HD 219134, an orange K-class star in Cassiopeia, is relatively close to the Sun (21 light years) and already known to have at least five planets, two of them being rocky super-Earths that can be tracked transiting their host. We know how significant the transit method has become thanks to the planet harvests of, for example, the Kepler mission and TESS, the Transiting Exoplanet Survey Satellite. It’s interesting to realize now that an entirely different kind of measurement based on stellar vibrations can also yield useful planet information.
The work I’m looking at this morning comes out of the Keck Observatory on Mauna Kea (Hawaii), where the Keck Planet Finder (KFP) is being used to track HD 219134’s oscillations. The field of asteroseismology is a window into the interior of a star, allowing scientists to hear the frequencies at which individual stars resonate. That makes it possible to refine our readings on the mass of the star, and just as significantly, to determine its age with higher accuracy.
KPF uses radial velocity measurements to do its work, a technique often discussed in these pages to identify exoplanet candidates. But in this case measuring the motion of the stellar surface to and from the Earth is a way of collecting the star’s vibrations, which are the key to stellar structure. Says lead author Yaguang Li (University of Hawaii at Mānoa):
“The vibrations of a star are like its unique song. By listening to those oscillations, we can precisely determine how massive a star is, how large it is, and how old it is. KPF’s fast readout mode makes it perfectly suited for detecting oscillations in cool stars, and it is the only spectrograph on Mauna Kea currently capable of making this type of discovery.”
Image: Artist’s concept of the HD219134 system. Sound waves propagating through the stellar interior were used to measure its age and size, and characterize the planets orbiting the star. Credit: openAI, based on original artwork from Gabriel Perez Diaz/Instituto de Astrofísica de Canarias. The 10-second audio clip transforms the oscillations of HD219134 measured using the Keck Planet Finder into audible sound. The star pulses roughly every four minutes. When sped up by a factor of ~250,000, its internal vibrations shift into the range of human hearing. By “listening” to starlight in this way, astronomers can explore the hidden structure and dynamics beneath the star’s surface.
What we learn here is that HD 219134 is more than twice the age of the Sun at about 10.2 billion years old. The age of a star can be difficult to determine. The most widely used measurement involves gyrochronology, which focuses on how swiftly a star spins, the assumption being that younger stars rotate more rapidly than older ones, with the gradual loss of angular momentum traceable over time. The problem: Older stars don’t necessarily follow this script, with their spin-down evidently stalling at older ages. Asteroseismology allows a more accurate reading for stars like this and provides a different reference point, providing that our models of stellar evolution allow us to interpret the results correctly..
We need to track this work because how old a star is has implications across the board. For one thing, understanding basic factors such as its temperature and luminosity requires a context to determine whether we’re dealing with a young, evolving system or a star nearing the transition to a red giant. From an astrobiological point of view, we’d like to know how old any planets in the system are, and whether they’ve had sufficient time to develop life. SETI also takes on a new dimension when considering stellar age, as targeting older exoplanet systems allows us to put our focus on higher priority targets.
Yaguang Li thinks the KPF work brings new levels of precision to these measurements, calling the result ‘a long-lost tuning fork for stellar clocks.’ From the exoplanet standpoint, stellar age is also quite informative. For the measurements have allowed the researchers to determine that HD 219134 is smaller than previously thought by about 4% in radius – this contrasts with interferometry measurements that measured its size via multiple telescopes. A more accurate reading on the size of the star affects all inferences about its planets.
That 4% difference, though, raises questions, and the authors note that it requires the models of stellar evolution they are using to be accurate. From the paper:
We were unable to easily attribute this discrepancy to any systematic uncertainties related to interferometry, variations in the canonical choices of atmospheric boundary conditions or mixing-length theory used in stellar modeling, magnetic fields, or tidal heating. Without any insight into the cause of this discrepancy, our subsequently derived quantities and treatment of rotational evolution—all of which are contingent on these model ages and radii—must necessarily be regarded as being only conditional, pending a better understanding of the physical origin for this discrepancy. Future direct constraints on stellar radii from asteroseismology (e.g., through potential breakthroughs in understanding and mitigating the surface term) may alleviate this dependence on evolutionary modeling.
So we have to be cautious in our conclusions here. If indeed the tension between the KPF measurements and interferometry is correct, we will have adjusted our calibration tools for transiting exoplanets but still need to probe the reasons for the discrepancy. That’s important, because with tuned-up measurements of a star’s size, the radii and densities of transiting planets can be more accurately measured. The updated values KPF has given us – assembled through over 2000 velocity measurements of the star – point to a significant aspect of stellar modeling that may need further adjustment.
The paper is Yaguang Li et al., “K Dwarf Radius Inflation and a 10 Gyr Spin-down Clock Unveiled through Asteroseismology of HD 219134 from the Keck Planet Finder,” Astrophysical Journal Vol. 984, No. 2 (6 May 2025), 125 (full text).
I don’t know the background science, but was our sun used to calibrate these vibrations for size, mass, age, etc.?
For stars with more frequent flares, does the impact of the flare[s] on the Doppler readings get filtered out as low-frequency noise?
I am always intrigued by a technique that seems almost outlandish, yet can be used as an orthogonal approach to validating inferences about the features of an object. We seem to have so few at present for the remote detection of life, mostly looking for biosignature gases and disequilibria of the atmospheric gases. If we are lucky, we may detect some features on the surface if life is similar to terrestrial life, like the chlorophyll “red edge”. I hope that in the future, we may develop other techniques to bolster these approaches.
I just briefly skimmed the paper. These acoustic signals have been collected for a small number of stars, of course including the Sun. They also only observed this K star for a few day. I expect that there is ample opportunity to refine the models to achieve better accuracy. They made an effort to calibrate their conclusions against, as you say, orthogonal techniques.
It is interesting that, if ultimately more reliable than other techniques, asteroseismology measurements change the determination of exoplanet masses and radii (from the paper conclusions). That could be important.
May 12
Want to find life? You’ll want to compare several exoplanets in the same system.
Most astronomers agree that life is likely common throughout the universe. While Earth is the only world known to have life, we know that life arose early in our world, and the building blocks of life, including amino acids and sugars, form readily. We also know there are countless worlds in the cosmos that might be home for life.
But just because life is likely, that doesn’t mean proving it will be easy. Many of the biosignatures we can observe can also have abiotic origins. So how can we be sure? One way is to compare our observations of a habitable world with other worlds in the system.
Our own solar system is a good example of this. If alien astronomers light-years away were to observe Earth’s atmosphere as it transited the sun, they would find the presence of oxygen, water vapor, and methane, all of which can indicate the presence of life. This would suggest the presence of life but not prove it. However, if they compared Earth’s atmosphere to those of Mars and Venus, Earth would stand out.
Our sibling worlds have dry atmospheres of mostly carbon dioxide. Since planets of a solar system have similar chemical compositions, the fact that Earth’s atmosphere stands out makes a strong case for the presence of life. If Earth, Venus, and Mars all had atmospheres rich in water and oxygen, that would weaken the case for life on Earth.
This is the idea behind a new study posted on the arXiv preprint server. The authors propose that rather than looking at the atmospheres of individual worlds, we should look at the atmospheres of several planets within a system. Since most of the worlds are likely barren, a world with life will stand out.
The team looked at our solar system and the Trappist-1 system. For our system, the results are glaringly obvious. Earth’s atmosphere is so unique that life is easy to detect compared to the rest of the system.
For Trappist-1, things are likely more subtle. The system has seven Earth-sized worlds, and, since they all orbit close to their red dwarf star, they are likely tidally locked. As the authors show, even if one of the Trappist worlds harbors life, its atmosphere won’t necessarily stand out the way Earth does.
So the authors propose using the atmospheres of all seven worlds to form what they call an abiotic baseline. They consider several molecules that are relatively easy to detect in exoplanet atmospheres, including oxygen, methane, nitrous oxide, and phosphine.
Taken individually, each of these molecules has both biotic and abiotic origins, so the fact that a planet’s atmosphere contains them is not conclusive. But with an abiotic baseline for the system, astronomers could identify a planet that is a statistical anomaly. If a Trappist planet has an anomalous number of these molecules, it would be strong evidence for the presence of life.
https://phys.org/news/2025-05-life-youll-exoplanets.html
If AI represents the final stage of advanced life, would most exist on extremely cold worlds where their superconducting circuits function optimally?
I think looking for AIs (or equivalent) in the outer regions of a planetary system would be interesting. IR sources that far out from a main sequence star of (around the Sun’s age) would be hard to assign to a natural object. Impacts between comets/asteroids should be rare, background sources could be eliminated after repeated observations.
One comment about this article – if their asteroseismology model results are 4% smaller than multiple interferometry measurements with multiple telescopes… I’d say that their model needs work, not the interferometry, which is a direct measurement.
I found a 2018 movie called *Clara* that is very similar to a film you recommended to me a year or two ago.
https://www.youtube.com/watch?v=-jwI0W-ZBOM
Both movies involve an extraterrestrial civilization orbiting a moon around another planet, which the astronomers in each film discover. They are so alike that I had to watch the end of *Clara* to see if it was the same movie! I can’t remember the name of the film you told me about, and I would greatly appreciate it if you could refresh my memory. Thanks
The link to the movie on Youtube is for “Voyager” which is the alternative title of “Clara”. It does seem very familiar…
This will be a good target for Plato (PLAnetary Transits and Oscillations of stars).
https://www.esa.int/Science_Exploration/Space_Science/Plato_factsheet
Could there be techno signatures if any effort to save the species on these planets? Like Noah’s Ark 2.
Second planet around Kepler 452 and it is in the habitual zone.
https://www.animalsaroundtheglobe.com/nasa-confirms-the-discovery-of-a-second-earth-like-planet-nearby-1-329738/
That 10 second audio clip if expanded 250,000 times will give 29+ days, almost a month. Very doubtful that it was watched continuous and uninterrupted for almost a month!
At what point between the “building blocks” and living organisms can one place the transition between non-life and life? Much might depend on the molecular biochemical sophistication of the observer. Unequivocal technosignature gases might be one of the few reliable indications of life.
Good point! I agree with your math, and according to the article they observed “approximately 9 hr of observations per night over four consecutive nights”, which gives 36 hours. But what my ears hear is about 67 putts at regular intervals; for those to be “every four minutes” the recording would have to be based on just 4.5 hours from one night. I don’t know that’s the case, but at least it would explain why we don’t see or hear a cut mentioned.