Spatial Analysis of Earthquake Fatalities in the Middle East, 1970-2008: First Results
Abstract
Earthquakes claim the lives of thousands of people each year and the annual number of earthquake fatalities in the Middle East (21 countries) is 20 % of the total yearly fatalities of the World. There have been several attempts to estimate the number of fatalities in a given earthquake. We review the results of previous attempts and present an estimation of fatalities using a new conceptual model for life loss that includes hazard (earthquake magnitude and focal depth), vulnerability (GDP value of countries and elapsed time since 1970 as proxy variables) and exposed population in the affected area of a given earthquake. PAGER_CAT is a global catalog (http://earthquake.usgs.gov/research/data/pager/) that presents information on casualties of earthquakes since 1900. Although, the catalog itself is almost a complete record of fatal earthquakes, the data on number of deaths is not complete. We use PAGER_CAT to assemble a Middle East (the latitude and longitude of 10°-42° N and 24°-64° E respectively) catalog for the period 1970-2008 that includes 202 events with published number of fatalities, including events with zero casualties. We investigated the effect of components of each event, e.g. exposed population, instrumental earthquake magnitude, focal depth, date (year of event) and GDP on earthquake fatalities in Middle East in the 202 events with detailed fatality estimates. To estimate the number of people exposed to each event, we used a fatality threshold for peak ground acceleration of 0.1g to calculate the radius of affected area. The exposed population of each event is the enclosed population of each circle calculated from gridded population data available on SEDAC (http://sedac.ciesin.columbia.edu/gpw/global.jsp) using ArcGIS. Results of our statistical model, using Poisson regression in R statistical software, show that the number of fatalities due to earthquakes is in direct (positive) relation to the exposed population and the magnitude of the earthquake at the epicenter. On the other hand, it is in inverse (negative) relation to elapsed time since 1970, focal depth and GDP of the country affected. These spatial and temporal patterns of life loss are consistent with the patterns expected within our conceptual framework in relationship with hazard, exposed population and proxies of vulnerability. Our findings suggest that for earthquakes with comparable physical characteristics, the number of fatalities has been falling since 1970 in the Middle East region. We interpret this as an overall reduction of vulnerability of the Middle East during 1970-2008. Ongoing research is focusing on more detailed analysis of particular indicators of vulnerability reduction such as the development of earthquake building codes and preparedness, and on the spatial disaggregation of exposed population and the attenuation of earthquake magnitude.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFMNH12A..04K
- Keywords:
-
- 6300 POLICY SCIENCES;
- 6600 PUBLIC ISSUES;
- 7299 SEISMOLOGY / General or miscellaneous