Risk-Based Decisionmaking in Water Resources X

This paper addresses disaster planning for infrastructures and their related societal entities through innovative modeling that coordinates multiple networks against the contingency scenarios that disrupt plans for mitigation, preparedness, response, and recovery. Across disciplines, interdependent infrastructures and related societal entities are increasingly amenable to characterization as mathematical networks. Effective disaster planning for complex systems is known to rely on the identification and management of a multitude of diverse contingency scenarios. Modeling of such contingencies in disaster planning for interdependent infrastructures and societal entities can proceed with a superposition of multiple network models. We study the interactions among contingency scenarios and diverse superposed networks. The developed theory and methodology will guide the multidisciplinary selection and management of relevant worst-case scenarios, an essential activity for disaster planning in large-scale, complex systems.

The fuzzy logic formulation of risk analysis is presented with application to a flood risk management case. If uncertainty in any element of risk analysis (exposure, resistance, and consequences) is expressed as a fuzzy set, the corresponding risk will also be calculated as a fuzzy set. The probabilistic formulation of risk analysis has several difficulties, including: the management of the low-probability/high-consequence case, the selection of the probability models, the scarcity of statistically meaningful data, the uncertainty of the consequence functions, and the covariances involved. Fuzzy logic formulation, as an alternative to statistical methods to define uncertainties, is applicable when all these difficulties must be dealt with simultaneously. A short review of the application of fuzzy logic illustrates the versatility of this approach. A simplified flood risk management example illustrates a case where probabilistic and fuzzy uncertainties are involved in exposure, resistance, and consequences; economic and ecological consequences are considered, and a tradeoff analysis can identify the preferable action.

The risk assessment, modeling, and simulation of critical infrastructure information technology (IT) security has been limited to broad, macro-level approaches. Concurrently, risk assessment in IT security has been limited to static analysis and modeling. This paper provides a dynamic risk framework and a model that synthesizes elements of an organizational decision model on both macro and micro levels. In the proposed dynamic model the focus is upon building the reactive capability of an infrastructure organization as well as preparing for time-based cascading effects.

One of this nation's largest ecosystem restoration projects is currently taking place in South Florida. The Comprehensive Everglades Restoration Project is a complex and expensive one, requiring a sequence of decisions over the next three decades in an attempt to preserve and enhance what remains of the Greater Everglades Region of South Florida. There is considerable uncertainty with respect to how this ecosystem functions and what decisions will be most effective in restoring it. This paper outlines some of the issues associated with estimating, quantifying, and communicating these uncertainties to various stakeholders, and to those who are responsible for decisionmaking.

Managing the risks posed by an intelligent and adaptable adversary is different from many other types of risk management. Thus, risk management in this context is a problem of game theory rather than decision theory. In other words, the defender wishes to choose the optimal strategy for defending against an optimal attack, and vice versa. In this paper, we apply game theory to help in characterizing optimal defensive strategies against intentional attacks. The results yield useful insights.

The purpose of this paper is to point out that the general theory of quantitative risk assessment (QRA) applies perfectly well to evaluating and quantifying the risk from terrorism. The main difference occurs during the "scenario identification" part of the risk assessment process. Whereas, in an "ordinary" QRA, we ask the question, "What can go wrong?," in terrorism risk assessment (TQRA) we ask, "If I wanted to, what could I make go wrong?" In answering this new question, the Theory of Scenario Structuring and the use of fault and event trees play the major roles as before. Also, the concept of "resources" now moves to center stage as part of the process of identifying terrorism scenarios. So also does the use of Bayes' theorem, not only to assess a priori the likelihoods of specific terrorism scenarios, but also as a crucial part of surveillance systems that have, potentially, the ability to quantify the likelihoods that such scenarios are in process.

The vulnerability of the nation's water infrastructure to terrorist threats is discussed, and the differences and similarities in the vulnerabilities to acts of terror and to acts of nature are noted. Haimes et al proposed hardening water resource systems by adding or augmenting security systems and by enhancing the redundancy, robustness, and resiliency of the systems. A general form of hardness perhaps may be given to the systems through a national water board that would coordinate the hardening activities of federal agencies and their responses to crises. Decisions depend on future expectations. The decision made to harden the infrastructure will be made in a state of uncertainty, given that the future cannot be clearly comprehended. The traditional presumption that uncertainty derives from imperfect knowledge and therefore can be addressed through the theory of probability may not be sufficient in the face of terrorist threats. Some decisions might be made in a state of uncertainty that derives from incomplete knowledge; this may be addressed through a theory of surprise.

The purpose of this paper is to provide an in-progress review of the development of a systems-based vulnerability assessment methodology for critical infrastructures. This methodology is a supporting framework for the Infrastructure Risk Analysis Model (IRAM), where vulnerability assessment was briefly discussed but not fully developed by the authors. This paper focuses exclusively on water supply as the complex organizational system in focus. However, further research, deployment, and refinement of the vulnerability methodology may prove the applicability of the framework for military installations and electric power, among other critical infrastructures. This paper discloses three main ideas. First, there has been very little published in the way of rigorous vulnerability assessment methodologies. In fact, there is no agreed-upon definition of vulnerability. Second, vulnerability assessment appears to be an ad hoc checklist of things-to-do, bereft of agreement regarding the system in focus, i.e., the boundaries and context of a water supply system. Last, the paper argues that a water supply system should be viewed as a complex organizational system. The implication of this paper is that classic risk assessment questions (What can go wrong? What is the likelihood? and What are the consequences?) should be preceded by the primal question: What is the system in focus? The system in focus must be understood before meaningful risk and vulnerability assessment is undertaken.

The paper discusses the demand-reduction I-O inoperability model to analyze the economic impact of demand reductions on a system of interconnected infrastructures. The propagation of inoperability depends upon the degree of interdependency of one infrastructure on another. A case study comprising twelve critical interconnected infrastructure sectors is presented. The interdependency matrix for this case study is derived via a transformation of the national industry-by-industry transactions data as published by the Bureau of Economic Analysis. The paper highlights the use of geographical and functional decompositions to tailor the demand-reduction analysis to specific US regions, or to components of a large-scale infrastructure, respectively.

Threats to water resources from terrorists are fundamentally different from threats from hurricanes and other natural phenomena because terrorists can choose their targets. As a result, effective approaches to disaster prevention should focus more on identifying and apprehending terrorists rather than reducing the vulnerability of water systems. Certain popular belief systems, however, are inconsistent with expanded government use of methods designed to spot terrorists, such as sensors, surveillance, and data mining. Progress in reducing risks requires wise organizational responses to these widespread concerns.

This paper describes a modular geographical information system (GIS) model that is being developed for estimating potential loss of life from natural and dam-failure floods. The model can be used to provide life-loss estimates for use in dam safety risk assessments, including dam failure caused by terrorism. It can be used to explore options for reducing life loss by reducing the probability of dam failure or by improving the effectiveness of emergency planning and response. The modeling system comprises the following internal modules: 1) loss of shelter, including prediction of structural performance of buildings, 2) warning and evacuation, and 3) loss of life based on empirical relationships developed from a wide range of case histories.

It is important to think of the future in a more comprehensive way. The present high-level water policies have good intentions, but little follow-through, giving them a Sysiphean aspect. We need a broad guiding national philosophy directly related to the principles of integrated water management (IWRM) and sustainable development (SD). Both IWRM and SD require a uniform set of planning and management principles, consistently applied. A national water policy would be a good starting-point. In the absence of a central federal water infrastructure agency or an interagency council, a President's Water Council may be the needed interim solution. There is a Pandora's box of current difficulties in federal water management, but we can always have hope that among the changes that are surely coming we will have positive solutions.

The terrorist attacks of September 11, 2001 revealed the vast scope of our vulnerability to terrorism. The reaction, and even overreaction, to these threats has led to costly disruption of both infrastructure and commerce, as the chaos in the airline industry so vividly displayed. A systematic framework is needed for analyzing the terrorist threat realistically and for coordinating the interaction among the responding organizations as well as the public. What has yet to be assimilated about this risk from terrorism is the human variable. Whether the source of terrorism is international like al Qaeda or domestic like Timothy McVeigh, terrorists learn from experience and alter their tactics constantly. Accordingly, the measures necessary to protect people and property must be constantly reevaluated and revised, as in a never-ending "game." To be predictive in this environment, new tools are needed to properly assess the vulnerability to terrorism of systems and assets. New methods are required for determining who and what is at risk, how great the risk is, what measures can be taken to minimize that risk, and at what costs. In this paper, a simple game structure is described.

On the last evening of each conference, a questionnaire was distributed to the participants. Recognizing that these are the tenth such proceedings, we have compiled the responses to the same three questions from all ten conferences. We have begun analyzing the trends that emerged, and the results should make an important contribution to the field of risk-based decisionmaking in water resources. The following questions were asked at each conference: (1) List the three most important issues/aspects/elements related to risk-based decisionmaking that were raised during this conference; (2) List the three most important new ideas/concepts that you learned during this conference which would be helpful in your job; and (3) List the three most important issues needing further study in risk-based decisionmaking in water resources.