Fovell and Dailey (2001) used a 3D, high resolution cloud model to examine the interaction between the sea-breeze front (SBF) and horizontal convective rolls (HCRs) in a simplified dynamical framework roughly resembling Florida. In this case, the vertically sheared initial environmental flow was directed parallel to coastline, and thus the HCRs which appeared by early afternoon were aligned parallel to the SBF. The SBF systematically encountered, and merged with, roll updrafts and downdrafts as it progressed inland.

Several of these encounters resulted in the appearance of deep convection. This is of interest because deep convection did not appear in simulations made excluding either the SBF or the rolls, suggesting both phenomena were needed in this case. In addition, the deep convection was initiated ahead of the SBF, above a boundary layer roll and at least 45 minutes prior to the front-roll merger. In the first event, the shallow cloudiness that had been present above the SBF for some time was suppressed following the appearance of the deep roll convection. Yet, that convection not only reappeared but also developed into deep convection itself... all prior to the front-roll merger. That merger took place as two deep, separately initiated updrafts combined to produce a strong if short-lived downdraft in between them.

Subsequent examination of this case has focused on three questions: (1) Why were both phenomena needed in order for deep convection to occur?; (2) Why was the roll convection spawned well before the SBF's arrival?; and (3) How did the SBF cloudiness manage to be reinvigorated following the roll convection-induced suppression? The first two questions directly relate to convective initiation, and will be the focus of this talk.

We will demonstrate why both phenomena, the sea-breeze and the HCRs, were each necessary yet not sufficient factors in deep convective development in this case. The sea-breeze circulation (SBC) moistens the middle troposphere out ahead -- over the land -- making that layer more convectively favorable. Yet, the convection still requires a spark that is provided, indirectly, by the rolls. As part of the SBC, horizontal flow directed seaward is induced above the mixed layer over land. This flow passes over the roll updrafts, perpendicular to the roll axes, and results in the generation of gravity waves in the midtroposphere via the obstacle effect (e.g., Mason and Sykes 1982). These gravity wave perturbations, superimposed upon the SBC's modifications, result in the local excitation of deep convection within the moistened layer. It is noteworthy that the roll cloud initially contains no air from the boundary layer.