Maroc - EURASIAREVIEW.COM - A la une - 23/Aug 10:28

The Digital Frontier Of Nuclear Sleuthing: Balancing Benefits And Risks – OpEd

On July 2, 2020, an American weather satellite detected some bright flames from space at 2:00 AM. Iran’s Atomic Energy Organization initially downplayed the news, calling the under-construction industrial shed caught fire. David Albright, a physicist by profession, is the founder of a nuclear non-proliferation nonprofit named the Institute for Science and International Security. The research team at this nonprofit had been closely following and analyzing the construction of the suspicious shed at Natanz, Iran, utilizing commercial satellite imagery. Meanwhile, a young researcher at the Center for Non-proliferation Studies named Fabian Hinz was tracking the fire, too. They were using images released by the Iranian government, commercial satellite imagery, and Google Earth to geolocate the scorched building. Based on their nuclear expertise, each concluded that the Iranian claims weren’t true. The shed was a nuclear centrifuge assembly building at Natanz, Iran's main uranium enrichment facility, and the fire might have been an act of sabotage. Albright and Hinz were not employees of the International Atomic Energy Agency (IAEA) or the U.S. intelligence community. They were independent observers equipped only with open-source information, satellite imagery, and computers. This is the new world of nuclear sleuthing. The commercialization of nuclear threat intelligence comes with its benefits and risks. Estimating Nuclear Dangers: Nuclear Intelligence falls into four broad categories: understanding the capabilities of known nuclear states, the spread of nuclear materials and know-how to non-nuclear states, understanding nuclear accident risks, and preventing strategic surprises like the 1962 Cuban missile crisis or the Indian nuclear tests of 1998. Surveillance of nuclear activities is very challenging because countries work so hard to conceal them. During the Cold War, the Soviet Union built three plutonium production reactors underground inside a mountain. Soviet premier Khrushchev's secret operation to deploy nuclear missiles to Cuba in 1962 employed a deception operation so elaborate that planning documents were hand-carried to a tighter inner circle, and even ship captains carrying the missiles were not told their final destination until they reached the middle of the Atlantic Ocean. Saddam Hussein hid some of his facilities in a large date palm grove and buried telltale power line connections underground to obscure them from overhead reconnaissance. Indian nuclear tests of 1974 were code-named ‘Smiling Buddha,” and the plutonium that was used in making a nuclear device was supplied from the CIRUS reactor. Canada provided the CIRUS reactor to India in 1954, and the United States used to supply heavy water (deuterium oxide) for this reactor. The reactor was not under the IAEA safeguards because the organization had not yet been formed when the reactor was sold to India. Although Canada stipulated the U.S. for the supply of heavy water, the contract with India explicitly specified that it could only be used for peaceful purposes. Assessing the record is also tricky because intelligence failures are public and well-known, but successes are often silent or obscured by events. Everyone remembers the U.S. intelligence community’s flawed Iraq WMD assessment. But almost nobody remembers how U.S. intelligence succeeded in getting Libya to relinquish its nuclear, chemical, and missile programs around the same time. “Intelligence was the key that opened the door to Libya’s clandestine programs,” noted CIA Director George Tenet in February 2004. The Democratization of Intelligence: Three trends have democratized nuclear threat intelligence collection and analysis. Rise in the quantity and capability of commercial satellites, the explosion of connectivity and other open-source information on the internet, and advances in automated analytics like machine learning. In the early Cold War, the United States flew U-2 photoreconnaissance airplanes over the Soviet Union to ascertain how many nuclear missiles and bombers the Soviets had and where they were deployed. In the 1950s, America’s CORONA satellite program ushered in the era of remote sensing from space. In the early 2000s, technological advances and commercialization opportunities converged, dramatically increasing the capabilities, quality, and number of satellites operated by private firms. The first CORONA satellite had a resolution of twelve meters, which meant that the image could not distinguish between two adjacent objects on the ground unless they were twelve meters apart. In 2000, one commercial satellite had a sub-two-meter resolution. By 2019, twenty-five commercial satellites were offering sub-two meter resolutions. Most of them provided resolutions under one meter. The WorldView-3 satellite has the sharpest commercially available imagery, about one foot. With resolutions that sharp, an analyst can detect manhole covers, utility lines, building vents, and even different car models driving on the road from space. Other improvements in satellite capabilities include video and observation of dynamic activities like vehicle movement, construction, and nuclear facility cooling plumes. The introduction of Synthetic Aperture Radar (SAR) has enabled imaging even in cloudy weather, through dense vegetation, and at night. SAR can also detect micro changes in the earth's surface over time, enabling better detection of hidden nuclear activities such as underground tunnel construction. Perhaps the most revolutionary change in satellite imagery is that just about anyone can use it. The cost of acquiring satellite images has plummeted from nearly 4000 USD to as little as 10 USD.  Double Edge Sword: “Knowing the launch location for a test or one that failed used to be the monopoly of intelligence agencies.” That’s no longer true. Nuclear sleuthing by independent observers and analysts offers more diverse perspectives, more hands on deck, and more sharable information. One of the most high-profile debunking successes of nuclear sleuthing involved the effort of an Iranian dissident group to derail negotiations on the Iranian nuclear deal. On February 24, 2015, a group calling itself the National Council for the Resistance of Iran (NCRI) tried to derail international negotiations by claiming that a company named Maritan was housing a secret nuclear facility in a Tehran office basement. NCRI’s evidence includes satellite imagery, photographs of its hallways, and a sizeable lead-lined door to prevent radiation leakage. Within a week, Jeffery Lewis's team at the Center for Nonproliferation Studies showed conclusively that all of NCRI’s evidence was fabricated. Lewis’s team found that Martin was a real company and even had LinkedIn employees. But Maritan had nothing to do with nuclear enrichment. It specializes in making secure documents like national identification cards.  Non-governmental nuclear sleuths have also uncovered important new information about North Korea's clandestine nuclear activities, which has aided intelligence agencies and influenced policy. In 2012, Stanford’s Siegfried Hecker and Frank Pabian determined the locations and supporting tunnels of North Korea’s first two nuclear tests using commercial imagery and publicly available seismological information. Their assessments proved highly accurate when North Korea revealed the actual test sites six years later. This non-governmental ecosystem also generates risks. Some arise when information is wrong. Others arise when information is right. The non-governmental nuclear sleuths can inject errors into the policymaking world. While the examples above highlight the best of non-governmental nuclear intelligence, the landscape is filled with questionable data, shoddy analyses, pet theories, and political agendas. Many amateur imagery analysts are well-intentioned but poorly trained. As Frank Pabian notes, there’s a common misperception that “anyone can look at pictures.” In reality, imagery analysis requires considerable skill and training to know how shapes, sizes, scales, textures, perspectives, and contexts can obscure or delineate different objects seen from space. Nuclear imagery requires understanding the nuclear fuel cycles so that the analysts know what to look for in the first place. The real value of satellite imagery lies in the analysis of what it means. Pictures may be commoditizing, but analysis is not. That’s why U.S. laws currently restrict the resolution of American commercial satellites so that government satellites still have an advantage.