Research Page

Long Baseline Interferometry

An schematic showing the 6-telescopes of CHARA (http://www.chara.gsu.edu/). The light from each telescope is combined in the lab, where we collect data on interference patterns. Baselines up to 330 meters give <1 milli-arcsecond resolution in H-band.

I use data from long baseline interferometers to develop techniques for studying exoplanets in unique regimes. I use the MIRC-X instrument at the CHARA array, as well as the GRAVITY instrument at VLTI. The CHARA array consists of six 1-meter telescopes on Mount Wilson in California. I use four of the 1.5-meter telescopes at VLTI in Chile. 

By combining multiple telescopes with interferometry, one can resolve binary stars down to separations much closer than possible with single telescopes. By targeting these close binary stars with interferometers, we are able to measure differential astrometry (angular distance between the two stars) of order 10 micro-arcsecond precision (Gardner et al, 2018)! This is sufficient precision to detect the "wobble" of a star due to the gravitational pull from a unseen exoplanet or low mass companion.


How big is a micro-arcsecond? Measuring an orbit to 10 micro-arcsecond precision is similar to measuring the width of a human hair from a distance of 2,000 km. 

1. Fringes formed by the CHARA array - each vertical line is a different baseline. With binary stars, we see a fringe packet from each star. This gives us the distance between the stars, resulting in extremely precise differential astrometry. 







2. A planet and star both orbit their center of mass. A Jupiter-mass planet causes a solar-mass star 10 pc away to "wobble" around this center by ~10 micro-arcseconds. 








3. (Top) By targeting binary stars with interferometers we precisely measure the position of one star relative to the other across an orbit. (Bottom) When there are additional previously unseen companions in the system, we have enough precision to detect them as additional "wobbles" on top of the binary motion.

With Project ARMADA (ARrangement for Micro-Arcsecond Differential Astrometry) at CHARA and VLTI, we target binary stars to search for "wobbles" from previously unseen companions. Our goal is to constrain the companion frequency down to the planetary mass regime around stars more massive than the Sun. These intermediate mass stars are extremely difficult to probe with other methods for detecting exoplanets. This multi-year survey is currently underway, with first results published in Gardner et al (2021) and Gardner et al (2022, submitted).


4. If an exoplanet is orbiting an individual star of a binary system, we can detect the planet as an added "wobble" to the binary motion. Above is a demonstration of this method on a newly discovered triple star system. We plot the position of the secondary star relative to the primary fixed at the origin. There is also a third star around this secondary, which we can detect as an added "wobble" to the overall binary motion (Gardner et al, 2021). The "wobble" from a planet would look similar, but the motion would be smaller.

5. With our ARMADA survey, we have discovered the "wobbles" from many newly detected companions (Gardner et al, 2022, submitted). So far, these companions all have stellar mass (i.e., they are triple or quadruple star systems rather than planets). We show the inner "wobble" motions for some of these detections, with the long period outer binary motion subtracted out. Many of these "wobbles" have a size <1 milli-arcsecond, with residuals of 20-50 micro-arcseconds. The green orbits show our best fits to the data points (plotted as crosses with 1-sigma error ellipses). The light grey orbits show other solutions consistent with our data. Since we model the outer binary motion and inner "wobble" motion simultaneously, we need a lot of data to constrain the orbits of such small "wobbles".

6. Zhao et al (2011) used MIRC at CHARA to obtain strict upper limits on a hot Jupiter direct detection. The instrument has since been upgraded, justifying another attempt in progress now!

In Project PRIME (PRecision Interferometry with Mirc-x for Exoplanets) we use the MIRC-X instrument at the CHARA array to directly detect the light from non-transiting hot Jupiter planets. With precision closure phase methods, in theory we can directly detect the flux from the planet as one would with a high-contrast binary star. Such a detection would result in low resolution spectra at all phases of the planet's orbit. This information would be very helpful in constraining exoplanet circulation models. 


To make such a detection, we need to detect star/planet flux ratios of about 10,000:1. We are currenly near this limit (see plot to the left), and I am working on collecting and analyzing new data on hot Jupiter systems to improve these constrast limits. We have some promising tentative detections, with results expected to be published soon!

Transmission Spectroscopy

I also use the 6.5-meter Magellan Baade telescope at Las Campanas Observatory in Chile in order to study the atmospheres of transiting exoplanets. We use multi-slit spectroscopy in order to study the transit depth as a function of wavelength. This transmission spectra is used to detect features in the exoplanet atmospheres. 

I am involved with the Michigan Optical Planetary Spectra Survey (MOPSS) started by Erin May (May et al, 2018). The purpose of this survey is to study the atmospheres of "inflated Neptune" exoplanets. 

7. Results from May et al, 2018 (1st MOPSS paper) show a fairly flat transmission spectra for WASP-52b, implying a cloudy atmosphere. Our data can distinguish between clear and cloudy atmospheres, and detect spectral lines from chemical species in the atmospheres when clouds are not present. We have since obtained data on many more planets, with results coming soon.

Publications