Read our COVID-19 research and news.

On 1 December 2020, the 900-ton instrument platform of the Arecibo Observatory crashed into its dish, which is cradled in a natural sinkhole.


How the famed Arecibo telescope fell—and how it might rise again

In the early morning of 10 August 2020, Sravani Vaddi, a postdoc astronomer at the Arecibo Observatory in Puerto Rico, was working from home, but her thoughts were at Arecibo’s giant radio telescope. At 2 a.m., she had one precious hour to focus the 305-meter dish on NGC 7469, a distant galaxy. At its center, two supermassive black holes wheeled around each other, following an earlier galaxy merger. Vaddi wanted to see whether having two dark hearts instead of the usual one made the galaxy shine more brightly by stirring up gases and stoking starbirth. Radio emissions from the glowing gases would help her find out.

When she checked in near the end of her observations, computer servers suggested the telescope wasn’t pointing at the galaxy anymore. She couldn’t get an on-site telescope operator on the phone, so she gave up and went to bed.

She woke up to a full inbox. At 2:45 a.m., toward the end of her slot, an 8-centimeter-thick steel cable, one of 18 suspending a 900-ton instrument platform high above the dish, had pulled out of its socket at one end and fallen, slicing into the dish. “I was totally shocked. How could a cable break?” she says. Although she didn’t know it at the time, the photons she gathered from NGC 7469 would be the last ones Arecibo would ever scoop up.

The rest of the story is now well known. A second support cable snapped 3 months later, on 6 November, and the National Science Foundation (NSF), which owns the observatory, said attempting repairs was too dangerous: Arecibo would be dismantled. On 1 December, fate took control as more cables snapped and the platform, as heavy as 2000 grand pianos, came crashing down into the dish.

The loss dismayed scientists worldwide. Although 57 years old, Arecibo was still a scientific trailblazer. Its powerful radar could bounce radio waves off other planets and asteroids, revealing the contours of their surfaces. Other antennas could heat plasma in Earth’s upper atmosphere, creating artificial aurorae for study. And for most of Arecibo’s life, it was the biggest radio dish in the world, able to sense the faintest emissions, from the metronomic beats of distant stellar beacons called pulsars to the whisper of rarefied gases between galaxies.

The public, familiar with the majestic dish from films such as Contact and GoldenEye, also felt the loss. And it was a bitter blow to the people of Puerto Rico, who embraced hosting the technological marvel. Some 130 people work at the observatory, and many more derive indirect economic benefits from it. Every schoolchild on the island goes on a field trip to see the telescope, and those experiences often lead to science careers, says astrobiologist Abel Méndez of the University of Puerto Rico, Arecibo. With its fall, “Puerto Rico loses much more than any other place,” he says.

Along with the grief have come sharper questions. After surviving numerous earthquakes and hurricanes, why did this scientific crown jewel collapse so unceremoniously on a calm winter morning? Some engineers and astronomers think manufacturing flaws or poor maintenance in a tropical, corrosive environment doomed the suspension cables. Others place blame at the feet of NSF’s astronomy division, which for more than a decade tried to offload Arecibo so it could divert funds to operating newer telescopes. “Somehow, we lost a $300 million instrument, a magnificent, really expensive instrument, for a few million dollars,” says Richard Behnke, an Arecibo staffer from 1970 to 1982 who went on to head the geospace science division at NSF. “Things should not collapse like that. It’s not acceptable stewardship at all.”

Meanwhile, astronomers are looking to the future. “First we mourned, then we had a wake, then we got down to work,” says Joanna Rankin, an astronomer at the University of Vermont. Together with Arecibo staff, researchers last month delivered a white paper to NSF describing plans for a new $400 million telescope on the same site. Although any rebuilding effort faces major political and financial hurdles, the proposal aims for an instrument with even more dazzling capabilities than the one that was lost. “There’s been a remarkable amount of commitment and energy,” Rankin says.

Originally, Arecibo had little to do with astronomy. The Pentagon’s Advanced Research Projects Agency funded its construction in the early 1960s as part of an effort to detect and intercept incoming Soviet missiles. Researchers thought radars might be able to spot missile trails left in the ionosphere, the upper part of the atmosphere where the Sun’s radiation ionizes air molecules. But little was known about the ionosphere at the time. Arecibo’s large dish, built in a natural sinkhole in Puerto Rico’s karst landscape, was meant to serve as a giant radar for probing it.

Hail and farewellBuilt into the hills of Puerto Rico, the Arecibo Observatory was a force in radio astronomy for 57 years. But its design left it vulnerable to a harsh tropical climate. Instrument platformThe platform weighs900 tons, stressing thesupport cables. Cableanchor Arecibo,Puerto Rico Gregorian domeAdded in 1997, thedome focuses radiowaves onto receiverscovering a range offrequencies. Line feedThis antenna bouncesradar beams off theionosphere, a mirrorlikelayer of charged particlesin the atmosphere. Tower 8 Suspension cables Tower 12 Access catwalk Karst terrainArecibo was built in a sinkhole in a Swiss cheese landscape of softlimestone rocks. The sinkhole supports the reflector dish and protectsit from radio interference. Reflector dishA 1974 upgrade replaced a wire meshwith 38,000 aluminum panels. Failed cables,detail view below Tower 4 Big eye on the skyArecibo was the world's largest radio dish until it was surpassed in 2016 by a radio telescope in China. Size offers sensitivity to faint objects. How Arecibo worked Radio astronomyArecibo’s spherical dish allows it to see objects that aren’t directly overhead; the telescope “points” by moving receivers on the platform. Runningtrack Arecibo dish 305 m 177 m Radar astronomyIn radar mode, the dish turns emitted waves into a collimated beam. The dish catches reflections from targets in the upper atmosphere or Solar System. Radio waves Radar waves Off-zenithobject Dish Dome Platform Azimuth arm Anatomy of a failureOn 10 August 2020, an auxiliary cable pulled out ofits socket. Then on 6 November, a main cablesnapped. In the weeks leading up to the1 December collapse, staff heardcable wires breaking aboutonce a day. Snapped main cable Auxiliarybackstay cable Main backstaycable View below Socket Rupturedauxiliary cable Main cable Auxiliary cable Dry air Cable wire 160 wires bundledin main cables,8 cm thick Protective paint Cable protectionTo prevent corrosion, cables are painted and air is blown through them. Fall from grace 1. After two cables connected to Tower 4 broke, the remaining four shouldered extra loads. 3. The platform crashed through the dish and onto the hillside. 2. On 1 December, those four cables snapped within seconds of each other. Tower 4 Platform

Upgrades after NSF took over the facility in 1969 made it alluring for more kinds of science. The original wire-mesh surface was replaced with aluminum panels that enabled observations at higher frequencies. NASA added a more powerful radar transmitter that could track Earth-threatening asteroids—and also used it to beam a message to possible civilizations among the stars. In the subsequent decades, a string of high-profile discoveries burnished the telescope’s reputation: a binary pulsar system whose subtly slowing pulses provided the first indirect evidence of gravitational waves, radar maps of Venus’s cloud-veiled surface that revealed evidence for volcanic repaving, and the very first planet outside our Solar System (albeit one orbiting a pulsar).

One of the telescope’s quirks is that the curve of its dish is spherical rather than parabolic like most other radio telescopes. That shape enables the telescope to track objects that aren’t directly overhead, even though the dish can’t tilt. But it also focuses incoming rays to a line rather than a point, requiring elongated receivers. A 1997 upgrade added the igloo-shaped “Gregorian dome,” which housed additional reflectors to focus the radio waves to a point where detectors and transmitters covering many frequencies could be mounted. “It became a completely different telescope and enabled it to stay on the cutting edge,” says Robert Kerr, who was observatory director for two spells in the past 15 years.

The beefed-up scope won a starring role in the NANOGrav project, which in 2007 began to monitor pulsar beats for fluctuations caused by passing gravitational waves. Arecibo also aided the hunt for fast radio bursts, short and powerful blasts that have been one of radio astronomy’s biggest mysteries of the past decade. In 2016, the telescope detected the first burst that repeated, showing that whatever produces the blasts is not destroyed in the process. (Highly magnetized neutron stars are the leading candidate.) “There was a new discovery every year,” remembers astronomer Joan Schmelz, who was deputy director from 2015 to 2018.

Although the 1997 upgrade kept Arecibo in the vanguard, it may also have contributed to its demise. The telescope’s cables were designed with a safety factor of just over two, so everyday loads on the cables would be less than half of the load that would break them. That surprises Robert Lark, a civil engineer at Cardiff University, who says that bridge cables typically have safety factors of six or more

The 110-ton Gregorian dome, added to Arecibo in 1997, boosted capabilities but the added weight may have hastened the platform’s collapse.


The Gregorian dome and other new equipment added 300 tons to the platform. Although six auxiliary cables were added to bring the safety factor back to two, Kerr says it never quite got there. It was one of these auxiliary cables that failed in August. “One of the difficulties of adding or replacing cables is the accurate distribution of load,” Lark says. “The new cable could have been bearing more than it should.”

The end of the cable pulled free from its socket at the top of one of the platform’s three support towers, says engineer Ramón Lugo, principal investigator for Arecibo at the University of Central Florida (UCF), which leads the consortium that now manages the observatory for NSF. Engineers make sockets by inserting the cable end into a cone-shaped steel cavity, splaying the cable’s wires, and filling the cavity with molten metal such as zinc. The zinc adheres to the wires and forms a plug that locks them in place.

Engineers from Cornell University, which managed Arecibo from its construction until 2011, got an unexpected glimpse into one of Arecibo’s sockets in the early 1980s, after an old cable was replaced and shipped to Cornell for inspection. Engineer Leigh Phoenix, who was on the team that carried out the postmortem, says the socket appeared to be faulty. The zinc was distributed unevenly and was poorly adhered to the splayed wires. “It provided an avenue for water to get in,” Phoenix says. The team also found broken and cracked wires in the socket. “It would be alarming if it had been allowed to continue,” he says.

After the August failure of the auxiliary cable, UCF brought in three engineering firms to assess the situation. Their suspicion was that similar manufacturing faults in this cable’s socket were to blame, Lugo says. They did not think the entire structure was at risk—even though staff were hearing individual wires break at a rate of about one per day across all of the telescope’s cables. The wires were known to corrode in the tropical environment, but with 160 of them bundled into each main cable, the breakage didn’t cause immediate alarm.

The lead engineering firm, Thornton Tomasetti, built a full structural model of the telescope. It showed that the four main cables running to the platform from the crippled tower, known as Tower 4, were now bearing a load equal to about 60% of their breaking strength: a safety factor of 1.67. After inspecting the structure, all three firms concluded it was stable and that the loss of another cable wouldn’t cause a collapse.

Thornton Tomasetti recommended replacing all the auxiliary cables because the socket failure made all of them suspect—and because inspections showed some other cables had slipped as much as 1 centimeter from their sockets. Lugo says Arecibo staff wrote up a 500-page proposal for the repairs in 2 weeks. NSF approved the $10.5 million request, and orders were placed for new cables. Then, on 6 November, the second cable broke: a main cable, with just six visible broken wires. And this time, it did not separate from its socket: It snapped.

Big dish, big science

For most of its 57 years, the 305-meter-wide dish of the Arecibo Observatory was the largest in the world. Researchers used it to study Earth’s upper atmosphere, the rocks and planets of the Solar System, and more distant astrophysical objects. Here are some of its milestones.

1974Finds first binary pulsar, a pair of neutron starsthat emits regular radio bursts. 1978Tracks inspiralingof pulsar pair—the first indirect evidence forgravitational waves. 1980Radar maps of 25% of Venus’s cloud-shroudedsurface reveal signs of volcanic repaving. 1989Radar reveals anasteroid’s peanutshape—a “contactbinary” held together by weak gravity. 1991Reflections inshadowed cratersnear Mercury’snorth pole suggest ice deposits. 1994Pulsarfluctuations pointto the tug of rockyworlds: the firstexoplanets. 2007NANOGrav beginsto use Arecibo tomonitor pulsars for signs of passinggravitational waves. 2016Finds a repeating“fast radio burst,”showing somesources survivethe outbursts. 1963Opens underCornell Universitymanagement. Built withDepartment of Defense funding. 1969National ScienceFoundation (NSF) takes over as owner. 2006NSF advisorypanelrecommends closure. 1974Message sent toa star cluster withpictures of theSolar System, DNA, and a human figure. 1997Gregorian dome added to platform, alongwith six auxiliarycables. 2011SRI Internationaltakes overmanagement. 2017Hurricane Mariabatters PuertoRico. “Line feed”antenna fallsinto dish. 2018University ofCentral Floridatakes overmanagement. 2020Suspension cables breakand platformcollapses. 1970 1980 1990 2000 2010 2020 HISTORY SCIENCE
M. Atarod and C. Bickel/Science

The mission to save the telescope was now urgent. The engineers had to reduce the load on the three main cables still attached to Tower 4, now shouldering more than 75% of their breaking load, but they couldn’t risk putting people on the towers or platform. They looked at using helicopters to install extra cables or sever platform components to reduce its weight. They even considered sacrificing the entire 110 tons of the Gregorian dome, but the violent recoil of the platform after the dome was cut loose would have been “a bad thing,” Lugo says. There was no good option.

One firm—Wiss, Janney, Elstner Associates—favored stabilizing the telescope by relaxing the backstays that stretch from the towers to the ground, installing extra support cables, and removing mass from the platform before starting restoration work. But Thornton Tomasetti and the third firm, WSP, concluded that, after two cables had broken well below their design strength, none of them could be trusted. “Although it saddens us to make this recommendation, we believe the structure should be demolished in a controlled way as soon as pragmatically possible,” principal engineer John Abruzzo of Thornton Tomasetti said in his report. So, at a 19 November press briefing, NSF called time on the telescope. “We understand how much Arecibo means to [the scientific] community and to Puerto Rico,” said Sean Jones, head of the Directorate for Mathematical and Physical Sciences. “There is no path forward that allows us to do so safely.”

On 1 December, less than 2 weeks later, Lugo, who had temporarily relocated to Puerto Rico, stopped to buy breakfast before driving up to the observatory. Just after 8 a.m., he got a call telling him the platform had collapsed. “I felt like throwing up,” he says. One hour later he was on-site talking to staff who had heard and felt the crash. “There were a lot of glazed over expressions, they were all crying,” he says. Cameras on a drone had caught the remaining Tower 4 cables snapping within seconds of each other while a fixed camera watched the platform fall. Arecibo’s giant telescope was no more.

So why did cables that had held up the platform for decades suddenly fail so spectacularly? Decades earlier, staff noted cable wires snapping and suspected that corrosion from water was to blame. In 1976, managers tackled the problem by painting the cables to seal them off from the elements and installing fans to blow dry air through the length of the cables. Phoenix says that reduced the rate of wire breaks, but it’s unclear how long those practices were maintained. Kerr says the fans weren’t in use when he took over in 2007, nor was he aware of when the cables were last painted. “Someone may have dropped the ball,” he says.

Lugo insists procedures were continued since UCF took over in 2018. “We were doing what was being done prior,” he says. “It was not poorly maintained,” Rankin agrees. “The Puerto Rico staff are incredible: They did every possible thing.”

Natural disasters hastened the end, Lugo says. Hurricane Maria battered Puerto Rico in 2017. Phoenix says it was “an opportunity for trouble,” because the storm’s winds could have picked up seawater, whose salt makes it especially corrosive, and dumped some on the telescope. The observatory was also shaken by a series of earthquakes in December 2019 and January 2020.

Others say the NSF astronomy division’s efforts to hand off the telescope didn’t help. In 2006, the division convened an independent panel of astronomers for one of its “senior reviews” of existing facilities. To pay for planned new telescopes, such as the Atacama Large Millimeter/submillimeter Array in Chile and the Daniel K. Inouye Solar Telescope in Hawaii, economies were needed. Among other measures, the panel recommended closing Arecibo by 2011 unless partners were found to share operating costs. The astronomy division began to ramp down its roughly $10 million annual spending on Arecibo. NSF’s atmospheric and geospace division increased its funding from $2 million to $4 million and NASA chipped in a few million dollars for tracking near-Earth asteroids. But Arecibo wasn’t out of the woods.

Following an open competition, management of the observatory was transferred in 2011 from Cornell to a collaboration led by SRI International, a nonprofit research institute. NSF’s astronomy division still wanted more savings, however. In 2018, UCF stepped up to take over management, with support from Puerto Rico’s Metropolitan University and the company Yang Enterprises, on the understanding that the astronomy division would gradually reduce its contribution to $2 million annually.

Two management changes in 7 years and the slow dwindling of funds took a toll, supporters say. “People would leave or retire when there are no raises. The best people would go elsewhere,” says planetary scientist Michael Nolan of the University of Arizona, who was Arecibo director from 2008 to 2011. And when old hands move on, something goes with them, Phoenix says. “Knowledge gets lost without that continuity.” In response to questions from Science, an NSF spokesperson says, “Funding from NSF covered scheduled maintenance for the facility and should not have negatively affected the observatory’s ability to maintain the 305-meter telescope.”

Although Kerr is convinced neglect was a factor, he believes the collapse had no single cause. “We drove that telescope hard. It’s an old piece of steel in the tropics, too heavy, it failed.” But he does think the 1997 upgrade, although scientifically valuable, was a mistake. “If it had not been upgraded, it would still be standing.”

After the shock of last month’s collapse wore off, observatory managers gave a group of staff and outside researchers 3 weeks to come up with a plan to replace the telescope. “We need something concrete to put in front of people,” Lugo says. “We want to develop a system that will be relevant for another 50 years.” The planners are aiming for a replacement that would surpass the capabilities of the original, be more flexible, and satisfy the needs of planetary and atmospheric scientists as well as astronomers. And they are trying to do that for less than $400 million—roughly the cost of making a Hollywood blockbuster.

First we mourned, then we had a wake, then we got down to work.

Joanna Rankin, University of Vermont

The researchers first considered a new fixed dish, along with an array of independently steerable smaller ones. But in the white paper delivered to NSF last month, they went with something more ambitious: a flat, 300-meter-wide, rigid platform, bridging the sinkhole, and studded with more than 1000 closely packed 9-meter dishes. The dishes would not steer but the disk would, with hydraulics tilting it more than 45° from the horizontal. At such an extreme tilt, one edge of the disk would be higher than Arecibo’s existing support towers. Steering “will be a great mechanical challenge,” says Anish Roshi, head of astrophysics at the observatory.

In this design, modern receivers built into each dish could cover a broader frequency range than its predecessor and, fired synchronously, the collective radar of 1000 dishes could send out a more powerful beam than a single transmitter. Dubbed the Next Generation Arecibo Telescope, it would be nearly twice as sensitive and have four times the radar power of the original. The steerable platform would enable it to see more than twice as much of the sky as its predecessor, while the field of view of its 1000 dishes would cover a swath 500 times larger.

The extreme tilt was designed to bring an important target within view: the supermassive black hole that sits in the galactic center. The 2020 Nobel Prize in Physics was awarded in part to astronomers who peered through a haze of dust and gas at the heart of the Galaxy to painstakingly track a star following a tortuous orbit in the grip of the black hole. If radio astronomers could discover a pulsar in a similar orbit, its steady clock would allow them to study the behemoth’s gravitational field in fine detail. “It would be a better probe than anything that exists now,” Roshi says.

But some think the plan is a pipe dream. When choosing major projects, NSF and funders in Congress traditionally follow the recommendations of the decadal survey in astrophysics, a priority-setting exercise that at the turn of each decade asks the field what it wants to do next. The current one is already complete and will report in the coming months. “If you skip to the front of the line, those other projects would be furious,” Behnke says.

In theory, Congress could choose to set aside extra funds for a pet project, as happened after the 90-meter telescope at Green Bank Observatory collapsed in 1988. West Virginia’s influential senator pushed through funding for a replacement, resulting in the Robert C. Byrd Green Bank Telescope, inaugurated in 2000 and the world’s largest steerable dish. But Puerto Rico, with only a nonvoting representative in Congress, has little clout, even though it could use a leg up after being battered by earthquakes and hurricanes. “In terms of economy, [Puerto Rico] needs it,” Méndez says.

Lugo says advocates for a new telescope are talking to private foundations. And late last month Puerto Rico Governor Wanda Vázquez Garced allocated $8 million to clean up the site and design a replacement. Lugo says the money will go to a feasibility study of the new design. “We have to be optimistic that we will make this happen.”

But for researchers who relied on data gathered by Arecibo’s big eye, it won’t happen soon enough, leaving them to cast around for other, less capable instruments to continue their work. “I had so many projects in mind,” Vaddi says. “Along with the cable, this broke all my projects.”

With additional reporting by Rodrigo Pérez Ortega.