HR: 1340h
AN: PA23A-1445    [Abstracts]
TI: Time-Varying Multi-Hazard Index Using Empirical Methods
AU: * Skorik, A
EM: acs2031@columbia.edu
AF: Columbia University, 2960 Broadway, New York, NY 10027 United States
AU: Isanuk, M
EM: mji2101@columbia.edu
AF: Columbia University, 2960 Broadway, New York, NY 10027 United States
AU: Lerner-Lam, A
EM: lerner@ldeo.columbia.edu
AF: Lamont-Doherty Earth Observatory, 61 Route 9W, Palisades, NY 10964 United States
AB: Natural hazards represent one of the greatest threats to the economic and social stability of developing countries. While many efforts have focused on studying and protecting against single hazards, these do not provide sufficient support for development decisions faced by countries at risk from multiple hazards. In order to improve resource allocation for the mitigation of multiple natural hazard impacts, we have developed a time-varying multiple hazard index using empirical methods. Our preliminary version of this index focuses on the four natural hazards with good historic data: earthquakes, floods, cyclones, and droughts. We create a statistic for each hazard computed from deviations away from observed "mean severity". For example, the mean earthquake severity is determined from average annual moment release in a specified geographic area (such as a country). Next, we compute the deviation of annual moment release by formulating a logarithmic statistic. A similar methodology involving an objective measure of hazard magnitudes and frequencies is used for the other hazards in our sample. We sum the individual hazard indices to form the time-varying multi-hazard index. Our early version of the Multi-Hazard Index has certain limitations. To be compatible with available socioeconomic data, we compute the index for whole countries regardless of the geographic scale of the hazard. This has the effect of oversmoothing the index for large countries. The use of sub national socioeconomic data (where available) would allow us to create the index on a smaller scale with improved accuracy. Furthermore, adjustments are needed in the weighting of different magnitude events of the same hazard type. Consideration also needs to be given to the weighting of the different individual hazard indices in the multi-hazard sum. We suggest several different approaches to these issues.
DE: 6699 General or miscellaneous
SC: Public Affairs [PA]
MN: 2004 AGU Fall Meeting