The ultimate goal of this study is to improve the representation of convective transport by cumulus parameterization for mesoscale and climate models. As Part 1 of the study, we perform extensive evaluations of cloud-resolving simulations of a squall line and mesoscale convective complexes in midlatitude continent and tropical regions using the Weather Research and Forecasting model with spectral bin microphysics (SBM) and with two double-moment bulk microphysics schemes: a modified Morrison (MOR) and Milbrandt and Yau (MY2). Compared to observations, in general, SBM gives better simulations of precipitation and vertical velocity of convective cores than MOR and MY2 and therefore will be used for analysis of scale dependence of eddy transport in Part 2. The common features of the simulations for all convective systems are (1) the model tends to overestimate convection intensity in the middle and upper troposphere, but SBM can alleviate much of the overestimation and reproduce the observed convection intensity well; (2) the model greatly overestimates Z e in convective cores, especially for the weak updraft velocity; and (3) the model performs better for midlatitude convective systems than the tropical system. The modeled mass fluxes of the midlatitude systems are not sensitive to microphysics schemes but are very sensitive for the tropical case indicating strong microphysics modification to convection. Cloud microphysical measurements of rain, snow, and graupel in convective cores will be critically important to further elucidate issues within cloud microphysics schemes.

Using 33 year European Centre for Medium-Range Weather Forecasts Re-Analysis Interim reanalysis in the months of August and September, we found that more than half of the low-level, moist vortices (called wet vortices) originating from south of the African easterly jet merged with a shallow, dry vortex from the north after leaving the West African coast. A dry vortex involved with the merger process is referred to as a D-vortex, and the process is referred to as a D-vortex merger. Dry vortices influenced by more intense African easterly waves moved southwestward and had a greater potential to serve as D-vortices in the merger process. The D-vortex merger occurred in the predepression stage of 70% of tropical cyclones (TCs) that formed in the Atlantic main development region and in 55% of nondeveloping systems. Further analysis showed that developing systems with the D-vortex merger (DM) were statistically dominated by a more intense wet vortex whose 500 hPa relative humidity was also significantly higher, while nondeveloping systems with the D-vortex merger (NM) were dominated by a more intense dry vortex. The average intensity of wet vortices for DM was more intense than that for NM, significant at a 95% confidence level. Moreover, warmer Saharan air was observed for DM than NM. While TC genesis is largely controlled by the large-scale environment over ocean, differences in vortex characteristics and environment over northwestern Africa between DM and NM could potentially help predict whether a tropical system associated with the D-vortex merger will ultimately evolve into an Atlantic TC.