In this paper a simulation approach to modeling the linkages between physical infrastructure systems and the urban economy is developed. A simulation approach based on probabilistically specifying the key model relationships is effective for situations that involve substantial uncertainty, and is particularly suited to assessing risk from natural hazards. In this paper, a model of economic losses from earthquakes is developed and applied to the Memphis, Tennessee, region of the United States. We focus on water as a critical infrastructure service supporting the urban economy. The methodological approach involves systems integration of natural-science, engineering, and social-science databases and models. The concept of infrastructure services provides the linchpin in this integration process. Key spatial, temporal, and functional dimensions of infrastructure services are explicitly modeled in the simulation framework. The resulting model permits the analyst to compare the effectiveness of alternative actions, including both predisaster mitigation and postdisaster emergency-response activities. The model is calibrated in part with data from the 1994 Northridge and 1995 Kobe earthquakes. Results for several scenario earthquakes indicate the likely range of loss from economic disruption as well as uncertainties associated with the loss estimates. Sensitivity analysis indicates that one type of risk-management strategy for the water system, retrofitting pump stations, appears to be highly effective in reducing expected losses from future disasters.
