Reflect Orbital wants to sell sunlight inside a circle five kilometres wide. The permission problem could stretch much farther. A new atmospheric model says light from one planned 54-metre mirror would leave the sky brighter than a moonlit night in the beam's direction for an observer 34 kilometres away. Yet the US regulator that approved the first test satellite says the reflector itself sits beyond its authority.
That mismatch matters before any mirror reaches orbit. The Federal Communications Commission's July 9 order lets Reflect Orbital deploy and operate Earendil-1, an 18-by-18-metre demonstration mirror, at an operating altitude of 625 kilometres. More precisely, it authorizes the spacecraft's radio links. In the same order, the FCC says it has no regulatory authority over the licensing and operation of the solar reflector.
A study submitted on August 6 now gives the light outside the target area a numerical shape. Its estimates are calculations, not observations of hardware in orbit, and most concern a future mirror three times wider than Earendil-1. Even with those limits, the work exposes an awkward sequence: radio approval has arrived before any public, location-specific process for deciding who may illuminate the ground and the surrounding sky.
One satellite, two different approvals
Earendil-1 is a single test, not a license for Reflect Orbital's proposed constellation. The FCC order permits radio operations in UHF, S-band and X-band frequencies so the company can command the spacecraft, test its steerable reflector and download mission data. The satellite is to fly in a near-polar orbit with an inclination of 88 degrees. Conditions in the order cover radio interference, orbital parameters and debris disposal.
The order draws a firm boundary around the bright part of the mission. It calls the solar reflector an activity supported by the licensed communications, then says that activity is outside the FCC's remit and cannot justify denying or adding conditions to the radio authorization. The agency also declined further review under the National Environmental Policy Act. It reasoned that operating the reflector was too far removed from its action of approving spectrum use.
The FCC offered a second argument in case that jurisdictional reading is challenged. It said the record did not show with enough specificity that one limited test would have a significant environmental effect. Much of the opposition addressed a future fleet of tens of thousands of mirrors, while Reflect Orbital had applied for one. The agency therefore refused to treat Earendil-1 as the cumulative start of a larger constellation.
That is legally tidy and operationally strange. Earendil-1 needs radios to deploy and aim its mirror, so a radio license is an essential part of the mission. The light produced by that commanded action is left for other authorities. Space lawyer Michelle Hanlon told MIT Technology Review that national and local permission would still be needed to direct sunlight onto the ground. If scattered light crosses borders, she said, more than one jurisdiction and aviation-safety rules may become involved.
The glow does not stop at the customer
The August paper by Miroslav Kocifaj, Gáspár Bakos and František Kundracik models how reflected sunlight would scatter through air molecules and aerosols, then bounce from the ground back into the atmosphere. The paper, accepted for publication in Astrophysical Journal Letters, examines a 54-by-54-metre mirror illuminating a circular patch with a radius of 2.5 kilometres. That is the proposed production design described by the authors, while the 18-metre Earendil-1 is a pathfinder.
Inside the target, their full-size mirror would appear as a point source about four astronomical magnitudes brighter than the full moon, a brightness ratio of roughly 40 to one. The model puts ground illumination near 9.5 lux, around street-light level. Diffuse scattering would make the sky resemble dusk shortly after sunset, with even the brightest stars lost in the simulated conditions.
Distance changes the view, though less abruptly than a five-kilometre sales footprint might suggest. At 5.4 kilometres from the beam axis, the brightest part of the scattered beam is about 15 times the moonlit-sky level. At 14.1 kilometres it is about four times that level in the beam direction, and most of the sky remains brighter than the paper's full-moon comparison. At 34.1 kilometres, the beam itself falls to about half the moonlit background at its brightest point, while the paper still finds the sky brighter than the moonlit case in the beam's direction once the wider luminous veil is considered.
Scale changes the issue again. The researchers calculate that 400 production mirrors aimed at the same patch would provide about 2,400 lux, roughly 10,000 times full-moon illuminance and 2.4 percent of sunlight. They estimate about 22 watts per square metre would reach the ground, close to the lower limit at which solar panels can operate. The resulting glow could remain visible from 80 kilometres away. Reflect Orbital's larger plan, as described in the study and MIT Technology Review's report, reaches about 50,000 mirrors. No such fleet has FCC approval.
The model has assumptions, and they matter
These figures do not describe a completed field test. The researchers assumed a mirror reflectance of 0.9, a clear atmosphere, an aerosol optical depth of 0.3 and a Rayleigh optical depth of 0.1. They ran cases for a dark surface, typical ground reflectivity and fresh snow, along with two aerosol-scattering patterns. Their baseline geometry places a mirror near zenith and uses a fixed five-kilometre spot.
Clouds are the largest unresolved variable. Dense clouds could block the beam from reaching the surface. Thin cirrus could remove some light and send some of it in new directions. The authors call their cloud-free calculation a best-case scenario for light pollution and say a lit cirrus layer could make the glow visible from farther away, but they do not calculate that distance. A proper cloud model would have to account for cloud type, height, thickness and the positions of the source and observer.
The prototype also needs its own accounting. In the paper's simple baseline calculation, Earendil-1 reflects one ninth as much light as a 54-metre mirror because its width is one third as large. The authors estimate about 1.05 lux at the ground under their prototype geometry, around four times the 0.26-lux full-moon reference. Their detailed three-dimensional sky maps focus on the larger production mirror. Treating the 34-kilometre result as a forecast for the first flight would therefore be wrong.
Reflect Orbital disputes the study. CEO Ben Nowack told MIT Technology Review that some assumptions were inaccurate and said the company's exclusion zones account for scattering. The company did not give the publication numerical details of those zones or its alternative model. In its own statement after the FCC decision, Reflect Orbital promised strict limits, public data and independent review, and said evidence from Earendil-1 would guide whether it scales.
A test can produce the missing evidence
There is a valid reason to fly a pathfinder: models need measurements. Earendil-1 could establish how accurately a thin film can be deployed and aimed, how its real reflectance changes, how the spot moves, and what atmospheric scattering looks like under different conditions. The FCC order says Reflect Orbital has described the satellite as a way to study environmental effects. Those data would be useful well beyond this company.
The test is also where process matters most. In comments filed before approval, 28 scientific and environmental organisations asked the FCC to deny the application or require an environmental assessment. They said the application did not explain how astronomy exclusion zones would be selected or enforced. The FCC received more than 1,800 individual letters as well as institutional filings, according to its order, but concluded that concerns about the single satellite did not justify extra conditions.
A credible demonstration should now publish more than a successful deployment video. Useful evidence would include calibrated brightness measurements inside and outside the beam, pointing and transition logs, weather and aerosol data, the geometry of every test, and records of how observatories and affected communities were notified. Reflect Orbital has promised public data. Earendil-1 will show whether that promise produces a dataset outsiders can audit.
The next decision will be more revealing than the first license. Watch for where and under whose authority the initial ground illumination is approved, whether exclusion zones are disclosed before testing, and whether independent measurements agree with the company's predictions. If Reflect Orbital later seeks permission for larger mirrors or a fleet, regulators will have observations instead of competing models. They will also have to decide which agency is responsible for the light, rather than only the radio signal that tells the mirror where to point.