An Observational And Modelling Study Of The Dynamics Of Marine Stratocumulus Clouds PDF Download

This work analyzes cloud processes on daily timescales in stratocumulus (Sc) cloud decks over eastern subtropical oceans. These clouds have a significant effect on climate. Subtropical Sc decks reflect abundant solar radiation, but emit infrared radiation at a temperature nearly as warm as the surface, causing a net cooling to the climate system. Climate models tend to poorly represent subtropical Sc, often producing too few clouds, which are too bright. The size and significance of these cloud decks combined with their poor representation in models motivates further research into how clouds are responding to their environment, and what internal processes drive observed behavior. To address this knowledge gap, a substantial Lagrangian framework is developed here, where thousands of cloudy parcels are followed for several days over the eastern subtropical oceans. These parcels are repeatedly sampled at 12-hour intervals, creating a large-scale dataset of environmental and cloud variables that can incorporate a time dimension with a variety of lead times. This framework is applied here in several ways: The first chapter incorporates the Lagrangian framework in two climate models as well as in observations for the same year. The model comparison shows which daily-scale cloud processes are well simulated and which are poorly simulated. The final two chapters address transitions in cellular structure in Sc cloud decks, first assessing which variables are associated with 24-hour transitions from classic, closed cell Sc to more broken open cells or disorganized cells, showing that transitions in cellularity are associated with separate mechanisms. In chapter 3, the results from chapter 2 are put in context using an expanded set of trajectories to study multi-step processes that incorporate interplay between environmental variables, cloud processes, and observed changes in cellular structure. The following paragraphs introduce the three chapters. A Lagrangian framework is developed and applied to two GCMs (CAM5 and HadGEM, a.k.a. UKMET) and an observational dataset in order to compare the daily-scale evolution of cloud cover and cloud variables between models and observations in the eastern subtropical oceans. Cloud cover in both models is less extensive than cloud cover seen by MODIS. Observed rain rates, as estimated by CloudSat-tuned AMSR/E 89 GHz brightness temperatures, is comparable to the HadGEM rain rates, while CAM5 rain rates appear too heavy. Inversion height estimated by MODIS and CALIPSO-observed cloud tops falls between the too-shallow CAM5 inversion and the too-deep HadGEM inversion. Lagrangian decorrelation timescales are similar in the modeled and observed environments, with e-folding times on the order of 12-36 hours for most cloud variables, shorter for cloud water path and cloud cover. Predictor variables, both meteorological and internal to the boundary layer, are tested as drivers of changes to cloud variables. Increased subsidence is modeled and observed to decrease cloud cover, inversion height, precipitation, and cloud water content. Modelled clouds tended to be oversensitive to changes in SST, while cloud microphysical and precipitation processes were poorly simulated by both models, indicating a need for improvement for the simulation of these processes. In chapter 2, mesoscale cellular convection (MCC) is classified by daytime MODIS L2 cloud liquid water path in 256km square boxes spaced 128 km apart in stratocumulus decks in the southeastern subtropical oceans. A Lagrangian framework is applied to MCC observations taken 24 hours apart in order to assess meteorological conditions and cloud properties associated with transitions from closed cell MCC to open cells, or from closed cells to disorganized cells more akin to trade cumulus. Results suggest that higher rain rates, observed by the CloudSat-tuned AMSR/E 89 GHz brightness temperatures, are associated with the closed-to-open MCC transition along with reduced cloud drop concentration as seen by MODIS and strong wind speeds, sourced from the ERA5 reanalysis. Strongly contrasting with the closed-to-open MCC transition, the closed-disorganized MCC transition is associated with entrainment warming and drying in a rapidly deepening boundary layer, observed by CALIPSO-tuned MODIS cloud top temperatures. In the closed-disorganized transition, the boundary layer appears to deepen in response to declining subsidence and reduced humidity in the lower troposphere, as well as a warmer sea surface. In chapter 3 the ability of wind speed to induce the closed-to-open MCC transition in subtropical marine stratocumulus is further investigated. This Lagrangian framework is expanded to use trajectories that span 96 hours. A new analysis is created to assess the power of a variable to predict a 24-hour closed-to-open or closed-to-disorganized MCC transition relative to a closed-to-closed case. Predicting power is compared for a large set of variables at various lead times up to 72 hours prior to the MCC transition. Results show that strong wind speeds precede heavy drizzle as a predictor of the closed-to-open transition. Further Lagrangian analysis shows that strong winds are associated with heavier rain in closed cell Sc and increasing rain rates over the 12 hours past the strong wind observation. The positive relationship between wind and rain is explained by rearranging the relationship between latent heating, humidity, and wind speed from LHF ~ wind/RH to wind ~ RH x LHF. This means that in a capped boundary layer, wind speed represents a combination of the moisture flux and moisture content, so that wind is pumping moisture by evaporating seawater into a closed system, which drives increased rainfall. The rainfall can initiate the closed-to-open transition through cold-pool convergence processes that have been shown to sustain open cells in prior modeling work. The closed-to-open MCC transition is compared to the closed-to-disorganized transition: Two different systems emerge, where the closed-to-open transition occurs when the boundary layer is overloaded with moisture, while the closed-to-disorganized transition occurs when the boundary layer dries due to excess entrainment. These results indicate that closed cell stratocumulus clouds rely on a balance between moisture input from wind and entrainment drying from the incorporation of free-tropospheric air. Excesses of moisture or drying can break apart the closed cells.