The map showing the Aleutians with respect to Hawai’i. The red and yellow arcs indicate the sections of the Aleutian subduction zones considered in the probability analysis. Stars and dates indicate epicenters of prior 20th century great earthquakes (Mw > 8). Credit: Butler et al., 2016.

The map showing the Aleutians with respect to Hawai‘i. The red and yellow arcs indicate the sections of the Aleutian subduction zones considered in the probability analysis. Stars and dates indicate epicenters of prior 20th century great earthquakes (Mw > 8). Credit: Butler et al., 2016.

A team of researchers from the University of Hawai‘i at Mānoa published a study last week that estimates the probability of a Magnitude 9-plus earthquake in the Aleutian Islands — an event with sufficient power to create a mega-tsunami especially threatening to Hawai‘i, according to a UH news release.

In the next 50 years, they report, there is a 9 percent chance of such an event. An earlier State of Hawai‘i report estimates the damage from such an event would be nearly $40 billion, with more than 300,000 people affected.

Earth’s crust is composed of numerous rocky plates. An earthquake occurs when two sections of crust suddenly slip past one another. The surface where they slip is called the fault, and the system of faults comprises a subduction zone. Hawai‘i is especially vulnerable to a tsunami created by an earthquake in the subduction zone of the Aleutian Islands.

“Necessity is the mother of invention,” said Rhett Butler, lead author and geophysicist at the UH School of Ocean and Earth Science and Technology. “Having no recorded history of mega tsunamis in Hawai‘i, and given the tsunami threat to Hawai‘i, we devised a model for Magnitude 9 earthquake rates following upon the insightful work of David Burbidge and others.”

Butler and co-authors Neil Frazer (UH SOEST) and William Templeton (now at Portland State University) created a numerical model based only upon the basics of plate tectonics: fault length and plate convergence rate, handling uncertainties in the data with Bayesian techniques.