Spatial Correlation of Losses: Impact of Earthquake and Tsunami Source Model Assumptions

Session: Forthcoming Updates of the USGS NSHMs: Hawaii, Conterminous U.S. and Alaska [Poster]
Type: Poster
Date: 4/28/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Post Tohoku, most countries have strived to ensure that they include the largest events that cannot be ruled out, as opposed to the largest event that has been observed plus some uncertainty. This included considering not only M9s on the Japan Trench, but also M9 on the Nankai trench, as well as a larger seismogenic area offshore Tokyo (Sagami trough, Chokkagata sources) and more recently M9 on the Kuril trench. Outside of Japan, M9s are considered for New Zealand (Hikurangi subduction) and Indonesia . Even for non-subduction countries, this “all that cannot be ruled out” perspective has led to the inclusion of very large multi-fault events (California). When such events are introduced, they change both the site-source distance and the partitioning of the seismic budget. Their geometrical aspects, which affect hazard and damage footprints, and the modelling of their recurrence impact the risk landscape of a country. For California, Fitzenz et al. 2018 looked at the impact of both segmentation assumptions and time-dependent modelling assumptions on the spatial correlations of expected losses in Bay Area and southern California counties. Loss spatial correlation has important implications for the willingness of an insurer to pursue business in a region.

When most of the large events newly introduced into a model can also cause tsunami losses, the question becomes more layered. What are the spatial correlations across coastal city wards in Japan? We present results obtained using the RMS High Definition Japan Earthquake and Tsunami model.

Such studies can help develop risk-based research priorities by pointing to where the need is the greatest to refine the physical characterization of the sources (lateral extent, up-dip and down-dip extent), recurrence model assumptions, slip distribution modelling, as well as bathymetry topography parameterization constraints that influence sea floor and coast deformation models. When equally plausible assumptions can be made that have very different consequences, none should be discarded.

Presenting Author: Delphine D. Fitzenz

Authors