There are many unknowns surrounding why convective storms occur exactly when and where they do, including environmental impacts on storm morphology and processes. In this study, we leverage 25 years of observations from spaceborne radars, along with coincident reanalysis data, to determine how the depth and width of precipitating convective storms are related to the large-scale environments in which they are observed. We find that the deepest convective features are observed in environments markedly different from the environments of the widest convective features. We identify eight large-scale environmental variables that serve to distinguish between storm modes, and then show that principal component analysis can be used to condense this information into just two scalar variables. The methodology presented offers a succinct way to describe a storm's environment and will allow us to better relate a storm's initial environment to its dynamical characteristics.

This study examines the characteristics of several model parameter perturbation methodologies for ensemble simulations of cloud microphysical processes in convection. A simplified 1D model is used to focus the results on cloud microphysics without the complication of feedbacks to the dynamics and environment. Several parameter perturbation methods are tested, including non-stochastic and stochastic with various distributions and parameter covariance. We find that an ensemble comprised of different time-invariant parameters (non-stochastic) exhibits little bias, but insufficient spread. In addition, its behavior does not respect the time evolution of convection through its various phases. Stochastic parameter (SP) methods in which no inter-parameter covariance is applied produce greater spread, but significant bias. The bias is particularly large for lognormal parameter perturbation distributions. The ensemble spread is retained and the bias reduced when time-varying parameter covariance is applied. In this case, the SP scheme is able to adapt to the time and state-dependent covariance structures and produce ensemble characteristics that are consistent with the specific microphysical processes operating at any given time. The results suggest that SP schemes would benefit from inclusion of parameter covariances, and specifically those that vary with the state of the system. It also suggests that a Normal or LogNormal SP scheme with no covariance may produce a significantly biased ensemble. Finally, the results indicate that high temporal and spatial resolution observations may be needed to characterize the variability in parameter values and covariance.