Some aspects of determining the stable boundary layer depth from sodar data

The question of estimating the height of the stable boundary layer (SBL) based on digitalized vertical profiles of sodar signal intensity has been re-examined. A simple one-dimensional numerical boundary-layer model is used to compute vertical profiles of the temperature structure parameterC _T ² . It is shown that especially at the beginning of the night (when stratification is weak) one can not expect any significant profile structure in the upper part of the SBL if its depth is determined in terms of common turbulent height scales. From this it is concluded that the SBL-height will be underestimated early in the night when derived from the maximum gradient in the signal intensity profiles. Later in the night in contrast, the computations often show elevated maxima or even zones with reduced, and above them enhanced, vertical gradients ofC _T ² , from which a SBL-height can be deduced that compares well with other common height scales. The computed profiles ofC _T ² are shown to be in qualitative agreement with observed profiles of sodar signal intensity for several analysed cases from the HAPEX-MOBILHY experiment.Comparing different SBL-depth scales with sodar observations, it is demonstrated that most of them are often closely related to a sodar-derived SBL-height only during certain phases of the night. Thus the sodar-SBL-height can, after a transition period, be perhaps associated with the lower turbulent layer of the growing surface inversion during the first part and with the height of the low-level wind maximum during the second part of the night.