Since the 11 September attacks, much of the literature on terror has focused on the psychology of terrorism rather than on identifying attacks that may occur. While the psychology of terrorists is important, it doesn’t do much to help those responsible for defending against terrorist acts. One thing should be remembered: terrorists, whoever they may be (Middle Eastern, Asian, European, or domestic), will use whatever tools are most suitable that they’re trained to use and are available.

In “A Terrorist Target Selection and Prioritization Model” (2003),1 the senior authors described a mathematical model they believed to be of the type used by al-Qaeda planners in selecting targets. By using this model, those responsible for safeguarding against acts of terrorism would have a heads-up as to which targets terrorists are likely to attack. To demonstrate the model, a terrorist assessment team of eight specialists played the role of terrorists in an action where the target was a Florida county with no obvious strategic sites. Using the model in a process of several steps, they identified 99 terrorist events, then determined the impact and likelihood for each event. The product of a given event’s impact and likelihood constitutes that event’s “expected value.” By ordering the 99 events according to their expected value, those responsible for providing safeguards reduce to a relatively small number the events for which to prepare.

Next, the assessment team developed three scenarios (involving six terrorist events). Using one scenario (involving two events), they easily carried out a simulated terrorist attack.

In 2003, Joshua Sinai2 concluded that the failure to anticipate 9/11 was not an intelligence failure but that 9/11 was simply beyond the imagination of those responsible for intelligence and law enforcement. He also emphasized the necessity to “adopt proactive measures to anticipate, defend against, and preempt new types of terrorist threats” and that “terrorists, especially al-Qaeda planners, always seek to exploit new vulnerabilities and new and innovative modes of warfare in order to evade detection and inflict maximum damage.” Firmly agreeing with Sinai’s conclusions, we believe that we must think like terrorists, determine what they are likely to undertake, and develop appropriate countermeasures.

Since 11 September, much has been done to bolster defenses in the United States. These steps have focused on such sites as airports, tunnels, and bridges. Consequently, these types of sites have become hard targets, a fact of which terrorists are well aware. Even so, the question is whether those focusing on these hard targets are focusing on the areas in which we truly are vulnerable.

In dealing with terrorists, particularly al-Qaeda, we cannot rely on historical trends with regard to sites and attack modes. Instead, we must anticipate unconventional methods, techniques, and targets—which means that we simply must think like terrorists. Furthermore, it would be a another failure in imagination if we as a nation focused exclusively, or nearly exclusively, on geographic areas with many potential hard targets. We should also consider areas where hard targets are scarce, where little thought has been given to the possibility of terrorist attacks.

Methodology

We came to the conclusion that to meaningfully identify potential domestic terrorist attacks, we would need to look at relatively small geographic areas (for example, county by county) rather than at the nation. For this reason we selected Charlotte County, FL, for testing the effectiveness of our Terrorist Target Selection and Prioritization Model. The county is devoid of strategic targets. People who live there, like those in many other areas of the United States, go about their business every day without being particularly concerned about terrorist attacks. Charlotte County3 has approximately 150,000 residents; 35% are age 65 or older. County government and medical and health providers are the leading employment segments, along with the varied retail industry. The county has another interesting and distinguishing feature: it accommodated several of the 11 September terrorists. Mohammed Atta visited restaurants, facilitated overseas cash transfers, and sought aid in a local hospital there. It is reasonable to assume that the county is well known to al-Qaeda planners.

To identify potential terrorist attacks in Charlotte County, we created a team of eight “terror planners” (the authors of this article) with backgrounds in physics-math, psychology, infectious disease, radiation oncology, emergency room medicine, emergency management, environmental health planning, and senior levels of the U.S. military.

Using the equations described in Stungis and Schori’s 2003 article on the Terrorist Target Selection and Prioritization Model, our team, thinking like terrorists, identified 99 possible terrorist events (target sites and modes of attack) that might take place in Charlotte County. The planning team judged the likelihood of success (from a terrorist perspective) for each event. Then the team made judgments, using the Stevens Power Function,4 of the impact that each event would have on three terrorist objectives (publicity, with a weight of 0.2; casualty count, with a weight of 0.3; and economic impact, with a weight of 0.5). Cross-impact matrices were also constructed. The product of the likelihood rating and the overall weighted impact associated with each terrorist event constitutes the expected value to the terrorists. Since the likelihood scores are probabilities and the impact scores can range from 0 (no impact on the terrorist objectives) to 8 (maximum on each terrorist objective), the resulting terrorist alert scores can range from 0 to 8.

Results

We believed that it would not be necessary to further consider potential terrorist events that had terrorist alert scores less than 3.0. The terrorist alert scores themselves can be thought of as a type of desirability index (from a terrorist perspective) and as a terrorist alert index (from the perspective of those responsible for safeguarding against terrorist attacks). Thus, we have been able to logically reduce the number of events so that countermeasures can be implemented (see Table 1).

The most attractive event (from a terrorist vantage point) would be a biological attack at a shopping mall in Charlotte County, followed by a suicide bomber attack at a major festival in the county.

Table 1. Potential Terrorist Events (Combination of Site and Mode of Attack) in Charlotte County Ranked by Terrorist Alert Score

For any given terrorist event (such as a biological attack at a shopping mall), terrorists could choose to carry it out in dozens of ways. To provide a look into the thinking of terrorists, we developed three scenarios for carrying out some of the 14 events.

Scenario 1: Vehicle bomb attack, then chemical attack, at a major festival.

Objective: To stage a terrorist attack in Charlotte County, FL, that will kill many and maim many more and thus instill hysteria.

Strategy: A truck bomb attack followed immediately by a chemical attack.

Plan: At a very popular Charlotte County festival, while the maximum number of people is present, three terrorists drive small pickup trucks, through separate festival entrances, directly into the crowds of people. Simultaneously, each terrorist will detonate a 500-lb. charge. Each explosion is expected to kill more than 50 persons and injure many more. Then, coinciding with the explosions, just as soon as survivors are huddling at a point equally distant from the three explosions, a fourth terrorist will use a fire hose to spray the survivors with anhydrous ammonia. At least 10% of those sprayed with the ammonia are likely to die. Many of those who survive will be blinded or horribly burned. So many burns will occur that the county’s health services will be unable to handle the volume and severity of the injuries.

These attacks will be completely unexpected and will result in mass panic, not only in the county but across the state and country as well. The immediate economic impact of these attacks will not be great. Only a few hundred people will be killed, and perhaps a thousand or so will be severely injured. The fear that will be created will far exceed the impact of any economic cost.

Scenario 2: Vehicle bomb attack, then a chemical attack at a major shopping mall.

Objective: To stage an attack that will convince the American public that government is powerless to protect them.

Strategy: A vehicle bomb attack will be used to cover an even more deadly chemical attack.

Plan: While crowds are at a maximum, three terrorists will simultaneously, at high speed, drive automobiles into separate entry points of a major shopping mall, directly into the crowds, where each terrorist will detonate 500 pounds of explosives. Each explosion is expected to kill 15 persons and injure another 125. Coinciding with the vehicle detonations, aerosolized sarin (which is more deadly than sarin applied to skin) will be dispensed throughout the mall through the ventilation system. At least 30% of those in the mall remaining alive after the vehicle detonations are likely to die from breathing the gas. There will be many deaths and so many injured that the county’s healthcare system will be sorely taxed.

That the attack will have occurred in a quiet, unassuming Florida county will frighten the American public far more than did the deaths and destruction that occurred in New York City on 9/11.

According to the 1997 Economic Census,5 49% of Charlotte County’s revenue is attributable to retail sales—a much higher proportion than for the country as a whole. County revenue, then, will be devastated by the economic impact of these attacks. But it will go far beyond Charlotte County. Throughout the country, people will stay away from shopping malls because if such an attack can happen in a county in Florida, it can happen anywhere. The recession from which the country is emerging will return with a vengeance.

Scenario 3: Biological attacks in hospitals followed by suicide bomber attacks in a popular shopping mall.

Objective: To paralyze the state of Florida and have an impact on the entire United States.

Strategy: To surreptitiously attack a key industry—health care—and overtly attack the retail industry.

Plan: Three individuals armed with two-ounce hairspray containers filled with a critical solution of smallpox virus in water would enter Charlotte County’s three hospitals approximately 15 minutes before the peak lunch hour. They would proceed to the hospital cafeteria and approach the salad bar, fruit bar, or suitable equivalent. They would carefully spray the solution over the bar, obtain a salad, and go to a nearby table and observe. Afterward they would leave the hospital at a leisurely pace. The process would be repeated before dinner—again in the cafeteria. The same steps would be taken the next day.

After 12 days, the infection rate for healthcare workers would be about 80%, while that for hospital patients would probably be around 90%; families of workers and patients would also have a very high rate of exposure that would, in all likelihood, lead to infection.

On the 12th day, at dinnertime, all three terrorists would independently proceed to the food court in Charlotte County’s main shopping mall. Sitting separately at key locations, they each would detonate explosive packs of 25 pounds of a nitrate compound. This would kill about 50 people and injure another 100 or so. The casualties would be rushed to hospitals that are severely biologically contaminated. At this point, hospital personnel and others would be at the threshold of showing the pox symptoms. County deaths could easily be in the thousands, and an ultimate financial loss of millions of dollars could be expected.

The terrorists would have released a videotape describing their mission and telling Americans to expect more. Across Florida and the United States, the healthcare system would be in panic and the retail industry would crumble. Two-thirds of the U.S. gross national product6 is accounted for by consumer spending. In all probability, the United States would rapidly go into a depression.

With all the discussion in the news media about terrorists and the potential for terrorist attacks, rational people might be led to believe that many safeguards have been put into place. Shortly after deaths started occurring from anthrax contamination of the U.S. mail, there was much discussion, for example, about the possibility of terrorists staging smallpox attacks. One might assume that hospitals, at least, would have implemented safeguards.

The three scenarios we described are just that—scenarios, not forecasts of doom. All in all, Charlotte County is no doubt a low-probability target for terrorist events. In fact, while a cutoff terrorist alert score of 3.0 may make sense for a given county in identifying what deserves to be safeguarded against, the government of the state in which that county is located should probably not become involved for county terrorist alert scores lower than 5. Likewise, the federal government should probably not become involved for county terrorist alert scores lower than 6.

Our intent was to demonstrate how an attack process unfolds and what the targets and modes of attack might be. Most of the counties in Florida (or any other state) could be handled, using the Terrorist Target Selection and Prioritization Model, in a similar way. Larger areas in Florida (such as Miami-Dade County or Tampa–St. Petersburg–Clearwater) would be evaluated somewhat differently.

The process shows the overall steps involved in planning an attack. Obviously, defenders cannot plan for all possible contingencies. The procedure is not a replacement for good intelligence data (which are not always available). Using the model allows a structure for attacks, a reduction of possible attack targets and modes, and the ability to view alternatives—facilitating defense planning and allocation of resources.

Conclusions

Starting with no a priori beliefs as to what terrorists might undertake, we were able to use the mode to cost-effectively identify and prioritize what terrorists might undertake within the county. By knowing what constitutes high-priority potential targets, those responsible for the at-risk entities (hospitals, festivals, etc.) have the opportunity to implement countermeasures designed to avoid (or reduce the severity of) terrorist attacks.

Not only political or military entities may become targets of terrorist attacks. Consider Illinois State University, with 22-story twin dormitories in which over 2,000 students reside, and State Farm Insurance Company’s national headquarters a mile or so away. It would make sense for planners in such entities to assess their vulnerabilities to terrorist acts but, unfortunately, they probably aren’t doing so. As Joshua Sinai indicated, as a nation, we don’t suffer from intelligence failures so much as from failures in imagination.

Recommendations

Because it is both effective and cost-efficient, the Stungis-Schori Terrorist Target Selection and Prioritization Model should be widely used by federal, state, county, and city governments and other entities to assess vulnerabilities to terrorist attacks.

If the methodology we used in Charlotte County is faithfully followed in assessing other entities, the resulting terrorist alert scores will be useful in comparing the vulnerabilities in one entity with those in another. In Charlotte County, a terrorist alert score of 3.0 is the point at which one should start developing countermeasures, as opposed to ignoring the threats.

Author Contact Information
George E. Stungis, Ph.D.
Thomas R. Schori, Ph.D.
Mark O. Asperilla, M.D.
Larry G. Beebe
Raymond A. James, D.O.
Rufus C. Lazzell
David J. Rice, M.D.
Wayne P. Salladé

References

1. George E. Stungis and Thomas R. Schori, “A Terrorist Target Selection and Prioritization Model,” Journal of Homeland Security, March 2003.

2. Joshua Sinai, “How to Forecast and Preempt al-Qaeda’s Catastrophic Terrorist Warfare,” Journal of Homeland Security, August 2003.

3. Robert Carpenter, Charlotte County, Florida, Statistical Prospectus, 2002-2003 edition, Charlotte County Chamber of Commerce, Punta Gorda, FL.

4. G.A.V. Borg and L. E. Marks, “Twelve Meanings of the Measure Constant in Psychological Power Functions,” Bulletin of the Psychonomic Society, vol. 21, pp. 73-75, 1983.

5. 1997 Economic Census, U.S. Census Bureau, Washington, DC.

6. 2002 Statistical Abstract of the United States, U.S. Census Bureau, Washington, DC.