The search for distant moons

In 1655, Christiaan Huygens pointed a self-constructed telescope at Saturn. The Dutch astronomer wanted to verify his assumption that the planet was surrounded by a single fixed ring whose orientation changed over the years. In the process, Huygens discovered the giant moon, Titan. Henceforth, Saturn was the third planet, along with Earth and Jupiter, of which a satellite was known. Today we know: Moons are even more common in our solar system than planets.

But is this equally true for the rest of the universe? In 2007, a network of automated telescopes observed a star about 433 light-years away in the constellation Centaurus. Its brightness plummeted noticeably for 54 days, then increased again. The cause was a giant gas planet surrounded by 37 rings. And like Saturn, this world called J1407b has a gap in its ring system. In it, a moon could move, which has approximately the mass of the earth.

Whether this is really so is unclear. If not there, there should be moons in the orbit of other exoplanets. Since the first discoveries in the 1990s, astronomers have now tracked down more than 4,000 planets around distant stars, thanks in particular to the Kepler Space Telescope, which was launched in 2009 and operated until 2018. Exoplanets seem to exist everywhere and in almost every imaginable size.

AT A GLANCE

INCONSPICUOUS COMPANIONS

• Moons are ubiquitous in the solar system. Planets in the vicinity of other stars are also likely to have satellites, except that none has yet been clearly identified.
• Our moon has probably made conditions on Earth more conducive to life. Accordingly, exomoons could create better conditions for extraterrestrial life - or even harbor it.
• Experts hope to soon detect the first of the relatively small celestial bodies. However, this will probably only be possible with the next generation of powerful telescopes.

Moons as planet stabilizers

Speculation about possible satellites began in the early 2000s. In the meantime, there are several candidates, but no evidence yet. Experts expect nothing less from discovered exomoons than a readjustment of our cosmic perspective. From moons in other star systems, we could find out how large such celestial bodies usually are and how they formed. From this, conclusions could be drawn as to whether the solar system is more ordinary or exotic - and what role the moon has played during terrestrial evolution.

To date, for example, our home planet is the only known one with plate tectonics. In addition, Earth has an atmosphere thick enough to keep water liquid, combined with a mild climate stable over eons. All of these conditions were important in the evolution of life and can be attributed, at least in part, to the influence of our satellite.

It entered the solar system in its infancy, 4.5 billion years ago. At that time, a celestial body the size of today's Mars probably collided with the precursor of the Earth. The violent crash left a glowing, elongated lump in its place and hurled a seething molten rock into space, from which the moon would soon form.

In the billions of years that followed, our satellite cooled and gradually moved away from the Earth. The Earth in turn became rounder and rounder the further the moon retreated. Its crust stretched under the resulting tidal forces - possibly the beginning of plate tectonics. The moon's migration also slowed the Earth's rotation, lengthening our day by almost two milliseconds per century.

Large moons with water could themselves be conducive to life

The moon's influence is related to its gravity: It has 1.2 percent of Earth's mass. Trabants of other planets are much lighter relative to their parent bodies. Titan, for example, although nearly 50 percent larger than Earth's moon, has only 0.02 percent of Saturn's mass. The moon's gravitational pull keeps Earth's axis at a constant 23.5-degree tilt to the sun. This configuration protects our climate from too rapid changes. The situation is quite different on Mars, for example, which has only two small moons. Its axis fluctuates between zero and 60 degrees every few million years, causing dramatic climatic disturbances.

If the terrestrial past is any guide, moons should provide more stable environmental conditions for exoplanets as well. And even if there is no life on the planets, possibly the exomoons are suitable habitats in their turn. Under the frozen shell of icy moons like Europa or Enceladus, experts have long suspected favorable conditions for simple life forms. Such scenarios are even conceivable for Titan. Saturn's moon is covered with seas of methane and ethane. At first glance, this is not compatible with life as we know it from Earth. But the moon has a dense atmosphere that keeps substances in a liquid state of aggregation, and so there is at least a chance for an exotic form of biology. Saturn, on the other hand, would hardly be a safe haven for life because of its strong gravity and toxic clouds of ammonia. The same is true of Jupiter's deadly radiation belts and gas layers, whereas its moons are quieter.

"We know from our solar system that Jupiter-sized planets have large moons that can potentially hold water on their surfaces. If such a planet orbits its star in the habitable zone, there could even be moons that resemble Earth. If moons are also far more common than planets in other star systems, they may be the most promising place to look for extraterrestrial organisms."

(Chris Fox of Western University in Ontario)

Long before the Kepler telescope took its first images, astronomers assumed that exomoons were everywhere in the cosmos. As early as 1999, Paola Sartoretti and Jean Schneider of the Paris Observatory proposed using the so-called transit method to search for satellites. Celestial bodies eclipse their star a little when they pass it as seen from Earth. The prerequisite for such a transit is the arrangement of the planetary system in a flat plane, where we look at the edge of the disk from the side.

How to find exomoons

The Kepler telescope has used the method for a decade to detect exoplanets. Sartoretti and Schneider suspected it could also be used to detect moons. At least if they orbit their planet at a great distance and are just next to the planet at the time of the transit. This would make the star appear somewhat dimmer than expected during the transit.

Even if exomoons follow close orbits around their planet, the satellites could probably still be registered, Sartoretti and Schneider realized. To do so, they would have to keep a precise record of whether the planet's periodic transit pattern changes over time. For most of the time, the eclipses repeat with metronome-like precision. Sometimes the time between two transits fluctuates a bit: Then the mini eclipses start or end a little earlier or later than expected. The so-called transit time variation usually occurs when there are several planets in the star's orbit and they tug on each other by means of their gravity. Moons should cause a similar effect in principle.

In addition, one must know that also our moon does not wander on a completely circular course around the earth. Rather both bodies circle their common center of gravity, the barycenter. It is in our case within the globe, but not exactly in its center. In the case of exoplanets and their moons, the wobble could be pronounced enough to be noticeable in a targeted search.

In 2017, Alex Teachey of the Academia Sinica Institute of Astronomy and Astrophysics in Taiwan and David Kipping of Columbia University in New York scoured the data from the Kepler telescope for evidence of exomoons. They did this by analyzing about 300 planets in hopes of finding suspicious transit signals. In the end, they came across only one candidate: Kepler-1625b.

They then successfully applied for observation time on the Hubble Space Telescope. For a year, they analyzed the measurements. And indeed, Kepler-1625b's transit began earlier than it should, suggesting a moon. In a five-year data set, the transit time also shifted by about 20 minutes. "This makes it clear that something is pushing the planet around," Kipping says. "We think it's a moon."

Discovery or wishful thinking?

Teachey and Kipping published their work in October 2018, suggesting that the exoplanet Kepler-1625b - which itself is significantly larger than Jupiter - could have a Neptune-sized moon. However, the two researchers did not claim a discovery. "I think some colleagues were frustrated with the way we presented our result," Teachey says. "We apparently gave the impression of being eager for credit for discovery on the one hand, but wary of seriously claiming it on the other. I understand that's frustrating; after all, we're all wondering if there's a moon now. There are just too many uncertainties still." René Heller of the Max Planck Institute for Solar System Research in Göttingen, Germany, reproduced some of Teachey's results shortly after publication but found insufficient overall evidence for a moon. Laura Kreidberg, who studies the atmospheres of exoplanets at the Max Planck Institute for Astronomy in Heidelberg, could not confirm an important part of the results.

Meanwhile, interest in the subject grew. Soon, other teams were sifting through the Kepler data, tracking variations in transits triggered by moons. Others turned to optical observing systems such as the Spectro-Polarimetric-High-contrast Exoplanet REsearch instrument (SPHERE) on the Very Large Telescope. Cecilia Lazzoni of the University of Padua in Italy, for example, believes she has found a giant exomoon using SPHERE data. In a paper published in the journal Astronomy & Astrophysics, she and her colleagues described it as the companion of a very low-mass brown dwarf. That's what astronomers call a faint object that, unlike a star, does not fuse hydrogen but is many times the size of Jupiter. Lazzoni's world and its companion body would therefore be more like two giant planets orbiting each other than a world with a satellite. Should such pairs become more common, it would be necessary to clarify how one actually defines what distinguishes planets from moons.

In 2019, Phil Sutton of the University of Lincoln in England reanalyzed the super-Saturn J1407b mentioned at the beginning of this article. He wanted to find evidence for moons that lie outside the ring, as is the case with most of Saturn's moons. So he looked to see if J1407b's 37 rings were shaped similarly to the gas planet in our solar system. Sutton saw no evidence of moons; more to the point, such would destroy the exoplanet's fragile rings, according to his analysis.

In the summer of 2020, Chris Fox of Ontario looked through more Kepler data. Together with his colleague Paul Weigert, he found a total of 8 planets with transit time variations that could be explained by exomoons in a sample of 13 detected by the space telescope. However, other causes are equally conceivable, from variable stellar activity to additional planets. "In many cases, we were able to attribute the pattern to a possible moon, but in all cases, it could just as easily be explained by a second planet," Fox concludes.

Conflicting analyses of the same data

Finally, in November 2020, Kipping organized the first international conference on exomoons, which was held virtually and brought together about 80 experts. One of the findings: The discovery of an exomoon is probably still a long time coming, in part because the matter has so far been at the limits of what is technically feasible. The tiny difference in brightness during transit is already hardly detectable in many cases if it originates from a planet. The measurement of the transit time also requires precision, which previous instruments only just deliver.

This frustrates. Laura Kreidberg, for example, is chagrined that she and Teachey did not arrive at the same answer when analyzing the exomoon candidate around Kepler-1625b. The two compared their analyses but simply could not reconcile the conclusions. "I learned a lesson from that: We're really pushing the limits of what the Hubble telescope can do," Kreidberg says. "It was designed to observe distant galaxies, not comparatively nearby exoplanets and their moons.

Other difficulties are geometric. Because of the fundamental laws of motion formulated by Kepler and Newton, the orbits of moons are stable whenever they are within a certain distance of the planet, known as the Hill sphere. Otherwise, the satellite runs the risk of being knocked out of its orbit by the star's gravity. This becomes a problem especially when the distance between the planet and the central star is small. So far, however, most known exoplanets orbit very closely around their star, often closer than Mercury around the Sun. Thus, there is a fairly good chance that they are traveling without companions. "The planets we see passing in front of their stars are mostly unlikely to have moons attached to them because of the unfavorable gravitational conditions," says Stephen Kane of the University of California at Riverside. He published a paper in 2017 suggesting that close-space planetary systems like TRAPPIST-1's, with its seven Earth-like planets, are unlikely to host satellites at all.

Planets located at greater distances from their stars, such as Jupiter and Saturn, on the other hand, are more likely to have moons, argues Alice Quillen, an astronomer at the University of Rochester who has studied super-Saturn J1407b. This is supported not only by the lesser influence of the star's gravity farther out. Planets are also more likely to capture stray asteroids and other celestial bodies at the edges of a system, as Neptune is thought to have done with its moon Triton. The latter may once have been a dwarf planet like Pluto, but then at some point fell into Neptune's catchment area.

Exoplanets far away from their star are difficult to detect, also because they need a very long time for one orbit. For comparison: Jupiter needs twelve terrestrial years for this. It is necessary to witness two or more transits to reliably identify a planet. On top of that, such signals cannot always be distinguished from those from binary systems, where both partners orbit and eclipse each other at appropriate distances.

Stellar activity drowns out weak signals

In addition, the stars themselves make the job more difficult. Our Sun is a comparatively quiet fireball; others, however, are more prone to bursts of radiation, repeatedly ejecting vast amounts of matter into space and forming many spots on their surfaces. "The more precisely you can measure the brightness of a star, the more you see of stellar activity," Kane says. The whole thing then resembles noise in the data, which is sometimes larger than the signal from any moon. "That creates a sensitivity limit that it's not clear can be overcome."

That makes the ideas for possible mathematical and instrumental ways out all the more creative. Apurva Oza of the University of Bern, for example, is on the lookout for volcanically active moons that might give themselves away more easily. The Jupiter satellite Io serves as a model here; it is one of the four moons already discovered by Galileo Galilei. Amateur astronomers only need a good pair of binoculars. For professionals with the right equipment, Io is one of the most striking objects in the sky. The vents on its surface spew sodium and potassium into space. These and other substances can spread out to a distance equivalent to 500 times the radius of Jupiter.

Such a signature would also be detectable from some distance, even independent of a randomly matching transit. After all, thanks to spectrographs attached to telescopes, gases can be detected in the vicinity of stars. For example, measurements have repeatedly revealed sodium, potassium, and other suspicious elements, with no simple explanation. "It is possible that a moon is simply responsible," Oza speculates.

Until more powerful observatories are available, only more skillful analysis remains

For the future of the field, astronomers are initially hoping for the James Webb Space Telescope (JWST), which is scheduled to launch in late 2021 after several delays. Currently, there are no suitable instruments in operation. So for now, all that's left is fiddling around with data processing. "Part of our work is to find better methods," Teachey says. Accordingly, a discovery is less like a clearly definable eureka moment that laypeople often envision. Rather, it's about patiently running one test after another. In the end, something interesting may emerge, but it will have to be further verified.

In the medium term, observatories on the ground could also provide new data for the search for exomoons, such as the Extremely Large Telescope, which is currently being built in the Chilean Atacama Desert. The European space telescope PLAnetary Transits and Oscillations of Stars (PLATO) could also help in the search and is scheduled to launch in 2026. Promising projects such as the James Webb successor LUVOIR (Large UV/Optical/IR Surveyor) are then planned for the 2030s.

"Right now, all we have left is Hubble until James Webb is finally up and running," Kipping says. At least that gives time to refine search strategies. Laura Kreidberg is confident of using JWST to pick up signals from exomoons but acknowledges that an unequivocal discovery could be a while coming. "You have to be optimistically inclined to explore exoplanets." That's even more true for their moons.