If you follow the fortunes of the stars closest to us, you know that Barnard’s Star has always excited interest, both because of its proximity to our system (about six light years) but also because of the early work on the star performed by Peter Van de Kamp at Sproul Observatory (Swarthmore College). That work, which ran until the early 1970s, initially appeared to show a Jupiter-class planet at the star but the results were later explained as instrumentation errors in Van de Kamp’s equipment.
It was a cautionary tale, but credit the astronomer for working tirelessly using astrometry to attempt to validate a conclusion we now take for granted: There are planets around other stars. In 2018 we seemed to have a solid detection of a much different planet candidate via Guillem Anglada-Escudé (Queen Mary University, London) and Ignasi Ribas (Institute of Space Studies of Catalonia and the Institute of Space Sciences, CSIC in Spain), indicating a super-Earth of 3.3 Earth masses in an orbit near Barnard Star’s snowline (see A Super-Earth Orbiting Barnard’s Star for that coverage), but no confirmation followed.
Indeed, we may have been looking at stellar activity in this second detection rather than a planet, according to a new paper announcing the discovery of a planet below Earth mass at the star. On the 2018 work, the paper notes that “ESPRESSO data does not support the existence of the 233 d candidate planet.” See Paul Robertson’s A very stealthy alias: the impostor planet of Barnard’s star for a detailed look at the detection and the stellar activity explanation.
But this new announcement of a Barnard’s Star planet looks to be solid. Lead author Jonay González Hernández (Instituto de Astrofísica de Canarias) and team, working at the European Southern Observatory’s Very Large Telescope (VLT) made the find with the help of ESPRESSO (Echelle SPectrograph for Rocky Exoplanet and Stable Spectroscopic Observations), the successor to the highly successful HARPS spectrograph, capable of teasing out the wobble induced in the star by a planet.
We now have a low-mass planet, as confirmed by HARPS at the La Silla Observatory, HARPS-N (on La Palma, Canary Islands) and CARMENES at the Calar Alto Observatory, Spain. Twenty times closer to Barnard’s Star than Mercury is to the Sun, the planet orbits in 3.15 Earth days and has a surface temperature around 400 K. The planet is about half the mass of Venus, or three times the mass of Mars. Says Hernández:
“Barnard b is one of the lowest-mass exoplanets known and one of the few known with a mass less than that of Earth. But the planet is too close to the host star, closer than the habitable zone. Even if the star is about 2500 degrees cooler than our Sun, it is too hot there to maintain liquid water on the surface.”
T
Image: This stunning panorama shows the Milky Way galaxy arching above the platform of ESO’s Very Large Telescope (VLT) on Cerro Paranal, Chile, where the work on the new Barnard’s Star discovery was performed. At 2635 metres above sea level, Paranal Observatory is one of the very best astronomical observing sites in the world and is the flagship facility for European ground-based astronomy. The extent of our galaxy’s cloudy and dusty structure can be seen in remarkable detail as a dim glowing band across the observation deck. Credit: ESO.
Indeed, Barnard’s Star b (which I see is being referred to simply as Barnard b) may not be the only planet here. The paper makes note of three other candidates currently under investigation using ESPRESSO. Here we have to be careful. The radial velocity data show several signals at periods less than 10 days: The paper reports periods of 3.15 d, 4.12 d, 2.34 d and 6.74 d, sorted by strength of the signals. The researchers cannot confirm these signals at this point, but are able to model a system that fits the data. Let me go a bit into the weeds here. From the paper:
[The modeled system] would correspond to a system of four sub-Earth mass planets with mp sin i = 0.32, 0.31, 0.22 and 0.17 M⊕. All candidate planetary orbits would be located inner to the habitable zone of the star, with orbital semi-major axes between 0.019 AU and 0.038 AU. Thus all the candidate planets would be irradiated more than the Earth with incident fluxes between 2.4 S ⊕ to 10.1 S ⊕, and their equilibrium temperatures, assuming albedo of 0.3, would be in between 440 K of the inner planet to the 310 K of the outer planet.
Let’s untangle this (this is how I learn things). The four potential planets that emerge from this model are described by mp sin i, which helps us determine a minimum mass (mp) for a planet. What is at stake here is the inclination angle (i) of the planet’s orbit as viewed from Earth, but because we cannot see such planets, we can go from an edge-on orbit (sin close to 1) to a face-on orbit, where sin i is small and the mass of the planet is much higher. So the numbers above refer to minimum masses that could be higher depending on how the system is tilted to our point of view. If these other worlds exist, they’re all too close to the star to fit the liquid water habitable zone. Indeed, the S value in the quote refers to solar flux, which in the case of the hypothetical planets would be 2.4 to 10.1 times the stellar radiation that Earth receives from the Sun.
In any case, the authors are careful to add that confirming an actual four-planet system at Barnard’s Star would take many more observations using ESPRESSO:
These observations would need to be done with sufficient cadence to sample these planet periods as well as with enough baseline to be able to properly model the activity of the star, in particular, those activity signals associated with the stellar rotation.
So the hunt continues, encouraged by the one newly confirmed planet, as we scour this and other nearby red dwarfs for evidence of small rocky worlds. We can look ahead to ANDES, the ArmazoNes high Dispersion Echelle Spectrograph, which will be used in conjunction with the European Southern Observatory’s Extremely Large Telescope, a 39-meter instrument that will be the largest visible and infrared light telescope in the world. Located at Cerro Armazones in Chile’s Atacama Desert, the telescope should see first light as soon as 2028.
The paper is Hernández et al., “A sub-Earth-mass planet orbiting Barnard’s star,” Astronomy & Astrophysics Volume 690 (October 2024). Full text.
Finally! Some good news about a more robust detection of a planet orbiting the nearby Barnard’s Star. For a detailed history of past efforts to find planets orbiting Barnard’s Star up to the now spurious detection in 2018, see this article: https://www.drewexmachina.com/2018/11/16/our-new-neighbor-orbiting-barnards-star-details-historical-background/
Great article and thank you for the fast publishing, I only saw it on Facebook 2 hours ago! It would be nice if someone could make an image of these four planets with the habitable zone. Since these planets may be small has anybody looked for their transits? How well do they fit for resonance orbits and what would be next in that order? Should be very interesting to see how this developers.
Found this image at the end of the paper “A sub-Earth-mass planet orbiting Barnard’s star,” Astronomy & Astrophysics Volume 690 (October 2024) that has the all four plant’s orbits and shows the habitable zone.
https://pbs.twimg.com/media/GY2RvY9acAAs5GT?format=jpg&name=medium
These planets seem tightly packed, the habitable zone beyond the predicted candidates. Would there be a resonance similar to the Trappist system? so tighty packed, they must influence each other.
I wonder if it ends before the habitable zone, and the star cannot have planets further out, or that the observation time was just to small to tease out additional candidates further out…
(My feeling is that with the limited search time in general, many planets with orbits longer then earth’s are still to be found, although that doesn’t apply to stars this small)
Some points:
We should be able to get a much better idea of the planets mass if we knew were Barnard’s equatorial plane was relative to Earth as the planets are almost certainly orbiting on the star’s equatorial plane. I did a quick search but couldn’t find it.
These planets are probably nearly airless rocky balls given i) the age of the system, ii) the amount of UV flaring they have been subjected to and iii) their level of instellation, but there is the possibility of an SO2-CO2 atmosphere particularly on the outer ones.
The big problem with characterizing planets to find Earth-like ones is that we can just detect Earth mass ones and we can detect super-Earths, but not sub-Earths (although, Barnard’s planets may qualify as super-Mars.) It appears as in the Trappist system that most Earth-sized planets come with a much higher volatile inventory than Earth rendering them at best an ocean planet. Sub-Earths, particularly those around M-dwarfs, are more likely to lose their excess volatiles are finish up like Earth. And, I think there’s a good chance that solitary M-dwarfs that we have not detected planets around will contain systems of sub-Earths.
I think I am getting “exoplanet discovery fatigue”. Each new discovery feels more like stamp collecting rather than any advances in planet taxonomics and development. Rather like a pre-Linnaean state of taxonomies of life.
What we need beyond size, mass, orbita, temperature, and a few atmospheric measurements is more detail on the planets so that we can better classify them, and even better, determine habitability and even whether they are inhabited. I look forward to data and details rather than speculations based on possibilities.
I get the sentiment, and I know your expertise from the comments here, but on the other end you argument feels like a demand for results rather then an exploration unknowns.
Astronomers are still only in the infant stages of finding planets, as not all formation patterns are known. The big take away here is how small the discovered exoplanet is, indicating there might be a lot of those out there and the focus is biased towards bigger planets because of the limitations of our tools: many small stars that seem to have no planets might harbour a litter of smaller unseen rocky worlds rather then a bunch of super earths.
Even without further characterisation, there is stuff to learn even if there’s no answers to the deeper questions of the possibilities for life, just like SETI is just a vehicle for science but should not be expected to deliver immediate or final results on it’s premise.
@Ivar, @Ron
At this point, we have in excess of 4000 exoplanets discovered, mostly from the Kepler telescope. It seems reasonable that this could easily reach 100,000 with more telescopes/observation time of the same basic type of instrument, plus the TESS data, plus other planet-finding specialist [space] telescopes.
JWST and other telescopes will add to the atmospheric data.
But we don’t yet have any clear way to establish the status of the terrestrial worlds – Earthlike, Venuslike, Marslike, Hycean, etc. We have a zoo of worlds, especially gas and ice giants that are hot rather than cold.
So suppose we acquire 100,000 exoplanets. Is that enough to really start classifying them more carefully to be able to create statistics on what their likely surface conditions are? What remote data would be sufficient to make informed statements about their surfaces? [Even with orbiters and landers, we cannot be sure Mars is/isn’t inhabited, but at least we know what the surface conditions are.] Are these worlds so remote that even SGL telescopes would be inadequate to determine surface conditions, or would future higher-resolution spectral analysis be sufficient?
Are there good overview sources of information to explain what is and is not possible with remote observation assuming ideal data is possible?
Are our models of planet formation still too limited to be able to explain how a type of exoplanet forms and predict its future in its system [barring catastrophes]?
I admit my interest is more in regards to astrobiology, especially detecting the presence of life, and ideally gaining some clues to its basic biology. What can we hope for with remote data apart from our system and possibly interstellar visitors, as we will not be able to get samples of life from exoplanets for a very long time? Biosignatures from atmospheric gas analysis provide hints, maybe even good ones, and we might get data from spectra directly from life on the surface. But beyond that?
Will we be able to build models to infer the probability of life on different types of worlds, especially types that do not exist in our system?
IOW, can we look forward to being able to be more systematic about planet types and even habitation, or are we forced only to make catalogs?
Alex, what’s the point of asking for the impossible? We can only do what we can do. But what we can do and have done is fabulous.
It will take time before we can do what you (and most of us) want. Overreaching without the prerequisite resources or technology typically creates fiascos that derail realistic projects that can make actual progress, incremental though it be.
If you disagree, what do you suggest be done that isn’t being done? There are examples though none that are really doable (at present) and that will achieve the objectives you enumerated.
Is it impossible?
Once upon a time, the stars were just points of light in the sky (“campfires of the Gods in the sky”) Some of those points moved. Careful plotting of their position, telescope observations (seeing things closer), and Laws of Motion ensured we could determine they were bodies orbiting our sun.
By the early 20th century, careful cataloging of stellar spectra resulted in the Hertzsprung–Russell diagram that provides a framework to classify stars, and also their evolution. No local samples are needed.
We now have several means of detecting exoplanets. Transits, doppler shift, and even direct observation, including the absence of dust in dust clouds of newly forming systems). Space-based telescopes with separate shades will enhance the ability to image planets.
With high-resolution spectra, we can expect to describe planetary atmospheres and surfaces. Monitoring planetary brightness can infer land and ocean distributions. Earth observation satellites can show objects on the ground, and spectra can distinguish vegetation types.
Therefore, with large enough telescopes, without any new science or speculative imaginary technology, we could, in principle, not just “stamp collect” planets, but determine their habitability and habitation.
It would be nice to have the equivalent of the H-R diagram for planets, albeit a more complex one for the greater complexity of planets.
I am not asking for this today, but as with other sciences, a path toward being able to answer “big questions” is desirable.
In biology, Darwin provided the theoretic framework to understand evolution. Genetics provided one means of understanding teh mechanism, while taxonomy provided the framework to classify species. Paleontology provided the “telescope” into the deep time of actual organisms. DNA and subsequent gene and genome sequencing can now almost instantly classify any organism – no field guide is needed if you can acquire the specimen. Indeed sampling the environment can detect species without direct acquisition, a technique that is now used to track infectious diseases in wastewater, and detect new species in fresh and seawater. We have surpassed the world of GATTACA in all but sequencing speed.
[Exo]planetology may be a new science still in the stamp-collecting phase, just as paleontology was before we understood deep time and evolution. Fossil marine shells in rocks on mountains were once confusing. Large dinosaur bones were interpreted by some as “dragons” and even in the C19th, reconstructions were badly in error as the sculptures in London’s Crystal Palace park attest.
“All” I am asking for is some overview education of where we are on planetology, and possible technology to develop to provide the frameworks to understand exoplanets, their development, and ultimately whether they can, or do, harbor life. Gene and genome sequencing was even a theoretical notion before the concept of genes was even developed, DNA’s structure elucidated, and the connection between DNA and genes understood (at the same time as its structure was determined. Just a decade earlier, Schroedinger speculated that genetic information would reside in some aperiodic structure).
When I was at university, geology seemed rather like a catalog, with just 3 broad rock classifications. That has changed in the last half-century. There is even a planetary evolution framework for minerals that includes the appearance of life that creates many of the minerals we know about. [If we had a 3-D [Raman] spectrum analyzer that could be used to detect all the minerals on a planetary surface and in its crust, we could, theoretically, determine if a world ever had life in its past, even if it appeared sterile.]
I may be impatient, but am I unreasonable?
“Is it impossible?”
You quoted me out of context. I am referring to what is possible *now*. I am no better at predicting the future than anyone else.
“I may be impatient, but am I unreasonable?”
Impatience is understandable (after all, we don’t live forever), however impatience is not a strategy. We want nice things and can’t always have them. At least not yet.
Better data will come with time, which brings new capabilities that we currently do not have. That’s really all that I’m saying.
We were misunderstanding each opther. I was channeling Arthur C Clarke’s “Profiles of the Future: An Inquiry into the Limits of the Possible”.
I don’t expect to see many possible technologies in the remaining decade or so of my life, but I do think most of the interesting ones I have read about are possible, even if unattainable with our current level of technology. Stanislaw Lem’s “Summa Technologiae” is remarkably prescient in outlining possible technologies, especially computers, given it was written in the 1960s and makes HAL 9000’s “intelligence” seem positively primitive.
Even if optical interferometry between telescopes is not possible today, the collation of data to first image a black hole using a “sneakernet” implies that we could build telescopes with the theoretical primary lens diameter of 100s of millions of km by harnessing space-based telescopes in orbit around the sun, a potential resolving power greater than the SGL we so fondly talk about. In brain science, for years the only organism that we had a complete connectome of is C. elegans with just 302 neurons and ~ 20 synapses per neuron). But just published is the connectome of the fruit fly D. melanogaster with over 139,000 neurons and a far greater number of synapses per neuron. How long before we map the connectomes of vertebrates from fish to humans? Does this suggest the possibilities of true “brain caps” that would make Neuralink look almost medieval in its crudeness and performance limitations? Is there any way to create artificial brains with the equivalent of that connectome density?
Perhaps your expectations are unreasonably high. We cannot expect some great revelation for each exoplanet discovery. That only leads to disappointment. However, the continuing stream of discoveries does lead to greater statistical significance for various classes and prevalence of planets and planetary systems. Each datum, other than the first number of them years ago, is not in itself very interesting.
I think it is natural for discoverers and other scientists to speculate about what is hinted at in the limited data set. But that also generates hypotheses and plans for new instruments to fill the data gaps. This is a marathon, not a sprint.
This reminds me of how I felt about astronomy in the 1980s. “What a boring science — all you can do is catalog more stars that you’ll never go to, and wonder if they have planets that you’ll never see!” Yet in the meanwhile, the good people were making crazy attempts to spot planets … everywhere. Lunatics claiming to see planets orbiting a pulsar, of all the impossible places! There’s a limit to science news, beyond which no reasonable person can be expected to believe, and the truth is usually “out there”.
Exoplanets are getting to be old hat, sure. But relative to stars, they have some things going for them to hold our interest. They’re made of a vast range of chemicals, which can be analyzed from that single pixel of light. By comparison to a metallicity rating, they might be described in more interesting ways: carbon planets, hcean worlds. Conclusions can be made about the atmosphere, ranging from “rain of molten iron” to “oxygen and water vapor in the atmosphere”. There is the possibility of finding technosignatures. Some people even have managed to wring out wind speed and direction from their one pixel of light. Imagine what they could do with a VLBI array spread around beyond the orbit of Jupiter…
For another way to look at this development:
It’s been a while since “doppler detection” of planets made headlines. When most of the attention was paid to the results was from the mid 1990s and a decade thereafter. But for another perspective, look at it this way: The stellar primary is observed to have doppler shifts of observable atmospheric lines. And if the lines are attributed to binary system motions with an unseen object, without a second means of detection, the inclination of the binary system motion remains unknown to the viewer.
Many astronomical charting conventions ( e.g., the HZ Diagram) would strike outsiders or even regular participants as counter intuitive. So, in this case inclination is calculated as tilt away from a cosmic plaster wall perpendicular to the line of sight. Were we talking about the ecliptic or an equatorial plane, we have all become used to speaking of inclination from that. And if we placed a satellite in orbit over the moon’s north and south pole, we would not describe it as inclination zero. So, ten to one I would bet, when one hears of doppler inclinations, the mind has a solar system ecliptic plane picture come into view.
And if that were the case, we could be talking about Barnard Star transits and only use the doppler data to check.
What we do have (evidently) are doppler shifts of Barnard’s star spectral lines due to several “reasons”, traceable to several planets. The doppler shifts are proportional to velocity shifts forward and aft of Barnard inversely proportional to the mass ratio with it of the individual planets ( more than one sine wave). Taking the planets individually, the star and planet are orbiting about a barycenter with the star much closer proportional to the mass ratio involved.
But the spectral shift does not reveal the projection of this radial motion.
Now if the inclination were nearly zero degrees, it would be hard to detect any doppler shift…. Unless the planet concerned was extremely massive. And when you pursue that notion, at some point you would have to say that the star could be better described as orbiting the planet rather than vice versa. Perhaps with a black hole involved, for example, that might work.
The other extreme is 90 degrees inclination. We see the doppler effect of a planet that orbits without up and down oscillation. Yet we have no transit…
Consequently, if astronomers divide the difference they end up with an inclination of about 45 degrees. Coming up with the most likelihood case. Which would give a mass the “square root of 2” times the value for a planet in a plane that would allow transit. 1.414 the mass if observed on edge. Unless there are other statistical considerations in the paper, I would suspect that the masses for the observed planets would be derived from an assumed 45 degree inclination angle.
What with transits the basis of the majority of exoplanet detections, I suspect that doppler’s role is still there but refining mass estimates on objects with diameter estimates based on extinction curves. Astrometric observations could do similar mass determinations, but would be a slower process. In fact, that’s what the Sproul Observatory had been doing. What I had heard was that sometime before the mid 1970s a maintenance overhaul and re-assembly had introduced a “systematic error” negating the cumulative results.
Barnard b 2024
Congratulations to the team who announced this finding.
And indeed, detection of Earth size planets is difficult even at low mass stars.
This one is so close it give a more clear signal, but they’re cautious and don’t declare any discovery for the other possible detections.
This is similar to the recent paper on planets at Teegarden where they only announced one, where I had a look at the data and was surprised they refrained from any announcement of the third planet of ½ Earth mass in the habitable zone.
But in the case with Barnard’s star we see the reason why, since the announcement of a super Earth planet in 2018 now is in question. And indeed it cannot be seen in the new dataset. But there’s also additional data here, with one peak ~125 days followed by one caused by Barnard’s star who is supposed to be at a 140 day cycle.
As for the low mass it is entirely correct it’s one of the least massive found so far, the smallest one no larger than our own moon. And it was detected by a different method, timing of a pulsar, so that one is a pulsar planet – or perhaps not even that, with such a low mass it might perhaps better be designated a dwarf planet.
As for the posts by Tolley above, I can only say that this science indeed is in the stamp collection phase. Since we simply do not have the instruments needed to get any information about these worlds. And only in the very best cases been able to even catch the faintest glimpse of these planets – where most either are forming right now, or so close to their star that they’re extremely hot so they provide a strong enough signal / IR radiation.
We will have to wait until there either is instruments like Pollux or a dedicated mission like HWO (Habitable Worlds Observatory) before we will get enough information to actually say a planet is potentially habitable.
For this we need at least thousands of samples, probably a magnitude more, for among the planets orbiting red dwarf M stars we’re trying to find an exceptionally rare case where the planet have gotten an nearly unreasonable amount of volatiles from the start so that not all have been boiled off during the active flare stage the M-stars initially got. And then one that also have some sort of asynchronous rotation, either through continuous winds, or by always showing one hemisphere to a planet with a nearby orbit ….or like Mercury having a 3/2 resonance – and since the entire orbit is counted in days that would actually be enough, but also this will be very rare indeed.
Something to read while we wait for HWO – or the Europa launch on Oct 10.
https://arxiv.org/abs/2410.00213
A bit of hope for possible asynchronous rotation
https://arxiv.org/abs/2410.00739
And lastly for those who get froth around the mouth over me using the term dwarf planet, lets label this one ‘Plutos revenge’: https://arxiv.org/abs/2409.16354
Suppose for example Woodward’s MEGA drive worked and we could build a probe that could get to Barnard’s star in about 20 years and send back data. I wonder how current sensor technology on such a probe would compare to the remote capabilities we might have in that time frame?
Barnard’s star should be in range of direct exploration relatively soon,
If we can use laser sails to send a swarm of tiny probes to Proxima Centauri in 20 years, we can send a similar swarm to Barnard’s star in 30 years, and send back data at roughly 1/2 the rate. If the expeditions are sent close together in time, the Barnard data will come back roughly a decade after the Proxima data.