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

Chidong Zhang

Second Committee Member

Brian Mapes

Third Committee Member

Paquita Zuidema

Fourth Committee Member

Jimy Dudhia

Fifth Committee Member

Pavlos Kollias


Tropical and subtropical clouds have a large influence in shaping the observed patterns of wind, temperature, and moisture in many time scales and different regions of the world. While shallow clouds have been traditionally investigated in trade wind regimes, shallow clouds in the deep tropics have recently gained attention because of their possible importance in the development of deep convection. Shallow clouds are ubiquitous in the tropics and connect the boundary layer and the lower troposphere. Difficulties to observe and model shallow clouds have resulted in unclear aspects about their interactions with the other elements of the tropical variability. It is a challenging task to address such interactions, but the objective is well worth to pursue. This study takes a three-pronged approach to advance the understanding of the role of shallow convection in the variability of the deep tropics. The first part of this study investigates the variability and feedbacks of shallow clouds in their large-scale environment from more than ten years of observations of a vertically pointing cloud radar and soundings over Manus Island, in the tropical western Pacific. To do so, it is developed a method to estimate bulk shallow cloud moistening due evaporation of cloud condensed water from observations of liquid water content, precipitation, and temporal cloud fractions. The second part of this study combines the Manus and reanalysis data to investigate the role of shallow convection in the evolution of large-scale convective events associated and not associated to Madden-Julian Oscillation (MJO). The third part of this work concerns the role of shallow convection in driving large-scale circulation. To investigate this problem, the region of study is focused on the tropical Atlantic. Surface equatorial winds over the equatorial Atlantic in models are notorious for their westerly biases and their association to Amazonian rainfall. This part of the study investigates the connection between the westerly biases and shallow convection over the Amazonia using simulations of several atmospheric global climate models. Our results indicate that shallow clouds provide a non-negligible amount of moisture to the lower troposphere. They, however, lack coherency with perturbations in low-tropospheric moisture, temperature, and wind circulation in synoptic time scales. In longer time scales, anomalies in their depth, occurrence, and estimated moistening can be, to a certain degree, related to those in lower tropospheric stability. Consistently, during the five to tend days prior to the rainfall peaks of MJO and non-MJO large-scale convective events at Manus, anomalous increases in low-level moisture are evident, but they cannot be attributed to moistening by shallow clouds. During this period, shallow clouds provide background moistening. The characteristic low-level moistening prior the rainfall peaks of MJO events is mainly caused by anomalous nonlinear zonal advection. The analysis about the role of shallow convection in the MJO revealed MJO signals that have not been observed before. They include an ultra slow (2.8 m s-1) structure in mid to upper tropospheric temperature and geopotential height anomalies that propagates eastward over the central and eastern Pacific with no discernable anomalies in precipitation associated to it. These results suggest the existence of a intrinsic structure of the MJO governed by dry dynamics. Different process might act to energize this structure. Moist convection, including shallow convection, can be an effective one. The importance of shallow convection in the tropical variability and circulation seems to be case-dependent and subject to the interaction with other factors. This is the case in the equatorial Atlantic. Our results indicate that westerly biases over this region can be associated to weak or absent Amazonian shallow convection, but they can also be associated to weak boundary layer entrainment.


shallow convection; tropical variability; MJO; westerly biases