My research interest is planetary atmospheres. Many meteorologists study only Earth's atmosphere, but I am fascinated by the vastly different weather and climate conditions that occur on all planets. I believe that in order to fully understand the physics that governs our own atmosphere, we must study the atmospheres of all planets. It is analogous to the study of the human body--biologists don't just study people to learn how our bodies work, the study all life forms. Studying a variety of atmospheres leads me to have a deep understanding of how fundamental processes change the atmospheric circulation and structure. Specifically, I use numerical modeling techniques to investigate the atmospheres of Pluto, Triton, Venus, Mars, Titan, and super-Earth exoplanets.
My educational background is in terrestrial atmospheric science and stellar occultations. I am a computational theorist and most of my work uses a tool called "general circulation models." The subject of my research is the large-scale circulation and structure of atmospheres of small (i.e. of order Earth sized) bodies.
Pluto — Stellar occultation light curves are one of the primary sources of information about Pluto's atmosphere, beginning with the first observed stellar occultation in 1988. We must decode the information contained in the light curves to determine the temperature, density, and winds globally in the atmosphere. Several models predict temperature and density under various assumptions. One of these is by Strobel et al. (1996) that includes heating and cooling by CH4 and CO given mixing ratios of these species, surface temperature, and surface pressure. I used the model to calculate a temperature profile (temperature as a function of height) and the resulting light curves. I found that the model light curves were most sensitive to surface pressure and CH4 mixing ratio. The model light curves were compared with the data for the years 1988, 2002, 2006, and 2008 to determine the CH4 mixing ratio and surface pressure on Pluto in these years. I followed up with this work by using an updated model that contained a troposphere (with troposphere height as a new parameter). We found the no-troposphere solutions to be in better agreement with the data. <br/>
General circulation models (GCMs) are global models of the atmosphere that predict the temperature, winds, surface pressure, and other parameters. They are especially important in the field of planetary atmospheres, where wind data is very sparse due to the difficulty in measuring it remotely. The winds on Pluto are completely unknown, so I haved developed a GCM for Pluto to predict them. I compare the temperature output from the GCM (based on the dynamical core of the MIT GCM) to stellar occultation light curves to validate the model. Frost-free versions of the model in 2D and 3D have aleady been analyzed. Adding a condensation cycle and multilayer subsurface are works in progress.Triton, Pluto's "sister world" is included in the 3D work. On the eve of New Horizons' flyby of Pluto, is it important to study both Pluto and Triton in tandem since they are so similar. What was learned about Triton during the 1989 Voyager 2 flyby informs us about Pluto, and the New Horizons flyby will inform us as to the state of Triton's atmosphere two decades later.
Venus— Like other planetary atmospheres, Venus' thermosphere is thought to have breaking gravity waves that deposit momentum and produce accelerations of the flow. I have been working to put a gravity wave drag parametrization into the Univeristy of Michigan Venus thermosphere general circulation model. This parameterization replaces the Rayleigh friction scheme, which was tuned to produce a result to match observations. However, the Rayleigh friction scheme does not describe the underlying physical mechanism to the momentum drag, whereas the gravity wave drag scheme does.
Mars — The zonally (i.e. longitudinally) averaged topography of Mars slopes downwards from the south pole to the north pole, and the average difference between the elevation of the southern and northern hemispheres is about 5 km. This slope has an effect on the heating of the atmosphere, where, at a given altitude aloft, the temperature over an elevated surface is warmer than the temperature over a lower surface. The slope induces a temperature gradient that modifies the Hadley cells. At equinox, the northern hemisphere cell is stronger than the southern hemisphere cell, and the latitude of the boundary between the two cells moves southward from the equator into the southern hemisphere. In the absence of topography we would expect symmetric cells because of the symmetric solar forcing at the equator, but the slope moves the peak heating southward. This shift in peak heating mimics a change of season, for which current Hadley cell theories already exist. To investigate this effect, I use a Martian version of the MIT atmospheric general circulation model. I also have a model based on that of Lindzen and Hou (1988), but modified to include topography, that solves for the poleward extent of both cells and the latitude of the boundary between them. Both of these models are forced by Newtonian relaxation to a prescribed radiative equilibrium state. In the first state ("pure" radiative equilibrium), the heating in the atmosphere does not depend on the height of the surface. In the second state (radiative-convective equilibrium), the heating depends on the height of the surface via convection. When the convective forcing is present, the effect of the sloping topography is reproduced in both the general circulation model and the modified Lindzen and Hou (1988) model. When convective forcing is not present, the slope has little effect on the circulation.
I am interested in continuing this work by adding the effect of atmospheric dust in the models, which will change the atmospheric heating and potentially the location and strength of the Hadley cells.
I am also interested in characterizing longitudinal variations in the meridional circulation, as the Hadley cells are not likely to be zonally uniform due to modulation from toporaphy.
Exoplanets — Planets around stars other than our Sun are quickly being discovered at smaller and smaller sizes. Currently, the smallest planets are "super-Earths", which are planets that are a few times larger than Earth, but potentially with a very different climate because they are tidally locked and very close to their parent star. I am using a general circulation model to determine the atmospheric circulation and temperatures of these super-Earth-type planets.
Titan — A stellar occultation occurs when a planet or moon passes in front of a star. By measuring the intensity of the starlight as it passes through the planet or moon's atmosphere, we can learn things about the vertical atmospheric structure such as temperature, pressure, and density as a function of altitude. This technique is very useful for studying the atmospheres of other planets and moons, since we can do it from Earth and don't have to use a spacecraft.The occultation we observed was of the star TYC 1343-1865-1 and occurred on 14 Nov 2003. Our data is superb in that we have light curves at three wavelengths (350, 480, and 770 nm) and high time resolution (0.033 s). By inverting the light curves we were able to obtain vertical temperature profiles of the equatorial mesosphere. These can only be lower limits on temperature, since our method assumes no atmospheric extinction, which is a very poor assumption since Titan has an optically thick layer of hydrocarbon haze shrouding its surface. Instead, we modeled the distribution of haze with height by comparing the temperature measured by spacecraft with our measured temperatures.
Because of the high time resolution of this data, we observed several scintillation spikes in the light curves. When comparing the three wavelengths we observed at, the spikes are offset in time, always in the same order. Had there been a terrestrial cloud or a glitch in the detector, the spikes would have occurred at the same time. The spikes are caused by density fluctuations in Titan's atmosphere, which refract the light by slightly different amounts, such that the wavelengths refracted the most take longer to arrive. By using the delay between the spikes, we are able to derive a temperature profile that is free from extinction effects.
We observed a central flash, which occurs when the center of the occulting body passes in front of the star and acts like a lens. Conventional theory predicts that a circular shaped planet produces a single peak, and an oblate planet produces two or four peaks. Imagine our surprise when our light curve had a three-peaked flash! When we applied a model for an irregularly shaped Titan, derived from a 1989 occultation (of which none of the data contained a three-peaked flash), to the parameters of our occultation, we achieved a three-peaked flash.
I am currently in the process of developing a Titan GCM. This work can be useful in interpreting stellar occultation data and Cassini/Huygens data.