INFO 04-1997: ESA's Hipparcos satellite revises the scale of the cosmos
14 February 1997The observable Universe may be about 10 per cent larger than astronomers have supposed, according to early results from the European Space Agency's Hipparcos mission. Investigators claim that the measuring ruler used since 1912 to gauge distances in the cosmos was wrongly marked.
This ruler relies on the brightnesses of winking stars called Cepheids, but the distances of the nearest examples, which calibrate the ruler, could only be estimated. Direct measurements by Hipparcos imply that the Cepheids are more luminous and more distant than previously imagined.
The brightnesses of Cepheids seen in other galaxies are used as a guide to their distances. All of these galaxies may now be judged to lie farther away. At the same time the Hipparcos Cepheid scale drastically reduces the ages of the oldest stars, to about 11 billion years. By a tentative interpretation the Universe is perhaps 12 billion years old.
Michael Feast from the University of Cape Town, South Africa, announces his conclusion about the Cepheids at a meeting devoted to Hipparcos at the Royal Astronomical Society in London today (14 February 1997). It will provoke much comment and controversy, because the scale and age of the Universe is the touchiest issue in cosmology.
The best hope for confirming or modifying the result now rests with studies using Hipparcos data on other kinds of variable stars. An investigation of the variables called Miras, by Floor van Leeuwen of Royal Greenwich Observatory, Cambridge, and his colleagues, is described at the same London meeting. Full scientific reports on both the Cepheids and Miras have been accepted for publication in a leading journal, the Monthly Notices of the Royal Astronomical Society.
European teams of scientists and engineers conceived and launched the unique Hipparcos satellite, which operated from 1989 to 1993. Hipparcos fixed precise positions in the sky of 120 000 stars (Hipparcos Catalogue) and logged a million more with a little less accuracy (Tycho Catalogue). Since 1993 the largest computations in the history of astronomy have reconciled the observations, to achieve a hundredfold improvement in the accuracy of star positions compared with previous surveys.
Slight seasonal shifts in stellar positions as the Earth orbits the Sun, called parallaxes, give the first direct measurements of the distances of large numbers of stars. With the overall calculations completed, the harvest of scientific discoveries has begun. Among those delighted with the immediate irruption into cosmology, from this spacecraft made in Europe, is ESA's director of science, Roger Bonnet.
"When supporters of the Hipparcos project argued their case," Bonnet recalls, "they were competing with astrophysical missions with more obvious glamour. But they promised remarkable consequences for all branches of astronomy. And already we see that even the teams using the Hubble Space Telescope will benefit from a verdict from Hipparcos on the distance scale that underpins all their reckonings of the expansion of the Universe."
The pulse-rates of the stars
Cepheid stars alternately squeeze themselves and relax, like a beating heart. They wax and wane rhythmically in brightness, every few days or weeks, at a rate that depends on their luminosity. Henrietta Leavitt at the Harvard College Observatory discovered in the early years of this century that bigger and more brilliant Cepheids vary with a longer period, according to a strict rule. It allows astronomers to gauge relative distances simply by taking the pulse-rates of the Cepheids and measuring their apparent brightnesses.
Nearby Cepheids are typically 1000-2000 light-years away. They are too far for even Hipparcos to obtain very exact distance measurements, but by taking twenty-six examples and comparing them, Michael Feast and his colleague Robin Catchpole of RGO Cambridge arrive at consistent statistics. These define the relationship between the period and the luminosity, needed to judge the distances of Cepheids. The zero point is for an imaginary Cepheid pulsating once a day. This would be a star 300 times more luminous than the Sun, according to the Hipparcos data. The slowest Cepheid in the sample, l Carinae, has a period of 36 days and is equivalent to 18 000 suns.
Applied to existing data on Cepheids seen in nearby galaxies, the Hipparcos result increases their distances. It pushes the Large Magellanic Cloud away, from 163 000 light-years, the previously accepted value, to 179 000 light-years with the Hipparcos Cepheid corrections, an increase of 10 per cent. Feast and Catchpole feed this result back to our own Milky Way Galaxy, and into calculations of the age of globular clusters, which harbour some of the oldest stars of the Universe. The reckoning involves another kind of variable star, the RR Lyraes, and the Hipparcos investigators arrive at an age of 11 billion years for the oldest stars. Other estimates of the oldest stars assigned to them an age of 14.6 billion years. This seemed, absurdly, to leave them older than the Universe. A team of astronomers using the Hubble Space Telescope recently declared the Universe to be only 9-12 billion years old. The Hipparcos Cepheid result increases that Hubble-inferred cosmic lifespan to 10-13 billion years.
"I hope we've cured a nonsensical contradiction that was a headache for cosmologists," Michael Feast says, "We judge the Universe to be a little bigger and therefore a little older, by about a billion years. The oldest stars seem to be much younger than supposed, by about 4 billion years. If we can settle on an age of the Universe at, say, 12 billion years then everything will fit nicely."
Feast and Catchpole have also cleared up a mystery about the nearest and most familiar Cepheid variable. This is Polaris, the Pole Star. Imperceptibly to the human eye, its brightness varies at a relatively high rate, every 3 days. That should make it, by the Cepheid rule, a feebler star than it appears to be.
Hipparcos fixes the distance of Polaris at 430 light-years, and the researchers conclude that Polaris pulsates with an overtone, at a rate 40 per cent faster than expected for a Cepheid of its size and luminosity. Several other Cepheids gauged by Hipparcos also exhibit overtones. Were these not recognized as fast pulsators they would give false impressions in the Cepheid distance scale.
The Miraculous Stars
Another famous variable star pulsates at more than twice the frequency that theorists would expect. This is Mira, the prototype of the class of stars investigated by Floor van Leeuwen and his colleagues, using the Hipparcos data. To an unaided eye, Omicron Ceti appears and disappears in a cycle of 11 months. In the 17th Century astronomers named it Mira, the miraculous star. Astrophysicists today interpret Mira as a senile star slightly more massive than the Sun. It has swollen into a red giant and started oscillating, as a prelude to greater instabilities that will in due course fling the outer layers of the star into space.
Hipparcos fixes Mira's distance at 420 light-years. Other astronomers have gauged the apparent width of the star, as seen from the ground, so the Hipparcos team can compute the diameter of Mira as 650 million kilometres - somewhat wider than the orbit of Mars. If the Sun were in Mira's state it would swallow up the Earth and all of the inner planets.
Astronomers knew that Mira was big, but the Hipparcos result confirms that it is too large to be oscillating in a simple fashion. Again its variation is an overtone, and the same is true of some other variable stars of the same type, known collectively as the Miras.
The sixteen Miras in the survey are mostly 300-1000 light-years away, at distances more comfortably within the grasp of Hipparcos parallaxes. Before Hipparcos, there was only one fairly good measurement of a Mira distance, for the star R Leonis. Even in that case, Hipparcos adjusts the distance from 390 to 330 light-years.
Patricia Whitelock of the South African Astronomical Observatory played a prominent part in the Mira study. In preparation for the Hipparcos data, observations of selected Miras from South Africa and Russia, with infrared instruments, assessed the extent to which they are dimmed by dust. Taking this effect into account, as well as the occurrence of overtones, the team arrives at a cosmic distance scale. As with the Cepheids, they can deduce distances by comparing the brightness of a Mira with its period of variation.
Applied to the Large Magellanic Cloud, where Miras have been detected, the Hipparcos Mira scale puts the galaxy at 166 000 or 171 000 light-years, depending on the method of calculation preferred. This result is intermediate between the commonly accepted distance to the Large Magellanic Cloud and the new result from the Hipparcos Cepheid scale.
"Frankly the Cepheids are at the limit of the useful range of Hipparcos, for distance measurements," comments Floor van Leeuwen, "And as for the Miras, ours is the very first attempt to gauge the absolute distance to another galaxy via parallax measurements on this type of star. So I think we should be grateful to Hipparcos, that our earliest answers are in the right ballpark and in fairly good agreement, without being hasty in drawing cosmological conclusions."
Only the Beginning
Michael Perryman, ESA's project scientist for Hipparcos, anticipates a warm debate among astronomers. Should the Hipparcos Cepheid results be taken at face value, with all their implications for the size and age of the Universe? He remains confident that the issue will be settled by other results quarried from the Hipparcos data.
Further Hipparcos studies of variable stars, including the RR Lyraes, are in progress. Also relevant to the distance scale are differing quantities of heavy elements present in stars of different ages, which can affect their luminosities. Any remaining confusion on this point will be dispelled by mainstream Hipparcos research devoted to the basic astrophysics of stars of different ages of origin, and at different stages of their life cycles.
"Until Hipparcos, the cosmic distance scale rested on well-informed guesses," Michael Perryman says, "The distances we now have, for stars of many kinds, provide for the very first time a firm foundation from which to gauge the distances of galaxies. The work has only just begun. If it should turn out that the Cepheids have given the final answer straight away, that might be surprising. But there will be no reason for astonishment when Hipparcos's direct measurements of stellar distances lead to a revised scale for the Universe."
The Hipparcos Cepheid scale is due to be debated in London today and in Seattle on 17 February, when Michael Feast will speak at the annual meeting the American Association for the Advancement of Science. It will also be one of the hot topics at ESA's Hipparcos Symposium in Venice, 13-16 May.
The Venice meeting will celebrate the release of the Hipparcos and Tycho Catalogues to the world-wide astronomical community. It will also offer the first overview of results obtained by the groups who have had early access to the data, by virtue of their contributions to the Hipparcos mission. The subjects range from the Solar System and the Sun's neighbours among the stars, through special stars and the shape and behaviour of the Milky Way Galaxy, to the link between the starry sky of Hipparcos and the wide Universe of galaxies and quasars.