Abstract
The bulk formulation for surface fluxes based on surface‐layer similarity theory is found to be a poor approximation for weak‐wind stably‐stratified conditions. These conditions are often non‐stationary, with very weak turbulence, and have often been excluded from analysis. As a consequence of such exclusion and the failure to account for self‐correlation in the bulk aerodynamic formula for the surface stress, the validity of the similarity theory has been overestimated for weak‐wind stable conditions.
Four months of turbulence data from FLOSSII are analysed, in order to generalize the bulk formula to include the influence of the non‐stationary mesoscale motions on the generation of turbulence. The generalization of the bulk formula leads to a more systematic variation of the drag coefficient. Excluding cases of significant non‐stationarity of the wind field also leads to a more systematic behaviour of the drag coefficient, although it removes most of the very stable cases, thus introducing bias. Additional data are required in order to establish confidence in the generalized bulk formula. Copyright © 2008 Royal Meteorological Society