How far are the stars we see every night? This was possibly one of the first questions early astronomers asked themselves and it is still among the most fundamental problems for present day astrophysics. In some sense, astronomers do build the Universe as the observing methods allow them to place neighboring stars, distant galaxies or trace the cosmological expansion. This cosmic distance ladder not only provides a three dimensional view of the cosmos but it is also essential to understand the physics of the observed phenomena. While we have been able to travel billions of light-years into the young Universe the first steps in our ladder may require a little work to be in perfect equilibrium, as discussed in a recent paper by Carl Mellis and collaborators 1.
The simplest method that can be used for distance measures is that of parallax. The parallax is the apparent motion of near objects due to the displacement of the observer on its orbit (typically, but not always, the Earth revolving around the Sun). Since more distant objects also show smaller parallaxes (as anybody staring through the window of a fast moving vehicle can check), you have to always choose the distant objects as your reference. This is essentially what the ESA Hipparcos Space Astrometry Mission did for more than 100,000 stars during the early 90s. And this is what the ESA Gaia space observatory is being doing for less than a year.
However, Hipparcos threw some intriguing results. For example, the Pleiades cluster seemed to be too close to the Earth, closer than anybody had previously determined. Most works, either using parallaxes or not, had agreed in slightly more than 135 pc, while Hipparcos retrieved an anomalously low value of some 120 pc. Even after the stirling job by van Leeuwen 2 in correcting errors and identifying systematic offsets, the Pleiades were simply too close to us. Too close means that stars should also be dimmer than we thought and here is where the problem starts. The complex distance ladder we had been building linked the invisible facts inside the heart of stars to the observable variations of size, luminosity and temperature via the stellar structure and evolution. And if the stars were closer than expected, then something very important was missing.
Brave scientists should never fear inexplicable facts since they reveal either an incomplete theory or inaccurate empirical data. The so-called ‘Pleiades controversy’ was investigated by many people, as for example David R. Soderblow and his collaborators 3 who in 2005 used the Fine Guidance Sensor onboard the Hubble Space Telescope to determine the parallax and proper motions for three members of the cluster. Again, the far Pleiades scenario was favored. And this was not the only case. Astronomers modeling the orbital motion of binary stars inside the cluster (Atlas, for example) or isochrone fitting of the cluster as a whole were finding the same. Everyone was retrieving values consistently higher than those of the most precise instrument we had ever had.
In this context, early this summer, a new work put what is possibly the last nail in the coffin of the controversy. Mellis et al. used the Very Long BaseLine Interferometry (VLBI) array of radiotelescopes during one year and a half to map the parallax motion of some radio sources inside the Pleiades cluster with respect to a supermassive black hole in the distant universe. What they retrieved was, surprise, surprise, again in disagreement with Hipparcos. Moreover, a weighted average taken into account the uncertainty of the measurements nailed the VLBI value.
One may argue that a 10% difference in the distance is not that important. However, as Mellis points out, this would require an increase of 20 to 40% of Helium in the Pleiades stars. Alternatively, we might have a bad description of the physics happening inside young stars. This would affect our understanding of stellar formation and evolution and, so far, no other observation had questioned those points to such extent.
This story does not end here 4. While most people would be convinced at this point that Hipparcos results can be taken as a bad measurement from a superb instrument, one cannot avoid feeling that random errors in the distance of stars may be contaminating our picture of the near universe. In particular when the GAIA mission, an update and extension of Hipparcos, is expected to provide the most precise information for the position of about 1% of the stars in the Galaxy. We need to know what was wrong with those data in order to be sure that this is not going to happen again in the future. Until someone is able to provide us with a convincing explanation there will be room for the doubt. This is how science works and the only way we have to support our theories on solid ground.
- C. Mellis et al. (2014). A VLBI resolution of the Pleiades distance controversy.Science 345, 1029. doi: 10.1126/science.1256101 ↩
- F. van Leeuwen (2009). Parallaxes and proper motions for 20 open clusters as based on new Hipparcos catalogue. Astronomy & Astrophysics 497, 209 – 242. doi: 10.1051/0004-6361/200811382 ↩
- D.R. Soderblom et al. (2005). Confirmation of errors in Hipparcos parallaxes from Hubble Space Telescope fine guidance sensor astrometry of the Pleiades. The Astronomical Journal 129, 1616 – 1624. doi: 10.1086/427860 ↩
- L. Girardi (2014). One good cosmic measure. Science 345, 1001. doi: 10.1126/science.1258425 ↩