Preprint Models Hypothetical 1‑Megaton Nuclear Blast in Space That Could Disable Nearly All LEO Satellites in a Week

A new arXiv preprint models a hypothetical 1-megaton nuclear detonation in low Earth orbit and concludes it could disable about 98% of the satellites in its simulated LEO fleet within a week. The study, posted as a preprint and not yet peer reviewed, does not describe a real event. It is a scenario analysis of what might happen if a nuclear weapon were detonated in space.

The paper, titled “Effects of a Nuclear Detonation in Low Earth Orbit,” was posted to arXiv as arXiv:2611.09827v1 after being submitted on Nov. 6, 2026. Its authors are Zachary Petersen, Hong Qin, Sarah Thiele and Sébastien Philippe, with affiliations listed as the University of Wisconsin-Madison and Princeton University.

The study’s central scenario examines a single 1-megaton detonation at an altitude of 500 kilometers over the South Atlantic, near the South Atlantic Anomaly, a region where Earth’s inner radiation belt dips unusually close to the planet. In plain terms, the paper argues that the main danger would not be the initial flash alone. Prompt radiation from the blast would damage some spacecraft immediately, but the much larger modeled effect comes from an artificial radiation belt: trapped high-energy electrons that would circle Earth and repeatedly irradiate satellites passing through them.

That distinction matters to the paper’s headline result. The authors write in the abstract that “repeated belt exposure disables approximately 98% of modeled satellites within one week in our central case.” By contrast, in this South Atlantic scenario, the initial X-ray burst alone disables 383 objects at t = 0, or about 2.9% of the satellites in the model.

For its orbital baseline, the study uses cataloged January 2026 data for 13,306 payloads. The paper notes that there are now more than 13,000 operating satellites in low Earth orbit, the region of space closest to Earth and home to most communications and Earth-observing spacecraft. Under the authors’ central assumptions, about 98% of the modeled satellites are disabled within one week. Even when the model triples the assumed radiation tolerance of spacecraft, it still shows about 97% disabled within two weeks.

The paper then traces what those failures could mean for the rest of the orbital environment. With many satellites disabled and unable to maneuver, collision risk rises sharply in the simulation. The study estimates about 20 collisions within two years and about 26 over a decade, compared with about 12 over a decade in a matched no-detonation control. Atmospheric drag would eventually pull many disabled satellites down from lower altitudes, the paper says, but replacement satellites could remain exposed to radiation and debris for years.

This was a hypothetical scenario analysis, not evidence that any nuclear detonation has occurred in space.

It is also important to read the numbers as model outputs, not settled forecasts. The paper is a numerical analysis, not an observation of a real-world detonation, and it presents sensitivity cases rather than confidence intervals. Its results depend on assumptions including satellite radiation tolerance, shielding, how many electrons are released into orbit, how collision risk is calculated and the solar and drag conditions used in the simulation.

Still, the mechanism behind the paper is not purely theoretical. The authors cite Starfish Prime, the 1962 U.S. high-altitude nuclear test, which created an artificial radiation belt and damaged satellites. What is new here is the attempt to estimate how a similar mechanism might play out in today’s far more crowded low Earth orbit, with thousands of commercial and government spacecraft sharing the same region.

Two longstanding treaties are also relevant. The 1963 Partial Test Ban Treaty prohibits nuclear explosions in outer space, and the 1967 Outer Space Treaty prohibits placing nuclear weapons or other weapons of mass destruction in orbit.

The paper argues that those rules matter even more in today’s orbital environment than they did in the early space age. In 1962, relatively few satellites were in orbit. Now, with more than 13,000 operating satellites in LEO, the study suggests a single detonation could do more than destroy individual spacecraft. It could, under the paper’s assumptions, cripple a large share of the infrastructure modern economies increasingly rely on, while leaving behind radiation and debris hazards that persist well beyond the initial blast.