Recent research casts doubt on whether nearby Kapteyn b, a supposed super-Earth circling in its star’s habitable zone, is a planet at all.

If the nearby exoplanet Kapteyn b exists, it would be the oldest-known planet to orbit its star in the coveted habitable zone, where water could flow on the planet’s surface. But this super-Earth could turn out to be nothing more than a trick of starlight.

Artistic representation of the potentially habitable exoplanet Kapteyn b as compared with Earth. Kapteyn b is represented here as an old and cold ocean planet with a network of channels of flowing water under a thin cloud cover. Planetary Habitability Laboratory
Artistic representation of the potentially habitable exoplanet Kapteyn b as compared with Earth. Kapteyn b is represented here as an old and cold ocean planet with a network of channels of flowing water under a thin cloud cover.
Credit: Planetary Habitability Laboratory

Kapteyn b is one of two recently-detected exoplanet candidates around Kapteyn’s star. Astronomers found the planets as part of the High Accuracy Radial Velocity Planetary Search (HARPS) project in the southern constellation Pictor. Their parent star is a cool, red dwarf and the nearest halo star to Earth at only 13 light-years away.

Both exoplanets appear to be super-Earths, but it’s Kapteyn b that had scientists in a tizzy. This exoplanet candidate had it all: a comfortable location in its star’s habitable zone, five Earths’ worth of mass, and an estimated age twice that of Earth — more than enough time for life to have evolved. Astronomers infer its age from the fact that the star probably originated in a dwarf galaxy that collided with the Milky Way eons ago, making it and its planets much older than their Milky Way counterparts.

But research from Paul Robertson (Pennsylvania State University) and colleagues posits that this goldilocks planet is actually an artificial signal created by stellar activity.

Seeing Planetary Ghosts?

Astronomers discovered Kapteyn b using the radial velocity method. This method relies on the fact that stars are not stationary — they orbit their system’s center of mass, responding to the tiny gravitational tugs of their planets, and that movement leaves a signature in the star’s light that astronomers can measure.

A star’s spectral lines move in the blueward direction if the star moves toward us, redward if the star’s moving away. Regular blue- and redshifts can indicate an exoplanet. But astronomers must take care. “Spots on the stellar surface and other magnetic phenomena can easily mimic a planetary signal as the stellar rotation brings these features in and out of sight,” says Tiago Campante (University of Birmingham, UK).

Usually, noise from starspots or other surface inconsistencies is small compared to the larger tug of an orbiting planet. But as technologies advance, scientists are able to look for finer details and smaller planets, so starspots become more relevant. In the case of Kapteyn b, the authors present a series of strong arguments that refute its existence, says Campante.

When Robertson and his team analyzed the Kapteyn b HARPS data, they were able to deduce the star’s rotation and found that Kapteyn b’s orbit was “worryingly close,” to an integer fraction of this rotation. At 48 days, the supposed exoplanet’s orbit is about a third of the star’s rotational period — consistent with a starspot being sampled at irregular intervals. The authors conclude that the exoplanet is nothing but a trick of starlight.

Is Kapteyn b No More?

Others disagree with this case of stellar mimicry.

If the signal that researchers originally attributed to Kapteyn b came from a starspot, that splotch should rotate with the star, assuming it lasts long enough. Along our line of sight, it would disappear behind the star as the star rotates, creating a signal with a period half that of the star’s rotation or some other fraction, says Xavier Dumusque (Harvard-Smithsonian Center for Astrophysics).

The authors argue that any signal at half-stellar rotation, as opposed to a third, is not seen in the data because of uneven sampling — telescopes simply couldn’t observe the star continuously. “It's very likely that there are signals at the [stellar] rotation period and other harmonics,” says Robertson.

Nevertheless, “a correlation between both signals is not a proof refuting the origin of the [planetary] signal,” Dumusque says, adding that the authors couldn’t show that the stellar activity caused the debated exoplanet signal.

So Kapteyn b’s existence is still up for debate. The only way forward is to keep observing, adding radial velocity measurements in order to get a better picture of Kapteyn’s star’s activity. The good news is that there are other similarly old halo and Galactic disk stars, should Kapteyn b be disproven conclusively. Kepler-10 and Kepler-444 are such stars, and even though they don’t host habitable-zone exoplanets, it “does not preclude the existence of habitable worlds around similarly old stars,” says Campante.

Reference:
Robertson et al. Stellar activity mimics a habitable-zone planet around Kapteyn’s Star, Astrophysical Journal Letters, published online 11 May 2015.


Read more about the exoplanet revolution from Sara Seager in our classic August 2013 issue.

Comments


Image of

June 10, 2015 at 9:09 am

Kapteyn b is only the latest in a series of extrasolar planets whose existence has been called into question as a result of detailed analysis of stellar activity performed by Robertson and his colleagues. The variations observed in the radial velocity of GJ 581 had been interpreted as being the result of as many as six planets including a couple of potentially habitable worlds. Many other astronomers have questioned this interpretation in independent analyses and work by Robertson et al. found only strong evidence for three planets with the others being the result of stellar surface activity modulated by the stars 130-day rotation period. This included the potentially habitable GJ 581d whose purported 66-day orbital period is almost exactly one half of the star period of rotation.

Then there is GJ 667C whose radial velocity variations had been explained by some by the presence of as many as seven planets. Again, many astronomers have questioned the existence of many of these purported planets (including the HARPS team) over the years. Another analysis by Robertson and Suvrath Mahadevan released last year could find only clear evidence for two planets with the rest being either undetectable through the noise or the result of natural "noise". For example, the signal attributed to GJ 667Cd with an orbital period of 92 days disappeared when the natural noise of GJ 667C modulated by its 105-day period of rotation (almost a 1:1 ratio with the planet's orbital period) was taken into account.

What are the odds of three different planets (i.e. Kapteyn b, GJ 581d, GJ 667Cd) orbiting three different stars having orbital periods that are an integer ratio of the stars' rotational period just by chance (i.e. 3:1, 2:1, 1:1 respectively)? Claims by Dumusque and the HARPS that these must be planets just because the precise mechanism of how stellar activity could mimic the radial velocity signature of an orbiting planet is not known simply does not hold water. The planetary interpretation of these finds is clearly suspect regardless of what the powerful mathematical techniques Dumusque et al. are telling them and new methods must be employed to characterize more accurately the natural noise of the stars being observed. Because of these and other examples, is obvious that we are quickly approaching the limits of the current radial velocity measurement and analysis technique's ability to detect small extrasolar planets.

You must be logged in to post a comment.

Image of Peter Rowen

Peter Rowen

June 10, 2015 at 10:25 pm

I would like to address the rhetorical question in the last paragraph. "What are the odds of three different planets (i.e. Kapteyn b, GJ 581d, GJ 667Cd) orbiting three different stars having orbital periods that are an integer ratio of the stars’ rotational period just by chance (i.e. 3:1, 2:1, 1:1 respectively)?" Well, assessing the precise odds is of course impossible, but we can certainly make meaningful estimates. Given that the number of known exoplanets and exoplanet candidates is in the thousands, and that the periods of their rotations cover a broad range that corresponds to typical stellar rotation periods as well, it is somewhere between unsurprising and inevitable that many planets have an orbital period that is an integer ratio of their stars' rotational period. The polemic tone of this paragraph and the rest of the paragraph ill serve the cause of science.

You must be logged in to post a comment.

Image of

June 11, 2015 at 6:35 pm

Peter, I am merely repeating the claims made by Robertson et al. as well as other astronomers in the community to place the current issue of Kapteyn b (as ably explained by Ms. Poore) into a broader context. You may label this as "polemic" but sometimes science is an adversarial process with competing interpretations of data and high stakes on who is right. But, as I stated, more data and better data analysis tools as well as independent assessments of the current results are needed to sort out the situation. But by the criteria typically used by astronomers in this field of work, the planetary interpretation of Kapteyn b and several other unconfirmed extrasolar planets like those I have mentioned are suspect and experts in the field are wrestling with the fact that we seem to be approaching a fundamental detection limit of the radial velocity technique for finding extrasolar planets.

You must be logged in to post a comment.

Image of Peter Wilson

Peter Wilson

June 11, 2015 at 10:23 am

I'm still amazed they can measure another star moving 1 m/s, from the surface of a planet rotating some 10 m/s, orbiting its own star at 30,000 km/s!

The word "data reduction" comes to mind.

You must be logged in to post a comment.

Image of Sandeep Joshi

Sandeep Joshi

June 13, 2015 at 9:49 am

I'd thought the rate of Earth's revolution around the Sun was 30 Km/s!

You must be logged in to post a comment.

Image of Tom Hoffelder

Tom Hoffelder

June 11, 2015 at 10:22 pm

Recent research casts doubt = the most common phrase seen at the beginning of science articles in the 21st century.

You must be logged in to post a comment.

Image of William Lamm

William Lamm

June 14, 2015 at 5:30 am

That's the purpose of good science - to always look further...and to admit when it is wrong : )

You must be logged in to post a comment.

Image of Peter Wilson

Peter Wilson

June 17, 2015 at 11:34 pm

But more research is needed = the most common phrase seen at the end of...

You must be logged in to post a comment.

You must be logged in to post a comment.