The Lyceum: Space Economy Weekly — Jun 25, 2026

This was a substantive week, anchored by one story that should reset a few threat models: a small California startup pointed a GPS receiver at Earth from orbit and confirmed, with hard numbers, that satellite navigation is functionally broken across large stretches of Europe and the Middle East — and that this is now a constellation-operations problem, not just a defense one. Around that, the plumbing of the space economy kept maturing: Japan's H3 didn't just return to flight, it immediately booked its next government navigation satellite; ESA's EarthCARE crossed from science mission to operational forecast input; and a quiet trio of regulatory and servicing moves — TraCSS going multinational, the FCC tightening Amazon's deployment clock, NASA's Swift rescue mission reaching launch prep — added up to governance and infrastructure catching up to the hardware already in orbit.

• Starlink Group (Falcon 9) (SpaceX): Two dozen more Starlink satellites reached low Earth orbit from Vandenberg on the night of June 24, with a second Florida batch in its window — at least four Falcon 9 missions in roughly a week.
• H3-30 maiden flight (JAXA): The first all-liquid, booster-free H3 configuration flew successfully on June 12, validating the adapter fix and clearing the path back to government payloads.
• Pulsar-0 (Xona Space Systems): An experimental LEO navigation satellite delivered the first space-based map of GPS jamming across Europe and the Middle East.
• EarthCARE cloud data (ESA/JAXA): The mission's simultaneous radar-lidar cloud profiles are now being folded into operational weather forecasting for the first time.
• Dragon CRS-34 (SpaceX): The uncrewed cargo Dragon splashed down off California on June 17, returning research samples from SpaceX's 34th resupply run to the ISS.

Every satellite operator knows GPS jamming is a problem. Nobody had measured it from orbit until now.

Pulsar-0 — operated by Burlingame, California startup Xona Space Systems, orbiting roughly 310 miles up as a pathfinder for a planned 300-satellite low Earth orbit navigation constellation — switched on a GPS receiver expecting some interference and found something worse. The satellite observed attacks that disrupt signals and broadcast false positions across Europe and the Middle East, with degradation particularly severe near Russia's western border and in Middle East conflict zones. Xona's team said the scale was "quite a bit more than we expected."

Here's why this isn't a phones-and-cars story. When a LEO satellite loses GPS, it can't determine its position to photograph a target area, can't orient itself, and can't accurately point its telecommand antenna at the ground. Starlink uses GPS for collision avoidance. Jamming is now a constellation-operations problem for every LEO operator. Separately, a University of Texas team led by Todd Humphreys traced some continental-scale interference to Russia's Kosmos 2546 satellite, noting the outages clustered on Tuesdays through Thursdays during business hours — scheduled operations, not accidents, per Gizmodo's account of the study.

The business case for GPS-independent navigation just hardened considerably. Xona — fresh off a $170 million Series C — plans to launch six satellites this October. The signal to watch: whether defense and intelligence customers show up in Xona's order book before that launch. If they do, alternative PNT (positioning, navigation, and timing) has become a procurement category, not a research thesis.

Flying a brand-new configuration while recovering from a failure is a double-or-nothing bet. On June 12, H3 took it.

The maiden flight of the H3-30 configuration — three LE-9 engines, zero solid boosters — placed its payloads into orbit and validated the adapter fix blamed for December's loss of the Michibiki-5 navigation satellite. The speed of what came next is the real signal. Within days, JAXA scheduled Michibiki-7 — Japan's next GPS-augmentation navigation satellite, the exact payload class lost in December — aboard a later H3 flight. That's not probation; that's a cleared program. The manifest behind it is unforgiving: HTV-X2 cargo resupply to the ISS targeting roughly July, followed by the Martian Moons eXploration mission, which must depart within a narrow planetary window in Japanese fiscal year 2026 or slip two years to the next Mars opportunity.

This resolves a trigger we'd been watching since December. The next proof point is whether HTV-X2 holds its July window — sustained cadence, not a single rescue flight, is what tells you Japan's launch capacity is reliable again for everyone who books on it.

Most Earth observation satellites take years to go from "launched" to "changes how forecasters work." EarthCARE just did it.

The ESA–JAXA mission carries four instruments — a cloud profiling radar, an atmospheric lidar (a laser that measures thin clouds and aerosols), a multispectral imager, and a broadband radiometer — that all observe the same column of atmosphere at once, something no prior mission has managed, per the Copernicus Atmospheric Measurement Techniques overview. That simultaneity is the innovation: it lets EarthCARE measure directly how clouds and aerosols influence Earth's energy balance. The European Centre for Medium-Range Weather Forecasts now wants to assimilate that data in real time, on the theory that a truer picture of cloud locations at forecast start yields better predictions of how weather systems evolve.

This is the moment a research mission becomes operational infrastructure. For commercial weather firms, reinsurers, and agricultural analytics shops licensing ECMWF outputs, it's a data-quality upgrade they didn't pay for. The verification to watch: whether ECMWF publishes forecast skill scores showing measurable gains attributable to EarthCARE. That's what turns a press release into a procurement argument — and what tells you whether this assimilation sticks or stays experimental.

Most on-orbit servicing coverage reads like a concept car: glossy, promising, rarely on an actual road. This one reached the launch pad.

Katalyst Space's LINK servicing spacecraft has arrived at NASA Wallops, awaiting encapsulation on a Pegasus XL, and is slated to rendezvous with NASA's Neil Gehrels Swift Observatory and raise its orbit. Swift is a working science satellite, not a disposable demo — which is exactly why this matters. If LINK can safely meet, capture, and boost a spacecraft already on orbit, life-extension moves from conference panel to a line item operators can price into fleet planning. Every extra year of useful service is a year an operator doesn't have to rush replacement hardware onto a crowded manifest.

Failure looks like a missed rendezvous or a botched capture, and servicing stays theoretical for another budget cycle. The observable signal either way: whether NASA and Katalyst disclose the rendezvous sequence and orbit-raise performance in operational detail afterward. Detail means we've entered the "judge the repeatability" phase. Silence means the demo wasn't clean.

Spectrum licenses are restaurant reservations with orbital mechanics attached: miss your slot and someone else wants the table. On June 5, the FCC granted Kuiper Systems — labeled "Amazon Leo" in the order — a limited waiver on the first-generation deployment milestone for its 3,232-satellite LEO broadband constellation, which had faced a July 30, 2026 halfway deadline.

The waiver is real relief; the conditions are the story. The FCC tied spectrum consequences to deployment progress, limiting priority treatment for satellites launched after the original July milestone unless Amazon meets specific build-and-launch conditions. Runway granted, but the sprint is now being timed. The constellation remains licensed and partially launched, but nowhere near half deployed — and the agency just turned that delay into a measurable penalty box.

Separately, the FCC's Space Bureau opened new processing rounds for non-geostationary broadband systems across Ku-, Ka-, and V-band spectrum, with a July 6, 2026 cutoff for applications to be considered together — a live deadline, not a past one. The thing to watch: whether Amazon discloses manufacturing or launch-booking rates before the conditions bite. Disclosure means the waiver is a genuine bridge. Continued silence means the milestone is already conceded and the spectrum penalty is the working assumption.

Space traffic management empties rooms until a conjunction alert ruins someone's quarter. This update matters more than it reads. The Office of Space Commerce says its Traffic Coordination System for Space has now onboarded 52 pilot users plus two national government accounts — the United Kingdom and Australia — together covering more than 11,125 satellites in orbit.

TraCSS is meant to become a civil, operator-facing source of conjunction and collision-avoidance data — a way to keep LEO from becoming a demolition derby. A traffic system with one country's operators is a neighborhood watch. With allied governments joining, it starts to look like the skeleton of an actual regime that multinational fleets plan around. The signal to track: whether more allied governments and large commercial operators keep joining over the coming weeks. That's how you tell a pilot from a standard — and whether the U.S. civil traffic system becomes the default coordinating layer or stalls as a demo with a handful of friendly users.

Not every important mission is a debut or a disaster. On June 17, NASA confirmed SpaceX's uncrewed Dragon splashed down off California, completing the company's 34th commercial resupply run to the ISS and returning cargo and research samples.

Why include something this routine? Because "routine" is the whole point of a transport market. A working LEO economy needs logistics that behave like freight — cargo up, cargo down, schedule met, hardware recovered — not one-off expeditions. That repetition is what supports stations, research customers, and the commercial platforms now in development. Boring logistics are what make orbital business possible; the day cargo delivery stops being newsworthy is the day the market actually exists. The next thing to watch: whether NASA's station logistics mix keeps diversifying or stays concentrated in a small number of proven providers — the difference between a competitive market and a single-vendor dependency.

A Russian satellite jamming GPS on a Tuesday-through-Thursday office schedule, a Japanese rocket that lost its boosters and got cheaper for it, and a cargo capsule splashing into the Pacific so reliably that its biggest achievement is being boring. The most honest signal of the week came from a startup that turned on a GPS receiver in orbit expecting noise and instead discovered the navigation backbone of two continents keeps banker's hours — which means somewhere a counterspace operator is taking the weekend off while half the world's drones lose their way.

Watch the sky, read the dockets.

Forward this to the satellite operator you know who still assumes GPS is always there and always clean.