The Night Is Getting Cancelled: What Space Mirrors Would Do to Our Skies

The Science

In August 2026, a startup called ReflectOrbital received Federal Communications Commission approval to launch a small pathfinder satellite called EARENDIL-1: an 18×18 meter reflective mirror positioned roughly 600 kilometers above Earth. The satellite would direct sunlight down onto a 2.5-kilometer radius circle on the ground—enough to illuminate a small town. But EARENDIL-1 is only a test. ReflectOrbital's actual ambition is a constellation of 50,000 far larger mirrors, each stretching 54 meters across, all working in concert to flood entire regions with reflected sunlight.

No one had previously calculated what this would do to the night sky. The FCC approved the launch without examining environmental impacts. So three researchers—Miroslav Kocifaj, Gáspár Bakos, and František Kundracik, affiliated with institutions in Slovakia and Hungary—decided to do the math themselves.

They used a sophisticated 3D radiative transfer model called the Modified Successive Orders of Scattering (MSOS) algorithm, originally developed for studying light pollution from ground-based sources. Adapting it to an incoming beam from space required treating the atmosphere as a stratified medium—air and aerosol densities decrease exponentially with altitude—and tracking how photons scatter through this layered system. They ran simulations across nine scattering orders, testing different aerosol conditions (fine-mode urban haze versus coarse maritime dust), three ground albedos (zero reflection, typical Earth, and fresh snow), and varying observer distances from the beam center. All calculations used 550 nanometers—the wavelength where both sunlight and human vision are most sensitive.

What They Found

The results are striking. A single 54-meter mirror in orbit, shining on a cloudless night, would appear as a magnitude -16.7 point of light—about four magnitudes brighter than the full moon. (Astronomers measure brightness on an inverse scale: lower (more negative) numbers are brighter. Venus at its peak reaches -4.7; the full moon sits at -12.7.) At this brightness, the mirror would be easily visible in daylight.

But the direct light is only part of the story. The real problem is what happens when that beam hits the atmosphere.

Sky luminance at different distances from beam center

Label	Value
Full Moon sky	0.02
5.4 km (α=0)	0.03
5.4 km (α=0.2)	0.06
5.4 km (α=0.8)	0.15
14.1 km (α=0.2)	0.015
34.1 km (α=0.2)	0.006

Values approximate based on Figure 2 and comparison to full moon. α=0 is absorbing surface, α=0.2 is typical Earth, α=0.8 is snow.

For an observer standing inside the illuminated patch, the diffuse sky glow—light scattered by air molecules and aerosols bouncing photons in every direction—would rival the twilight sky just after sunset. "Too bright to see even the brightest stars," the authors note. The sky itself would become a source of light pollution, not just the mirror. On snow-covered ground, the effect amplifies further: reflected sunlight bouncing off the white surface re-enters the atmosphere as a secondary source, creating a luminous veil that persists even when looking away from the beam.

The pollution extends far beyond the lit circle.

Distance at which mirror effect exceeds full moon sky brightness

Label	Value
1 km (inside beam)	0.1
5.4 km	0.06
14.1 km	0.02
34.1 km	0.008
80 km (400 mirrors)	0.0015

Distances at which the mirror sky brightness remains comparable to or exceeds full moon illumination. The 80 km figure assumes 400 simultaneous mirrors.

At just 5.4 kilometers from the beam center—roughly 3 kilometers outside the illuminated area—the sky luminance near the horizon can reach approximately 0.15 candelas per square meter for snow-covered terrain. This is an order of magnitude brighter than the moonlit sky. Move to 14 kilometers out, and the glow still exceeds the luminance of the full moon for most of the sky. Even at 34 kilometers—more than 13 times the beam's radius—the sky remains brighter than a typical moonlit night in the direction of the beam.

Effect of ground albedo on sky brightness

Label	Value
α = 0.0 (absorbing)	1
α = 0.2 (typical)	3.5
α = 0.8 (snow)	10

Luminance multiplier compared to absorbing surface at 5.4 km distance, aerosol asymmetry parameter g=0.6.

The contrast between surface types matters enormously. Over fresh snow (albedo 0.8), the sky can be 5 to 10 times brighter than over typical terrain (albedo 0.2), because reflected light becomes a secondary illumination source. A winter city beneath one of these mirrors would face vastly worse light pollution than a summer desert.

And these numbers are for a single satellite. ReflectOrbital's plans envision thousands operating in concert. If 400 mirrors illuminated the same patch simultaneously—which the company would likely do for practical illumination purposes—the glow would be obvious from 80 kilometers away. The combined brightness would reach roughly 2,400 lux: about 10,000 times brighter than full moonlight, and approaching 2.4% of actual sunlight. At that point, you're not illuminating the night—you've recreated daytime in a 5-kilometer radius.

Why This Changes Things

Space mirrors occupy a legal gray zone. The 1967 Outer Space Treaty requires nations to authorize and supervise space activities, but environmental review isn't mandatory for satellite approvals in most jurisdictions. The FCC approved ReflectOrbital's pilot without assessing atmospheric impacts—an omission this paper makes impossible to ignore.

The ecological consequences extend well beyond astronomy. Nocturnal ecosystems have evolved around darkness. Migrating birds navigate by stars; insects orient by polarized moonlight; predator-prey dynamics shift when shadows disappear. A constellation of 50,000 mirrors creating perpetual artificial twilight across swaths of Earth could disrupt behaviors that have been stable for millions of years. Sea turtle hatchlings, which orient toward the brightest horizon (traditionally the ocean under a moonlit sky), would instead head toward an illuminated city. The effect would be concentrated in the beam, but the scattered glow reaches much further.

For astronomers, the implications are starker. Even the first prototype—much smaller than the planned constellation—would saturate ground-based telescopes within its footprint. The scattered light would compromise observations across regions far larger than the directly illuminated area. Professional observatories in New Mexico, Arizona, and Chile could find their skies compromised if large constellations become operational.

The authors note that these calculations assume cloud-free conditions. In reality, roughly 67% of Earth is cloud-covered at any given time. Clouds illuminated from above would behave as bright diffuse sources, potentially making the effect worse—or at least far less predictable. The paper explicitly flags this as a limitation; a full environmental assessment would need to model cloud scattering via Mie scattering theory, a more complex calculation the authors leave for future work.

There's also the question of orbital mechanics. At 600 kilometers altitude, these mirrors would orbit Earth roughly every 90 minutes. The beam would sweep across the surface at about 7.6 kilometers per second—fast enough that any single location would experience brief illumination rather than permanent daylight. But at the timescales relevant to ecological disruption (nightly migration patterns, pollination cycles, predator-prey encounters), even transient pollution could matter. An animal that pauses behavior during a 10-second flash of artificial twilight is still disrupted.

What's Next

This paper raises questions that the approval process failed to ask. The authors have done the community a service by publishing detailed, peer-reviewable calculations—but their work is only the beginning.

Several gaps remain. The cloud scenario needs modeling: how do thin cirrus versus cumulus decks change the light distribution? What happens when multiple beams overlap, as they would in a dense constellation? How do different wavelengths scatter differently—the paper uses 550 nanometers, but sunlight contains a spectrum that would behave differently? And crucially: what are the ecological thresholds? At what illumination levels do behavioral disruptions become population-level extinctions?

There's also the question of regulation. The FCC's approval came without environmental review; other nations have no framework whatsoever for assessing the atmospheric impacts of reflective satellite constellations. This paper could serve as a technical foundation for future rulemaking—if policymakers choose to use it.

The irony is that ReflectOrbital frames its mirrors as an environmental solution: cheaper than solar panels, providing light without carbon emissions. But a technology that transforms the darkness of night into an extension of daytime carries ecological costs that carbon accounting doesn't capture. The night sky is not unused real estate. It is an ecosystem.

Before 50,000 mirrors go up, someone needs to count what we'd be losing.