Â鶹ÒùÔº

The search for life beyond our planet continues, and one of the most underappreciated tools in an astrobiologist's toolkit is statistics. While it might not be as glamorous as directly imaging a planet's atmosphere or finding a system with seven planets in it, statistics is absolutely critical if we want to be sure that what we're seeing is real and not just an artifact of the data, or of our observational techniques themselves.

A new paper by Caleb Traxler and their co-authors at the Department of Information and Computer Science at UC Irvine takes on that challenge head-on by statistically analyzing a set of about 10% of the total number of exoplanets found and judging their habitability. The work is on the arXiv preprint server.

Statistics is a numbers game, and over the past few decades we have collected plenty of numbers on around 5,700 confirmed exoplanets. However, so far at least, we haven't detected any definitive signs of life on any of them. So, finding which ones are the most likely to potentially support life as we know it is a necessary step. That would ensure the resources that could potentially detect that life, like the James Webb Space Telescope, which is capable of detecting exoplanet atmospheres, are pointed in the right direction to most accurately find something if it's there.

Breaking down how to calculate where the most likely candidates are is precisely what statistics is good at. But so far, astrobiologists have focused on an exoplanet's habitability using essentially one dimension—the "habitable zone" that is so frequently discussed when talking about exoplanets. Essentially, the habitable zone is just a calculation of the average temperature a planet experiences and whether or not that could support the existence of liquid water—an absolutely critical medium for life as we know it.

The authors make the argument that such a system is too general to be practically useful in finding a planet with a high probability of supporting life. They suggest looking at the characteristics of both the planet and its parent star, and then analyzing those characteristics compared to Earth, which is still the baseline for a "habitable" planet.

They analyze an exoplanet based on its radius, temperature, insolation flux (i.e. how much sunlight it gets), and density. Each of these values could have a major impact on a planet's habitability, according to various other studies. For the exoplanet's host star, they analyzed its effective temperature, radius, mass, and metallicity—a common measurement in stars that is the ratio of its iron content to its hydrogen content.

Using those eight parameters, they split 517 exoplanets for which the data existed into four different categories. Excellent Candidate means the planet is close enough to Earth to be of interest. Good Planet, Poor Star meant that at least one of the star's parameters were significantly different from our sun. Good Star, Poor Planet meant the characteristics of the planet were significantly different from those of Earth. The final category—Poor Candidate—fits the bill with neither the star nor the planet.

Good Star, Poor Planet was actually the category that most of the exoplanets fell into, with 388 systems (75%) falling into that category.

According to the authors, this might be more of a "detection bias" than an actual physical reality, as the techniques used to find exoplanets (such as transits) are heavily biased to finding large planets with short orbital periods, which would place them firmly in this category. They mention that given longer observational times, there is a good chance that exoplanet hunters would find more planets that would fit in the Excellent Candidate bucket, but that observational time hasn't been forthcoming yet.

Out of the full 517 in the data set, only 3 were classified in that Excellent Candidate bucket—Earth itself, Kepler-22 b, and Kepler 538-b. Kepler-22b especially seems like a great candidate, with only a 3.1% difference in temperature and 1% difference in insolation compared to Earth. According to the paper, it has the highest likelihood of harboring life and should be a prime candidate for atmospheric observation by Webb, which should be capable of doing so despite the 635 light-year distance.

Kepler 538-b is larger than Earth, and has a much higher temperature, but it still falls within the realm of potentially habitable. But that rarity points to another important finding from the paper—Earth is a statistical rarity in terms of planets, but not one that requires some miraculous confluence of planetary and stellar characteristics. Using a statistical technique called the Mahalanobis distance analysis, the authors found that Earth is around 69.4% different in terms of "statistical unusualness," making it rare, but not too rare.

Another rare type of planet were those that fell into the Good Planet, Poor Star category. Six planets ended up there because their host stars (which were all M-dwarfs, the most common stars in the galaxy) fell outside the defined habitable range for temperature. However, the authors point out that despite lying outside the generally accepted framework, these candidates had a good chance of harboring life given their other physical parameters. Many are already under observation from JWST, and if it turns out they do have viable habitable conditions, this could turn the field of astrobiology on its head due to the prevalence of their host stars in the galactic population.

This statistical analysis further cements some important points that avid astrobiologists have known for some time. Kepler 22-b is a prime candidate for further observation, and offers our best chance at seeing evidence of life on another planet. Conditions on Earth are relatively rare, but not so rare as to be considered a miracle. And there's a significant bias in the exoplanet dataset towards planets that wouldn't be habitable due to their large size and short orbital periods.

As the science of astrobiology and exoplanets moves forward, continuing this type of statistical analysis will provide valuable context that could otherwise mislead or obfuscate the areas that have the most potential to answer one of the most important questions to humanity—are we alone? With increasingly powerful observational equipment pointed in the right direction, we might soon have a definitive answer to that question.