nature 02 November 2000
News Feature
Nature 408, 12 - 15 (2000); doi:10.1038/35040743

Does size matter?

Large telescopes are starting to dominate astronomy, putting their smaller predecessors under pressure to close. But if astronomers can adapt their ways of working, says Alexander Hellemans, both big and small could thrive.


In focus: the Very Large Telescope can be configured to act as a 16-metre mirror.

On remote mountain tops across the world, a new breed of optical and infrared telescopes stands sentry, watching the heavens. They are the colossi of modern astronomy. Twice the size of their predecessors, they are giants equipped with mirrors measuring up to ten metres across.

The first was the 10-metre Keck I telescope, sitting at the summit of Mauna Kea in Hawaii. It began science operations in May 1993 and was joined three years later by its partner, Keck II. Other behemoths are following close behind. "Until two years ago, we had only one or two large telescopes," says Shri Kulkarni, an astrophysicist at the California Institute of Technology (Caltech) in Pasadena. "A few years from now we will have almost ten."

This year, the eight-nation European Southern Observatory (ESO) finished building its Very Large Telescope on Mount Paranal in Chile. This consists of four 8-metre mirrors that can be operated independently or as a single telescope with the light-gathering capacity of a 16-metre mirror. On La Palma, one of the Canary Islands, Spain's 10.4-metre Gran Telescopio Canarias should begin routine operations by 2003. And on Mount Graham in Arizona, the Large Binocular Telescope, with two 8.4-metre mirrors, should be working by 2004.

But these giants do not come cheap. The Keck telescopes cost around US$100 million each, and as astronomers strive after ever-larger mirrors, costs will continue to rise. A team at the ESO's headquarters in Garching, Germany, has hatched a plan for the OverWhelmingly Large telescope, which will have a mirror measuring 100 metres across. The project's leader, Roberto Gilmozzi, is confident that the team can keep the price tag for this telescope within the bounds of reason. But even so, it could not be built for much less than US$1 billion.

Not surprisingly, budgetary constraints mean that some tough questions are now being asked. Even in the United States, where philanthropic institutions such as the W. M. Keck Foundation have bankrolled large telescope projects, some astronomers are questioning the wisdom of keeping the entire current roster of smaller telescopes in operation. And in Europe, where large and small telescopes compete directly for the same pots of taxpayers' money, the debate is particularly intense.

At its heart lie difficult questions about the relative scientific productivity of different telescopes. Astronomers cannot agree, for example, on whether it is better to operate one 10-metre telescope, or a series of 2-metre devices with the same light-collecting area. But it is becoming clear that, for small telescopes to survive, they need to become more technologically sophisticated and must be deployed more effectively. The era of individual astronomers competing for time on these telescopes for their own pet projects could soon be over.

A problem resolved

Little and large: is there scope for all sizes of observatory, or will the planned OverWhelmingly Large telescope (top) leave smaller set-ups such as that at Pic du Midi out in the cold?

At the most simplistic level, bigger is clearly better. Large mirrors gather more light, and so allow astronomers to observe fainter objects. They also have better resolution, giving sharper images. But there has been a gulf between theory and practice. Most ground-based optical and infrared telescopes have not achieved their maximum resolution because of the distortion effects caused by atmospheric turbulence. As a result, the Keck I telescope could not match the Hubble Space Telescope's resolution, despite the latter using a mirror just 2.4 metres across.

But adaptive optics could change all that. Developed by the US military1 and by astronomers cooperating with industrial groups in France2, adaptive optics detects the effects of atmospheric turbulence on a 'guide' star. It then uses this information to compensate for the distortion by deforming a flexible mirror onto which light collected from the telescope's primary mirror is reflected. The technique's power was demonstrated when the 3.6-metre Canada-France-Hawaii telescope on Mauna Kea produced images of the galaxy NGC 7469 that matched the resolution of those from Hubble3.

Last year, the technology was applied successfully at Keck II4, and the first adaptive optics images from Gemini North, an 8.1-metre telescope at Mauna Kea, were released last month. The technique is now central to the planning of all the main large-telescope projects.

Some astronomers believe that combining large mirrors with adaptive optics will make many small telescopes obsolete. "In my opinion there are too many small observatories that now are too expensive," says Lodewijk Woltjer of the Haute-Provence Observatory in France. "You can't expect to keep doing what you used to do and at the same time start up new things."

The French government seems to agree. Last year, it decided to stop core funding to telescopes smaller than two metres. Only strong protests by astronomers prevented the telescopes from being shut down. But in the absence of central government funding, observatories now have to seek alternative income. Stressing their value for training astronomy students, some telescopes are now being funded by individual universities or local authorities. The Pic du Midi Observatory high in the Pyrenees, meanwhile, has opened its facilities to 'scientific tourists'. "We now have only two national telescopes in France, one at Pic du Midi, the other at Haute-Provence," says Roland Bacon, director of the Lyon Observatory, who is seeking alternative finding for his observatory's 1-metre telescope.

In Britain, smaller telescopes are also coming under financial scrutiny, as the UK Particle Physics and Astronomy Research Council negotiates terms for Britain to become a member of the ESO. An announcement on whether Britain will join is expected soon. And depending on the precise terms of any membership deal, Britain might have to stop funding some smaller telescopes, such as the 4-metre Anglo-Australian Telescope at Siding Spring in New South Wales and some members of the Isaac Newton Group of Telescopes on La Palma.

Mirror man: Roger Angel's work is helping to drive the development of ever-larger telescopes.

Meanwhile, the ESO is itself threatening to axe some of the smaller telescopes on its site at La Silla in Chile unless other organizations offer to take them over. "ESO is involved in facilities that are beyond the capabilities of a single country," says Piero Benvenuti, an astronomer who works for the European Space Agency but is based at the ESO. "If a small telescope is only there for historical reasons, ESO should discontinue its use."

The same questions surround many of the telescopes run by the US National Optical Astronomy Observatories — in fact, some of the organization's smallest telescopes are already being transferred to consortia of universities. "There is a lot of handwringing about what to do with our national observatories," says Roger Angel, whose work on mirror technology at the University of Arizona in Tucson is helping to drive the development of ever-larger telescopes.

Small scopes take a stand
Some astronomers who use or work at the observatories now coming under threat feel that small telescopes are getting a raw deal, particularly as the costs of keeping them open are not huge. "A new telescope may cost US$5 million a year to run, and the small telescope will cost US$50,000," says Helmut Abt of the Kitt Peak National Observatory in Arizona.

There are also fears that politics and egos are being allowed to dominate over a reasoned assessment of the scientific merits of the different telescope sizes. "Many of the people who are in senior positions have been using large telescopes," says Richard Ellis, an astronomer at Caltech who is himself involved in a proposal to build a 30-metre device called the California Extremely Large Telescope. "In the political arena, the large telescopes get a lot of attention; they always get prominence in the newspapers. The smaller telescopes deserve to be emphasized more than they have been," he adds.

Given these conflicts over priorities, some astronomers argue that the productivity of different telescopes should be subjected to rigorous quantitative analysis. Several researchers have attempted 'scientometric' studies over the years, counting publications or analysing patterns of citation in the astronomy literature.

Sizing up the issues
The most thorough study since the introduction of the Keck telescopes was done by Chris Benn, manager of the Anglo-Dutch 4.2-metre William Herschel Telescope on La Palma, and his colleague Sebastián Sánchez5. They analysed the 452 astronomy papers published in Nature between 1989 and 1998, and a sample of 1,000 papers making up the top 125 for citations in each of the years from 1991 to 1998. They reasoned that these papers represented the cream of the scientific output of the astronomy community, and hoped that analysing the contributions made by different telescopes would indicate the relative scientific productivity of different size classes.

A similar picture emerged both for the citation data and counts of Nature papers, with the Keck telescopes making an immediate impact from the mid-1990s. For papers published in 1995–98, for instance, the citation impact of Keck I was eight times greater than that of a typical 4-metre telescope, and the mean citation impact of telescopes in the 2-metre class was four times lower still. These ratios roughly reflect the differences in mirror collecting area between the telescopes. So although larger telescopes were more productive, they did not seem to have a disproportionate advantage.

When Benn and Sánchez compared different size classes of telescope according to the total mirror surface area available within each class, the productivity of 2-metre telescopes held up well, even compared with the two Keck telescopes. "The strong showing by 1-metre and 2-metre telescopes in the 1990s augurs well for the continued scientific impact of 4-metre telescopes in the era of 8-metre telescopes," Benn and Sánchez's paper concludes. Whether this conclusion will hold up once adaptive optics is in widespread use remains to be seen. But for now, Benn argues that "it is very important to maintain a variety of telescopes of different sizes and types".

Other researchers point out that certain subdisciplines within astronomy are particularly suited to smaller telescopes. In fact, one of the most famous astronomical discoveries of recent years, the detection of the first extrasolar planet around a Sun-like star, was made by Michel Mayor and Didier Queloz of the Geneva Observatory using the 1.9-metre telescope at Haute-Provence6. Most of the subsequent discoveries in this area have been made using telescopes with mirrors less than three metres in diameter.

Detecting the 'wobble' of a star caused by an orbiting planet does not require a large telescope. But it does need up to 50 nights of observations, which are impossible to obtain on one of the giants. "Work with a small telescope is much cheaper, and people can get much more observing time," says Abt.

Any project that needs to make repeated observations of an astronomical object depends on small telescopes for this very reason. Brian Marsden of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, who leads a team that acts as a clearing-house for data about comets, asteroids and other objects in the Solar System, adds that small telescopes continue to dominate his field. "There are very few observations of anything that comes our way made with a telescope larger than four metres," he says.

In reflective mood
Nevertheless, limited budgets mean that the least productive of the small telescopes currently in operation may face the chop — or at least will get taken out of the front line of research and be used for training. To avoid that fate, the operators of small telescopes could do worse than to imitate the current star performers within the medium and smaller size classes.

In Benn and Sánchez's citation analyses, the Canada-France-Hawaii telescope was the most productive telescope in the 4-metre class. It combines two attributes that are likely to emerge as the keys to survival. These are a strong focus on developing technology such as instrumentation, and a policy of concentrating on a particular scientific problem, rather than the usual practice of allowing astronomers working on a plethora of projects to compete for slots of observing time.

"It doesn't pay to let everybody and his dog use your telescope," says Virginia Trimble of the University of California at Irvine, who noted the high productivity of the Canada-France-Hawaii telescope in an earlier scientometric study7. This telescope not only helped to pioneer adaptive optics, but its operators also took the decision to focus primarily on assigning redshifts to large numbers of objects, so determining their distances from the Earth.

Among the telescopes with mirrors smaller than two metres, Benn and Sánchez identified the 1.3-metre telescope at the Mount Stromlo Observatory near Canberra in Australia as the strongest performer. This has also been used for a specific purpose — the Massive Compact Halo Object project, which is searching the halo of our Galaxy for dark matter in the form of 'failed' stars called brown dwarfs8.

Caltech's Kulkarni is convinced that these examples point the way to the future for small telescopes. "The strategy for the coming decade is to take these old telescopes and do very focused experiments with them — a creative new style of doing astronomy," he says. But Kulkarni warns that established ways of working, which do not readily lend themselves to such focused collaborations, are entrenched among researchers who are used to letting other people run the telescopes while they just compete for observing slots. "Most astronomers are still thinking in an old-fashioned way — 'your time, and my time'," he says.

Adapt and survive
If small telescopes are to be adapted to specialize on particular scientific problems, they need to be fitted with specific instrumentation to optimize them for the task in hand — rather than the 'one size fits all' instrumentation typically fitted at the moment. And that, says Carl Akerlof of the University of Michigan, may require a shift in emphasis on the part of funding agencies, and in attitudes among astronomers. "Scientists who get involved in instrumentation are regarded as lower in the pecking order," says Akerlof, who believes that this view delayed the development of adaptive optics by "many years".

Akerlof heads a project which suggests that new small telescopes will still be around even when the current generation is too old to be worth maintaining. His Robotic Optical Transient Search Experiment (ROTSE), based at the Los Alamos National Laboratory in New Mexico, last year captured the first-ever optical images of a gamma-ray burst as it occurred9. In terms of optics, ROTSE consists of nothing more sophisticated than an array of four telephoto lenses. The key to its success was again a combination of a narrow scientific focus and the latest technology — in this case, automated operation that allowed ROTSE to home in almost instantaneously on the burst, when it was triggered by data received from NASA's orbiting Compton Gamma-Ray Observatory.

Astronomers who fear that mammoth telescopes are about to consume their field's financial resources to the exclusion of other observatories have good justification for arguing that size is not everything. But if their voices are to be heard, they will need to heed the other half of that saying: "It's what you do with it that counts."

Alexander Hellemans is a science writer in Naples.


1. Hardy, J. W. Proc. SPIE 332, 252-259 (1982).
2. Roddier, F., Nortcott, M. & Graves, L. N. Publ. Astron. Soc. Pacif. 103, 131-149 (1991). | Article | ISI |
3. Rigaut, F. Publ. Astron. Soc. Pacif. 110, 152-164 (1998). | Article | ISI |
4. Wizinowich, P. Publ. Astron. Soc. Pacif. 112, 315-319 (2000). | Article | ISI |
5. Benn, C. R. & Sánchez, S. F. Publ. Astron. Soc. Pacif. (in the press).
6. Mayor, M. & Queloz, D. Nature 378, 355-359 (1995). | Article | ISI | ChemPort |
7. Trimble, V. Scientometrics 36, 237-246 (1996). | ISI |
8. Alcock, C. et al. Astrophys. J. 486, 697-726 (1997). | Article | ISI |
9. Akerlof, C. et al. Nature 398, 400-402 (1999). | Article | ISI | ChemPort |

Macmillan MagazinesNature © Macmillan Publishers Ltd 2000 Registered No. 785998 England.