Precise Radial Velocities in the Near Infrared

Open Access
Redman, Stephen Lewis
Graduate Program:
Astronomy and Astrophysics
Doctor of Philosophy
Document Type:
Date of Defense:
March 04, 2011
Committee Members:
  • Dr Larry Ramsey, Dissertation Advisor
  • Lawrence William Ramsey, Committee Chair
  • James Kasting, Committee Member
  • Kevin Luhman, Committee Member
  • Mercedes Richards, Committee Member
  • Steinn Sigurdsson, Committee Member
  • Alexander Wolszczan, Committee Member
  • Jason Wright, Committee Member
  • instrumentation
  • spectrograph
  • infrared array
  • exoplanets
  • near infrared
  • uranium
Since the first detection of a planet outside our Solar System by citet{1992Natur.355..145W}, over $500$ exoplanets have been found to datefootnote{url{}, 2010-12-01}, none of which resemble the Earth. Most of these planets were discovered by measuring the radial velocity (hereafter, RV) of the host star, which wobbles under the gravitational influence of any existing planetary companions. However, this method has yet to achieve the sub-m/s precision necessary to detect an Earth-mass planet in the Habitable Zone (the region around a star that can support liquid water; hereafter, HZ) citep{1993Icar..101..108K} around a Solar-type star. Even though Kepler citep{2010Sci...327..977B} has announced several Earth-sized HZ candidates, these targets will be exceptionally difficult to confirm with current astrophysical spectrographs citep{2011arXiv1102.0541B}. The fastest way to discover and confirm potentially-habitable Earth-mass planets is to observe stars with lower masses - in particular, late M dwarfs. While M dwarfs are readily abundant, comprising some $70\%$ of the local stellar population, their low optical luminosity presents a formidable challenge to current optical RV instruments. By observing in the near-infrared (hereafter, NIR), where the flux from M dwarfs peaks, we can potentially reach low RV precisions with significantly less telescope time than would be required by a comparable optical instrument. However, NIR precision RV measurements are a relatively new idea and replete with challenges: IR arrays, unlike CCDs, are sensitive to the thermal background; modal noise is a bigger issue in the NIR than in the optical; and the NIR currently lacks the calibration sources like the very successful thorium-argon (hereafter, ThAr) hollow-cathode lamp and Iodine gas cell of the optical. The PSU Pathfinder (hereafter, Pathfinder) was designed to explore these technical issues with the intention of mitigating these problems for future NIR high-resolution spectrographs, such as the Habitable-Zone Planet Finder (HZPF) citep{mahadevan2010habitable}, and forms the core of my dissertation. I have investigated and quantified several aspects of making precision radial velocity measurements in the NIR using Pathfinder. Between $2006$ and $2008$, I made precise measurements of the Earth's rotational velocity with respect to the solar spectrum, with which we were able to achieve precisions of less than $10$ m/s. In late $2008$ and $2009$, I worked on optimizing the spectrograph and reduction code in preparation for our first on-sky tests. I also began characterizing a new calibration source for the NIR, the emission spectrum of a uranium-neon hollow-cathode lamp. During $2010$, Pathfinder saw first light at the Hobby-Eberly Telescope (hereafter, HET), where we observed almost a dozen radial velocity standard stars and bright planet-hosting stars. Using uranium-neon as a calibration source, we were able to achieve a precision of $20$ m/s in the Y band. In collaboration with Colorado University and the National Institute for Standards and Technology (NIST), we fed Pathfinder with a laser frequency comb, and were able to achieve precisions of less than $5$ m/s in the H band. These are some of the highest-precision radial velocity measurements in the Y and H bands to date, and represent an enormous advancement in our ability to make precision measurements in the NIR.