New Candidates from K2 from Planet Hunters and Their Nearest Neighbors

By Yale grad student, Joey Schmitt

In the 10th paper(!) from the Planet Hunters citizen science program, a stupendously great number, we independently discovered 10 new planet candidates in the K2 *Kepler* data (Campaigns 1 and 3). However, simply discovering them was not the main goal of the new paper. We wanted to explore their neighborhoods.

The environment in which a star is created has a large and enduring impact on how planets form. Under standard planet formation theory, when a star collapses, it forms a disk, called a protoplanetary disk, due to the conservation of angular momentum. It is in this disk of material orbiting the infant star that planets are formed. Solid material clumps together and forms planets. In the inner disk, the material is hotter, so the only solid material is metallic or rocky. In the outer disk, the material is cooler, which allows molecules like ice and frozen ammonia to clump together as well. This extra solid mass in the outer solar system allows the outer planet to grow bigger and eventually capture gas. Interactions between all these planets can then jumble them around.

However, most stars are not born alone. They more often come in pairs or triplets or even larger clusters. If two stars are forming too close together, each star could disrupt or even completely destroy the other’s
protoplanetary disk, making one or both stars devoid of planets. Conversely, it’s at least hypothetically possible that, at certain distances, a star could funnel its protoplanetary disk material into the protoplanetary disk of a neighboring star, giving the star more material to make planets out of. Current research has suggested that the destructive effect dominates. We aimed to test this suggestion and to further examine the potential effects of stellar neighbors to planetary systems. There are similarly interesting questions exploring the effect of a third star in eclipsing binary (EB) systems.

In this paper, we made a selection of many planet candidates, several from Planet Hunters and several others from previously published journal articles, and also many EB candidates, all of which were discovered through Planet Hunters volunteers, for a total of 75 targets. In order to find nearby stellar companions to these planet or EB systems, one has to take very high resolution images. Typically, this is impossible due to atmospheric turbulence blurring the starlight (seeing). To get around this,
we used two telescopes, SOAR in Chile and Keck in Hawaii, that get around this problem. The SOAR telescope uses speckle imaging, which takes hundreds images so quickly that the air doesn’t have time to move around and blur the image and then combines them. The Keck telescope, on the other hand, uses lasers to measure the air turbulence and then deforms its mirrors many times per second to correct the light before it reaches the camera.

With these techniques, we were able to find three stellar companions to our planet-host stars and six companions to our EBs. While we did not have a large enough set of targets to definitively measure the overall effect of nearby neighbors on planetary and EB formation, the results were suggestive
of two things. First, we found just one very close companion to a planet-host, strengthening the hypothesis that nearby stars are in fact destructive to planet formation. Secondly, we discovered several new stars
near very short-period EBs, implying that the shortest period EBs necessarily need a third star in the system. The third star steals energy from the close pair, which pushes those two stars on a shorter and shorter orbit.

The six companion stars found by the SOAR telescope are shown in the image below:

In the meantime, we are continuing to show data from the original *Kepler* set of stars. This current project will allow us to calculate the frequency of planets in long-period orbits around *Kepler *stars, something
that no other research project is yet capable of doing. An integral part of this is displaying synthetic (or “fake”) planets in the data. The synthetic transits allow us to measure how good Planet Hunters are at
finding planets of different sizes and periods around different kinds of stars. This knowledge is *required* to know how frequent planets occur because it allows us to correct for the planets that are there but *not*
found.

We would like to thank everyone involved in this program! The volunteers here at Planet Hunters are simply wonderful. This is one of the most popular *and scientifically productive* of the Zooniverse projects. We’re
also looking into if and how we can reincorporate K2 data and, in the future, TESS data. We hope that you continue to contribute to astronomical research.

The (not *quite* final) public version of this paper is here.
.