The structure of gradually transforming marine stratocumulus during the ASTEX first Lagrangian experiment

Abstract

AbstractAn analysis of the mean and turbulent structure of the planetary boundary layer is presented, using aircraft data obtained during the Atlantic Stratocumulus Transition Experiment first Lagrangian experiment. The analysis of the mean vertical structure relies mostly on slant profile data, however, part of the analysis also uses so‐called ‘porpoise’ runs in the vicinity of the cloud top. The analysis of the turbulence structure utilizes all of the data, horizontal flight legs as well as slant profiles, and presents both scaled turbulence statistics and a spectral analysis. The purpose is to investigate details in the temporal development of the first Lagrangian experiment, and to study the turbulent structure in order to find scale relations that can be utilized in improved modelling of the stratocumulus‐capped boundary layer. The hypothesis is that so‐called ‘cloud decoupling’ has a significant impact on the turbulence scaling.The measured cloud water is corrected to become consistent between the two participating aircraft. The corrected maximum cloud water is more constant in time than in previous studies of the first Lagrangian; it is suggested that previously analysed temporal variability may be an artefact of using two different instruments to measure cloud water. The cloud‐top jump in equivalent potential temperature is positive in the first three flights, sufficiently large for a stable cloud top, but decreases and becomes negative for the last two flights in line with the breaking up of the cloud field. The upper part of the boundary layer warms and dries more rapidly than the lower, suggesting entrainment to be a dominant factor in the development of the cloud. The analysis of the porpoise flight legs illustrates how sharp features at the cloud‐top inversion are smoothed out while analysing several profiles, flight legs or even entire flights, rather than normalizing each profile with the height to the local cloud top.Scaling of the turbulent fluxes and the velocity variances shows that the cloud layer is decoupled throughout almost the entire first Lagrangian. It is also shown that while surface‐layer scaling is appropriate for the marine atmospheric boundary layer (MABL) the cloud‐layer turbulence scales with convective scaling. The low‐frequency spectrum has no spectral gap, with the exception of the vertical velocity, and continues to increase with decreasing frequency with only a change in slope; this indicates the presence of mesoscale motions. The cospectra of the turbulent fluxes are very noisy throughout. When averaging several spectra into height intervals, the scales of the turbulence as indicated by the peak in the averaged spectra appear to vary with height above the surface in the MABL but to be height‐independent in the cloud layer. An attempt to calculate a dissipation length‐scale shows scatter, but is consistent with a dissipation length‐scale increasing linearly with height in the MABL, but being roughly constant or possibly parabolic with height in the cloud layer. Copyright © 2003 Royal Meteorological Society