Colleen M. Kaul
Associate Research Scientist, California Institute of Technology
I currently work as an Associate Research Scientist with the Climate Dynamics Group at the California Institute of Technology. As a postdoctoral scholar, I have had the opportunity to work with Tapio Schneider at ETH Zürich, João Teixeira at the Jet Propulsion Laboratory, and Heinz Pitsch at the Center for Turbulence Research at Stanford University. I completed my Ph.D. in Aerospace Engineering at the University of Texas at Austin, where I analyzed and improved models for scalar mixing in large eddy simulations under the guidance of Venkat Raman.
The unifying goal of my research, from my work on turbulent combustion to my current interests in modeling the atmospheric boundary layer and clouds, is to increase our understanding and improve our ability to make predictions about multi-scale, multi-physics fluid flows. Numerical simulations are essential for grappling with intricate interactions between disparate physical processes operating across a wide range of temporal and spatial scales. However, accurate simulations themselves require well-crafted closures for unresolved processes and appropriate numerical discretization choices, which again depend on the underlying physics. My work addresses aspects of each of these issues.
Earth's climate is among the ultimate examples of such complex multi-scale, multi-physics systems, and reduced uncertainty in simulating its future evolution is urgently needed. Much of this uncertainty is traceable to the representation of turbulence and clouds in general circulation models. My latest efforts center around the formulation of a unified turbulence and cloud parameterization that is extensible from dry convection to stratocumulus to shallow cumulus to deep convection. To realize this theoretical framework, I developed a novel single column model called SCAMPy. Additionally, I leverage state-of-the-art large eddy simulations performed with PyCLES to gain deeper understanding of the dynamical processes of clouds and their interactions with their environment as well as to inform key process closures, such as entrainment and detrainment rates, needed to complete the cloud parameterization.
The image above is a visualization of coherent convective structures in a shallow cumulus-topped boundary layer simulated by PyCLES.