Geoscience Reference
In-Depth Information
Like northern California, the Oregon and open-ocean Washington coastlines are threat-
ened by tsunamis generated by the Cascadia subduction zone earthquakes, and the different
modeling approaches taken by each state illustrate the disparities in tsunami source determi-
nations. Inundation modeling for the open-ocean coast of Washington related to a Cascadia
scenario is based on tsunami source determinations conducted over 10 years ago in Oregon
(Priest, 1997). Modeling efforts for the Oregon coast related to Cascadia scenarios are based on
newer methods and source determinations (González et al., 2006; Priest et al., 2009), but those
in the state of Washington were not. Input received by the committee from NTHMP representa-
tives from Oregon and Washington revealed differences in opinions on the costs and beneits
of newer modeling approaches. Oregon representatives believed improvements in inundation
modeling approaches are essential for hazards assessments, whereas Washington representa-
tives consider the existing maps to be suficient and instead focus NTHMP funds on education
and other preparedness efforts. The committee cannot fully comment on which philosophy is
more appropriate because there has not been adequate discussion within the NTHMP on the
relative costs, beneits, and utility of higher order models compared to existing methods. Until
these discussions occur, individual states will likely continue their independent approaches. In
addition, not all these source determinations are subject to rigorous peer-review. One conse-
quence of these independent efforts is that different tsunami-source determinations in two
states for the same tsunami scenario (e.g., Cascadia) complicate national efforts to compare
societal risk from tsunamis.
With so many unknowns, what overall strategies are likely to deine the sources of
tsunamis that threaten the United States the most? One strategy is to focus on sources for
near-ield tsunamis. This focus is suitable for Alaska, the Cascadia coast of Washington, Oregon,
and northern California, and the Caribbean. It can be justiied by existing simulations of
tsunami inundation that show greater looding from near-ield tsunamis than from far-ield
tsunamis in Alaska, Oregon, and northern California (Appendix C). It can also be justiied by the
importance of public education about tsunami hazards as a means of saving lives from near-
ield tsunamis (Chapter 3). Yet another reason to focus on near-ield tsunamis is their sensitivity
to local properties of their sources. These include the coastwise extent, depth range, and local-
ized concentration of slip on a fault plane, and the orientation and speed of a landslide.
A second overall strategy is to look further back in time. A tsunami hazard assessment
based on U.S. written history alone would overlook nearly all the tsunami hazards from earth-
quakes on the Cascadia subduction zone and the Seattle fault, and from most of the land-
slides off southern California and the U.S. Atlantic coast. Time intervals, or recurrence intervals,
between earthquakes causing catastrophic tsunamis can be as long as centuries or millennia.
The 2004 Indian Ocean tsunami, for example, seemingly lacked historical precedent because
its most recent predecessor occurred 550-700 years ago (Jankaew et al., 2008). The preliminary
hazard assessment by Dunbar and Weaver (2008) explicitly recognizes this issue and addresses
it partially by drawing on national probabilistic seismic hazard mapping that incorporates
earthquake histories, inevitably incomplete, that are drawn from geological records that span
centuries or even millennia.
