Kaboomly / Methods

How It Works

This page explains how the Kaboomly nuke simulator turns a yield and a burst type into the rings on the map: which sources it uses, how each effect is calculated, how the results are checked, and where the method stops being reliable.

Overview

The simulator estimates six distances for one explosion: the maximum fireball radius, the 20, 5, and 1 psi blast overpressure radii, the distance for a 50% chance of third-degree burns on bare skin, and the distance at which the initial radiation dose is 500 rem. Everything is calculated in your browser. Each effect uses its own relationship from the literature, because blast, heat, and radiation do not scale with yield in the same way.

Method for each effect

Fireball

Glasstone and Dolan give the maximum fireball of a 1-megaton explosion as about 5,700 feet across (§2.05) and state that fireball radius scales with about the 0.4 power of the yield (§2.127). The simulator uses that rule. For a surface burst, where the fireball meets the ground, it follows the book's guidance to treat the explosion as having twice the yield.

Blast overpressure

Blast uses an analytic curve fit by H. L. Brode (1986) for peak overpressure on the ground as a function of distance and burst height, ported from the open-source glasstone library. Distances scale with the cube root of the yield, as described in Glasstone and Dolan chapter III. The simulator searches for the distance at which the overpressure falls to 20, 5, and 1 psi. A sea-level atmosphere and flat, open ground are assumed.

Thermal radiation (burns)

The heat model reads the curves in Glasstone and Dolan Fig. 7.42, which give the distance at which the ground receives a given amount of heat for air bursts with 12 miles of visibility. The amount of heat needed for a 50% chance of third-degree burns comes from Fig. 12.65; it rises from about 6 calories per square centimeter at 1 kiloton to about 11 at 10 megatons (our reading of the figure), because a larger explosion delivers its heat over a longer pulse. For surface bursts the model uses a thermal fraction of 0.18 instead of the air-burst 0.35 (§7.101). The distance along the line of sight is converted to a distance along the ground using the burst height.

Initial nuclear radiation

Initial radiation is the radiation released within the first minute. The model adds three parts: neutrons, secondary gamma rays, and gamma rays from fission products, using digitized curves from Glasstone and Dolan chapter VIII (Figs. 8.123, 8.127, 8.130, and 8.132) provided by the glasstone library. Yields up to 100 kilotons use the fission weapon curves; larger yields use the thermonuclear curves. Neutron and gamma doses are counted equally, which is a simplification. As with heat, the line-of-sight distance is converted to a ground distance.

The fixed airburst height

The simulator has exactly two burst options. A surface burst happens at ground level. An airburst always uses one example from the literature: Hiroshima, where Glasstone and Dolan give a burst height of 1,670 feet for a yield of about 12.5 kilotons (§2.24).

To apply that one example to other yields, the height is scaled with the cube root of the yield, the same scaling the book uses for burst height in its blast charts. That works out to about 720 feet times the cube root of the yield in kilotons, so the geometry looks the same as the historical example at every size. For example:

• Little Boy (15 kt): about 0.54 km (0.34 mi) above the ground.
• B83 (1.2 Mt): about 2.3 km (1.4 mi) above the ground.
• Tsar Bomba (50 Mt): about 8.1 km (5.0 mi) above the ground.

The height is not chosen to increase any effect, and the simulator does not offer a height setting or any "optimum height" calculation. This is a deliberate boundary for an educational tool. One visible result is that for the largest yields the burst is so high that the 500 rem initial radiation dose does not reach the ground at all, so that ring has no size.

Fallout: an idealized illustration

Fallout is shown only for a surface burst. Glasstone and Dolan explain that after an airburst the weapon residues rise so high within about a minute that their radiation no longer reaches the ground in significant amounts (§9.04), so the airburst option has no fallout pattern.

For a surface burst the simulator uses the book's own idealized fallout contours (chapter IX, "Fallout Predictions for Land Surface Bursts"). Table 9.93 gives, for a burst on the ground of any yield, the size of eight contours of the "unit-time reference dose rate", the dose rate one hour after the explosion if all the fallout had already landed: 3,000, 1,000, 300, 100, 30, 10, 3 and 1 rad per hour. For each contour it gives the distance downwind, the maximum width and the width at ground zero, each as a simple power of the yield. For example, the 1 rad per hour contour reaches 40 miles downwind for 1 kiloton, growing with the yield to the power 0.45. Upwind, each contour extends about half its ground-zero width (§9.93).

The table assumes one fixed "effective wind" of 15 miles per hour (about 24 km/h) with 15° of wind shear, and the simulator uses exactly that wind. It is not adjustable, and the pattern always points in one fixed direction (east). It is an idealized illustration, not a weather forecast. The shape follows the book's cigar-shaped Fig. 9.93; the exact outline between the tabulated widths is our drawing choice. Contour sizes follow the total yield, while the dose-rate values follow the fission share (§9.94). We assume pure fission up to 100 kilotons and 50% fission above that, the same split used for initial radiation. No dose to people and no casualty figures are derived from it.

People living inside each ring

Under each distance the results panel shows about how many people live inside that circle. It is a count of residents, taken from a population map; it is not an estimate of deaths, injuries, or anyone harmed, and the simulator never makes one.

The population map is GHSL GHS-POP R2023A for the year 2020, on a grid of cells about 1 km across (30 arc-seconds), where each cell holds the estimated number of residents. We keep that 1 km grid and cut it into 2° by 2° tiles. When you detonate, your browser loads only the tiles that touch the largest ring and adds up the cells: a cell fully inside a circle counts in full, one fully outside counts zero, and a cell on the edge counts in proportion to the share of its area inside the circle. Distances are measured along the Earth's surface. Each ring's figure is everyone inside that circle, so the smaller rings are part of the larger ones; the figures are never added together.

Limits: the counts are where people live (night-time residents), not where they are at a given moment, so a business district holds far more people on a weekday than it shows here. The grid is about 1 km, so rings much smaller than a cell are approximate. Figures are rounded to two significant digits.

Sources

Every number the simulator shows comes from published, public literature. No code from NUKEMAP or any other unlicensed project is used.

• Glasstone, Samuel, and Philip J. Dolan, eds. The Effects of Nuclear Weapons. 3rd ed. Washington, D.C.: U.S. Department of Defense and U.S. Energy Research and Development Administration, 1977 (sold by the Superintendent of Documents, U.S. Government Printing Office). Used for:
  • Fireball size: §1.32 and §2.05 (1-megaton fireball about 5,700 ft across), §2.127, eqs. (2.127.1) and (2.127.2) (radius scales as yield^0.4; a surface burst behaves like twice the yield).
  • The fixed airburst example: §2.24 (Hiroshima: burst height 1,670 ft, yield about 12.5 kilotons), scaled with the cube root of yield as in the notes to Figs. 3.73a to 3.73c.
  • Blast (chapter III): Figs. 3.72 and 3.73a to 3.73c and their worked examples, used as reference cases for the blast model; §2.34 and §2.35 and §5.53 as further checks.
  • Thermal radiation (chapters VII and XII): Fig. 7.42 (distance for a given heat dose vs. yield, air bursts, 12-mile visibility), Table 7.88 and §7.101 (heat fraction 0.35 for air bursts, 0.18 for surface bursts), Fig. 12.65 (heat dose for 50% probability of third-degree burns).
  • Initial nuclear radiation (chapter VIII): Figs. 8.123a and b, 8.127a and b, 8.130a and b, and 8.132 with §8.124 and §8.131 (neutron, secondary gamma, and fission-product gamma doses); Figs. 8.33a and b split fission (up to 100 kt) and thermonuclear weapons.
  • Fallout (chapter IX, "Fallout Predictions for Land Surface Bursts"): §9.04 (no significant local fallout from an air burst), §9.83 to §9.85 (idealized patterns and effective wind), §9.93 with Table 9.93 and Fig. 9.93 (idealized unit-time reference dose-rate contours for a contact surface burst, 15 mph effective wind, 15° shear), §9.94 (fission fraction), the Fallout Example, and §9.95 and §9.99 to §9.101 (limitations). The Fallout Example is used as a reference case.
• GOFAI, glasstone: nuclear weapons effects modelling in Python, github.com/GOFAI/glasstone, Copyright © 2016 GOFAI, released under the MIT License. The blast model (a curve fit by H. L. Brode, Airblast From Nuclear Bursts: Analytic Approximations, Pacific-Sierra Research Corporation, 1986, as cited by glasstone) and the digitized initial-radiation curves from Glasstone and Dolan chapter VIII are ported from this library. The MIT license text ships with our source code (LICENSES/glasstone-MIT.txt).

Fallout uses Glasstone and Dolan's own idealized contours (Table 9.93). The simplified "Miller" fallout scaling mentioned by other simulators is not used, because we could not locate and fully cite the original report.

How the numbers are checked

The model is tested automatically against about thirty values printed in Glasstone and Dolan (1977): worked examples in the text and readings taken from its figures. Each check has a tolerance of ±10% for numbers printed in the text and ±15% for values read from graphs. Values the model itself produces are never used as reference values.

Three cases fall outside those tolerances, and they are kept in the test file (tests/effects.test.ts) as known deviations rather than hidden. For example, the 1 psi distance of a small surface burst is about 20% larger than one of the book's graphs, although it agrees with the book's own "twice the yield" rule. The fireball has no independent check, because the book's one clear fireball figure was used to set the formula.

The distances quoted in the text and tables on this site are generated from the same model when the site is built, so they always match what the map shows.

Limitations

These are rough, idealized estimates for education. Real outcomes could be quite different.

• Flat, open ground and a standard sea-level atmosphere are assumed. Hills, buildings, and weather are ignored, and they can change results a lot.
• Heat (burn) distances assume a clear day with 12 miles (about 19 km) of visibility and bare skin facing the flash. Haze, cloud, smoke, or any shade reduce them; snow or cloud above the burst can increase them.
• Radiation distances assume a typical weapon design; the source rates its curves as reliable only to within roughly a factor of two (0.5 to 2 times for fission weapons, 0.25 to 1.5 times for thermonuclear weapons). The 500 rem contour counts neutron and gamma doses equally. Yields up to 100 kt use fission-weapon curves, larger yields use thermonuclear curves, so there is a small jump at 100 kt.
• The airburst option always uses one fixed example burst height: Hiroshima's (1,670 ft for about 12.5 kt), scaled to the chosen yield. No burst height optimization or damage-maximizing features are offered, by design.
• Very small (below 1 kt) and very large (above 10 to 20 Mt) yields go beyond the source charts; results there are extrapolations and are flagged in the results panel.
• Rings are sharp lines, but real effects fade gradually with distance.
• Fallout is an idealized pattern for one fixed textbook wind (15 mph, one direction), not a forecast. Glasstone and Dolan warn that real patterns differ in detail: the idealized shapes assume little wind shear, which tends to maximize how far the contours reach downwind; with more shear the pattern is wider and shorter, and in extreme cases it spreads in two or more directions (§9.99). Real fallout has unpredictable hot spots, and rain can wash it out unevenly (§9.100, §9.101). The dose rates are for a perfectly smooth open surface; ordinary ground roughness lowers them to about 70% and hilly terrain to about 50 to 60% (§9.95), and buildings lower them further. The book says these patterns are meant only for overall planning.
• Population figures show people living inside a zone (residents in 2020), never casualty estimates. See how they are counted.
• Not affiliated with NUKEMAP or any government agency.

Data and attribution

• Map tiles and buildings:OpenFreeMap, using data © OpenStreetMap contributors (Open Database License). The attribution is also shown in the corner of the map.
• City search: GeoNames (cities15000 dataset), licensed under CC BY 4.0, trimmed by us to cities with a population of at least 500,000.
• Effects model: blast fit and radiation curve data ported from GOFAI/glasstone, MIT License, Copyright © 2016 GOFAI.
• Map rendering: MapLibre GL JS (BSD 3-Clause License).
• Population: Schiavina, M., Freire, S., & MacManus, K. (2023). GHS-POP R2023A - GHS population grid multitemporal (1975-2030) [Dataset]. European Commission, Joint Research Centre (JRC). doi:10.2905/2FF68A52-5B5B-4A22-8F40-C41DA8332CFE. We use the epoch 2020, 30 arc-second (WGS84) file GHS_POP_E2020_GLOBE_R2023A_4326_30ss_V1_0. © European Union, 1995-2026, licensed under CC BY 4.0. Changes: we re-aligned the grid to whole 30-arc-second cells (a shift of about 46 m), rounded counts to whole people per 2° tile (tile totals preserved), and split it into 2° × 2° tiles. The world total after processing is 7,840,952,941, the same as the source to within 10 people.

Want to see the methods in action? Try the nuke simulator, or read the nuclear blast radius guide for a table of every preset.

For a historical example, read the Hiroshima bomb blast radius guide, which distinguishes model estimates from historical observations.