After Tohoku: Re-evaluating Japanese Earthquake Hazard

Japanese Hazard Map_Subduction-zones

The magnitude of the 2011 Tohoku earthquake came largely as a surprise to the seismologist community and revealed certain shortcomings in previous hazard studies. Helping to address this, in 2012 and 2013, the Headquarters for Earthquake Research Promotion (HERP) in Japan released provisional studies incorporating new research findings and in December 2014 issued the new National Seismic Hazard Maps. These maps and the accompanying report, currently only available in Japanese, furnish important context for the offering that commercial modeling companies (such as RMS and AIR) provide with regards to Japanese earthquake.

Key Updates

What are the main changes in the new HERP hazard maps and report?

  • Re-evaluation of the magnitude and long-term probabilities of several earthquakes (for example, those triggered along the Sagami Trough) on the basis of new data and considering potentially larger uncertainties
  • Increases to potential magnitude of earthquakes triggered by unknown source faults (such as the 1968 Hyuga-nada Plate Earthquake and the Yonaguni Island Earthquake of 1998)
  • Expansion and clarification of explanations regarding the fundamental principles, methods of probability evaluation, earthquake category, etc.

In addition, special attention has been paid to the Tohoku-Oki earthquake. At this point, HERP estimates a maximum magnitude of 9.0 with an average recurrence interval of 600 years; they see negligible probability of reoccurrence within the next 50 years.

How do the Vendor Models Compare?

Both AIR and RMS rely heavily on HERP analysis to delineate and characterize their seismic source models. But the two models were last updated in 2013 and 2012, respectively. Given the increased magnitudes of HERP’s 2014 report, it is reasonable to ask whether the current vendor models sufficiently contemplate the risk.

Our Model Research and Evaluation team investigated this question and determined that for the seismic sources of most interest, i.e. Eastern Japan, the model vendors have incorporated sufficient uncertainty and generally include events in their catalogues with magnitude potential equal to or larger than what is assumed in the 2014 HERP report. This is not the case for the Sagami Trough, where the vendors appear to be using lower magnitudes. However, the probability that HERP assigns to this event is very low.

Business Implications

Overall, we find that the most important changes in the 2014 HERP hazard maps are the further uncertainties considered and the increased maximum magnitude of earthquakes triggered by subduction zones and unknown source faults. For the most part, existing models already include events of equal or greater magnitude and we see no need for vendor model adjustments at this time.



Myrto Papaspiliou, Lin Ke, and John Alarcon Guest bloggers Myrto Papaspiliou, Lin Ke, and John Alarcon are with the Model Research and Evaluation (MR&E) team of Willis Re International and Specialties. The MR&E team is in charge of assessing, comparing, validating and adjusting catastrophe model vendors (AIR, EQE, ERN, RMS) for all perils/territories in the international business. Dr. Myrto Papaspiliou joined Willis in October 2012 and is a Senior Earthquake Research Analyst. Dr Lin Ke joined Willis Re Japan K.K. in January 2015 as a Catastrophe Risk Analyst. Dr. John E. Alarcon is Executive Director in Willis Re and head of the MR&E team.

Categories: Analytics, Natural Catastrophe | Tags: ,

Leave a Reply

Your email address will not be published. Required fields are marked *