Hipparcos by the Numbers

For a direct sense of what Hipparcos learned, ESA offers a sky map on its website that can easily fit in the palm of your hand.34 The “Hipparcos Star Globe” represents the brightest stars and the major constellations measured by the satellite as charts that can be printed out on two sheets of paper and then assembled into a sky sphere. In fact, for ease of construction, the sky is projected onto an icosahedron, a polyhedron with 20 triangle-shaped faces. Instructions for constructing this astronomical origami are also provided on the ESA website. This simple sky chart conveys no more than the “bony skeleton” of the night sky’s stars; Hippar – cos mapped the anatomy in exquisite detail. Its main instrument charted the positions of 118,218 stars with the highest precision.35

In addition, a beam-splitter was used with a secondary detector to map out the sky with slightly lower precision—the resulting Tycho Catalog lists 1,058,332 stars.36 Years after the satellite ceased op­eration, astronomers produced the definitive Tycho-2 Catalog, containing a prodigious 2,539,913 stars.37 That number is 99 per­cent of the stars down to 11th magnitude, which is a level 100,000 times fainter than the brightest star, Sirius. With this exquisite level of detail, Hipparcos has mapped our location in the “city of stars” called the Milky Way galaxy (plate 14).

Space missions produce data of such complexity and abundance that it’s often years before all the results are known. Hipparcos is no exception. The Tycho-2 Catalog was published in 2000, seven years after the satellite returned its last data. As recently as 2007, Dutch astronomer Floor Van Leeuwen re-analyzed the Hipparcos data. He diagnosed many small effects that had been overlooked in the original analysis, such as tiny jogs in the spacecraft’s orienta­tion due to micrometeorite impacts and subtle changes in the image geometry each time the satellite went into Earth’s shadow and then emerged into sunlight. He also took advantage of great gains in the power of computers to improve the calculation of positions. With a million stars, the number of angles between any star and all of the others is a million squared, or a trillion. To pin down the errors in positions, those trillion angles must be calculated many times, a procedure that took six months at the end of the mission but only a week when Van Leeuwen did his work using much faster proces­sors. His analysis has shrunk the errors by a factor of three from the initial goal of 0.002 arc seconds, and a factor of ten for the brightest stars.38 This minuscule angle would be formed by draw­ing lines from the top and bottom of Lincoln’s eye on a penny in New York and having them come to an apex in Paris.

The trick of Hipparcos is to measure positions across the entire sky rather than picking off stars one by one. Thus it gains from the power of large numbers. Imagine you had to cover the floor of a large room with irregular but similar-sized tiles. You could lay them out tile by tile with a good chance of keeping the separa­tion of adjacent tiles uniform, but as you covered a larger region it would become very difficult to control the uniformity. Whether you worked from the center out or the edges in or from side to side, it’s likely you’d either have tiles left over or leave a gap. The optimum solution would be to be aware of the distance between any two tiles and regulate it over the whole area, thereby filling it uniformly. If the tiles are now stars on the sky, that’s what scien­tists did with the Hipparcos data. They made an optimal solution for all stars simultaneously. In practice it was a calculation that taxed the best computers at that time.

Behind the numbers are the people who work to make a mission successful, often devoting their entire scientific careers to the task. Michael Perryman was born in the dreary industrial town of Luton, just north of London, and he was interested in math and numbers from an early age. His math teacher at school advised him to study a subject with better employment potential. He ignored the advice and studied theoretical physics at Cambridge University. For his PhD he stayed at Cambridge but switched to radio astronomy, joining a group that was still buzzing with excitement from the discovery of pulsars and the award of the Nobel Prize in Physics to Martin Ryle and Tony Hewish in 1974. At this point he has spent thirty years of his career on the unglamorous but very important work of mapping star positions, exceeding the amount of time the illustrious Tycho Brahe spent on his observations.39 Appropriately, in 2011 he was awarded the prestigious Tycho Brahe Prize by the European Astronomical Society.

Perryman was just twenty-six when he was selected to be the project scientist for the Hipparcos mission, a great honor and re­sponsibility for someone so young.40 Soon he found himself re­sponsible for the coordination of two hundred scientists and for all the headaches that go along with a complex multinational project. The biggest challenge came soon after launch when a motor on the Ariane launch rocket malfunctioned and the satellite didn’t reach its desired geostationary orbit. The unplanned orbit exposed Hip- parcos to high levels of radiation twice a day and it was thought the satellite might not last more than a few months. The team adjusted and made the best of the situation, but for over two years Perry­man lived under a “sword of Damocles” as gyros were knocked out by the radiation. In the end, the mission exceeded its design goal of both lifetime and science. Away from the project, Perryman enjoyed hiking and caving, choosing to escape underground from his upward-looking day job.