Effects of a Nuclear Detonation in Low Earth Orbit

Abstract: More than 13,000 satellites operate in low Earth orbit (LEO), and concern has grown that nuclear weapons could be used in space to target them. Beyond the initial X-ray burst, electrons released by radioactive fission products can become trapped in Earth's magnetic field, repeatedly exposing satellites as an artificial radiation belt. Here we couple prompt radiation, electron trapping and belt evolution to satellite failures, collisions and atmospheric drag. For a 1-Mt detonation at 500km over the South Atlantic, repeated belt exposure disables approximately 98% of modeled satellites within one week in our central case. Tripling the assumed radiation tolerance delays widespread failure by only days: approximately 97% are still disabled within two weeks. The loss of maneuverability sharply increases collision risk, with approximately 20 collisions expected within two years. Drag removes many disabled spacecraft from lower altitudes, but replacement satellites could remain exposed to radiation and debris for years. These findings show how a single detonation could devastate orbital infrastructure and support efforts to prevent the deployment and use of nuclear weapons in space.

• The paper investigates the consequences of a nuclear detonation in LEO, revealing that while prompt radiation is local, delayed exposure to trapped electrons can disable nearly the entire modeled satellite fleet within weeks
• The study identifies a high susceptibility of fleet manueverability to subsequent collision generation as artificial radiation belts form
• The centralized location of a derection of approximately 500 km can optimize volume results with controlled cross-section values.
• The paper concludes that the dominant hazard might be the prompt radiation pulse, and that detailed electron retention time and radiation tolerances make a real impact on orbital system functioning

The paper develops an end-to-end model of the consequences of a nuclear detonation in low Earth orbit (LEO), linking prompt X-ray exposure, fission-product electron production, geomagnetic trapping, artificial radiation-belt evolution, spacecraft failure, collision generation, fragmentation, and atmospheric drag. Its central claim is that the dominant hazard need not be the prompt radiation pulse. Under favorable geomagnetic and orbital conditions, delayed exposure to trapped beta-decay electrons can disable nearly the entire modeled LEO payload population within days to weeks, after which the loss of maneuverability produces a pronounced but transient increase in collision rates. The study is presented as a first-order population assessment rather than a prediction of specific conjunctions or individual collision events (2611.09827).

Study design and reference scenario

The reference calculation assumes a 1-Mt detonation at 500 km altitude over $20^\circ$S, $35^\circ$W, with a 50% fission fraction. This corresponds to approximately $7.25\times10^{25}$ fissions. The modeled radiation population contains 13,306 payloads derived from January 2026 orbital elements, while the coupled collision inventory includes payloads, rocket bodies, pre-existing debris, and other catalogued objects. Spacecraft are assigned 2 mm of aluminium-equivalent shielding, and accumulated electron dose is evaluated against illustrative thresholds of 40 and 120 krad(Si).

The framework combines several reduced-order components. Prompt X-rays are attenuated through the atmosphere and spacecraft shielding. Radioactive fission products are modeled using beta-activity histories generated with ONIX and ENDF/B-VIII.0 nuclear data. The emitted electrons are distributed over magnetic shells obtained from IGRF-14, with trapping determined by whether their drift paths remain above a 120-km atmospheric boundary. The artificial belt is then evolved using radial diffusion and prescribed energy- and shell-dependent lifetimes. Satellite trajectories are propagated with SGP4 at 20-minute intervals for one year, and the resulting dose histories are passed to an altitude-resolved collision and fragmentation model extending ten years.

The source-term uncertainty is represented by three release cases: 0.40, 1.29, and 5.21 emitted electrons per fission. These correspond approximately to radioactive-debris retention over 2 s, 20 s, and six months, respectively. The 20-s case is the central scenario; the six-month case is explicitly treated as an extreme upper bound rather than a realistic expectation. The radiation tolerances are also deliberately stylized: uniform fleet-wide thresholds are used to approximate a population comprising civilian and military spacecraft, although the authors emphasize that these values are not spacecraft-specific specifications.

This architecture is valuable because it preserves causal dependence between subsystems. Radiation failures remove avoidance capability; disabled spacecraft remain collision targets; collisions create fragments; and drag subsequently removes part of the affected population. At the same time, each link introduces assumptions that limit the interpretation of the numerical results, particularly the reduced-order radiation-belt representation and the population-level collision-rate formulation.

Burst location controls belt formation

A major result is that yield and altitude alone do not determine the outcome. Electron trapping depends on the entire drift path around Earth, not merely on the field strength at the detonation point. Electrons that mirror above the atmosphere near injection can later encounter a weaker magnetic field elsewhere along the same drift shell and be lost through the drift loss cone. In the adopted geomagnetic field, some equatorial burst locations produce little persistent trapping, whereas locations near regions of weak magnetic field can populate compact shells that intersect dense LEO constellations.

The paper compares two 1-Mt detonations at approximately 500 km. One location is selected to maximize prompt exposure near a high instantaneous satellite density. The other is selected near a weak-field region to maximize the intensity of an artificial belt intersecting the 500–600 km population. The prompt-optimized case retains approximately 79% of emitted electrons on shells $L=1.74$–2.19 and reaches a peak shell-average dose rate of about 12 rad s$^{-1}$. The belt-optimized case retains only approximately 47% but concentrates the electrons on the more compact shells $L=1.31$–1.51, producing a peak dose rate of about 30 rad s$^{-1}$.

Figure 1: Comparison of prompt-exposure geometry and artificial-belt geometry for two 1-Mt detonations, showing that compact trapped shells can produce higher dose rates despite lower total electron retention.

This contrast establishes an important physical point: total trapping efficiency is not a sufficient predictor of fleet damage. Shell volume and intersection with populated orbital altitudes can dominate the dose rate. The location survey reports peak mapped dose rates varying by more than two orders of magnitude. In particular, bursts near the magnetic equator over portions of the Atlantic and Pacific can populate shells whose mirror points descend below the atmospheric boundary over the South Atlantic Anomaly, suppressing long-lived trapping. Conversely, injection near the modeled reference location produces compact shells overlapping heavily populated LEO altitudes.

Figure 2: Geographic dependence of artificial-belt intensity, including regions where drift-loss-cone effects prevent persistent trapping and regions that produce intense compact belts.

The implication is operationally significant but conditional: a detonation in LEO could produce substantially different consequences depending on latitude, longitude, geomagnetic configuration, and the spatial distribution of the orbital population. The calculation uses IGRF-14 for a single epoch and an assumed isotropic electron source. It therefore resolves the leading-order geometry but does not establish how realistic pitch-angle scattering, time-varying geomagnetic activity, or ion-by-ion debris transport would alter the location map.

Delayed electron exposure dominates fleet losses

For the belt-optimized reference burst, prompt X-rays disable 383 objects, or 2.9% of the modeled 13,306-object fleet. This initial loss is small compared with the subsequent radiation-belt effect. In the central source case and at the 40 krad(Si) threshold, approximately 25% of the fleet is disabled after one day, 78% after three days, and 97.8% after seven days. Half the fleet crosses the failure threshold after approximately 1.9 days, while 90% does so after approximately 4.1 days. By day 15, approximately 99.4% is disabled.

Figure 3: Time evolution of radiation-driven fleet loss, showing the transition from localized prompt failures to near-total fleet disablement through repeated belt passages.

The result is robust in direction across the three electron-source assumptions, although the timing varies substantially. At the 40 krad(Si) threshold, 90% fleet loss occurs after approximately 10.4 days for the lower source, 4.1 days for the central source, and 2.1 days for the upper source. Nevertheless, all three cases disable more than 99% of the modeled fleet within 30 days. Thus, the source uncertainty primarily affects the temporal profile in the 1-Mt cases rather than whether widespread failure occurs.

Increasing the representative tolerance to 120 krad(Si) delays but does not prevent fleet-scale losses. In the central source case, 71.8% of the fleet is disabled by day 7, 97.2% by day 15, and 99.3% by day 30. The lower source produces 90.0% loss by day 30, while the upper source produces 99.8%. This supports the paper’s deliberately strong claim that tripling the assumed radiation tolerance changes the timing by days to weeks but does not avert catastrophic fleet loss under the reference detonation.

The dependence on yield is more substantial. With the central source and 40 krad(Si) tolerance, a 100-kt detonation disables 90% of the fleet in approximately 33 days and 99.4% within one year. At 120 krad(Si), 90% loss occurs after about 93 days and 96.7% after one year. A 10-kt detonation is more sensitive to the tolerance assumption: at 40 krad(Si), 92.2% of the fleet is disabled after one year, compared with 18.6% at 120 krad(Si). The study therefore does not support an unconditional claim that every orbital nuclear detonation would eliminate nearly all LEO spacecraft. Its strongest result applies to the specified yield, location, geomagnetic configuration, electron source, shielding model, and threshold assumptions.

The spatial distribution of the population is central to these numbers. The modeled objects are concentrated near 500–600 km, where the reference shells intersect the fleet. The dose calculation also treats satellites as having 2 mm of aluminium shielding and applies a uniform failure threshold. Real spacecraft vary in shielding topology, component-level hardness, redundancy, operational state, and failure criteria. Consequently, the reported fleet fractions should be interpreted as scenario outputs from a common threshold model, not as a spacecraft-by-spacecraft vulnerability estimate.

Loss of maneuverability creates an early collision surge

Radiation failure is coupled to collision risk by treating disabled spacecraft as unable to maneuver. The collision model calculates expected rates across altitude-resolved populations rather than simulating individual conjunctions. Its global CRASH Clock is the inverse of the instantaneous expected collision rate across all modeled pairs, including spacecraft, debris, rocket bodies, and collision fragments. It is a population-level waiting-time indicator, not an individual satellite lifetime.

In the central 1-Mt case, the clock falls from approximately 250 days immediately after the prompt burst to about four days within two weeks. The corresponding loss of avoidance capability produces a 95% probability of at least one modeled collision within 15 days. Approximately three collisions are expected by that time, accompanied by about 2,400 newly generated fragments at least 10 cm in characteristic length. By day 30, approximately six collisions and 4,400 new fragments are expected.

Figure 4: Coupled radiation, collision, fragmentation, and drag response, with an early collision surge followed by atmospheric removal of lower-altitude targets.

Over two years, the reference detonation produces approximately 20 expected collisions, compared with 12 in the matched no-detonation model. Over ten years, the detonation case produces approximately 26 expected collisions versus 12.4 in the control, with about 75% of the detonation-case events occurring during the first year. Approximately 22 of the ten-year events are catastrophic and four are cratering events; satellite–satellite collisions account for roughly 19 events, while satellite–debris collisions account for about five.

The early surge follows directly from the model’s avoidance prescription. A pair involving an operational spacecraft receives a 99% residual-risk factor, whereas pairs involving disabled spacecraft do not. This is an optimistic treatment for operational systems because it assumes effective avoidance against every modeled partner for which avoidance credit is applied. It also does not resolve detectability, conjunction-warning latency, maneuver authority, or the ability to discriminate intact spacecraft from fragments.

The collision totals are highly sensitive to geometric assumptions. The central object-dependent geometric-cross-section model yields 26.1 collisions over ten years. Alternative solar-flux paths produce approximately 24.5–28.2 collisions, and Starlink drag-property variations produce approximately 17.3–34.8. Smaller fixed encounter areas yield about 16.4 collisions, whereas the larger 10-5-10 prescription yields approximately 88.4. A deliberately sustained low-density atmosphere produces approximately 477 collisions and 97,035 new fragments. These cases are not confidence intervals; they demonstrate that atmospheric removal and collision geometry can dominate the long-term numerical magnitude.

The model’s more stable conclusion is therefore qualitative and temporal: radiation-induced loss of maneuverability produces an early collision-rate increase, followed by a reduction as drag removes lower-altitude targets. The precise number of collisions is much less robust than the existence of this transient phase.

Drag clears lower altitudes but does not restore the environment

Atmospheric drag gradually removes many disabled spacecraft, particularly at lower altitudes. Under central solar forcing, approximately 5,600 intact radiation-disabled payloads remain after one year, 3,200 after two years, and 2,100 after ten years, weighted by orbital residence below 2,000 km. As targets are removed, the modeled global CRASH Clock lengthens to about 470 days after two years and 580 days after a decade.

This apparent late-time improvement is not a recovery of orbital services or safety. It is primarily a consequence of population depletion. The detonation case has fewer collision targets because disabled spacecraft and fragments reenter, whereas operational spacecraft in the control retain their orbits. Between years 2 and 10, the control produces approximately ten additional collisions compared with approximately six in the detonation case, despite the detonation having produced the larger early hazard.

At year 10, the central case retains approximately 3,900 new fragments and 6,200 pre-existing debris objects in the modeled population. Breakup velocity kicks distribute fragments above and below their formation altitudes, and objects at higher altitudes persist longer. This altitude dependence means that drag can reduce risk in lower LEO while leaving a substantial long-lived debris population at higher altitudes.

Figure 5: Altitude distribution of the radiation-modeled satellite population, showing the concentration of objects near the altitudes intersected by the modeled artificial belt.

The paper also examines a hypothetical replacement satellite inserted at different altitudes and times. At lower altitudes, its collision rate can eventually fall below the matched control because drag has cleared disabled spacecraft and debris. At higher altitudes, the rate remains elevated for longer. However, a lower modeled collision rate does not establish that operations can safely resume. The insertion analysis assumes that the replacement spacecraft survives radiation, excludes debris smaller than 10 cm, and does not resolve inclination, conjunction phase, or operational constraints.

The radiation environment creates an additional obstacle to replacement. Historical observations from Starfish Prime indicate that artificial-belt electrons persisted for several years. Therefore, even after atmospheric drag reduces collision activity, replacement spacecraft may remain vulnerable to accumulated electron dose. The paper’s coupled result is consequently asymmetric: drag can eventually remove some collision targets, but it cannot be interpreted as rapid environmental recovery.

Limitations and open questions

The principal uncertainty is the retention time of radioactive debris ions above the atmosphere. The central and lower source cases are informed by Starfish Prime observations of debris arrival in the conjugate atmosphere, while the six-month case is an intentionally extreme upper bound. The model distributes the source across magnetic shells rather than tracking individual ions, their trajectories, and their time-dependent release locations. A more resolved transport calculation could change both the temporal injection profile and the spatial overlap between the belt and satellite orbits.

The radiation-belt model is also reduced order. It uses a radial diffusion operator with fixed $K_p=2$, prescribed electron lifetimes, approximate shell volumes, and an assumed injected energy spectrum. Wave-driven pitch-angle scattering, detailed energy-dependent drift and bounce dynamics, and data assimilation are not explicitly calculated. The geomagnetic field is evaluated for a single date, and the electron source is assumed isotropic. These choices are adequate for a scenario survey but do not provide a validated prediction of belt evolution under an actual detonation.

The spacecraft-failure model applies uniform thresholds and shielding to a heterogeneous population. It does not represent component-level dose margins, shielding anisotropy, electronics architecture, redundancy, temporary upsets, recovery, or spacecraft-specific failure modes. Prompt neutrons and gamma rays are excluded from the coupled failure population. Radiation histories are also fixed after day 60, and dose is not recomputed as spacecraft descend.

The collision model estimates population-level rates from altitude bins and geometric proxies. It assumes isotropic relative velocities, does not resolve inclination or orbital precession, and treats collision avoidance through simplified residual-risk factors. Debris smaller than 10 cm is excluded from collision rates, even though such material can damage spacecraft. The 99% avoidance benefit is optimistic for poorly tracked fragments, and the model does not simulate conjunction-level responses. The ten-year horizon excludes longer-term replenishment, planned launches, controlled disposal, and potential century-scale debris evolution. Numerical tests verify conservation and discretization stability, but they do not independently validate the physical assumptions.

The paper also identifies consequences outside the modeled orbital system. A high-altitude nuclear detonation could generate a broad electromagnetic-pulse footprint, and the reentry of thousands of spacecraft could inject substantial metal vapor and aerosol material into the atmosphere. These effects are noted but not quantified. The open questions that follow most directly from the study are how realistic ion transport and pitch-angle scattering would modify the belt, how spacecraft-specific hardness distributions would alter fleet-loss times, and how a longer-duration orbital evolution would couple persistent fragments, replacement launches, and atmospheric variability.

Conclusion

The paper presents a coupled scenario model in which a nuclear detonation in LEO produces consequences through two distinct pathways. Prompt X-rays disable a relatively small local fraction of spacecraft, while trapped fission-product electrons can repeatedly irradiate satellites on intersecting magnetic shells and disable most of a densely populated LEO fleet. For the reference 1-Mt burst, approximately 98% of the modeled fleet is disabled within one week at the central 40 krad(Si) threshold, and approximately 97% is disabled within two weeks even at 120 krad(Si).

The subsequent loss of maneuverability produces an early collision surge, with approximately 20 expected collisions within two years and 26 over ten years in the central case. Atmospheric drag later removes many lower-altitude targets, but this reduces collision rates through population depletion rather than restoring orbital functionality. The numerical collision totals remain sensitive to geometric and atmospheric assumptions, whereas the principal conclusion—that radiation failure can couple rapidly to collision and debris hazards—is more persistent across the tested scenarios.

The specific unresolved issue is not whether a detonation can affect satellites beyond the prompt X-ray footprint, but how the interaction among debris retention, geomagnetic trapping, spacecraft-specific radiation tolerance, and long-term orbital transport determines the scale and duration of that effect.