Environmental variability during TOGA COARE and its effect on mesoscale convective systems: Observations and modeling

Two-dimensional model experiments using the Goddard Cumulus Ensemble model are performed in order to examine the role that environmental profiles of wind and humidity play in influencing the structure of mesoscale convective systems (MCSs) over the tropical oceans. The initial environments used in this study are derived from the results of a cluster analysis of the TOGA COARE sounding data. The model data are analyzed with methods and measurements similar to those used in observational studies.; Experiments to test the sensitivity of MCSs to the thermodynamic profile focus on the role of humidity in the free troposphere. As humidity is increased, model storms are found to transition from small, weak systems with relatively little precipitation to strong, upshear tilted systems with large amounts of rainfall. This behavior is hypothesized to be related to the entrainment of environmental air into the updrafts. Supplemental experiments show that the model is most sensitive to moisture in the lowest 3 km.; Experiments to test the sensitivity of MCSs to the kinematic profile focus on the amount of vertical wind shear between approximately 2 and 10 km. As mid-level shear was increased, the total amounts of condensation and rainfall decreased and systems acquired more of an upshear tilt. The dynamical and microphysical characteristics of the runs changed dramatically in different shear environments. Increased shear led to weaker updrafts in the mid-troposphere, reducing the amount of graupel produced and generally higher brightness temperatures at 85 GHz. Shear in the mid-levels affects the updrafts primarily by tilting the upward momentum into the horizontal.; A comparison of the properties of updraft and downdraft cores and cold pools produced in the model with observations of those quantities is very favorable, suggesting that the model captures the dynamical aspects of reality rather well. However, the ice-scattering signatures are too strong. This is believed to occur because graupel is produced at too high an altitude in the model. Observations of ice processes in real tropical clouds are needed to improve this aspect of the model.