LEARN · EXPLORE · UNDERSTAND

Methods, sources and model limitations

Published by Nuclear Effects Explorer · Updated 31 October 2026

Model v3.0 is an educational reference, not a prediction of a real incident.

The model supports yields from 1 to 10,000 kt in a standard sea-level atmosphere. It separates pressure, fireball and a simplified thermal-energy estimate from optional idealised surface fallout. Real outcomes may differ substantially.

Surface-burst pressure

The surface-burst ground-range fit is documented in the Nuclear Bomb Effects Computer (CEX-62.2), equation 2.5. With W in kilotons, pressure P in psi and x = log₁₀(P):

r(km) = 1.609344 × W^1/3 × 10^{(−0.1307982 − 0.6836211x + 0.1091296x² − 0.0167348x³)}

The interface shows 1, 5 and 20 psi contours. Cube-root scaling assumes comparable scaled conditions; it does not resolve terrain, buildings or individual structural response. The documented reference coefficients can be inspected in Alex Wellerstein’s effects reference.

Fixed-reference airburst

The airburst option uses the reflected-airblast fit by H. L. Brode, Airblast From Nuclear Bursts: Analytic Approximations (1986), adapted from GOFAI/glasstone. The upstream work is used under its MIT licence.

The height is fixed by a scaled reference: 508 m × (W / 12.5 kt)^1/3, using the example in Glasstone & Dolan §2.24. Pressure contours are found from the outer threshold crossing. There is no burst-height optimisation. The model assumes an idealised surface and sea-level atmosphere.

Fireball and thermal-energy contour

Maximum fireball radius is estimated as 0.0792449 × W^0.4 km for a surface burst or 0.0609611 × W^0.4 km for the airburst reference. In airburst mode, the ring is the maximum horizontal envelope aloft projected onto the map; it is not ground contact.

The thermal-energy contour uses an illustrative point-source budget with 18% effective thermal partition for a contact surface burst, 35% for an airburst, 70% fixed transmission and a 10 cal/cm² threshold. Airburst ground range is projected from slant range. The surface partition is discussed in Glasstone & Dolan §7.101.

This is not a burn or ignition model. Atmospheric path variation, pulse duration, obstructions, cloud, curvature and local exposure are not resolved.

Idealised surface fallout and wind

The optional surface-fallout layer uses Table 9.93 and §§9.94, 9.97–9.100. Table dimensions scale with total yield. The selected fission fraction rescales H+1 reference-rate labels. Downwind length uses the approximate effective-wind correction, limited to 8–45 mph; width and upwind distance are held fixed under the source’s approximation.

Each smooth envelope is reconstructed from downwind length, downwind width and ground-zero width. Its width peaks at 35% of downwind length as a drawing convention; that location is not a tabulated physical prediction. Low-yield empirical width relationships may not be mutually exact. H+1 rates describe idealised smooth, unshielded ground, not dose inside a building.

What is not calculated

No real weather, rainout, wind-shear field, particle-transport forecast, actual fallout arrival time, prompt ionising radiation, EMP, fire spread, population exposure, casualties, evacuation routes or safety zones are calculated. A real event needs information and assessment outside this tool.

Yield references and uncertainty

Sourced US warhead examples cite Federation of American Scientists, United States nuclear weapons, 2025, table 1. They are dated public estimates, not current loadout claims. The additional international list is explicitly user-supplied and unverified. Missile names, warhead variants and historical test yields are not interchangeable.

Checks and validation limits

Implementation checks cover surface-model regression, numerical inversion of airburst pressure thresholds, cube-root scaling, reference height, fallout-table dimensions, wind factors, fission-rate scaling and supported input limits. These checks do not amount to independent validation against real-world events. Rounded figures should not be read as a claim of corresponding real-world precision.

Maps, place names and privacy

Distances use a spherical Earth with radius 6,371 km. Map tiles are supplied by Esri; rendering uses Leaflet. The bundled place directory is from Natural Earth’s public-domain populated-places data, accessed 30 October 2026. Country and territory labels follow source conventions and are not a forecast of risk.

Online searches are sent to Photon/OpenStreetMap, and map requests go to the tile provider. Google Analytics 4 is used for website usage analytics. Saved views and preparedness checklists are stored in the user’s browser. Scenario links can include the chosen location name and coordinates, so review a link before sharing it.

Primary reading

Return to the simulator or browse the blast and fallout guides.