I am broadly interested in problems of natural systems that are amenable to mathematical modelling. Most of my work has been in fluids, and I am currently working on more applied oceanography. Below are brief summaries of several past and current projects.
If you have landed here from my 2020 Ocean Sciences poster, the poster and additional details of the work can be found here.
That poster is part of the work on a larger effort:
Modeling the effects of climate: biological impacts of (sub)mesoscale upper-ocean physics
Together with Kelvin Richards, Frank Bryan, Matt Long, Dan Whitt, and Kate Feloy, I am studying the effects of resolution and climate change on the biological environment in ocean models. I developed a set of idealized tracers that give insight into atmosphere-ocean gas exchange, nutrient availability, phytoplankton production, and carbon export to depth. Other team members developed a method of running high-resolution global future climate projections without running the years between present and future states. Together, these methods will allow us to determine how changes in ocean stratification with climate will affect the upper-ocean biogeochemical environment. We will run at a wide range of resolutions to understand the effects of including different physical processes on the results.
We have 3 papers: 1 on submesoscale activity in a regional high-res run nested inside the mesoscale global runs, 1 on global variation in projected phytoplankton production, and 1 in prep on the regional submesoscale impact on phytoplankton production.
With Jim Gunton and his group, I studied the self-assembly of Janus particles. Janus particles are manufactured ellipsoids that have two surfaces: one is mutually attracting (with the same side facing, two particles move towards eachother) and the other is mutually repelling (with the same side facing, two particles move away from eachother). When suspended in a fluid, the fluid and Janus particles together is a colloid. Using Monte Carlo simulations, we examined the behavior of the particles, which generally form clumps with the attracting faces oriented inward and the repelling faces oriented outward. This formation of clumps is called self-assembly. We also studied the ability of these clumps of Janus particles to enclose neutral particles, called encapsulation. The efficiency of self-assembly and encapsulation were related to enthalpy/energy and entropy as drivers. Results showed that flatter Janus particles formed nearly-spherical clumps with fewer particles and a smaller interior, while more spherical Janus particles formed clumps with irregular shapes, including chains. An intermediate aspect ratio was found to be most efficient at encapsulation. 2 publications.
Ocean wave breaking and mixing
With Greg Gerbi and Sam Kastner, I studied the effects of turbulent kinetic energy injected by breaking waves in the rate of vertical mixing in the surface ocean. Generally, the ocean is stratified, meaning less dense water is situated above more dense water. At the surface, the mixed layer has fairly constant density. The depth of this mixed layer changes both due to heating and cooling by the atmosphere and solar radiation and due to mechanical efforts by wind and waves. When surface waves break, their overturning begins a cascade of turbulent motions that increase the kinetic energy of the near-surface ocean. This added kinetic energy can be converted to potential energy by mixing denser water into the mixed layer, thereby deepening the mixed layer. We found that breaking waves cause the mixed layer to begin deepening faster in a wind-driven situation, leading to a mixed layer deeper by 10-15% after 2 days of steady wind. See publication in Journal of Physical Oceanography.
Competition between stirring and mixing in a small overturning ocean eddy
Recently there has been a noticeable effort to study and quantify chaotic stirring in the ocean. Chaotic motions are aperiodic, sensitive to initial conditions, and are generally present in limited regions of a system. As with all stirring, chaotic stirring can quickly filament any marker fluid (like coffee in cream), creating large gradients. The large gradients then promote fast mixing of the marker throughout the rest of the flow. I became interested in the relative effects of stirring and mixing on the distribution of the marker. This work with coauthors Larry Pratt, Irina Rypina, and Peng Wang quantifies these relative effects using several methods in the context of a small (submesoscale) overturning ocean eddy. We found that chaotic stirring is the dominant process in large chaotic regions and that chaotic stirring can double the effective mixing rate of the system despite only acting in some regions. See publication in Nonlinear Processes in Geophysics.
Controls on the properties of the Western Alboran Gyre
This work, with Larry Pratt and Irina Rypina, endeavors to explain the water properties and maintenance of the Western Alboran Gyre (WAG) in the western Mediterranean Sea. The WAG is located near the Strait of Gibraltar, and is typically observed to have a core of water with temperature and salinity similar to the Atlantic. The WAG also is slightly unsteady: about once a year, it disappears and re-forms. Using the MIT general circulation model, I simulated the region for about one year. Analysis of the 5-month period with a steady WAG showed that advection of water from the surroundings into the gyre generally moved the water properties toward Mediterranean values and slowed the rotation of the gyre. It seems that the Atlantic water properties are from the formation of the gyre, when the current from the Atlantic moves more slowly and mixes with the Mediterranean waters less. The advective exchange with surrounding waters is much faster at the edges of the gyre than the center, leading to the center retaining the Atlantic properties better over the lifetime of the gyre. The continued rotation of the gyre is driven primarily by the inflow and outflow currents connected to the Strait of Gibraltar. This work is presented in my thesis, and a publication has been submitted for review at the Journal of Physical Oceanography (January 2020).
Tracing the Missing Physics of Submesoscale Entrainment andSubduction
Together with Jacob Wenegrat and Baylor Fox-Kemper, we plan to improve the scientific understanding of submesoscale ocean physics and improve the existing parameterization of mixed layer instabilities, one type of submesoscales.
Submesoscale physics connects the mesoscale, or ocean weather scale, to the smaller turbulent motions where energy is dissipated.
There are multiple submesoscale dynamical features with overlapping scales, and part of this work will use state of the art large eddy simulations to improve our understanding of the multi-scale interactions between these processes.
We will also use more idealized simulations of ocean fronts to quantify the way mixed-layer submesoscale motions impact the fluxes across the base of the mixed layer.
Current mixed-layer instability parameterizations explicitly do not include these fluxes, causing large-scale ocean models to miss a portion of the effects of these scales on ocean stratification and biogeochemical cycles.
I will work with Matt Mazloff (UCSD) and Baylor's student to test the impact of our improved parameterization in a realistic model of the Southern Ocean. This model, B-SOSE, currently contains no parameterizations of the submesoscale.
By analyzing runs with the existing and new parameterization, we will quantify the impact of the submesoscale effects on heat, oxygen, and plankton biomass.
We will thereby learn whether including our parameterization in this global-scale model brings it closer to observations.
The Physics-informed AI Climate Model Agent Neuro-symbolic Simulator (PACMANS) for Tipping Point Discovery"
PIs Jennifer Sleeman and Anand Gnanadesikan
We are using a combination of box models, global general circulation models, and state-of-the-art AI/ML methods to understand whether the Atlantic Meridional Overturning Circulation is approaching a tipping point due to the warming climate. Listen to the related DARPA podcast below. More information on our work is available here.
Approved for public release; distribution is unlimited. This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Agreement No. HR00112290032.