A Megaton Burst Can Wreck an Asteroid Without a Hit

At 68 milliseconds, a 1-megaton burst 25 metres off a 160-metre asteroid had fully damaged 92.6 percent of a Chelyabinsk-like body, against 78.1 percent from 10 metres.

The warhead never had to hit. X-rays from the miss spread over more of the surface even though less energy arrived. Astrophysicist Isaiah Santistevan and colleagues at Lawrence Livermore National Laboratory ran the cases in 3D and, on 14 August 2026, published them in The Planetary Science Journal.

A 1-Megaton Burst Does Not Need to Strike the Rock

The team used Spheral, a lab hydrodynamics code written by co-author J. Michael Owen, to park a 1-megaton nuclear explosive device above a homogeneous forsterite asteroid 160 metres across. The body took its outline and porous structure from Bennu, the near-Earth asteroid NASA’s OSIRIS-REx spacecraft sampled in 2020. Real Bennu is a third of a mile wide at its equator. The simulated rock is a scaled copy of that diamond-top shape, not Bennu itself.

In vacuum there is no air to carry a Hollywood blast wave. About 70 to 80 percent of a nuclear burst’s energy can come out as X-rays, a figure the paper takes from Glasstone and Dolan’s 1977 effects handbook. Those X-rays soak a thin skin of rock, boil it, and throw it off. The departing gas shoves the rest of the body the other way. The same dump of energy drives a shock into the interior that can crack it.

Height of burst, the gap between the device and the surface, sets how wide that X-ray footprint is. Yield sets how hard it hits. Together they are the two knobs a last-resort mission would actually have. The Livermore 3D nuclear disruption simulations sit near the fuzzy line the field uses for a clean breakup, not in the easy, overkill regime.

A 1D check against older mesh models agreed to within 0.8 percent across fluence, pulse length, composition, and porosity. Full 3D was coarser, and drifted by about 18.1 percent from earlier cylindrical runs, because a whole asteroid eats far more computer time than a slice.

The Farther Shot Damaged More of the Surface

Two Chelyabinsk-based runs kept the rock and the 1-megaton yield fixed and moved only the burst. At 10 metres, 55.4 kilotons of energy made it into the body. At 25 metres, that fell to 30.0 kilotons. The farther shot still lit a larger share of the face, the way a torch thrown back paints a wider, dimmer circle.

At the shared 68-millisecond mark, the 25-metre Chelyabinsk case had more fully damaged material than the 10-metre case. Full damage here means the damage tensor had already gone to 1. It is a crack map, not a promise that every chip has left the scene.

By 145 milliseconds, the 10-metre Chelyabinsk run had fully damaged 98.2 percent of the mass, and about 97 percent of that mass was moving faster than the asteroid’s own escape speed. Large volumes on opposite sides were heading opposite ways, which is what a body looks like when it is being pulled apart. The 25-metre Chelyabinsk run also posted a slightly higher speed change relative to escape, 8.46 times against 8.37, so the extra standoff was not a weaker shove.

Zooming the device in as close as possible is the instinct from every asteroid movie. These snapshots say that instinct can be wrong for disruption. A miss of tens of metres is not a fumble. It is the aiming method.

Why X-Rays Matter in a Vacuum

Landing a warhead on a tumbling, poorly mapped rock, then holding it there, is the hard engineering problem the films skip. A standoff burst sidesteps the landing. The useful work is radiation on a surface, then rock acting like a short-lived rocket.

THE X-RAY STANDOFF SEQUENCE

• The burst: The device detonates metres off the face, far from Earth so debris and radiation from the device itself are not a local hazard.
• The soak: X-rays dump energy into a thin skin of rock, ice, or metal, and that skin vaporizes.
• The shove: Escaping gas that beats local gravity leaves, and the leftover body recoils the other way.
• The crack: The same energy launches a shock that can damage the interior; if enough mass then moves faster than escape and in opposed directions, the object can come apart.

Lab machines are starting to put numbers on that soak. The paper points to X-ray pulse shots on small quartz and rock targets at Sandia’s Z machine and at OMEGA, plus work at the National Ignition Facility. Those coupons are centimetres, not city-scale, but they are the first real check on the coupling the codes assume.

The field still splits two jobs that sound alike. Deflection keeps one body and adds a small speed change. Disruption breaks it so no leftover chunk is still a hazard, with fragments around 10 metres still treated as dangerous. Unintentional breakup is often flagged when the shove exceeds about 10 percent of escape speed. A common rule of thumb for a thorough disruption is about 10 times escape. Santistevan’s Chelyabinsk cases landed at 8.37 and 8.46 times escape, under that 10-times line and well above the 10 percent fragment warning. That is why the team calls the outcome likely, not settled.

Chelyabinsk and Aba Panu Set the Fracture Rules

The 160-metre body is in a different class from the rocks that supplied its crack model. The paper’s Chelyabinsk meteoroid was about 18 metres across when it burst over Russia in 2013 and blew out windows across a wide region. Tunguska in 1908, a 50 to 60 metre object, flattened more than 2,000 square kilometres of trees. A 160-metre stony body is a city-scale strike if it reaches the ground intact. The fracture settings still come from real meteorites, not invented stone.

Chelyabinsk supplied one Weibull pair, m of 2.8 and k of 1.2 × 10¹² per cubic centimetre. Aba Panu, which fell in Nigeria in 2018, supplied a much tougher pair, m of 5.93 and k of 4.2 × 10²⁵. The third run kept the 25-metre burst and swapped in Aba Panu. Damage was slower and thinner. At 34 milliseconds, 34.2 percent of that body was fully damaged, against 60.2 percent in the Chelyabinsk 25-metre case and 45.0 percent in the Chelyabinsk 10-metre case at the same clock.

THE THREE SPHERAL RUNS

The three clocks are not equal. Aba Panu’s 34.2 percent is an earlier snapshot, not a fair score against 98.2 percent at 145 milliseconds. The filled Bennu-shape model at this size and a bulk density of 2.228 grams per cubic centimetre masses about 3.73 × 10⁹ kilograms. The 10-metre and 25-metre meshes came in at 3.65 × 10⁹ and 3.56 × 10⁹ kilograms. Specific energy in the 10-metre and 25-metre Chelyabinsk runs sat 50 to 100 times above typical porous S-type disruption thresholds, which on paper says half the mass should not pile back up. The team still would not call that a finished fragment catalogue.

Based on the extent of damage, the directionality of the material motion, and the velocity change that we impart on these asteroids, we suggest that disruption is highly likely for two of the three scenarios.

Isaiah B. Santistevan and colleagues, Lawrence Livermore National Laboratory, The Planetary Science Journal

The holdout is Aba Panu. A tougher flaw distribution is enough to push a 25-metre, 1-megaton shot down to 6.74 times escape and a thinner damage map. Composition is not a detail you can guess from Earth and then ignore.

DART Already Handles the Slow Threats

Nuclear disruption is the option you reach for when a kinetic hit is too weak or too late. NASA already flew the gentler method. On 26 September 2022, the Double Asteroid Redirection Test spacecraft struck Dimorphos, the small moon of asteroid Didymos, and altered Dimorphos’s orbit by 33 minutes, plus or minus one minute.

Nicola Fox, then associate administrator for NASA’s Science Mission Directorate, said she cheered when DART slammed head on into the asteroid, and that the impact was only the start. Later work found Dimorphos slowed by about 2.7 millimetres per second along its orbit. Ejecta recoil multiplied the spacecraft’s own shove by a factor of 2.2 to 4.9, depending on the moonlet’s mass. Targeting worked without a prior scout flight. The papers still said a real defence shot wants years of warning, and preferably decades.

Santistevan’s introduction is blunt about the gap DART leaves. One heavy spacecraft, or a flock of smaller ones, may be impractical against a truly massive object. The paper also flags 2024 YR4 as the sort of late discovery where breaking the body, rather than nudging it, may be the only practical mission. Kinetic impact remains the first tool when the clock is long. The Livermore runs are for the clock that is not.

Supercomputers Stop at a Tenth of a Second

The longest Spheral case, the 10-metre Chelyabinsk run, covered 145 milliseconds of physical time and took 59 days on 1,680 processors. The 25-metre Chelyabinsk case used 36 one-day jobs. Aba Panu used 23. That is why the paper cannot say what the pile of rock does an hour later, let alone two months later.

Pieces might spread and become harmless. Some chunks might stay large enough to threaten Earth. Gravity might pull debris back into a weaker rubble pile. Those three endings are all still live at 145 milliseconds. The scatter of large fragments is the objection that keeps coming back whenever a nuke in space is proposed, and it is also the thing this study is not yet long enough to close. Readers chasing whether the rubble stays gone are asking the question the 59-day run could not finish.

THE CLOCKS BEHIND THE LAST-RESORT CASE

1. 15 February 2013: An 18-metre meteoroid bursts over Chelyabinsk and later becomes one of two fracture models in the 2026 runs.
2. 5 October 2021: Livermore describes Patrick King’s late-time study of a 1-megaton standoff a few metres off a 100-metre Bennu-shaped body.
3. 26 September 2022: DART hits Dimorphos and proves a kinetic impactor can change an asteroid’s motion.
4. 14 August 2026: Santistevan’s team publishes the 160-metre Spheral disruption cases, received 22 January 2026.

King’s earlier work is the longer look these new runs cannot afford. He took a 1-megaton device a few metres off a 100-metre Bennu-shaped asteroid, one-fifth the scale of Bennu, then followed the fragment cloud around the Sun. For all five orbits he tried, a breakup just two months before impact cut the hitting mass by a factor of 1,000 or more, so 99.9 percent of the mass misses Earth. On a larger body the spread would be weaker, but even speeds reduced by a factor of ten still put 99 percent of the mass wide if the shot went off at least six months out.

King, then a Livermore graduate fellow and now a physicist at the Johns Hopkins Applied Physics Laboratory, said nuclear disruption is a very effective defence of last resort, and that when warning time runs to decades, kinetic impactors are the preferred tool. Owen, who wrote Spheral, said that if a hazardous object were spotted too late to divert, the remaining option would be to break it so thoroughly the fragments would largely miss Earth. He also said that cloud stretches into a curved stream, and that how fast it spreads, plus how long until it crosses Earth’s path, decides how many pieces still hit.

Megan Bruck Syal of the Livermore planetary defence group tied that 2021 study to a White House Office of Science and Technology Policy goal on near-Earth object modelling. The 2026 paper’s last job, as the authors wrote, is to inform whether nuclear options belong in that toolkit and to support emergency response planning. It does not prove a 1-megaton miss would save a city. It does show that the miss can be the method, that 25 metres can outperform 10 metres on the damage map, and that a tougher meteorite can take the same shot and lag. The rubble’s next hour is still uncomputed.