Scripps Institution of Oceanography
University of California, San Diego
La Jolla, CA 92093
Marine stratocumulus play an important yet still poorly modeled role in the climate system (Charlson et al., 1987, Roeckner, 1994). These clouds cool the planet, having a large albedo, but little infrared effect. A fundamental question is whether such clouds will exist at a given time and location. Stratocumulus is often formed at higher latitudes as stratus and advected equatorward until it breaks up. Possible mechanisms for cloud breakup include strong subsidence, cloud top entrainment instability (CTEI), drizzle, solar heating and resultant boundary layer decoupling, and surface forcing. The Atlantic Stratocumulus Transition Experiment (ASTEX) was conducted (AŤores, June 1992) to investigate these potential cloud breakup mechanisms.
A secondary question is the importance of cloud texture on the radiative properties of the stratocumulus, and in particular on the cloud albedo. Simple calculations (using the delta-Eddington and independent pixel approximations) were made with a fixed total liquid water content which was distributed in space sinusoidally with varying amplitude. For a mean optical depth of 10, the albedo decreases by ‰ 15% in going from a homogeneous cloud to one in which the optical depth becomes zero at the troughs of the sinusoid. Hence, the effect of cloud texture is indeed significant.
Convection can be driven in the stratocumulus layer by cloud-top longwave radiative cooling and warming below. This warming may be due to solar heating or to longwave cloud-base warming. This produces a cellular structure in the cloud. The dominance of cloud top cooling leads to closed cells (Getling, 1991; Agee, 1987). A quantitative understanding of radiatively induced convection is thus required, both to correctly estimate the cloud top entrainment rate (and the potential for cloud breakup) and to estimate texture statistics for the albedo calculation.
Norris presented an example of some liquid water and radiative flux data from aircraft during ASTEX. The cloud optical depth was estimated from aircraft profiles of cloud liquid water content and effective radius. A delta- Eddington calculation of the cloud albedo agreed well with the albedo measured during an overflight of the cloud. An interesting result was the observation of significant energy at the scale of ‰ 600 meters in the cloud albedo spectrum and possibly the liquid water content spectrum. This is thought to be related to radiatively generated convective cells in the cloud. It is planned to use ASTEX data to test modeling efforts underway.
An existing three-dimensional Rayleigh-Bé nard code (Hathaway and Somerville, 1986) has been modified to perform simulations of a cloud-topped marine boundary layer. Convection is driven in a virtual cloud layer characterized by an effective differential radiative forcing (top cooling and base warming). The cloud layer is above an initially neutral subcloud layer and below a strong capping inversion. As convection sets in, negatively buoyant cloud top air is mixed down into the top of the neutral subcloud layer and destabilizes it, causing the model to jump to a new steady state with a characteristic profile of convective kinetic energy. Figure 12.1 shows the growth of convective cells in the mid-cloud horizontal plane. Lighter shades indicate upward motion. Clearly the cellular pattern is closed, as is expected for a dominantly top-cooled system. Figure 12.2 shows a vertical slice through the model in three variables. The upper panel shows the potential temperature perturbations generated by convection. The convective cells are clearly visible. The second panel shows that the downdrafts generated at cloud top extend all the way to the surface, thereby forcing the whole boundary layer. Figure 12.3 shows the profiles of horizontally averaged potential temperature, both initially and after the convective steady state is established. This final state is characterized by an unstable surface layer, a near neutral mixed layer, a slightly stable transition layer below cloud, a cloud layer which is unstable, but less so than before convection was established, and a strong capping inversion. All of these features are also found in the actual observed marine boundary layer.
The Rayleigh-Bé nard boundary layer model will be used to investigate the amount of convective kinetic energy and cloud top mixing generated as a function of such factors as the strength of radiative forcing (cloud Rayleigh number), the depth of the subcloud layer, and the strength of the capping inversion. Such a study will provide a simple dynamical framework for more complicated models including realistic clouds and turbulence. The long-term aim is to make realistic predictions of observed cloud spatial scales and entrainment rates at cloud top and base.
References
Agee, E. M., 1987: Mesoscale cellular convection over the oceans. Dynam. Atmos. Oceans, 10, 317-341.
Charlson, R. J., J. E. Lovelock, M. O. Andreae, and S. G. Warren, 1987: Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature, 326, 655-661.
Getling, A. V., 1991: Formation of spatial structures in Rayleigh- Bé nard convection. Sov. Phys. Usp., 34, 737-776.
Hathaway, D. H., and R. C. J. Somerville, 1986: Nonlinear interactions between convection, rotation and flows with vertical shear. J. Fluid. Mech., 164, 91-105.
Roeckner, E., 1994: Parameterization of clouds and radiation in climate models, this volume.
