Binary Host Star Imaging
Complementary to my spectroscopic study of binary host stars is my interest in using high spatial resolution imaging to detect planets in multiple stellar systems and determine the influence of multiplicity on planet formation, migration, and evolution.
In 2017 and 2018, the California-Kepler Survey (CKS) used precisely measured stellar radii (combining high resolution spectroscopy and Gaia) and precisely measured planet radii from Kepler to show that the most common types of planets are relatively small, < 4 Rearth, but also seem to come in two flavors — those that are scaled-up versions of the Earth (super-Earths, about 1-1.6 Rearth), and those that are scaled-down versions of Neptune (about 2-3.5 Rearth). To be clear, since we lack planets of either of these sizes in the Solar System, we really do not know how similar or different their other properties are to Earth and Neptune. For example, all of the Kepler exoplanets in these studies orbit their stars within 100 days, so the stellar radiation could play a larger role in shaping their final radii and compositions. But this was still an incredibly interesting detection, one that had been predicted for a few years but not observationally confirmed before.
From Fulton et al. (2017). Completeness-corrected histogram of planet radii for planets with orbital periods shorter than 100 days. Uncertainties in the bin amplitudes are calculated using the suite of simulated surveys described in Section C. The light gray region of the histogram for radii smaller than 1.14 R suffers from low completeness. The histogram plotted in the dotted grey line is the same distribution of planet radii uncorrected for completeness. The median radius uncertainty is plotted in the upper right portion of the plot.
However, one thing that is challenging with detecting planets from photometry, especially far away planets like those from Kepler (most are many hundreds if not thousands of light years away), is potential contamination in the photometric aperture by additional stars. These stars could be bound companions to the primary, brighter star, or they could be unrelated background stars at a different distance. But if the extra light from another star is not accounted for, that can bias the inferred transiting planet radius — you think all the light in the aperture is coming from one star, and thus all being blocked by the same fraction during transit, but it could be that some of the light is not changing during transit if there is a second star that the planet is not transiting. Since planet radii are derived from the transit depth, (Rplanet/Rstar)^2, we need to know Rstar to know Rplanet!
In Teske et al. (2018), I investigated the fact of close companions, both detected and undetected, on the observed (raw count) exoplanet radius distribution, and demonstrated that the valley between the super-Earths and sub-Neptunes is fairly robust to undetected stellar companions, given that all of the systems in the CKS underwent some kind of vetting with high-resolution imaging. However, while the valley in the distribution was not erased or shifted, it was partially filled in after accounting for possible undetected stellar companions.
From Teske et al. (2018). This figure shows a comparison of three colored histograms with our results, as well as the histogram of observed exoplanet radii from the Fulton et al. (2017) sample of 900 Kepler Objects of Interest (KOIs) as a grey dashed line. The colored histograms represent different simulations of undetected stellar companions to the KOI host stars, assuming different probabilities that the primary versus secondary star hosts the planet (e.g., 90/10 = 90% change that the primary hosts the planet). Each colored histogram also accounts for effect of the actually-detected stellar companions (“real”).
This result was interesting because the position and depth of the radius valley can be matched with mass loss/evaporation models with different assumptions, one of which is the composition of the planet population. Owen & Wu (2017) and Jin & Mordasini (2018), each using slightly different evaporation/mass loss models, found that the radius distribution of Fulton et al. (2017) was well matched by models populated with planets having uniformly rocky cores, composed of a silicate-iron mixture similar to the Earth’s bulk density, and not by planets with cores having a substantial mass fraction (>~75%) of ice/water or made purely of iron. These authors, as well as Lopez & Fortney (2013), note that heterogeneity in the core composition would smear out the gap in the radius distribution.
By accounting for possible undetected companions, we observe a slight smearing out of the observed radius distribution gap (particularly in the oprob=90/10, which we think is the most realistic ratio). Our results suggests that, if there are undetected companions around the KOIs in the Fulton et al. (2017) sample, there could also be more heterogeneity in the core composition of most super-Earth and sub-Neptune planets than would be inferred from the original distribution. Specifically, a nonzero fraction of the cores could be composed of ice/water. Potential undetected companions complicate the origin story of these planets, as the addition of ice/water in the core opens up the possibility that they formed beyond the water ice line and migrated inwards, rather than only forming and migrating locally within the water ice line. Other factors not explored in my paper, like the relative importance of X-ray/UV flux over time as a function of stellar mass, could also contribute to the radius distribution being smeared out. This is one thing I’m hoping to investigate more with our MTS!
My paper also provided a cautionary tale: Even for targets at TESS-like distances (tens to hundreds of parsecs), vetting with high-resolution imaging is needed to infer the correct planet radius distribution.
From Teske et al. (2018). Similar figure to the histogram above, only this analysis assumes more TESS-like target distances, and no high-resolution image vetting. You can see how the pink “observed” histogram skews towards bigger planets than the grey “true” histogram.
I also published a comparison of high resolution optical spectroscopic versus high contrast imaging techniques for detecting close companions to Kepler objects of interest (Teske et al. 2015b), finding that the two techniques often do not overlap in the properties of companions they detect, suggesting that many KOIs may have more than one companion. This work was done as part of the Di fferential Speckle Survey Instrument (DSSI) team, the only group conducting high resolution, high contrast speckle "snapshot" observations (60 millisec) at large telescopes for Kepler-related follow-up (Horch et al. 2012, 2014; Everett et al. 2015; Howell et al. 2011).
Plots from Teske et al. (2015), created by UC Berkeley graduate student Lea Hirsch, showing the results of the imaging data analysis for KOI 5 in section 3.1 of our paper. Left: Primary KOI absolute photometry contours, and companion photometry contours, calculated from observed F692 magnitude and assuming it lies at the same distance and has the same age and metallicity as the KOI, mapped on the same (primary KOI) isochrone. The red point represents the absolute magnitude and "true" color for the companion (assuming it is bound), calculated from relative color information. The spread in color of the contours represents the spread in the normalized probability distribution, ranging from 1 (red) to 0 (dark blue). Middle: Same as left, but with companion photometry contours calculated from K magnitude. Right: A comparison of the overlap between the relative photometry contours of the companion. The red point here is the same as in the left and middle panels.
Plots from Teske et al. (2015), showing parameter comparisons for companions found through Kobl et al. (2015)'s analysis, and derived from our analysis of imaging data. THE TAKE AWAY IS there is not good agreement! The labels refer to the companion's primary KOI. Blue circles indicate KOIs with imaging data in multiple bands, and red symbols indicate the imaging-detected companion is >0.8" away from the primary KOI. The values plotted here for the companion to KOI 3471 assume a subigant primary star. Note that the "imaging" values are those derived from our analysis, and thus may be incorrect if the companion is unbound. Top left: Teff values of the companions. Top right: Flux ratios (companion/primary) of the companions. The flux ratios measured from imaging data are in the Kepler bandpass. Bottom left: Diff erence in derived Teff values, versus the separation as measured from imaging data (averaged over all detections). Dashed horizontal lines designate the separation limits reported in K15. Bottom right: Di fference in derived flux ratios of companions, versus the separation as measured from imaging data (averaged over all detections). Dashed horizontal lines designate the separation limits reported in Kobl et al. (2015).