Power-Law Scaling of Turbulence Cospectra for the Stably Stratified Atmospheric Boundary Layer

Abstract

Surface turbulent fluxes provide a key boundary condition for the prediction of weather, hydrology, and atmospheric carbon dioxide. The turbulence cospectrum is assumed to typically follow a −7/3 power-law scaling, which is used for the high-frequency spectral correction of eddy-covariance data. The derivation of this scaling is mostly grounded on dimensional analysis. The dimensional analysis or cospectral budget analyses, however, can lead to alternative cospectral scaling. Here we examine the shape of turbulence cospectra at high Reynolds number and high wavenumbers based on extensive field measurements of wind velocity and temperature in various stably stratified atmospheric conditions. We show that the cospectral scaling deviates from the −7/3 scaling at high wavenumbers in the inertial subrange of the stable atmospheric boundary layer, and appears to follow a −2 power-law scaling. We suggest that −2 power-law scaling is a better alternative for cospectral corrections for eddy-covariance measurements of the stable boundary layer.