Space Mirrors
Frequently Asked Questions

A space mirror (or orbital mirror) is a large reflective surface deployed in orbit around Earth, designed to redirect sunlight toward a specific location on the ground. Unlike a solar panel, which converts sunlight to electricity, a space mirror simply reflects sunlight — delivering it as enhanced ambient light to the target area below.

Modern proposals use thin aluminised Mylar film, which is extremely lightweight and folds compactly for launch. Once in orbit, it unfurls into a large reflective sheet. The mirror is steerable — it can be tilted to point its reflection at a specific ground target.

EARENDIL-1 is the first satellite planned by US startup Reflect Orbital — a demonstration mission for the company's commercial orbital mirror service. It consists of an 18×18 metre Mylar reflective sail deployed in sun-synchronous orbit at 625km altitude, targeting a mid-2026 launch.

Each pass of EARENDIL-1 over a ground target delivers approximately 3.5 minutes of enhanced sunlight to a ~5km footprint. The demonstration mission's primary customers are expected to be solar farms wanting supplemental light after sunset, and entertainment venues wanting dramatic lighting effects.

Yes. Russia launched two orbital mirrors from the Mir space station:

Znamya-2 (1993) — a 20m aluminised Mylar reflector that successfully deployed and traced a ~5km beam of sunlight across nighttime Europe. It was the first and only fully successful orbital mirror in history.

Znamya-2.5 (1999) — a larger 25m follow-on mirror that snagged on an antenna during deployment and failed to fully unfurl. The programme was cancelled when Mir was deorbited in 2001.

China announced a conceptual "artificial moon" satellite in 2018 (the Chengdu proposal), but no hardware was ever built. For the complete history, see our full timeline page.

Reflect Orbital is a US commercial space startup founded to deliver orbital solar illumination as a service. The company has raised $20 million in a Series A round led by Lux Capital and Sequoia Capital. Notable backers include Baiju Bhatt, co-founder of the Robinhood trading platform.

The company has also received a $1.25 million contract from the US Air Force through the Small Business Innovation Research (SBIR) programme, validating military illumination use cases including force protection, remote operations, and search and rescue.

OrbitalSolar.ai is independent and not affiliated with Reflect Orbital or any mirror operator.

At peak brightness (directly overhead with optimal orientation), EARENDIL-1 would appear approximately as bright as Venus at its maximum — roughly apparent magnitude −4 to −5. This is easily visible to the naked eye, conspicuous at dusk, and significantly brighter than any star.

At lower elevations or non-optimal angles, apparent brightness drops considerably. The mirror is designed to be steerable — it can tilt away from Earth when not actively delivering a pass, which would make it much fainter or invisible during non-service periods.

For detailed magnitude comparisons, see our How Bright Are Space Mirrors? page.

EARENDIL-1's design target is approximately 20% of midday solar irradiance over its 5km footprint. Full midday sunlight is roughly 1,000 W/m². At 20%, EARENDIL-1 would deliver around 200 W/m² — meaningful for solar farms but not full illumination equivalent.

The pass lasts approximately 3.5 minutes per orbit. For a solar farm to receive one continuous hour of supplemental light at this intensity, approximately 15 coordinated satellites would be needed. Reflect Orbital's ultimate 4,000-satellite constellation targets this kind of continuous delivery to multiple ground sites simultaneously.

Anyone within a large viewing arc would be able to see the mirror as a bright moving point of light — similar to seeing the ISS or a bright Starlink satellite. The concentrated illumination beam that hits the 5km target footprint is visible only within that zone, but the reflective satellite itself is visible across a much wider area of the night sky.

During a targeted pass, the mirror would be among the brightest objects in the sky for observers over a broad region, lasting the ~3.5 minutes of the pass before fading as it moves out of optimal angle or the operator steers it away.

No — at EARENDIL-1's scale and design, this is not a realistic risk. The mirror is not a focusing concentrator; it produces diffuse ambient light across a 5km area, not a focused beam. 20% of midday sun intensity is comparable to slightly overcast daylight — not dangerous to human eyes under normal circumstances, and not capable of igniting fires.

A very different class of device — an orbiting solar concentrator designed to focus sunlight to a tight spot — would pose different risks. EARENDIL-1 is not that device.

This is one of the more serious technical concerns raised about orbital mirrors. An uncontrolled tumbling large-area reflector sweeps its reflection across Earth in unpredictable arcs. Depending on geometry, the worst-case flare brightness could reach magnitude −10 to −12 — comparable to or brighter than the full Moon — lasting several seconds per rotation.

NASA's Advanced Composite Solar Sail System (ACS3) began spinning uncontrollably in 2024, demonstrating that this failure mode is real. A mirror at 625km altitude without active deorbit capability could remain in orbit for months to years if attitude control is lost.

Reflect Orbital has described attitude control as a core challenge but detailed public failure mode analysis has not been published.

Artificial light at night (ALAN) is a documented ecological stressor. It disrupts insect navigation and reproduction, affects bird migration, alters predator-prey dynamics, and interferes with plant flowering cycles that respond to day length. Nocturnal animals particularly depend on dark night conditions.

A single EARENDIL-1 pass over a 5km footprint for 3.5 minutes represents a brief, targeted addition to the light environment — its ecological impact is likely minimal at this scale. A 4,000-satellite constellation delivering near-continuous illumination over large regions would be a substantially different ecological intervention, and no environmental impact assessment in the public domain addresses this at scale.

Professional astronomers are nearly unanimous in opposing large orbital mirror constellations. Their objections centre on three issues:

Sky background brightness: Even faint satellite flares degrade observatory sensitivity by adding to the sky background. An orbital mirror optimised for ground-level brightness is far more damaging than a typical satellite.

Governance and precedent: Unlike Starlink (which astronomers also object to), orbital mirrors are explicitly designed to be bright from the ground. Allowing this sets a precedent that the night sky can be commercially exploited for brightness delivery.

Irreversibility at scale: A constellation in orbit cannot be recalled once deployed. If a 4,000-satellite constellation degrades dark sky quality globally, that damage is essentially permanent for decades.

For a comprehensive analysis, see our Controversy Explained page.

EARENDIL-1 alone is unlikely to cause material damage to professional observatories. A single satellite passing through a field of view for under 4 minutes represents a manageable, predictable event — observatories can schedule around known passes, as they already do for ISS and bright Starlink satellites.

The concern is precedent, not immediate damage. Astronomers worry that approving EARENDIL-1 without any regulatory framework opens the door to constellations that would cause material harm.

Currently, no body specifically regulates orbital mirrors. In the United States, commercial satellites are regulated by the FCC (for communications/spectrum) and the FAA (for launch). Neither body has a mandate to evaluate the astronomical, ecological, or light pollution impact of a reflective payload.

The Outer Space Treaty (1967) governs state actors and requires that space activities be conducted "for the benefit of all countries," but provides no mechanism for objection or veto over commercial missions that alter the shared night sky.

The International Astronomical Union (IAU) has called for moratoriums and environmental impact assessments, but has no regulatory authority.

For a US-based operator, the FCC could theoretically revoke an experimental licence, and the US government could order a company to deorbit satellites. However, this has never been done, and the legal pathway is untested.

If satellites are in orbit with functional deorbit capability, compliance is technically achievable. If a satellite fails or lacks deorbit capability, enforcement becomes a physical rather than legal problem — there is no mechanism to retrieve or disable a satellite against an operator's will.

For non-US operators (such as China), US regulatory authority is entirely absent.

The economics are challenging at current projected service prices. A solar farm receiving 4 passes per day of 3.5 minutes each at 20% solar intensity gains roughly 14 minutes of enhanced generation — valuable at the margins but small relative to a full day's production.

At a service price of $5,000/mirror-hour, a 50MW solar farm would typically spend more on the service than it earns in additional electricity revenue. The break-even point requires either very large farms, high electricity prices, or significantly lower service prices than current projections suggest.

Use our interactive Economics Estimator to model specific scenarios.

The economics are more favourable for applications where the value of light is not measured in kilowatt-hours:

Entertainment events: A stadium, festival, or outdoor concert operator might pay a substantial premium for a dramatic, memorable light effect. The per-pass cost amortises differently against entertainment budgets than energy budgets.

Military and emergency response: The US Air Force SBIR contract validates this. In tactical scenarios, the ability to illuminate a specific 5km area on demand — without aircraft, drones, or ground infrastructure — has value that cannot be priced in $/kWh terms.

Search and rescue: Illuminating a remote crash site or disaster zone for rescue teams at night represents a similar high-value, low-frequency use case.

Once EARENDIL-1 launches and its TLE (Two-Line Element) orbital data is published, it will be trackable via any satellite tracking service that uses TLE data — including OrbitalNodes.ai, which is building a dedicated solar mirror pass mode for exactly this purpose.

OrbitalNodes.ai already tracks all catalogued satellites in real time in your browser without any install required. A Solar Mirrors toggle is in development and will go live at EARENDIL-1's launch.

Visit OrbitalNodes.ai to track satellites now, and check back here for mirror-specific pass predictions once EARENDIL-1 is in orbit.

EARENDIL-1 is targeting a mid-2026 launch. If it launches on schedule, it would be the first visible orbital mirror since Znamya-2 (1993) and Znamya-2.5 (1999).

There is currently no space mirror in orbit. The live constellation counter on this site (and at OrbitalSolar.ai) will update to reflect this the moment that changes.

The Znamya programme was a Soviet / Russian space mirror research initiative conducted from the Mir space station in the 1990s. It produced two missions:

Znamya-2 (1993) — a 20m reflector successfully deployed from Progress M-15 that traced a 5km beam of sunlight across Europe. The only fully successful orbital mirror ever deployed.

Znamya-2.5 (1999) — a 25m follow-on that failed to deploy when the reflector snagged on an external antenna. The programme ended when Mir was deorbited in 2001.

For the full documented history, see our Complete Timeline.

No. In 2018, Chengdu city government announced plans for an "artificial moon" satellite that would illuminate the city with 8 times the brightness of the full Moon. The proposal received extensive global media coverage but no satellite hardware was ever confirmed, funded, or built. Chinese aerospace experts publicly questioned its technical and economic feasibility. The proposal appears inactive as of 2024.