Initially little was known about the dispersion of nuclear fallout on a global scale. The AEC assumed that fallout would be dispersed evenly across the globe, dispersed by atmospheric winds and will gradually settle to the earth's surface after weeks, months, and even years as worldwide fallout. Nuclear products were deposited in the Northern Hemisphere becoming "far more dangerous than they had originally been estimated."
The radio-biological hazard of worldwide fallout is essentially a long-term one because of the potential accumulation of long-lived radioisotopes (such as strontium-90 and caesium-137) in the body as a result of ingestion of foods containing the radioactive materials. This hazard is much less serious than those which are associated with local fallout, which is of much greater immediate operational concern.
There will be large amounts of particles of less than 100 nm to several millimeters in diameter generated in a surface burst in addition to the very fine particles which contribute to worldwide fallout. The larger particles spill out of the stem and cascade down the outside of the fireball in a downdraft even while the cloud rises, so fallout begins to arrive near ground zero within an hour, and more than half the total bomb debris is deposited on the ground within about 24 hours as local fallout.
The chemical properties of the different elements in the fallout will control the rate at which they are deposited on the ground. The less volatile elements will deposit first.
Severe local fallout contamination can extend far beyond the blast and thermal effects, particularly in the case of high yield surface detonations. The ground track of fallout from an explosion depends on the weather situation from the time of detonation onwards. In stronger winds, fallout travels faster but takes the same time to descend, so although it covers a larger path, it is more spread out or diluted. So the width of the fallout pattern for any given dose rate is reduced where the downwind distance is increased by higher winds. The total amount of activity deposited up to any given time is the same irrespective of the wind pattern, so the overall casualty figures from fallout will generally be independent of the winds. But thunderstorms can bring down activity as rain more rapidly than dry fallout, particularly if the mushroom cloud is low enough to be below, or mixed with, the thunderstorm.
Whenever individuals remain in a radiologically contaminated area, such contamination will lead to an immediate external radiation exposure as well as a possible later internal hazard from inhalation and ingestion of radiocontaminants, such as the rather short-lived iodine-131, which is accumulated in the thyroid.
There are two main considerations for the location of an explosion: height and surface composition. A nuclear weapon detonated in the air, called an air burst, will produce less fallout than a comparable explosion near the ground. Less particulate matter will be contaminated by an air burst. Detonations at the surface (surface bursts) will tend to produce more fallout material.
In case of water surface bursts, the particles tend to be rather lighter and smaller, producing less local fallout but extending over a greater area. The particles contain mostly sea salts with some water; these can have a cloud seeding effect causing local rainout and areas of high local fallout. Fallout from a seawater burst is difficult to remove once it has soaked into porous surfaces because the fission products are present as metallic ions which become chemically bonded to many surfaces. Water and detergent washing is effective on removing less than about 50% of this chemically bonded activity from concrete or steel (complete decontamination requires aggressive treatment like sandblasting, or acidic treatment). After the Crossroads underwater test, it was found that wet fallout needs to be immediately removed from ships by continuous water washdown (such as from the fire sprinkler system on the decks).
For subsurface bursts, there is an additional phenomenon present called "base surge". The base surge is a cloud that rolls outward from the bottom of the subsiding column, which is cause by an excessive density of dust or water droplets in the air. For underwater bursts, the visible surge is, in effect, a cloud of liquid (usually water) droplets with the property of flowing almost as if it were a homogeneous fluid. After the water evaporates, an invisible base surge of small radioactive particles may persist.
For subsurface land bursts, the surge is made up of small solid particles, but it still behaves like a fluid. A soil earth medium favors base surge formation in an underground burst. Although the base surge typically contains only about 10% of the total bomb debris in a subsurface burst, it can create larger radiation doses than fallout near the detonation, because it arrives sooner than fallout, before much radioactive decay has occurred.
Meteorological conditions will greatly influence fallout, particularly local fallout. Atmospheric winds are able to bring fallout over large areas. For example, as a result of a Castle Bravo surface burst of a 15 Mt thermonuclear device at Bikini Atoll on March 1, 1954, a roughly cigar-shaped area of the Pacific extending over 500 km downwind and varying in width to a maximum of 100 km was severely contaminated. There are three very different versions of the fallout pattern from this test, because the fallout was only measured on a small number of widely spaced Pacific Atolls. The two alternative versions both ascribe the high radiation levels at north Rongelap to a downwind hotspot caused by the large amount of radioactivity carried on fallout particles of about 50-100 micrometres size
After Bravo, it was discovered that fallout landing on the ocean disperses in the top water layer (above the thermocline at 100 m depth), and the land equivalent dose rate can be calculated by multiplying the ocean dose rate at two days after burst by a factor of about 530. In other 1954 tests, including Yankee and Nectar, hotspots were mapped out by ships with submersible probes, and similar hotspots occurred in 1956 tests such as Zuni and Tewa However, the major U.S. 'DELFIC' (Defence Land Fallout Interpretive Code) computer calculations use the natural size distributions of particles in soil instead of the afterwind sweep-up spectrum, and this results in more straightforward fallout patterns lacking the downwind hotspot.
Snow and rain, especially if they come from considerable heights, will accelerate local fallout. Under special meteorological conditions, such as a local rain shower that originates above the radio-active cloud, limited areas of heavy contamination just downwind of a nuclear blast may be formed.
The unit of actual exposure is the Röntgen which is defined in ionisations per unit volume of air, and all ionisation based instruments (including geiger counters and ionisation chambers) measure exposure. However, effects depend on the energy per unit mass, not the exposure measured in air. A deposit of 1 joule per kilogram has the unit of 1 gray. For 1 MeV energy gamma rays, an exposure of 1 röntgen in air will produce a dose of about 0.01 gray (1 centigray, cGy) in water or surface tissue. Because of shielding by the tissue surrounding the bones, the bone marrow will only receive about 0.67 cGy when the air exposure is 1 röntgen and the surface skin dose is 1 cGy. Some of the lower values reported for the amount of radiation which would kill 50% of personnel (the 'LD50') refer to bone marrow dose, which is only 67% of the air dose.
In the 1950s, the LD50 for gamma rays was set at 3.5 Gy, while under more dire conditions of war (a bad diet, little medical care, poor nursing) the LD50 was 2.5 Gy (250 rad). There have been few documented cases of survival beyond 6 Gy. One person at Chernobyl survived a dose of more than 10 Gy, but many of the persons exposed there were not uniformly exposed over their entire body. If a person is exposed in a non-homogeneous manner then a given dose (averaged over the entire body) is less likely to be of a lethal dose. For instance if a person gets a hand/low arm dose of 100 Gy which gives them an overall dose of 4 Gy then they are more likely to survive than a person who gets a 4 Gy dose uniformly over their entire body. A hand dose of 10 Gy or more will likely result in loss of the hand; a British industrial radiographer who got a lifetime hand dose of 100 Gy lost his hand because of radiation dermatitis. Most people become ill after an exposure to 1 Gy or more. The fetuses of pregnant women are often more vulnerable than the host body and may miscarry, especially in the first trimester. Though the human biology resists mutation from large radiation exposure; grossly mutated fetuses usually miscarry, and this often causes gene-faults.
One hour after a surface burst, the radiation from fallout in the crater region is 30 grays per hour (Gy/h). Civilian dose rates in peacetime range from 30 to 100 µGy per year.
Fallout radiation decays exponentially relatively quickly with time. Most areas become fairly safe for travel and decontamination after three to five weeks.
For yields of up to 10 kt, prompt radiation is the dominant producer of casualties on the battlefield. Humans receiving an acute incapacitating dose (30 Gy) will have their performance degraded almost immediately and become ineffective within several hours. However, they will not die until 5 to 6 days after exposure assuming they do not receive any other injuries.
Individuals receiving less than a total of 1.5 Gy will not be incapacitated. People receiving doses greater than 1.5 Gy will become disabled; some will eventually die.
A dose of 5.3 Gy to 8.3 Gy is considered lethal but not immediately incapacitating. Personnel exposed to this amount of radiation will have their performance degraded within 2 to 3 hours, depending on how physically demanding the tasks they must perform are, and will remain in this disabled state at least 2 days. However, at that point they will experience a recovery period and be effective at performing non-demanding tasks for about 6 days, after which they will relapse for about 4 weeks. At this time they will begin exhibiting symptoms of radiation poisoning of sufficient severity to render them totally ineffective. Death follows at approximately 6 weeks after exposure, although results may vary.
Blast injuries and thermal burns from the use of nuclear weapons for military action in many cases will far outnumber radiation injuries. However, radiation effects are considerably more complex and varied than are blast or thermal effects and are subject to considerable misunderstanding.
The closer to ground an atomic bomb is detonated, the more dust and debris is thrown into the air, resulting in greater amounts of local fallout. From a tactical standpoint, this has the disadvantage of hindering any occupation/invading efforts until the fallout clears, but more directly, the impact with the ground severely limits the destructive force of the bomb. For these reasons, ground bursts are not usually considered tactically advantageous, with the exception of hardened underground targets such as missile silos or command centers, however "salting" enemy territory with a fallout-heavy atomic burst can be used to deny ill-equipped civilians/military personnel access to a contaminated area.
Below is shown a comparison of the calculated gamma dose rates in open air from the fallout of a fission bomb and of the Chernobyl release. It is clear that average half-life of the Chernobyl release is longer than that for the bomb fallout.