Detecting New Planets via Radial Velocity
It is incredibly satisfying when different research projects overlap, and allow you to answer multiple science questions at once. I was lucky enough to have that happen in a project to detect planets via the radial velocity technique with the Planet Finder Spectrograph on Magellan. In Teske et al. (2016b), my collaborators and I showed that there are planets around two stars in the ``twin'' binary system HD 133131 A&B. This is a really cool finding because the stars are old (~9.5 Gyr), and metal-poor (I measure an [Fe/H]~ -0.30 for both), which means their protoplanetary disks were significantly different than that of our Sun (even though they're each about the same mass as the Sun) and that the planets we detected are much older than those in our Solar System. In particular, that we found *giant* planets around these stars is interesting, because the prevelance of giant planets is supposed to decrease with [Fe/H], at least in terms of giant *close-in* planets. Based only on what has been measured, the probability of HD 133131A/B hosting planets is < 1%...and we found them! But the planets we detected with PFS (and help from the other optical high resolution spectrograph on Magellan, MIKE) are long period -- HD 133131Bb has a period of almost 17 years (with a semi-major axis of a bit more than 6 AU), and HD 133131Ab and Ac have periods of 1.8 years (~1.4 AU) and almost 10 years (~4.5 AU), respectively. Especially with the Bb planet, we're starying to approach the separations at which directly image planets are found.
This cartoon shows the orbits of the planets around HD 133131A and B roughly to scale. Thanks to T.J. Rodigas for making it!
My differential abundance analysis of HD 133131A & B, showing a small but significant difference in the abundances of high-T_c elements is detected between these ``twin'' stars, which may be related to the formation and evolution of their giant planets. The relative abundances of HD 133131A/ are in orange circles/blue stars. Red lines connect multiple ionization states of the same species, or in the case of C, abundances derived from C I and CH features. A green dashed line shows zero difference; dotted lines show the weighted mean of elements, split at 1000 K.
In addition to detecting three new planets, I conducted a differential abundance analysis on the two host stars, which indicates a depletion in refractory elements in HD 133131A versus B (with standard errors of ~0.017 dex). In the paper I tried to not settle on one explanation for this trend, but explored the possible timings of rocky material accretion/depletion in the system, implementing a realistic estimate of giant planet metallicity (from my collaborators and friends, Thorngren et al. 2016). I think this is a good approach to take in the future when discussing ``twin'' star differences, because it lets the reader make up their own mind in this (butterfly-collecting) stage of studying "twin'" host stars. Plus it's fun to do the exploratory calculations!