A space mirror is a large reflective structure placed in orbit to redirect sunlight to a chosen place on Earth — lighting a city after dark, illuminating a disaster zone, or shining on a solar farm at night. The idea is old, the engineering is hard, and the argument about whether it should be done at all is older than the technology. Russia tried it twice in the 1990s; a California startup called Reflect Orbital is trying it now, with a demonstration satellite authorised in July 2026 and a first flight targeted for 2027. Nothing of the kind has yet delivered useful light to a paying customer.[1][2]
The physics of a bright patch
The numbers explain both the appeal and the scepticism. Noon sunlight delivers roughly 100,000 lux at the surface; a full moon delivers about 0.1 lux. Reflect Orbital’s first target is 0.1 lux over a spot a few kilometres across — genuinely moon-bright, and useful for visibility, but nowhere near enough to run a solar panel, because photovoltaics need hundreds of watts per square metre rather than a dim wash of light. The company’s published roadmap does project far higher levels later: about 2 lux in 2027, up to 100 lux in 2028 and 36,000 lux by 2035, along with thousands of satellites. Those are company targets, and no source reviewed for this article reports a flight measurement.[3][4]
The geometry is unforgiving in a second way. A satellite at 650 km moves at about 7.5 km per second, so it can hold a spot for only about four to five minutes before disappearing over the horizon — which is why the roadmap needs hundreds and then thousands of satellites to offer anything resembling continuous service. Each satellite is also a steerable structure that must hold its shape to within a fraction of a degree to keep the beam on target.[5]
The Russian precedent
The idea was tried first, and for a practical reason. Vladimir Syromyatnikov had been working on solar sails and needed an economic justification; he repurposed the reflective sail as a mirror to light Siberian cities through the polar winter, extending the working day and cutting lighting costs. The project, called Znamya — Russian for banner — was funded by the Space Regatta Consortium and overseen by Roscosmos.[6]
Znamya 2 flew in February 1993: a 20 m mirror of aluminised Mylar unfurled from a Progress cargo craft by centrifugal force, casting a spot about 5 km wide that raced from southern France through Switzerland, Germany and Poland into western Russia. The light was reported at roughly the brightness of a full moon, with some ground observers seeing a flash through cloud. It worked, but the reflected light was dimmer and more diffuse than expected, the structure was hard to keep stable, and the beam’s rapid sweep made it of little practical use. The mirror burned up over Canada.[7][8]
Znamya 2.5, with a 25 m mirror and a beam intended to be five to ten times brighter, failed in February 1999: an extra command deployed the Progress craft’s docking antenna during unfurling, the foil caught on it and tore, and the assembly burned up over the Pacific the next day. A planned 70 m Znamya 3 never secured funding. Even before that failure, the Royal Astronomical Society had written to the project’s sponsors objecting that large mirrors would add to light pollution and jeopardise observatories built at great cost in dark places, and ecologists warned about the effect of artificial light on Arctic plants and on animals whose hibernation and migration are triggered by day length.[9][10][11]
Every version of this idea has run into the same objection, and it has never been an engineering one: the sky is a shared resource, and nobody asked the other eight billion people whether they wanted a second moon.
The current attempt
Reflect Orbital, based in Los Angeles and led by Ben Nowack, is pursuing the same goal with modern deployable structures. Its first spacecraft, Eärendil-1, is a 134 kg satellite carrying four triangular reflector panels that unfurl into an 18 m square — roughly twice the size of a microwave oven before deployment. It is to launch to about 500 km, deploy and check out, then raise its orbit to 650 km for demonstrations that would put a moon-bright spot on a target for about five minutes, covering something like 24 square kilometres.[12][13]
The regulatory record is unusually specific. The FCC granted conditional authority on 9 July 2026 for one demonstration satellite — not a constellation — using UHF at 401–402 MHz plus S- and X-band for telemetry and testing, with deployment at 510 ± 20 km, operations at 625 ± 25 km, an inclination of 88 ± 2 degrees, a two-year licence term beginning after successful placement, and a hard deadline of 9 July 2032 to launch and operate. Disposal is by uncontrolled re-entry. The Commission expressly declined to assess the environmental impact of a larger fleet, because only one prototype was applied for.[14]
Money has followed. On 10 November 2026 the company announced $73.5 million in new funding, taking its total above $100 million, to support initial launches and commercial partnerships in energy and lighting. It has said its inaugural demonstration will fly in the second quarter of 2027 and has bought two dedicated launches from India’s Skyroot Aerospace — an arrangement that would also require Indian regulatory approval through IN-SPACe and could, in principle, let the company operate outside US licensing.[15][16]
What has actually been demonstrated
Not the light, so far. What the company did demonstrate, on the night of 10 October 2026, was a rehearsal: two large spotlights hauled up a mountainside above Lone Pine, California, flooding the Owens Valley with light about twice as bright as a full moon over roughly 3 by 10 miles, so that Inyo County search-and-rescue crews could run the same grid searches with and without it. The rescuers reported that the extra light made moving over rough ground easier and improved peripheral vision, but that for the essential task of finding people the benefit was marginal. It is a more honest result than a press release would have produced, and it cuts against the most sympathetic use case for the technology.[17]
The objections
Astronomers were the first critics in the 1990s and remain so. A moon-bright artificial source moving across the sky is a calibration nightmare for telescopes, and a constellation of thousands would be worse; the concern is not a single demonstration but what it licenses. Ecologists point to the same problem in biological terms — light is a cue for migration, hibernation, flowering and predator-prey behaviour, and the argument that Arctic nights are unproductive is not the same as the argument that they are unimportant.[18][19]
There are two further questions with no settled answer. The first is consent: a satellite that illuminates a region crosses many jurisdictions, and the licensing so far has been national — one FCC grant for one American spacecraft — while the sky it affects is not. The second is debris and traffic, since a fleet of thousands of large thin structures in low orbit would be a new population of objects to track, and a torn reflector is a large area of uncontrolled material.[20]
The other meaning
“Space mirror” also refers to the primary mirrors of space telescopes — the segmented 6.5 m beryllium mirror of the James Webb Space Telescope being the best-known example — and to reflective solar sails used for propulsion, the technology from which Znamya was derived in the first place. Those are different engineering problems with different stakes, and neither raises the questions above; a mirror that looks outward at the universe does not change anyone’s night. The phrase is worth using carefully for that reason: one kind of space mirror is an instrument, and the other is an intervention.[21]
Select a year and month above to explore cold archived discussions.
Please select an archived topic above to view the discussion logs.