On 25 September 2016, in a remote corner of Guizhou province in south-west China, engineers switched on a radio telescope so large that it fills a natural bowl in the hills. Ten years later, the Five-hundred-meter Aperture Spherical radio Telescope, known as FAST and nicknamed Zhongguo Tianyan, the “China Sky Eye”, has become one of the most productive instruments in world astronomy. According to the National Astronomical Observatories of the Chinese Academy of Sciences (NAOC), which runs it, it has found more than 1,300 pulsars. It has also recorded thousands of fast radio bursts and helped reveal the gravitational-wave background, one of the decade’s biggest results in gravitational-wave physics.
FAST is also a useful lens on how modern China does science. It shows the strengths of a state willing to commit money and land to a long-term project, and the costs, including the relocation of thousands of villagers. It shows scientific openness, with observing time offered to researchers of any nationality, alongside a strong national narrative. And its story is still unfolding: at the time of the anniversary, the telescope was being turned into the centre of a much larger array. This is an account of what FAST is and how it works, what it has found, what it cost the people who lived around it, and what it tells us about China’s scientific ambitions.
A telescope that begins with geology
Most big dishes are built on flat ground and held up by steel. FAST sits inside the earth. Its site is Dawodang, a deep, roughly circular depression in the karst landscape of Pingtang County, where limestone has been dissolved over millions of years into sinkholes, towers and caves. The shape of the land did much of the structural work.
The choice was the result of a long search. The project’s founding scientist, the astronomer Nan Rendong, proposed the telescope in 1994. Chinese accounts describe him spending about 12 years looking for a suitable site, using satellite remote-sensing images and visiting dozens of karst depressions on foot. A NAOC researcher told China Daily in 2017 that Dawodang was chosen from more than 900 candidate depressions after ten years of investigation. Chinese reports give slightly different figures for the length of the search and the number of sites inspected, but they agree on its thoroughness.
The karst setting offered several advantages. It saved a great deal of excavation: a feature published by CGTN for the tenth anniversary estimates the saving at about 500 million yuan. The surrounding hills block much of the radio noise of modern life. The porous limestone drains rainwater quickly, so the electronics beneath the dish are not left standing in water. The site has also proved geologically stable enough for a structure with very fine tolerances.
The Chinese government approved the project in July 2007, and construction began in March 2011. Its cost is usually given as 1.2 billion yuan. That figure has a history of its own: a 2013 China Daily report cited an initial budget of 667 million yuan and quoted a NAOC official as saying that a rise was “inevitable” as costs grew. FAST saw first light in September 2016 and, after years of commissioning, passed national acceptance and began formal operations in January 2020.
Nan Rendong did not see that point. He died of lung cancer on 15 September 2017, a year after the telescope was completed. He has since become a national figure in China, a symbol of patient, unglamorous dedication to science.
How the dish actually works
FAST is often compared with the Arecibo Observatory in Puerto Rico, whose 305-metre dish was the largest single-aperture radio telescope for more than half a century. Both are fixed dishes set into the landscape. The differences show what FAST’s engineers set out to improve.
A surface that changes shape
The reflector is 500 metres across. It is made of 4,450 panels, most of them triangles about 11 metres on a side, mounted on a net of steel cables. Beneath the net, about 2,225 actuators pull on the cable nodes. By adjusting them, the telescope’s control system can deform a patch of the spherical surface, about 300 metres across, into a paraboloid that focuses radio waves onto a single point.
That is the key innovation. A spherical dish does not bring radio waves to one sharp focus. Arecibo corrected the problem with a heavy system of secondary mirrors hung above the dish. FAST corrects it at the main surface instead. As a target moves across the sky, the illuminated patch moves across the dish with it. So although the whole surface measures 500 metres, the effective aperture at any moment is about 300 metres.
A cabin on six cables
The second innovation hangs above the dish. The receivers sit in a feed cabin weighing about 30 tonnes, about 140 metres above the reflector. There is no rigid platform or support tower beneath it. Instead, six cables, each about 600 metres long, run from towers around the rim of the depression and move the cabin like a puppet on strings. Inside the cabin, a secondary mechanism corrects residual movement, for example from wind, so the receivers stay precisely at the focus.
This lightweight design allows FAST to observe objects up to 40 degrees from the zenith, the point directly overhead. Its main instrument is a receiver with 19 beams at the L-band, around 1.0 to 1.5 gigahertz, which lets it survey 19 patches of sky at once. It is this combination of a huge collecting area and a multi-beam receiver that makes FAST so effective at finding faint objects.
Sensitivity versus sharpness
FAST is described as the most sensitive single-dish radio telescope in the world, and in its main frequency range no single dish matches it. But sensitivity is not the same as sharpness. A single dish, however large, has limited angular resolution: at L-band FAST’s beam is about three arcminutes wide, according to a review of its first decade published in The Innovation in 2026. That is enough to detect a source but often not enough to pin it to a particular galaxy. FAST also cannot see the whole sky. From Guizhou, and limited to 40 degrees from the zenith, it misses much of the southern sky, including the centre of the Milky Way. These limitations explain the next stage of the project.
A decade of discoveries
FAST’s results fall into a few main areas: pulsars, fast radio bursts, gravitational waves and neutral hydrogen gas in galaxies. Xinhua reported on 24 September 2026 that research based on FAST observations had produced 13 papers in Science and Nature. It also said FAST results had been listed twice among Science’s top ten breakthroughs of the year and once among Nature’s top ten discoveries.
Pulsars by the hundred
Pulsars are the collapsed cores of dead stars, spinning neutron stars that sweep beams of radio waves across space like lighthouses. They are extraordinary natural clocks, and they are FAST’s speciality. In November 2024 NAOC announced that the telescope had identified more than 1,000 new pulsars. By its tenth anniversary, Xinhua reported more than 1,300, which it said was more than all other telescopes in the world had found over the same period.
These totals should be read with some care. Different sources, published a few months apart, give different figures. The Innovation review, for example, speaks of more than 1,200 pulsars and says they increased the number of known pulsars by about 30 per cent. The difference mainly reflects timing and whether candidates have been fully confirmed. Either way, the scale is remarkable. One survey alone, the Galactic Plane Pulsar Snapshot survey led by Han Jinlin, reported 751 pulsars in a 2025 paper, after covering only about a quarter of its planned area.
Some individual finds are scientifically important in their own right. In 2023 a team using FAST reported in Nature a binary pulsar, PSR J1953+1844, with an orbit of just 53 minutes. That is the shortest known for this kind of system, and it helps fill a gap in the evolution of so-called “spider” pulsars, which slowly strip material from their companion stars. In 2025 the Chinese Academy of Sciences announced another rare system: an eclipsing pulsar orbiting a helium star every 3.6 hours, thought to have formed after the two stars shared a common envelope of gas.
Fast radio bursts
Fast radio bursts, or FRBs, are flashes of radio waves lasting a few milliseconds, arriving from far beyond our galaxy. Their origin is one of the big open questions in astrophysics. FAST’s sensitivity makes it very good at catching them, including faint ones that other telescopes miss.
In 2021 a team led by Li Di reported in Nature that FAST had detected 1,652 bursts from a single repeating source, FRB 121102, in 59.5 hours of observation over 47 days. That was a larger sample from one source than had previously been available. It allowed the researchers to study the distribution of burst energies in detail.
The most recent headline result came in January 2026. A team led by the Purple Mountain Observatory, working with NAOC, reported in Science a sharp spike and subsequent fall in the Faraday rotation measure of a repeating FRB. That measure shows how the radio waves’ polarisation is twisted by magnetised gas along the way. The researchers and Chinese media presented the result as strong evidence that at least some FRBs come from binary star systems. It is a significant finding, although the question of what produces FRBs in general remains open.
Gravitational waves
Pulsars also allow astronomers to search for gravitational waves with wavelengths of light years. If gravitational waves pass between Earth and a set of very stable millisecond pulsars, they cause tiny, correlated changes in the arrival times of the pulses. Detecting that pattern requires years of extremely precise timing.
In June 2023 the Chinese Pulsar Timing Array, using FAST observations of 57 millisecond pulsars over about three and a half years, reported evidence for a background of nanohertz gravitational waves. The signal’s characteristic correlation pattern was measured with a significance of 4.6 sigma. The announcement was coordinated with similar results from teams in North America, Europe, Australia and India. It is important to note the wording: all the groups described their results as evidence, not a definitive detection, which in physics conventionally requires five sigma. Still, the Chinese team reached that level with only about three and a half years of data, far less than the longest-running programmes elsewhere, which shows how powerful FAST is for this work.
Listening for other civilisations
Not every story has been straightforward. In June 2022 an article in Science and Technology Daily, a state-owned newspaper, reported that FAST had picked up candidate signals in a search for extraterrestrial intelligence. The report was deleted soon afterwards, and scientists involved said the signals were most likely radio interference. A paper on a candidate signal from the direction of the exoplanet system Kepler-438 later concluded that it showed the characteristics of man-made interference, and follow-up observations found nothing. The episode was a reminder that sensitivity cuts both ways: the more a telescope can hear, the more carefully it has to screen out human noise.
The cost on the ground
FAST requires radio quiet. To protect it, the authorities created a radio quiet zone and moved people out of the area immediately around the dish. In February 2016 Guizhou officials said they planned to relocate 9,110 residents living within 5 kilometres of the telescope in Pingtang and Luodian counties. Each was to receive a subsidy of 12,000 yuan, with additional help for ethnic minority households with housing difficulties. By September 2016 Xinhua reported that about 8,000 people from eight villages were being moved, with a relocation budget of about 1.8 billion yuan drawn from poverty-relief funds and bank loans. Official figures therefore vary between about 8,000 and just over 9,000, depending on the date and the area counted.
The area lies within the Qiannan Buyi and Miao Autonomous Prefecture, home to large ethnic minority communities. Officials presented the move as part of poverty alleviation: about a third of those being resettled were said to be living below the poverty line, and new homes were built in settlements about 10 kilometres away. The 12 families who had lived in the Dawodang depression itself received three-storey houses and substantial payments. China Daily reported in 2017 that these ranged from 520,000 to 3 million yuan per household, depending on land held. Coverage in state media has emphasised improved roads, housing and jobs. Foreign reporting, including by the BBC, drew attention to the scale of the move. Independent assessments of how relocated residents have fared over the longer term are limited.
The telescope has also reshaped the local economy. Guizhou planned an astronomy-themed tourist area around the site from early on; a 2013 China Daily report cited a planned investment of 4.9 billion yuan in new towns. Today the Pingtang Astronomy Town, with its astronomy museum and visitor facilities, is a well-known destination for school trips. Because the telescope must stay quiet, visitors to the viewing platform are asked to hand over mobile phones and cameras first, an unusual rule in a country where digital payment and constant connectivity are the norm.
Open to the world, within limits
One of the more striking aspects of FAST is how openly its observing time is allocated. The telescope began accepting proposals from international researchers on 31 March 2021. By March 2024, Xinhua reported, it had approved nearly 900 hours for foreign teams from 15 countries. By the tenth anniversary, CCTV reported that 149 projects from 59 foreign institutions had been approved. Calls for proposals invite “users of any nationality or affiliation”, and data from regular programmes becomes public after a 12-month proprietary period.
Demand is intense. The call for the 2024–25 observing cycle reported an average oversubscription rate of 5.6, rising to 7.1 for the times of day when FAST can see the Galactic plane. About 1,600 hours were available for regular proposals in that cycle and about 1,980 hours in 2025–26. The telescope runs for more than 5,000 hours a year. CGTN reported that in its latest full year it observed for over 5,400 hours and produced more than 17.5 petabytes of data.
There are limits. The large survey programmes that take up much of the telescope’s time are led by Chinese teams, and the 2025–26 call specifies that principal investigators for medium-sized programmes must be affiliated with an institution in China. That is not unusual for a national facility, and many Western observatories give priority to domestic users. It does mean that FAST’s biggest surveys are firmly Chinese-led. Foreign scientists interviewed by Chinese media for the anniversary, including researchers from Australia and Germany, praised the quality of the data and the professionalism of the operating team. Such interviews are naturally selective, but they reflect a real international user base.
From a single eye to a compound eye
FAST’s main weakness, its limited resolution, is now being tackled. In September 2024 Jiang Peng, FAST’s chief engineer and deputy director of NAOC, announced that construction had begun on the FAST Core Array. The plan is to surround the big dish with smaller 40-metre antennas and combine their signals through interferometry. This technique lets widely spaced telescopes act, for resolution purposes, like a single instrument as large as the distance between them.
Published plans set out two phases. The first calls for 24 antennas within about 5 kilometres of FAST by 2027. The second would add 40 more within about 30 kilometres by 2030, making 64 in all. Different documents give different performance figures. A 2024 Chinese-language paper projected a resolution of about 5 arcseconds for the first phase, some 33 times sharper than FAST alone. The Innovation review gives about 1 arcsecond for the full array. At the anniversary, Jiang told CCTV that the completed array would improve FAST’s resolution 200-fold and double its sensitivity. CGTN reported his analogy: FAST alone is like sketching with a thick pencil, while the array will be more like a high-resolution digital camera. The goal is precise enough positions to link fast radio bursts to their host galaxies. Coverage of the southern sky, including the Galactic centre, would also improve, because the smaller antennas can point much lower towards the horizon.
The timetable looks demanding. The Innovation review describes a pathfinder of two 40-metre antennas as already built and being commissioned. On 25 September 2026, CCTV reported that the main structure of the fifth antenna of the experimental array had been completed. That is real progress, but still a long way from 24 antennas by 2027. As with many large projects, the dates may slip.
What FAST says about Chinese science
FAST fits a wider pattern in Chinese research: very large, state-funded facilities built to give domestic scientists a lead in specific fields. Other examples include the China Spallation Neutron Source in Dongguan and the Jiangmen Underground Neutrino Observatory, as well as the lunar programme. The approach concentrates money and long-term commitment on a few big bets. In FAST’s case, the bet has clearly paid off. China went from a minor player in radio astronomy to operating a world-leading instrument in a decade.
The official language surrounding it is unapologetically national. “Ten years of dedicated work has secured China’s leading position in global radio astronomy,” Jiang Peng told Xinhua on the eve of the anniversary. Yet FAST is also a reminder that science at this scale depends on international networks. Its pulsar timing results were coordinated with teams abroad; its engineering drew on international design studies, including simulations with a German engineering firm; and its users include scientists from many countries. China is also a member of the international Square Kilometre Array Observatory, the next-generation radio array being built in Australia and South Africa.
Timing has also played a part. The collapse of Arecibo’s instrument platform in December 2020, after cable failures, left FAST as the only giant fixed dish of its kind. For research that needs raw sensitivity at these frequencies, astronomers from around the world now often have to go to Guizhou. That gives China a degree of scientific influence that is not a propaganda slogan but a practical fact.
Conclusion: ten years in the bowl
FAST’s first decade has been, by most measures, a success. It works reliably, runs for thousands of hours a year, and has turned out pulsars, fast radio burst data and gravitational-wave evidence at a rate few instruments can match. Its engineering, with a surface that reshapes itself and a cabin flown on cables, has solved problems that constrained earlier dishes such as Arecibo.
The picture is not without complications. The telescope was made possible by moving thousands of people from their homes, and the long-term effects on them are not well documented independently. Some of the most eye-catching numbers come from institutions with an interest in presenting them favourably, and they differ from source to source. Its access policy is open but not unconditional. And its next stage, the Core Array, is only just taking shape.
Still, the basic story is clear. In a karst hollow chosen after a decade of searching, China has built an instrument that the rest of the world’s astronomers want to use. Whether the dish at Dawodang becomes the centre of a still more powerful array will depend on the next five years of engineering. The first ten years have already changed where radio astronomy is done.
