Paris, 5 January 1998

Hipparcos pinpoints an amazing gamma-ray clock

The position in the sky of the "silent" neutron star Geminga is now known to within about 10 millionths of a degree (0.04 arc-second) thanks to results from ESA's Hipparcos star-fixing satellite. These frame observations of Geminga made by the Hubble Space Telescope. Geminga emits pulses of gamma rays like a ticking clock, but its apparent rate changes because of the Earth's motion in orbit around the Sun. Using the Hipparcos position to correct this effect, astronomers have made a continuous reckoning of some 3,200,000,000 pulses in the gamma-rays emitted by Geminga, going back to observations by NASA's SAS-2 and ESA's COS-B gamma-ray satellites in the 1970s.

Following a preliminary report at a symposium on Hipparcos results in Venice in May, the full story of the pinpointing of Geminga is told in a paper to be published in Astronomy and Astrophysics in January 1998.

Patrizia Caraveo of the Istituto di Fisica Cosmica in Milan is the lead author, and other astronomers in Milan, Turin, Garching, Copenhagen and Noordwijk contributed to this aspect of the work (see footnote). The result made it possible to to use the observation of gamma-ray pulses to time the rotation of Geminga with extreme accuracy, as described in a paper by John Mattox of Boston University, Jules Halpern of Columbia University, and Patrizia Caraveo. It is due to appear in the Astrophysical Journal in February 1998, and is already accessible on the Internet.

Geminga is a unique object: a highly compressed, spinning neutron star which does not emit radio beeps like the well-known pulsars. Yet it is a powerful source of pulsating gamma-rays and X-rays. Geminga is probably the prototype of millions of radio-silent neutron stars in the Milky Way Galaxy, so far unrecognized.

"We needed Hipparcos to finish a long and complicated task of tracking down Geminga," Patrizia Caraveo comments. "Never was so faint an object pinpointed so precisely. Now we can say that we have more exact knowledge on the position of Geminga than of any other 'classical' neutron star -- even the famous Crab pulsar."

Closing in on Geminga

When first observed in a systematic way by COS-B, Geminga's place in the sky was known only to within half a degree -- an uncertainty in position as wide as the Moon. X-ray observations by other satellites narrowed the field and led to the detection of Geminga by visible light, as an extremely faint star. Last year, the same Milan-based Italian team was able to determine the distance of Geminga at 500 light-years, by a succession of observations with the Hubble Space Telescope (see ESA Information Note 04-96).

The next task was precisely to situate a Hubble Space Telescope image, including Geminga, within a much wider field of stars. The Hubble image was only one-hundredth of a degree wide, but to find enough Hipparcos stars for accurate positioning required a region more than a degree wide, as seen in a photographic plate obtained with an astrometric telescope at Turin Observatory. This plate included four stars from the Hipparcos Catalogue (118,000 stars plotted with the highest precision) and fifteen more from the Tycho Catalogue (a million stars plotted from Hipparcos data with somewhat less precision).

To link the widest and the narrowest images, four intermediate steps were needed. Two other starfields from Turin, half a degree wide and one-sixth of a degree wide, zoomed in towards the location of a starfield from the New Technology Telescope (NTT) of the European Southern Observatory at La Silla, Chile, one-26th of a degree wide. Within the NTT image lay Geminga and the location of the Hubble image.

The Hipparcos/Tycho stars in the widest field served to fix the positions of a number of other stars within the next image, and these in turn became the references for fixing other stars within the third image -- and so on, until references in the NTT image pinned down some of the background stars in the Hubble Geminga image. In this precise work, corrections had to be made for the slight motions of the reference stars across the sky in the interval between the first (1984) and the last (1996) of the images in the series. To complicate the task, Geminga itself is travelling at 120 km per second in a northeasterly direction.

The astronomers are now able to declare that on 17 March 1995 (the date of the Hubble image), Geminga was at the celestial longitude of 98.47563 degrees and celestial latitude (or declination) of + 17.77025 degrees within the Hipparcos Reference Frame. The uncertainty in the figures is only +/- 1 in the last decimal place.

Taking Geminga's pulse

Geminga rotates like a lighthouse, flashing a beam of gamma-rays and X-rays towards the Earth 252 times a minute. The first detection of pulsations in Geminga's emissions came from the German-US-UK Rosat X-ray satellite in 1992. NASA's Compton gamma-ray observatory detected the same pulses in gamma-rays and observed them for several years up to 1996. Re-examination of the gamma-ray counts for Geminga, from NASA's SAS-2 satellite (1972-73) and ESA's COS-B satellite (1975-82) led to the detection of the same pulses retrospectively. Astronomers then had the tantalizing prospect of reconstructing, from sporadic periods of observations, every tick of Geminga's clock over 24 years, 1972-1996.

Positioning Geminga accurately enough was the remaining hurdle. The gamma-ray clock seems to run fast or slow depending on the Earth's motion in orbit. At the end of March each year the Earth, and any attendant gamma-ray satellites, are travelling towards Geminga in the constellation Gemini at about 30 kilometres per second. Geminga's timekeeping speeds up by 9 seconds a day. Six months later, on the other side of the Sun, the Earth is receding at the same speed, and Geminga's pulses seem slower by the same amount. To correct completely for this seasonal effect required the more exact position of Geminga provided by Hipparcos data.

Motions in Earth orbit of the SAS-2, COS-B and Compton spacecraft at the times of the intermittent gamma-ray observations also had to be taken into account. The outcome is a coherent 24.2-year reconstruction of more than 3 billion rotations of Geminga.

"The Hipparcos result allowed us to correct for the effect of the Earth's motion," comments John Mattox of Boston. "It is also fortunate that a major glitch in the rotation of Geminga has not occurred. Some pulsars often suffer internal rearrangements or starquakes that change their rotation rates suddenly. Instead we see a gradual spindown as Geminga radiates its rotational energy."

Created by the collapse of the core of an exploding star about 300,000 years ago, Geminga has no renewable source of energy. Nevertheless it is more luminous in its gamma-rays and X-rays than the Sun is by visible light. Interaction between the spinning neutron star and a surrounding magnetosphere of ionized gas powers the emissions, by extracting energy from the rotation and slowing the star down. The new timings suggest that Geminga is like a watch that loses less than one less than one microsecond a year. But the rate of slowdown is increasing faster than expected by comparison with other young pulsars.

Does Geminga possess a planet? Another puzzle concerns a slight rhythmic change in the pulse-rate of Geminga, in a cycle of 5 years, seen most clearly in the recent Compton observations. While this could be a fluke due to errors in the rather sparse data, a physical explanation could be the presence of a planet with about twice the mass of the Earth orbiting around the neutron star every 5 years, and causing it to wobble.

ESA's framework for astronomy

Michael Perryman, ESA's project scientist for the Hipparcos mission, is a co-author of the Hipparcos-Geminga report in Astronomy and Astrophysics. He sees the multiplicity of instruments and wavelengths used in the Geminga study as an illustration of the overarching role of Hipparcos.

"The results from Hipparcos provide a framework for every branch of astronomy and bring new precision to all of them," Perryman comments."Hipparcos never saw Geminga, because it is far too faint. Yet when used to calibrate other observations in visible light, the Hipparcos and Tycho Catalogues give a position for Geminga far more accurate than could ever be expected from the X-ray and gamma-ray observations alone. Similarly Hipparcos relates the entire Universe seen by radio and infrared telescopes to the local frame of bright stars."

Other ESA astronomy missions besides Hipparcos have played key roles in the Geminga saga, starting with the gamma-ray satellite COS-B in the 1970s, and continuing with the sharp observations by the NASA-ESA Hubble Space Telescope, seeing Geminga as a faint visible star. The story will broaden into a hunt for other silent neutron stars like Geminga, using ESA's super-sensitive X-ray astronomy satellite XMM, due to be launched in 1999. Two years later XMM will be followed into orbit by Integral, ESA's successor to COS-B as a gamma-ray astronomy satellite of vastly enhanced performance.

Another Hipparcos-Geminga co-author is Giovanni Bignami. He named Geminga in 1976 and has hunted it, with his colleagues in Milan, for more than 20 years. (In the Milanese argot, "gheminga" means "it's not there" and referred to Geminga's invisibility at radio wavelengths.)

Bignami is a principal investigator in the XMM mission, and is director for science in the Agenzia Spatiale Italiana, Rome. "Geminga is a shining example of how multinational collaboration in astronomy pays off," Bignami says. "We in Europe have done very well out of ESA in space astronomy, and also in our collaborations in ground observatories. We have world-class facilities that none of our home countries could offer us on their own. While governments ponder ESA's future role in science they should sense the surge of excitement throughout Europe, as astronomers beat all expectations in using the superb opportunities that come from ESA membership."

Note 1. The full authorship of the Hipparcos-Geminga paper in Astronomy and Astrophysics is: P.A. Caraveo (IFC, Milan), M.G. Lattanzi (OAT, Turin), G. Massone (OAT, Turin), R.P. Mignani (MPE, Garching), V.V. Makarov (CUO, Copenhagen), M.A.C. Perryman (ESA, Noordwijk) and G.F. Bignami (ASI, Rome).

Note 2. A postscript version of the Geminga timing paper for the Astrophysical Journal is accessible on the www.

Back to ASTRONET's home page
Terug naar ASTRONET's home page