Bayesian probabilities for Mw 9.0+ earthquakes in the Aleutian Islands from a regionally scaled global rate. Issue 5 (17th May 2016)

• Record Type: Journal Article
• Authors: Butler, Rhett; Frazer, L. Neil; Templeton, William J.
• Is Part Of: Journal of geophysical research. Volume 121:Issue 5(2016:May)
• Page Start: 3586
• Page End: 3608
• Publication Date: 2016-05-17
• Subjects: Bayesian -- Aleutian -- paleotsunami -- Mw 9 -- paleoseismology; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1002/2016JB012861
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: We use the global rate of Mw ≥ 9.0 earthquakes, and standard Bayesian procedures, to estimate the probability of such mega events in the Aleutian Islands, where they pose a significant risk to Hawaii. We find that the probability of such an earthquake along the Aleutians island arc is 6.5% to 12% over the next 50 years (50% credibility interval) and that the annualized risk to Hawai'i is about $30 M. Our method (the regionally scaled global rate method or RSGR) is to scale the global rate of Mw 9.0+ events in proportion to the fraction of global subduction (units of area per year) that takes place in the Aleutians. The RSGR method assumes that Mw 9.0+ events are a Poisson process with a rate that is both globally and regionally stationary on the time scale of centuries, and it follows the principle of Burbidge et al. (2008) who used the product of fault length and convergence rate, i.e., the area being subducted per annum, to scale the Poisson rate for the GSS to sections of the Indonesian subduction zone. Before applying RSGR to the Aleutians, we first apply it to five other regions of the global subduction system where its rate predictions can be compared with those from paleotsunami, paleoseismic, and geoarcheology data. To obtain regional rates from paleodata, we give a closed‐form solution for the probability density function of the Poisson rate when event count and observation time are both uncertain. Key Points: Bayesian probabilities are derived for estimating the recurrence of great Mw 9 earthquakes Fault length and convergence vector may serve as proxy where paleodata are lacking The credibility interval for a Mw 9.0+ Aleutian earthquake in the next 50 years is 6.5% to 12%