Research interests


1. Storm Track Dynamics

Atmospheric variability in the extratropics is dominated by baroclinic eddies. These eddies carry out the bulk of the transport of momentum, heat, and moisture in the mid-latitudes and therefore play a vital role in Earth's climate. Eddies are generated by baroclinic processes, preferentially in regions of strong temperature gradients and vertical wind shear near the entrance of the storm tracks, the regions of strong eddy activity. Even small shifts in the location of the storm tracks can have significant effects on regional climate, for example, through changes in hydrologic balances, especially in regions that in the present climate are on the flanks of the storm tracks. Our studies focus on understanding the mechanisms that set the latitudinal position, longitudinal extent and seasonal variability of storm tracks and what controls the formation and intensity of baroclinic eddies. We use idealized general circulation models (GCMs) to isolate specific physical processes.  Particularly we focus on the effect of stationary waves and water vapor on the evolution, extent and location of the storm track.



Verically integrated Eddy kinetic energy (left) and temperature gradient (right) for aquaplanet GCM simulations in response to NH surface heating perturbation (top), and NCEP reanalysis data averaged over 40 years (bottom).

 More information:

The role of stationary eddies in shaping midlatitude storm tracks [download]

Downstream self-destruction of storm tracks [download]

Winter cold of eastern continental boundaries induced by warm ocean waters [download]


2. The gravity signature of internal dynamics on Jupiter and Saturn

In 2016, NASA's Juno mission will arrive at Jupiter, and obtain for the first time a high precision gravity spectrum of this planet. In preparation for the arrival of this data, we develop dynamical models that will allow us to invert this gravity data to determine the sub-cloud atmospheric dynamics of this planet. Our focus is on the high order (n > 10), and odd gravity harmonics for which the signal from the dynamics will dominate the static signal of the planet if the dynamics are deep enough. We do this through developing a hierarchy of dynamical models and using adjoint inverse methods. The same type of analysis will be done for the proximal orbit gravity data of Cassini at Saturn planned for 2017.



Left: Artist image of Juno at Jupiter (NASA); Right: our calculated gravity spectum for different e-folding zonal wind decay depths (in km). Green squares are the caluculated solid body gravity harmonics. Full (open) dots denote positive (negative) values. Black plus signs are the measured J2, J4 and J6 harmonics.

 More information:

An improved method for estimation of Jupiter's gravity field using the Juno expected measurements [download]

Inferring the depth of the atmospheric circulation on Jupiter and Saturn through gravity measurements by Juno and Cassini and an adjoint based dynamical model [download]


3. Atmospheric Dynamics on Giant Planets

Jupiter and Saturn's atmospheres are dominated by multiple zonal jets with strong superrotation around the equator. We have built a new general circulation model (based on the MITgcm dynamical core), for studying giant planets. The model's geometry is a full 3D sphere down to a small inner core (unlike the traditional atmospheric thin spherical shell). It is non-hydrostatic, anelastic � thus includes the full radial variation in density, uses a equation of state suitable for hydrogen-helium mixtures (SCVH), and is driven by internal heat. In the parameter regime suitable for giant planets the convective plumes tend to align with the direction of the rotation axis. We show that compressibility effects cause significant baroclinic shear  and therefore to diminished interior winds. Interior convection drives convection columns parallel to the axis of rotation which propagate  eastward, driven by dynamics similar but opposite to that of atmospheric Rossby waves. This leads to up-gradient eddy angular momentum fluxes which drive the equatorial superrotation. We further study these interior dynamics using a simplified barotropic annulus model, which shows that the planetary vorticity radial variation causes the eddy angular momentum flux divergence, which drives the superrotating equatorial flow.

Simulations from the Jupiter-MITgcm (JCM): Zonal velocity at the 1 bar pressure level (left); Entropy anomalies on the equatorial plane (right, red represents higher entropy).

 More information:

Equatorial Superrotation on Gas Giants Driven by Internal Convection [download]

A General Circulation Model for Deep Circulation on Gas Giants: Internal Convection, Solar Heating and Zonal Flows Part I: Axisymmetric Calculations [download]

MITgcm including the anelastic code: here


4. Jets in Geostrophic Turbulence

In geophysical systems due to the planetary rotation, long temporal scales and large special scales, turbulent processes cascade energy from small to large scales forming coherent large scale structures such as vortices and jets. Particularly we focus on formation of multiple zonal jets by baroclinic instability in geostrophic turbulence. Examples of such multiple zonal jets are found in earth�s oceans and in the atmospheres of the giant planets. The atmospheres of the giants have weak meridional temperature gradients and therefore have been assumed not to be controlled by baroclinic processes. However, we proposed that due to the deep geometry of these atmospheres, baroclinic instability of even a weak shear may play an important role in the generation and stability of the strong zonal jets. Using a two layer quasigeostrophic model we show that due to the deep circulation and the negative beta effect the dominant most unstable modes for weak baroclinic shears have a high modal meridional structure driven by the nonlinear interaction between the eddies. We use linear stability analysis, a nonlinear analytical model truncated to one growing mode and a full nonlinear pseudospectral model to show that multiple jets induced by the baroclinic instability can be dominant and stronger than the eddy field and thus forming stable multiple east-west zonal jets.

                            click figure for animation

The zonal velocity in the upper layer of a 2 layer quasigeostrophic, beginning with a random potential vorticity perturbation , through baroclinic instability (in this case with the fasting growing mode k=9, l=5), and an inverse energy cascade resulting in stable baroclinic multiple zonal jets.

more information: [animations]

Baroclinic Instability as a Source for Zonal Jets on Giant Gas Planets [download]

Poleward migration of eddy driven jets [download]


5. Convection in Rotating geophysical flows

Many geophysical systems are dominated by rotating convection. Examples include Earth�s mantle, giant planets and stars. This problem has a long history and has been often treated within the framework of the Boussinesq approximation and for Rayleigh-Benard convection type problems. We focus on anelastic systems, thus systems which contain a vertical variation in density. We characterize conditions under which the flow aligns with the axis of rotation , and forms convection columns. In the low Prandtl and Rayleigh number regime we show the transition from weak convection cases with symmetric spiraling columnar modes similar to those found in previous analytic linear theory, to more turbulent cases which exhibit similar, though less regular and solely cyclonic, convection columns which manifest on the surface in the form of waves embedded within the superrotation. We develop a mechanical understanding of this system and scaling laws by studying simpler configurations and the dependence on physical properties such as the rotation period, bottom boundary location and forcing structure.

         click figures for animation

Simulations showing the effect of rotation on convection. Looking at meridional slices (radius - latitude). Left simulation is rotating with a rotation period of 10 hours and the right one has a rotation period of 100 hours. Brighter colors are higher entropy. In the faster rotating case plumes align with the axis of rotation.

The magnitude of the superrotation as function of the rotation period. Low rotation periods have superrotation while high rotation periods are subrotating at the equator.