| description abstract | An investigation of a cloud-topped boundary layer during a wintertime cold air outbreak over Lake Michigan has been conducted based on research aircraft 1 and 20 Hz data for five vertically-stacked levels within the boundary layer. Mean east-west cross sections for this northerly flow event have been prepared for heat, moisture, and momentum fields along with spectral decomposition and turbulent flux statistics. Results have been interpreted on the basis of three generalized spatial domains: mesoscale ≥5 km, large eddies from 100 m to <5 km, and higher frequency turbulence <100 m. Mesoscale structures identified in the mean cross-sections include the lower Michigan land breeze, a strong offshore convergence snow band, cloud streets and precipitation bands over the lake, and entrainment spikes in the vicinity of the inversion layer. Analyses of particle concentrations, sizes, and liquid water content have also been presented in mean cross sections, which correspond well with all identifiable mesoscale structures. This study has also examined a ?hand-picked? data set in the mixed layer over the middle of the lake not affected by the land breeze and inversion layer phenomena. Analysis of this 20 Hz data set shows convergence in the turbulence statistics for different 15 km flight segments, which also introduces an analysis technique called telescoping skewness. Also included are two independently derived vertical profiles of vertical velocity skewness that are seen to be nearly identical. Differences and similarities in numerical model results and observations are discussed that suggest a modeling weakness in predicting convective turbulent transport in the cloud layer when driven by strong surface heating. Analysis shows the subcloud layer of the mixed layer to be characterized by large eddies arranged in accordance with open cell geometry (narrow strong updrafts with weak broad regions of downdraft). This positive skewness in the subcloud layer is about twice that in the cloud layer indicative of a modulating effect by cloud layer processes such as condensational heating. The vertical profile of the normalized buoyant production of turbulence kinetic energy is also presented, which agrees well with Deardorff's model results with strong surface heating and cloud-top cooling. The role of the mean vertical velocity is also considered in the mathematical expansion of skewness, which is seen to agree with Deardorff's model results with strong surface heating and cloud-top cooling. The role of the mean vertical velocity is also considered in the mathematical expansion of skewness, which is seen to agree with Deardorff's results if w? is set equal to zero (a feature of numerical models, but not necessarily a feature of the real planetary boundary layer as evidenced by this study). | |
