Hot Jupiter Atmospheres

While hot Jupiters were the first exoplanets to be discovered around main-sequence stars more than 25 years ago, their origins and evolution remain largely unknown. While demographic surveys can help us understand the population as a whole, they offer little insight into individual planets.
I work on understanding individual hot Jupiters by using high-resolution near-infrared spectroscopy to study their atmospheres. Hot Jupiters are too close to their host stars for planet and stellar light to be separated, but over a few hours, spectral features associated with a planet will shift in projected velocity by tens of kilometers per second due to the planet's orbital motion, while spectral features of the star will be mostly fixed. This allows planet features to be isolated by cross-correlating an observed high-resolution time series with a planet template. This technique (known as high-resolution cross-correlation spectroscopy or HRCCS) works for both transiting and non-transiting planets,and can offer detailed
insight into an exoplanet's atmosphere. HRCCS breaks the mass-inclination degeneracy of stellar RV techniques, and also enables the identification of molecules, measurements of species abundances, and constraints on the wind speeds. Using a cross-correlation to log-likelhood mapping and nested sampling, I perform atmospheric retrievals to measure chemical abundances in hot Jupiter atmospheres. The ratios of heavy elements in a planet's atmosphere, in particular carbon, oxygen, and sulfur, are sensitive to where and how a planet formed, and will improve our understanding of the origins and evolution of hot Jupiters.
Since moving to UCLA, I've been working on HRCCS techniques with the Keck/KPIC instrument. While KPIC offers a lower throughput than a seeing-limited spectrograph, the stability improvements from the use of single-mode fibers is a major advantage for HRCCS. From initial on-sky results, we estimate as many as 40 hot Jupiters can be detected and characterized with KPIC HRCCS. We have begun a survey to characterize 20-30 of these planets and place constraints on the underlying distributions of atmospheric metallicity and C/O ratio in this population. Observations are ongoing, with a planned completion date in early 2025.
Our first target in this program was the ultra-hot Jupiter WASP-33 b, and the analysis was published in Finnerty et al. 2023. We confirmed the presence of a dayside thermal inversion and found abundances consistent with a 2-15x solar atmospheric metallicity and C/O ratio of ~0.8. This combination is at odds with predictions that metallicity and C/O ratio should be inversely correlated, and suggests WASP-33 b somehow accreted material enriched in both solids and carbon during its formation.

We next analyzed our observations of HD 189733 b, published in Finnerty et al. 2024. Our results are consistent with a 3-5x solar metallicity atmosphere with C/O = 0.3+/-0.1. We also placed upper limits on the abundances of methane and ammonia in the planet's atmosphere. This composition is consistent with core accretion models of planet formation and significant late accretion of solid material during the formation of HD 189733 b.

Improvements to our retrieval pipeline are ongoing. Specific areas planned for our next paper include improved treatment of the KPIC line-spread function (LSF) in order to measure vsini and changes to our free retrieval model to account for vertical variations in molecular abundances. I am also working on commissioning KPIC's L-band capability, which will enable constraints on additional molecular species and improve constraints on the pressure-temperature profile when combined with K-band observations.
Picture
Figure 3 from Finnerty et al. 2024, showing the retrieved pressure-temperature profile, emission contribution function, and spectrum for HD 189733 b. The bottom panel shows the opacities for several molecular species in the observed wavelength range.

Fiber-Fed Infrared Spectrographs

I'm also a member of the KPIC team. KPIC is a series of upgrades to Keck II Adaptive Optics and Keck II/NIRSPEC in order to enable diffraction-limited, high-resolution spectroscopy with single mode fibers in the K and L bands. Phase I finished in November 2021, Phase II was deployed in March 2022, and Phase III will be deployed in spring 2024. During phase I, I developed a set of scripts to enable reliable and repeatable calibration of the fiber injection and extraction units before observing. For phase II, I modified these scripts to perform NCPA compensation, and development is ongoing. Optimizations to these scripts are ongoing, with several improvements planned for the phase III upgrade in preparation for KPIC facilitization. My work on KPIC is described in several SPIE papers.

Dusty Galaxies

During my post-bacc, I also took a brief break from exoplanets to work on dusty galaxies at z ~ 2-3, specifically hot, dust-obscured galaxies (Hot DOGs). These sources were selected based on their colors in the WISE survey, and include the most luminous known galaxies. Working with Tom Soifer, Kirsten Larson, and Lee Armus, I reduced and analyzed Keck/NIRES observations of 24 Hot DOGs, which we published in 2020. The plot to the right shows the full NIRES spectra of 8 hot DOGs from that paper, with prominent emission features marked. We found Hot DOGs commonly host fast, massive ionized outflows, and commonly show Eddington ratios greater than unity, in addition to hosting ongoing star formation. The simultaneous presence of these phenomena suggest Hot DOGs may be a short-lived phase in galaxy evolution as feedback from the central black hole quenches star formation.

At the end of my gap year, I also led the initial port of the CAFE fitting tool, intended for fitting continuum and line features in Spitzer spectra of luminous infrared galaxies, from IDL into python. This project is part of JWST ERS program 1328, which will release a version of the software optimized for fitting IFU cubes.
Picture
Full NIRES spectra of 8 Hot DOGs, excerpted from Finnerty et al. 2020.