A line-up of rocky, habitable zone planets discovered by Kepler. Image Credit: NASA/Ames/JPL-Caltech.

Sep 12, 2015 Are rocky habitable zone planets dependent on stellar composition?

Sen—Three astronomers from the Universidade do Porto, Portugal have begun tackling a very large and multi-dimensional question: How is the frequency of small habitable zone planets dependent on the properties of the host star? Knowing the answer to the question proposed in Vardan Adibekyan and his co-authors' study would help determine which types of stars are most suitable for habitable worlds, providing a component of the infamous Drake Equation. This study makes a commendable attempt at proposing an answer to this, though as with many early attempts it is limited by the data and techniques available.

Adibekyan and collaborators set out to find how the frequency of small, rocky planets in the habitable zone is affected by the composition of the host star. One of the first exoplanet properties noted was that the frequency of giant planets increases for host stars with more heavy elements. This study is one of the first attempts at figuring out this relation for Earth-like planets.

The composition of a star is typically broken up into three main categories: The fraction of a star composed of hydrogen gas, the fraction composed of helium gas, and the fraction composed of any element heavier than helium (also in gas form). While it might seem strange to lump over one hundred chemical elements into a single category, hydrogen and helium make up over 98 per cent of a star's total mass. While elements like oxygen, silicon, magnesium, and iron are much heavier atoms than hydrogen or helium, they are just not abundant enough to make a significant contribution on their own. For general purposes, they usually get grouped together along with the remaining elements into the term "metals."

Astronomers assume that most stars have similar fractions of hydrogen and helium, so stellar composition is typically defined only by the quantity of metals. Hence, "metallicity" is used as a proxy for "composition." A scale relative to the Sun is used to quantify the metallicity of a star. Stars with fewer heavy elements than the Sun are termed "metal-poor" stars and those with more heavy elements than the Sun are termed "metal-rich."

A ratio of iron-to-hydrogen abundance is used in almost all astronomical studies as a proxy for the fraction of all metals, as iron is typically observable in a star and is one of the more abundant metals. Where there is little iron to be found, silicon and magnesium are often found to be more abundant instead.

These three elements are crucial in the formation of rocky planets: the Earth is composed of a large iron-nickel core with a silicate rock mantle and crust on top. The amount of these three elements available to make planets and where those elements can be found in a proto-planetary disk, should determine where exactly rocky planets can form and what their internal structures will be. Adibekyan seized on the importance iron, silicon, and magnesium to support the hypothesis that habitable zone rocky planet formation would depend on the composition of the host star.

Under this assumption, the authors compiled data on all low-mass exoplanets around stars larger than half the Sun's mass that were detected via radial velocity. They further culled this sample by removing any small planet that had a companion planet larger than 10 times Earth's mass, since the larger planet would affect where the small planet naturally tended to form. This gave them 25 small planets in 12 star systems from radial velocity measurements.

The radial velocity sample was supplemented by transiting planets smaller than twice the size of Earth around stars meeting the same criteria as the radial velocity targets. This list was then filtered to remove non-confirmed planets and highly-irradiated planets that may have lost mass. This left them with 45 transiting planets in 20 star systems, for a total combined sample of 60 planets in 32 star systems.

They split their sample of stars into metal-rich and metal-poor categories, split at 80 per cent of the Sun's metallicity. Though the reason for this choice of delineation is not made clear in the report, it is likely because the average metallicity of stars in the Milky Way is slightly lower than the Sun. Their choice reflects the average Galactic metallicity rather than the Solar value.

They examined how the orbital distances of their final sample are related to the planetary mass and/or the planetary radius, depending on which measurements are available. Of the 60 planets examined only 15 of them fall within the habitable zone. They chose to use a single orbital distance to determine the interior edge of the habitable zone. While not ideal, it is an understandable choice since some of their target stars lack measured temperatures. They chose to use the most optimist location for the inner edge of the habitable zone; likely, some of those 15 planets are not really habitable.


Habitable zone planets around Sun-like stars, plotted with their orbital distance versus either planetary mass or radius. Red symbols correspond to planets orbiting metal-poor stars and blue symbols represent planets orbiting metal-rich stars. Asterisks indicate mass measurements (and should be compared to the left axis), and closed circles indicate radius measurements (and should be compared to the right axis). Three planets with both mass and radius measurements are connected by red lines. Three planets orbiting the coolest stars are marked by open large circles.

They find that of their 15 rocky habitable zone planets, 10 of them orbit metal-poor stars, while only five of them orbit metal-rich ones. Of those five, they note that three of them orbit the three coolest stars in their sample. Those three stars have the least accurately measured metallicities and might, in fact, be metal-poor. From this the authors conclude that rocky habitable zone planets around Sun-like stars are more likely to form around metal-poor stars. They then make a tenuous speculation that, since older stars are generally metal-poor, rocky habitable zone planets may have formed more frequently when the galaxy was younger.

With only 15 habitable zone rocky planets to work with, the authors caution that their results may be dominated by the detection biases of the radial velocity and transit surveys. Since our telescopes and detection programs have only begun being sensitive enough to find a small planet in the habitable zone, it is possible the ones in this sample are not truly representative of the whole planet population, but are just the easiest ones to find. Future exoplanet missions like NASA's Transiting Exoplanet Survey Satellite (TESS), and European Space Agency's CHaracterising ExOPlanet Satellite (CHEOPS) and PLAnetary Transits and Oscillations of stars (PLATO-2.0) missions will bring in even more data than Kepler and help us fill out the ranks of small habitable zone planets.

While the authors do caution that their connection of habitable planet frequency with stellar age is speculation, it is a very large leap to make based on these data. Stellar metallicity, along with age, are two of the most difficult stellar parameters to nail down. Age is difficult because once a star begins burning hydrogen, it stays relatively unchanged for over 90 percent of its life. Sure, the star may become slightly larger and brighter over billions of years as it ages, but it is extremely difficult to tell whether a hydrogen-burning star began life with those characteristics or evolved there over time.

Metallicity is tough to determine for a few reasons. Firstly because the effects of slightly changing the composition of a star can easily be confused with the effects of altering other stellar properties. Without having ultra-precise stellar spectra, the exact ratio of heavy elements to hydrogen can only be broadly guessed at. Indeed, broad guesses are all we have for the majority of the stars in the Kepler Input Catalogue from which the authors draw much of their data. Many of the planet host stars included in this study have large uncertainties on their metallicities which might swing them into either the metal-rich or metal-poor category. Even with precise spectra, metallicity measurements can be obscured by looking at the star through the Earth's atmosphere, which contributes its own highly-variable spectrum to the mix.

The composition of a star is determined first and foremost by the particular interstellar cloud from which it formed. The overall heavy-metal abundance of the galaxy does change over billions of years as stars process lighter elements into heavier ones and recycle them back into star forming material at the end of their lives. Stars born at the beginning of the galaxy usually have fewer heavy metals than young stars. But since the radial velocity and transit targets span large, diverse regions of the galaxy, the observed range of stellar metallicities could simply be a product of their environments, and not of age.

Assuming that the frequency of observed habitable zone planets really is a result of stellar age, this begs the question of whether the planets currently observed in the habitable zone really formed there or whether they were brought into it as their star evolved. As the hydrogen-burning star evolves it pushes the habitable zone farther out. So, a planet that is exterior to the habitable zone at formation might end up in the habitable zone over time. Without having an accurate sense of the host stars' ages, we cannot know whether the planets around metal-poor stars truly formed there, as per the authors' conclusion.

Overall this study introduces the interesting idea that the formation of a small habitable zone planets may be controlled largely by the host-star's composition. However, our exoplanet surveys have only just begun to detect enough habitable zone planets to attempt this type of analysis, and our understanding of stellar metallicity and planet formation is still too limited to make any firm conclusions. For sure, this idea should be revisited after future large-scale exoplanet surveys begin raking in results.