Barnard’s star is an unassuming M-class dwarf: small, red, several billion years older than the Sun, and quiet in its dotage. Yet, at about 6 light-years from Earth, it has a certain fame in some circles.
That fame culminated in 2018 with the discovery of a super-Earth, dubbed Barnard’s star b, hidden in the star’s meager light. Now, a new study to appear in the Astronomical Journal (preprint available here) challenges that claim.
A Planet for Barnard’s Star?
Due in part to its proximity, Barnard’s star has the highest proper motion of any star, so its movement across the sky is clearly visible to the patient astronomer, including E. E. Barnard, whom the star is named after. Perhaps that’s part of why it caught the imagination of science fiction writers, most notably serving as a stellar waystation in both Hitchhiker’s Guide to the Galaxy and The Garden of Rama.
Those sci-fi authors appeared prescient when a team of astronomers led by Ignasi Ribas (Institute of Space Sciences, Spain) discovered a planet with at least three times Earth’s mass circling its red sun every 232 days. The researchers reported a 99% confidence level in the discovery, meaning there was a 1% chance that the signal was a statistical fluke.
Assuming it wasn’t a fluke, the planet would be too chilly for life. At 0.4 astronomical units (a.u.), less than half Earth’s distance from the Sun, it would nevertheless be outside its cooler star’s habitable zone.
The find utilized 20 years of radial velocity measurements collected by seven different instruments as they monitored the star’s slight motions toward and away from us. Within that data, Ribas and colleagues saw the repeated signal that indicated a planet’s gravitational back-and-forth pull as it orbits its star.
When the velocity changes are so slight, though, a star’s roiling surface complicates matters. Red dwarfs are infamous for magnetic flare-ups and starspots that can introduce velocity changes that have nothing to do with a planet. Barnard’s star’s age gives it an advantage: It seems to have for the most part given up the fits of youth. Ribas’s team nevertheless searched for the possibility of stellar interference in their data — and found none.
A Challenge to the Claim
But a new analysis of that data, with the addition of observations from the newly minted Habitable-zone Planet Finder on the Hobby-Eberly Telescope in Texas, puts a challenge to that claim. Graduate student Jack Lubin (University of California, Irvine) and colleagues find that stellar activity might indeed be masquerading as a planetary signal.
Lubin led a team in analyzing 856 days’ worth of data from the new spectrograph. Even after three exoplanet-years’ worth of data, the planet’s signal didn’t become evident. The team also re-analyzed the data published by Ribas’s team, breaking the data into thirds and then into smaller chunks to show that the signal strengthens and weakens over time — something a planet’s signal would not do.
In fact, the researchers find that almost all of the data supporting the signal consists of 211 data points collected over less than three years, from 2011 to 2013 . And during this time period, Lubin says, the team found a chemical signature associated with stellar activity that also has a 233-day period.
The team concludes the ephemeral signal is more likely to be caused by something like a starspot. Such a feature might last for hundreds to thousands of days on small, elderly Barnard’s star, persisting over multiple 145-day rotations. Combined with uneven sampling from Earth, plus Earth’s own radial velocity signature due to its orbit around the Sun, the stellar activity could mimic the 233-day planetary signal.
Lubin and his colleagues note that this would be the first time this kind of mimicry would have occurred over periods longer than the star’s own rotation. That means the planetary community needs to be on guard regarding farther-out planet candidates, as even wide orbits do not necessarily make a planetary signal immune to statistical interference from its host star.
Barnard’s Star: To b or Not to b
However, Ribas stands by his team’s detection. “I think it is extremely healthy that all published science is scrutinized by independent groups, and therefore we welcome Lubin et al.’s analysis,” Ribas says. But he adds that his team already looked at the possibility that the star’s rotation period affected their results and found it wasn’t likely. “To our dismay, the new study seems to disregard that we ran all those checks.”
Ribas also takes issue with Lubin’s team’s re-analysis of the two decades of archival data, which must be combined ever so carefully to avoid introducing or removing signals. “The authors do not really discuss in detail how they handle the different instrument zero-points,” he says. “Improper handling of offsets can lead to cancelling a low-frequency signal such as that induced by the candidate planet.”
At the end of the day, the existence of the planet rests on the ability to detect a radial velocity that is on the cutting edge of what modern detectors can do. The planet, if it exists, induces a stellar wobble of 1.2 meters per second, while the instruments Ribas’s team used were sensitive to individual velocity changes of about 0.9 to 1.8 meters per second; the detection was possible because of data summed over a 20-year baseline. Lubin notes that the Habitable-zone Planet Finder can sense changes as small as 0.8 meters per second, but Ribas argues that its long-term stability at this level hasn’t yet been proven, and so far, it has collected only about two years’ worth of data on Barnard’s star.
To settle the debate, astronomers will need more data. “The more data points we collect, the more confident we become of trends or signals,” Lubin says. Indeed, several groups are already — or soon will be — adding to the data pile using high-resolution spectrographs all over the world. The fascination with Barnard’s star is far from over.