Exoplanet Stuff

[lpetrich]

Oops: In the Solar System, Mimas-Tethys: 1.6 yr (free) 71 yr, Enceladus-Dione: 4 yr (free?) 11 yr, Titan-Hyperion: 19 yr (free) 360 yr.

Conjunction directions:

• Mimas - Tethys: halfway between the two moons' ascending-node directions
• Enceladus - Dione: Enceladus's pericenter (has a forced eccentricity)
• Titan - Hyperion: Hyperion's apocenter (has a forced eccentricity)

Planetary Satellite Mean Elements

Planets also have resonances or near-resonances.

Jupiter and Saturn are close to a 5:2 orbital resonance, and tney have a "great inequality" with a period of about 850 years and an amplitude for Jupiter of 20m of arc and for Saturn of 48m, in opposite phase.

Venus and the Earth are close to an 8:5 resonance, but it's much weaker. Neptune and Pluto are in a 3:2 resonance.

 

[lpetrich]

I now get into how Neptune was discovered.

In the 18th and 19th centuries, astronomers worked out how to solve Isaac Newton's laws of motion with his law of gravity. Since they could not solve the three-body problem in closed form in most cases, they contented themselves with doing perturbation series expansions.

In the process, they solved a discrepancy that made Newton's head ache: the Moon's pericenter precession rate. For both the pericenter and the nodes, one find 17.83 years from the lowest-order calculation, though the pericenter (and apocenter) precess forwards and the nodes precess backwards. The actual value for the nodes is 18.60 years, not far off, while the actual value for the pericenter is 8.85 years, twice as much.

But that was from the *lowest* order of perturbation expansion. The next few terms in the perturbation series expansion were enough to get close to the observed value, thus solving the problem that was was so difficult for Newton.


How to measure the masses of Solar-System bodies?

For the Earth, that's easy. Measure its acceleration of gravity. It is slightly larger at the poles than at the equator, and one has to measure the size of the Earth's equatorial bulge with precision surveying to factor out that effect, because it makes its own amount of gravity. One finds the Earth's mass in what I call gravitational units: (gravitational constant) * (mass) with dimensions (length)³ / (time)².

Over the last half-century, a high-precision method has emerged: tracking artificial satellites. Using the Kepler-Newton 1-2-3 law, as I call it, one can get the Earth's mass, and with more careful analysis, the departure of its gravity from spherical symmetry, including the gravity of its equatorial bulge.

That can be extended to the rest of the Solar System, for anything that is orbited by something.

For objects that aren't orbited by something, it's much more difficult. For the Earth's Moon, one can find the effects of its pull on the Earth. It makes the Earth move in a miniature version of the Moon's orbit, and that effect can be measured by observing an asteroid that the Earth goes near. That also happens to the other planets that have moons, but that effect is very tiny and very difficult to observe.

Needless to say, spacecraft missions have helped there also.

 Timeline of discovery of Solar System planets and their moons
The Sun, the Earth, Mercury, Venus, Mars, Jupiter, and Saturn are all pre-telescopic.

• 1610: Jupiter: Io, Europa, Ganymede, Callisto -- the Galilean moons
• 1655: Saturn: Titan
• 1781: Uranus
• 1787: Uranus: Titania, Oberon
• 1846: Neptune
• 1846: Neptune: Triton
• 1877: Mars: Phobos, Deimos

I'm only giving the first moons discovered.

 

[lpetrich]

So by the early 19th cy., one could get good mass values for the Sun, the Earth, Jupiter, Saturn, and Uranus. But what of Mercury, Venus, and Mars? Urbain Leverrier found their masses by working from how much the inner planets' orbits precess. Comparing his mass values with present-day ones, Venus and Mars are very close (2% and 16% too large), and Mercury's mass is twice the present-day value, though it is small.

He is notable for identifying some excess precession of Mercury, and he proposed that it was due to some intra-Mercurian planet or planets. But no such planet was ever discovered, and an alternate solution was to modify the law of gravity, like making its exponent a little different from 2. Modifying the law of gravity was what ultimately succeeded, in the form of Einstein's General Theory of Relativity. It states that space-time is curved, and that this curvature is controlled by the mass density of matter in it.

But in the 1840's, he and another astronomer, John Couch Adams, made another planet prediction. By then, Uranus had been observed for almost a complete orbit, and its motion did not quite fit what could be predicted from the rest of the Solar System. The two astronomers worked from those discrepancies to predict another planet, and that planet was soon discovered: Neptune.

Astronomers have searched for additional trans-Uranian planets, because taking Neptune into account did not quite account for the discrepancies in Uranus's motion.  Planets beyond Neptune has more. Early in the 20th cy., Percival Lowell called a possible trans-Neptunian planet "Planet X", and in 1930, Clyde Tombaugh discovered it: Pluto. Or so it seemed. Pluto turned out to be too low in mass to make these orbital discrepancies, and that issue was finally resolved when its largest moon Charon was discovered in 1978. Around then, the two Voyager spacecraft were launched to Jupiter and Saturn, with the second one continuing to Uranus in 1986 and Neptune in 1989. That gave good mass values for the four outer planets, and that made the discrepancies in Uranus's motion disappear. The trouble was that earlier estimates of Neptune's mass were a little off.

 

[lpetrich]

After this long discussion of the Solar System's planet-planet and moon-moon effects, I get into exoplanets that have been discovered by analyzing transit timing variations.

The first of them was a planet of Kepler-19, discovered by Sarah Ballard and her colleagues.  Kepler-19 The Kepler-19 System: A Transiting 2.2 R <SUB>⊕</SUB> Planet and a Second Planet Detected via Transit Timing Variations - NASA/ADS with preprint [1109.1561] The Kepler-19 System: A Transiting 2.2 R_Earth Planet and a Second Planet Detected via Transit Timing Variations

That star is at a distance of 220 parsecs (717 light years), and it has a mass of 0.936 solar masses, a radius of 0.85 solar radii, and a surface temperature of 5541 K. I estimate a luminosity of 0.61 solar luminosities.

The first planet discovered was discovered with transits, and named Kepler-19b. Its radius is about 2.2 Earth radii and its orbit period about 9.287 days. I estimate its semimajor axis as 0.085 AU and its equilibrium temperature as 810 K.

The transit timing variations have period 316 days and amplitude 5 minutes.

This looks like a resonance -- the period is 30 times the period of planet b.

So Sarah Ballard joined John Couch Adams and Urbain Leverrier in discovering a planet using its gravitational effects on known planets.

 

[lpetrich]

But it was difficult to pin down anything further about Kepler-19c.

Radial-velocity observations, however, confirmed by b and c, and provided evidence of an additional planet, d.

The Kepler-19 System: A Thick-envelope Super-Earth with Two Neptune-mass Companions Characterized Using Radial Velocities and Transit Timing Variations - IOPscience

Planet	Period	Major Axis	Equil Temp	Mass	Radius
b	9.287 d	0.085 AU	840 K	8.4 ME	2.2 RE
c	28.73 d	0.18 AU	580 K	13.1 ME
d	63.0 d	0.30 AU	450 K	20.3 ME

All three planets are *hot*. Also, planet b is a little less dense than the Earth, and from Planet Models it must have a thick volatile layer, like a super ocean.

Planet c is in a 3:1 resonance with b, and d in an approximate 2:1 resonance with c.

 

[lpetrich]

I think I'll call it a wrap for planet-detection methods. I'll now consider what Loren Pechtel linked to a while back:
Bad Astronomy | Some exoplanets are actually so massive they're really stars | SYFY WIRE
noting
Look! Up in the sky! Is it a planet? Nope, just a star | MIT News | Massachusetts Institute of Technology
noting
Revisiting Kepler Transiting Systems: Unvetting Planets and Constraining Relationships among Harmonics in Phase Curves - IOPscience

Phil Plait seems to have moved his "Bad Astronomy" column to syfy.com - he's been in other places over the years.

 

[lpetrich]

I'll discuss the journal paper, since it's the primary source. It starts off with mentioning statistical tests for planethood, and notes that these tests sometimes have false positives, like from new parallax measurements.

Parallax is the gold standard of interstellar distance measurement. Parallaxes are primary, and all other techniques depend on parallaxes either directly or indirectly: the cosmic distance ladder. The best-known kind of parallax is over the Earth's orbit, and numerous stars have parallax measurements for them.

(distance) = 1 / (parallax)
(distance error) = (distance)² * (parallax error)

So there is a maximum distance for parallax that is the result of one's measurement errors.

Stellar Distances - Modern Developments and Limits to Parallax and Gaia Mission Science Performance - Gaia - Cosmos

• Ground-based: 0.01 arcsec - maximum distance 40 parsecs (pc)
• Hipparcos: 0.001 arcsec - max dist 400 pc
• Gaia: 0.00001 arcsec - max dist 40,000 pc

Maximum distance: for parallax 2.5 * (parallax error)

Hipparcos was an astrometric satellite that observed over 1989 - 1993, and Gaia is one that is still observing, starting in 2014.


I note in passing some additional parallax techniques.

• Moving cluster - with proper motions and radial velocities
• Visual binaries - with radial velocities
• Illumination of surrounding material by a supernova - notably for SN 1987 A

But they are much more specialized than Earth-orbit parallaxes.

 

[lpetrich]

Back to the journal paper.

Similarly, revised stellar parameters, notably those provided by the Gaia mission (Gaia Collaboration et al. 2018, 2021), can reveal misclassified planets as we will be demonstrating in this paper.

Beyond RV follow up and revised stellar parameters, phase curves, which denote photometric modulations that are synchronous with the orbital period, provide a powerful tool for gaining insights into the mutual gravitational and radiative interactions between objects in binary systems. The three primary astrophysical phenomena driving these modulations are (1) ellipsoidal variation originating from the tidal interaction, (2) beaming (or Doppler boosting) due to relativistic Doppler effects, and (3) mutual illumination due to the mutual heating of the components' atmospheres. In the past, phase curves have been used to validate exoplanets, such as Kepler-41 b (Quintana et al. 2013) and Kepler-78 b (Sanchis-Ojeda et al. 2013). For short-period candidates (typically shorter than 5 days), the phase curve can provide an independent estimate of the companion mass and complement other planet characterization techniques without requiring further observations (e.g., Faigler & Mazeh 2011; Shporer 2017).

In this paper, we reassess the planetary nature of Kepler companions in light of updated stellar catalogs and then consider how phase-curve photometry can be used to mitigate the occurrence of misclassified companions. We find that three of them, and possibly a fourth, cannot be planets or brown dwarfs and are thus stellar objects that have been misclassified as planets. Then, we use a sample of Kepler eclipsing stellar binaries to show that the higher-order harmonic seen in the phase curve of one of the misclassified Kepler planets is likely to be a manifestation of the host star's tidal distortion.

Relativistic beaming is simple. The relative luminosity change:
\( \displaystyle{ \frac{\Delta L}{L} = 4 \frac{v_{rad}}{c} } \)

It is small unless v is close to c.

For phase illumination, we have this integral:

\( \displaystyle{ L(n_i,n_f) = \int R(n_i,n_f,n) (n \cdot n_i) (n \cdot n_f) d\Omega_n } \)

integrating over the body's surface, with i = initial and f = final directions n. For Lambertian reflectance, R = constant, we have

\( L(\alpha) = R \cdot ( (\pi - \alpha) \cos \alpha + \sin \alpha ) ,\ \cos \alpha = n_i \cdot n_f \)


Turning to ellipsoidal variation from tidal effects, it is not only from the star's tidally-distorted shape, but also from the different strengths of gravity at different points on the star's surface, an effect called gravitational darkening or brightening: stronger gravity -> brighter.

•  Gravity darkening
• Rapid Rotation Distorts Bright Star Vega | Space -  Vega
• Appearance of a gravity darkened star. Upper row: Surface appearance of... | Download Scientific Diagram
• [PDF] Gravity darkening and brightening in binaries | Semantic Scholar

The equator of a rotating star has less gravity from centrifugal force, and is thus fainter. For a slow-rotating star, this effect is too small to be noticeable, but for a fast-rotating star like Vega, it is. Poles: 10,000 K, Equator: 8,000 K. From the Stefan-Boltzmann law, the equator is only 2/5 as bright as the poles. The Sun has a period of 27 days, Vega of 12.5 hours -- 88% of the star's maximum possible rotation rate. As a result, Vega has a (1-rp/re) flattening of 16%.

This also happens with tidal distortion. The poles of the resulting football shape are cooler and fainter than the equator. Thus, when looking perpendicular to the poles, the star is brighter from looking more at the equator than the poles in addition to its shape.

 

[lpetrich]

Back to the paper.

Usually, the amplitude of the ellipsoidal variation in optical wavelengths does not exceed 200 ppm in a planet–star system (see Millholland & Laughlin 2017), allowing us to rule out some of the obvious false positives.

So even for a very close Jovian planet, it's not a very large effect.

The authors then considered revisions of parameter values for the planets' stars. They compared the Kepler Input Catalog (original, 2011) and the TESS Input Catalog (version 8, 2019), catalogs of stars that these space telescopes were to look at. They found that the TESS one's radius values are biased toward larger values than the Kepler one's:

R(T)/R(K) = 1.05 + 0.25 - 0.14

with 3.5% of the stars having discrepancies larger than 2. The changes were due to using the Gaia DR2 parallaxes for the TESS catalog. Gaia DR2 contents - Gaia - Cosmos

• Ground-based: 10 mas - 0.1 kpc
• Hipparcos: 1 mas - 1 kpc
• Gaia DR2:
  • G <= 14: 0.04 mas - 25 kpc
  • G = 17: 0.1 mas - 10 kpc
  • G = 20: 0.7 mas - 1.4 kpc

G = Gaia visual magnitude (3 <= G <= 21 -- avoiding the brightest stars). Distance = reciprocal of error. mas = milliarcsecond, kpc = kiloparsec.

 

[lpetrich]

Figure 1 shows these four graphs:

1. Histogram of number of Kepler planets for each (TESS radius) / (Kepler radius) value
2. (TESS radius) / (Kepler radius) as a function of (TESS distance)
3. (TESS radius) as a function of (Kepler radius)
4. (TESS radius) / (Kepler radius) as a function of (insolation: received luminosity per unit area)

TESS and Kepler: input catalogs TIC and KIC.

The TESS/Kepler values scatter around 1 until around 1000 parsecs, then it starts to increase to around 3 at 3000 pc.

There is also some overestimation for the closer planets, though that effect is not as prominent. This may be due to some demographic effect, like more massive stars tending to have closer planets, more massive meaning more luminous and thus easier to see over great distances.

The paper then got into the three effects: ellipsoidal variations, beaming effects, and mutual illumination.

 

[lpetrich]

Now the four reclassified planets.

Kepler-854 b - "However, based on the range of estimated masses derived from the measured ellipsoidal effect in the Kepler photometry, we conclude that this companion is not a planet."

Most likely a red dwarf star with mass 0.2 Msun and radius 0.3 Rsun.

Kepler-840 b - too large and too luminous to be a planet, likely also a red dwarf star.

Kepler-699 b - too large to be a planet.

Kepler-747 b - too large to be a planet?

 

[lpetrich]

I'll collect my method classification into a big table.

Method	Observation Type	Time Behavior	Visibility Effect	Gravity Effect
Microlensing	Photometry	One-time		Bending starlight
Disk kinematics	Imaging	Stationary		Clearing a region
Direct Imaging	Imaging		Direct
Transits	Photometry		Eclipsing
Phases	Photometry		Direct
Reflection	Photometry		Reflected
Astrometric	Imaging			Offset - Transverse
Pulsing	Photometry			Offset - Radial
Radial velocity	Spectroscopy			Velocity - Radial
Relativistic beaming	Photometry			Velocity - Radial
Ellipsoidal Variations	Photometry			Tidal distortion
TTV's	(from transits)			Offset

• Time behavior is persistent and changing unless stated otherwise.
• Visibility and gravity effects are none unless stated otherwise.
• Pulsing includes pulsar pulses, normal-star pulses, and binary-star eclipses.

 

[lpetrich]

 Superhabitable planet
Source article:
In Search for a Planet Better than Earth: Top Contenders for a Superhabitable World | Astrobiology

"We argue that there could be regions of astrophysical parameter space of star-planet systems that could allow for planets to be even better for life than our Earth."

Table 2. Most Valuable Planets—Planets That Might Be More Habitable Than Earth

• In orbit around a K dwarf star
• About 5–8 billion years old
• Up to 1.5 more massive than Earth and about 10% larger than Earth
• Mean surface temperature about 5°C higher than on Earth
• Moist atmosphere with 25–30% O2 levels, the rest mostly inert gases (e.g., N2)
• Scattered land/water distributed with lots of shallow water areas and archipelagos
• Large moon (1–10% of the planetary mass) at moderate distance (10–100 planetary radii)
• Has plate tectonics or similar geological/geochemical recycling mechanism as well as a strong protective geomagnetic field

The authors then propose 24 exoplanets as possibly good candidates.

Higher mean surface temperature means that high latitudes are not very cold by Earth standards, something like what happened on the Earth itself during the Jurassic and Cretaceous, where some dinosaurs were found at high latitudes.

Lots of shallow water and lots of islands means a lot of biodiversity. Supercontinents like Pangaea are the opposite, with large desert areas in their interiors.

Turning to a large moon, the Earth's Moon is where it is from tidal drag over the Earth's history.  Tidal acceleration and  Tidal locking

\( \displaystyle{ \frac{1}{t_{drag}} \sim \frac{1}{Q} \frac{T_{rot}}{(T_{surf})^2} \left( \frac{m'}{m} \right)^2 \left( \frac{R}{a} \right)^6 } \)

Q - dissipation factor, T(rot) - rotation period, T(surf) - surface-satellite period, m - planet mass, m' - perturber mass, R - planet radius, a - perturber distance

For the Earth, t(surf) = 1.4 hours. For the Earth-Moon system and Q = 1, I find t(drag) ~ 3 billion years, and for the Earth-Sun system, t(drag) ~14 billion years.

For a larger planet or a relatively more massive moon, the tidal-drag timescale becomes less.

 

[lpetrich]

For planet modeling, I've found Planet Models and Seager2007.pdf

A size of 1.1 Earth radii for 1.5 Earth masses is about right for an Earthlike composition. That gives a surface velocity of 1.24 Earth's and an escape velocity of 1.17 Earth's.

Since mountain sizes and organism sizes are limited by 1/g (g = surface gravity), that means that mountains and trees will be somewhat shorter than on Earth, and that one's jump height is a little less.

Now to what a K star would be like. From  Main sequence

Spectral type	Temp (K)	Radius	Mass	Luminosity	MS Lif (Gyr)	Hab dist (AU)	Hab per (yr)	Log10(tdrag)
O2	50,000	12	100	800,000	0.001	890	2700	13.7
O6	38,000	9.8	35	180,000	0.002	420	1500	12.7
B0	30,000	7.4	18	20,000	0.009	140	400	10.4
B5	16,400	3.8	6.5	800	0.08	29	59	7.1
A0	10,800	2.5	3.2	80	0.4	8.9	15	4.7
A5	8,620	1.7	2.1	20	1	4.5	6.5	3.3
F0	7,240	1.3	1.7	6	3	2.4	2.9	1.9
F5	6,540	1.2	1.3	2.5	5	1.6	1.7	1.0
G0	5,920	1.05	1.10	1.26	9	1.1	1.1	0.2
G2 (Sun)	5,780	1	1	1	10	1	1	0
G5	5,610	0.93	0.93	0.78	12	0.89	0.87	-0.2
K0	5,240	0.85	0.78	0.40	20	0.63	0.57	-1.0
K5	4,410	0.74	0.69	0.16	40	0.40	0.30	-2.1
M0	3,800	0.51	0.60	0.072	80	0.27	0.18	-3.0
M5	3,120	0.18	0.15	0.0027	600	0.052	0.031	-6.1
M8	2,650	0.11	0.08	0.0004	2000	0.020	0.010	-8.0
L1	2,200	0.09	0.07	0.00017	4000	0.013	0.056	-9.0

 

[lpetrich]

The James Webb Space Telescope is proving its worth for studying exoplanets.

NASA’s Webb Detects Carbon Dioxide in Exoplanet Atmosphere

After years of preparation and anticipation, exoplanet researchers are ecstatic. NASA’s James Webb Space Telescope has captured an astonishingly detailed rainbow of near-infrared starlight filtered through the atmosphere of a hot gas giant 700 light-years away. The transmission spectrum of exoplanet WASP-39 b, based on a single set of measurements made using Webb’s Near-Infrared Spectrograph and analyzed by dozens of scientists, represents a hat trick of firsts: Webb’s first official scientific observation of an exoplanet; the first detailed exoplanet spectrum covering this range of near-infrared colors; and the first indisputable evidence for carbon dioxide in the atmosphere of a planet orbiting a distant star. The results are indicative of Webb’s ability to spot key molecules like carbon dioxide in a wide variety of exoplanets – including smaller, cooler, rocky planets – providing insights into the composition, formation, and evolution of planets across the galaxy.

A CO2 absorption line at 4.3 microns made the planet look slightly larger, going up from blocking 2.15% of its star's light to blocking 2.25% -- 0.1% more.

 

[lpetrich]

One can find which molecules have infrared vibrational lines by looking to see if those vibrations make an electric dipole. For CO2, symmetric stretching will not, but asymmetric stretching will, and the line that we see in that exoplanet is an asymmetric-stretch line. - Vibrational Modes of Carbon Dioxide - I note that CO2 has a bending line, but it's lower in frequency, higher in wavelength.

That's why we won't see IR vibrational lines for same-element diatomic molecules, like H2, N2, O2, ... But different-element ones like CO do have such lines.

Vibrational Modes of Water - H2O - three vibrational IR lines.

Since N2 does not show up very well in the infrared, one must look for small molecules that contain it and other elements, like NH3, NOx, and HCN.

Carbon dioxide presents a paradox. It requires an oxidizing environment, as opposed to a reducing one, like a hydrogen-rich one, what one would expect a gas-giant planet to be. A lot of H2 makes CO2 + 4H2 -> CH4 + 2H2O -- thus making the carbon be present in methane. Likewise the nitrogen will be present in ammonia, NH3.

 

[Swammerdami]

The James Webb Space Telescope is proving its worth for studying exoplanets.

A CO2 absorption line at 4.3 microns made the planet look slightly larger, going up from blocking 2.15% of its star's light to blocking 2.25% -- 0.1% more.

This seems very impressive, especially considering the tininess of the 2.15% to 2.25% change.

Now, if only they will spot O₂, probably both symptom of, and prerequisite for advanced life!

One can find which molecules have infrared vibrational lines by looking to see if those vibrations make an electric dipole....

That's why we won't see IR vibrational lines for same-element diatomic molecules, like H2, N2, O2, ... But different-element ones like CO do have such lines.

So, scratch my point about detecting O₂ ?

 

[Jarhyn]

The James Webb Space Telescope is proving its worth for studying exoplanets.

A CO2 absorption line at 4.3 microns made the planet look slightly larger, going up from blocking 2.15% of its star's light to blocking 2.25% -- 0.1% more.

This seems very impressive, especially considering the tininess of the 2.15% to 2.25% change.

Now, if only they will spot O₂, probably both symptom of, and prerequisite for advanced life!

One can find which molecules have infrared vibrational lines by looking to see if those vibrations make an electric dipole....

That's why we won't see IR vibrational lines for same-element diatomic molecules, like H2, N2, O2, ... But different-element ones like CO do have such lines.

So, scratch my point about detecting O₂ ?

In this case, you would need to look for oxides that only exist when there is an oxygen-rich environment, like weak oxides that in any other situation would be grabbed up by free carbon or more oxygen-hungry systems.

Unstable oxides can only exist when they can get access to the free oxygen which doesn't happen in heavily reducing environments.

 

[bilby]

The James Webb Space Telescope is proving its worth for studying exoplanets.

A CO2 absorption line at 4.3 microns made the planet look slightly larger, going up from blocking 2.15% of its star's light to blocking 2.25% -- 0.1% more.

This seems very impressive, especially considering the tininess of the 2.15% to 2.25% change.

Now, if only they will spot O₂, probably both symptom of, and prerequisite for advanced life!

One can find which molecules have infrared vibrational lines by looking to see if those vibrations make an electric dipole....

That's why we won't see IR vibrational lines for same-element diatomic molecules, like H2, N2, O2, ... But different-element ones like CO do have such lines.

So, scratch my point about detecting O₂ ?

In this case, you would need to look for oxides that only exist when there is an oxygen-rich environment, like weak oxides that in any other situation would be grabbed up by free carbon or more oxygen-hungry systems.

Unstable oxides can only exist when they can get access to the free oxygen which doesn't happen in heavily reducing environments.

Given the massive dominance of Hydrogen in the universe, most planetary environments are strongly reducing; The detection of CO₂ is itself something of a hint that the environment is more oxidising than might be expected.

Though that could have more mundane explanations, such as a higher than average metallicity for the system (which may also cause larger number or sizes of planets). If there's lots of carbon, or lots of other metals competing for hydrogen, you would expect to see lots of low-hydrogen carbon compounds, including oxides.

Of course, lots of complex carbon compounds could also imply a higher probability that life might evolve...

It's all tantalising hints at this stage.

 

[lpetrich]

• C/O AND Mg/Si RATIOS OF STARS IN THE SOLAR NEIGHBORHOOD - IOPscience
• C/O vs. Mg/Si ratios in solar type stars: The HARPS sample | Astronomy & Astrophysics (A&A)

Stars with known planets are not very different from stars in general.

C/O ratios are 0.5 +- 0.1, with the Solar System having 0.55. By number of atoms?

No C/O ratio greater than 0.8 was found.


Mg/Si ratios 1.0 +- 0.1 or 1.1 +- 0.1 (the two papers differ on the average value), with the Solar System having 1.0.

This will affect planets' mineralogy.