There are Probably Many More Earth-Sized Worlds Than Previously Believed

In the past decade, the discovery of extrasolar planets has accelerated immensely. To date, 4,424 exoplanets have been confirmed in 3,280 star systems, with another 7,453 awaiting confirmation. So far, most of these planets have been gas giants, with about 66% being similar to Jupiter or Neptune, while another 30% have been giant rocky planets (aka. “Super-Earths). Only a small fraction of confirmed exoplanets (less than 4%) have been similar in size to Earth.

However, according to new research by astronomers working at NASA Ames Research Center, it is possible that Earth-sized exoplanets are more common than previously thought. As they indicated in a recent study, there could be twice as many rocky exoplanets in binary systems that are obscured by the glare of their parent stars. These findings could have drastic implications in the search for potentially habitable worlds since roughly half of all stars are binary systems.

For the sake of their study, the research team examined 517 exoplanet-hosting stars that were identified by NASA’s Transiting Exoplanet Survey Satellite (TESS) during its three years in operation. When compared to data from the twin telescopes of the international Gemini Observatory and the WIYN 3.5-meter Telescope at Kitt Peak National Observatory, they found that over 100 of these stars likely had a binary companion.

An artist’s rendition of the Transiting Exoplanet Survey Satellite (TESS). Credit: NASA’s Goddard Space Flight Center

The paper that describes their findings has been accepted for publication in the Astronomical Journal. Dr. Kathryn Lester, a postdoctoral researcher at NASA Ames Research Center, led the research effort with the assistance of colleagues from NASA Ames, the U.S. Naval Observatory, the NASA Exoplanet Science Institute, the NSF’s National Optical-Infrared Astronomy Research Laboratory (NOIRLab), the Lowell Observatory, as well as Georgia State and Standford University.

The Trouble with Transits

To date, the vast majority of confirmed exoplanets (roughly 75%) have been discovered using the Transit Method (aka. Transit Photometry). This consists of observing stars for periodic dips in their brightness, which can be the result of a planet passing in front of their face (transiting) relative to the observer. Like its predecessor, Kepler, TESS relies on the Transit Method to determine the presence of exoplanet systems around thousands of stars at any given time.

Unfortunately, binary companions have always been challenging when it comes to detecting transiting exoplanets. Transit Photometry requires that star systems being observed edge-on in order for exoplanets to be detected. But in binary star systems, where two stars orbit each other, dips in luminosity are a regular occurrence and are the result of one companion eclipsing the other.

As a result, it can be very difficult to spot smaller exoplanets that orbit closer to their stars, which is where astronomers expect to find rocky planets in the stars’ circumsolar Habitable Zone (HZ). Instead, using the Transit Method with binary star systems is likely to reveal only gas giants and/or planets that have distant orbits from their parent stars. This is why Dr. Lester and her colleagues set out to determine whether some of the exoplanet-hosting stars were in fact binaries.

Searching for Stellar Companions

The team relied on a technique called Speckle Imaging, where large numbers of short exposure images are combined and analyzed to greatly enhance the resolution of ground-based telescopes (similar to interferometry). Of the 517 TESS Objects of Interest (TOIs) they examined, they found that 73 exoplanet host stars that had previously appeared as a single point of light actually had a stellar companion.

They similarly found that 29 TOI stars that had produced false positives in the past also had stellar companions. Said Dr. Lester in a recent NOIRLab press release:

“With the Gemini Observatory’s 8.1-meter telescopes, we obtained extremely high-resolution images of exoplanet host stars and detected stellar companions at very small separations… Since roughly 50% of stars are in binary systems, we could be missing the discovery of — and the chance to study — a lot of Earth-like planets.”

The next step was to take the exoplanets that have been detected in these systems and comparing them to the sizes of exoplanets detected in single-star systems. From this, the team was able to demonstrate that while the TESS spacecraft was able to identify both Jupiter and Neptune-like (“large”) exoplanets and Super-Earths and Earth-like (“small”) exoplanets orbiting single stars, it found only large planets in binary systems. 

These results imply that there could be a population of Earth-sized exoplanets in binary systems that have gone undetected by missions like TESS, Kepler, and other exoplanet surveys that rely on Transit Photometry. For some time, scientists have suspected that transit surveys have been missing small planets in binary systems because of the potential for interference from a companion star.

However, this new study provides the first observational support for this suspicion while also showing what kind of exoplanets are affected. It is also significant because of the way the Transit Method has been seen as the most effective means of exoplanet detection to date – accounting for 3343 of the 4424 confirmed exoplanets. But if these results are correct, there could be up to 1600 rocky exoplanets that were missed with transit surveys.

This means that moving forward, astronomers will need to rely on a variety of observational techniques before they conclude that a binary system has no Earth-like planets. As Dr. Lester stated:

“Since roughly 50% of stars are in binary systems, we could be missing the discovery of — and the chance to study — a lot of Earth-like planets. Astronomers need to know whether a star is single or binary before they claim that no small planets exist in that system. If it’s single, then you could say that no small planets exist. But if the host is in a binary, you wouldn’t know whether a small planet is hidden by the companion star or does not exist at all. You would need more observations with a different technique to figure that out.”

“We have shown that it is more difficult to find Earth-sized planets in binary systems because small planets get lost in the glare of their two parent stars,” added Dr. Steve Howell, the leader of the speckle imaging effort at NASA Ames and a co-author on the paper. “Their transits are ‘filled in’ by the light from the companion star. This is a major finding in exoplanet work. The results will help theorists create their models for how planets form and evolve in double-star systems.”

An illustration of the Kepler-47 circumbinary planet system. Credit: NASA/JPL Caltech/T. Pyle

Another aspect of the study involved Dr. Lester and her colleagues analyzing the distance between binary companions in systems where TESS detected large planets. What they found was that exoplanet-hosting pairs were typically farther apart than binary pairs that have no known exoplanets. This could be interpreted as an indication that planets do not form around stars that have close stellar companions.

In the future, this could be used to place added constraints on where astronomers should look for rocky planets. The speckle imaging survey conducted by Dr. Lester and her team also illustrates how exoplanet studies are transitioning from exoplanet discovery to characterization. In addition to characterizing exoplanet atmospheres and surface environments, there’s also the vital task of characterizing planetary systems.

By knowing which types of stars are most likely to support rocky exoplanets, astronomers and astrobiologists can narrow the search for planets that are most amenable to “life as we know it.”

Further Reading: NOIR Lab