Hubble The Telescope That Revolutionised Our Understanding Of Space

There are 100 billion stars in our galaxy, the Milky Way. Or maybe there are 200 billion. Or maybe 400 billion. No one really knows. In the universe, there are 100 billion galaxies. And as technology improves, that number could increase to 200 billion galaxies. No one is certain. At the start of the 20th century, there was only one galaxy. Our own. Clearly, over the past 100 years, our knowledge of the universe has expanded. But there's still a long way to go. After more than 40 years, Voyager 1 has left the influence of the solar wind and is heading for interstellar space. Most of its instruments have been switched off, and its nuclear battery is running low. It will be many thousands of years before it reaches anything other than interstellar debris. By then, the craft will be long dead. Cosmic distances are so vast that sending a probe to investigate anything beyond our solar system is not practical. The only realistic way we have of exploring the universe beyond our neighborhood is by examining the light that reaches us. The telescope is the main tool we have for this investigation. And for centuries, astronomers have been building ever larger instruments to see deeper into the cosmos. In the 20th century, conventional telescope design reached its limit. Past 8 metres in diameter, the curved primary mirror becomes so heavy that it will sag. As it is aimed at different targets, its precise curve is distorted by the effects of gravity. In addition, any extra resolution gained by the larger mirror is not delivered because of atmospheric distortion. One way to solve both these problems is to put a telescope in space. The Hubble Space Telescope is the best known example. While the absence of atmospheric distortion was a breakthrough, producing very sharp images, both operational and maintenance costs are considerable, and there are launch constraints on the size of a telescope that can be delivered to orbit. But engineers have developed a new generation of Earth-based telescopes with segmented mirrors that overcome the limits to size and can nullify much of the distortion caused by the atmosphere. The twin Keck telescopes on Mauna Kea in Hawaii were early adopters of this technique. Both have 10-metre mirrors, each in 36 hexagonal segments. Sensors monitor the mirror's shape, and actuators behind each segment continually correct for any distortions to the curve. The system is called active optics. A similar system that works at a more rapid rate monitors the stability of a bright guide star, or an artificial guide star generated by a laser. The adaptive optics system reads wavefront changes in the atmosphere, then adjusts the mirror's shape to minimise any distortion. This image of Jupiter using adaptive optics rivals those captured by an orbiting telescope. Observatories are located at high altitude, where the air is dry and there is no light pollution from nearby cities. The European Southern Observatory has clusters of telescopes at several sites, high in Chile's Atacama Desert. At Paranal, four identical large telescopes each have 8.2-metre mirrors. They can combine their light using interferometry. In addition, four auxiliary 1.8-metre telescopes can link into the cluster. Collectively, they are the Very Large Telescope Interferometer, or VLTI. Visible light is just a small part of the electromagnetic spectrum. ESO has another facility north-east of Paranal, on the Chajnantor Plateau. Here, an array of 66 radio dishes observe in millimetre and submillimetre wavelengths. They are linked and also function as an interferometer. Observing in the microwave part of the spectrum is particularly useful, as these frequencies penetrate gas and dust clouds that remain opaque to higher energy signals. The more widely the telescopes are separated, the better the angular resolution they can deliver. This property led radio astronomers around the globe on a unique collaboration. If radio telescopes around the world could link up, perhaps they could see the unseeable. General relativity predicted bodies so massive, with such strong gravitation, that not even light could escape from them. They were called black holes, and at first they were regarded as a mathematical anomaly. But evidence mounted suggesting that black holes were more than just a theoretical curiosity. At the centre of some galaxies, light far more intense than could be expected from stars has been detected. Massive bodies such as young stars attract diffuse matter, which begins swirling and flattens into what's called an accretion disk. Just as planets orbit stars, it's now accepted that galaxies orbit supermassive black holes. The very bright light comes from the accretion disk surrounding the black hole. Matter in the accretion disk is accelerated to incredible speeds, generating intense heat. Along the axis of rotation, jets of ionised material are ejected at close to the speed of light. Computer modelling indicated that the vast energy generated by a black hole's interaction with material in the accretion disk could be detected. The silhouette is the event horizon, the boundary beyond which light cannot escape. If you want to make a test of the fundamental theories of the universe, you want to go to the most extreme laboratories in the universe. And a black hole is that. Astrophysicists knew that by linking widely separated radio telescopes, the resolving power of the combined instruments would increase dramatically. The first attempt at imaging the black hole at the centre of our own galaxy was made in 2006 using instruments in Arizona and Hawaii. It was not successful, but researchers learned a lot, realising more telescopes were required. By 2009, an international collaboration between eight different telescopes had coalesced. They aimed to capture an image of the supermassive black hole at the centre of Messier 87. So for this experiment it was clear that we need expertise from completely different areas. And so we needed someone with imaging capabilities, we needed someone with numerical relativity capabilities and in my case we needed someone who can try to find this additional information via pulsars. And so just being part of this team means you work completely differently. The project was named the Event Horizon Telescope. Instruments in North and South America, Europe, Hawaii and even in Antarctica would unite to form a telescope with a virtual diameter equal to the Earth. None of these facilities was designed to work as part of such a large array and the technical difficulties facing the team were immense. Signals from all the sites would have to be processed at a central location, yet the distances that separated them were so great that a direct line linking them was impractical. Data transfer would take place via computer hard drives that were moved from the place of acquisition to two locations, the Max Planck Institute in Bonn and the MIT Haystack Observatory in Massachusetts. The Event Horizon Telescope collaboration is a collaboration of more than 200 scientists with different backgrounds. They all come from different parts of the world. They have different experiences, different set of skills. These are engineers, observers, theoreticians, and they all work together not only to image the event horizon of a black hole, but also to understand what we're seeing. Theoreticians already understood that while a black hole could only ever appear as a silhouette, its influence upon light emanating from surrounding matter would be dramatic. By using mathematical simulation, NASA came up with a moving image of what they thought the EHT project would reveal. Viewed from above, the swirling accretion disk appears without distortion. The smooth inner circle is the photon ring, light in orbit just outside the event horizon. When viewed from the side, the black hole's gravitation bends light from the accretion disk so it can be seen from several angles at the same time. But no one really knew what they would see. For decades, evidence had accumulated that a black hole at the center of galaxy M87 was consuming huge amounts of material. Unusually high emissions across the electromagnetic spectrum were detected in this region, and images from the Hubble Space Telescope revealed a huge jet of matter extending out like years. In April 2017, the EHT was scheduled to make a series of coordinated observations of the heart of the M87 galaxy. Each observatory had been equipped with its own atomic clock, as all data had to be meticulously time-stamped. Good weather had to happen simultaneously at each site, and a go-no-go decision was made several hours before each observation. April 5th, 6th and 7th had good conditions, with data being collected. Observations were suspended on the 8th and 9th due to strong winds at the sites in Mexico and Arizona, with the final night's observations made on the 10th and 11th of April. We have this terrible period of waiting where we don't know if it's all worked. We send all of our data together, and only when it's truly combined do we know if it's worked. And then, the even harder part begins, of analyzing that data and being very, very careful, doing all the checks and balances to know that we got it right. During the observation period, 5 petabytes, that's 5,000 terabytes, of data was recorded. The hard drives upon which it was stored weighed half a ton. They were transported to correlators in Westford, Massachusetts, and to Bonn in Germany. While this was relatively simple for most sites, limited access to the South Pole telescope delayed its data for several months. Interpreting radio data as a picture relies upon the black hole's gravitation bending, passing electromagnetic waves. The EHT will only be able to see the individual points from rays that are bent toward the Earth. It is the correlation of all these points that builds them into a recognizable image. Seeing a black hole actually allows us to not only know they exist, and not only know an event horizon exists, it also allows us to test some of the very basic predictions of the theory of general relativity of Albert Einstein, which really describes space and time in its completeness. And that has never been tested before. In April 2019, the Event Horizon Telescope team released their image of the active center of the M87 galaxy. At its heart was the silhouette of one of the largest black holes we know. It's roughly 38 billion kilometers across, the diameter of our solar system. Its mass is 6.6 billion times greater than our Sun. After predictions flowing from the general theory of relativity more than 100 years ago, the world finally has proof that black holes do exist. But this is not the end for the EHT team. Expanding the number of telescopes in the network will enhance resolution, and there are plans to investigate the black hole at the center of our own galaxy. The future of this project is amazing. We have done something extraordinary. We've made the first picture of a black hole, but now we want to do even more. Now we want to make the first movie. Now we want to understand how space-time rotates around the black hole. We'll do that by putting more telescopes around the world to make our virtual lens even better. Investigation into black holes is not restricted to radio astronomy. The New Technology Telescope at La Silla in Chile, the first instrument equipped with active optics, began long-term observation of the stars at the center of our own galaxy in 1992. The NTT observed in the infrared part of the spectrum, because only that light was able to penetrate the heavy clouds of gas and dust shrouding the central regions of the Milky Way. A team from the Max Planck Institute plotted the positions of the stars, building up an image of the way they moved. They began using the larger telescopes of the European Southern Observatory at Paranal when they came online. Over 16 years, they mapped the paths of 28 stars. Well, you see, the Milky Way Center is one of the most important laboratories we have to study in very great detail what's happening in centers of galaxies, in much more detail than we can ever hope to do in all other galaxies. Yet, here we are, we can study whether there's a central black hole, what happens around it, and so forth. All very general issues which you would like to explore, and which you cannot really study that much in detail in other galactic nuclei. Results show that the stars are orbiting an invisible point, which must be our own galaxy's black hole. Analysis of their highly eccentric paths, suggests the Milky Way's black hole is slightly greater than 4 million times the mass of our Sun. Of particular interest is the star S2, which is in a 16-year orbit of the black hole. Its elliptical path was due to take it to its closest pass, and its highest speed, in April 2018. We want to use it as a tool to test whether general relativity, the theory of Einstein, is actually wrong or right. General relativity predicts that the black hole's influence would distort S2's light toward the infrared as it reached 3% of the speed of light, which happened, yet another vindication for Einstein. This test is important to researchers, because general relativity predicts that time stops inside a black hole's event horizon, and they want to know if the laws of physics break down approaching these extremes. Since Copernicus proposed that the planets orbited the Sun, astronomers have speculated about planets orbiting other stars. Advances in telescope technology allowed researchers to see dips in the brightness of stars that could only be caused by a transiting planet. These fluctuations in intensity gave astronomers clues about the planet's size and its orbit. A different method monitored the spectral shifts in a star's light caused by a consistent change in its position, again only explicable by an orbiting planet. Using these systems, telescopes around the world began cataloguing exoplanets. But the methods used were biased toward finding large planets, similar in size to Jupiter, that had an easily detectable influence on their star. And of all the planetary systems out there, researchers could only use these methods for planetary systems that were edge-on to the Earth. Even so, the number of exoplanets we knew about continued to grow. There were surprises. In 2012, a French team discovered an orphan planet wandering across our galaxy. It had no star. It was detected by its emissions in the infrared. It's difficult to predict the orphan planet's size, but it's certainly still at the very least several times the mass of Jupiter. Naturally, researchers were keen to see Earth-sized planets, planets that could possibly nurture life. When a new and highly accurate spectrometer called HARPS was installed at ESO's La Silla 3.5-metre telescope, this became possible. Among its discoveries are two planets orbiting Gliese 667 c, about 20 light-years from Earth. The star is a red dwarf, about one-third the mass of our Sun. Of particular interest is Gliese 667 cc, a planet about five times the mass of the Earth, and at a distance from its star that puts it in the habitable zone. Statements about conditions on Gliese 667 cc have such a large margin of error that much more work needs to be done before it could be declared a possible venue for life as we know it. Another star, 40 light-years away, has a greater range of possibilities. TRAPPIST-1 is an ultra-cool red dwarf star orbited by seven planets, all of them close to Earth in size. To date, this is the largest number of exoplanets detected orbiting an individual star. At least three of the planets are considered in the habitable zone. Many different observatories have been gathering data about the TRAPPIST-1 system, but we still know so little about what kind of environments the different planets might have. However, the coming generation of telescopes should be able to analyse the atmospheres of the TRAPPIST's planets, so they will remain a region of close study into the future. At the moment, around 3,800 exoplanets have been catalogued, and that number continues to grow. In 2018, the satellite TESS was launched to search for more exoplanets. To have a clear view of both the northern and southern skies, TESS was placed in a 14-day orbit around the Earth, at a 40-degree angle to the plane of the Moon's orbit. It will systematically map 26 segments of the sky, recording fluctuations in the brightness of nearby stars that reveal exoplanets. During its close passes of the Earth, the Planet Hunter downlinks its data to sites across the United States. Ground-based observatories will then follow up leads in a more detailed way. This type of collaboration between orbiting and terrestrial telescopes is becoming common. An interesting discovery came from the young star Beta Pictoris. Fluctuations in its light did not bear the rhythmic signature of an orbiting planet. Further investigation by the European Southern Observatory revealed three comets orbiting the young star. Beta Pictoris is only around 20 million years old, and still displays a protoplanetary ring of dust. In TESS's first year of operations, at least 29 new exoplanets were confirmed, with a further 1,000 objects of interest listed in publicly accessible databases. During its two-year life, TESS will image 200,000 celestial bodies, including 1,000 of the closest red dwarf stars. And there will soon be a new Planet Hunter in orbit. The Wide Field Infrared Survey Telescope, WFIRST, will use a different technique known as gravitational microlensing to detect much smaller exoplanets. As one star passes in front of another, its gravitation distorts the distant star's light. By monitoring the light over time, a spike in what should be a regular curve reveals an exoplanet orbiting the lensing star. But that's not WFIRST's only objective. There's a gap in knowledge that some cosmologists find embarrassing. When quantifying the speed at which galaxies rotate, and the amount of mass they contain, there's a baffling mismatch. Our laws of physics say that spinning galaxies do not contain enough matter, and therefore enough gravitation, to prevent them from flying apart. And yet, they don't. Astrophysicists came up with the dark matter hypothesis. This is not just matter that doesn't glow. It is matter that interacts with nothing else. The only influence it has is via gravitation. Payloads of dark matter surrounding galaxies is the current method to explain why they move the way they do. Estimates say there is five times more dark matter than matter. Yet no one can explain it. Another fascinating anomaly is that early galaxies, galaxies whose light has taken more than 10 billion years to reach us, rotate at a slower rate. The influence of dark matter is not as pronounced. New generations of telescope are being built to help researchers answer some fundamental questions about the laws governing the universe. But it seems that cosmology has reached a stage where it is just discovering how little we know about what lies beyond the solar system. Arguably the most significant scientific instrument in history has been the Hubble Space Telescope. It has changed our understanding of the universe. Yet it has posed serious new questions about the nature of matter itself. Hubble has confirmed the widespread distribution of black holes and has viewed light from galaxies more distant than anything previously seen. As well as its scientific discoveries are the stunning images. In 1609, Galileo built an early telescope and soon turned it to the night sky. The telescope was a powerful tool that led to a complete re-evaluation of the Earth's place in the universe. In 1668, Newton invented the Reflecting Telescope to eliminate the problem of uneven refraction of the different wavelengths of light. The Reflecting Telescope became the design of choice for astronomers. And in 1781, William Herschel used an instrument he had built himself to discover the planet Uranus. The Newtonian Telescope was scaled up to immense proportions and in 1924, it was with the 2.5 meter Hooker Telescope at the Mount Wilson Observatory that Edwin Hubble realized the Milky Way was not the universe, but just one of countless galaxies. But there was still a problem. No matter how mathematically perfect a telescope is, its images are distorted by the Earth's atmosphere and some wavelengths cannot reach the ground. In 1946, astrophysicist Lyman Spitzer proposed a telescope in orbit above the Earth's atmosphere. The idea clearly outstripped the technology of the time. But by 1966, NASA began launching a series of orbiting astronomical observatories. Only two were successful, but the telescopes, in low Earth orbit, were the first to see the night skies in the ultraviolet part of the spectrum. In the 1970s, plans were drawn up for a large orbiting telescope. Part of its brief was for it to be visited regularly by maintenance crews, something that would become possible when the Space Shuttle went into service. The design called for a 2.4 meter primary mirror, ground within very fine tolerances, because it was required to function well into the ultraviolet spectrum. Originally known as the Large Space Telescope, it was slated for launch in 1979, but delays in construction led to several postponements and the Challenger disaster led to more delays. In 1983, the name Hubble Space Telescope was adopted in honor of the man who confirmed that the universe was expanding. Finally, in April 1990, Hubble was ready for launch. Sound suppression water system has started. T-minus 13 seconds. T-minus 10. Go for main engine start. We are go for main engine start. T-minus 6, 5, 4, 3, 2, 1 and liftoff of the Space Shuttle Discovery with the Hubble Space Telescope, our window on the universe. Hubble has a unique relationship with the Space Shuttle, which would revisit the telescope on five different occasions. At this stage, nobody understood just how vital these missions would be. To deploy Hubble, the Shuttle Discovery set a new altitude record of more than 600 kilometers. Release of the telescope was routine. It took several weeks methodically checking Hubble's control and communication systems before astronomers working at the Space Telescope Science Institute in Baltimore could see Hubble's first images. Results were disappointing. Images were blurred. The telescope's mirror had been accurately ground, but to the wrong shape. An incorrectly assembled test instrument had been relied upon during manufacture and nobody had double-checked. Soon a fix was proposed. Because the error was understood, a corrective device could be fitted to the telescope. A series of small mirrors would compensate for the primary mirror's defect. It was called COSTAR, another of those cumbersome NASA acronyms. Training astronauts for the job of accurately installing COSTAR began. The high-speed photometer would be scrapped to make room for it. In addition to correcting the optics, a number of other modifications were set in train. A new wide-field planetary camera would be installed. The original was obsolete. The telescope's solar panels would be replaced. There would be a new electronics processor, extra magnetometers, and two gyroscopes would be replaced. Ten, nine, and we have a go for main engine start. And we have liftoff. Liftoff of the Space Shuttle Endeavour on an ambitious mission to service the Hubble Space Telescope. Finally, in December 1993, the mission to repair Hubble began. It took the Endeavour three days to catch up with Hubble, and it was clear from the beginning that one solar panel was damaged. The work on the telescope was considerable, and there were five spacewalks scheduled. Firstly, the shuttle's robotic arm grabbed the Hubble so it would remain stable while the work was carried out. The work would be shared between two teams of two astronauts. Over five days, the allocated tasks were carried out, with the Space Telescope Operations Control Centre monitoring the Hubble's performance as each new component came online. Most of the parts replaced were stowed in the cargo bay for return to Earth, the one exception being the damaged solar panel which was set adrift in space. All five spacewalks went according to plan, with the only major problem being the difficulty involved in closing the telescope's doors. On the ninth day of the mission, Hubble was released. It would take controllers on the ground another month to fully check the telescope's new systems. Hubble had been repaired in December 1993, yet it would take close to two months for technicians on the ground to run through a complex series of optical alignments before they could be certain that the telescope was performing correctly. When astronomers finally saw results, they were stunned at the quality it was delivering. The repair mission had been successful. With the services of a powerful new tool finally at their disposal, astronomers at the Space Telescope Science Institute began addressing a pressing list of demands for time on Hubble. In March 1993, a comet had been discovered orbiting Jupiter. Its path suggested that it had only recently been captured by Jupiter's gravitation and that it would soon crash into the planet. No one had ever seen a collision between bodies within the solar system, and opinions differed about how visible the impact would be. Soon, Hubble captured this image of a chain of fragments. Comet Shoemaker-Levy 9 had been shattered by Jupiter's gravitation. The fragments ranged in size from several hundred metres to several kilometres across. Over six days in July 1994, Hubble observed the pieces crash into Jupiter's unlit face. As the planet revolved, the impact region showed a series of black swirls. Some of the dark shadows were as large as the Earth, and they persisted in Jupiter's atmosphere for months. Spectral analysis from Hubble revealed diatomic sulphur and carbon disulfide. These had never been seen at Jupiter, yet the amounts detected were too great to have come in with the comet. They had been stirred up from the planet's interior. Scientists learnt so much from observing these impacts, yet the observations almost didn't happen. Just days before the first impact was due, Hubble began acting unpredictably, and then went into safe mode. It would not execute any of its instructions. Technicians suspected a memory problem, and took measures to switch across to new memory installed on the recent service mission. Things started to improve until the spacecraft reported it had entered zero gyro some point. It appeared that two gyroscopes had failed, and the situation had dramatically deteriorated. The problem was traced to a counter that had run out of numbers with which to count. The issue was understood and quickly resolved. The incident underlined just how complex the operations of the Hubble would be, and that the ground engineers still had to learn how to operate the new hardware. Though Hubble functions as a telescope, it is also a spacecraft that must be controlled with exquisite accuracy. Unlike other spacecraft, it has no thrusters. Chemical residue from rocket engines would quickly contaminate the precision optics. To enable it to move, Hubble has four reaction control wheels. As these spin, the telescope will rotate in the opposite direction. Each one has a mass of 45 kilograms, and they are controlled by the telescope's computer. Combinations of spin in any three wheels will allow the telescope to point accurately in any direction. They are mounted in angled pairs around Hubble's centre of gravity. In 1997, on the space telescope's second servicing mission, one of the reaction control wheels was replaced. It had developed an electrical fault. So that Hubble can remain pointed accurately, three fine guidance sensors positioned toward the back of the telescope will lock onto any of a series of bright guide stars. Hubble only needs two guidance sensors to point with accuracy. They are so sensitive, they can detect wobble in the motion of closer stars. The third vital aspect of the telescope's pointing system is its ability to detect the rate and direction of its movement. Hubble is equipped with six gyroscopes that register its orientation. These are essential when pointing the telescope in a new direction. The gyros spin at 19,200 rpm, and they do wear out. This is why there are six units, even though when designed, Hubble only needed three to function properly. With new algorithms, Hubble can now point with only two gyros, though less accurately. Work has been done to enable it to work at reduced capacity with only one gyroscope. Hubble can point with an accuracy better than two millionths of one degree. This ability to stay fixed on one narrow region of the sky for a very long time led astronomers to perform a unique observation. For ten days in 1995, they pointed Hubble at a small, empty region of the sky near the constellation Ursa Major. To some, this was folly, a waste of valuable observation time. The results astounded everyone. This empty part of the sky was packed with irregular-shaped galaxies. Some were as old as 13 billion years. Hubble had looked back in time at the formation of new galaxies in a range of shapes not seen closer to our own Milky Way. It became known as the Deep Field Survey. It was the first of a series of similar explorations of areas of the sky in which nothing had previously been seen. Animators have added depth to these images by using spectral information, known as redshift, which indicates a body's distance from Earth. This was a completely new area of astronomy, and it was one reason why Hubble had been built, but the Space Telescope was already running into design constraints. Engineers began work on equipment that would upgrade the telescope's performance in the near-infrared part of the spectrum. This would be fitted during the next servicing mission, along with updated support equipment such as a solid-state recorder, which would replace the original reel-to-reel recorder. Yet even at this early stage of Hubble's life, astronomers were realising its limitations. The distant galaxies it had seen were approaching the extent of its view in the infrared end of the spectrum. Even with enhancements, the telescope could not be kept cold enough to observe the large redshift wavelengths revealing very old, very distant objects. Plans were drawn up for a new, larger telescope, known as the Next Generation Space Telescope, that could explore the most distant parts of the Universe. But there was still plenty that Hubble could do better than any other telescope, and there appeared no reason that it wouldn't continue being refitted with the latest technology as it became available. Early in 1997, the Space Shuttle Discovery visited Hubble for a second servicing mission. Astronauts fitted new instruments to improve its reach into the infrared. In 1999, Discovery again visited Hubble. This mission had been brought forward as four of the gyroscopes had failed and the telescope had gone into safe mode. To lift off on Space Shuttle Columbia to broaden our view of the Universe through the Hubble Space Telescope. In 2002, Columbia lifted off with a new instrument, the Advanced Camera for Surveys. It would replace the faint object camera, the last of Hubble's original instruments. When it departed, Hubble was again in peak condition. Less than a year later, at the close of its next mission, Columbia burnt up during re-entry. Foam insulation had come loose during launch, damaging thermal protection tiles. After an exhaustive search for debris and a thorough inquiry, it was decided that the shuttle system was aging. It would be retired upon completion of the International Space Station. All other missions were ruled out. For the Hubble Space Telescope, this was a death sentence. In the corridors of power, Hubble had friends. The sheer beauty of images from the telescope gave it a public profile and key senators began campaigning for one last servicing mission. After a new administrator took charge at NASA, a final shuttle flight to Hubble was reinstated. Work began on what would amount to a rebuild of Hubble's operational systems as well as another update of scientific instruments. New fine guidance sensors would be fitted. All gyroscopes would be replaced. And a new set of batteries would replace the originals that Hubble was still using. Within the imaging spectrograph, a power supply had failed and training began to open up the unit and replace an electronics board. The advanced camera for surveys had also malfunctioned and its repair work was seen as even more detailed. Because the shuttle now had to operate with a backup and because Atlantis could not reach the International Space Station from Hubble's orbit, the shuttle Endeavour was prepared for a rescue mission. On 11 May 2009, Atlantis sat on NASA's Pad 39A ready for Hubble's final servicing mission. This was to be the most complex servicing mission. It had to deliver the longest possible life for the orbiting telescope. The 14-day operation went according to plan. When problems arose, they were solved and Hubble was eventually released as the best telescope it had ever been. On the ground, astronomers from the Space Telescope Science Institute waited for the first pictures from Hubble's new, more sensitive instruments. This means the telescope can operate more efficiently, needing less observing time than earlier incarnations. The team was quick to release examples of the new generation images. In the Eagle Nebula, the Pillars of Creation, thousands of light-years from Earth, immense clouds of hydrogen and dust are giving birth to new stars. The five galaxies of Stephan's Quintet in the Pegasus constellation. Four of these are colliding. 16,000 light-years from Earth. Omega Centauri in the constellation Centaurus is a globular cluster. Stars here are so densely packed that on average there is just one-tenth of a light-year between them. While Hubble has been adding to its list of discoveries, everyone understands that without further servicing, the telescope will eventually die. That slow death has already started, and it's the gyroscopes that are failing. Engineers have been careful, operating just three gyros, keeping the other three as spares. By mid-2018, three of the units had ceased functioning. As the last of the three gyroscopes held in reserve was brought online, it misbehaved, and all science operations of the telescope were suspended. By recycling power to the unit, a measure akin to switching it off and back on, normal function resumed. As further gyros fail, the telescope will use its star trackers to help it point accurately. Hubble is expected to function well into the 2020s, by which time a new space telescope should be in orbit. The next-generation space telescope is now called the James Webb Space Telescope, named after NASA's Apollo-era administrator. The telescope has been completed, and is undergoing exhaustive checks before it is launched. It has a giant mirror of 18 hexagonal segments that is folded before deployment. Compared with Hubble, it is huge, yet the complete spacecraft weighs considerably less than Hubble. NASA built the Webb Telescope in collaboration with the European and Canadian space agencies, and it will launch on an Ariane 5 from the European spaceport in Kourou, French Guiana. It will orbit at the second Sun-Earth Lagrangian point, a place 1.5 million kilometres from the night side of the Earth, where the telescope can maintain a stable position. A large, very thin sun shield will protect the James Webb, allowing the optical components of the telescope to cool to around 50 degrees Kelvin. Its operational temperature must be very low to allow it to see deep into the infrared end of the spectrum. The mid-infrared instrument needs to be colder still. A cooler using helium as its refrigerant will enable the sensor to function at just 7 degrees Kelvin. The Webb Space Telescope is designed to build on the work of Hubble by seeing further back to the earliest emergence of galaxies, stars and exoplanets. NASA Jet Propulsion Laboratory, California Institute of Technology