HD 140283: The Star Older Than the Universe
HD 140283 has been known to astronomers since before the start of World War I. William S. Adams measured its spectrum in 1912. He noted its exceptionally high proper motion, its angular movement across the sky due to its velocity through space. At approximately 350 kilometers per second relative to the local standard of rest, HD 140283 is racing through the solar neighborhood on a highly elongated orbit. This orbit will eventually carry it back out into the galactic halo, the ancient, roughly spherical distribution of old stars that surrounds the Milky Way's flattened disk.
Spectroscopic analysis in the 1950s by Joseph Chamberlain and Lawrence Aller revealed that HD 140283 contains approximately 250 times less iron than the Sun. This marked it as a Population II halo star formed from gas that had undergone minimal chemical enrichment from previous stellar generations. Yet despite its low metallicity, the star contains detectable lithium, a fragile element that is destroyed in stellar interiors at relatively low temperatures. The presence of lithium indicates that HD 140283 has not yet evolved into a red giant. During the red giant phase, the first dredge-up, a mixing event that brings processed material from the stellar interior to the surface, would have destroyed its surface lithium. The star is a subgiant, caught in the brief evolutionary phase between hydrogen core burning on the main sequence and the red giant phase.
In 2013, Howard Bond and colleagues published Hubble Space Telescope parallax measurements for HD 140283 with unprecedented precision. Combined with the star's measured temperature, spectroscopic composition, and the detailed physics of stellar evolution, these observations yielded an age estimate of 14.46 billion years with a formal uncertainty of ±0.8 billion years. The problem was immediately apparent. The best cosmological measurements at that time, based on observations of the cosmic microwave background by the Wilkinson Microwave Anisotropy Probe (WMAP), indicated a universe age of 13.77 billion years. The Planck satellite later refined this to 13.787 billion years. At face value, HD 140283 appeared to be older than the universe itself, a logical impossibility.
The scientific community did not immediately dismiss this result as an error. Bond and colleagues had been meticulous in their analysis. The uncertainty bars, while substantial, represented only the random measurement errors. More problematic were potential systematic errors in both the stellar age determination and the cosmological age measurement. Either HD 140283 was genuinely old enough to create tension with Big Bang cosmology, pointing to fundamental problems in our understanding of the universe's expansion history or the formation of the first stars, or unaccounted systematic effects were biasing one or both age determinations.
The resolution required years of follow-up work attacking the problem from multiple angles. Interferometric observations from the CHARA Array at Mount Wilson Observatory measured HD 140283's angular diameter directly. This yielded a radius determination independent of stellar models. When combined with the star's luminosity, calculated from its parallax and apparent brightness, and effective temperature, this radius provided a tighter constraint on the star's evolutionary state. Meanwhile, high-resolution spectroscopy determined tailored elemental abundances specifically for HD 140283. This moved beyond generic low-metallicity composition assumptions.
A critical refinement concerned oxygen abundance. HD 140283 exhibits substantial alpha-element enhancement. This means that elements like oxygen, magnesium, and silicon, which are produced by massive stars and ejected in supernova explosions, are over-abundant relative to iron-peak elements produced by Type Ia supernovae. The oxygen-to-iron ratio is about twice that of the Sun. This reflects the star's formation from gas enriched primarily by core-collapse supernovae before Type Ia supernovae had contributed significantly to the chemical inventory. Opacity tables, the data describing how stellar material absorbs radiation at different wavelengths and temperatures, must account for this unusual composition. Oxygen absorption features significantly affect energy transport in the stellar envelope.
In 2021, Jianling Tang and Meridith Joyce published revised age estimates using the Modules for Experiments in Stellar Astrophysics (MESA) stellar evolution code, interferometric radius measurements, and updated Gaia parallaxes from Gaia Data Release 3. These measurements placed the star at 201.15 light-years with a fractional uncertainty of only 0.16 percent. Their best estimate was 12.01 billion years with a statistical uncertainty of ±0.05 billion years. When modeling uncertainties, primarily the treatment of convection and element diffusion, were included, the age was 12 billion years ±0.5 billion years. A 2024 study by Cédric Guillaume and colleagues, incorporating custom opacity tables for HD 140283's specific abundance pattern, found ages between 12 and 14 billion years. These ages depended on assumptions about the initial helium abundance and mixing-length parameter.
Most recently, in 2025, an asteroseismic analysis detected stellar oscillations in HD 140283 using data from NASA's Transiting Exoplanet Survey Satellite (TESS). Asteroseismology, the study of internal stellar structure through observations of pulsational frequencies, provides an independent mass determination. The measured oscillation frequencies, combined with the star's other observed properties, yielded a mass of approximately 0.82 solar masses and an age of 14.2 billion years with an uncertainty of ±0.4 billion years. While this central value is higher than some other recent estimates, it remains consistent with the age of the universe within the combined uncertainties.
The evolution of HD 140283's age determination from 14.5 billion years to approximately 12-14 billion years illustrates several crucial points about precision astrophysics. First, seemingly small improvements in observational constraints can propagate into gigayear-scale changes in inferred age for very old stars. A 10 percent reduction in parallax uncertainty or a 50-Kelvin refinement in effective temperature can make this difference. Evolutionary timescales are long for these stars, and models are sensitive to input parameters. Second, composition matters profoundly. Custom opacity tables accounting for HD 140283's specific oxygen enhancement yield systematically different ages than generic low-metallicity opacities. Third, multiple independent constraints are essential. Parallax, interferometric radius, spectroscopy, and asteroseismology reveal and quantify systematic uncertainties that would otherwise remain hidden.
The case of HD 140283 transformed age-dating methodology for the oldest stars. Rather than reporting single-valued ages with optimistically small error bars, the modern approach emphasizes comprehensive uncertainty quantification. This includes both random measurement errors and systematic modeling uncertainties. The goal is not merely to determine whether a particular old star is 12 or 14 billion years old. Rather, we must establish whether the ensemble of oldest stellar ages is consistent with cosmological constraints. Any remaining tension can then probe subtle effects in stellar physics or cosmology.