A new study posted on Oct. 8 suggests that heavy rocky planets, known as mega-Earths, might be an illusion. Astronomers Casey Brinkman and Nicolas Cowan said stellar activity often creates fake signals that make planets look heavier than they are. This noise makes it hard to measure their true mass accurately.
A new study suggests that some of the heaviest rocky planets ever reported may not be as heavy as they look. These so-called 'mega-Earths,' rocky planets weighing more than about 10 Earths, could be an illusion created by activity on their host stars. The paper was posted to the arXiv preprint server on Oct. 8.
A shaky transition
Planets stop being bare rocky worlds and start developing a gas envelope by accreting gas from their disk at roughly 1.5 to 1.75 times Earth's radius, corresponding to about 4 to 5 Earth masses. Theory suggests a much higher upper limit on this mass range. Even then, simulations suggest a purely rocky planet is unlikely to grow beyond about 10 Earth masses. This is because a core that massive should be pulling a thick envelope of gas from the disk it formed in.
Yet astronomers have found a few planets above this threshold that are dense and gas-free. If they are real, they would need exceptional pathways of formation, such as collisions after the gas disk disappeared, or they would have to be the stripped cores of former giant planets.
Astronomers also look at another possibility: the host star's activity. Planet masses are measured using the radial-velocity (RV) method, which measures how much a planet tugs its star back and forth. Stars with strong magnetic activity make it hard to measure the mass accurately this way. As star spots and bright patches on their surfaces rotate in and out of view, they produce wobbles that can imitate a planet's gravitational pull. This stellar 'noise' can drown out the faint signal of a small planet or sometimes exaggerate it, so planets may end up with mass estimates that are too high.
In this study, astronomers Casey Brinkman and Nicolas Cowan set out to test whether the mega-Earths are real. They started with small planets under four times Earth's size from NASA's Exoplanet Archive and kept only those whose masses were measured by the radial-velocity method. Dropping planets with weak or indirect mass estimates, they were left with 333 planets in their sample.
Eight oddballs
To spot any strange behavior, they plotted planet mass against size. Doing so, they found eight planets with rare combinations. All eight weighed more than about 10 Earth masses. Four were found to have rocky densities: WASP-84 c, TOI-1347 b, TOI-2322 c and HD 63433 b. The other four were dense planets that could hold water or other heavy volatiles but little hydrogen-helium gas: GJ 523 b, TOI-815 c, HD 18599 b and GJ 143 b.
They then checked how magnetically active each host star is, using the glow of calcium in its atmosphere. They used calcium because "Ca II H & K lines are particularly sensitive to magnetic heating in the chromosphere, producing greater flux in the cores of these lines," the researchers explain in the paper. They also looked at TESS light curves for the flicker of spots rotating in and out of view.
All eight host stars turned out to be active. The team calculated the probability of finding the four rocky mega-Earths around active stars, which turned out to be about 1%, and for all eight, the probability was just a fraction of a percent. They also found no physical reason why young, active stars should favor such planets. If anything, the formation of one such heavy planet should take longer, which would favor older stars.
"We therefore would not expect young stars to preferentially host these worlds, and predict at minimum a uniform distribution of mega-Earths across host star age and activity level if not a preference toward older systems," the researchers say.
An illusion created by stars
The researchers also pointed out some signals that looked unreliable. For three planets, the signal was too weak relative to the noise for the standard technique, called Gaussian process modeling, to recover it reliably. Other planets had orbital periods close to the star's rotation period, which made it difficult to disentangle the planet's signal from the star's. Additionally, one team measured the mass of HD 63433 b to be 37 Earth masses, while another team's data revealed around 21, making it a striking case.
Hence, the researchers concluded that the planets could appear massive because of the host stars' activity. "The most likely culprit for these larger-than-expected masses is artificial inflation from stellar activity," the researchers write, adding, "particularly for systems that have similar stellar rotation and planetary orbital periods."
They also found no "compelling evidence" that rocky planets should form above 10 Earth masses. The researchers are skeptical about the reliability of the standard technique for planets around active stars, as their activity can contaminate the mass measurements.
Further, the researchers call for repeated radial-velocity measurements with shorter time intervals between observations to track changes and incorporate the effects of stellar activity with precision. They stress the importance of planet mass measurements given how a misclassification could challenge theories of planet formation.