The Ratio of Rural to Urban People Killed in Earthquakes Needs to be Assessed for Countries Separately, the Example of Colombia

Abstract

Abstract The hypothesis that very large earthquakes kill predominantly rural people is tested for the case of Colombia. For models of the eight largest earthquakes that have occurred in Colombia, the hypothetical ratios of rural to urban fatalities are calculated. Choosing these historic ruptures does not mean that they are expected to reoccur soon; instead, they are selected to sample fatalities in realistic models. The fatalities due to the assumed earthquakes are calculated by the tool Quake Loss Alerts for Rescue and Mitigation (QLARM), using the population data from the 2018 Columbian census. The minimum population for a settlement to be classified as urban is 35,000. The results do not depend on this limit. The condition for a model calculation to be accepted is that it matches the maximum intensities of shaking reported in the literature for the earthquake in question. Of the eight hypothetical earthquakes, today four would predominantly kill rural people, three would be urban, and one is about evenly split. However, the sum of the hypothetical fatalities in the eight test earthquakes is 79% urban and only 21% rural. This means that the observation that worldwide more than 90% of earthquake fatalities are rural does not apply to Colombia. The reasons for this contrast are that on average the Colombian test earthquakes are only about half as long as the ruptures modeled in the worldwide sample. Relatively short ruptures (75 km on average in Colombia) can rip through industrial areas without many villages because Colombia is strongly industrialized. By implication, this result means that the observation that most large earthquakes kill mostly rural people cannot be applied universally. The likely ratio of rural to urban earthquake fatalities has to be determined for each country separately for the assessment of the ratio of mitigation efforts to be allocated locally.