The long view: engineering the world's largest telescope
An enormous ground-based telescope could provide unparalleled images of our universe
According to some, we are in a golden era for astronomy. The voyages of NASA’s deep-space probes in the 1970s and 1980s gave us stunning images and undreamed-of information about our neighbours in the solar system. The advent of the space telescope has given us access to sharper images than ever of the planets, the stars of the Milky Way and distant galaxies. Since the 1950s, radio telescopes have told us about the stars we can’t see, emitting different frequencies of radiation. With all these advances, our knowledge of the processes that power stars, and that formed the universe itself, continue to grow.
But even as plans advance for new space telescopes - most notably the James Webb Telescope, successor to Hubble, slated for launch in 2014 or 2015 - a new era in ground-based observation is set to begin. Plans are advancing for optical telescopes that will provide sharper images even than Hubble, to be sited on high mountaintops away from artificial light. These telescopes will be huge, and the largest of them all, depending heavily on technologies being developed by UK scientists, will be the European Extremely Large Telescope (E-ELT), to be built in Chile by a consortium of 15 countries. The design for E-ELT, which is expected to cost around €1bn (£836m), was finalised late last year by the European Southern Observatory (ESO) and the full authorisation to begin construction is expected in the next few months.
When it comes to telescope design, there is a simple maxim: size matters. The more light the telescope is able to focus, the sharper the images will be. The large telescopes that dominate astronomy are all reflecting instruments that focus light using mirrors, and the mirrors are getting larger. Hubble’s primary mirror is 2.4m in diameter; the largest single-piece mirror currently in space is on the European Space Agency’s Herschel observatory and is 3.5m wide. James Webb’s mirror will be made from segments and will total 6.5m in diameter. On Earth, the largest telescope currently operating is the Gran Telescopio Canarias in the Canary Islands, whose primary mirror is 10.4m in diameter. The jump in size to E-ELT is considerable. Its main mirror will be 42m in diameter.
Considering the incredible images seen from Hubble, it might be asked why there’s any need for ground-based telescopes at all. The main reason is one of access. The difficulties of maintaining orbiting telescopes are extreme; it was only possible with Hubble thanks to the Space Shuttle, and after this year there will be no more Shuttle flights. Herschel and James Webb will be inaccessible to humans, orbiting at the Lagrange Point between the Earth and the sun, 1.5 million kilometres away. They can’t be upgraded or repaired. No matter how hard it is to build a 42m telescope on top of a mountain in Chile, it will still be possible to maintain it and to evolve its design.
The size of the telescope is driven by what it is designed to do. With the discovery of exoplanets - worlds orbiting stars other than our own - astronomers want to be able to study these planets in detail, including looking at the atmospheres of the planets to determine their composition. ’To do this, they need to collect more photons from the object they are observing by using a higher-aperture telescope, and to gain higher spatial resolution, again by using a bigger telescope,’ said Colin Cunningham of Roehampton University, who is leading the UK’s E-ELT programme. This could be done by combining the images from several telescopes. However, Cunningham explained, astronomers also want to look at very large objects, such as galaxies, and this is best done with a single large telescope, as these are more sensitive than groups of smaller instruments.
Earth-based telescopes all suffer from one problem that has put a limit on their size in the past: they have to look through the atmosphere. This is one reason that the biggest instruments are on mountains - less atmosphere to look through - but even so, the view is through a curtain of moving gas, which causes the images gathered by the huge mirrors to ripple and distort.
“Telescopes have to be built in remote locations so there has to be a high degree of modularity”
John Lyle, Arup
The technology to overcome this problem is the key to building the next generation of large telescopes - distort the mirror to counter the distortions of the atmosphere. Known as adaptive optics, this technique is currently being tested at various large telescopes around the world ahead of its adoption on E-ELT.
E-ELT will have five mirrors, several of which will be controlled using controlled optics. The primary mirror will consist of 984 hexagonal segments up to 2m across, each weighing about 150kg, mounted on three actuators that will adjust their position to take account of gravitational and low-speed wind forces affecting the telescope support structure - as the structure flexes, the mirrors stay still. This is known as active optics and has been used on large telescopes since the 1980s - ESO operates four telescopes with mirrors 8.2m in diameter, which use active optics to fine-tune the mirror’s shape once per minute. E-ELT’s mirror will be controlled in a similar way, with the actuators working once per second.
But the truly challenging mirror is the fourth one in the sequence of five. Mirror M4, the atmospheric correction mirror, is 2.6m across, 2mm thick, and uses 8,000 actuators moving up to a thousand times per second. While the primary mirror control is designed to keep the mirror as much as possible in the same shape - to an accuracy of about 10nm across its diameter - by moving various separate elements, the M4 system has to keep the surface of a single-component, flexible mirror in constant and complex motion.
“When viewing huge objects, a single large telescope is more sensitive than groups of smaller instruments”
Adaptive optics works by monitoring a reference star close to the location that the telescope is observing, and translating its flickering into a series of computer commands to move the adaptive mirror so that it cancels out those flickers. In fact, the reference star does not have to be real. In the late 1980s, ESO pioneered the use of ’laser guide stars’, a very bright point of laser light projected into the upper atmosphere wherever the astronomers wish to observe. The laser’s energy is chosen so that it excites sodium atoms in the upper atmosphere, providing a visible spot of light whose wavelength is known precisely. The interaction between the steady, controlled light and the moving atmosphere mimics how a star would behave if it were observed through the atmosphere at that point, providing reference where there are no suitable stars.
It’s the scale of the challenge on the E-ELT that makes the difference. M4 is more than twice the size of any existing adaptive-optics mirrors.
ccording to Jason Spyromillo of the ESO, who is working on the control system for M4, 15 times as many computations are needed for the adaptive control algorithms as for the active control on M1, despite the difference in size between the two mirrors. Even the enclosure for the E-ELT will present a challenge. The dome that covers the telescope will be 90m across, making the whole structure about twice the width and height of the Royal Albert Hall.
’The biggest difficulty in building a structure like that is logistic,’ said John Lyle of Arup, who was part of a team that submitted an enclosure design to ESO - many elements of which have been incorporated into the final design, which is undergoing a front-end engineering design process at ESO. ’Telescopes have to be built in remote locations, and access to mountaintop sites tends to be tortuous, so there has to be a high degree of modularity.’
eyes on the sky
Larger telescopes will sharpen up our image of space
The ambitions of astronomers for E-ELT are commensurate with the size of the telescope. As we discover more about the universe, we also discover that there are huge gaps in our knowledge: we have found new phenomena, but we have no idea what is causing them.
One of the reasons for looking harder at the universe is that we have recently discovered that the universe is not only expanding but, contrary to expectations, that expansion is accelerating. Cosmologists have postulated that some sort of ’dark energy’ is responsible for this acceleration but there is no evidence for this. There may be some clue in the rapid expansion of the universe following the Big Bang - and to look for that, you need to look far back in space and time to dim, distant objects whose light began travelling towards Earth (or, rather, the space where Earth now exists) before that inflation.
Meanwhile, the current generation of large telescopes is discovering the existence of planets around other stars - but we know very little about them. Some of these planets are gas giants so large that they can generate their own light. E-ELT’s ability to form images 100 times sharper than Hubble will allow astronomers to study the emissions from these planets and analyse their electromagnetic spectra, which will tell us what they are made of; this may allow us to study smaller, rocky exoplanets as well. It may even tell us whether these planets have atmospheres capable of supporting life.
There are also gaps in our knowledge about when stars first formed, and how nebulae condense into stars and solar systems. Current telescopes are unable to resolve galaxies relatively close to the Milky Way into individual stars; E-ELT will be able to look at stars 10s of millions of light-years away, and will also allow us to look at the faintest, lightest stars in our own galaxy, which will tell us how long they have been shining.