Nearby Ancient Star is Almost as Old as the Universe

A metal-poor star located merely 190 light-years from the Sun is 14.46+-0.80 billion years old, which implies that the star is nearly as old as the Universe!  Those results emerged from a new study led by Howard Bond.  Such metal-poor stars are (super) important to astronomers because they set an independent lower limit for the age of the Universe, which can be used to corroborate age estimates inferred by other means.

In the past, analyses of globular clusters and the Hubble constant (expansion rate of the Universe) yielded vastly different ages for the Universe, and were offset by billions of years! Hence the importance of the star (designated HD 140283) studied by Bond and his coauthors.

“Within the errors, the age of HD 140283 does not conflict with the age of the Universe, 13.77 ± 0.06 billion years, based on the microwave background and Hubble constant, but it must have formed soon after the big bang.” the team noted.

Metal-poor stars can be used to constrain the age of the Universe because metal-content is typically a proxy for age. Heavier metals are generally formed in supernova explosions, which pollute the surrounding interstellar medium. Stars subsequently born from that medium are more enriched with metals than their predecessors, with each successive generation becoming increasingly enriched.  Indeed, HD 140283 exhibits less than 1% the iron content of the Sun, which provides an indication of its sizable age.

HD 140283 had been used previously to constrain the age of the Universe, but uncertainties tied to its estimated distance (at that time) made the age determination somewhat imprecise.  The team therefore decided to obtain a new and improved distance for HD 140283 using the Hubble Space Telescope (HST), namely via the trigonometric parallax approach. The distance uncertainty for HD 140283 was significantly reduced by comparison to existing estimates, thus resulting in a more precise age estimate for the star.

Age estimate for HD 140283 is 14.46+-0.80 Gyr.  On the y-axis is the star's pseudo-luminosity, on the x-axis its temperature.  An evolutionary track was applied to infer the age (credit: adapted by D. Majaess from Fig 1 in Bond et al. 2013, arXiv).
HD 140283 is estimated to be 14.46+-0.80 billion years old. On the y-axis is the star’s pseudo-luminosity, on the x-axis its temperature. Computed evolutionary tracks (solid lines ranging from 13.4 to 14.4 billion years) were applied to infer the age (image credit: adapted from Fig 1 in Bond et al. 2013 by D. Majaess, arXiv).

The team applied the latest evolutionary tracks (basically, computer models that trace a star’s luminosity and temperature evolution as a function of time) to HD 140283 and derived an age of 14.46+-0.80 billion years (see figure above).  Yet the associated uncertainty could be further mitigated by increasing the sample size of (very) metal-poor stars with precise distances, in concert with the unending task of improving computer models employed to delineate a star’s evolutionary track.  An average computed from that sample would provide a firm lower-limit for the age of the Universe.  The reliability of the age determined is likewise contingent on accurately determining the sample’s metal content.  However, we may not have to wait long, as Don VandenBerg (UVic) kindly relayed to Universe Today to expect, “an expanded article on HD 140283, and the other [similar] targets for which we have improved parallaxes [distances].”

As noted at the outset, analyses of globular clusters and the Hubble constant yielded vastly different ages for the Universe.  Hence the motivation for the Bond et al. 2013 study, which aimed to determine an age for the metal-poor star HD 140283 that could be compared with existing age estimates for the Universe.  The discrepant ages stemmed partly from uncertainties in the cosmic distance scale, as the determination of the Hubble constant relied on establishing (accurate) distances to galaxies.  Historical estimates for the Hubble constant ranged from 50-100 km/s/Mpc, which defines an age spread for the Universe of ~10 billion years.

Age estimates for globular clusters were previously larger than that inferred for the Age of the Universe from the Hubble constant (NASA, R. Gilliland (STScI), D. Malin (AAO))
Age estimates for the Universe as inferred from globular clusters and the Hubble constant were previously in significant disagreement (image credit: NASA, R. Gilliland (STScI), D. Malin (AAO)).

The aforementioned spread in Hubble constant estimates was certainly unsatisfactory, and astronomers recognized that reliable results were needed.  One of the key objectives envisioned for HST was to reduce uncertainties associated with the Hubble constant to <10%, thus providing an improved estimate for the age of the Universe. Present estimates for the Hubble constant, as tied to HST data, appear to span a smaller range (64-75 km/s/Mpc), with the mean implying an age near ~14 billion years.

Determining a reliable age for stars in globular clusters is likewise contingent on the availability of a reliable distance, and the team notes that “it is still unclear whether or not globular cluster ages are compatible with the age of the Universe [predicted from the Hubble constant and other means].” Globular clusters set a lower limit to the age of the Universe, and their age should be smaller than that inferred from the Hubble constant (& cosmological parameters).

In sum, the study reaffirms that there are old stars roaming the solar neighborhood which can be used to constrain the age of the Universe (~14 billion years). The Sun, by comparison, is ~4.5 billion years old.

The team’s findings will appear in the Astrophysical Journal Letters, and a preprint is available on arXiv.  The coauthors on the study are E. Nelan, D. VandenBerg, G. Schaefer, and D. Harmer.  The interested reader desiring complete information will find the following works pertinent: Pont et al. 1998, VandenBerg 2000, Freedman & Madore (2010), Tammann & Reindl 2012.

51 Replies to “Nearby Ancient Star is Almost as Old as the Universe”

  1. Please help me out. I do see a conflict in this sentence:”age inferred for the star (14.46+-0.80 billion years) does not conflict with the latest estimates (2013) for the age of the Universe stemming from WMAP’s survey of the cosmic microwave background, which implies an age of 13.74+-0.11 billion years”.
    So what is the real non conflicting age?

    1. Taking into account the uncertainty if each number, there is no conflict. When the ranges of two values overlap, they are said to be consistent.

      1. It can’t be and it isn’t. See Stephan’s comment and re-look at the error bars for the two independent methods of age determination. The WMAP estimate of the universe’s age has the smaller error bars.

      2. Addendum: if you estimate one number to be 10 plus/minus two and another estimate puts it at 8 plus/minus 1.6 then the answers are consistent because there is commonality among the estimates. (8 to 9.6)You can estimate something to be older than the universe as long as the minus takes it back within agreed parameters.

    2. Oldest possible age of the Universe 13.74+0.11 = 13.85 billion years.
      Youngest possible age of the star 14.46-0.8 billion years =13.66 billion years.
      The star could have formed 0.19 billion “years” after the big bang.

      1. Clearly the actual age of the star must be near the smaller value. This seems to put some bounds on cosmic time scales. This star has low metallicity, but it must have enough to be main sequence. The question is whether it comes from the remains of ancient PopIII stars. This star must have enough elements heavier than hydrogen to be opaque and exist as a main sequence or PopII star.


    3. Damn, I was thinking the same thing when I 1st read the headline on this topic. Then I thought about a black hole picking up a star in its gravity. The more we find out things, the more it adds and harbors more questions. Speculations galore!

    4. UNCERTAINTY!!! Notice those “+/-” signs? They’re not there for show, they are there to give you a measurement of error. The fact that the two ranges overlap without even going two standard deviations away, there is no statistical difference between the age given by this star, and the age given by WMAP.

    5. I think I can see the problem people are having with this result. It isn’t just about error bars – it’s more how could a model that predicts the age based on the evolution of metals from the ancient and short-lived pop III stars shortly after the Big Bang predict a date Before This Process Even Started? I, myself, did have a short WTF moment. But, there is a logical explanation…

      Before there were stars, there was hydrogen, some deuterium, maybe a tiny amount of helium, and very little else. So that would normally be our time zero point. It is possible that there are pockets of that primordial gas surviving today. If it were all rolled into a star, it would make one of those primordial stars. The gas would be much colder today, so a little star would hold together, where the original pop III stars condensed from hotter gas and had to be really massive or the gases would just stream away. But there is no mechanism that we know of today that would collect this thin cold material and concentrate it in one place to form a new star.

      The earliest stars we see today are pop II stars. These are seeded by the shockwaves and the material when the pop III stars exploded. If we model the rates these pop III stars went supernova, and assume their metals get mixed in, then we can come up with a graph with how the metal content should have varied with time. if we measure the spectra of the star, we can guess the metal content it has now, and knowing how big and bright the star is, we can extrapolate back a long way, and guess the amount of metals it probably started with. And from our model, and from measuring real stars, we get a graph that estimates the age the star formed.

      In this case, we end up with a point that is just off the bottom of our graph. This does not mean that time is working backwards, just that the star is a little cleaner than we would have predicted even if it was one of the first pop II stars. There is a long chain of calculations and many different measurements, so some error is to be expected.

      So, what do we have here? We have a star that must have been amongst the first of the pop II stars. It may be the metals emitted by the pop III stars did not always mix, and an early set of coincidences swept together a star out of the primordial gas later we would normally expect, and it is particularly clean. It is one star out of many millions we can see, and it is an odd one, so it may be a bit off the main sequence. But the most likely explanation is it is a very early star, and our measurements are a bit out.

      This is still very exciting. This star is much older than the galaxy it is in. Go figure.

      1. “how could a model that predicts the age based on the evolution of metals from the ancient and short-lived pop III stars shortly after the Big Bang predict a date Before This Process Even Started”.

        Simply by having those being two independent observations.

        That is also why the overlap, the absence of difference, is so extraordinary neat!

    1. It would be if they had better than a .8GyR uncertainty. I’d bet quite a lot that as they tighten their uncertainty, they’ll get an age closer to WMAP’s data.

    1. The accompanying link mentions that the fine guidance sensors on the Hubble Space Telescope were used in conjunction with Hipparchus data to determine accurate parallax distances.

    2. It’s almost 60 parsecs – since 1 parsec is one arc second angle and we have two AUs to play with (without doing the trig) it’s probably a manageable fraction of an arc second. If you have Celestia you can pick a nearby star and wind time up so fast you can see it wobble from the parallax shift.

  2. Unless I missed it, I am assuming this is likely a red dwarf to be that old. It would be interesting to find out if there are any planets orbiting it. That would certainly put a kink in a few theories about whether planets could for around such metal poor stars.

    Just thinking out loud.

  3. Very interesting! At 7th magnitude, this star should be visible in many amateur astronomer’s telescopes @ 15h 43m 03.1′, -10*56’01” (Libra) and would make for an interesting star party object! Unfortunately(?) at this time of year, rising in the wee late or early morning hours.. Maybe someone can take a shot and post it during the Star Party?

    1. Why would its location be “all wrong”? The Milky Way started out at ~ 400 million years as all other galaxies, see the first image. It has to have the oldest stars, same as the other galaxies.

    2. Why do you say that? There’s no reason a small enough Pop II star from shortly after the big bang couldn’t survive and still be around today in our neighborhood. Remember that the big bang occurred everywhere (or, put another way, everywhere was once where the big bang occurred), so remnants of it could be found anywhere. And 14 Gyr isn’t so long for a red dwarf, say, to so exist.

      Perhaps you are thinking about the fact that to see things as they were shortly after the big bang we have to look far way. But nobody is claiming that we’re seeing this star as it was 14 Gya; they’re claiming that they are seeing a 14 Gyr-old star.

    1. not technically, the error on the estimate of the star’s age allows for the potential age of the star to be from 13.64 to 15.24 billion years. the age of the universe is 13.77+-.06. Meaning the star was either formed at t=0 in respect to the universe up to t = .19 billion years. Meaning yes this could be a really interesting star that could provide more obtainable data to piece together events in the beginning of the universe but it could also be a dud as a lot of shit happens in .19 billion years

    2. Read the article:

      “A metal-poor star located merely 190 light-years from the Sun is 14.46+-0.80 billion years old, which implies that the star is nearly as old as the Universe!”

      ““Within the errors, the age of HD 140283 does not conflict with the age of the Universe, 13.77 ± 0.06 billion years, based on the microwave background and Hubble constant, but it must have formed soon after the big bang.” the team noted.”

      Error estimates are useful, but especially in these cases.

      1. Hi TL, thanks for taking the time to help explain uncertainties to folks. For you (& others) who appear rather interested in this topic, here is some other aspects to consider. I believe that on all fronts the uncertainties are likely optimistic, and are probably understated (perhaps you agree). For example, the 0.8 billion year uncertainty for the star’s age does not include systematic uncertainties from the model used to deduce that estimate, and the WMAP-based age does not consider a Hubble constant of 63 km/s/Mpc (e.g., as advocated by Tammann & Reindl 2012 and Allan Sandage, incidentally, the latter was a student of Hubble’s and died recently). As you may know, I would likewise note that some folks do not believe in dark matter or the lCDM model, and thus they question the interpretation of the WMAP results. Hence the importance of examining as many independent approaches to constraining the Universe’s age. The fact that we have yet to find dark matter has partly motivated folks to explore other ideas. For example, Pavel Kroupa et al. 2012 recently published a study titled “The Failures of the Standard Model of Cosmology Require a New Paradigm” (, and G. Verschuur “has serious consequences for the cosmological interpretation of the WMAP data.” (…22050402V). Personally, I believe it is important to be familiar with the diverse range of perspectives being presented, no matter what side (or undecided) of the fence one is on. I know the WMAP team cited a time to reionization of around 400 Myr+, so that may put some strain on the above result of having the star discussed (HD140283) emerge sooner after the Big Bang. But again, the uncertainties cited are optimistic and don’t tell the whole story. However, as mentioned above, the team plans to study a larger sample in order to beat down the age uncertainties estimated from metal-poor stars, so that it can provide a more reliable test of the WMAP-based age for the Universe.


  4. If there are any, I’d guess that they would probably be gas giants formed at the same time, or rogue planets that have been caught by the system’s gravity well.

    1. If there were terrestrial planets present, and they could be shown not to be rogues, what a story they could tell. The cratering alone would be treasure trove of the universes history.

  5. Unfortunately, despite its very old age, it’s still a Population II star. We’ve yet to discover any Population III stars, leaving them merely theoretical.

    1. We’re going to have to look a lot farther away than 190 ly to find a Population III star…

  6. The
    age of the Sun nearly the age of the universe.

    metal-poor star located merely 190 light-years from the Sun is 14.46+-0.80
    billion years old, which implies that the star is nearly as old as the
    Universe! “

    I calculate and written in my book “Complete Unified Theory”, (page-150-152,
    total page-424, 1998), that the age of the Sun is 14.6156 billion years. So,
    the sun is 0.1556 times older than the HD 140283 star. The birth of the sun is indicating
    nearly the birth of the universe.

    I want to request to all scientists in this fields, please verify the
    age of the sun / universe again.

    Nirmalendu Das.

    Dated: 24-02-2013.

  7. Kudos for using the latest, most precise, age for the universe, even if it moved [gasp!] the age from ~ 13.7 years to ~ 13.8 years! What is +/- 60 million years among friends? I like how cosmology has become a precision science.

    The oldest discovered planet to date should still bePSR B1620-26 b @ ~ 12.7 billion years. PSR B1620-26 metallicity is unknown, but it inhabits a low metallicity star cluster M4, with [Fe/H] ~ -1.20 [ ].

    Naively, since frequency of terrestrials shows no correlation to metallicity, another order of magnitude (oom) lower metallicity shouldn’t be a problem. (Another oom, since [Fe/H] is a logarithmic measure.) These early stars could have planets!

  8. It is only gas giants that correlates with metallicity, see Kepler & Corot data, they are rarer in metal rich stars.* Terrestrials form independently of metallicity.

    Maybe you are referring to pre-survey hypotheses?

    * Apparently gas giants don’t have as much time to form by core collapse, if the core is formed as terrestrials by aggregation and they are equally frequent everywhere. Meaning the disk scatters faster around metal rich stars.)

  9. Maybe Torbjörn Larsson can answer this one, are they saying this star existed before the Milky Way and was drawn into it, or is the Milky Way possibly older than thought?

  10. After so many years stating that the Universe is 13.73 BYO, how can they now say a single star is 14.46 BYO? Is everybody using different parameters to date objects in the Universe? I thought it was one procedure agreed on decades ago that the entire Science used…..

    1. Hi Steve, for many years there was considerable debate as to the Universe’s age, namely because various methods (& interpreters of those methods) implied vastly different results (as much as a 10 billion year offset between them!). However, that spread has reduced considerably in recent years, and I would say that these days a sizable fraction of astronomers favour the WMAP-based results, but certainly not all astronomers. The age cited in the above article has a larger uncertainty than the WMAP-based age, although I personally believe uncertainties cited for both results are likely optimistic. Essentially what you’d like to do is have as many different methods pointing to a consistent age for the Universe as possible, thereby reducing the impact of systematic uncertainties tied to any one given method. In the coming years the uncertainties will continue to be mitigated.


  11. I’m still waiting for someone to find a “white” hole which would explain where this stuff originally came from – another universe’s black hole.

Comments are closed.