Sonic anemometers are used to determine the vertical fluxes of momentum and scalars in the atmospheric surface layer. In practice the coordinate systems of the sonic anemometer and the surface will not be perfectly aligned so that a tilt-correction for these small deviations of the anemometer coordinates is needed. Three options are available: (1) to use known (e.g. measured) angles, (2) to impose requirements on the means and stresses within an averaging period (the 'classic method'), and (3) to assume that the mean flow over a certain period, as observed by the sonic, is in a plane parallel to the surface (planar fit method).
Data from five 3-D sonic anemometers (three different types) collected in June 2001 have been used to assess these three methods. The main questions were:
- are the rotation angles of the 'classic' method significant (relative to their tolerance)?
We find that for all five systems the run-to-run variation of classic pitch angles is significant relative to its tolerance but the variation shows no correlation among adjacent instruments, suggesting that there is no meteorological cause for the variation. The mean and run-to-run variation of the roll angle are in most cases insignificant.
Comparison of the correlation of the shear stress between different systems, shows that day-by-day (modified) planar fit rotation increases the correlation for most combinations of systems. To our opinion this enhanced consistency shows the superiority of the (modified) planar-fit method to the classical tilt-corrections in the context of surface-flux estimation.
Supplementary URL: