Astrometry and Images and Spectra, Oh My! A New Nearby Super-Jupiter

Title: Direct Imaging and Astrometric Detection of a Gas Giant Planet Orbiting an Accelerating Star

Authors: Thayne Currie, G. Mirek Brandt, Timothy D. Brandt, et al.

First Author’s Institution: University of Texas-San Antonio

Status: Accepted for publication in Science (Closed Access). Available on arXiv (

Most of the 5000+ known exoplanets were discovered with indirect methods, that is, they were discovered by observing the subtle influences they have on their host stars. The transit method and the radial velocity method are the two standout successes, but direct detections of exoplanets are becoming more common as new and more powerful instruments come online. Today’s paper reports the discovery of the directly imaged planet HIP 99770 b, only 40 parsecs away and discovered by the Subaru Telescope, aided by the Hipparcos and Gaia space missions.

An Astrometric Waltz

Contrary to the concept of the celestial spheres, the stars we see are not actually fixed in space and time. With careful measurements and a good deal of patience, we can actually see stars moving across the sky (with their so-called proper motion). The technique of precisely measuring stellar positions is called astrometry, and by tracking this data over long periods of time, we can not only identify the bulk motions of stars, but also any higher-order motion, such as a slight wobble due to the orbit of a planet. Observed for long enough, rather than tracing a straight path across the sky, the astrometric signal of a planet-hosting system looks rather like the path of a pair of dancers waltzing, with subtle loops as the planet orbits its star.

Figure 1: (Left) Dancers in various waltz positions, from Thomas Wilson’s Correct Method of German and French Waltzing (1816). (Right) The simulated astrometric signal of a sun-like star with a large planet. (ESA)

HIP 99770 was identified by the Hipparcos and Gaia missions to have an astrometric signal slightly differing from a straight line path. While this extra acceleration did not show an entire planetary orbit, it was enough to motivate further study to try and find out whether this acceleration was due to a large planet in the system.

Subaru, Not Just For Rock Climbers

In order to do this, the authors used the Subaru Telescope, operated by the National Astronomical Observatory of Japan, and its high-contrast imaging/spectroscopic instrument SCExAO/CHARIS. Four images of HIP 99770 taken over the course of a year show a faint point source (the planet HIP 99770 b) comoving with the star. By combining the direct images of the planet (which show its position around the star) with the historical astrometric data (showing the motion of the star), the team was able to model the orbit of the planet, finding an orbital semimajor axis of 16.9 AU, a planet mass of 16 Jupiter masses, and an orbital eccentricity of 0.25. 

Figure 2: (Left) Two direct images of HIP 99770 b. (Right) The orbital analysis of the planet including the astrometric data. Figures 1 and 2 in the paper.

So far so good, but that mass measurement puts this planet right up against the deuterium-burning limit, a rough mass boundary separating planets and brown dwarfs. Who’s to say this is really a planet, and not a brown dwarf?

Some Like It Hot

Mass alone isn’t enough to distinguish large planets from small brown dwarfs. Instead of drawing a line at a particular mass, sometimes it’s better to look at the mass ratio between an object and its host star. Here, HD 99770 b’s mass ratio is smaller than all other comparable unambiguous brown dwarfs, and much more similar to other comparable exoplanets. In addition, orbital architectures between star/brown dwarf binaries tend to differ from planetary systems, and HD 99770 b’s orbit is fairly circular, while most other brown dwarfs have significantly eccentric orbits.

Spectra help here, too, and with CHARIS the authors obtained a low resolution spectrum of HIP 99770 b between 1.2 and 2.4 micron. Compared to template objects like the exoplanet HR 8799 d and a known L-type brown dwarf, HIP 99770 b appears to be very much like other planets in the L-T transition region. While L-type objects have thick clouds and little visible methane, T-type objects have clear atmospheres and abundant methane absorption. The best-fitting models find a temperature between 1300 and 1500K, with relatively thin cloud cover. Given the planet’s measured luminosity, it likely formed from the direct collapse of gas and dust in its protoplanetary disk, predicting an age between 80 and 200 million years, consistent with estimates of the age of its host star.

Figure 3: (Left) Observed spectrum of HIP 99770 b and template spectra. (Right) The landscape of RV/DI exoplanets and brown dwarfs, showing HIP 99770 b is less like the unambiguous brown dwarfs and more like the confirmed planets. Figures 3 and 4 in the paper.

While previous direct imaging discoveries have leveraged indirect detection methods like radial velocities, HIP 99770 b is the first true exoplanet to go from astrometric hint to directly imaged planet, proving this powerful combination for discovery and characterization can succeed. In the next few years, as the full epochal astrometric Gaia data are released, observatories like Subaru will be critical in confirming and characterizing the hundreds and thousands of new candidate exoplanets Gaia will identify.

Astrobite edited by William Balmer

Featured image credit: ESA, NASA/JPL-Caltech, and Robert Linsdell

About Yoni Brande

I'm a third year PhD candidate at the University of Kansas, working on exoplanet discovery and characterization. I primarily work with TESS transit data and Hubble Space Telescope exoplanet transmission spectroscopy data, and I'm also interested in enabling more collaborative science with open source astronomical software tools. When I'm not doing research or writing astrobites, I can be found in a sci-fi streaming binge, running, lifting, cooking, or on Twitter @YoniAstro.

Leave a Reply