August 21, 2002

Richard L. Garwin
Senior Fellow for Science and Technology
Council on Foreign Relations, New York
(914) 945-2555
FAX: (914) 945-4419
Email: [email protected]
Web: http://www.fas.org/rlg

The loss of 3000 Americans to Al Qaeda terrorism September 11, 2001 brought to many the sudden recognition that America was no longer leading a charmed life. Since then, a great deal of hand wringing and discussion has ensued, but the problem is a serious one and won't go away. Not that it was unrecognized and unpublicized. For instance, in 1999 the Commission chaired by former U.S. senators Gary Hart and Warren Rudman reported

"There will...be a greater probability of (catastrophic terrorism) in the next millennium...Future terrorists will probably be even more hierarchically organized, and yet better networked than they are today. This diffuse nature will make them more anonymous, yet their ability to coordinate mass effects on a global basis will increase...Terrorism will appeal to many weak states as an attractive option to blunt the influence of major powers...(but) there will be a greater incidence of as hoc cells and individuals, often moved by religious zeal, seemingly irrational cultist beliefs, or seething resentment...The growing resentment against Western culture and values...is breeding a backlash...Therefore, the United States should assume that it will be a target of terrorist attacks against its homeland using weapons of mass destruction. The United States will be vulnerable to such strikes."

--U.S. Commission on National Security/21st Century, New World Coming: American Security in the 21st Century, September 1999, p. 48

The concept of megaterrorism was well known; the warning was there; only the date, place, and nature of the deed were in question to those who had looked at the prospects.

NOT TERRORISM, BUT GENOCIDE. It is important to recognize that some of the most important threats do not have the object to instill terror in the population, but simply to destroy vast numbers of people and even their accomplishments. The Nazi campaign of extermination of gypsies, homosexuals, and Jews was not oriented toward terror and publicity, but only on the goal of genocide. As will be seen, the empowerment of the individual by modern technology (which increases lethality and which has increased vulnerability), together with the willingness or the desire to die in the act of destruction, can greatly reduce the effectiveness of some of the most important tools of counter terrorism: post-attack attribution and punishment- e.g., by the promise of reward. Nevertheless, such tools are vitally important in other cases and can serve as an important deterrent to terrorist organizations and acts.

This paper is focused largely on the threat and only secondarily on the potential remedies. And it concentrates on vulnerabilities of U.S. society. But this is only by example; other societies are at least as vulnerable, and may find themselves also the target of terrorism and megaterrorism. Furthermore, the solution to terrorism does not lie in actions by the United States alone. The talents and efforts of many throughout the world can reduce the threat of terrorism. In particular, the World Laboratory may provide a mechanism for initiating such contributions to the good of humanity.

Before concentrating on analysis of terrorism and potential solutions, it is a good idea to introduce and define terrorism and its nature. There are many types of terrorists, many types of victims and targets, and several categories of actions against those targets. Thus we consider actors, targets, and actions.

First, the actors-- humans. A robot, an emplaced explosive, or a trained animal can conduct terrorism, but for the near term we count those as mechanisms and the terrorists themselves as human beings. There can be state-sponsored terrorism, individual terrorists with their own agendas, and a whole range between. State-sponsored terrorism might use individuals of the sponsoring state; it is as likely to depend upon other individuals who act for hire, or who are motivated by similar causes as those that impel the state sponsoring the terrorism.

The individual terrorist, such as Ted Kaczynski, can have a range of grudges ranging from strong feelings against abortion to equally strong feelings against the use of animals in research or commerce, to the preservation of trees, to race hatred, to a grudge against a movie star or a president. Such groups or individuals can be loosely affiliated, or isolated. "Copycat" terrorism can spread from one group to another without direct communication between the groups.

Then there can be terrorist organizations, either sponsored by a state or not, more or less closely organized to carry out a campaign which if not dictated is at least outlined from the top.

There are distinctions among actors-- emphasized most recently to the American consciousness by the fact that some individuals don't mind or even prize losing their lives in carrying our terrorist acts. This should be no surprise, since suicide bombers in Israel have been doing this for many years. The professional anti-terrorism community for decades observed that not only did terrorists not wish to die, but also they really did not wish to kill very many people. Their purpose was to bring their cause to world attention in a favorable light, and that does not occur by victimizing and killing large numbers of innocent civilians. The terrorists and their cause must appear to be the victims. The professionals were largely correct-- for that era.

A hired killer will take a risk of death, but it is a rare one who will commit himself or herself to death for the cause-- and then only if the fee for hire goes to the family, for instance. But the Kamikaze pilots of the Zeros in WW II, and suicide bombers in Israel, and the 19 who died (or at least the four who were in charge of the aircraft hijackings on 9/11) show that there is no shortage of people who fit this mold. This is important not only because it shows increased dedication to the cause, but also because it enables means of action that are otherwise outside the capability of many individuals or groups.

Thus, to bring ten kilograms of explosives into a crowded mall and to detonate them at the desired moment where the crowd is most dense while remaining safe, oneself, is more difficult technically than pressing a button to detonate explosives on one's own body.

Now, the targets. If they were governments and government activities, the deed might not be so much terrorism, but, for instance, revolution. An act of one government against another government constitutes an act of war, declared or undeclared. In any case, governments in general have the resources to protect their personnel and facilities better than do corporations and citizens. Attacks on families of government workers, including families of personnel in the armed services are not a direct attack on government facilities or powers. However, they are little different from attacks on the population, selected groups, or on facilities not belonging to the government, so this paper does not discriminate between government facilities and those belonging more generally to the population and infrastructure.

Terrorists may attack infrastructure, such as transportation nodes, water supply facilities, sewage facilities, electrical transmission, and the information or the financial infrastructure. Such would not have large immediate death tolls, but could bring a society to its knees and focus the attention of the population on the immediate problem rather than on economic activity, health, and the survival of the society.

Some terrorist attacks would deny substantial areas to occupancy or even to transit. Among such are the widespread scattering of antipersonnel mines; the dissemination of intense radioactivity; the spreading of anthrax spores or of other durable BW agent. In part, the denial may be self-denial, because the risk of reoccupancy may not be very great. Yet in a society which has felt at peace for a long time, it is difficult to feed into the calculus of everyday life even a small probability of being blown up in the marketplace, or a tiny enhancement in the probability of dying of cancer.

Some terrorist acts, on first thought, would get a substantial amount of moral support from the population attacked. The Internal Revenue Service appears to be a focus of domestic terrorist resentment. But the IRS impacts economic activity a year ahead. An attack on the system by which the government pays its bills would affect commerce and government within days.

At this point, I give my judgments, which are that the world should fear first megaterrorism by biologic agents-- smallpox, anthrax, and the like; and next in line is nuclear megaterrorism-- either from the clandestine transport of a nuclear weapon stolen in Russia, or from the assembly in the United States of an improvised nuclear device (IND) based on high-enriched uranium. More about this when we get to the "actions."

Not all agree on the importance of megaterrorism-- BW or nuclear, or even chemical agents, as compared with small-scale terrorism using explosives, for instance.

One pole of this debate is represented by a particularly well-informed participant:

"The main trend (in international terrorism) is the trend away from state involvement and state sponsorship (compared to) 15 years ago...I believe that the specter of terrorists, especially international terrorists, using chemical, biological, radiological, or nuclear means has...diverted our attention from what...will continue to be the main threat, which is the infliction of loss of life through conventional means...(There were) two attacks on the United States. The one that used box cutters and aircraft hijacking is the one that killed almost 4,000 people; the one that used anthrax spores has so far killed five. We ought to reflect on that; I can assure you the terrorists will reflect on that."

--Paul Pillar, National Intelligence Officer for the Near East and South Asia and former Deputy Director for Counter-Terrorism, CIA, National Journal, January 28, 2002

Whatever the purpose of terrorism, terrorists have limited resources. Such limits were clearly not financial in the case of the damage inflicted by the hijacked aircraft 9/11/2001; there are estimates that the overall direct cost of this activity was $65,000. Be that as it may, if terrorists are limited in lives to be expended, the 9/11/01 events provided the grand benefit-cost ratio of 3000/19. This derives from 3000 dead in the terrorist collisions with the two World Trade Towers and the Pentagon, and 19 terrorists who sacrificed themselves in the process, including those on the aircraft which went down in Pennsylvania because of the heroic actions of the passengers and crew. Although these attacks succeeded-- according to Bin Laden beyond his wildest expectations-- the result is still a benefit-cost ratio of 3000/19, or about 150:1.

From a cold-blooded reckoning of lives taken against terrorist lives lost, this is "only" 150 to 1. However, the impact on American society, economy, and the world was very much larger than would have been the death by accident of 3000 on an ocean liner, which shows that lives lost are far from the only metric on either side.

And so Pillar is right, in part. It is a lot easier to cause deaths of this magnitude by conventional means than by the display of imagination and organization required for the events of 9/11/01. In particular, if terrorists are satisfied with such an exchange ratio, the tried and true means of obtaining it is to bring down aircraft by the use of explosives on board. And if explosives in luggage unaccompanied by the passenger won't do, then an explosive charge secreted on the body (in the shoes!) or in the carry-on luggage of a witting passenger is available. Such passengers would be more likely to fly on heavily laden flights than on the lightly loaded flights of 9/11/01. And the simultaneous destruction of about a dozen U.S. aircraft, planned by Ramsi Ahmed Yousef in 1995, was designed to obtain just this impact. The chill would begin with the air transport system, and have a substantial impact on the economy. But it would not have the impact of the highly visible destruction of the World Trade Center in New York City.


We have already discussed the agent of a hijacked large aircraft, crashing into a building of symbolic significance and/or which houses many people. This is self-induced vulnerability, but one which can hardly be eliminated at this stage of our society, and in any case not without some years of action. But it can be ameliorated-- by strengthened and locked cockpit doors on large passenger aircraft, and by greater control over freight aircraft and private aviation.

My fears-- megaterrorism by means of BW agents or nuclear weapons-- deserve a bit more description.

Biological warfare agents perfected by the major powers in the immediate post-war period include diseases of plants, animals, and humans. They are further divided into diseases that are only infectious, and those that are also communicable, or contagious. A disease such as anthrax will cause disease in its target by infection with the bacterium or virus which is disseminated in the attack. But there will not be significant spread from one victim to another target in the population.

This does not mean, of course, that there is zero probability of contracting anthrax by contact with a person who has the disease. The difference between a contagious disease and one which is infectious, really lies in the magnitude of probability of cross-infection, rather in its existence. If the magnitude is substantially less than one-- say, 10% probability of a secondary case-- the impact of a BW attack with that agent is little increased by the possibility of transfection.

On the other hand, if each primary victim infects (by contagion) three secondary victims, then each secondary victim might be expected to infect another three, so that after about 14 days the number of primary victims is multiplied by three, then again by three after the next 14 days, and so on.

Depending upon the state of immunity in the society, and the planning and organization for countering such an infection or for public health in general, the initial infection of 100,000 people might lead to infection of many tens of millions, and (with a smallpox fatality rate of 30%) the death of 30 million Americans within four months.

Anthrax was weaponized by the United States and a number of other countries in the years before President Nixon's 1969 Executive Order banned offensive work on biological warfare agents, forbidding the development, stockpiling, and of course the use of such agents. The Biological Weapons Convention of 1972 has been signed by 145 states and ratified by almost all of these. It is, of course, far from perfect protection against the use of BW by a state in violation of its pledge, or by terrorists who have no standing.

Anthrax has, for the weaponeer, the desirable feature that the bacterium forms a spore, which is extremely durable not only in storage, but even in the ground for decades. Under appropriate conditions, the spores revert to the vegetative state, and the bacteria then multiply rapidly in the host, as is the nature of bacteria.

The five deaths caused by anthrax-bearing letters-- probably from a single domestic terrorist incited by 9/11-- and a number of illnesses have taught us a lot about inhalation anthrax which we did not know before. Quite apparent, and not surprising, is the fact that the spores may reside for weeks and months in the host, before vegetating-- even in the presence of effective antibiotics, which have no effect on spores-- and if antibiotics are withdrawn from the healthy person, an ensuing infection is then largely lethal within days, unless promptly treated. The likelihood of prompt treatment has, of course, increased greatly, since a disease almost unknown to the ordinary medical practitioner is now at the top of everybody's alert list.

A U.S. Army publication shows a number of the agents weaponized by the United States or other countries, with their LD-50, and their division between morbidity and mortality-- the causing of illness on the one hand, or death on the other2.

Of course, there is some immunity to some of these diseases, but widespread immunity due to smallpox has largely disappeared (especially in the United States) since vaccination stopped in 1972, in the expectation that smallpox would be eliminated in its natural state worldwide by 1980, as was the case. And immunity to smallpox in a well-vaccinated individual is expected to decay after 10 years.

Clearly, if the United States and the world considered it a priority to develop vaccines against all potential BW agents, this would be a useful contribution to the reduction of terrorism.


One might counter terrorism by intervening at any point in the chain, Actors-Actions-Targets. Nothing the U.S. government says or does outside the narrow counter-terrorism field will dissuade all terrorists; but such government behavior may well reduce the number of candidate terrorists by a large factor.

A similar factor is available if access to terrorist targets or to the mechanisms of terrorism is limited to "trusted persons." Such measures have worked effectively (but not perfectly) in the granting of security clearances. More particularly, they have provably worked to secure El Al aircraft against an intense terrorist threat. In 1989, the author had a stressful two-hour interview at Tel Aviv airport with airport security, in part because a sudden change in schedule forced him to buy a one-way ticket to London and to leave his wife in Israel to return separately to New York.

The cost of such interviews would be much reduced, and effectiveness increased if they formed part of a database of information on trusted individuals, together with personal history. A person would be trusted for some months or a year after a satisfactory interview, and access (to aircraft, vulnerable and hazardous facilities, working as a truck driver, etc.) would be possible only if the person passed a biometric test (iris scan, fingerprints, or the like) to verify that the person's identity was as claimed, and matches that of a trusted person.

Beyond that, of course, is the necessity to ensure that a trusted person is not acting under coercion, as was the case in some of the truck-bomb attacks in Northern Ireland-- where an ordinary citizen was blackmailed into cooperating by the threat that his family would be murdered if he did not comply.

WHAT TO DO ABOUT HIJACKED PASSENGER AIRCRAFT AS WEAPONS. The instigators of the 9/11 attack knew better than to imagine that they could break into the cockpit and depend upon a threatened commercial pilot to drive the aircraft into its target. Evidently no pilot would do that at the certain cost of his life and those of the entire crew and passengers. To prevent a repetition, all that is necessary is to strengthen the cockpit door. That and the informed determination of cabin crew and passengers using improvised weapons will sufficiently lower the chance of success that such an attempt will rarely succeed. Thus, it will not even be attempted.

But there are other scenarios of aircraft as manned missile. A commercial pilot might kill the co-pilot within the locked cockpit, and this could be done within 30 minutes of a scheduled landing in the area of the desired target. A partial solution is in the trusted person approach, including its extension to foreign pilots entering the United States as well.

Fighter aircraft are unlikely to be able to respond in such a short time to destroy an airliner-become-missile, although one could envisage almost instant warning as an aircraft diverted from its computer-monitored declared flight path-- a system which is long overdue and which would help to improve the capacity of the airways and safety and reliability as well3.

FREIGHT AIRCRAFT AND PRIVATE AVIATION AS MANNED MISSILE. Hundreds of very large aircraft are used in the United States and elsewhere in freight service, where there would be no cabin crew or passengers (or on-board hijackers). With better piloting skills than were shown by the 9/11 hijackers, such an aircraft, fully fueled, could be stolen from an airfield and flown to its target. Evidently there are many opportunities for countering this action. They range from improved "ignition key" locks to an additional device recognizing a secure token carried by the intended pilot, if the device is removed or not satisfied, it would disable a portion of the aircraft essential to takeoff.

HOW TO COUNTER PRIVATE AVIATION AS A FLYING BOMB. Light aircraft and their fuel load cannot do damage comparable with that from a half-million pound gross weight airliner. But a large fraction of the light aircraft are capable of carrying half ton of explosives-- either dynamite or military explosives or, more likely, the ammonium nitrate used in the Oklahoma City bombing. Many light aircraft take off from fields where there is no full-time human presence. It would be helpful for every takeoff to be registered by telephone or computer first, and then confirmed by radio, including the flight plan and the identity of the pilot. Biometrics in the form of a thumbprint reader in the aircraft would be helpful together with the trusted-person interview of all with a pilot's license in the United States. Such a device and its interface with the radio might sell for $200, including the appropriate encryption/validation mechanism.

Half a ton of explosives detonated against the side of a building might initiate progressive collapse, or it might not. Government buildings of great prestige such as the White House or the Capitol could be protected with very short-range antiaircraft missiles. Nuclear reactors might also be in this category; France has deployed such missiles for the protection of the reprocessing plant at La Hague, in Normandy.

COUNTERING HIJACKED 18-WHEELERS AS WEAPONS. Bill Wattenburg has proposed and tested a means of preventing the use of a hijacked 18-wheeler as a weapon. The scenario is that a truck driver is accosted and killed as he or she is about to reenter the truck after a rest stop. The hijacker then drives the hazardous cargo (possibly gasoline, but there are much worse) at high speed toward a pre-arranged target. Large trucks are extremely difficult to stop at a road block; Wattenburg's idea instead is to fit each truck with an exposed air-brake hose at the rear, equipped with a cutter than can be activated by a push-bumper now found at the front of many highway patrol cars. Within a minute or so after the act, the truck brakes set, and the truck becomes immobile.

TRUCK BOMBS OR HAZARDOUS CARGO PILOTED BY TERRORIST DRIVERS. In this case, no hijacking is required, since the driver has spent years in training and getting assigned by a transport company to driving a truck. It could be loaded with many tons of explosives by accomplices, or the normal cargo could be explosives, chlorine, or the like. A solution is the trusted-person approach, whereby not only the driving skills but also the background would be determined in an interview and investigation, before the driver's license (biometric-enabled) would be issued. In addition, as will be discussed under "securing targets," only approved combinations of driver, truck, and cargo might be admitted to sensitive areas such as Manhattan. Such an extension of EZ-Pass might be called a SafeTpass and would not slow vehicles so equipped.


At a hearing of the Senate Foreign Relations Committee March 6, 2002 Henry Kelly, President of the Federation of the American Scientists, presented extensive testimony on the impact of radiological dispersal devices (RDD). Kelly's testimony was based on a paper by physicists Michael Levi and Rob Nelson, and is available on the Web4. RDDs might be used to disperse radioactive material most hazardous if inhaled and deposited in the lung. For this purpose, the particles should be in the range of one to five microns in diameter. In our book we report a hypothetical attack on Munich with one kilogram of plutonium dispersed by high explosives5. Assuming a very pessimistic low wind speed so that the cloud remains over the city for 12 hours, the net result is that 120 people would die of cancer after 40 years or so.

Radioactivity is easily detectable in the most minute amounts. The average person every second is host to some 8000 disintegrations of potassium-40, created in the super-novae which produced all of the heavy elements of which we are formed. And this is only 5% of the background to which we are normally subject-- from the rocks around us, cosmic rays, and from naturally occurring radon.

The effect of radiation on people is to add to the death rate from cancer. Some 20% of us die of cancer, but only about 0.5% of us from cancers presumably induced by the normal background radiation.

Yet the deposit of radioactive materials sufficient to triple the annual average exposure to an individual, as the result of the Chernobyl disaster of 1986, was sufficient to provoke general evacuation of the population by the Soviet authorities. For a lifetime of exposure in the contaminated area, an additional 1% of the population would die from cancer.

Kelly's testimony included the example of the radioactivity of a single cobalt-60 pencil (or "rod") from a food irradiator, explosively dispersed in lower Manhattan with a wind aligned with the axis of the island. The contours of exposure are shown in Fig. K2. (Page 11 of 14-page Ref. 4). There are many uncertainties in the calculation, including the rate of deposition (fallout) of the finely dispersed material, the depth of penetration into the soil, and the like. The point is that except in the immediate vicinity of the explosion, the populace need take no early action, since a week of exposure would be 250 times less dangerous than the lifetime of exposure (here the lifetime of cobalt-60-- about 5 years) that would lead to the cancer toll indicated. .

A single Co-60 rod or "pencil" contains typically 10,000 curies of radioactivity. The pencil could be stolen, transported in a shield cask of lead or iron, and removed from the shield only shortly before it is to be dispersed. Hundreds of millions of curies (a curie is 37 billion disintegrations per second) of Co-60 have been shipped for use in clinical sterilization machines and food irradiators. The spices in our kitchen cupboards are irradiated before shipment from their origin in India or elsewhere. The actor for many RDDs can be following a script, rather than have thought out the plot himself. The radioactive material might have been stolen within the country by other members of the cell, or even imported by aircraft or shipping container.

OPPORTUNITIES FOR COUNTERING CO-60 RDDS. Of course, protection at the source is highly desirable, and that includes not only food irradiators and the like the world over, but medical radiotherapy devices such as that which in 1987, taken from scrap metal, contaminated much of the town of Goiania, Brazil. The villagers used their fingers to spread the glowing powder on their skin and some ingested it with their food. Fifty-four people were hospitalized; four died; and the cleanup of the town by the International Atomic Energy Authority (IAEA) required the disposition of 4000 tons of contaminated buildings and soil.

There is also the opportunity of interdiction of the material on the way to its dispersal target. Co-60 is widely used also as a high-energy x-ray source for radiographic industrial castings, structures, and the like. The penetrating radiation can be shielded in increasing amounts by increasing thickness of shield, but highly sensitive detectors can detect the specific gamma-ray lines of Co-60 or the other common industrial and medical sources. Such detectors are being deployed in various areas, and are increasingly linked in a network, so that a moving source that does not spend enough time near a single detector can still be reliably detected and identified as it moves through the system. But an effective system of this kind depends upon larger and more widely deployed detectors that are commonly present.

Improved security over intense sources of radioactivity in the health and industrial sector is evidently necessary to counter the threat of terrorism, since the existing measures were directed largely towards safety and not security. Increased costs associated with such protection will in some cases drive the process to replace the radioactive source with a compact electron accelerator, which can pose no threat of use as a radioactive dispersal device.


My chief concern is not with 10,000 anthrax letters, or even the equivalent of 100,000 distributed in a fashion conducive to inhalation anthrax-- 80% fatal if untreated. I am concerned about the use of biological agents that are contagious as well as infectious.

Smallpox will make the point. The world has a lot of experience with smallpox, which by foresight and aggressive action on the part of the World Health Organization was rendered extinct in 1980, since it has no host other than human. Two stocks of smallpox virus were officially maintained-- one in the United States and one now in Russia-- but testimony from one of the workers in the Russian BW program attests to the fact that the Soviet Union had weaponized smallpox. Numerous ballistic missile warheads filled with BW agents were reportedly deployed, as was discovered in 1991 to have been the case with Iraq. But here we are concerned not with these BW agents in wartime, but its use by terrorists.

Some of the Soviet military BW agents may have been stolen or diverted; the U.S. has never had access to these former military BW installations in Russia. It is also likely that some individual researchers whether in military or civil programs did not destroy their stocks of smallpox virus when their nation signed the Biological Warfare Convention of 1972, but kept some for a rainy day-- perhaps without any malevolent intent. Some of these stocks may have fallen into the hands of terrorist groups; stored, they could be multiplied by the same techniques used to grow viruses for human or animal vaccines, and could be available for widespread dispersion by nebulizer from a moving car or truck. The BW agent is difficult to detect in a sealed container, even in amounts of 100 lbs or so.

At the time vaccination was abandoned in the United States in 1972, I argued strongly in the President's Science Advisory Committee that because of the coming extinction of smallpox, society would be too vulnerable to intentional attack, and that vaccination should be continued, despite the two or three people per year who might die from the continued vaccination program. The U.S. now has 15 million standard doses of smallpox vaccine, but recent experiments have shown that the vaccine is effective in doses five times as dilute. And a further economy may be achieved by a bifurcated needle, which holds the liquid, rather than relying on liquid which has been applied to the skin. If exposure is known to have taken place, the vaccine is apparently still effective if given within the next four days. In late March, 2002, it was revealed that some 85 million additional doses of smallpox vaccine exist in the United States; these also can be extended in the by dilution, so that there is now more than enough for every US resident, even without the new-production vaccine planned for availability in 2003.

But there are many potential BW agents, such as that of glanders (Burkholderia mallei), for which there is no vaccine. They might be disseminated within a large building, and distributed by the circulating air in the HVAC (heating, ventilating, and air conditioning) system-- or outdoors, to expose a whole city. It is of interest that a person within a building without filters will receive the same dose of BW or CW agent as a person outside directly exposed to the passing cloud. If the windows are open, this is not surprising, but even if the windows are closed so that there is a typical exchange of air only every 30 minutes, the slow influx of BW or CW agent is matched by the slow efflux, so that the integrated dose is the same. This is strictly true only if the deposition time of the BW or CW agent is very long.

The situation is different if there are in place effective filters against BW. High efficiency particulate air filters (HEPA) are available with face velocities (air flow per unit area) identical with that of normal air filters used in circulating systems (500 ft-170 meters--per minute) and a pressure drop which is the same as that of a normal filter partially loaded with dirt. HEPA filters trap 99.97% of particles larger than 0.3 microns in diameter-- compared with an anthrax bacillus of perhaps one micron diameter by three microns in length.

Although the smallpox virus is much smaller (as is the toxin of botulinum), such tiny particles agglomerate, attach to dust or droplets, and may also be effectively filtered by HEPA. More research is necessary to determine the practical effectiveness of these widely available filters. An example is a normal home with a blower circulating air at 2000 cfm (cubic feet per minute) and occupied by several people, each breathing at a resting ventilation rate of 8 liters per minute or 0.3 cfm. A particle infiltrating the home through a crack in the door is 6000 times more likely to be caught in the filter than in a particular person's lungs. But the person's exposure is not reduced by a factor 6000 because of the circulating system. In a normal home of 2000 sq ft and volume of 16,000 cubic ft, the circulating air changes every eight minutes, while air exchange with the outside occurs on the average every 30 minutes. Thus particles which infiltrate are in the air for about 8/30 as long as if there were no circulation or filtering. The improvement in indoor air quality due to filtering circulating air in this case is a modest factor four, and that is the factor by which exposure to the BW agent will be reduced.

POSITIVE PRESSURE COLLECTIVE PROTECTION. But if the living space is maintained at positive pressure so that any leakage of air through the cracks of the enclosure is out instead of in, HEPA-filtration of this "makeup air"-- that required to maintain the necessary pressure against infiltration-- can reduce the exposure by a factor thousand or more. The required pressure is just the "stagnation pressure" at the wind speed V-- 0.5 V2. For a brisk wind of 20 miles per hour, this amounts to about to a pressure of 0.2 iwg in practical units (inches of water-- gauge), or about 0.5 millibar. It is also relevant that this is about 1.1 lbs per sq ft (psf), so that the force on an exterior door is about 20 lbs.

Such positive pressure with filtered makeup air can be readily applied under certain circumstances to office suites. It is more complicated with a centralized building HVAC system and implementing it would require professional advice.

In a home, special attention must be paid to vents from furnaces and water heaters, although such appliances can experience similar circumstances if a door is left open on the upwind side of the house in a strong breeze.

In any case, I strongly advocate positive-pressure collective protection. Of course, this does nothing for people who are outside during the typical 20 minutes or so of cloud passage, but most people are inside. And it is entirely possible to provide collective protection with positive-pressure filtered air in public transport or even private automobiles.

BW agents act as individual microbes, despite the widespread confusion about a "5000-spore lethal dose" for anthrax. This simply means that a single anthrax spore residing in the lung has about 1/5000 probability of causing death, if untreated.

So if air filtration reduces the dose to the average person by a factor 100, and if the BW agent was efficiently distributed so that unprotected people received no more than the average lethal dose, the number of people dying would be reduced by a factor 100.

A citizen (or especially a politician) might say that 1/300 chance of dying in a BW attack is too great a risk, but from the point of view of protecting people, 0.3% instead of 30% is a tremendous factor. If instead of one million deaths one has ten thousand-99% protection--this is the difference between life and death for 990,000 people.

The search for perfection often stands in the way of tremendous benefit.

VACCINES AGAINST BW. Before the AIDS epidemic and the resurgence of tuberculosis, it was a common expectation that humanity was well on the way to vanquishing infectious diseases. In reality, microbes readily acquire resistance to antibiotics and in this sense do learn to fight back.

Vaccines can be very effective against infection, although to prevent epidemics and in favorable cases to bring about extinction of the microbe, it is only required to reduce the transfectivity below one on the average. If it were taken as three for smallpox, as in the previous example, then a vaccine that was 80% effective would do the job. Greater effectiveness is sought against BW agents, but in reality modest effectiveness can make all the difference.

But vaccines may not be effective against different strains of the same microbe, as is universally recognized with influenza. Each year brings the need for a different vaccine, because of the rapid mutation of the influenza virus, and the peculiarities of its transfection and modification in passing from bird to pig to human.

In addition to vaccines, there is the prospect of antibiotics to interfere with the growth of pathogenic organisms, or even antitoxins to counter the specific mechanism by which anthrax, for instance, or botulinum toxin claims its victims. Apparently anthrax produces three proteins which, together account for the virulence of the disease. These proteins are very specific, and with modern techniques it may be possible to block their action and thus convert anthrax to a nuisance rather than a highly dangerous infectious agent.

TERRORIST NUCLEAR EXPLOSIVES. A terrorist nuclear explosive would devastate a city, whether detonated in the hold of a ship in harbor, in a cargo container, in a cellar, or in an apartment. The essential ingredient for a nuclear explosive is fissile material-- highly enriched uranium (HEU) or plutonium. Although the yield of the uranium bomb that devastated Hiroshima was 13 kilotons (13,000 tons of TNT equivalent), and the plutonium bomb which destroyed Nagasaki yielded 20 kilotons, nominal U.S. and Russian strategic weapons now are in the range of 150 kt. A recent report details the damage of what we expected from explosions of 1, 10, and 100 kt at ground level in a city. The Table taken from NCRP shows the approximate radii to which the quality or destruction extends, for the 1 kt and 10 kt yields6.

Consider a 1-kt explosion. This might occur from a gun-type device with less material than was used at Hiroshima, or a plutonium implosion-type device made from reactor-grade plutonium and yielding only a "fizzle" because of a large neutron background from the reactor-grade plutonium. On the other hand, the plutonium device might yield 10 kt, so both are shown in the Table.

Table-- Summary of ranges for significant effects (in meters).
Yield (kt) (a)* (b)* (c)* (d)*
1 275 610 790 5500
10 590 1800 1200 9600
a* Range for 50% mortality from air blast (m)
b* Range for 50% mortality from thermal burns (m)
c* Range for 4 Gy initial nuclear radiation (m)
d* Range for 4 Gy fallout in first hour after blast (m)

Considering the numbers for 1 kt, we see that people out to 275 m (900 feet) are likely to die from the blast. We can transform the first three columns into the number of Manhattan city blocks which would be destroyed, simply by equating the area within the circle of 50% effect to a number of city blocks7.

The conversion was made by noting that Central Park is 836 acres, and there are 247 acres in a sq km. Thus Central Park is 3.38 sq km. Extending from 59th St. to 110th St., it is 51 blocks north-south and three large blocks east-west. Thus it has 153 large Manhattan blocks. There are thus 45 Manhattan blocks per sq km.

The city blocks destroyed by air blast (50% mortality in the "cookie cutter" approximation-- 100% lethality out to the 50% line, and 0% mortality beyond that): 11

City blocks in which almost everyone would die from thermal burns: 53

City blocks in which people would get a lethal dose of prompt nuclear radiation: 88

For the 10 kt explosive, the results are 49, 457, and 203 city blocks.

To convert these areal measures into fatalities, we might take a particularly high local daytime Manhattan population density of 125,000 per sq km or an average of about 2360 people per Manhattan block. So for the 1-kt explosion, some 210,000 people would die-- mostly from prompt radiation within a week or so. Of these, 30,000 would have died from blast earlier, and about 100,000 from burns.

For the 10-kt explosion, about a million people will die from burns. Less than half of these would have died from radiation exposure.

As for fallout, the Table is to some extent misleading, since this provides the distance at which lethal fallout within one hour might be deposited, but it is not a circle of that radius. From the 1977 "Effects of Nuclear Weapons," Table 9.93 (p. 4.30) we see that for a reference dose rate (i.e., for a 1 kiloton explosion) of 3 Sv per hour (300 rads/hr), the downwind distance would be 4.5 miles, and the width about 0.15 miles, for a region affected on the order of 0.7 square miles or 1.5 square kilometers, or 80 Manhattan blocks. So the fallout, although lethal, would not totally dominate the casualties from a nuclear explosion.

Compared with an air burst of a large nuclear weapon at an altitude designed to maximize the blast damage, the prompt radiation and the fallout are far worse with a terrorist explosion. This comes about because the bomb detonated at or near surface of the Earth throws up an enormous amount of earth and vaporized structure, which descends in the immediate neighborhood, providing lethal fallout, which is essentially absent when the fireball does not touch the ground.

If it were known that a nuclear explosion was to take place, evacuation would be highly desirable. And as in the case of potential reactor accidents (with or without terrorist involvement) it would be very useful to have distributed and ready for use potassium iodide (KI) tablets or capsules. A 130-mg dose would block the uptake of radioactive iodine to a young thyroid (or to a nursing mother), and avoid many thyroid cancers which would destroy the thyroid and might be lethal.

Of course, hospitals would be overwhelmed with the number of people actually injured by flying glass, suffering from radiation exposure, and the like. Furthermore, transit in the city would be disorganized in the regions effected. With buildings down over a square kilometer or so, as was evident in the case of the World Trade Center collapse covering 1% of that area, severe damage to the communications and transportation infrastructure would be expected.

Organized medicine would be unable to cope. A volunteer emergency medical corps, with adequate planning and practice, could save some people who would otherwise die.

Nevertheless, a terrorist nuclear explosion would explode in one place, or a very few, compared with the nuclear attack which we feared for many years and decades from the Soviet Union. So other localities could send personnel and supplies and be a destination for evacuation from contaminated areas.

Public safety personnel would need to use radiation detectors to determine places which posed no continuing radiological problem; regions in which people could not stay for even an hour or five hours without a high likelihood of dying within weeks from radiation damage; and regions in which radioactivity was clearly evident, but which would add perhaps only 1% to the 20% of American citizens who ultimately die of cancer instead of from some other disease.

The effects of a nuclear detonation in a city are so horrendous that it is clear that most effort should be placed on preventing access by terrorists to nuclear materials or weapons; to interdicting the transportation of weapons or the building of improvised nuclear devices; and to keeping them out of areas of large population density8.

Unlike the case of large-scale nuclear war, a single terrorist nuclear explosion would not eliminate the resources of the rest of the country, so healthy survivors could be accommodated elsewhere. Those in the regions subject to substantial fallout could receive expedient medical care, but little can be done for those exposed above the levels shown in the Table. Unlike BW attack, a nuclear explosion is evidently far better prevented than treated.

Stolen or diverted military nuclear weapons are rugged, but they are usually provided with substantial protection against unauthorized detonation, so considerable skill might be required to employ one. On the other hand, an improvised nuclear device (IND) would not have this problem, but can be difficult to carry off. The fissile material is not an article of commerce and itself would have to be stolen or diverted. The first plutonium bomb incorporated 6 kg of weapon-grade plutonium, of which more than 250 tons has now been made worldwide-- enough for 40,000 such crude weapons. Almost all was produced by the United States and the Soviet Union.

In addition, every large nuclear power reactor produces annually on the order of 200 kg of plutonium, which is not and need not be weapon grade to make an improvised nuclear device. In January 1997 the U.S. Department of Energy stated of reactor-grade plutonium, "Proliferating states using designs of intermediate sophistication could produce weapons with assured yields substantially higher than the kiloton-range possible with a simple, first-generation nuclear device."

At the March 6, 2002 hearing of the Senate Foreign Relations Committee, Senator Joseph Biden quoted former Los Alamos National Laboratory Director Harold M. Agnew to the effect that "If somebody tells you that making a plutonium implosion weapon is easy, he is wrong. And if somebody tells you that making an improved nuclear device with highly enriched uranium is difficult, he is even more wrong." Plutonium metal can be safely accumulated in spherical form up to the so-called "critical mass" of 10 kg for weapon-grade plutonium or 13 kg for reactor-grade plutonium. The analogous critical mass for 94% U-235 is 52 kg, and these numbers set the scale for the amount of fissile material required for a nuclear weapon.

Instead of being assembled by high explosive as in the plutonium bomb (which can also be used for assembly of a uranium core) the Hiroshima bomb was two solid masses of highly enriched uranium metal, one of which was propelled in a shortened, converted naval gun to form more than a critical mass with the stationary uranium metal. Although less efficient, this is far simpler than is the plutonium IND.

With the enriched-uranium gun-type weapon, there is an additional means of preventing significant nuclear energy release. Guns are exquisitely sensitive to the presence of neutrons-a relatively few neutrons will guarantee that only a very small amount of fission energy is released. This is evident from the prompt criticality accidents at Los Alamos, which killed people in the room by acute radiation sickness over a period of weeks, but did not even disrupt the fissile assembly. It is possible with a neutron source of one kind or another to flood the uranium remotely with neutrons at such a rate that, even if the smokeless powder is fired to assemble the uranium, no significant yield will result.

The best single protection against the terrorist use of such weapons is to deny the acquisition of the necessary plutonium or enriched uranium. The low-enriched uranium used in U.S. nuclear reactors (typically 4.4% U-235) can in no way be used directly to make a nuclear explosive. That is true up to about 20%, for which the critical mass is 800 kg. HEU is used not only in nuclear weaponry, but in some research reactors and in fuel for naval reactors, such as propel our aircraft carriers and many of our submarines. Likewise, Russian nuclear-propelled ships use HEU. And in Russia particularly, stocks of HEU and plutonium (even weapon plutonium) do not have nearly the security provided to their nuclear weaponry.

After some months of denigrating U.S. programs which have existed since 1994 to help Russia protect weapon-usable materials, the Bush Administration in December 2001 recognized the seriousness of this problem and that something can be done to solve it, and has increased the budget for such Cooperative Threat Reduction activities.

The U.S. is buying 500 tons of HEU (diluted in Russia to LEU to fuel U.S. reactors) over 20 years, at a cost of about $12 billion. Here is a threat which will persist for mush longer than necessary. It would be a simple matter for the United States and/or the international community to advance Russia the much smaller amount of money required to blend down the remaining 370 tons (and perhaps another 700 tons of HEU not included in this deal) to 19.9% U-235-- thus essentially unusable for nuclear weaponry. This could be done quite readily in about two years, and the money would be repaid by Russia with or without interest at the same time this 19.9% materials (remaining in Russia) was later further blended to the 4% range for transfer to the United States. These funds should come from the G-7.

Weapon-usable materials might be detected in transit. Normal uranium metal is detectable primarily because of gamma rays of near 1 MeV energy, although relatively few are emitted, since the half-life of U-238 is 4.5 billion years. U-235 has a shorter half-life (700 million years) but its decay scheme is far more difficult to detect.

The intense radioactivity of plutonium is largely alpha-particle emission in its decay to U-235. The half-life of 24,000 years means that 6 kg of Pu is about 500 curies. Pu is a serious hazard if inhaled, but has very little external radiation. Nevertheless, appropriate counters detect it at a considerable distance, although it is easier to shield than is U-238.

Weapon uranium is only very weakly radioactive-- with U-235 having a half-life 30,000 times that of plutonium.

There have been many hoaxes in the United States, mostly extortionists demanding money in order not detonate a nuclear weapon they say is ready to explode someplace in Boston, New York, or elsewhere9. To find such emplaced explosives and to disable them, the government created the Nuclear Emergency Search Team (NEST) which now has the ability to deploy about 600 people with appropriate detection and disabling devices.

But a terrorist with a mission actually to kill people would certainly not alert the authorities to the existence of a nuclear explosive. It would need to be detected either following intelligence tips, or by generalized search, or in transit to its emplacement.

This is a tall order for NEST, even granting substantial improvement in their capabilities.

SUMMARY. In looking at a particular terrorist act as involving an actor, an action, and a target, we find that different acts may be impeded in a variety of ways. But there is some generality to the solutions that will reduce terrorism, even though it will not be possible to eliminate it. The U.S. and other states should act to

  1. Directly and indirectly to minimize the number of people who wish to become terrorists against us.
  2. Move against the actions and to harden targets so that the actor needs more training to carry out the deed.
  3. Introduce trusted-person databases and a biometric-based personal identification to reduce the access of un-vetted persons to hazardous areas or hazardous tools.
  4. Modify aircraft standards (strengthen cockpit doors) against hijacking; harden freight and private aircraft against unauthorized use.
  5. In the bioterrorism area, urgently expand development and production of vaccines, not only in the most highly industrialized states but also in India, for example.
  6. Develop and implement in the government and civil economy collective protection by positive-pressure filtered air and filtered circulating air (and masks where desired). Reduction by a factor hundred or even 20 of deaths due to bioterrorism is a worthy goal.
  7. Improve security over radioactive sources to reduce the threat of radiological dispersal devices (terrorism by contamination with radioactivity). The cost of adequate security will encourage the substitution of radioactive sources by electron-beam accelerators.
  8. Have contingency plans and public education so that people will not move precipitously and dangerously when there would be no significant hazard in remaining in place and living a normal life for a week or more.
  9. Replace absolute limits for radiation protection by a market-based approach, in which, for instance, full disclosure and inspection would quantify the increased cancer hazard for a particular home in a contaminated area, and it could be transferred at a market price to people who would on the average be older and have fewer children. Recall that 20% of us will die from cancer. Life is too short for the individual to worry about an additional 1% probability in the remote case of terrorist attack, although it is an important topic in public health.

These partial remedies are not now available, and they will not exist unless the United States creates a technical organization responsible for evaluating the terrorist threat, identifying potential remedies, and evaluating capabilities at any time. This needs to be done with wartime urgency, the same urgency that drove the creation during World War II of the radar lab at MIT and the Manhattan Project for the development of the nuclear weapon. Sections of a small number of existing government or national laboratories might initially be put under the firm control of a homeland-defense analogue of J. Robert Oppenheimer-a person with technical leadership and total dedication to the cause of reducing the vulnerability of society.

A homeland security institute is one of the major recommendations a recent report10. I was a member of the authoring committee and of its panel on nuclear and radiological issues. The United States is creating a Department of Homeland Security; if headed by a 21st-century counterpart to General Leslie R. Groves of Manhattan Project fame, it could in principle realize some of the near-term remedies advocated here. It could also mount a longer-term program of research and development to reduce the likelihood of catastrophic terrorism and-in the case of bioweapons and radiological dispersal devices-to reduce the economic and human costs in the event of actual attack. The solution is not in more organization but in ensuring that competent people can do their jobs.

The peril is global; and so should be the response. Vaccine development can proceed in India as well as in the United States. Advances in techniques for cleanup of radioactive contamination can proceed worldwide. Intense sources of radioactivity are present in many countries; an effort led by the IAEA could improve their security against theft. If the resources of the world are to be marshalled for the benefit of humanity, they should not be diverted unnecessarily to countering terrorism. Hence the need for collaboration and efficiency in solving this new and urgent problem.


  1. This paper is an expanded version of an article to appear in the September 2002 issue of MIT Technology Review, titled "The Technology of Megaterror."
  2. Appendix C: BW Agent Characteristics, of USAMRIID's Medical Management of Biological Casualties Handbook, available at http://www.vnh.org/BIOCASU/26.html
  3. Boeing "free-flight" approach. Also Air Traffic Control Panel Report of the President's Science Advisory Committee, R.L. Garwin, Chairman (1971).
  4. At http://www.fas.org/ssp/docs/kelly_testimony_030602.pdf
  5. "Megawatts and Megatons: A Turning Point in the Nuclear Age?" by R.L. Garwin and G. Charpak (Alfred A. Knopf, October 2001), pp. 341-342.
  6. Extracted from "Management of Terrorist Events Involving Radioactive Material," Table 3.7 on p. 23 of NCRP Report no. 138 of 10/24/01, Recommendations of the National Council on Radiation Protection and Measurements, 7910 Woodmont Avenue, Bethesda, MD 20814-3095.
  7. In the "cookie cutter" approximation, assuming that damage beyond the 50% damage expectation contour is equal to the less than total damage within that contour.
  8. A set of pages portraying the course of a terrorist nuclear explosion in New York: http://www.atomicarchive.com/Example/ExampleStart.shtml
  9. "Defusing Nuclear Terror," Bulletin of the Atomic Scientists, March/April 2002, pp. 39-43.
  10. "Making the Nation Safer: The Role of Science and Technology in Countering Terrorism" The National Academies Press, Washington, DC, June 25, 2002. (Available at http://www.nap.edu).