A new measurement of the distance to the Pleiades star cluster is casting doubt on the reliability of one of astronomy鈥檚 most widely used databases, reigniting a seven-year controversy over how cosmic distances are determined.
The issue is a key one, says Valeri Makarov, at the California Institute of Technology: 鈥淲hatever we know about the distances to astronomical objects is based on a few distances to the nearest open clusters.鈥� Open clusters, like the Pleiades, are loose groupings of stars.
Measuring astronomical distances is not simple. For example, distant bright objects can look the same as nearby dim ones. So astronomers have built a step-by-step system that starts by using several independent methods to accurately determine the distance to nearby objects 鈥� the open clusters. They then use these measurements to define a more distant cosmic yardstick, and so on.
Advertisement
Traditional measurements put the Pleiades at 430 light-years away, give or take a dozen light-years. These were derived by comparing plots of the colour and brightness of some its stars with similar plots from the much closer Hyades cluster, whose distance is known from simple geometric measurements.
But things got more complicated in 1997. That spring, an enormous database of the distances to more than 100,000 stars was published based on data from the European Space Agency鈥檚 Hipparcos satellite. It calculated the distances using parallax, i.e. observing a target at widely separated points along our orbit around the Sun and measuring the angle between them to a precision of one milliarcsecond.
Astronomers immediately noticed the distances of about a dozen open clusters differed from traditional figures by as much as 20 per cent. Hipparcos鈥檚 distance to the Pleiades was about 385 light years, 10 per cent lower.
Completely different
鈥淭he issue was very controversial from the start, with Hipparcos claiming there was no way systematic errors could be that large,鈥� says Bohdan Paczynski of Princeton University, New Jersey. 鈥淥thers claimed that if Hipparcos were correct, this would require a major revision in our understanding of stars and clusters. The best way to resolve the debate is to measure the distance very accurately but with a completely different method, and that鈥檚 what this latest work did.鈥�
The lengthy study was carried out by Xiaopei Pan, an astronomer at NASA鈥檚 Jet Propulsion Laboratory in California, and two colleagues. They used linked telescopes, called interferometers, to monitor a bright double star in the Pleiades cluster for a decade.
They determined how frequently the stars orbit each other and how far apart the stars appear to be in the sky. The stars鈥� intrinsic brightness gave an estimate of their masses, and they plugged all this information into a simple equation to arrive at a cluster distance of about 440 light-years.
鈥淭his paper doesn鈥檛 completely resolve the debate but makes a strong case against Hipparcos,鈥� says Paczynski. But he says another team may have evidence before the end of 2004 to verify Pan鈥檚 result.
The study confirms that traditional approaches to estimating distances are good, Pan told New 杏吧原创: 鈥淎nd Hipparcos is good too, but only to one milliarcsecond.鈥�
Unknown error
Paczynski agrees: 鈥淚n this case, they tried to stretch the instrument beyond its means.鈥� That is because all of the controversial clusters were too far away for the satellite to measure accurately using a single star. So the distances to a few dozen stars in each cluster were averaged. Normally that would work, Paczynski says, but some unknown error intervened.
鈥淚 don鈥檛 think it鈥檚 an instrument error,鈥� says Makarov, who was involved in compiling the Hipparcos database. He believes it lies in the equations used reduce the data and has published a possible solution.
That may be important, as Hipparcos is likely to remain the best distance data available until next-generation astrometry missions such as ESA鈥檚 Gaia and NASA鈥檚 Space Interferometry Mission go up in the next decade. These will be 1000 times more accurate than Hipparcos.
Journal reference: Nature (vol 427, p 326)