A Functional Approach to Vertical Turbulent Transport of Scalars in the Atmospheric Surface Layer

Eddy covariance measurements of surface/atmosphere scalar transport are based on the Reynolds decomposed, time averaged scalar transport equation, assuming stationarity and horizontal homogeneity. The eddy covariance method must strike a balance between the desire for flux measurements with improved temporal resolution and the loss of information because of non-stationarity/low frequency fluxes. Here we dispense with time as the primary independent variable and present a method, functional covariance, in which flux-appropriate independent variables are used to determine averaging of the decomposed transport equation. The 30 minute fluxes resulting from this method show magnitude increases in line with low frequency contributions. This method has the additional benefits of direct estimates of stochastic uncertainty, improved functional relationships of fluxes with driving variables, and the potential to negate the time rate of change term. We also discuss the additional data requirements and potential pitfalls of this approach.