Coming To Terms With The End of The Falcon 9

Falcon 9 launch A Falcon 9 lifts off September 13 carrying three O3b mPower satellites for SES. (credit: Katalyst Space) Coming to terms with the end of the Falcon 9 by Jeff Foust Monday, September 21, 2026 Last week, space industry executives gathered in Paris for the annual World Space Business Week conference, an opportunity to discuss the state of the commercial space field as well as make deals and announce contracts. “Without naming names, we all know who the go-to easy button was for many years” when it came to launch, said Crain. This year, there were few contracts to announce since the conference was preempted by the International Space Summit, held the preceding week also in Paris. The French government, which organized the summit, encouraged French and other European space companies to make their major announcements there, ranging from launch contracts by Arianespace to satellite manufacturing awards for IRIS², the European secure connectivity constellation. There was, though, plenty to talk about at World Space Business Week beyond contracts. The biggest theme to emerge from both onstage and offstage discussions was that the launch constraints the industry was facing, with demand for launches outstripping supply, may soon get much worse before it gets better. “It seems like all anyone is talking about this week is access to launchers,” said Jeff Thornburg, CEO of Portal Space Systems, a company developing highly maneuverable spacecraft, on one conference panel. The reason it was a hot topic in Paris was not just the problems with existing vehicles, like Blue Origin’s New Glenn and ULA’s Vulcan, which remain grounded, or new vehicles whose first flights keep getting delayed (see “Wishing for rockets”, The Space Review, August 17, 2026.) It was the realization that the rocket that has been the workhorse for the commercial space industry in recent years, SpaceX’s Falcon 9, is going away. SpaceX has not made a formal announcement about the future of the Falcon 9. Earlier this year, though, companies said that SpaceX had stopped taking reservations for rideshare launches beyond late 2028 or early 2029. At World Space Business Week, executives said privately that it was now effectively impossible to buy a Falcon 9 for any commercial mission, either for a rideshare flight or a dedicated launch of a larger satellite or constellation of spacecraft. That is becoming a problem for an industry that had relied on the Falcon 9 as much for its availability as for its price. The Falcon 9 flew 165 times last year and SpaceX executives has previously said they were willing to balance its own demand on the rocket for the Starlink constellation with commercial customers. “Without naming names, we all know who the go-to easy button was for many years” when it came to launch, said Tim Crain, senior vice president and chief technology officer of Intuitive Machines, at the conference. His company has used the Falcon 9 for launching its lunar landers while customers of its satellite manufacturing unit, the former Lanteris Space Systems acquired last year, turned to Falcon 9 for launching large GEO communications satellites. Access to space is “the critical supply chain issue” for the industry, he added. “If we don’t obsolete our own products and services, someone’s going to find a way to obsolete them for us,” Shotwell said. “Look at what we did to the market, right?” That leaves companies worried about where they’ll go for launch when that easy button no longer works. “It would be very undesirable that the key launch service provider for us disappears from the market, so we obviously don’t want to see that,” said Peter Olsen, CEO of Space Norway Satcom, an operator of communications satellites that has previously used the Falcon 9 with a contract for an upcoming launch as well. While the withdrawal of the Falcon 9 from the commercial launch market was something of a revelation at the conference, it was, to some, not a surprise. “This situation was totally predictable. Everybody knew it was going to happen,” said Jean-Luc Maria, CEO of Exotrail, a French company developing orbital transfer vehicles. “You could really see this coming a few months ago. At least, we did,” said Thornburg, whose company announced last month it signed a contract for a Falcon 9 launch of its first Supernova spacecraft—perhaps one of the last commercial Falcon 9 contracts. Portal is working with launch aggregator Maverick Space Systems to sell excess capacity on the launch for rideshare payloads, including a deal announced at the conference to fly ten satellites for Kepler Communications’ data-relay constellation. SpaceX has dropped hints that it would phase out the Falcon 9 in the coming years. “While we have steadily increased our Falcon 9 launch cadence over recent years, we expect Falcon 9 launches to decrease over time,” the company said in May in the prospectus for its initial public offering. That document also suggested that commercial launches would be the first to be wound down for the Falcon 9 while the vehicle continues to fly for government customers. “As Starship progresses toward full operational utilization, the Falcon 9 and Falcon Heavy platforms will remain key assets for specialized missions, including NASA crew rotations and national security payloads,” it stated. SpaceX was absent from launch discussions at World Space Business Week, skipping the conference’s traditional launch panel. Gwynne Shotwell, president and COO of SpaceX, instead appeared at the All-In Summit where she suggested the company was willing to disrupt itself by ending Falcon 9 to focus on Starship. “If we don’t obsolete our own products and services, someone’s going to find a way to obsolete them for us,” she said. “Look at what we did to the market, right? The analogy is there. They were caught flatfooted. We crushed them. Now we want to make sure we are not flatfooted.” The simplest answer is that customers who had used the Falcon 9 will shift to Starship, which is expected to make its first orbital launch as soon as next Monday. However, there remain several concerns in the industry about that, including how available Starship will be for commercial customers. The vehicle will be busy launching next-generation Starlink satellites and Starmind orbital data center spacecraft, as well as supporting NASA’s Artemis missions. There are also issues about the compatibility of commercial satellites with Starship, which initially will have a unique slot-shaped payload bay door. That works well for the “flatsat” designs SpaceX has developed for Starlink and Starmind, but not for more traditional commercial satellites. The company has offered a version with a larger clamshell payload bay door for larger spacecraft, but it’s unclear how available that will be. Starship also does not help on the smaller end of the market, those smallsats that had relied on Falcon 9 rideshare missions for regular and relatively inexpensive access to Sun-synchronous and mid-inclination orbits. SpaceX has hinted in the past it might offer a similar rideshare program for Starship, although the vehicle would likely be oversized for that application. Another simple answer is for other launch companies to step into the gap created by the departing Falcon 9. That has proven to be easier said than done, though, as companies face delays with the first launches and challenges scaling up launches. “There’s two years to go displace what Falcon 9 did, and there’s a couple providers out there trying to do that,“ said Jenkins. “I think they need to hit the gas pedal on it.” “Every single design decision has been thought through to ensure very rapid producibility,” said Jordan Charles, senior vice president of New Glenn at Blue Origin, during the conference’s launch panel. He cited work on an upgraded version of that rocket, called the New Glenn 9x4. “Our vision, our goal for 9x4, is to be operating north of 100 launches a year. That system—the design, everything about it—is intended to scale to those rates.” He did not, though, disclose a timeline for getting to 100 launches a year or intermediate steps along the way, other than reiterating that the company was still working to resume New Glenn launches by the end of this year after a pad explosion in late May. Other launch companies have more modest goals. Arianespace is seeking to reach the current limit of nine to ten Ariane 6 launches a year in 2027, as it works with ESA on studies of potentially increasing that launch rate to as high as 20 per year. “Currently, we are looking at the investment required and, let's say, the overall business case for a cadence increase,” Arianespace CEO David Cavaillolès said on the panel. That would require investments in increased production of vehicles, including solid-propellant boosters, and potentially upgrades to the launch site in French Guiana. “We'll see if we go to 11, 12, 15, or beyond,” he said. “But clearly, we want to be there for our clients.” (Arianespace set a goal this year of seven to eight launches, but has done only four to date, with engine issues reportedly limiting the company to just one or two more launches this year.) At the conference there was a desire by some launch customers, bordering on impatience, for Falcon 9 alternatives. “We have launch vehicles that have not met the timelines they originally promised to be competitive with Falcon 9,” said Thornburg. “There’s got to be a lot of advancement in the next two years,” said Matt Jenkins, chief space systems officer at Vantor, an Earth imaging company previously known as Maxar Intelligence. “There’s two years to go displace what Falcon 9 did, and there’s a couple providers out there trying to do that. I think they need to hit the gas pedal on it.” There are some signs of progress. Earlier this month, Stoke Space announced it raised $1 billion to advance work on its fully reusable launch system called Nova. The company also announced it would scale up the original Nova, now called Nova Pathfinder and capable of placing several tons into low Earth orbit, to the larger Nova Block 2, with a capacity of 15 metric tons. “What we’ve seen in the last year is an explosion in demand for more capable and higher quantities of satellites,” said Stoke Space CEO Andy Lapsa. “We think that we can fly at rate and fly full payloads at the 15-ton capacity.” However, the Nova Block 2 won’t be ready until 2029, with Nova Pathfinder making “multiple” launches in 2027 and 2028 ahead of Block 2’s debut. Based on the experience of past vehicles, it will take several years for Nova Block 2 or other new vehicles in the same payload class—Relativity Space’s Terran R, Rocket Lab’s Neutron, or the joint Firefly Aerospace-Northrop Grumman Eclipse—to start flying at a useful cadence. The problem is worse for smallsat customers who had relied on Falcon 9 rideshare launches, with few options open to them. In an earnings call earlier in the month, Giulio Ranzo, CEO of Avio, said his company has seen a steady stream of smallsat customers interested in his company’s Vega C. “We have a whole bunch of customers coming to us and saying, ‘Oh my God, can you fly us? Because SpaceX no longer gives us a chance to launch,’” he said. “Guess what? The party is finished, and they don’t do that anymore.” However, Avio’s ability to serve those customers is limited. The Vega C has flown twice this year, including a launch of two Earth science satellites for ESA and the European Commission last week, with one more launch expected this year. At the International Space Summit, Ranzo said Avio was looking at ways to increase Vega C’s launch rate to six per year, and then to nine, but did not offer a schedule for doing so. The SpaceX rideshare missions also helped undermine the business case for many small launch vehicles. “What’s happened over the last decade is that one company has dominated this market and it has chilled investment,” said Chris Kemp, CEO of Astra, during one conference panel. Astra had a small launch vehicle, Rocket 3.3, but withdrew it from the market in 2022 after several failures. Kemp argued on the panel that Astra ended that Rocket not because of its checkered flight history but because it was too small to serve a market that had shifted from cubesats to larger spacecraft. He said Astra’s larger Rocket 4, capable of putting up to a ton into LEO, will make its first launch early next year from Wallops Island, Virginia. Astra is planning quarterly launches of Rocket 4 next year as it works to ramp up to weekly launches “within five years.” “What’s happened over the last decade is that one company has dominated this market and it has chilled investment,” said Kemp. In Europe, Isar Aerospace successfully reached orbit for the first time September 5, when the German company’s Spectrum rocket launched for the second time from Norway’s Andøya Spaceport. At a briefing during World Space Business Week, company executives said they are ready to begin commercial launches of the rocket, with a capacity of up to one ton to LEO. Stella Guillen, Isar’s chief commercial officer, said at the briefing that Spectrum’s manifest is fully booked for 2027 and 2028. The company is planning four launches next year and “a little bit more than doubling” that in 2028. Isar’s new factory near Munich can produce up to 40 Spectrum rockets a year. “The market is there and demand is super high,” she said. “What we’re trying to do is scale up as much as we can.” She said that customers are increasingly worried about how they will get their satellites to orbit as the Falcon 9 phases out. “Now there is suddenly this panic mode for 2028 and 2029,” she noted. Others, though, said companies that once pressed the easy button of SpaceX should avoid now hitting the panic button. “We seem to see a lot of panic in the industry over the lack of capacity,” said Portal’s Thornburg. “That’s what gets written, but I don’t believe that’s really the case. I think it’s the lack of ability of the current capacity to meet the growing demand.” “Realistically, people are going to jump into that opportunity and come in to replace that Falcon 9 capacity,” he predicted. That capacity, most in the industry expect, will eventually be replaced, either by SpaceX itself with Starship or competing vehicles. However, they also expect a crunch around the end of the decade as Falcon 9 exits the commercial market while those replacements are still working to fly at sufficient rates to replace that capacity. “Don’t panic. I don’t think this is a structural problem. I think it’s a temporary gap,” said Renato Panesi, chief commercial officer of D-Orbit, which develops orbital transfer vehicles that have flown on many SpaceX rideshare missions. “The critical moment will be 2029 and 2030, and maybe 2031,” he predicted. “That specific window is risky because I think new entrants will not have the launch cadence needed to meet the market demand.” D-Orbit has worked to mitigate its risk by buying launches from others, including Isar Aerospace and Spanish small launch company PLD Space. “I’m really confident that there will be just a small period of time where it will be difficult to have access to space,” Exotrail’s Maria said. “For me, the problem will be solved.” The solutions that emerge, though, may not be everyone. “You have the very small players, cubesats. I think some of them are in trouble because it will be very difficult to access space,” he said. The industry left Paris last week less hopeful about the future than hoping that alternatives will emerge in time. “There’s a lot of development going on,” said Intuitive Machines’ Crain. “What we’re seeing now is the opportunity for the rest of the market to step up and expand and fill the gap.” Jeff Foust (jeff@thespacereview.com) is the editor and publisher of The Space Review, and a senior staff writer with SpaceNews. He also operates the Spacetoday.net web site. Views and opinions expressed in this article are those of the author alone.

The E-3 Samos Program

Samos The first Samos E-1 reconnaissance satellite was launched in October 1960 but failed to reach orbit. The E-1 and E-2 satellites used film-readout technology. The Samos E-3 was a successor proposed in July 1959 using a newer technology and having higher resolution. It was never built. (credit: Peter Hunter Collection) To E-3 or not to E-3: the liminal Samos E-3 program by Dwayne A. Day Monday, September 21, 2026 Samos was the United States’ first reconnaissance satellite program, emerging from a US Air Force project known as Weapons System 117L in 1957. Briefly named Sentry, the renamed Samos quickly expanded into more than a half-dozen different satellite types, resulting in 11 flights, including several launch failures. Because Samos was unsuccessful, it is mostly forgotten in histories of the early space program. But Samos is also problematic because although the Air Force preserved substantial programmatic records, relatively few technical records, illustrations, and photographs survive. It is therefore possible to write a history about when Samos was started and evolved and what decisions were made along the way, but more difficult to describe how each of the spacecraft worked, or show what they looked like Samos A September 1959 Lockheed illustration of the Samos E-3 satellite. The primary optics, which would have gathered the light to focus inside the camera, are omitted. Unfortunately, the source documents are blurry and not high quality, making some of the labels difficult to read. The Samos E-3 was canceled by December 1959. (credit: NRO) From October 1960 to November 1962, the Air Force launched nearly a dozen Samos satellites with designations E-1, E-2, E-5, and E-6. But no Samos E-3 or E-4 satellites were ever launched, making them the most emblematic spacecraft in the Samos program. The E-4 was a mapping camera that was built but never flown, whereas the E-3 only existed for the second half of 1959, never even getting off the drawing board. The Samos E-3 emerged because the limitations of film readout were becoming apparent by summer 1959, but some in the Pentagon did not want to abandon a requirement to return imagery to the ground soon after the photographs were taken, even if very few photographs were actually taken. The Samos E-3 was the ultimate liminal system, caught between two worlds: the readout technology that was fading due to its limitations and the film-return technology that would dominate US satellite reconnaissance for the next two decades. Samos The Samos program consisted of visual reconnaissance satellites with an "E" designation and signals intelligence satellites with an "F" designation. The E-4 and E-5 were never built or flown. In addition to the satellites listed here, the Samos E-6 was also built and flown, but was unsuccessful. (credit: NRO) The Samos E series Samos was initially the overall name for a series of subsystems including the satellite, propulsion system, and payloads that were given letter designations A through G: A for airframe, B for propulsion, C for auxiliary power, D for guidance and control, and G for infrared (missile warning). Most of these designations were quickly abandoned or forgotten by those working on the program except for Samos E, which designated photographic reconnaissance satellites, and Samos F, which designated signals intelligence collection satellites. Both the E and F subsystems soon produced multiple variants. Samos The Samos program had major subsystems designated A-G (later adding subsystem H). In practice, most of these designations were ignored other than Samos E and Samos F. (credit: NRO) In early 1958, a project was spun-off from the Samos E-series. It was a small recoverable satellite that would return its film to the ground. It was given the codename CORONA, and the camera system was designed and built by Itek and managed by the CIA, with the Air Force responsible for launch and tracking and spacecraft operations and recovery. CORONA was covert, meaning that its existence was classified, unknown even to most of the Air Force and contractor personnel working on Samos. Samos E-1 and E-2 were “film-relay” satellites that recorded images on film, developed the film inside the satellite, and scanned that film for relay to Earth via radio transmission. The Samos E-1 was primarily a proof-of-concept system, with the Samos E-2 being the operational system. Up until late 1959, the Air Force expected that eventually multiple Samos E-2 satellites would fly and become the primary Air Force reconnaissance satellite. Samos The Samos E-1 satellite. Film-readout technology had existed in some form for many years. Samos was applying it to a satellite. But it had substantial limitations, leading to a search for better technologies. (credit: Peter Hunter Collection) By 1959, due to the limitations of the E-2’s film-readout technology, the Air Force began considering Samos film-return systems as well, soon leading to the E-4 and E-5 systems in the first half of 1959, and later the E-6 system. Samos E-5 was a high-resolution “film-return” system that took photographs on film and then returned both the film and the camera to Earth inside a large reentry vehicle. “High resolution” meant the ability to spot images on the ground five feet (1.5 meters) large. The Samos E-6, which did not start until 1960, was a broader area reconnaissance satellite. Its photos covered more territory than the E-5, were taken by two cameras operating in stereo mode, and were also recorded on film, but the E-6’s ground resolution was not as good as the E-5. The Samos E-6 film, but not the camera, would be returned to Earth inside a reentry vehicle smaller and pointier than the E-5. The E-4 was a mapping system that would use the same large reentry vehicle as the E-5. Although E-4 mapping cameras were apparently produced, the program was canceled at least in part due to bureaucratic in-fighting, with the US Army insisting that it, and not the Air Force, was responsible for mapping the Earth. Samos This 1959 illustration shows the overall layout of the Samos vehicles. The payloads were different for each of the E-series. (credit: NRO) The Samos E-3 system Lockheed, which was building the Agena spacecraft that supported all the Samos payloads, proposed the Samos E-3 on July 29, 1959, according to an official history of early satellite programs. The history indicates that Lockheed may have assumed that the Advanced Research Projects Agency (ARPA) then running the overall program was more interested in readout satellites like the E-1 and E-2 than in film-return satellites. Lockheed therefore proposed the E-3 as the next step in technology that could return images to Earth relatively quickly, with the goal of achieving five-foot resolution. One of the limitations of the Samos E-2 was that the film supply was finite: once it was used up, the satellite could no longer operate. In contrast, the Samos E-3’s electrostatic storage tape could be reused. Lockheed asserted that the E-2 was “based on pre-1959 concepts” and a new requirement for five-foot ground resolution required new readout technology. The E-2 was a hybrid system, using both mechanical and film development as well as electronic readout. By using a new “electronic tape,” the Samos E-3’s “all-electronic approach would provide the highest possible performance in the earliest time period at minimum cost.” Lockheed asserted that electronic tape systems had already been proven under Aeronautical Research Laboratory contracts at the Wright Air Development Center in Ohio, and claimed that the performance was excellent. Samos The Samos E-3 would have used a new electrostatic tape system for recording images onto a tape storage medium. The images would then be extracted from that tape and transmitted to the ground. The tape was intended to be reusable. (credit: NRO) For the Samos E-3, an image would be recorded on photoelectric-sensitive electrostatic tape. It would then be read out “by deflecting the modulation of an electron beam to scan a portion of the tape, and the view signal amplified and then applied as a modulating signal for transmission to ground stations.” Using a transmission system with a bandwidth of 12 megacycles per second—which would require new tubes to be developed—the system could achieve a readout time of 8.7 seconds per frame. On the ground, the images would then be read on to film for the best possible reproduction. Samos Images taken by a Samos E-3 satellite in space would be transmitted to the ground and then directly read-onto film. Film was a storage medium capable of holding a tremendous amount of data. (credit: NRO) One of the limitations of the Samos E-2 was that the film supply was finite: once it was used up, the satellite could no longer operate. In contrast, the Samos E-3’s electrostatic storage tape could be reused. This, combined with a higher orbit, meant that the Samos E-3 satellite could operate for a year in space. The goals for the E-3 Visual Reconnaissance System were: Active life of approximately one year. Ground resolution of five feet, or 2.5 feet (0.75 meters) with a power changer [unexplained]. Optics of 144 inches (366 centimeters) effective focal length, with F/4.5 optics. Sensor electrostatic storage tape, which would be 60 feet (18 meters) long and reusable. Readout would be an electron beam directly from tape. Ground reconstruction would be an electron beam film recorder operating in vertical, oblique, and stereo modes. The Samos E-3 satellite would operate in a nominally circular orbit at an altitude of 483 kilometers (300 miles). This was a tradeoff. It would be easier to achieve higher resolution in a lower orbit. But a lower orbit would have greater atmospheric drag, decreasing the satellite’s lifetime. The 483-kilometer orbit therefore required a more powerful optical system because the satellite was farther from the targets it was photographing. Samos E-3’s proponents expected that having an all-electronic visual reconnaissance equipment would also maximize optical resolution. Samos A schematic showing the Samos E-3's camera system. Lockheed's illustrations omitted the optics that would focus the image onto the storage medium. Lockheed referred to this as an "all electronic" design because it did not require the film and chemicals used for the E-1/E-2 film-readout technnology. (credit: NRO) As a satellite moves over the Earth, the image inside its camera system also moves. Reconnaissance satellites required some method of compensating for the image motion inside the camera. For the Samos E-3, the exposure period would be only one millisecond, which would hopefully simplify the vehicle attitude control and the camera image-motion compensation. According to Lockheed, the approach provided a very high degree of flexibility in operational use and performance capability. “The system provides a very high resolution capability and, through the use of the reusable image storage medium, permits a payload design of minimum size and weight.” “Breadboard designs of this camera have already been built and tested by Radio Corporation of America, Astro-electronic Products Division.” Lockheed claimed that they determined that it was “technically feasible.” The proposal contains no discussion of the payload optics other than the bare basics. The Samos E-3 was the ultimate liminal system, caught between two worlds: the readout technology that was fading due to its limitations and the film-return technology that would dominate US satellite reconnaissance for the next two decades. The available technical schematics for the proposed satellite payload are simple and difficult to read due to poor reproduction. The “primary optics” that would have gathered light and sent it to the electrostatic tape camera are not described. The overall proposal reads like the kind of reconnaissance payload one would expect from a spacecraft designer rather than a camera designer: for instance, Lockheed lists the focal length as 44 inches (112 centimeters) in one part of a document and 144 inches in another part. That was one of the problems that plagued the Samos program, particularly the E-5 which had a camera that was compromised from the start. In contrast, CIA-led reconnaissance programs like the CORONA tended to maximize camera performance over other criteria. Samos The system for converting the images to stored data in the E-3 camera. Unfortunately, the source documents are blurry and not high quality, making some of the labels difficult to read.(credit: NRO) A short life, a quick death The Samos E-3 would have used the same Atlas rocket that was used for the E-1, E-2, E-5, and E-6. The Air Force’s view was that the Atlas was available, and capable. But Atlas was not cheap, which is why both CORONA and later signals intelligence satellites used the less expensive Thor instead. In December 1959, the Air Force approved the recoverable Samos E-5 to satisfy the five-foot ground resolution requirement and ordered Lockheed to cancel any work on the Samos E-3. ARPA had annoyed the military services with its overall management of military space programs and was removed from having any control over Samos by the middle of 1960. Samos A general early timeline for the American reconnaissance satellite program. Although a major contract was awarded to Lockheed in 1956, it was not until after the October 1957 Sputnik launch that the program began to receive the necessary funding to proceed. (credit: NRO) The first Samos E-1 launched in October 1960 but failed to reach orbit. The second launched in January 1961 and had limited success. A Samos E-2 blew up on its launch pad in September 1961, putting an end to the film-readout program. Three Samos E-5s and five Samos E-6s were launched between November 1961 and November 1962, and all suffered various launch, operational, or recovery failures. But Samos E-3 never even got that far, being canceled while still in the laboratory stage, and quickly forgotten. Sources HEXAGON Mapping Camera History, National Reconnaissance Office, June 1979, pp. 43-44. LMSD Satellite Systems Briefing Part II, The SAMOS Program, September 14, 1959. Dwayne Day can be reached at zirconic1@cox.net. Special thanks to Steve Schultz.

The Orbit Is The Territory

OneWeb Territory takes on new meaning in space when everything is in motion but some things, like satellite constellations, offer permanent infrastructure. The orbit is the territory: rethinking spacepower in a domain where nothing stays still by Michael Uhall Monday, September 21, 2026 A satellite in low Earth orbit circles the planet roughly every 90 minutes, and a constellation of such satellites is never at rest. Each spacecraft continually shifts position relative to the Earth, to its neighbors, and to whatever ground systems it serves. What the constellation produces, however, can be remarkably stable: persistent communications coverage, navigation, remote sensing, early warning, or surveillance. The objects move without pause while the pattern they compose endures, which means that orbital infrastructure occupies space without occupying it in the sense we are used to, where occupation means staying put. This is a basic problem for how spacepower gets conceptualized. For well over a century, thinking about geopolitical power has been organized partly by domain, with land power, sea power, and air power each naming a way in which geography, technology, and political organization all condition how power is generated and exercised. Spacepower operates through infrastructures whose recurrent movements help constitute the usable geography of the domain itself. Spacepower, put bluntly, builds the territory it occupies. Since the beginning of the Space Age, strategists and theorists have understandably tried to extend these traditions upward, comparing outer space to the sea, sometimes to the air, occasionally to the ultimate geographical “high ground.” The analogies remain useful, but they also carry assumptions inherited from terrestrial geography that may not survive the trip. Outer space has its own physical and operational structure, and that structure changes what occupation, access, control, and territory can mean. My suggestion is that what distinguishes spacepower lies in a peculiar relationship between movement and territorial order: spacepower operates through infrastructures whose recurrent movements help constitute the usable geography of the domain itself. Spacepower, put bluntly, builds the territory it occupies. From sea power and air power to spacepower Alfred Thayer Mahan’s classic theory of sea power is the obvious point of departure. For Mahan, maritime power was considerably more than possession of a strong navy. It linked battle fleets to commerce, merchant shipping, overseas bases and coaling stations, lines of communication, coastal geography, national productive capacity, and the policy focus of government. Command of the sea meant that one’s own maritime activity could proceed while an adversary’s was constrained or throttled. This was, in fact, a theory of an entire spatial system. In it, shipping routes connect ports and markets, bases sustain fleets across distance, and straits and chokepoints acquire strategic weight because geography concentrates movement through them. As such, sea power means the capacity to protect these circulatory systems and, when necessary, to block an adversary’s. The current standoff over the Strait of Hormuz offers an unusually stark example: naval power is being used simultaneously to preserve access through one of the world’s most important maritime chokepoints and to restrict the movement of designated adversary shipping. Air power theory developed around a different set of geographical possibilities. Giulio Douhet and Alexander de Seversky both emphasize that aircraft could pass over the obstacles constraining armies and navies, and Seversky in particular argue—with an enthusiasm for strategic bombing that the historical record has not always been kind to—that air power could simply bypass the defensive and logistical “tentacles” of traditional military systems and strike directly at the industrial and political centers on which those tentacles depended. Whatever the fate of the strongest version of that claim, the broader conceptual point holds: access to the air introduced a new vertical dimension into strategy, so that range, altitude, speed, overflight, air superiority, and the projection of force across terrestrial barriers all became central strategic considerations. Spacepower theory inherited elements of both sea power and air power theory. James Oberg’s Space Power Theory emphasizes the distinctiveness of the space environment, the planetary vantage available from orbit, the importance of orbital positions, and the value of situational awareness, while Colin Gray stresses continuity among the forms of strategic power. Geography changes from one domain to another, but strategy remains bounded by political purpose, physics, logistics, and the requirement to secure freedom of action. Geography changes from one domain to another, but strategy remains bounded by political purpose, physics, logistics, and the requirement to secure freedom of action. Other theorists have turned explicitly toward maritime theory. John Klein’s “Corbett in Orbit: A Maritime Model for Strategic Space Theory” draws on naval strategist Julian Corbett to think about celestial lines of communication and the necessarily situational character of any “command” of space. Bleddyn Bowen describes contemporary spacepower as littoral, meaning that near-Earth space remains closely tethered to the surface, to terrestrial infrastructure, and to Earth-based vulnerabilities. Further still, Brad Townsend pushes the terrestrial connection furthest, treating space primarily as an enabling domain for military operations whose consequences remain on Earth. Each of these approaches captures something important about spacepower. Sea power theory directs attention toward communications, circulation, command, infrastructure, and strategic position; air power theory highlights access, altitude, speed, and the consequences of opening a new dimension of movement; the largely terrestrial orientation of Gray, Bowen, and Townsend serves as a standing reminder that most existing space systems remain deeply connected to political purposes and infrastructures on Earth. The difficulty starts when these similarities stop functioning as heuristics and start functioning as substitutes for examining the domain and its constraints and prospects on its own terms. Strange geographies of orbit Outer space is a peculiar geographical environment, and even near-Earth space is already heavily structured by gravity, the Earth’s rotation, orbital mechanics, radiation, atmospheric drag, and the delta-v cost of moving from one trajectory to another. Low Earth orbit, medium Earth orbit, geosynchronous regimes, and cislunar space each present different combinations of accessibility, persistence, usefulness, and exposure, so that what appears from the ground to be an undifferentiated void is in fact patterned throughout by physical relationships. Low Earth orbit is comparatively accessible and increasingly crowded. Medium Earth orbit hosts the major navigation constellations. Geostationary orbit allows a spacecraft to keep a persistent relationship to a particular region of the surface. Lagrange points and other cislunar positions carry their own advantages and constraints. Hohmann transfers, launch windows, maneuver costs, communications delays, and stationkeeping burns that hold a geostationary satellite inside its assigned slot round out the practical geography through which spacecraft actually move. Nearly every politically meaningful interaction within this environment is also heavily mediated by technology: rockets for access, tracking networks for position, ground stations and relays for communication, computers and propulsion systems for modeling and altering trajectories, and sensors for making remote objects visible and usable at all. Extraterrestrial geography therefore has an unusually strong infrastructural character. The environment supplies the physical topology, while technical systems transform portions of that topology into usable pathways, positions, and relationships. The upshot is a domain in which movement itself becomes unusually important to spatial order as such. A ship can stop in a harbor, an aircraft can land at an airfield, and a state can draw a boundary on a map and claim jurisdiction over what lies inside it. But orbital spacecraft operate under different conditions, in which remaining usefully “in place,” in fact, means remaining in motion. Howard Kleinberg captures the strangeness of this well. “Just being in space,” he observes, “requires nonstop motion at extremely high velocities and altitudes,” along paths that take a spacecraft around the whole globe. In other words, outer space hates you: “It is a domain where speed alone can kill.” Orbital infrastructure survives by continually falling around the Earth, and our usual image of territory becomes inadequate at exactly this point. Territory as a pattern of movement Territory ordinarily evokes bounded land—a state, a border, a parcel, a jurisdiction—but political geographers have already complicated that image considerably. Stuart Elden, for example, treats territory as a kind of political technology, produced through practices of mapping, measurement, surveying, regulation, administration, and enforcement that make physical space legible and governable. A border by itself does very little, and asserting and maintaining territorial order requires continuous activity. This broader conception becomes especially useful once we leave the ground, the surface of the Earth. Taken together, these systems produce a political geography whose most durable feature may be the recurrence of movement, and this is where spacepower becomes decisively different from its terrestrial predecessors. Territorialization can be understood as the process through which movements, access, activities, and relationships are organized into a persistent spatial order, so that territory consists partly of trajectories. Terrestrial territory already works this way more than we usually notice, since roads organize movement, shipping lanes channel commerce through Malacca and Suez, borders impede or regulate passage, infrastructure connects some places and bypasses others, and military positions shape who can safely move where. Territory is sustained through routines and repeated practices more than through lines printed on maps (although, in fact, the maps and the practices coproduce each other as often as not). Orbit simply makes the kinetic character of territory much harder to ignore. Consider a single satellite. The object continually changes position while its trajectory remains highly predictable, and that trajectory creates recurring relationships with particular places and other objects. A remote sensing satellite periodically returns over the same regions. A communications satellite establishes recurring windows of connectivity. A navigation constellation coordinates many moving spacecraft so that users on the ground experience a service that is, for all practical purposes, continuous. The object moves and the pattern persists and recurs, which suggests a useful way of expanding the concept. Territory can be understood as a persistent spatial order generated through recurrent movement, access, and use. In orbital space that persistence depends more on the repetition of trajectories than on the fixedness or immobility of anything. A territory in this sense becomes a processual map of access cycles: who can reach a Sun-synchronous orbit, and at what cost; how often a spacecraft passes over a given area; which geostationary slots the ITU has already coordinated, and with whom; and how fast a constellation absorbs the loss of a node, as Iridium did after the 2009 collision with Cosmos 2251. These are territorial questions once territory is understood as an organized regime of movement. These questions also help clarify a distinction between territorialization and legal sovereignty that is easy to blur. Describing orbital space as territorialized does not necessarily require treating orbital shells as national property or proposing sovereign ownership of outer space, and political ordering can occur without converting a domain into conventional sovereign territory. Access can be structured, routes maintained, infrastructure defended, and forms of exclusion can arise even where formal national appropriation remains legally prohibited, which is why outer space may become increasingly territorialized while remaining quite unlike the territorial system familiar on Earth. Spacepower as the production of orbital order This expanded conception of territory suggests a correspondingly broader definition of spacepower. Spacepower is frequently discussed in military terms—space superiority, freedom of action, denial, surveillance, communications, targeting, force projection—and these are in many ways the determinative capabilities. But each also depends on a much larger spatial order. A functioning satellite system requires access to launch infrastructure, communications links, tracking by something like the US Space Surveillance Network, command systems, supply chains, and terrestrial users. Constellations depend on coordinated patterns of movement. Navigation systems depend on timing and geometry. Remote sensing depends on predictable relationships among sensors, targets, and ground stations. Political and military power in space therefore arises partly from the ability to create and sustain these patterns, and spacepower can be understood as the capacity to produce, maintain, alter, exploit, defend, or disrupt the spatial orders through which activity in the extraterrestrial domain becomes possible in the first place. That definition places orbital infrastructure near the center of the concept. As examples, Starlink produces a territorial effect through the coordinated movement of thousands of satellites across a handful of low-orbit shells. A launch complex at Kourou, five degrees off the Equator, a site inherited from the French penal colony, creates cheaper access to geostationary transfer than Baikonur’s higher latitude and constrained azimuths permit .NASA’s Deep Space Network, its antennas spaced roughly 120 degrees apart at Goldstone, Madrid, and Canberra, connects moving spacecraft to a particular national state. These tracking architectures make otherwise inaccessible regions of space legible, and the communications links join widely separated nodes into an actionable whole. Taken together, these systems produce a political geography whose most durable feature may be the recurrence of movement, and this is where spacepower becomes decisively different from its terrestrial predecessors. Mahan’s ocean exists as an element before the fleet sails across it. Hence, sea power organizes and exploits maritime geography through fleets, bases, commerce, and communications. Similarly, air power exploits a preexisting atmospheric medium whose relationship to the surface remains comparatively direct. Spacepower, by contrast, encounters a domain in which the usable political geography is extraordinarily dependent upon the infrastructure operating within the void of the extraterrestrial domain. Orbital routes, constellation architectures, communications systems, launch cycles, and positional relationships all help create the practical territory through which spacepower operates, with the result that territory and infrastructure become unusually difficult to separate. This condition will only grow more pronounced as activity moves further beyond near-Earth orbit. Cislunar operations will depend increasingly on transfer trajectories, rendezvous points like the near-rectilinear halo orbit that would have been used for the lunar Gateway, relay satellites like Queqiao at Earth-Moon L2 (without which Chang’e-4 could not have operated on the far side at all), and carefully timed departures and arrivals. Farther out into the solar system, distance and maneuver costs will privilege pre-positioning, timing, and sustained infrastructure over anything resembling fixed position, and movement will continue to generate structure. Persistence in motion Strategic analogies are indispensable, since nobody begins thinking about a new domain from scratch. The history of sea power provides concepts of command, communications, strategic position, commerce, and circulation, while the history of air power shows how access to a new spatial dimension can transform military imaginaries and political possibility alike. Analogy works well as a heuristic but becomes dangerous when the inherited concept starts to obscure the environment it was supposed to explain. Spacepower builds the territory it occupies through its very exercise, and understanding that territory requires learning to recognize persistence in movement. If orbital territory is constituted partly through recurrent movement, then several familiar strategic questions shift in emphasis. Control depends increasingly on the ability to maintain, modify, or interrupt patterns. Strategic position involves relationships among trajectories, timing, infrastructure, and access. Persistence can mean preserving a functioning architecture even while every individual component moves or is eventually replaced. The question “who controls this area?” thus gives way to a different set of questions: who can reach it, how frequently, along which trajectories, with what supporting infrastructure, who can sustain that access, who can deny it, and how quickly a disrupted pattern can be reconstructed. They reflect the distinctive geography of a domain in which presence usually means motion. Here we see a transformation of the classic political questions of “who should govern?” and “who can govern?” into something materially novel. Accordingly, the lesson extends beyond military doctrine. Politics has always involved the production, occupation, and contestation of spatial orders, and entering the extraterrestrial domain expands the range of spatial orders available to political action while placing pressure on concepts that developed under terrestrial conditions. Spacepower theory—and, indeed, thinking about outer space in general—should take that pressure more seriously. Outer space resembles the sea in some ways and the air in others, and each comparison illuminates part of the picture. But the extraterrestrial domain presents its own combination of scale, motion, gravity, technological mediation, and infrastructural dependence. The Space Age, in fact, has already begun to transform portions of that domain into politically structured space. What results is a territorial order whose stability resides, more often than not, in motion itself. Spacepower builds the territory it occupies through its very exercise, and understanding that territory requires learning to recognize persistence in movement. Michael Uhall (https://www.michaeluhall.com/) is a political theorist at Indiana University East. His research spans political ecology, critical geography, philosophical anthropology, intelligence and security studies, and film and literary criticism. His work has appeared with Bloomsbury, Routledge, and Stanford University Press, in Contemporary Political Theory, Nature and Culture, Critical Horizons, and Configurations, and in public-facing venues such as 3:AM Magazine, Ancillary Review of Books, Extrapolation, Philosophy Now, Radical Philosophy, and Vault of Culture.

A Crater Of Our Own Making-What The SpaceX Lunar Impact Reveals About Moon Governance

crater A Lunar Reconnaissance Orbiter image of the crater (center, partially overlapping another crater) created when a Falcon 9 upper stage hit the Moon in August. (credit: NASA GSFC/Intuitive Machines) A crater of our own making: what the SpaceX lunar impact reveals about the future of Moon governance by Rachita Agrawal Monday, September 21, 2026 On August 5, humanity left another mark on the Moon. A Falcon 9 upper stage, used in the January 2025 launch of two commercial lunar landers, struck the lunar surface after more than a year in space. NASA’s Lunar Reconnaissance Orbiter subsequently photographed the resulting crater, approximately 18 meters across and no more than 3 meters deep. The images revealed bright and dark ejecta extending from the impact site, exposing material that had remained beneath the Moon’s weathered surface. The legal regime governing the Moon was designed for exploration. The future Moon will require a regime for management. The immediate reaction is understandably scientific: What does the crater tell us about the Moon? But there is a more uncomfortable legal question: What does the crater tell us about humanity’s emerging authority to alter a celestial body that international law declares to be the province of all humankind? The impact was not a deliberate act of lunar modification. Nor does it, by itself, establish a straightforward case of international liability. Yet that is precisely why it matters. The event exposes a regulatory problem that will become increasingly difficult as lunar activity moves from isolated missions toward sustained commercial and governmental operations. The legal regime governing the Moon was designed for exploration. The future Moon will require a regime for management. From exploration to alteration The 1967 Outer Space Treaty remains the foundation of international space law. Article I provides that the exploration and use of the Moon and other celestial bodies shall be carried out for the benefit and in the interests of all countries. Article II prohibits national appropriation. Article VI makes states internationally responsible for national space activities, including activities conducted by non-governmental entities, and requires authorization and continuing supervision of such private activities. Article IX is particularly significant. States must conduct lunar activities with “due regard” to the corresponding interests of other states and must pursue exploration so as to avoid harmful contamination of the Moon and other celestial bodies. It also provides for international consultation where an activity could cause potentially harmful interference with another state’s peaceful exploration and use of outer space. At first glance, these provisions appear broad enough to address the SpaceX impact. But they leave a fundamental question unanswered: When does physical alteration of the lunar surface become a legally relevant harm? A crater created by a discarded rocket stage may seem insignificant against a Moon covered by billions of years of natural impacts. Yet the legal significance of the event cannot be measured solely by its physical size. The problem is cumulative. Hundreds of missions could produce hundreds of impacts, as well as discarded components, excavation sites, and construction zones. Lunar activity could eventually create an artificial landscape layered over the natural one. The law has not yet developed a clear concept of the Moon as an environment requiring cumulative protection. The limits of liability The Convention on International Liability for Damage Caused by Space Objects establishes a framework for state responsibility when space objects cause damage. The Outer Space Treaty similarly provides that a launching State is internationally liable for damage caused by its space object or component parts, including on the Moon. But the concept of “damage” presents a difficulty. The conventional liability framework is fundamentally remedial: an identifiable space object causes identifiable damage, and an affected state may seek compensation. That model works relatively well when the victim is a person, spacecraft or other property. It is less clear, though, when the “victim” is the lunar environment itself. The Moon could develop something resembling an environmental externality problem. A mission may be economically rational for the operator while imposing costs on future users. Who has suffered legally cognizable damage when an abandoned rocket creates a crater on an otherwise uninhabited part of the Moon? Can a state claim compensation merely because a lunar geological feature has been altered? What if no state-owned spacecraft, installation, or personnel is affected? More importantly, who represents the collective interest of humanity in the integrity of the lunar environment? The Outer Space Treaty’s language of the Moon as the province of all mankind suggests a collective interest. But the treaty does not establish a dedicated institution capable of acting as the legal guardian of that interest. This creates a paradox. The Moon cannot be nationally appropriated, but individual states remain responsible for activities conducted by their governmental and private actors. A state therefore exercises regulatory authority over its operators while the physical environment being altered is legally characterized as belonging to no state. That gap will become increasingly consequential. The lunar surface is becoming infrastructure The significance of the crater is not to argue or establish that it is the largest or most destructive human intervention on the Moon. It is that it demonstrates how lunar surface alteration can occur even without an actor intending to alter the Moon. The next generation of lunar exploration will involve repeated landings, commercial missions, resource extraction, scientific experiments and permanent infrastructure. NASA’s Artemis program and other national and commercial lunar programs will increase the number of objects moving between Earth and the lunar surface. The Moon could therefore develop something resembling an environmental externality problem. A mission may be economically rational for the operator while imposing costs on future users: dust contamination, altered terrain, debris, loss of scientific information, interference with landing zones, or damage to sites of exceptional historical or scientific importance. The existing principle of “due regard” is valuable, but broad principles alone may not be enough. The international community needs to begin distinguishing between ordinary operational impacts and unacceptable cumulative alteration. From “space debris” to “lunar debris” Space debris governance has largely focused on Earth orbit. The growing lunar economy requires the concept to expand. A discarded upper stage that eventually strikes the Moon is no longer merely orbital debris. It becomes lunar debris. This distinction matters because lunar debris can become permanent. On Earth, environmental systems, weather, and human cleanup can gradually alter or remove waste. The Moon has no comparable natural processes. An artificial object placed on its surface may remain for extraordinarily long periods. International space law should develop a dedicated framework for lunar environmental and surface protection with practical implemented mechanisms. The legal framework should therefore require operators planning lunar missions to undertake a lunar end-of-mission assessment: Where will the spacecraft, upper stage, lander, or other component ultimately go? Could its trajectory create an avoidable impact? Could it interfere with future missions or scientifically significant locations? Such requirements need not prohibit lunar exploration. They would simply translate the principle of responsible activity into operational standards. Protecting the Moon before it becomes crowded The most important lesson from the Falcon 9 upper stage impact is that conversation cannot alone fight the practical reality and address the gray areas in the current space law regime. International space law should develop a dedicated framework for lunar environmental and surface protection with practical implemented mechanisms. Such a framework could establish internationally recognized categories of protected lunar areas, impact-risk assessments for lunar missions, notification requirements for planned or unplanned impacts, standards for disposal of lunar-bound hardware, and mechanisms for sharing information about surface alteration. The objective should instead be to ensure that modification is deliberate, proportionate, transparent and internationally accountable. Article IX of the Outer Space Treaty already provides the legal foundation for such an approach through its requirements of due regard, avoidance of harmful contamination, and international consultation. The next step is to give those principles operational content. A new lunar social contract The crater created last month will eventually become another feature on the lunar map. Future missions may land nearby. Scientists may study the exposed material. In the vastness of the lunar landscape, the physical scar may appear insignificant. Legally, however, it should be treated as a warning. Humanity has entered an era in which the Moon is being visited, landed upon, mapped, used, and increasingly commercialized. The central question is consequently changing from whether humans may explore the Moon to how humans may collectively govern the consequences of exploration. The SpaceX impact illustrates the inadequacy of waiting for a dramatic catastrophe before developing rules. By the time a lunar base is damaged, a scientifically significant site is contaminated, or an accumulation of debris obstructs a valuable landing region, regulation will be reacting to a problem rather than preventing one. The Moon does not need to be treated as a museum, nor should every crater created by human activity become an international legal dispute. But the absence of national sovereignty over the Moon cannot mean the absence of responsibility for what happens to it. The future of lunar governance will depend on whether international space law can move from regulating objects on the Moon to regulating humanity’s cumulative footprint upon the Moon. Dr. Rachita Agrawal is a law and policy scholar specializing in outer space law, policy, and governance. Her research examines emerging legal and regulatory challenges surrounding satellite-based systems, lunar activities, space sustainability, and international space governance. With a PhD in Law, she brings an interdisciplinary perspective to contemporary questions at the intersection of international law, technology, public policy, and the evolving commercialization and exploration of outer space. Note: we are now moderating comments. There will be a delay in posting comments and no guarantee that all submitted comments will be po

The Gulf States Did Not Buy Imagery Independence

Göktürk-1 Göktürk-1 is a Turkish imaging satellite that could be used as part of the Mecca Joint Defence Agreement with Pakistan and Saudi Arabia. (credit: Turkish Defense Ministry) The Gulf states did not buy imagery independence by Kelevitch Monday, September 21, 2026 The most quoted number to come out of the Persian Gulf war this year was assembled from photographs the enemy published. In May, the Washington Post documented damage to at least 228 structures and pieces of equipment at 15 US military sites across six Gulf states. It did so by geolocating 128 satellite images put out by Iranian state-affiliated media and checking 109 of them against Sentinel-2 frames from Copernicus. American commercial imagery had been withheld since April at the request of the US government. A European civil program helped determine what had happened on Arab soil, along with pictures supplied by the state doing the bombing. I argued elsewhere that this should worry the commercial imagery industry. It should also prompt a harder question, and it is the one being asked in Riyadh rather than in Washington: what does a state actually need to own before it can say it can see? The answer arriving in the region is not the one the region says it is getting. The inventory nobody counted On August 7, Saudi Arabia signed the Mecca Joint Defence Agreement with Turkey and Pakistan. The announcement highlighted three assets: NATO’s second-largest army, the only Muslim nuclear arsenal, and the world’s largest oil exporter. In other words, troops, warheads, and money. Nobody counted the satellites, which is odd because, on the specific problem that the war exposed, the orbital ledger is more interesting than the order of battle. What does a state actually need to own before it can say it can see? The answer arriving in the region is not the one the region says it is getting. Turkey operates Göktürk-1, at 0.7-meter panchromatic resolution imager with a roughly two-day revisit over Turkish territory; Göktürk-2, which provides imagery at 2.5 meters and has been operational since December 2012 and flown by the Air Force from Ahlatlıbel; and İMECE, a sub-meter imagery satellite launched in February 2023 and taken into Air Force Command inventory as Göktürk-2B in May 2025. Göktürk-3, a synthetic aperture radar (SAR) satellite promising sub-meter all-weather imaging, is in development. Pakistan’s record is more striking for its pace. PRSC-EO1 launched in January 2025 as the country’s first domestically produced electro-optical satellite. PRSS-2 launched in July 2025 as Pakistan’s first SAR platform. A hyperspectral satellite followed in October, along with PRSC-EO2 in February 2026 and PRSC-EO3 in April 2026. Five satellites in 16 months, two of which launched while the war was running. Saudi Arabia’s own dedicated earth observation assets are SaudiSat 5A and 5B, launched from Jiuquan in December 2018. KAUSTSat, a 6U cubesat carrying a hyperspectral imager at 30-metre resolution, operated for 14 months from 2023 and is no longer in service. Neo Space Group, the Public Investment Fund’s commercial space champion, was founded in 2024 and has nothing comparable in orbit. Read the pact against that and a conclusion suggests itself: Riyadh has connected itself to the two states in the Islamic world with functioning military imaging at precisely the moment it discovered that borrowed imagery can be switched off. I think that reading is broadly right, but it is missing the two things that matter most. The state that solved it is the state that is not in the room The first problem with the argument is Abu Dhabi. Space42’s Foresight constellation now comprises five SAR satellites at 25-centimetre resolution, offering all-weather imagery day and night. Foresight-1 launched in August 2024, Foresight-2 in January 2025, and Foresight-3, -4 and -5 went into mid-inclined low Earth orbits in November 2025, entering full operation on June 9. Imagery feeds an AI analytics platform that the company says converts raw data into usable intelligence within minutes. The full constellation is targeted for completion in 2027. That is better than anything Turkey flies today. It is a working ISR constellation owned by a Gulf state. And yet the UAE is not in the Mecca pact. This is where the story stops being about satellites. Abu Dhabi is, by any operational measure, the Gulf state most deeply committed to this war. The Wall Street Journal reported in July that the UAE provided targeting intelligence and defensive air cover for the Bahraini and Kuwaiti strikes into Iran. It absorbed more Iranian fire than any other country. Its ambassador in Washington wrote in March that a ceasefire would not be enough. So, the one Gulf state with sovereign imaging, and the one with the deepest operational involvement, is outside the arrangement that the region’s other capitals built in response to the same war. American shutter control was replaced, or is being replaced, by Finnish export policy, Chinese launch access, and Turkish tasking priority. The imagery gap in the Gulf is a political, not technology, gap. Riyadh did not lack a regional partner with radar satellites 300 kilometers up its own coastline. It chose Ankara and Islamabad instead. Anyone reading the Mecca agreement as a response to a capability shortfall has to explain why the nearest available capability was passed over. The explanation is the fault line running between Riyadh, Ankara, and Doha on one side and Abu Dhabi on the other, which this war has widened rather than closed. chart A chart of satellite imagery capabilities of countries in the region. (larger version) What "sovereign" survives contact with The second problem is the word everyone in the region is using. Space42 describes Foresight as sovereign Earth observation. In the sense that matters most day to day, it is: the UAE holds the tasking authority, and nobody in Washington can switch it off. That is a real and hard-won difference from what the Gulf had in March. But the Foresight satellites were manufactured in partnership with Iceye, a Finnish company, under a joint venture announced in December 2024. What Abu Dhabi has localized is assembly, integration, and testing—genuinely significant, and the region’s first such facility—rather than design. The company is explicit that localization is a gradual pathway. It is not yet an independent capability, and Helsinki sits inside an EU export-control regime that, as of July 13, has already demonstrated it will delay imagery over the Gulf of Oman when Washington submits a démarche. Apply the same test elsewhere and the pattern holds. Göktürk-1, Turkey’s highest-resolution optical satellite, was built by Telespazio and Thales Alenia Space, with a payload derived from the Pléiades high-resolution imaging satellites. Göktürk-2 is a TÜBİTAK UZAY and TAI platform, but its primary multispectral imager came from Satrec Initiative in South Korea. Pakistan’s satellites are SUPARCO-built and Chinese-launched, every one of them, from Jiuquan, Xichang, Taiyuan and Yangjiang; PRSS-1, the 2018 predecessor, was built outright by China’s CAST. Saudi Arabia’s SaudiSat 3 used imagers from SUNspace in South Africa. The UAE’s DubaiSat-1 and DubaiSat-2 were built by Satrec Initiative; the Hope Mars orbiter was developed with the University of Colorado’s Laboratory for Atmospheric and Space Physics, Arizona State University, and the University of California Berkeley, and launched on a Japanese H-2A. This is understandable. Nearly every space program in history has been built on somebody else’s components, and buying capability is how capability begins. But it does mean that the story the region is telling itself—that the war exposed a dependency and the region is now correcting it—describes the first half of the process but skips the second. What the Gulf has done is change creditors. American shutter control was replaced, or is being replaced, by Finnish export policy, Chinese launch access, and Turkish tasking priority. Those are better bets, because the interests diverge less often and less abruptly. However, they are not the same thing as independence. There is one clear exception, and it deserves discussion precisely because it is so rare. İMECE’s electro-optical camera was designed and built in Turkey, at TÜBİTAK’s optical systems laboratory: a Korsch-type telescope with CMOS TDI sensors developed domestically. Space-grade optics are the threshold. Buses can be bought, integration can be learned, launch can be contracted, but a sub-meter camera you designed yourself is the point at which nobody can quietly stop selling to you. That is what sovereignty looks like in this domain, and across the entire region it currently amounts to one instrument on one satellite. Why the distinction is not academic It would be easy to treat this as a purist’s objection. It is not, and the war showed why. Consider what actually failed between March and July. Gulf air defenses performed: the UAE alone intercepted 537 ballistic missiles and 2,256 drones by early April. Interception was not the problem. The problem came afterwards, when six states discovered they could not establish what had been hit on their own territory, could not attribute strikes, could not contest false claims, and could not assess the effect of their own retaliation. Every one of those functions had been outsourced, and the supplier withdrew mid-conflict without notice. What made that possible was not the absence of satellites in the sky but rather the presence of a decision point outside the region. The question the war posed was not whether the region could obtain imagery but whether anyone outside the region could take it away. On that question, the answer has changed only by degree. A capability with a foreign decision point in it has not removed that vulnerability. It has moved it. If the relevant question is “who can stop me seeing?”, then the honest answer for the Gulf in 2026 is fewer people than in March, and not zero. For Foresight, the answer runs through Finland and, by extension, through Brussels. For Pakistan’s constellation, it runs through Beijing, which controls the launch manifest. For anything the Mecca pact eventually shares, it runs through Ankara’s own tasking queue, which is calculated over Turkish territory and has its own priorities in Syria, Iraq, and the Aegean. None of those chains is likely to be pulled. American shutter control was not likely either, until March 9. What to watch Three things will indicate whether the region is building capability or sovereignty, and they are all observable within a year. The first is whether the Mecca agreement produces an imagery annex. The published text says nothing about space. Turkish officials have described a ministerial committee and a secretariat based in Saudi Arabia. If capability instead moves quietly through the existing Turkish basing relationship—Tariq bin Ziyad in Doha, whatever is agreed in Syria—that tells you the arrangement is in the form of bilateral favors rather than collective capability. The second is whether anyone in the Gulf funds domestic optics. This doesn’t mean integration facilities, constellations bought as a service, or joint ventures, but instead a national program to design and build a sub-meter space camera. Saudi Arabia has the money and, in KACST, an institution with four decades of satellite work behind it. Whether the Public Investment Fund puts capital into an instrument program or into buying more constellations is the single clearest signal of which problem Riyadh thinks it has. The third involves positioning, navigation, and timing (PNT). Imagery answers what happened, while PNT answers where you are. There is no regional augmentation system anywhere in the Gulf, although Turkey’s national space program includes a regional positioning ambition. If the region is serious about seeing without permission, positioning is the second half of the problem, yet nobody has started on it. The Gulf spent this war learning that its eyes belonged to somebody else. The response has been fast, expensive, and, on its own terms, effective. Riyadh will have better imagery in 2027 than it had in February, and it will have it from partners who are less likely to withdraw it. But the question the war actually posed was not whether the region could obtain imagery. Instead, it was whether anyone outside the region could take it away. On that question, the answer has changed only by degree. The distinction between owning a capability and holding a subscription to one is the difference between a state that can see and a state that is currently being allowed to. Kelevitch writes on space policy and Gulf security, working from Arabic and Turkish primary sources. Disclosure: the author has no commercial or contractual relationship with any satellite operator, imagery provider or government named in this article.

The Letters Tsiolkovsky Asked For

Tsiolkovsky Konstantin Tsiolkovsky‘s archive was digitized years ago, but not in a machine-readable or searchable form. (credit: Konstantin E. Tsiolkovsky State Museum of the History of Cosmonautics) The letters Tsiolkovsky asked for A 1934 fireball, 970 pages of replies, and the first machine-readable edition of the father of astronautics’ personal archive by Vladimir Beskorovainyi Monday, September 21, 2026 On May 14, 1934, a fireball passed over the Moscow region. Five weeks later, on June 21, the newspaper Izvestia printed a short note by Konstantin Tsiolkovsky, titled “Who saw the bolide?”, asking eyewitnesses to write to him in Kaluga. As of August 2026, the entire fond—2,019 archival files, 51,008 scanned pages—exists for the first time as a machine-readable corpus. They did. His personal archive preserves 221 files, or 970 pages, of replies: a physics teacher at an agricultural technical school, a livestock specialist, a foreman at a chemical plant, a head of a technical control department, a magazine editor, a village council. The responses spanned weeks: 195 letters in June, 20 more in July. Tsiolkovsky annotated the letters and began an article, “On the bolide of May 14, 1934,” which he never finished. For 90 years those letters have been sitting in fond 555 of the Archive of the Russian Academy of Sciences in Moscow, Tsiolkovsky’s personal fond. They were scanned and posted online years ago. But “scanned” turned out to be a long way from “readable,” and the story of closing that gap is what this article is about. As of August 2026, the entire fond—2,019 archival files, 51,008 scanned pages—exists for the first time as a machine-readable corpus: catalogued, dated, page-classified, and fully transcribed, released into the public domain under CC0 with a permanent DOI and a public search interface. The bolide correspondence is one small, human-scaled corner of it, and for the first time anyone can search it and read it. One caveat up front: the transcription was produced by machine and has not been verified by a human on a single page. Its accuracy has been measured—that measurement, I would argue, is the most interesting part of the project—and the corpus is a finding aid, not a citable edition. Every file links back to the archive’s own scans, which remain the source of record. Nothing previously unknown was “discovered.” Rather, what changed is that the material became usable. Why “digitized” did not mean “accessible” The archive’s portal serves the scans, but working with it as a data source turns up obstacles that only appear when you try to use it. The identifier in a page’s web address does not match the archival file number: id=300 corresponds to file 297 of the first inventory, because 31 files carry lettered numbers (145a, 077b, 585a, and others) that occupy positions in the running numeration without shifting the archival one. Large scans, above roughly 800 kilobytes, are served incomplete, cut off around 130 kilobytes, and can only be retrieved with byte-range requests; a browser silently saves the truncated file. And, in at least four cases, the “variant number” in a file’s description contradicts its own dating: a “second variant” dated September 1, 1920, precedes a “first variant” dated October 19 of the same year. None of this is a complaint about the archive, which did the essential work of scanning and posting 51,008 pages. It is instead an illustration of the distance between putting images online and making a fond queryable. Closing that distance took a catalog of all 2,019 files (published July 30, 2026, with a DOI), dating for 1,969 of them (155 marked as tentative), a per-page classification—34,903 manuscript pages, 14,585 typescript, 1,520 covers and annotations—and, finally, transcription of every page. The measurement problem, and a trick the typewriter era left us Machine transcription of handwriting is easy to produce and hard to trust. To measure its accuracy you need ground truth: the same text keyed in by a human. For an archival fond, no such ground truth exists; if it did, machine reading would be unnecessary. The existing workarounds lean on a model’s internal confidence signals. Dating 1,969 files from the portal’s catalog cards makes it possible, apparently for the first time, to look at the whole fond on a time axis. Personal archives of the typewriter era offer a way out. They often preserve one text twice: the autograph manuscript and a typescript copy made from it, filed together. You can read both with the same pipeline and compare the transcriptions. The text is the same and the pipeline is the same; what differs is only the legibility of the page. The disagreement between the two readings is, almost entirely, a measure of how hard the handwriting is. Fond 555 contains 1,759 such autograph-typescript pairs across 224 files. The two readings agree, at the median, on 37% of the words. The median longest verbatim run shared between them is 10 words, and in 43% of pairs no shared run exceeds that. Those numbers are the honest calibration of what “machine-read manuscript” means here. Is the trick itself trustworthy? It can be checked against real ground truth in the minority of cases where one exists: files whose text also survives in a printed edition. There the paired-reading estimate is unbiased to within a percentage point, and it ranks pages by difficulty the same way the truth does: a rank correlation of 0.67 across all 55 available pairs, rising to 0.92 on the 32 pairs where the printed edition is demonstrably the same text, and 0.97 on the 17 most certain. Against printed editions directly, the earlier pipeline measured 98.1% character accuracy on typescript and 81.1% on manuscript (91.7% and 73.7% at word level). On the completed corpus the picture splits and cannot be averaged: where an archival file contains the same redaction that was printed, character accuracy is 92.3%, and where the file holds a draft or working materials toward an article, agreement with the edition drops to 24%. But that is the distance between a draft and its published form, not reading error. Reporting the negative results A project like this generates findings that would be tempting to leave out. They belong in the article as much as the successes do. Comparing authorial redactions is impossible at this reading quality. The fond preserves “The Space Ship” in two variants, exactly the material for textual comparison. The variants agree with each other on 19% of words, or less than two machine readings of the same page agree with each other. Authorial revision cannot be distinguished from reading error, so that line of work was closed rather than published. The prohibition is built into the tool: the software refuses to display differences that fail the measured threshold. The page classifier passed its test and was still wrong. It separates manuscript from typescript by the spread of ink-stroke lengths, was checked against 19 hand-labeled pages, and got all 19 right. On the 4,675 pages where an independent signal existed to check against, its accuracy is 80%: carbon copies and faded typescript read as manuscript. Two pages cannot be read by any model at all: a German typescript review from 1927 of one of Tsiolkovsky’s brochures trips a built-in filter against verbatim reproduction of known printed text. They were read by a separate route and are flagged in the corpus. What the dating shows Dating 1,969 files from the portal’s catalog cards makes it possible, apparently for the first time, to look at the whole fond on a time axis. Eighty-six percent of the files fall in the last 18 years of Tsiolkovsky’s life. The 1930s account for 58% of the files but only 40% of the pages. In his final years his work does not stop, but it shortens from extended treatises to brief notes. The 1920s contribute 557 files and the 1910s 145. Everything before 1890 amounts to single files, and the post-1935 dates belong to materials about Tsiolkovsky rather than by him. Against that curve, the bolide correspondence of 1934 stands out. A 77-year-old who had moved to short notes still turned a newspaper into a distributed observation instrument, and several hundred readers—teachers, technicians, a village council—answered within weeks. The unfinished article the letters were meant to feed is in the fond too. Transcribing history for $39 The remainder of the fond—41,212 pages—was transcribed in one overnight batch run for approximately $39. The model was chosen by measurement rather than by price list: candidates were compared on manuscript pages against the typescript copy of the same text, with the previous pipeline as the baseline on the same pages. On typical handwriting no model clearly outran the rest; on a 48-page sample the more expensive model was better on 37 pages and worse on 11—a real but modest edge that the material, not the price, turns out to limit. The fond that documents his life is now, for the first time, something anyone can search, with its accuracy printed on the label. One practical finding deserves passing on. Newer models enable “reasoning” by default, and it spends the same output budget as the answer. With a 4,096-token limit, up to 3,929 tokens went to reasoning; the transcription broke off mid-page, and the stub looked like bad reading rather than a truncated response. With reasoning explicitly set to minimum, quality did not drop and output volume halved. What this is and is not The corpus is published only for files transcribed in full, because a partial transcription reads as continuous text with a missing middle. Uncertain readings are flagged word by word and illegible passages, authorial deletions (38,465 of them), insertions and marginalia are marked; pre-reform orthography is preserved as written. The dataset is CC0 with a permanent DOI (10.5281/zenodo.21705221, always resolving to the latest version, currently v1.4.2). The code is on GitHub. The full-fond search includes an English catalog search alongside the Russian. A preprint describing the method is on the arXiv. The scans remain in the archive: the corpus links to them rather than republishing them. An English translation of the corpus was deliberately not made, though its cost was measured at $29 for the whole fond: translating a machine transcription stacks error on top of error and uncertainty flags vanish in translation, leaving an English text that would look more reliable than the Russian it came from. Expert verification has not been performed on any page and is not planned here: that is work for a manuscript specialist, not a machine. The 61st Tsiolkovsky Readings opened in Kaluga on September 15, two days before what would be his 169th birthday. The fond that documents his life is now, for the first time, something anyone can search, with its accuracy printed on the label. Vladimir Beskorovainyi is an independent researcher. The project is not affiliated with or endorsed by the Archive of the Russian Academy of Sciences. Contact: admin@besk.tech; ORCID 0009-0004-7005-6242.

The Martian Anomaly

The Martian Anomaly Mars may be hiding a major temperature imbalance deep beneath its surface, with a new study suggesting that the planet’s southern interior is up to 750 degrees warmer than the north. A research team uncovered the anomaly after analyzing archival data from three NASA spacecraft – Mars Global Surveyor, Mars Odyssey and Mars Reconnaissance Orbiter – looking for subtle changes in their orbital speeds caused by variations in Mars’ gravitational field. Using a technique called “tidal tomography,” lead author Alexander Berne and his colleagues accounted for how the Sun’s gravitational pull on Mars changes as the planet follows its elliptical orbit. The resulting model revealed an interior that looks considerably less uniform than scientists generally assume. "Scientists usually assume that the interiors of planetary bodies are generally spherically symmetric, but this is not necessarily true," Berne explained in a statement. To account for the gravitational differences, the researchers calculated that the mantle beneath the southern hemisphere must be between 390 and 750 degrees Fahrenheit hotter than its northern counterpart – which suggests that parts of it could even remain partially molten. Mars already has a striking north-south divide on the surface: Its northern hemisphere consists largely of lowland plains that may once have contained a vast ocean, while the southern crust is on average 15.5 miles (25 kilometers) thicker and dominated by cratered highlands. The newfound heat difference could help explain other Martian mysteries. NASA’s InSight lander previously found that seismic waves dissipate faster in the south, something a hotter mantle could account for. It could also shed light on unusual magnetic remnants in southern iron-bearing minerals. Mars lost its global magnetic field billions of years ago, but hot material rising from the mantle could have heated parts of the crust beyond their “Curie temperature,” altering their magnetic properties, according to Space.com. But why exactly the southern interior is hotter remains uncertain. One possibility is that a colossal impact more than four billion years ago carved out the northern lowlands and allowed heat to escape, cooling the northern mantle faster. Alternatively, the south’s thicker crust may have acted as an insulating lid, trapping heat below. "The dichotomy that we see between north and south is important to understand because it gives information about processes that may have influenced the hydrology of Mars, including the formation of basins that may have held water," co-author Amirhossein Bagheri said in the statement. The technique could eventually be used to probe other worlds, including Mercury and Jupiter’s large moons, giving scientists another way to investigate planetary interiors without ever touching the surface.