SRA Logo (print)

Society For Risk Analysis Annual Meeting 2009

Risk Analysis: The Evolution of a Science

Session Schedule & Abstracts

* Disclaimer: All presentations represent the views of the authors, and not the organizations that support their research. Please apply the standard disclaimer that any opinions, findings, and conclusions or recommendations in abstracts, posters, and presentations at the meeting are those of the authors and do not necessarily reflect the views of any other organization or agency. Meeting attendees and authors should be aware that this disclaimer is intended to apply to all abstracts contained in this document. Authors who wish to emphasize this disclaimer should do so in their presentation or poster. In an effort to make the abstracts as concise as possible and easy for meeting participants to read, the abstracts have been formatted such that they exclude references to papers, affiliations, and/or funding sources. Authors who wish to provide attendees with this information should do so in their presentation or poster.

Common abbreviations

Poster Platform: Applications of Risk Analysis to Terrorism Security Sponsored by DARSG

Room: Salon A   3:30-5:00 PM

Chair(s): Henry Willis

T4-C.1  15:30  What can be learned about the public’s perception of terrorism risk? Rosoff H*, John R; University of Southern California

Abstract: Traditional risk perception approaches base their analyses on mean risk ratings, rather than individual subject ratings. This produces results that hide individual risk perception differences and generates patterns that may not be representative of any individual or of a majority of individuals in a sample. Our study focused on mapping the mental models of individual subject’s risk perceptions of terror and non-terror events. Data was collected through an evaluation of the perceived risk relationships across various event characteristics, both cognitive and emotional. Researchers also investigated whether perceived risk perceptions vary by subgroups defined in terms of demographic variables. The study design employed hierarchical linear modeling and used a heterogeneous sample (roughly 1,000 subjects) to allow for a systematic investigation of general and subgroup risk relationships. Study results suggest that judgments about attack likelihood and dread were the best predictors of perceived risk of terror events. Subgroup analyses found there to be stronger relationships between predictors and perceived terror risk compared to non-terror events. More specifically, personal annual income, level of education, and age were sensitive to terror attacks, while geographic location had a stronger association with non-terror events. Gender was equally sensitive to both. Data patterns providing additional insight into how individuals and subgroups think about risk were translated in policy recommendations that might assist policy makers in their development of terror preparedness and response programs.

T4-C.2  15:30  Using risk analysis to support law enforcement intelligence targeting. Lundberg R*, Willis HH; RAND Corporation

Abstract: While intelligence has often been used to inform risk analysis in the allocation of resources, intelligence collection is itself a constrained asset that can benefit from applied risk analysis. Intelligence collection is constrained by limited personnel, limited technical assets, and limited legal authorities. These assets can be applied to examining many threats (terrorism, drug trade, organized crime, etc.), gathering many types of intelligence (financial reports, wiretaps, etc.) for many purposes (disrupting activities, accumulating evidence to support a conviction, understanding modus operandi, etc.). This research is developing and evaluating a process to improve the allocation of intelligence collection assets. Using an approach based upon probabilistic risk analysis and integer programming, we have developed an initial decision framework to quantify the value of assigning a specific asset to a specific crime. Next steps involve validating this decision framework with law enforcement practitioners to ensure that the decision realistically reflects law enforcement processes. This will allow us to demonstrate proof of concept for this approach, both for consideration of homeland security concerns within law enforcement priorities, and to be extendable to a wider range of intelligence allocation contexts.

T4-C.3  15:30  Response reliability as a measure of preparedness for disaster response. Jackson JB*, Goulka J, Sullivan K, Willis HH; RAND Corporation

Abstract: In the years since September 11, 2001, the question “is the United States sufficiently prepared for future natural disasters or terrorist attacks?” has been prominent in national policy debate. Though a variety of efforts have been made to assess preparedness, these methods cannot answer the fundamental question of policymakers and the public: How certain should we be that the systems we have put in place to respond to damaging events will be able to deliver when called upon in the future? To address this concern about response performance, we are exploring the concept of response reliability, a measure of the confidence that response operations will go as expected at a future incident. The assessment relies on defining the nature of the organizational system that will be relied upon at an incident and, via a failure mode/fault tree analysis, making an assessment of the likelihood of events that would threaten response performance. In this pilot effort, we are examining the response to a major chlorine release. Drawing on the both response literature, after-action reports of past response operations, and expert elicitation, potential failure modes will be identified and a prototype assessment performed to explore the utility of this approach for preparedness assessment.

T4-C.4  15:30  A game theoretic approach for randomization in security: A report from the trenches. Tambe M*, Kiekintveld M, Taylor M, Pita J, Ordonez F; University of Southern California

Abstract: Security at major locations of economic or political importance is a key concern around the world, particularly given the threat of terrorism. Limited security resources prevent full security coverage at all times, which allows adversaries to observe and exploit patterns in selective patrolling or monitoring, e.g. they can plan an attack avoiding existing patrols. Hence, randomized patrolling or monitoring is important, but randomization does not mean haphazard operations --- we must still provide security guarantees. To that end, randomization must consider the differences in priorities of targets requiring security coverage, the responses of the adversaries to the security posture and potential uncertainty over the types of adversaries faced. To that end, we provide a game-theoretic approach to security randomization. Casting the problem as a Bayesian Stackelberg game and solving such games allows us provide randomized patrolling or inspection strategies that address differences in target weights, different adversary types and adversary responses to the security posture. The resulting algorithms have been successfully deployed in the form of ARMOR (Assistant for Randomized Monitoring over Routes) since August 2007 at the Los Angeles International Airport (LAX). ARMOR is a software scheduler that randomizes checkpoints on the roadways entering the airport and canine patrol routes within the airport terminals. Another scheduler for randomized deployment called IRIS for the Federal Air Marshals Service (FAMS) required significant scale-up in underlying algorithms, and is currently undergoing evaluation by the FAMS. In this presentation, we discuss the underlying algorithmic techniques, experiences from our deployment, and the significant effort focused on evaluation. We will also discuss continued evolution into newer applications, including at the port of Los Angeles and others.

T4-C.5  15:30  CTRA medical mitigation model: Assessing the benefits of the public health response. Good K*, Montello B, Von Niederhausern M, Hawkins B; Battelle Memorial Institute

Abstract: A model for estimating the benefits of the public health response to injuries resulting from a chemical terrorism attack on the general public was developed as part of the Department of Homeland Security Chemical Security Analysis Center 2010 Chemical Terrorism Risk Assessment. The resulting model applies a stock-and-flow modeling approach, similar to that used in Susceptible-Exposed-Infected-Recovered (SEIR) models of the public health response to biological outbreaks. The layering of treatment and use of the best available treatment options is captured using a tiered treatment concept, with each tier containing numerous treatment options based on the victim’s injury level. Four levels of injury are considered, ranging in severity from lethal to mild, as well as a “worried well” population. The model is able to handle uncertainty in the countermeasure efficacy versus time-to-treatment relationships in order to better reflect the current state of data. This model can be used to quantify the benefits to victims (e.g., number of lives saved and number of injured victims who benefitted from timely treatment) of a chemical terrorism attack based on the actions of first responders, medical personnel, and state, local, and national authorities, as well as examine potential mitigation strategies. An overview of the model will be presented.

T4-C.6  15:30  The Chemical Terrorism Risk Assessment: A Biannual Assessment of Risk to the Nation. Roszell LE*, Cox J, Whitmire M; Department of Homeland Security

Abstract: The Department of Homeland Security's Chemical Security Analysis Center (CSAC) is responsible for science based knowledge management and characterization of chemical risk to the nation from a terrorist event. In this capacity, the CSAC has been tasked with conducting a biannual assessment of risk associated with such an event. In 2008 the CSAC published the first end-to-end Chemical Terrorism Risk Assessment (CTRA). The CTRA is a probablilistic risk assessment that allows the threat, vulnerability, consequences, mitigation techniques and their associated uncertainties to be processed together to yield a comprehensive risk to the nation for the compounds of concern. In 2008 the risk from 57 chemicals was assessed; in 2010 these same chemicals will be assessed, as well as an additional 43, for a total of 100 chemicals. These chemicals include chemical warfare agents, toxic industrial chemicals, and other chemicals of high concern. The results from the CTRA aids policy makers and other officials in making risk informed decisions regarding detectors, countermeasures, consequence management plans and capabilities as well as identify knowledge gaps. This presentation will give an overview of the methodology utilized to conduct the 2008 CTRA, as well as modifications and improvements that will be included in the 2010 CTRA.

T4-C.7  15:30  The Chemical Infrastructure Risk Assessment: Assessment of Risk to the Chemical Supply Chain. Roszell LE, Gooding R*, Kolakowski J; Department of Homeland Security

Abstract: The Department of Homeland Security's Chemical Security Analysis Center (CSAC) is responsible for science based knowledge management and characterization of chemical risk to the nation.The CSAC has initiated the Chemical Infrastructure Risk Assessment, a study of the acute risk to human health and the national economy due to a terrorist initiated event within the US chemical supply chain. A probabilistic risk assessment of events that could occur at a chemical plant making, using or storing highly toxic industrial compounds or toxic industrial materials, is being conducted. Representative facilities have been developed to capture the breadth of the U.S. Chemical Industry, and serve as the basis for scenario and event tree development. The event tree considers a number of possible types of terrorist groups, target selection, and chemical or chemical classes. The impacts of interdiction, medical mitigation and event mitigation are also considered as part of the overall risk and economic impact. This study is currently underway; and the methodology and scope will be presented, as well as information on how the study results can be utilized by decision makers.

T4-C.8  15:30  CTRA foodborne contamination consequence model: Assessing health consequences of a foodborne chemical terrorism attack. Min S*, Luedeke J, Knebel N, Hawkins B; BATTELLE MEMORIAL INSTITUTE

Abstract: A model for estimating the human health consequences of a foodborne chemical terrorism attack was developed as part of the Department of Homeland Security Chemical Security Analysis Center 2010 Chemical Terrorism Risk Assessment. Specifically, a stock-and-flow modeling approach was used to mathematically simulate foodborne scenarios from the point of contamination (e.g., a storage tank), through food processing (e.g., pasteurization), packaging or bottling, distribution to retail outlets, sales of contaminated product, and consumption. The possibility of intervention is considered in the form of a recall announcement, using a recall timing model based on data gathered from literature on thirty-five past foodborne outbreaks and associated recalls involving foodborne contamination with both chemical and biological agents. The model is able to consider the variability associated with production batch sizes, processing parameters, package sizes, distribution times and temperatures, retail storage times and temperatures, purchasing habits, consumption timing, and serving sizes. Additionally, the model is able to consider the uncertainty regarding recall timelines and recall effectiveness. With the resulting model, contaminant-specific data, such as dose-response data, hydrolysis rate data, and time to symptom onset, can be used to provide estimates of potential consequences in terms of deaths and injuries of varying severity. An overview of the model and discussion of current and potential future applications will be presented.

[back to schedule]