Probabilistic tsunami hazard assessment of El Salvador

Student: Beatriz Brizuela
Supervisors: Dr Carlo Lai, Dr Mario Fernandez, Dr Carlos Pullinger

Abstract

El Salvador can be catalogued as a potential tsunamigenic country. Being in the Pacific, where the majority of tsunamis have been recorded, the country can be considered likely to be hit by a tsunami. This thesis illustrates the tsunami hazard potential in El Salvador. A theoretical frame containing general concepts and definitions, geological settings of El Salvador and previous studies related to tsunamis in Central America and El Salvador is presented. Historical tsunami information contained in the Central America tsunami catalogue and more specifically in the Salvadorian tsunami catalogue was gathered and levels of accuracy and reliability of both catalogues were established. The information contained in the Salvadorian tsunami catalogue was used to perform a probabilistic tsunami hazard assessment of El Salvador after estimating the completeness of the catalogue. Based on historical information, two locations at the Salvadorian coast (Acajutla and La Union) were identified as tsunami hazard scenarios. Maximum run-up values for return periods of 250 and 475 years were estimated at each location. Logic tree analysis was performed to associate some level of likelihood of those estimated run-ups being correct. Estimated run-up values at Acajutla are 9.62 and 12.08 meters for return periods of 250 and 475 years respectively whereas at la Union run-up values are expected to reach 3.70 and 4.68 meters for the same return periods. According to the results obtained from the analyses, destructive tsunamis are likely to hit El Salvador, therefore a tsunami warning system, evacuation plan and tsunami educational programs are needed in order to avoid high death tolls if a tsunami occurs. Tsunami warning systems that are currently operating around the world are described. Some conclusions regarding to tsunamicity of the country obtained from the analyses and recommendations for future research are made.

You may download a digital version of this MSc dissertation here.