At this point, we’ve spotted thousands of planets orbiting other stars. But we still don’t have a good picture of whether exosolar systems are similar to our own Solar System. Things like asteroids and comets are far too small to resolve given our current technologies. But there is a chance we could detect the presence of a major feature of our Solar System elsewhere: exomoons.
But a group of researchers have suggested there may be an exomoon orbiting a planet discovered by the Kepler mission, and the researchers’ data was compelling enough to get them time on the Hubble Space Telescope. But the additional data both makes the case for the moon better and worse (don’t worry, we will explain that). And the model that fits the data best involves a Neptune-sized moon orbiting a super-Jupiter.
Timing is everything
There are two ways to potentially identify an exomoon. The first involves the amount of light blocked while the exoplanet orbits its host star. When the moon isn’t eclipsing or being eclipsed by its planet, it should block a small bit of additional light during the transit. Look at enough transits, and there could be a regular pattern of additional light blocked. But this method relies on the moon being large enough to block a significant amount of light, which is something that’s far from guaranteed.
The alternative is what are called transit-timing variations. These are generated because, as the moon orbits its planet, it exerts a gravitational tug on it. In some configurations, that tug will be in the direction of the planet’s orbit around its star, causing the transit to occur sooner than expected. During other orbits, the moon will trail behind its planet, slowing it down and making the transit start later than expected. The challenge with these is that other planets in the same exosolar system can also cause transit-timing variations. Knowing when you’re looking at a moon and not a planet is tricky.
So, in theory, detecting an exomoon is difficult but not impossible. For practical reasons, however, we probably shouldn’t expect to.
That’s because most of the exoplanets we know about orbit close to their host stars. This comes about because of the methods we use to detect them. Methods based on gravitational interactions between the planet and star have more to work with when the two are closer together. And those that follow a planet’s passages in front of its star have more of these passages to work with when planets are closer and their orbits are shorter.
The problem is that plant-moon-star setups aren’t gravitationally stable when they’re too close. In most cases, the moon ends up getting ejected from the system. So it would be surprising if we actually did see an exomoon in most of the systems we’ve studied so far.
Is that a moon?
One of the researchers behind the work, Columbia’s Alex Teachey, was involved in a previous study that looked for exomoons despite the long odds (Teachey is joined by David Kipping for this new paper). Out of 284 planets they looked at, only one called Kepler-1625b seemed to show transit-timing variations that looked consistent with the presence of an exomoon. Kepler-1625b is a Jupiter-sized planet that orbits far enough away from its star—about the same distance Earth is from ours—that a moon could be stable. And, with a year between orbits, the researchers had enough time to arrange for the Hubble Space Telescope to image the planet’s next pass in front of its star.
With that data in hand, they first repeated their former analysis. And here things got somewhat odd. In the intervening time, the Kepler data has been cleaned up to control for some minor biases in its instruments. Repeating the initial analysis with the updated data actually weakens the case for an exomoon at Kepler-1625b.
But the Hubble data doesn’t. The planet started its passage in front of the star over 20 minutes earlier than would be expected. And, based on their earlier data, the researchers would expect the moon to be trailing Kepler-1625b at this point in its orbit. And Hubble shows an additional dimming later in the planet’s transit. Although the presence of an additional planet could explain the transit-timing variations, it would not explain this dimming.
The authors then took all their data and tried to find a model that was consistent with it. The results they got, in their own words, were “jarring.” The moon would have to have a mass and radius that made it similar in size to Neptune, one of the larger planets in our Solar System. It would also orbit roughly 45° outside the plane defined by the planet’s orbit and at an unusually large distance from the planet. The planet itself would be roughly three times the mass of Jupiter. None of these facts seems normal.
Quite reasonably, the authors conclude, “All in all, it is difficult to assign a precise probability to the reality of Kepler-1625b-i [the moon].” We don’t have evidence that exomoons are common, so the authors recognize any claims regarding the existence of exomoons should bet met with skepticism. And the tenuous nature of the supporting data, combined with the bizarre properties of the system, give us reasons to take that skepticism seriously, even if a moon is still the best explanation for them.
Faced with that skepticism, Teachey and Kipping are undoubtedly applying for more Hubble time during its next transit.