Existence of habitable exoplanet questioned

In late September, an experienced group of U.S. astronomers made headlines with news of the first extrasolar planet likely to be hospitable to life. The planet lies at a distance from its parent star at which water could be liquid (SN: 10/23/10, p. 5).

But a Swiss team of veteran planet hunters has now cast some doubt on that finding. On October 11, Francesco Pepe of the Geneva Observatory in Sauverny, Switzerland, announced at an extrasolar planet meeting in Torino, Italy, that a combination of old and new data acquired by his team shows no sign of the planet, dubbed Gliese 581g.

“If a signal corresponding to the announced Gliese 581g planet was present in our data,” Pepe says, “we should have been able to detect it.”

Other astronomers say that only time, and more studies, will tell if the first exoplanet in the habitable zone has truly been found or not. “I don’t know if we should be in such a hurry to say one way or the other,” says MIT astronomer Sara Seager. “We will have consensus at some point; I don’t think we need to vote right now.”

In 2009, when the Swiss team made its last report about planets orbiting the red dwarf star Gliese 581, they had 119 measurements of the star’s wobble — a telltale sign of unseen planets tugging back and forth on the star — recorded over four years. Those measurements revealed a total of four planets circling the star. The team’s latest report includes an additional 60 measurements made with the sensitive HARPS spectrograph on a telescope at La Silla, Chile, for a total of 6.5 years of observations.

“From these data we easily recover the four previously announced planets,” says Pepe. “However, we do not see any evidence for a fifth planet in an orbit of 37 days,” he said, referring to the planet in the habitable zone that Steven Vogt of the University of California, Santa Cruz, and his colleagues announced last month.

The underlying problem, Pepe says, is that despite the extreme sensitivity of the instruments used by the U.S. and Swiss teams to monitor the motion of the star, the signal of the potentially habitable planet is extraordinarily tiny, about the same level as the noise in the instruments. “Simulations on the real data have shown that the probability that such a signal is just produced ‘by chance’ … is several percent,” he says. “Under these conditions we cannot confirm the presence of the announced planet Gliese 581g.” Pepe adds, though, that his team can’t formally refute it either.

Vogt says he and his colleagues stand by their results, which analyzed a larger set of data and also found evidence for a sixth planet. “I feel confident that we have accurately and honestly reported our uncertainties and done a thorough and responsible job extracting what information this dataset has to offer,” he says.

Vogt’s team combined the Swiss group’s 119 measurements with 122 measurements recorded by his team over a span of 11 years using the HIRES spectrograph at the Keck Observatory atop Hawaii’s Mauna Kea. The wobble induced in the parent star by Gliese 581g can be detected only when the two data sets are combined, Vogt and his colleagues note in an article posted September 29 at arXiv.org and scheduled to be published in Astrophysical Journal.

“I am not overly surprised,” says Vogt, that even with some new data, the Swiss team does not find evidence of Gliese 581g “as these are very weak signals, and adding 60 points onto 119 does not necessarily translate to big gains in sensitivity.”

“As the Swiss group has our data,” Vogt adds, “I am also wondering why they have not already combined all the data together into a more complete analysis themselves.”

Pepe declined to say when and if he and his colleagues would add the U.S. Keck data to their analysis.

Seager, who is attending the conference, called Pepe’s presentation “striking.” She notes that, unlike Vogt’s, the Swiss team’s analysis doesn’t assume that all the planets orbiting Gliese 581 have perfectly circular orbits.

That difference in approach could be the crux of the disagreement between the teams, says Alan Boss of the Carnegie Institution for Science in Washington, D.C., who is also at the Torino conference. Allowing the known planets orbiting Gliese 581 to have eccentric, or elongated, orbits could mask the signal from an additional planet, he notes. On the other hand, Seager says, assuming perfectly circular orbits, as Vogt’s team has done, could produce false signals that appear to be additional small planets.

Mercedes López-Morales of the Carnegie Institution puts it this way: “The solutions from both groups might in fact fit the data, depending on what one does with the eccentricities. More follow-up data and a clear explanation of the different eccentricity assumptions made are in order before [Gliese] 581g can be either confirmed or discarded.” López-Morales coauthored a paper in the Jan. 20 Astrophysical Journal on the unintended role that assigning an orbital eccentricity to a planet may play in masking the presence of other extrasolar planets.

Planet hunter Ray Jayawardhana of the University of Toronto comments that “the detection of Gliese 581g was less than comfortably secure even in the original Vogt et al. paper.… Of course, it’s not easy to definitively rule out something, but the HARPS results presented here at Turin are at least raising some doubts. Given the extremely interesting implications of such a discovery, it’s important to have independent confirmation.”