Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Meteorology and Physical Oceanography (Marine)

Date of Defense


First Committee Member

Brian E. Mapes

Second Committee Member

Stefan N. Tulich

Third Committee Member

Benjamin P. Kirtman

Fourth Committee Member

Chidong Zhang


This work uses a two-pronged approach to study the form-function relationship of convective organization and the larger scale. Form is simply the visual shape of convection and function is how the convection and larger scale interact. First, CloudSat observations are used to study cloud modulation during the Madden-Julian Oscillation (MJO). Second, a cloud systems resolving model (CSRM) with parameterized large-scale dynamics is used to examine how convective organization affects the interdependence of convection and the larger-scale. Using CloudSat observations, cloud type, total cloud cover, and temperature and moisture evolution are document across MJO phases. Deep cloud types were classified as wide or narrow as a proxy for designating organized and unorganized convective systems. For locally defined phases, the MJO exhibits a familiar progression of cloud types from shallow clouds mixed with deep, isolated convection in the growing stages of the MJO, to deep, widespread, organized convection during the mature stages, to more anvil dominated conditions during the decay stages. Comparison to the convectively coupled Kelvin wave reveals both wave types exhibit similar cloud type evolution, though, the MJO was found to be modulated more by moisture variation, while the Kelvin wave was modulated more by temperature variations. In terms of globally defined MJO phases, the wide deep precipitating systems were modulated more than other cloud types by MJO phases, with the well-known progression of cloud cover from the Indian Ocean to the central Pacific. The narrow deep precipitating systems only propagated from the Indian Ocean to the Maritime Continent. The modeling component of this work involved periodic domains, where convective organization was controlled by adding shear to a three-dimensional (3D) isotropic CSRM domain or by altering the 3D domain to be longer and narrower, until eventually becoming a 2D domain. Snapshots of convective activity in various domains reveal the 3D isotropic domain has scattered, unorganized convection, while the addition of shear leads to squall lines, and the increasingly 2D domains lead to widely spaced infinite lines (in the periodic sense) of convective blobs. When the CSRM domain mean state is coupled to a large-scale wave equation, the coupled CSRM-wave system responds to the change in organization. The shear simulations are very similar to the no-shear simulations, suggesting the convectively coupled wave is indifferent to organized convection. However, wave amplitude at equilibrium increases as the domains are stretched from 3D to 2D, though only up to a certain xy aspect. Why a particular xy aspect gives the most responsive domain remains a mystery despite many sensitivity and forcing experiments (described in appendices here). Nonetheless, this framework for studying the convective, large-scale system offers uniquely tractable and controllable ways forward to understanding the multi-scale nature of atmospheric convection. The interdependence of convective organization and the larger-scale should help guide the future of cumulus parameterization and/or other options for representing the convective scale in general circulation models (GCMs; e.g. superparameterization techniques)


tropical convection; multi-scale interactions; cloud resolving model; cloud radar; Madden-Julian Oscillation; convectively coupled waves