Observations of drizzle cells in marine stratocumulus

Airborne radar and in situ observations collected during the Second Dynamics and Chemistry of Marine Stratocumulus (DYCOMS-II) project are examined with respect to the presence and role of mesoscale organization within the apparently homogenous stratocumulus layer. This work is divided into two parts: the first part describes a modified dual-Doppler analysis used to derive velocity components lying in a single-plane that is described by the orientations of the radar beams and the aircraft velocity vector. This technique, referred to as Single-Plane Dual-Doppler (SPDD) analysis, is then used to examine the dynamic structure of the stratocumulus-topped boundary layer.; The SPDD analysis differs from traditional dual-Doppler analysis in that, due to variations in aircraft attitude, the radar beams trace out a curtain rather than a plane. Thus, it is necessary to form a three-dimensional solution that incorporates the Doppler velocity measurements from the two beams along with an a priori velocity estimate. Two velocity components, presumably those defining the idealized observation plane, which contain a minimal contribution from the a priori velocity estimate are then extracted from the three dimensional velocity field. The velocity components retrieved using this technique are affected by a number of errors including: noise in the Doppler velocity measurements, spatial and temporal displacements between the velocities measured by the two beams, errors in the aircraft motion correction of the Doppler velocities, errors in and variations about the a priori velocity estimate. Random noise is found to be a minor contribution for high SNR, but is the dominant error source for low SNR and effectively limits the use of the SPDD analysis to SNR > 0 dB. Displacements between the beams increase approximately as the square of the range from the aircraft to the region of interest and are a significant error source for ranges > 1km. Errors due to incorrect aircraft motion correction remain poorly quantified, but are believed to be < 0.5 ms-1 and result primarily in biases in the retrieved velocity components.; The DYCOMS-II observations focus primarily on a single case, in which drizzle allows velocities to be retrieved throughout the cloud layer and for most of the subcloud layer. Inspection of the vertical-plane velocity fields reveals the presence of mesoscale circulations with a typical horizontal of 3--5km. The most compelling evidence of these circulations is a pattern The strongest drizzle is found in the updraft branches of the mesoscale circulations, thus the mesoscale structures are referred to as drizzle-cells. Most of the variance in the horizontal velocities is associated with the mesoscale circulations. However, at most half of the variance in the vertical velocity is associated with these circulations and then, only in the middle of the boundary layer---a disparity that results from the larger width of the updrafts and downdrafts relative to the depth of the inflow and outflow branches of the circulations.