The Effects of Nuclear War

CASE 3: A COUNTERFORCE ATTACK AGAINST THE UNITED STATES

Prompt Effects

The blast damage from a counterforce attack is concentrated on military installations. Attacks on submarine bases and bomber bases would cause considerable blast damage to nearby populations and urban structures; attacks on silos would cause relatively little civilian blast damage. Unlike ICBM silos, many bomber bases and fleet ballistic missile submarine (SSBN) support facilities are near cities. (See figure 15.) For example, an attack on Griffiss Air Force Base, near Utica and Rome, N. Y., would place nearly 200,000 people at risk from prompt effects; attacking the SSBN support facility near Charleston, S. C., would place more than 200,000 people at risk; attacking Mather Air Force Base, near Sacramento, Calif., would place more than 600,000 people at risk. The additional attacks would simultaneously reduce the number of people able to provide aid and increase the number of injured or evacuees re quiring aid. The attacks would make it harder for people able to provide aid to sustain those needing it.

Countersilo attacks would probably detonate some weapons at or near the Earth’s surface to maximize the likelihood of destroying ICBM silos. Surface bursts produce intense fallout, causing most of the damage to the civilian population, economy, and society. The principal civilian impact of adding attacks on bomber and SSBN bases is the large increase in urban destruction.

The Period Before Fallout Deposition

Fallout would begin to reach closer populated areas in a few hours; it would reach many others in a few days. As fallout arrives, radiation levels rise sharply and rapidly. People would therefore have to take any protective actions —shelter or evacuation — before the fallout arrives. This prearrival period would thus be one of intense activity and intense confusion. How would people react? Training could help, but people trained in how to behave under fallout conditions would fare poorly if they could not get to shelters or if shelters were unstocked. To what extent would people panic, seek other family members, or evacuate spontaneously, and what would be the consequences of such actions?

Evacuation would probably be a poor choice, since it would be difficult or impossible to predict which would be the safe areas and which the hot spots, and since a car in a traffic jam would offer poor shelter indeed. The decision on whether or not to evacuate, however, is complicated because evacuation is a reasonable response for people who would be at risk from blast from further attacks even though evacuation is a poor strategy for people at risk from fallout alone.

Shelter would in theory be available to a majority of people, although the best available shelter might not be good enough in areas where the fallout proved to be very intense. However, the practical difficulties of fallout sheltering could be very great. The time to seek shelter could be very limited (and people would not know how long they had), and people would want to get their families together first. A shelter must have a sufficient protection factor. Fallout particles must be kept out of the shelter, which requires a ventilation system more complicated than an open window or door, and if anybody enters a shelter after fallout has fallen there must be some means of decontaminating the new arrival. Water is necessary; heat may be necessary depending on the time of year; sanitation is a problem. Finally, people could not tell how long it was necessary to stay in the shelter without radiation rate meters.

It is obvious that the time of day, the time of the year, and the degree of emergency preparations during the hours or days before the attack would all affect the level of deaths. Whatever the circumstances, the few hours after the attack would see a frantic effort to seek shelter on the part of most of the American population. Then, in densities and locations determined by the attack parameters and the weather, the fallout would descend. Many Americans would be lucky enough to be in areas where the fallout level was low. Many others (between an estimated 2 million and 20 million), would be caught without shelter, or with inadequate shelter, and would die. Still others would suffer from a degree of radiation that would make them sick, or at least lower their life expectancy, but would not kill them. The trials of living in fallout shelters would be intensified by the fact that many people would not know which category they and their families were in.

A comprehensive counterforce attack would impose a greater burden than a countersilo attack. Many more people would be injured by prompt effects, and people near bomber and SSBN bases would have only a few minutes warning in which to seek shelter.

Cities in the blast area –those near SSBN or bomber bases–would be heavily damaged. A few cities, such as Charleston, SC., and Little Rock, Ark., could suffer consequences similar to Detroit in Case 1 (chapter 11) or Philadelphia in Case 2 (above in this chapter); most would not. People in blast areas would face hazards as noted in Case 1 — injuries from blast, initial nuclear radiation, and thermal radiation, and from such secondary effects as falling buildings and fires. As in other cases, rescue would be difficult, with streets blocked by rubble, water pressure gone, and emergency vehicles destroyed.

People in areas damaged by blast and in the path of fallout would be in greatest peril. injuries, damage to prospective shelters, damage to transportation, and damage to power and water could make them highly vuInerable. Little Rock, Ark., for example, the site of an ICBM base and a bomber base, would receive both blast damage from a pattern attack (designed to destroy bombers in flight) and intense fallout radiation from the attack on ICBMs.

People in areas neither damaged by blast nor threatened by fallout would believe themselves to be at risk from blast or, at a minimum, from fallout until it was clear that attacks had ended. To these people would fall the burdens of producing necessities and caring for the injured and evacuees. Yet people in these areas, believing themselves to be at risk, would feel compelled to seek shelter or, especially in unattached cities, to evacuate spontaneously. These actions would reduce the flow of aid to damaged areas. Indeed, the economy would probably shut down until people were certain that the war had ended and until most people could get back to work, probably until the end of the shelter period. Even if some people reported to work, production would be difficult with many absentees. There would be large credit, monetary, contractual, and legal problems. If production stopped even for a week, the loss would be tremendous. This attack would disrupt the economy less than Case 2, however, because most productive resources would remain intact.

Casualty Estimates

In seeking to estimate prompt damage from the attacks, fatalities are the most important component of damage and the most calculable. To estimate fatalities, the critical questions are which areas would be damaged by blast, and to what extent? How much fallout would there be, and where would it be deposited? These questions cannot be answered with great confidence because estimates of deaths from these attacks are highly sensitive to attack parameters and civilian shelter assumptions. However, reference can be made to several recent executive branch studies of counterforce attacks.

OTA drew on studies, conducted between 1974 and 1978, of counterforce attacks. These studies differed widely in their results, primarily because of differences in the assumptions they made. OTA felt that it would be more useful to look at the ways in which these assumptions affect the results than to attempt to determine the “correct” assumption for each uncertainty. Consequently, a range of results is presented; it is believed that if OTA had done a new study of this case the results would have fallen somewhere within this range. ⁵

The executive branch countersilo studies that OTA drew on indicated that between 2 million and 20 million Americans would die within the first 30 days after an attack on U.S. ICBM silos. This range of results is so wide because of the extent of the uncertainties surrounding fallout. The key uncertainties are:

• Height of Burst.– If the fireball touches the ground, it vaporizes some dirt, irradiates it, and draws it up into the mushroom cloud. This material condenses to become fallout. The lower the height of burst, the more of the fireball touches the ground, and the more fallout that is produced. An air burst in which none of the fireball touches the ground creates negligible fallout. Because ICBM silos are very hard, a surface burst offers the greatest probability of destroying the silo with one explosion; it also maximizes fallout. The probability of destroying an ICBM silo is increased if two warheads are targeted against it; opinions differ as to whether the most effective tactic is to use two surface bursts, which doubles the amount of fallout, or one air burst and one surface burst.
• Weapon Design.— Some weapons derive a greater portion of their energy from fission (as opposed to fusion) than others; the more fission, the more fallout. The weapon yield affects the amount of fallout; the higher the yield of a given explosion, the greater the fallout.
• Wind.– The speed and direction of the wind at various altitudes determines the directions and distance from the explosion at which fallout is deposited, and influences fallout concentration. Winds typically vary with the season; indeed, this variance is so great that it can affect casualties by about a factor of three, as figure 16 shows. The hourly and daily variation of winds also affects casualties. It is important to bear in mind, when considering possible civil defense measures, that winds could not be accurately predicted even after an attack had taken place, much less in advance.
• Rain. – Raindrops collect fallout particles from the radioactive cloud, thereby creating areas of intense fallout where it is raining, and reducing fallout elsewhere.
• Terrain. — Hills, buildings, and ground temperature gradients (such as are caused by highways and small lakes) affect the exact pattern of fallout, creating hot spots in some places and relatively uncontaminated spots nearby.
• Distance.—Other things remaining constant, fallout decreases with distance from the explosion beyond roughly 50 miles [80 km].

As chapter II explained, radiation from fallout in large doses causes death, in smaller doses causes illness, and in still smaller doses creates a probability of eventual illness or death (hence, lowers life expectancy). As chapter III explained, protection can be obtained when matter is placed between the fallout and people— in general, the more matter (the greater the mass) between a source of radiation and a person, the greater the protection. The degree of protection offered by various materials is described as a protection factor (PF). The adequacy of a given PF depends on the intensity of the fallout. For example, a PF of 20 (typical of a home basement with earth piled over windows and against the walls) would reduce an outdoor radiation level of 60 rem per hour to an indoor level of 3 rem per hour. In this case, a person outdoors for 10 hours would almost certainly be killed by radiation, and a person in the basement shelter would have a good chance of survival. But if the outdoor level is not 60 rems per hour but 600 rems per hour, a PF of 20 is inadequate to save lives.

Calculations of deaths from fallout are made by combining:

• an assumed distribution of fallout, with various intensities at various locations;
• an assumed distribution of population within the areas where fallout is assumed to be deposited; and
• an assumed distribution of PFs for the population.

Some computer models use a grid (perhaps 4,000 yards on a side for a fine-grained model, but much larger in other cases) and assume that within each square of the grid the fallout intensity and population density are constant, with PFs mixed. Other calculations use regional or nationwide averages. In general, the calculations show lower numbers of deaths when they assume that the population is widely dispersed, and higher numbers when they take into account concentrations of population. The calculations also show lower numbers of deaths when they assume high PFs; in general, increasing PFs above 40 does not reduce casualties much in the calculations, but that does not mean that raising a PF above 40 might not save an individual’s life in reality. The calculations also show lower numbers of deaths when the winds do not blow fallout into densely populated areas.

The studies mentioned previously made separate calculations for attacks including bomber and missile submarine bases, as well as silos. Assuming that there is no preattack evacuation, calculated deaths range from a low of 2 million to a high of 22 million. The differences result primarily from variations in assumptions regarding fallout protection: the high figure assumes approximately to degree of protection which people receive in their daily peacetime lives (PF of 3), and the IO W figure assumes that the entire population moves after the attack to fallout shelters with a PF of at least 25. A more reasonable assumption, that the fallout shelters which now exist are utilized by people living near them, produces a calculation of 14 million dead. The same studies also assessed the effects of extensive preattack evacuation (crisis relocation), and found that it reduced the range of predicted deaths. However, the assumptions regarding fallout protection, both for those who are assumed to evacuate and for those who are assumed to remain near home dominate the results. Further detail is in appendix D.

Given the threat U.S. bombers pose to the Soviet Union, a Soviet preemptive counterforce attack on bomber bases would probably seek to destroy the aircraft and supporting facilities rather than cratering the runways. To destroy airborne bombers launched on warning of attack, an attacker might detonate weapons in a spaced pattern over the base. Airbursting weapons rather than ground-bursting them could reduce the threat of fallout but increase casualties from blast and thermal effects; if the weapons were detonated much above the optimum height of burst for maximizing overpressure on the ground, fallout would be negligible and blast damage would be reduced. The attacks against missile submarine bases are much less complex. Ususally a single high-yield weapon with medium-togood accuracy will destroy docks, piers, cranes, and other facilities — and nearby cities, factories, and people as well.

Accordingly, it is certain that if the only difference between two attacks is that one attacks only ICBM silos and the other attacks bomber and missile submarine bases as well, the latter attack would kill more people. However, the variations in assumptions made about attack design, weather, and fallout protection obscure this. Since these variations reflect genuine uncertainties, it is not possible to determine which set of assumptions and which fatality calculation is most probable. However, some of the extreme assumptions do appear implausible. One Defense Department study notes that its highest fatality figure assumed the use of Soviet weapons larger than those which U.S. intelligence estimates the Soviets possess. Very low fatality estimates assume abnormally low winds, an absence of surface bursts, and /or virtually perfect fallout protection. On balance, it does not appear possible to sustain greater precision than to say that “studies of hypothetical counterforce attacks show deaths ranging from 1 million to 20 million, depending on the assumptions used.” However, the low-end of this range (deaths below the 8 to 10 million level) requires quite optimistic assumptions, while the high end of the range is plausible only on the assumption that the attack is not preceded by a crisis period during which civilians are educated about fallout protection.

The data on injuries contained in the executive branch studies are quite limited; for the counterforce attacks, however, the results suggest that injuries would about equal fatalities.

5. For example, after the OTA analysis, was completed, a new study was completed showing fatalities from a counterforce attack with the current U.S.civil defense posture to be 8 to 12 million without warning, and 5 to 8 million with warning. See Roger Sullivan et al , “Civil Defense Needs of High-Risk Areas of the United States” (Arlington, Va System Planning Corporation, 1979), p. 22

---