If Kepler’s task was to give us a first statistical cut at the distribution of exoplanets in the galaxy, TESS (Transiting Exoplanet Survey Satellite) has a significantly different brief, to use its four cameras to study stars that are near and bright. Among these we may hope to find the first small, rocky planets close enough that their atmospheres may be examined by space telescopes and the coming generation of extremely large telescopes (ELTs) on Earth.
Thus the news that TESS has found its first planet of Earth size is heartening, even if the newly found world orbiting HD 21749 is in a tight 7.8 day orbit, making it anything but clement for life. What counts, of course, is the demonstrated ability of this mission to locate the small worlds we had hoped to find. Diana Dragomir is a postdoc at MIT’s Kavli Institute for Astrophysics and Space Research, as well as lead author on the paper describing the latest TESS planet:
“Because TESS monitors stars that are much closer and brighter, we can measure the mass of this planet in the very near future, whereas for Kepler’s Earth-sized planets, that was out of the question. So this new TESS discovery could lead to the first mass measurement of an Earth-sized planet. And we’re excited about what that mass could be. Will it be Earth’s mass? Or heavier? We don’t really know.”
Image: NASA’s Transiting Exoplanet Survey Satellite (TESS), shown here in a conceptual illustration, will identify exoplanets orbiting the brightest stars just outside our Solar System. Credit: NASA’s Goddard Space Flight Center.
Dragomir considers the new planet around HD 21749, some 52 light years from Earth, a milestone in being the mission’s first planet of Earth size, though the expectation is for at least a few dozen more among the nearest and brightest stars as TESS continues sweeping the sky in overlapping sectors. This is a two-year mission that has already discovered 10 planets smaller than Neptune, four of which now have estimated masses. We’re in the early days of this mission, and that’s a good sign. As noted in the paper, “All of these discoveries are based on only the first two sectors of TESS data, suggesting many more are to be found.”
The star is a K-class dwarf in the southern constellation Reticulum that is known to host a second planet, recently confirmed, that is about three times the size of Earth. The paper reports on the discovery and confirmation of HD 21749b and the discovery of HD 21749c, but it is the latter, given its small size, that is receiving the lion’s share of coverage.
The paper notes that spectroscopic and photometric data have made the confirmation of the larger planet possible, while the Earth-sized HD 21749c would be a challenging observation for radial velocity confirmation, if possibly feasible with a dedicated campaign using the combination of the Very Large Telescope (VLT) and the ESPRESSO spectrograph. But density measurements of both planets should be useful. From the paper:
…the density of HD 21749b indicates that it is likely surrounded by a substantial atmosphere. By measuring the density of these two planets (and other similar planets that TESS may find) more precisely, we can begin to observationally constrain the maximum core mass a planet can reach during its formation before accreting a volatile envelope.
The paper is Dragomir et al., “TESS Delivers Its First Earth-sized Planet and a Warm Sub-Neptune,” Astrophysical Journal Letters Vol. 875, No. 2 (15 April 2019. Abstract.
“a tight 7.8 day orbit, making it anything but clement for life”…
It would indeed, if it implies a proximity to the host that could result in tidal locking and/or incineration. For life as we can imagine.
More clement conditions elsewhere could still be inadequate for our life, while permitting life as we have (or haven’t) imagined.
Well said, Robin.
Among planets in the galaxy, the question remains, how often do you get an “earth”? How many exo’s have now been found? Something over 4000? So one is tempted to say that earths are less than 1 in 4000, but with the sampling bias of transiting planets, this may not be so.
Given how hard it is to spot small planets in wider orbits for radial velocity methods, it’s going to be a number much larger than 1 in 4000. See Ravi Kopparapu’s article for more. He’s investigated the question thoroughly:
https://centauri-dreams.org/2014/10/03/how-common-are-potential-habitable-worlds-in-our-galaxy/
Quoting from the Kopparapu article re M-dwarfs alone:
“M-dwarfs are the most prevalent stars in our Galaxy. About 77% of stars in our Galaxy are M-dwarfs. Within 30 light-years of the Sun, there are nearly 250 M-dwarfs (whereas, there are only about 20 Sun type stars). So I recalculated Dressing & Charbonneau’s estimate of the prevalence of potential habitable planets, using my newly determined HZs (Figure 1). And the number I got is a BIG increase from 15%: a conservative estimate showed that about 48% of M-dwarfs should have Earth-size planets in the HZ. That means, nearly 1 out of 2 M-dwarfs (i.e, approximately 50% of M-dwarfs in our Galaxy) may have Earth-size planets in the HZ!” But he also treats Sun-like stars in the above link.
The eta-Earth (occurrence rate of Earth-like planets, ??) calculations vary greatly in literatures, and more than half of researchers have given values differ from each other by more than an order of magnitude, including Ravi Kopparapu. I have spent a lot of time trying to dig deeper into this subject. It is important to examine where the discrepancy comes from and what factors are causing it.
While the ?? calculations around M-dwarfs and late K-dwarfs are relative consistent, the value, 0.24-0.60 defined by conservative HZ and radius between 0.5 and 1.4Re, given in Ravi Kopparapu (2014) using Kepler dataset has been significantly revised down. Ravi Kopparapu (2014)’s result was based on the same samples (K01422.02 and K02626.01) used in Dressing & Charbonneau (2013) but adopted a less strict HZ, thus expanding the HZ samples (Adding two more K02418.01 and K01686.01) and getting higher ??.
In 2015, Dressing & Charbonneau updated their result from 2 years ago by adopting more accurately measured stellar properties and rigorous Kepler pipeline detection efficiency. With the new planet properties, two of HZ samples in Kopparapu (2014) turned out to be mini-Neptunes and K01686.01 was confirmed as false positive, leaving K02418.01 the only HZ sample. This updated M-dwarf planet catalog led to a smaller M-dwarf ?? of 0.11-0.39.
While Dressing & Charbonneau (2015)’s result is the current best estimate of ??, significant revision is still needed after new photometry and Gaia DR2 become available in 2018 and 2019. After Gaia DR2, K02418.01 is now a mini-Neptune instead of a super-Earth, and in the Kepler dataset there is no M-dwarf planet that meets the definition of an Earth-like planet (conservative HZ and 0.5-1.4 Re) anymore.
This is interesting. Now we know that the past M-dwarf ?? calculations are actually no longer valid due to the use of inaccurate stellar properties.
I believe the true M-dwarf ?? is likely to be smaller than current estimate but not by too much.
Excellent work, Nicky. Thank you.
Since Kepler didn’t detect any Earth-like planet around M-dwarfs, ?? could not be directly measured. Past constraints on ?? using Kepler data, such as Dressing & Charbonneau and Kopparapu, even if they accidentally cover the actual ??, lose their empirical evidence support and meanings.
There are other statistics of ?? using radial velocity method. HARPS team (Bonfils et al 2013) had monitored over 102 nearby M-dwarfs from 2003 to 2009 discovering two low-mass planets, GJ 581 d and GJ 667C c, orbiting within the HZ. The team adopted a very loose definition (minimum mass in between 1Me and 10Me and optimistic HZ) of Earth-like planet and obtained ?? = 0.28-0.95 (0.41). But now the signal of GJ 581 d is widely regarded as stellar activity instead of an actual planet (Robertson et al 2014), the exclusion of this planet reduces ?? to 0.22-0.76 (0.33). This estimate still only represents the optimistic upper limit of the true ??, because the ?? definition used by the team covers wide range of planetary compositions including mini-Neptunes.
Based on all these calculations using Kepler transits data and radial velocity, the true ?? most likely falls in between 0.1 and 0.2
So, just doing the math there might be as many as 125 Earth sized planets around M-dwarfs within 30 light years of us. It would be very interesting to test that hypothesis. If we could get either radial velocity or transit measurements for even 20% of the 125 (or about 25 planets) we might be able to produce a better estimate.
Looking at the Planetary Habitability Index tables I think there are 8 known potential rocky planets in the habitable zones of any stars within 30 light years of us with a fairly large degree of uncertainty so we have a long, long way to go!
Given the prevalence in nature that seems to be that small is not only beautiful but ubiquitous, TESS, and the (finally) JWST, should show that smaller worlds are likely more common that current search techniques have indicated. If planets like Earth were not that common, then the chances of Earth and Venus in the same solar system would be even more intriguing. We look at Mercury, and it appears as if it has lost the bulk of it’s outer layers and was possibly once a world significantly larger than today.
If we want to know what other planetary systems are like then I would suggest we pay far more attention to the largest planets, because both Saturn, and especially Jupiter, are more likely typical of larger bodies, from Gas planets to brown dwarves to stars. What makes this interesting is the 4 planet sized moons of Jupiter so close together, so can we assume this a statistical anomaly or something rather common? The similarity between the Jovian system and the solar system as a whole, for me, points the way to what we can expect to find.
Perhaps all planetary systems will have a number of large planets and then a retinue of smaller ones that range in size from diminutive Ceres/Pluto to those mightier than Jupiter and even brown dwarves, with the larger ones all having their retinue of smaller bodies.
The next 20 years will be very interesting and informative.
I agree James. Smaller objects typically vastly outnumber the larger ones in what can often be fitted into mathematical inverse power law distribution curves. Why would (or how could) planetary mass distribution buck this natural trend?
I’ve often wondered about super jupiter planets. Shouldn’t they be orbited by supersized moons as a common outcome?