The second-closest star system to Earth is one of the most intensively studied targets in the galaxy. Astronomers have been measuring minuscule characteristics of the star for decades, searching for signs of orbiting planets, and over the years, hints of a possible world tugging on the star have been gleaned here and there—but nothing was ever considered conclusive.

Now, however, astronomers have finally revealed strong signs that Barnard’s Star, only about six light-years away, has one or more planets orbiting around it.

A team of researchers at the exoplanet-hunting institution Red Dots published a study today in Nature that compiles more than 20 years of observations of the star. The work drew out a quite unusual planetary candidate: a super-cold planet, about 3.2 times the mass of Earth, that circles its star every 233 days.

Barnard’s Star, a small red dwarf, has long been a prime planet-hunting target. From 1963 to 1972, the star was widely believed to host one or more gas giants, accounting for some of the earliest extrasolar planet claims that received widespread attention. The “planets,” however, ended up being nothing more than an instrumentation problem with the Sproul Observatory in Swarthmore, Pennsylvania, whose director, Peter van de Kamp, first claimed the existence of planets around the star.

Since that time, however, planet hunter after planet hunter have pointed their telescopes to the star for signs of anything out of the ordinary. Although Barnard’s Star is the second-closest star system to Earth—after the three stars of the Alpha Centauri system—astronomers continued to come up empty handed until many years of data could be acquired.

Now, this cold Super-Earth is the second-closest known exoplanet to Earth, after Proxima Centaur b .

“In the end, we believe firmly enough the object is there,” Ignasi Ribas of the Institut de Ciències de l’Espai and lead author of the paper says. “We always have to remain a bit cautious, and of course more data should strengthen the case for it, but we were sure enough that we were willing to go forward with publication.”

Part of the challenge of finding the planet comes from the method that astronomers used: radial velocity (RV). The radial velocity method looks for gravitational changes as a planet pushes or pulls on its star. The method is great for finding large planets, and has been effectively used to find smaller, nearby planets like Proxima Centauri b—but for finding small, far-out planets, it’s a different matter entirely.

Barnard’s Star b, as the new planet is known, causes just a 1.2 meter per second change to its star’s radial velocity. The first planets discovered by the RV method caused 13 meter per second changes, making this signal less than 10 percent the strength of other successful finds.

Cullen Blake, a University of Pennsylvania professor who was not involved in the study, says that some of the RV data could potentially have been muddied by stellar activity from the star. But nevertheless, the sheer number of observations builds a fairly compelling case that the planet is there.

“It seems like there’s some substantial noise in terms of stellar activity that they’ve modelled out,” he says. “[The candidate] is very strong in terms of the statistical significance.”

The first hints that Barnard’s planet was there came in 2015, when one of the co-authors of the paper, Mikko Tuomi, looked at existing radial velocity data from the star and found something changing according to a 233-day period. At the time, the case for a planetary object (instead of some kind of activity intrinsic to the star) wasn’t strong enough to stand on its own.

Ribas and his team used 800 different observations of Barnard’s Star to drive down the uncertainty that the planet existed. After crunching the numbers of decades' worth of data, the existence of Barnard’s Star b seems fairly certain—though the ever-cautious Red Dots team is still calling it a planet candidate.

And the researchers also found something else in the data: faint evidence of another planet, which would be known as Barnard’s Star c. This ghostly second planet in the system, if it exists, is estimated to have a long orbital period and about the mass of Neptune. The existence of the gas planet is even harder to prove, as its 10-year period means it produces a weaker signal. It is worth noting, however, that the measurements place the planet at a similar period to van de Kamp’s claims in the 1960s.

Van de Kamp thought he saw at least two planets, one in a 12-year and one in a 25-year orbit, both around the mass of Jupiter. The clues came from astrometric signals, which look for deviation in the path of a star from a center line due to the influence of an unseen planet.

The new signal, on the other hand, seems to indicate something about 15 Earth-masses, which is unlikely to show a noticeable astrometric signal from Earth. (The new paper mentions that Gaia, a space-based astrometry observatory, might be able to confirm the planet.) Because of its size, the planet makes about a 3 to 4 meter per second radial velocity change—stronger than the other planetary candidate, but without the sheer number of observations to rule out an alternate explanation.

“We’re not making the claim that this is in any way related to the van de Kamp astrometric planets,” Ribas says. He does note, however, that it is an interesting coincidence. “We do see some variation. The stars are known to show activity cycles, so this might be a cycle of stellar activity [rather than a planet].”

The 233-day period for the snow world, Barnard’s Star b, is far beyond the habitable zone of its star. Even though the planet is closer to its star than Earth is to the sun, Barnard's Star is very small and cold by comparison. The planet orbits in an area known as the “snow line,” where rocky bodies tend to give way to icier bodies. The paper also says that there is no evidence of anything around the mass of Earth within the habitable zone of Barnard’s Star—though Mars-mass objects haven’t been ruled out.

But even if this world is too cold for life as we know it, it marks an intriguing discovery, as the planet is much farther away from its host star than other planets found orbiting small red dwarfs (such as Proxima Centauri b, which has a period of just 11 days). From an exploration perspective, the planet is closer than any known exoplanet save one, and in 10,000 years, Barnard’s Star will actually have moved in to become the closest star to our sun. Physicists are already thinking about possible ways to launch a miniature probe to Proxima Centauri b , about 4.2 light-years away—and in the future, Barnard’s Star b could become another target.

“My feeling is that these things are probably fairly common, but extracting them is just much harder,” Blake says of distant, small, long-period planets like Barnard’s Star b. “My guess is that there will be more campaigns like this.”