College of Engineering | Apply to CSU | Disclaimer | Equal Opportunity Statement | Privacy | Search CSU

Convective Cold Pools: Characterization and Soil Moisture Dependence

You must be on the CSU network—either physically or using VPN—to watch this or any of the videos on this site.

July 21, 2016
Aryeh Drager
Hosted by Sue van den Heever (advisor), Wayne Schubert, Christopher Davis (affiliate), Michael Kirby (Mathematics)


Convective cold pools play an important role in Earth’s climate system. However, a common framework does not exist for conceptually defining and objectively identifying convective cold pools in observations and models. The first part of this thesis begins with a review of the identification methods used in previous works. This is followed by an investigation of convective cold pools within a high-resolution simulation of rainforest convection simulated using the Regional Atmospheric Modeling System (RAMS), an open-source cloud-resolving model with a coupled land-surface model. Multiple variables are assessed for their potential for identifying convective cold pool boundaries, and a novel technique is developed and tested for identifying and tracking convective cold pools in numerical model simulations. This algorithm is based on surface rainfall rates and radial gradients in the density potential temperature field. The algorithm successfully identifies near-surface cold pool boundaries and is able to distinguish between connected cold pools. Once cold pools have been identified and tracked, composites of cold pool evolution are then constructed, and average cold pool properties are investigated. One novel result is the presence of moist patches that develop within the centers of cold pools where the ground has been soaked with rainwater. These moist patches help to maintain cool temperatures and prevent cold pool dissipation, which has implications for the development of subsequent convection.

The second part of this thesis explores how the properties of convective cold pools are modulated by soil moisture. Three high-resolution simulations of tropical rainforest convection are performed using the RAMS, and the initial soil moisture is varied between 25% and 75% saturation. The cold pool identification algorithm developed in the first part of the thesis is used to construct composites of cold pools within each simulation, and the composites are compared. When soil moisture is decreased, stronger convective cold pools result. These stronger cold pools are also smaller because increased sensible heat fluxes in the reduced soil-moisture simulations cause the cold pools to dissipate more quickly as they expand. Finally, the rings of enhanced water vapor that have been documented in previous studies of tropical cold pools disappear when soil moisture is reduced. These results emphasize the role that land surface properties can have in modulating convective cold pool properties.