IAG-IASPEI 2017

講演情報

Poster

Joint Symposia » J04. Geohazard early warning systems

[J04-P] Poster

2017年8月4日(金) 15:00 〜 16:00 Shinsho Hall (The KOBE Chamber of Commerce and Industry, 3F)

15:00 〜 16:00

[J04-P-09] Revisiting the 1985 M8.1 Michoacan earthquake: Tsunami simulations and synthetic GPS data to test rapid response

Angel Ruiz-Angulo1, Diego Melgar2, Carlos Mendoza3 (1.Centro de Ciencias de la Atmosfera, UNAM, Mexico City, Mexico, 2.UC Berkeley, California, USA, 3.Centro de Geociencias, UNAM, Juriquilla, Mexico)

The 1985 Michoacan earthquake was a catastrophic event for Mexico. In terms of lives it had a toll of 10,000-20,000 casualties and the total economic cost was more than 5 billion dollars. The earthquake generated a local tsunami, which at the time did not generate much attention since most of the damage from the event was due to strong shaking in Mexico City. However, the extent of the tsunami was significant in the coast of Michoacan. In the industrial port city of Lazaro Cardenas, wave heights reached up to 2.5 m, with inundation distances of about 500 m. A segment of railroad of 1.5 km was destroyed and several fishing boats were reported missing. In Zihuatanejo, the impact was very similar and several beach front restaurants were washed away. In spite of this, at that time, in 1985, urbanization and development of the coastline was still somewhat sparse. However, much has changed in the intervening 30 years. We argue that the impacts from a repeat of an event such as this one would be much more significant today. The current population along the Mexican Pacific coastline has increased considerably and infrastructure has developed as well. Many locations have already suffered from flooding events during hurricanes and tropical storms, suggesting a rather high risk in case of a tsunami.

We use a published finite fault inversion for the earthquake and its largest aftershock as the initial condition of tsunami simulations and study the impacts along the coastline. Furthermore we generate synthetic high-rate GPS data for stations that exist today and show that the magnitude and extent of the earthquake could have been known within the first 60s of the event origin time. We show that rapid tsunami models that rely on these fast earthquake source products can provide useful forecasts of tsunami intensity in the first minutes after a large event.