes of EC data and support near-surface turbulence transport analyzes in open and forested landscapes.