Determining with some precision when and where disasters are likely to strike is the focus of an Oak Ridge National Laboratory effort that ultimately could save lives. The project takes advantage of a combination of tools developed by a team of researchers headed by Auroop Ganguly of the Department of Energy lab’s Computational Sciences and Engineering Division. Ganguly believes the effort could have tremendous significance. “Our goal is to save lives by providing policy-makers with the information they need to anticipate and perhaps prevent catastrophic disasters,” Ganguly said. “The insights we’re providing can be acted upon to prevent the disasters from becoming catastrophic in terms of loss of lives and property or to the critical civil infrastructure of a nation.” The visual decision aids developed by Ganguly and colleagues are derived from sophisticated statistical methodologies combined with population distribution and climate models. For this study, scientists examined daily observations from 7,900 stations in South America from 1940 to 2004 available at 2.5-degree spatial grids. They paid special attention to the number of consecutive two- and three-day rainfall extremes and regions with upward trends in extremes – especially regions with dense populations. “What we did next was classify the extremes according to their volatility and variability,” said co-author George Ostrouchov, a mathematician in ORNL’s Computational Sciences and Mathematics Division. “By looking at extremes from the past, we were able to come up with a predictive tool that helps us set priorities according to the likelihood of an extreme event -- such as a 100-year or 200-year rainfall event -- combined with population distribution and gross domestic product.” Researchers focused on catastrophic disasters caused by rainfall extremes – particularly in the tropical and sub-tropical regions, which receive heavy rainfall in the summer and minimal rainfall in the winter. To develop their model, researchers took rainfall observations from the National Oceanic and Atmospheric Administration and combined that information with the LandScan Global Population Database, a unique high-resolution tool developed and maintained by ORNL, and the gross domestic product of each nation. The results were presented recently at a meeting of the American Geophysical Union in a session titled “Catastrophic risk from natural perils.” Given the information on what could be called potentially “catastrophic disasters,” policy-makers at world bodies like the United Nation or at federal agencies can determine how budgets should be allocated among individual nations or among local and state agencies, respectively. For example, money could be spent to develop advanced disaster warning or public education systems as well as disaster management infrastructures and evacuation strategies in areas determined to be the most vulnerable. Ganguly, however, emphasizes that the effort extends beyond responding to disasters. “The ability to develop predictive knowledge goes a step further than merely performing relief-and-rescue operations once a disaster has occurred,” Ganguly said. “Advances in disaster-readiness imply a cost to the taxpayer and allocation of funds to one region or country, usually at the expense of another. This is where our tools can prove especially useful for policy-making. “Specifically, while weather extremes may continue to occur and perhaps even grow more intense because of inherent climate variability and/or climate change, the risks and impacts from such extremes may be reduced or even eliminated in some situations through well-designed policies.” Researchers also noted that damages caused by extreme amounts of rainfall are influenced by a variety of factors other than just the amount of rain. These factors include surface and sub-surface hydrology because floods and flash floods cause most of the damage. The other key factors are population density and the ability of a region to respond. The team chose to study South America for a variety of reasons, including the availability of recent comprehensive rainfall data from NOAA, the lack of adequate prior research focused on South America and because the region is relatively geo-politically homogenous. The case study can be extended in a straightforward manner to other regions where similar data are available. While this study looked at rainfall, similar methodologies can in principle be applied to earthquakes, tsunamis and pandemics. Others involved in this research include Chris Fuller, Shiraj Khan, Gabriel Kuhn and Aarthy Sabesan of the Computational Sciences and Engineering Division, and David Erickson and Marcia Branstetter of the Computational Sciences and Mathematics Division. The research was funded by ORNL’s Laboratory Directed Research and Development program. UT-Battelle manages Oak Ridge National Laboratory for DOE.