In this paper we examine the salient features of the drizzle-cells and discuss the implications of these features. The centers of the cells are marked by: a pattern of divergence (convergence) in the horizontal velocity field within the cloud- (subcloud-) layer, higher reflectivity, and higher cloud- and echo-top heights. The boundaries between cells are more clearly defined than the centers and are marked by: convergence (divergence) in the horizontal velocity field within the cloud- (subcloud-) layer, lower reflectivities, and lower cloud-/echo-top heights. The circulation implied by the horizontal velocity field indicates that drizzle occurs primarily in and near the updraft branch of the mesoscale circulation. The updraft-branch of the mesoscale circulation cannot be observed directly in the vertical component of the velocity field for two reasons: (i) the vertical air motion is masked by the fallspeed of the drizzle droplets (the effect of the particle fallspeed on the vertical velocity is not simply a bias due to correlations between the vertical air velocity and the particle fallspeed), and (ii) the relative width of the updraft-branch of the mesoscale circulation (~1km) compared to the depth of the inflow and outflow branches (< 0.5km ), both of which are constrained to lie within the depth of the boundary layer. Because of the disparity between the depth of inflow/outflow branches and the width of the updraft branch, the vertical air velocity is dominated by eddies with a horizontal extent comparable to, or smaller than, the depth of the boundary layer. Indeed, the updraft-branch of the mesoscale circulation can be more appropriately thought of as a region in which there is a higher probability of finding a small-scale updraft rather than a weak, monolithic updraft.
The observed features of the drizzle-cells have several implications for the characteristics and evolution of the stratocumulus-topped boundary layer including: (i) The co-location of high-reflectivity cores with the updraft-branch of the mesoscale circulation means that most of the drizzle falls into upward moving air and, as a consequence, evaporates more slowly, more completely, and closer to cloud-base. (ii) This co-location also suggests that traditional cold-pool models in which evaporating drizzle leads to divergence/outflow are not applicable to stratocumulus. (iii) The mesoscale circulations, which carry a large fraction of the TKE within the boundary layer, allow the cloud- and subcloud-layers to remain strongly coupled despite the stabilizing effects of evaporating drizzle.
The DYCOMS-II observations suggest the following conceptual model of the interaction between the mesoscale circulations and drizzle: mesoscale circulations are present in all cases, as drizzle begins to form locally it evaporates completely before reaching the surface. This leads to increased moisture in the subcloud layer, and consequently a lowering of cloud-base in the updraft-branch of the mesoscale circulation. It is widely recognized that drizzle increases extremely rapidly with increasing cloud-depth (or liquid water path) thus, the increased cloud-depth in the updraft-branch of the mesoscale circulation leads to a disproportionate increase in drizzle in the neighborhood of the updraft-branch which, in turn largely evaporates, leading to further localization in both liquid water content and drizzle in the neighborhood of the updraft-branch. In essence, drizzle allows water to short-circuit the mesoscale circulations thereby providing a mechanism for creating localized inhomogeneities in the cloud liquid water content and, by extension, in drizzle itself.