U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Steering Committee for the Symposium on the Medical Implications of Nuclear War; Solomon F, Marston RQ, editors. The Medical Implications of Nuclear War. Washington (DC): National Academies Press (US); 1986.

Cover of The Medical Implications of Nuclear War

The Medical Implications of Nuclear War.

Show details

Radioactive Fallout

Charles S. Shapiro, PH.D.

San Francisco State University, San Francisco, California

Lawrence Livermore National Laboratory, Livermore, California

Ted F. Harvey, PH.D., and Kendall R. Peterson, M.S.

Lawrence Livermore National Laboratory, Livermore, California


Potential radiation doses from several scenarios involving nuclear attack on an unsheltered United States population are calculated for local, intermediate time scale and long-term fallout. Dose estimates are made for both a normal atmosphere and an atmosphere perturbed by smoke produced by massive fires. A separate section discusses the additional doses from nuclear fuel facilities, were they to be targeted in an attack. Finally, in an appendix the direct effects of fallout on humans are considered. These include effects of sheltering and biological repair of damage from chronic doses.

Radioactivity From Nuclear Weapons


In this paper the potential doses associated with the radionuclides created by nuclear explosions are assessed. Our focus is on the areas outside the zone of the initial blast and fires. Prompt initial ionizing radiation within the first minute after the explosion is not considered here, because the physical range for biological damage from this source for large-yield weapons is generally smaller than the ranges for blast and thermal effects.

The contributions from local (first 24 hours) and more widely distributed, or global fallout, will be considered separately. Global fallout will be further subdivided into an intermediate time scale, sometimes called tropospheric, of 1 to 30 days, and a long-term (beyond 30 days) stratospheric component. Mainly the dose from gamma-ray emitters external to the body is considered. Contributions from external beta emitters are not estimated because of the limited penetration ability of beta radiation, but there is the possibility that in areas of local fallout, beta radiation can have a significant impact on certain biota directly exposed to the emitters by surface deposition (Svirezhev, 1985). Potential internal doses from ingestion and inhalation of gamma and beta emitters are estimated in only an approximate manner, as these are much more difficult to quantify.

The total amount of gamma-ray radioactivity dispersed in a nuclear exchange is dominated by the weapon fission products whose production is proportional to the total fission yield of the exchange. Exposure to local fallout, which has the greatest potential for producing human casualties, is very sensitive to assumptions about height of burst, winds, time of exposure, protection factor, and other variables. For global fallout, the dose commitments are sensitive to how these fission products are injected into various regions of the atmosphere, which depend on individual war-head yield as well as burst location.

For local fallout, aspects of the baseline scenario outlined in the Scientific Committee on Problems of the Environment-Environmental Effects of Nuclear War (SCOPE-ENUWAR) Study (Pittock et al., 1985) are considered. For global fallout, both the 5,300-megaton (Mt) baseline scenario reported by Knox (1983) and the 5,000-Mt reference nuclear war scenario described by Turco et al. (1983; also known as the TTAPS study) are considered.

Local Fallout

Local fallout is the early deposition of relatively large radioactive particles that are lofted by a nuclear explosion occurring near the surface in which large quantities of debris are drawn into the fireball. For nuclear weapons, the primary early danger from local fallout is due to gamma radiation.

Fresh fission products are highly radioactive and most decay by simultaneous emission of electrons and gamma rays. An approximate rule of thumb for the first 6 months following a weapon detonation is that the gamma radiation will decay by an order of magnitude for every factor of seven in time (Glasstone and Dolan, 1977).

If the implausible assumption is made that all of the radioactivity in the fresh nuclear debris from a 1-Mt, all-fission weapon arrives on the ground 1 hour after detonation and is uniformly spread over grassy ground such that it would just give a 48-hour unshielded dose of 450 rads, then approximately 50,000 km2 could be covered. Given such a uniform deposition model, it would require only about 100 such weapons to completely cover Europe. In reality, because of a variety of physical processes, the actual areas affected are much smaller. Most of the radioactivity is airborne for much longer than an hour, thus allowing substantial decay to occur before reaching the ground. Also, the deposition pattern of the radioactivity is uneven, with the heaviest fallout being near the detonation point where extremely high radiation levels occur. When realistic depositional processes are considered, the approximate area covered by a 48-hour unshielded 450-rad dose is about 1,300 km2, i.e., nearly a factor of 40 smaller than the area predicted using the simplistic model above. This large factor is partially explained because only about one-half of the radioactivity from ground bursts is on fallout-sized particles (Defense Civil Preparedness Agency [DCPA] 1973). The other portion of the radioactivity is found on smaller particles that have very low settling velocities and therefore contribute to global fallout over longer times. Portions of this radioactivity can remain airborne for years. For airbursts of strategic-sized weapons, virtually no fallout-sized particles are created, and all of the radioactivity contributes to global fallout.

Lofted radioactive fallout particles that have radii exceeding 5 to 10 µm have sufficient fall velocities to contribute to local fallout. Some particles can be as large as several millimeters in radius. Settling velocities range from a few centimeters per second to many tens of meters per second for these particles. They are lofted by the rising nuclear debris cloud and are detrained anywhere from ground level to the top of the stabilized cloud. Horizontal wind speeds usually increase with height up to the tropopause, and frequently, wind directions have large angular shears. Nuclear clouds disperse due to atmospheric shears and turbulence. The arrival of radioactivity at a given location can occur over many hours, with large particles from high in the cloud usually arriving first at a downwind location.

Rainout effects have been suggested as being potentially significant contributors to local fallout effects from strategic nuclear war (Glasstone and Dolan, 1977). The inclusion of rainout processes would probably not significantly affect the answers to generic questions pertaining to large-scale nuclear war phenomena (for example, What percentage of Western Europe would suffer lethal levels of gamma radiation from local fallout in a large-scale nuclear exchange?), especially if a substantial portion of the weapons are surface burst. This is particularly true for strategic weapon yields of greater than 30 kilotons (kt), because the radioactivity on the small particles most affected by rainout rises above all but the largest convective rain cells. Thus potentially lethal doses from rainout should occur only from large convective rain cells, and this should occur only over relatively small areas (i.e., beneath moving convective cells). However, for any given radioactive air parcel, the overall probability of rainout the first day from a convective cell is quite low for yields greater than 30 kt. Rainout also may occur over large areas associated with frontal systems, but in the case of strategic weapons yields, the radioactivity on small particles must diffuse downward from levels that are often above the top of the precipitation system to produce rainout. As a result, radiological doses from debris in precipitation would be substantially lower than early-time doses associated with local fallout. In either case (frontal or convective rainout), for a large-scale multi-burst exchange, the size of the expected lethal-dose rainout areas should typically be small (i.e., well within the range of modeling uncertainty) compared to the size of the fallout areas created by particles with large settling velocities. Thus, to first order rainout areas can be ignored in calculating the radiological hazard from a large-scale nuclear war scenario. However, for lower yield ( tactical war scenarios, or for scenarios at specific locations, rainout could lead to important and dominant radiological effects.

Single-Weapon Fallout Model

For this work the KDFOC2 computer model (Harvey and Serduke, 1979) was used to calculate fallout fields for single bursts, which in turn were used to develop a semiquantitative model for preparing rough estimates of fallout areas for typical strategic weapons. A wind profile (including shear) characteristic of midcontinental Northern Hemisphere summer conditions was selected from observations, and baseline fallout calculations were performed for several explosion yields under the assumption that all-fission weapons were used. As an example of the results, a 1-Mt fallout pattern is shown in Figure 1. Figure 2 gives the area versus minimum dose relationship for several different yields. Fallout areas are shown rather than maximum downwind extents for various doses since areas are less sensitive to variations in wind direction and speed shears and should be more useful for analysis. These areas correspond to unshielded doses associated with external gamma-ray emissions. All of the local fallout estimates given below are based on the KDFOC2 model and the wind pattern used for Figure 1.

Figure 1. 48-hour dose predictions for a 1-Mt all-fission weapon detonated at the surface.

Figure 1

48-hour dose predictions for a 1-Mt all-fission weapon detonated at the surface. A midcontinental Northern Hemisphere summer wind profile was used. The double-lobed pattern is due to a strong directional wind shear that is typical during this season. (more...)

Figure 2. Fallout areas versus minimum 48-hour doses for selected yields from 30 kt to 5 Mt.

Figure 2

Fallout areas versus minimum 48-hour doses for selected yields from 30 kt to 5 Mt. The weapons were surface burst and all fission. The wind was that used in the calculation to produce Figure 1. These curves include an instrument shielding factor of about (more...)

To convert from areas for the 48-hour curves shown in Figure 2 to areas for minimum doses over longer times, an area multiplication factor, AMF, is given in Figure 3. For example, if the 2-week, 300-rad area is needed, first the 48-hour, 300-rad area is found from Figure 2 and then the appropriate AMF is read from Figure 3. The 2-week, 300-rad area is the product of the 300-rad, 48-hour area and the 2-week, 300-rad AMF. For example, a 1-Mt, all-fission weapon, has a 2-week, 300-rad area of

Figure 3. Area multiplication factors to extend the dose integration time from 48 hours to longer times.

Figure 3

Area multiplication factors to extend the dose integration time from 48 hours to longer times. These factors must be used in conjunction with the areas given in Figure 2. Source: Pittock et al. (1985, p. 244). Reprinted with permission from the Scientific (more...)

There are two scaling laws that allow weapons design and various sheltering to be factored into dose calculations. The first scaling law permits consideration of weapons that are not all fission. Most large-yield weapons (> 100 kt) are combined fission-fusion explosives with approximately equal amounts of fusion and fission (Fetter and Tsipis, 1981). The fission fraction (ρ) is the ratio

To find a 48-hour minimum dose area for a particular fission fraction using Figures 2 and 3, the dose of interest, D, should be multiplied by 1/p before reading the values of the area and the area multiplication factor. For example, to obtain the 450-rad, 48-hour dose area for a 50 percent fission weapon, the area for the scaled dose of 900 rads would be obtained from Figure 2. For a 1-Mt, 50 percent fission weapon, the estimated 450-rad dose area is found to be 720 km2. The rationale for this scaling law is that the thermodynamics and hydrodynamics of fallout development are insensitive to fission fraction because particle characteristics and lofting altitudes are determined predominantly by total energy yield. For yields that are only part fission, each particle has a fraction of the gamma radioactivity that it would otherwise have if the weapon were an all-fission weapon. This scaling law is appropriate for fission fraction ratios above ~0.3; smaller ratios can lead to situations where neutron-induced radioactivity becomes a significant factor. For such cases, careful consideration of surrounding materials may be necessary to produce accurate fallout estimates.

The second scaling law accounts for protection factors (K) against ionizing radiation that would be provided by sheltering. The 48-hour minimum dose areas given in Figure 2 are appropriate for a person or other organism located on a rolling grassy plain. In other configurations, radiation exposure varies according to how much shielding is obtained while a person remains in the area. For example, a person leading a normal lifestyle is likely to achieve an average K of 2 to 3 for gamma radiation from time spent inside buildings and other structures. Basements can provide K values of 10 to 20. Specially constructed shelters can provide K values of 10 to 10,000 (Glasstone and Dolan, 1977).

To determine the radiation area for a dose of D when shielding with a protection factor K is available, the scaled dose KD from Figure 2 should be used. For example, for those in an undamaged basement with K = 10 for the first 48 hours, Figure 2 indicates that the effective dose area of 450 rads or more from a 1-Mt, all-fission weapon is about 130 km2. This is obtained by using a scaled dose of 4,500 rads. For comparison, the 450-rad minimum dose area is about 1,300 km2 for people with no shelter, greater by a factor of 10 than the area for those with a K of 10.

Other factors that could reduce the effects of fallout on the population over long time periods (month) include weathering (runoff and soil penetration), cleanup measures, relocation, and the ability of the body to repair itself when the dose is spread over time or occurs at lower rates. These considerations can be taken into account with existing computer models but are not treated here. Several factors that could enhance the effects of fallout are mentioned below.

Dose Estimation From Multiple Explosions

In a major nuclear exchange, thousands of nuclear warheads could be detonated. For such an exchange, realistic wind patterns and targeting scenarios could cause individual weapon fallout patterns to overlap in complicated ways that are difficult to predict and calculate. Even though acute doses are additive, a single-dose pattern calculated for a weapon cannot be used directly to add up doses in a multiweapon scenario, except under limited conditions. For example, if the wind speed and direction are not approximately the same for the detonation of each weapon, then different patterns should be used. In addition, the number of possible fallout scenarios far exceeds the number of targeting scenarios. This is because, for each targeting scenario that exists, the possible meteorological situations are numerous, complex, and varying. Thus, only under limited conditions may a single dose pattern be moved around a dose accumulation grid to obtain the sum of total doses from many weapons.

Two relatively simple multiburst models can be developed for use in conjunction with the semiquantitative model presented here. These cases can provide rough estimates of fallout areas from multiple weapons scenarios; however, their results have an uncertainty of no better than a factor of several, for reasons explained below, and are neither upper nor lower case limits. The no-overlap (NO) case is considered first; this could occur when targets are dispersed, there is one warhead per target and the fallout areas essentially do not overlap. Second, the total-overlap (TO) case is examined where multiple bursts are assumed to be at the same burst location. This approximation would arise when targets are densely packed and warheads of the same size are used against each. A large number of warheads used against, say, a hardened missile field site would be more closely modeled by the TO model than the NO model.

As an example of the use of the NO and TO approximations, a case with 100 1-Mt, 50 percent fission, surface-detonated explosions is considered, and estimates are developed for the 450-rad, 48-hour dose areas for both cases. For the NO case the fallout area can be obtained by determining the area for a single 1-Mt weapon (900-rad scaled dose from Figure 2) and multiplying by 100. This gives 7.2 × 104 km2 for the 450-rad, 48-hour dose contour. For the TO model, the area is obtained for a single 1-Mt weapon, 9-rad scaled dose from Figure 2. One hundred of these, laid on top of each other, would give 450 rads for 50 percent fission weapons. The area in this case is 3.3 × 104 km2. These results differ by about a factor of two, with the NO case giving a larger area.

Although these models are extremes in terms of fallout pattern overlap, neither can be taken as a bounding calculation of the extremes in fallout areas for specified doses. It is very possible that a more realistic calculation of overlap would produce a greater area for 100 weapons than either of these models. Such a result is demonstrated by a more sophisticated model prediction that explicitly takes overlap into account (Harvey, 1982). In this study, a scenario was developed for a severe case of fallout in a countervalue attack on the United States where population centers were targeted with surface bursts. Figure 4 shows the contours of a 500-rad minimum 1-week dose where overlap was considered. The 500-rad area is about three times greater than that predicted by the NO model and six times that of the TO model. Note also that the distribution of radioactivity is extremely uneven. About 20 percent of the United States is covered with 500-rad contours, including nearly 100 percent of the northeast, approximately 50 percent of the area east of the Mississippi River, 10 percent of the area west of the Mississippi River, and only a few percent of the area in the Great Plains.

Figure 4. A fallout assessment that explicitly takes fallout pattern overlap into account.

Figure 4

A fallout assessment that explicitly takes fallout pattern overlap into account. Shown are 500-rad, 1-week minimum isodose contours. This scenario was intended to emphasize population dose. Approximately 1,000 population centers in the United States were (more...)

Results of these scenarios, as well as those postulated by others, clearly show that such estimates are very scenario dependent and that detailed estimates should be made with care. For example, the regional results shown in Figure 4 could be significantly different if military targets (e.g., intercontinental ballistic missile [ICBM] silos) were included as well. Although the NO and TO cases presented in this paper are simple to apply, they must be used only to develop rough estimates of total area coverage within regions with relatively uniformly dispersed targets. When the density of targets of one area is as large as that in the northeastern United States and another is as dispersed as that in the western United States, regional models should be used to develop specific regional estimates. Even then, multiple-weapon fallout estimates should be considered to have uncertainties no smaller than a factor of several, with the uncertainty factor increasing as the model sophistication decreases.

Sample Calculation Of Multiple-Weapon Fallout

To illustrate the fallout prediction method presented here, an escalating nuclear exchange scenario, which is consistent with that described in the SCOPE-ENUWAR study (Pittock et al., 1985), is used to estimate fallout areas. In this scenario there are four sequential phases of attack against five different regions. The five regions are Europe (both east and west), western USSR (west of the Ural Mountains), eastern USSR, the western United States (west of 96° W latitude), and the eastern United States. The four phases of attack are initial counterforce, extended counterforce, industrial countervalue, and a final phase of mixed military and countervalue targeting. The weapon yields and the number of warheads that are employed for just the surface bursts during each phase are shown in Table 1. Airbursts are omitted since they do not produce appreciable local fallout.

Table 1. Surface-Burst Warheads in a Phased Nuclear Exchange.

Table 1

Surface-Burst Warheads in a Phased Nuclear Exchange.

In the first phase, land-based ICBMs are the primary targets. These are assumed to be located in the western United States and the USSR at sites containing 125 to 275 missiles. The geographical distribution of missile silos in the USSR is assumed to be 50 percent east and 50 percent west of the Ural Mountains. Each missile silo is attacked with a surface-burst and an airburst weapon. For a given site, the TO model is used to calculate the fallout pattern. All U.S. ICBM sites are attacked with 0.5-Mt weapons. Each of five U.S. ICBM complexes is presumed to have 200 missile silos, while each of six USSR complexes is presumed to have between 125 and 275 missile silos, with a total of 1,300. The Soviet sites are attacked with 1-, 0.3-, and 0.1-Mt weapons. During this phase, each side employs a total of about 1,000 Mt. Besides the attack on Soviet missile silos, 425 0.1-Mt weapons are assumed to be surface-burst against other Soviet military targets, with approximately 28 Mt west of the Urals and 14 Mt to the east. The 425 fallout patterns from these weapons have been modeled with the NO model.

In the second phase of the attack, there are an additional 1,000 Mt of surface-burst weapons employed. These are employed against each region with 20, 40, and 40 percent of the weapons being used against targets in Europe, the United States and the USSR, respectively. Here, Europe includes both the North Atlantic Treaty Organization (NATO) and Warsaw Pact countries. To roughly account for population distribution, the weapons employed against the United States are divided up as two-thirds in the eastern U.S. and one-third in the western United States; for Soviet targets it is assumed that two-thirds are detonated west of and one-third are detonated east of the Ural Mountains.

For all the weapons employed in the second, third, and fourth phases, the fallout pattern is calculated using the NO model. The results, in terms of percent of land covered by at least a 450-rad, 48-hour dose, are shown in Table 2. No shielding has been assumed in calculating these percentages. Similar areas were found for 600 rads over 2 weeks.

Table 2. Percentage of Land Mass Covered by a Minimum 450-rad, 48-hour Dose.

Table 2

Percentage of Land Mass Covered by a Minimum 450-rad, 48-hour Dose.

Care must be taken in interpreting these results. To begin with, there is an uncertainty factor of several in the NO and TO modeling schemes, as discussed earlier. Another substantial bias is introduced by neglecting the radioactivity that is blown into or out of a region. For example, the western USSR would likely receive substantial amounts of radiation from weapons detonated in eastern Europe because the wind usually blows from Europe toward the Soviet Union. Thus, the area percentages shown in Table 2 for Europe would be expected to decrease since some of the area credited to Europe would actually be in the Soviet sector. Similarly, the percentage of radiation over the western United States is probably overestimated, assuming typical wind conditions. For the eastern United States the area covered would be increased by radioactivity originating in the central United States and decreased as a result of radioactivity blowing out over the Atlantic Ocean.

There are a number of factors that could change these local fallout assessments.

  • Shielding is probably the most sensitive parameter in reducing the effective dose to a population. This effect has been ignored in these calculations. Protective measures could substantially reduce the impact of fallout on humans.
  • Choosing a scenario that exacerbates local fallout (e.g., surface bursts over cities) could increase lethal areas by factors of several.
  • Large differences in doses could arise because of irregularities in fallout patterns in the local fallout zones that could range over orders of magnitude. Relocation could substantially reduce a population's dose.
  • Debilitating, but not lethal, radiation doses (~200 rads or more) would be received over much larger areas than areas receiving lethal doses.
  • Fission fractions of smaller modern weapons could be twice the baseline assumption of 0.5. Adding these to the scenario mix could increase lethal fallout areas by up to 20 percent of the baseline calculation.
  • Tactical weapons, ignored in the baseline scenario, could increase lethal local fallout areas in certain geographical regions, particularly within Europe, by about 20 percent of the baseline calculations.
  • Internal radiation exposure could increase the average total doses to humans by about 20 percent of the external dose.
  • External beta exposure, not treated here, could add significantly to plant and animal exposures in local fallout areas.
  • Targeting of nuclear fuel cycle facilities could contribute to radiation doses.

Global Fallout

Global fallout consists of the radioactivity carried by fine particulate matter and gaseous compounds that are lofted into the atmosphere by nuclear explosions. One may distinguish two components to global fall-out—intermediate time scale and long term. Intermediate-time-scale fallout consists of material that is initially injected into the troposphere and is removed principally by precipitation within the first month. The fractional contribution to intermediate-time-scale fallout decreases as the total weapon yield increases above 100 kt. The importance of intermediate-time-scale fallout has grown with reductions in warhead yields. Long-term fallout occurs as a result of deposition of very fine particles that are initially injected into the stratosphere. Because the stratosphere is so stable against vertical mixing and the fine particulate matter has negligible fall velocities, the primary deposition mechanism involves transport of the radioactivity to the troposphere through seasonal changes in stratospheric circulation. Once within the troposphere, these particles would normally be removed within a month by precipitation scavenging.

Given a specific nuclear war scenario, it is possible to use experience gained from atmospheric nuclear tests to estimate the fate of both inter-mediate-time-scale and long-term fallout particles if the atmosphere is not perturbed by smoke. GLODEP2 (Edwards et al., 1984), an empirical code that was designed to match measurements from atmospheric testing, has been used. The model contains two tropospheric and six stratospheric injection compartments. By following unique tracer material from several atmospheric nuclear tests in the late 1950s, combined with subsequent balloon and aircraft measurements in the stratosphere and upper troposphere and many surface air and precipitation observations, it was possible to estimate the residence time of radioactivity in the various stratospheric compartments and the interhemispheric exchange rate in the stratosphere. Radioactive material that is placed initially into the troposphere is also handled by the GLODEP2 model (Edwards et al., 1984). From this information, surface deposition tables were prepared. The GLODEP2 model has never been tested against atmospheric nuclear tests in middle latitudes since no extensive series of explosions have occurred in this region. As a result, there is some uncertainty in the results of explosions centered around the Northern Hemisphere middle latitudes, but little uncertainty in the Northern Hemisphere subpolar latitude calculations since the stratospheric fallout there would deposit much the same as the global fallout from the polar bursts used to generate the polar deposition tables in the model.

Global Dose In An Unperturbed Atmosphere Using Specific Scenarios

A variety of scenario studies have been performed using GLODEP2 (Knox, 1983; Edwards et al., 1984). Dose calculations for scenarios A and B, which are described in Table 3, are presented in detail in Table 4. The atmospheric compartments in Table 3 refer to those used in the GLODEP2 model.

Table 3. Nuclear War Scenarios.

Table 3

Nuclear War Scenarios.

Table 4. Global Fallout Dose Assessments (rads) for an Unperturbed Atmosphere with No Smoke.

Table 4

Global Fallout Dose Assessments (rads) for an Unperturbed Atmosphere with No Smoke.

From a comparison of GLODEP2 results for the A and B scenarios for a Northern Hemisphere winter injection (Table 4, columns A1 and B1 ), it is seen that the Northern Hemisphere averages for scenarios A and B are about 16 and 19 rads, respectively, while Southern Hemisphere averages are more than a factor of 20 smaller. The maximum appears in the 30 to 50µN latitude band, where scenarios A and B yield 33 and 42 rads, respectively. All the doses reported here for global fallout are integrated external gamma-ray exposure over 50 years and assume no sheltering, no weathering, and a smooth plane surface.

For scenario A, 55 percent of the dose emanates from the tropospheric injections. The corresponding value for B is 75 percent. This emphasizes the sensitivity of dose to the yield mix of the scenario. As individual warhead yields decrease, the fractional injections into the troposphere increase, resulting in much larger doses on the ground due to more rapid deposition. Tropospheric radioactivity injections per megaton of fission can produce doses on the ground about a factor of 10 greater than those resulting from lower stratospheric injections, which in turn contribute about 3 to 5 times higher doses compared to upper stratospheric injections (Shapiro, 1984). Injections of radioactivity above the stratosphere as a gas or as extremely fine particles would produce relatively negligible doses at the ground.

Table 4 includes calculated values for the global human population dose. This quantity is calculated by multiplying the dose in each 20°-wide latitude band by the population of the latitude band, and then summing over all latitudes. For a given scenario, this number is one measure of the potential global biological impact. The global population dose as calculated by GLODEP2 for scenarios A and B are 7 × 1010 and 8 × 1010 person-tads, respectively. Essentially all of this dose occurs in the Northern Hemisphere because 90 percent of the world's population and higher doses prevail there.

Figure 5 illustrates the time behavior of the buildup of the dose to the 50-year lifetime value as a function of latitude for scenario A. The bulk of the dose is caused by deposition (mainly from the troposphere) and exposure during the first season after the war, followed by a gradual rise to the 50-year value.

Figure 5. Global fallout: accumulated whole-body gamma dose (rads) from 6,235 explosions totaling 2,031 Mt of fission products (scenario A).

Figure 5

Global fallout: accumulated whole-body gamma dose (rads) from 6,235 explosions totaling 2,031 Mt of fission products (scenario A). An 8-day tropospheric deposition decay constant, characteristic of a winter injection, is assumed. Source: Pittock et al. (more...)

A comparison of the GLODEP2 results for the TRAPS scenario (B) and the results of Turco et al. (1983) (using an entirely different methodology) reveals that GLODEP2 doses are 19 rads for the Northern Hemi sphere average and 42 rads for the 30-50°N latitude band, while estimates of Turco et al. give corresponding doses of 20 rads and about 40 to 60 rads.

Other studies that have been undertaken using GLODEP2, and the 5,300-Mt scenario A have led to the following conclusions:

Winter versus Summer Injection Because of a decrease in the frequency and intensity of large-scale precipitation systems in summer, the doses from the troposphere and lower polar stratosphere are reduced somewhat in comparison to those in winter, while the upper stratospheric contribution is increased. The total dose differences between summer and winter are not large, and other sources of uncertainty would predominate.

Scenarios with Smaller-Yield Devices The long-term consequences of the shift in the nuclear arsenals from larger-to smaller-yield devices has been assessed. Table 5 presents results comparing the 5,300-Mt baseline scenario with two variations. In scenario Aa, the number of devices in the baseline scenario A is increased from 6,235 to 13,250, while the total yield is held at 5,300 Mt. In scenario Ab, smaller yields have been used, but the number of devices is constant at 6,235 (the total yield consequently is reduced by 25 percent, from 5,300 to 4,000 Mt). The figures presented are for the 50-year gamma-ray dose. For the same total yield, it is seen that a shift to smaller weapons in the baseline scenario has approximately doubled the dose (scenario Aa). For Ab, the dose remains about the same even with a 25 percent drop in the total yield.

Table 5. Global Fallout: Sensitivity of Dose to Warhead Yield.

Table 5

Global Fallout: Sensitivity of Dose to Warhead Yield.

Global Fallout In A Perturbed Atmosphere

Following a large-scale nuclear exchange, the large quantities of smoke and soot lofted to high altitudes could decrease the incoming solar radiation, resulting in tropospheric and stratospheric circulation changes. Over land in the Northern Hemisphere, the presence of smoke and soot would probably result in less precipitation and a lowering of the tropopause; these changes could decrease the intermediate-time-scale (tropospheric) fallout and, depending on changes in stratospheric circulation, could alter the stratospheric contribution to fallout in the Northern Hemisphere. However, before the stratospheric burden is carried into the troposphere, a sizeable fraction would be transported to the Southern Hemisphere by the accelerated interhemispheric transport, resulting in doses there that are likely to be increased over those calculated for an unperturbed atmosphere.

Both the GLODEP2 and the Turco et al. (1983) models assumed fission product depositions from a normal atmosphere in calculating global fallout. Preliminary studies have been conducted with radionuclides in a perturbed atmosphere using a three-dimensional version of the GRANTOUR model (see MacCracken and Walton, 1984). GRANTOUR is a three-dimensional transport model driven by meteorological data generated by the Oregon State University (OSU) general circulation model (Schlesinger and Gates, 1980). Particulate matter appearing as an initial distribution or generated by sources is advected by wind fields, locally diffused in the horizontal and vertical, moved vertically by convective fluxes and the re-evaporation of precipitation, and removed by precipitation scavenging and dry deposition. It is assumed that the fission products are in the form of particulate material in two size ranges: greater than and less than 1 µm in diameter. Coagulation from small to large particles is not treated in the version of the model used here.

Studies focused on comparisons of radiation dose assessments with smoke in the atmosphere (interactive atmosphere) and without smoke (noninteractive); other relevant parameters were also explored, including consideration of particle size distribution, source location, different initial meteorology, and averaging doses over land areas only. All of the GRANTOUR simulations reported here are for the Northern Hemisphere summer season and use five radioactivity and smoke source locations of equal strength. The locations include two in the United States, two in the USSR, and one in western Europe. This division of sources is similar to that assumed in our earlier discussion on local fallout. Sources were initially injected with a Gaussian distribution whose amplitude was 10 percent of the maximum at a radius of 15° along a great circle. The total amount of smoke injected was 150 teragrams (equivalent to the urban smoke contributions used by Turco et al. [1983] and the National Research Council [NRC, 1985]). MacCracken and Walton (1984) describe the induced climatic perturbations. The vertical distribution of the radioactivity injections were distributed, as was the smoke, with the same vertical distribution as the source term injections calculated using the GLODEP2 injection algorithm. Deposition was followed for 30 days in most calculations. A single 60-day run indicated that 30 days is sufficient to account for 90 percent of the deposition. Results are compared for a 50-year unsheltered, unweathered, external gamma-ray dose.

GRANTOUR treats only the troposphere and splits it into three vertical layers extending from 800 to 1,000, 400 to 800, and 200 to 400 mbar. In a normal atmosphere, these layers reach up to 2.0, 7.1, and 11.8 km. In the comparisons, GLODEP2 was used to estimate the dose contributions from the stratospheric injections, which were added to the doses calculated by GRANTOUR assuming altered climatic conditions.

Scenarios A and B were used in the calculations. Columns A2 and B2 in Table 4 display a comparison of the predictions of GLODEP2 for these two scenarios. Columns A3 and B3 list the results from GRANTOUR, assuming an unperturbed atmosphere (no smoke; no climatic perturbation) for the same two scenarios. There is reasonable agreement (i.e., generally within about 50 percent) between the GLODEP2 only and GRANTOUR/ GLODEP2 methodologies for an unperturbed atmosphere (scenarios 1 and 3), providing some confidence that the results of GLODEP2 and GRANTOUR can be combined for simulations with a perturbed atmosphere, although the initial accelerated interhemispheric mixing of radionuclides in the stratosphere has not yet been considered. This may lead to a small underestimate of the long-term Southern Hemisphere dose.

Table 6 compares calculations for a perturbed atmosphere (interactive smoke) with estimates for normal July conditions. These results are also shown in Figure 6 and indicate that the perturbed atmosphere lowers the average dose in the Northern Hemisphere by about 15 percent. Because the principal mechanism for radionuclide removal from the troposphere is precipitation, the GRANTOUR calculations are roughly consistent with the thesis that precipitation is inhibited when large amounts of smoke are introduced. The transfer of fission product radionuclides to the Southern Hemisphere is somewhat enhanced by the perturbed climate, resulting in higher doses than for the unperturbed case. The increases in Southern Hemisphere dose, however, are not large, and the resulting doses are still about a factor of 20 lower than those in the Northern Hemisphere. This is because the increased transfer to the Southern Hemisphere is mitigated by the decay in activity during the time before the radionuclides are deposited on the ground.

Table 6. Global Fallout Dose Using the Three-Dimensional GRANTOUR Model (summer scenario).

Table 6

Global Fallout Dose Using the Three-Dimensional GRANTOUR Model (summer scenario).

Figure 6. Comparison of radionuclide global dose distribution for cases with unperturbed and smoke-perturbed climates (tropospheric contributions only).

Figure 6

Comparison of radionuclide global dose distribution for cases with unperturbed and smoke-perturbed climates (tropospheric contributions only). Source: Pittock et al. (1985, p. 261). Reprinted with permission from the Scientific Committee on Problems of (more...)

Figure 6 reveals longitudinal, as well as latitudinal, details that are not apparent in the averages of Table 6. Scenario B is illustrated here since the changes due to smoke-induced effects are more apparent. The five original sources have produced four discernible peaks in the tropospheric dose distribution, and the two U.S. sources have merged in the 30-day dose distribution. The tabulated values presented in Table 6 are averages over 20° latitude bands. The dose in hotspots can be examined by looking at peaks on the 10° × 10° grid. Typically the highest value for a grid square (~5 × 105 km2) is about a factor of 6 to 8 higher than the Northern Hemisphere average dose. There will also be local areas much smaller than the 10° × 10° grid size where the peak doses would be considerably higher.

As GRANTOUR tests only the troposphere and GLODEP2 has been used for the stratospheric contributions (which assumes an unperturbed stratosphere), additional calculations using a computer model that includes the perturbed stratosphere should be undertaken.

Internal Dose Due To Inhalation And The Food Chain

One serious problem following a large-scale nuclear exchange is radioactive contamination of drinking water. Those cities that are damaged would undoubtedly lose their water system due to power loss and ruptured supply pipes. Suburban residents within the local fallout pattern would encounter heavily contaminated water supplies and would have to rely on stored water. Surface water supplies would be directly contaminated by fission products.

During the first few months in areas extending several hundred kilometers downwind of an explosion, the dust, smoke, and radioactivity could cause severe water pollution in surface waters. The dominant fission product during this time would be 131I (iodine-131). Beyond a few months, the dominant fission product in solution would be 90Sr (strontium-90) (Naidu, 1984). Many of the fission products would remain fixed in fallout dust, river and lake sediments, and soils. In rural areas, intermediate- and long-term fallout would pollute water supplies to a lesser extent than the city and suburban supplies. In the absence of additional contamination from runoff, lakes, reservoirs, and rivers would gradually become less contaminated as water flowed through the system.

Initially groundwater supplies would remain unpolluted but they may be difficult to tap. Eventually, however, some groundwater could become contaminated, and remain so for some tens of years after a nuclear war. It would take hundreds or thousands of years for an aquifer to become pure (or nearly so) (van der Heijde, 1985). Doses from drinking this water would be small but, nonetheless, possibly above current water quality standards. In the long term, 90Sr and 137Cs (cesium-137) would be the major radionuclides affecting fresh water supplies.

The GLODEP2 fractional deposition rates have been used to calculate 90Sr surface concentrations. The results are given in Table 7 for the Northern Hemisphere winter and summer seasons. The values are based on the Knox (1983) 5,300-Mt baseline scenario A, and are expressed in mCi/ km2 for a 6-year period over 20° latitude bands. The maximum deposition occurs between 30 and 70°N. The concurrent deposition values for 137Cs can be obtained by multiplying the 90Sr values by 1.6. These values assume an unperturbed atmosphere. As stated earlier, introducing smoke and soot into the troposphere and stratosphere would probably slightly reduce Northern Hemisphere values and slightly increase 90Sr deposition in the Southern Hemisphere.

Table 7. Average Accumulated 90Sr Deposition (mCi/km2) after 6 Years, as a Function of Latitude.

Table 7

Average Accumulated 90Sr Deposition (mCi/km2) after 6 Years, as a Function of Latitude.

Significant doses to individual human organs can also arise from specific radionuclides via food pathways. Such doses are caused by consumption of radioactively contaminated milk, meat, fish, vegetables, grains, and other foods. For a normal atmosphere, various researchers (International Commission on Radiological Protection Publication 30, 1980; Kocher, 1979; Ng, 1982; Lee and Strope, 1974) have provided means to calculate organ doses for a number of radionuclides and food pathways. However, in a post-nuclear-war atmosphere perturbed by large quantities of smoke, the results of the above studies may not be valid since the dose (in rads/ Ci) from soil to animal feed to humans are highly variable geographically and depend upon the degree of perturbation of weather and ecosystems.

However, the internal total body dose (the sum of the dose to each organ weighted by the risk factor due to consumption of various foods) has been very roughly estimated by J. Rotblat (private communication) to be about 20 percent of the external dose from local fallout, about equivalent for intermediate-time-scale fallout, and somewhat greater than the external dose from long-term fallout. These estimates are very uncertain.


For radionuclides, the most important short-term consequence is the downwind fallout during the first few days of relatively large radioactive particles lofted by surface explosions. The deposition of fresh radioactive material in natural and induced precipitation events also could contribute to enhanced surface dose rates over very limited areas (hotspots) both near to and far away from detonation sites. For both local fallout and distant hotspots, dose rates can be high enough to induce major short- and long-term biological and ecological consequences.

Calculations of local fallout fields were performed using the KDFOC2 model and an escalating nuclear exchange scenario. In this illustrative example, where simple assumptions are made about the overlap of fallout plumes, these estimates indicate that about 7 percent of the land surface in the United States, Europe, and the USSR would be covered by external gamma-ray doses exceeding 450 tads in 48 hours, assuming a protection factor of 1 (i.e., no protective action is taken). A similar area estimate is obtained for doses exceeding 600 tads in 2 weeks. More realistic overlap calculations would suggest that these areas could be greater (by a factor of 3 in one specific case). For those survivors protected from radiation by structures, these areas would be considerably reduced. Areas of sub-lethal debilitating exposure ( rads in 48 hours), however, would be larger. A good approximation is that these areas are inversely proportional to the 48-hour dose. In local fallout fields of limited area, the dose from beta rays could be high enough to significantly affect surviving biota. Variations in fallout patterns in the local fallout zones could range over orders of magnitude. If large populations could be mobilized to move from highly radioactive zones or take substantial protective measures, the impact of fallout on humans could be greatly reduced.

The uncertainties in these calculations of local fallout could be factors of several. In addition, the use of different scenarios (e.g., all surface bursting or little surface bursting of weapons) could modify the calculated lethal areas by factors of several. There are a number of other factors that could change these local fallout assessments. Fission fractions of smaller modem weapons could be twice the baseline assumption of 0.5. Adding these to the scenario mix could increase lethal fallout areas by about 20 percent of the baseline calculation. Tactical weapons, ignored in the baseline scenario, could increase lethal local fallout areas in certain geographical regions, particularly within western Europe, by up to 20 percent of the baseline scenario. Internal radiation exposure could increase the average total doses to humans by up to 20 percent of the external dose. Targeting of nuclear fuel cycle facilities could contribute to radiation doses.

For global fallout, different computer models and scenarios have been intercompared. The calculations predict that the 50-year unsheltered, unweathered average external total body gamma-ray dose levels in the Northern Hemisphere would be about 10 to 20 rads, and about 0.5 to 1 rad in the Southern Hemisphere. The peak doses of 20 to 60 rads appear in the 30 to 50°N latitude band. Values predicted for the global population dose using the assumptions made in this study are typically about 6 × 1010 person-rads. The doses in the maxima grid points using a 10° × 10° latitude and longitude mesh size, are a factor of 6 to 8 higher than the Northern Hemisphere averages. From 50 to 75 percent of the global fallout dose would be due to the tropospheric injection of radionuclides that are deposited in the first month. These results were obtained assuming a normal (unperturbed) atmosphere and have an estimated confidence level of a factor of 2 for a given scenario. The most sensitive parameter that affects global fallout levels is the scenario (e.g., total yield, yield mix, surface or airburst, burst locations).

Additional calculations involving a perturbed atmosphere indicate that the above dose assessments would be about 15 percent lower in the Northern Hemisphere and marginally higher (to approximately 1 rad) in the Southern Hemisphere compared to predictions for the unperturbed atmosphere. These results are consistent with the projection that smoke injections can increase vertical stability, inhibit precipitation, and increase interhemispheric transport.

Estimates of dose contributions from food pathways are much more tenuous. Rotblat (private communication) has estimated roughly that internal doses would be about 20 percent of the external dose from local fallout, about equivalent to the external dose from intermediate fallout, and somewhat greater than the external dose from long-term global fallout.

Radioactivity From Nuclear Fuel Cycle Facilities

The possible targeting with nuclear warheads of nuclear fuel cycle facilities arouses considerable controversy. There is general agreement that enormous reservoirs of long-lived radionuclides exist in reactor cores, spent fuel rods, fuel reprocessing plants, and radioactive waste storage facilities. Disagreement arises when the feasibility and extent of such a targeting strategy are considered. Even if one adopts the view that ''what if" questions must be considered, there is still disagreement over the quantitative treatment of the potential dispersal of the radioactivity contained in these sources. In the present treatment, some of the assumptions regarding radioactivity release are considered highly improbable by a number of researchers. The results, therefore, should not be separated from the assumptions and the large uncertainties associated with these assumptions.


A gigawatt nuclear power plant may be a valuable industrial target in a nuclear war. If a targeting rationale is proposed that the largest possible amount of gross national product be destroyed in an attack on a nation's industry (one measure of the worth of a target to a nation), then large (~1000 MW(e)) nuclear power plants could become priority targets for relatively small ( kt) strategic weapons (Chester and Chester, 1976). In the United States there are about 100 such targets, and worldwide there are about 300. There are also military nuclear reactors and weapons facilities that could be targeted. Since these facilities may be targeted, reactor-generated radioactivity should be considered as part of the potential postattack radiological problem.

Whether the radioactivity contained in a reactor vessel can be dispersed in a manner similar to a weapon's radioactivity is debatable. Nuclear reactor cores are typically surrounded by a 1-m-thick reinforced concrete building that has about a 1-cm-thick inner steel lining, many heavy steel structural elements inside the containment building, and an approximately 10-cm-thick reactor vessel. Inside the reactor vessel are fuel rods and cladding capable of withstanding high temperatures and pressures. For the core radioactivity to be dispersed in the same way' as the weapon's radioactivity, all of these barriers must be breached. The core itself must be at least fragmented, and possibly vaporized, and then entrained into the rising nuclear cloud column along with possibly hundreds of kilotons of fragmented and vaporized dirt and other materials from the crater and nearby structures, including the thick concrete slab that supports the reactor building. Under certain conditions of damage, there is a possibility of a reactor core meltdown resulting in the release of some of the more volatile radionuclides to the local environment. If this were to occur, however, the area of contamination would be relatively small compared to the contamination by a reactor core if it were to be pulverized and lofted by a nuclear explosion.

The primary contributor to the long-term dose at a nuclear power plant would not be the core. The most hazardous radioactivity, when assessing long-term effects ( year after attack), is that held in the spent-fuel ponds, if the reactor has been operating at full power for a few years. Since the spent-fuel storage usually has no containment building nor reactor vessel to be breached, it is much more vulnerable to being lofted by a nuclear weapon than are the core materials. Unless spent fuel is located at sufficient distance from a reactor, it could potentially become part of the local fallout problem.

Other nuclear fuel cycle radioactivity may also be significant. Reprocessing plants, although not as immediately important economically as power plants, contain a great deal of radioactivity that could significantly contribute to the long-term doses. Also, military nuclear reactors developing fissile material and their reprocessing plants might be important wartime targets. They also hold significant amounts of radioactivity in their waste ponds and reactor cores.

Military ships fueled by nuclear power could be prime targets as well. Ships' reactors typically produce less power (~60-250 MW(t) than commercial reactors (Ambio Advisors, 1982). They could, however, have substantially radioactive cores, depending on the megawatt hours of service a shipboard reactor has produced since refueling. A large nuclear-powered ship with more than one reactor, designed for years of service without refueling, can have nearly as much long-lived radioactivity (e.g., 9°Sr) on board as an operating commercial reactor (Rickover, 1980). Such shipboard reactors may also be more vulnerable to vaporization than commercial reactors.

Figure 7 shows the gamma radiation dose rate-area integrals from a 1-Mt, all-fission nuclear weapon and from possible commercial fuel cycle facilities. In the first few days, the higher activity of the nuclear weapon debris dominates over the gamma radiation of the reactor. Likewise, gamma-radiation levels from a light water reactor (LWR) are greater than those of 10 years worth of stored spent fuel for about 1 year after the detonation.

Figure 7. Gamma-ray dose rate-area integral versus time after shutdown or detonation (Chester and Chester, 1976, p.

Figure 7

Gamma-ray dose rate-area integral versus time after shutdown or detonation (Chester and Chester, 1976, p. 333). Copyright 1986 by the American Nuclear Society, LaGrange Park, Illinois.

Subsequently, the spent fuel would be relatively more radioactive. Similarly, the gamma radiation from 10 years of spent fuel is greater than the radioactivity of a 1-Mt fission weapon after about 2 months because of the greater abundance of long-lived gamma emitters in the spent fuel.

Thus, for doses from a 1-Mt, all-fission weapon detonated on a reactor, the core gamma radiation would be comparable to the weapon's radiation at about 5 days. By 2 months the gamma radioactivity from the weapon would have decayed by a factor of over 1,000 from its value at 1 hour. Beyond about 1 year the gamma radiation from the weapon is insignificant compared to a reactor's radiation; however, the dose levels are no longer acutely life threatening.

Local Fallout

For dose estimates from local fallout, two time frames are considered—the short term, where there is acute and potentially lethal radiation, and the long term, when chronic doses become important. In the short term, the gamma radiation is the main hazard. Later, specific radionuclides become important concerns for doses via food pathways.

For doses received within the first 48 hours, the nuclear weapon gamma radiation pathway for a high-yield (~1-Mt) warhead dominates the fuel-cycle gamma radioactivity, even if one assumes a worst case assumption in which all the radioactivity from the attacked nuclear fuel cycle facility is lofted with the weapon products. For lower yields and thermonuclear weapons, the core gamma radiation becomes more important and could be potentially greater than the dose from the nuclear weapons, even at very early times. However, since there are now only approximately 100 nuclear power plants available for targeting in the United States and possibly a few hundred shipboard reactor targets which are dispersed over the globe (Ambio Advisors, 1982), and because there are typically more than 1,000 other U.S. targets in major nuclear-exchange scenarios, the impact of fuel cycle radiation to the total U.S. 48-hour external gamma-ray dose would likely be less than 10 percent.

In the long term, the radioactivity from the core and spent-fuel ponds could be a dominant effect, both around the reactor and at substantial distances downwind. After about 1 year, the products from the nuclear fuel cycle could make a substantial contribution to the total gamma-ray dose fallout patterns over the United States. Certainly, if released, fallout gamma radiation from a large reactor would dominate the dose of a 1-Mt weapon over the long term (see Figure 8).

Figure 8. Contours of 100-rad fallout dose during 1 year's exposure, starting 1 month after the detonation of a 1-Mt bomb (A) and a 1-Mt bomb on a 1-GW(e) nuclear reactor (B).

Figure 8

Contours of 100-rad fallout dose during 1 year's exposure, starting 1 month after the detonation of a 1-Mt bomb (A) and a 1-Mt bomb on a 1-GW(e) nuclear reactor (B). Source: Rotblat (1981). Reprinted with permission from the Stockholm International Peace (more...)

In terms of radiological effects, individual radionuclides (e.g., 90Sr) become more important over the longer time frame than the whole-body gamma radiation. Assuming 50 percent fission weapons, it is possible to have more 90Sr in a single reactor and its spent-fuel pond than that produced in a 1,000-Mt attack. Most of the 90Sr is in the spent-fuel pond and thus could be more easily lofted as fallout than the 90Sr in the heavily shielded reactor core. Accordingly, in the long term, the fuel cycle 90Sr contribution can dominate over the weapon contribution. For example, Chester and Chester (1976) calculated levels of 90Sr much higher than the current maximum permissible concentration (MPC) over much of the U.S. farmland 1 year after an attack on the projected nuclear power industry of the year 2000. Scaling down their results to an attack on a 100-MW(e) nuclear power industry, they calculated that about 60 percent of the U.S. grain-growing capacity would be in areas that exceed current 90Sr MPC levels.

Global Fallout

In calculation of the potential global fallout, assumptions have been made that facilitated calculations and allowed estimation of expected dose. For example, it was assumed that each nuclear facility would be surface targeted by a high-yield, accurately delivered warhead that would completely pulverize and vaporize all of the nuclear materials and that these materials would then follow the same pathways as the weapon materials (a worst-case assumption). It was assumed further that the major nuclear facilities in a 100-GW(e) civilian nuclear power industry would also be attacked. The results should be viewed as providing estimates that approach maximum global fallout for an attack on a commercial nuclear power industry of 100 GW(e). Higher estimates would be obtained, however, using the same assumptions by including military nuclear facilities and a larger civilian industry.

This hypothetical reactor attack scenario assumed that, as part of the 5,300-Mt exchange of Knox (1983), some of the warheads would be targeted on nuclear power facilities. Specifically 0.9-Mt weapons would be surface burst on 100 light water reactors, 100 10-year spent-fuel storage (SFS) facilities, and one fuel reprocessing plant (FRP). With a 0.9-Mt surface burst on each facility, 2 percent of the radioactive fission products would be injected into the troposphere and 48 percent into the stratosphere. The remaining activity (50 percent) would contribute to local fallout. Such large yields were assumed because of the hardness of the nuclear reactor. If smaller-yield weapons were used to target the nuclear facilities, the relative injections of radioactivity into the troposphere would be much greater. While the weapons radioactivity would result in higher doses on the ground, this would not be true for the nuclear facility radioactivity. This is because of the relatively slow decay of the facilities' radioactivity. Hence, a faster deposition time would not significantly affect the 50-year dose. The patterns and local concentrations of fallout deposition would, however, be affected.

Using GLODEP2 and a Northern Hemisphere winter scenario, the resulting unsheltered, unweathered doses are shown in Table 8. The largest value of 95 rads for the total of weapons plus the nuclear power industry occurred in the 30-50°N latitude band. The doses obtained for the Southern Hemisphere were about a factor of 30 smaller than in the Northern Hemisphere. The majority of the dose contributions came from the spent-fuel storage facilities and the high level waste in the reprocessing plant.

Table 8. Fifty-Year External Gamma-Ray Global Fallout Dose, in Rads, for Nine Latitude Bands.

Table 8

Fifty-Year External Gamma-Ray Global Fallout Dose, in Rads, for Nine Latitude Bands.

Figure 9 is a plot of accumulated dose in the 30-50°N latitude band as a function of time to 50 years (200 quarter years) for the 5,300-Mt scenario (Northern Hemisphere winter injection) with and without the targeting of nuclear power facilities. The bulk of the dose from the weapons alone for this scenario resulted from deposition in the first year. The relative contributions of the nuclear facilities were minimal in the first year, but became larger with time. At 50 years, the contribution of the nuclear facilities would be approximately double that of the weapons alone. In addition, while the weapons-only curve at 50 years is almost flat, the nuclear facilities curve has a positive slope with the radioactivity continuing to directly affect future generations.

Figure 9. Accumulated dose at 30-50°N versus time scenario A, with (A2) and without (A1) an attack on nuclear facilities.

Figure 9

Accumulated dose at 30-50°N versus time scenario A, with (A2) and without (A1) an attack on nuclear facilities. Source: Pittock et al. (1985, p. 273). Reprinted with permission from the Scientific Committee on Problems of the Environment (SCOPE). (more...)

An attack on all of the world's civilian nuclear fuel cycle facilities (approximately 300 GW[e]) would scale the above results up by about a factor of 3, although this scenario is even less likely. The potential effect is growing in time; the world's nuclear capacity has been projected to grow to 500 GW(e) by 1995. A significant contribution could also come from the targeting of military nuclear facilities, with results qualitatively similar to those obtained from attacks on power plants.

In summary, using some worst-case assumptions for a speculative nuclear war scenario wherein 100 GW(e) of the nuclear power industry is included in the target list, the 50-year global fallout dose is estimated to increase by a factor of 3 over similar estimates wherein nuclear power facilities are not attacked.

If one adds the internal doses necessarily accompanying the external doses (perhaps doubling or tripling the latter) and considers that localized hotspots can be formed with up to 10 times the average dose, it seems that moderate to heavy attacks on civilian and military nuclear facilities could result in significant long-term radiological problems for humans and ecosystems. Many of these problems involving the radiological assessments associated with nuclear facilities are unresolved and uncertain but deserve more thorough attention.

Appendix: The Impact Of Fallout On Humans

In the main body of this paper the focus was an estimation of unprotected doses due to fallout. The focus of the SCOPE-ENUWAR fallout calculations (Pittock et al., 1985) was on assessing the impact on nonhuman biota; direct effects on humans was specifically excluded. Hence, the calculations made were predictions of the unprotected dose, and it is these that have been reported on earlier in this paper. Here, we are more concerned with direct effects of fallout on humans. Consequently, this appendix extends our previous discussion of unprotected doses to focus on the latter subject. We begin with a short discussion about the impact of global fallout on humans. The remainder of this appendix discusses the more serious impact of local fallout.

Global Fallout

As we have reported above, our GLODEP2 calculations for strategic nuclear exchanges of about 5,000 to 6,000 Mt predict that the 50-year unsheltered, unweathered, external total body gamma-ray dose levels average about 15 rads in the Northern Hemisphere and about 0.5 rads in the Southern Hemisphere. The maximum longitude-averaged dose of 30 to 40 rads appears in the 30 to 50°N latitude band. Values predicted for the global population (chronic) dose are typically about 6 × 1010 person-rads. The dose in rainout hotspots, obtained by using 10° latitude and longitude areas, are a factor of 6 to 8 higher than the Northern Hemisphere averages, or 90 to 120 rads, respectively. These results have an estimated confidence level of a factor of 2. From 50 to 75 percent of the global fallout dose is due to tropospheric injections of radionuclides that are deposited in the first month.

Additional calculations, utilizing GRANTOUR and assuming a perturbed nuclear winter atmosphere, indicate that the above dose assessments would be about 15 percent lower in the Northern Hemisphere and marginally higher (to approximately 1 rad) in the Southern Hemisphere than in an unperturbed atmosphere.

These calculations have been presented at a number of scientific meetings, including the ICSU-SCOPE workshop on radiation held in Paris, October 1984. There, internationally known radiation experts carefully reviewed this work, which subsequently became the basis of the chapter on radioactivity in the SCOPE-ENUWAR report (Pittock et al., 1985).

For radiation exposure that is protracted in time, biological repair of the resulting damage is significant in mitigating the effects. Dose effectiveness factors from 0.1 to 0.5 for chronic exposures have been suggested (National Council on Radiological Protection Report 64, 1980). This means that a large chronic dose will have an effect equivalent to a much smaller acute dose. This phenomenon has particular relevance here in assessing the impact of global fallout, which is chronic, low-dose-rate irradiation received over many decades.

The effects of the above levels of global fallout, even including the hotspots referred to earlier, were summarized in the Report of the Pads Commission on Radiological Dose Assessments and Biological Effects (SCOPE-ENUWAR Newsletter, 1984). It concluded that "the long-term increase in genetic and carcinogenic effects on humans from global fallout is of the order of 1% of the natural incidence and should be considered a second order effect." No mention was made of prodromal effects on humans because at these lifetime (50 years) dose levels, and assuming biological repair mechanisms, prodromal effects would not be observed. This result is far from that pictured in the On the Beach syndrome.

Local Fallout

As we have seen, projections of the intensity and extent of local fallout are highly sensitive to a number of variables, which helps explain why many assessments have produced widely different results. Uncertainties in these projections can be divided into three categories: those due to the targeting scenario, the fallout calculations model, and the selected meteorological conditions.

The targeting scenario contains variables such as the number of weapons and their yield mix, fission fractions, heights of burst, and precise target locations. The height of burst (HOB) is of particular significance because airbursts do not produce significant local fallout, except for rainout of debris from tactical yield weapons. Only when the fireball interacts with the ground (a ground or near-ground burst) does significant local fallout ensue. A widely used and reasonable assumption is that hardened military targets are targeted with ground bursts. For the softer industrial and other military targets, maximum damage is accomplished by airbursts where the HOB can be optimized. The fires hypothesized in urban areas in nuclear winter studies are assumed to be initiated by airbursts since ground bursts are not efficient in initiating large fires. Uncertainties in dose calculations in the best fallout models originate from several sources. These include limited experimental data, whether the modeled radioactivity is rigorously conserved, whether time of arrival is properly accounted for, and other inaccuracies of the model. Assumptions about selected meteorology (e.g., wind velocities, shears, precipitation patterns) affect the results. Hence, local fallout assessments can vary greatly, depending on these many assumptions.

For assessing the impact of local fallout on humans, additional factors must be considered. By far the most sensitive of these is the protection factor afforded by homes, buildings, basements, and other shelters. These structures can dramatically mitigate the unprotected dose assessments normally cited and used previously in this paper. In Table 9, structure protection factors from fallout gamma rays are listed.

Table 9. Fallout Gamma-Ray Dose Protection Factors for Various Structures.

Table 9

Fallout Gamma-Ray Dose Protection Factors for Various Structures.

An additional important consideration for humans is the assumption of what are the lethal acute external whole-body dose levels (50 percent lethal dose [LD50] values from 220 to 600 rads of external gamma radiation have been reported). Finally, for local fallout delivered over days and weeks, biological repair will reduce the damage from the dose by a significant factor, vis-à-vis an instantaneously delivered dose by a significant amount.

Our calculations of the total fatalities produced by large-scale attacks on the continental United States have produced estimates of fallout fatalities (after subtracting those already killed by blast and thermal effects) that range over almost 2 orders of magnitude. This large variation in fallout fatalities is well understood in terms of variations in the parameters discussed above.

In one study, fallout fatalities resulting from a massive countervalue attack of 1,000 Mt against U.S. urban population centers was estimated (Harvey, 1982). The scenario contained 1,000 surface-burst warheads each of 1 Mt, 50 percent fission yield. This population-destroying scenario was not purported to be realistic; rather, it was part of a parameter study to estimate the effects of evacuation and/or sheltering on fatality estimates. In this study, we used realistic overlap of fallout from multiple weapon bursts, the U.S. Census Bureau population distribution, and a probability of death from fallout with 500 rads received in 1 week with no sheltering. The total number of fatalities was estimated at about 160 million, of which 16 million were attributed to fallout. This study illustrated the great sensitivity of fallout fatalities to the choice of parameters.

Physicians are more concerned with nonfatal injuries. Radiation effects become apparent in humans with acute doses greater than about 100 rads. We can estimate the extent of the areas that are covered with a minimum dose by referring to Figure 2. The slope of the 48-hour dose versus area curves for strategic-sized weapons yield are approximately -1, meaning that the minimum dose area contours are inversely proportional to the 48-hour dose. As an example, the SCOPE-ENUWAR study reported that about 7 percent of the land masses of the United States, the USSR, and Europe would receive a minimum of 450 rads within 48 hours. The figure for the continental United States was about 8 percent. This result assumed no shielding and applied to an unsheltered population. Our inverse approximation would then project that the area covered by a minimum dose of 100 rads would be 4.5 times larger, or 36 percent of the total land area of the contiguous United States.

However, minimum dose contours over land areas do not relate in a simple manner to human exposure. Here, both protection factor distributions and population distributions must be considered to make a proper assessment.

In summary, global fallout is not expected to result in prodromal symptoms from radiation exposure because of both the magnitude of exposures and the chronic (long-term) exposure rate. Global fallout would result in a small statistical increase, of the order of 1 or 2 percent, above the current incidence of cancers and genetic mutations in the decades following the occurrence of a nuclear war. Local fallout can produce significant numbers of injuries and fatalities from radiation exposure, but numerical estimates are highly uncertain and are very sensitive to the assumptions made to obtain these estimates. Attempts to make these assessments as realistic as possible by including credible population distributions (relocated and/or sheltered) should be made. Superficial attempts at reality will yield an artificially large spread in the results.


This work was performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under Contract No. W-7405-Eng-48.


  • Ambio Advisors. 1982. Reference scenario: How a nuclear war might be fought. Ambio 11:94-99.
  • Chester, C. V., and R. O. Chester. 1976. Civil defense implications of the U.S. nuclear power industry during a large nuclear war in the year 2000. Nuclear Technol. 31:326-338.
  • Defense Civil Preparedness Agency (DCPA). 1973. Response to DCPA questions on fallout. DCPA Research Report No. 20, November 1973. Washington, D.C.: U.S. Defense Civil Preparedness Agency.
  • Edwards, L. L., T. F. Harvey, and K. R. Peterson. 1984. GLODEP2: A computer model for estimating gamma dose due to worldwide fallout of radioactive debris. Report UCID-20033. Livermore, Calif.: Lawrence Livermore National Laboratory.
  • Fetter, S. A., and K. Tsipis. 1981. Catastrophic release of radioactivity. Sci. Amer. 244(4): 41.
  • Glasstone, S., and P. Dolan. 1977. The Effects of Nuclear Weapons. Washington, D.C.: U.S. Department of Defense and U.S. Energy Research and Development Administration.
  • Harvey, T. F. 1982. Influence of civil defense on strategic countervalue fatalities. Report UCID-19370. Livermore, Calif.: Lawrence Livermore National Laboratory.
  • Harvey, T. F., and F. J. D. Serduke. 1979. Fallout model for system studies. Report UCRL-52858. Livermore, Calif.: Lawrence Livermore National Laboratory.
  • ICRP Publication 30. 1980. Limits for Intakes of Radionuclides by Workers. New York: Pergamon.
  • Knox, J. B. 1983. Global scale deposition of radioactivity from a large scale exchange. Proceedings of the International Conference on Nuclear War, 3rd Session: The Technical Basis for Peace, Erice, Sicily, Italy, August 19-24, 1983. Servizio Documentazione dei Laboratori Frascati dell'INFN, July 1984, pp.29-46. Also Report UCRL-89907. Livermore, Calif.: Lawrence Livermore National Laboratory.
  • Kocher, D.C. 1979. Dose-rate conversion factors for external exposures to photon and electron radiation from radionuclides occurring in routine releases from nuclear fuel cycle facilities. Health Phys. 38:543-621. [PubMed: 7410079]
  • Lee, H., and W. E. Strope. 1974. Assessment and control of the transoceanic fallout threat. Report EGU 2981. Menlo Park, Calif.: Stanford Research Institute.
  • MacCracken, M. C., and J. J. Walton. 1984. The effects of interactive transport and scavenging of smoke on the calculated temperature change resulting from large amounts of smoke. Proceedings of the International Seminar on Nuclear War 4th Session: The Nuclear Winter and the New Defense Systems: Problems and Perspectives, Erice, Italy, August 19-24, 1984. In preparation. Also Report UCRL-91446. Livermore, Calif.: Lawrence Livermore National Laboratory.
  • Naidu, J. R. 1984. Impact on water supplies—II. SCOPE/ENUWAR meeting, New Delhi, India, February 1984. Draft manuscript.
  • National Council on Radiation Protection and Measurement. April 1, 1980. Influence of dose and its distribution in time on dose-response relationships for low-LET radiations. Washington, D.C.: U.S. Government Printing Office. (Report No. 64.)
  • National Research Council (NRC). 1985. The Effects on the Atmosphere of a Major Nuclear Exchange. Washington, D.C.: National Academy Press.
  • Ng, Y. C., C. S. Colsher, and S. E. Thompson. 1982. Transfer coefficients for assessing the dose to man from radionuclides in meat and eggs. Lawrence Livermore National Laboratory Report NUREG/CR-2976.
  • Pittock, A. B., T. A. Ackerman, P. Crutzen, M. MacCracken, C. S. Shapiro, and R. P. Turco. 1985. Environmental Consequences of Nuclear War. Volume I. Physical and Atmospheric Effects. SCOPE 28. Chichester, U.K.: John Wiley & Sons.
  • Rickover, H. G. 1980. Naval nuclear propulsion program—1980. Statement before the Procurement and Military Nuclear Systems Subcommittee, 96th Cong. Washington, D.C.: U.S. Government Printing Office.
  • Rotblat, J. 1981. Nuclear radiation in warfare. Stockholm International Peace Research Institute (SIPRI). London: Taylor and Francis.
  • Schlesinger, M. E., and W. L. Gates. 1980. The January and July performance of the OSU two-level atmospheric general circulation model. J. Atmos. Sci. 37:667-670.
  • SCOPE-ENUWAR Newsletter. 1984. Scientific Committee on Problems of the Environment. University of Essex, England.
  • Shapiro, C. S. 1984. Scenario and parameter studies on global deposition of radioactivity using the computer model GLODEP2. Lawrence Livermore National Laboratory Report UCLD-20548.
  • Svirezhev, Y. M. 1985. Long-term ecological consequences of a nuclear war: Global ecological disaster. Moscow: Computer Center of USSR Academy of Sciences. Draft manuscript.
  • Turco, R. P., O. B. Toon, T. P. Ackerman, J. B. Pollack, and C. Sagan. 1983. Nuclear winter: Global consequences of multiple nuclear explosions. Science 222:1283-1292. [PubMed: 17773320]
  • van der Heijde, P. K. M. 1985. Groundwater contamination following a nuclear exchange, SCOPE/ENUWAR Workshop Report. Delft, The Netherlands, October 3-5, 1984.
Copyright © 1986 by the National Academy of Sciences.
Bookshelf ID: NBK219147


Related information

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...