Imaging and Characterizing Exoplanets with the Roman Coronagraph and Open-Source Tools

Jason Wang / Northwestern University – Evanston, PI

I propose to join the Roman Coronagraph team to assist in the target selection, data simulation, data reduction, instrument performance characterization, and science demonstration, with a focus on adapting current open-source tools to support the Roman Coronagraph. I am a lead developer of two open-source tools popular in the community: “pyKLIP” for stellar point spread function subtraction and exoplanet characterization and “orbitize!” for Bayesian orbit fitting of imaged exoplanets. I propose to add functionality to these packages so that they can be used to assist in the technology demonstration activities. This will ensure open and reproducible science and will mature the tools to analyze visible-light coronagraphic data from space.

In the area of target selection, “orbitize!” can be used and upgraded to predict the locations of known imaged planets and planets detected through indirect methods such as Gaia astrometry. This will allow us to know which objects will be in the Coronagraph field of view and which need ground-based follow-up to refine their orbits. In the area of data simulation and reduction, I will work on optimizing “pyKLIP” for Roman Coronagraph data. pyKLIP is widely used for coronagraphic data from ground-based telescopes and JWST and was the top performing algorithm in the Roman Exoplanet Imaging Data Challenge, detecting faint planets that other algorithms did not. Using simulated data, I will further improve pyKLIP such as by using orbital constraints as priors in the detection framework and by adapting it for the spectroscopic mode of the Coronagraph. In the area of instrument performance characterization, I have a decade of experience calibrating and characterizing high-contrast instruments such as the Gemini Planet Imager and the Keck Planet Imager and Characterizer, and will bring my expertise to the Roman Coronagraph. I will create tutorials and pipelines based on pyKLIP to easily assess the final planet sensitivity of the Roman Coronagraph in simulation, in the lab, and in space. I can also help develop astrometric and spectrophotometric calibration strategies for the instrument. In the area of science demonstration, I will create recipes and pipelines to measure the photometry, astrometry, orbits, and spectral properties of exoplanets with simulated and real data from the Coronagraph. For reflected light planets, orbital and photometric information are covariant, and I will add functionality to “orbitize!” to include constraining the phase function of reflected light planets in orbital fits. In all these cases, in addition to contributing to the Coronagraph Instrument’s technology demonstration, the software modifications would be accessible to the entire astronomical community in well-established software packages, and the same functionality can be used in future space missions such as the Habitable Worlds Observatory.