Evaluation of Tsunami Sources with the Potential to Impact the U.S. Atlantic and Gulf Coasts: An Updated Report to the Nuclear Regulatory Commission

The 2004 Sumatra tsunami, which took place in an area with no historical
record of a similar event, has brought awareness to the possibility of
tsunamis along the U.S. Atlantic and Gulf of Mexico coasts. While these rare
events may not have an impact on tsunami probability calculations for flood
insurance rates, they need to be considered in long-range planning, such as
for the placement of nuclear power plants. The U.S. Geological Survey was
tasked by the Nuclear Regulatory Commission to prepare an evaluation of
tsunami sources and their probability to impact the U.S. Atlantic and Gulf of
Mexico coasts. This report is an updated evaluation based on additional data
analysis and modeling. It provides a general review of potential tsunami
sources, and provides a geotechnical analysis and hydrodynamic model for
one landslide offshore North Carolina. The evaluation also identifies
geographical areas with limited information and topic for further study.
Finally, the updated report present new theoretical developments that may
aid in quantitative evaluation of tsunami probability.
The work included in this report represents both review of published
work and original work. Eight of the 14 topical chapters in this report
represent original work, and the remaining 6 chapters are based on literature
reviews. The original work is in the process of being published as eight
papers in a special issue of Marine Geology, an international peer-reviewed
scientific journal.
The main findings of the study so far include:
1. Landslides along the U.S. Atlantic margin have the potential to cause
tsunami locally. These landslides are concentrated along the New
England and Long Island sections of the margin, outward of major
ancient rivers in the mid-Atlantic margin and in the salt dome
province offshore North Carolina. The landslides generally removed
a relatively thin (a few 10s of meters) layer of mostly unconsolidated
sediments. The mapped landslides follow a log-normal size
distribution centered at a volume of about 1 km3
. However, some parts of the upper continental slope, particularly off Long Island New England, have not yet been mapped in detail. Relatively few large landslides from the entire mapped inventory (9 landslides over 500
km2 and 16 landslides over 10 km3) could have caused a damaging
tsunami. The criteria for devastating tsunami is presently based solely
on modeling of the Currituck slide offshore North Carolina.
Additional simulations off New England are needed to refine the
threshold for damaging tsunamis. A review of known ages of
submarine landslides around the Atlantic Ocean and worldwide
ii
shows a factor of 1.7-3.5 lower frequency of occurrence during the
past 5000 years relative to the last glacial period and the first 5000
years after the end of glaciation, suggesting that the majority of the
observed landslides along the U.S. Atlantic margin are older than
5000 years.
2. Earthquake sources that can generate trans-oceanic tsunamis, are
located west of Gibraltar and in the Puerto Rico trench. Tsunami
simulations from the 1755 Lisbon earthquake show that seafloor
topography between the source area and the Azores Islands plays a
major role in scattering the wave energy traveling toward the U.S.
East Coast. This conclusion matches the lack of tsunami reports
from parts of the U.S. East coast that were populated at the time
(Boston, New York, Chesapeake Bay, Savannah, Charleston).
However, simulations show that should a large tsunamigenic
earthquake take place in the Puerto Rico trench, the resultant tusnami
may be destructive to many parts of the U.S. East Coast. To date, no
evidence was found for large historical or pre-historical earthquakes
in the Puerto Rico trench, and the ability of this plate boundary to
generate large earthquakes is being investigated.
3. Far-field landslides, such as in the Canary Islands, are not expected to
cause a devastating tsunami along the U.S. Atlantic coast.
4. Large landslides in the Gulf of Mexico are found in the submarine
canyon and fan provinces extending from present Mississippi and
other former larger rivers that emptied into the Gulf. These large
landslides were probably active before 7500 years ago. In other areas,
landslides continue to be active, probably because of salt movement,
but are small and may not pose tsunami hazard. Very little is known
about the threat of landslide-generated tsunamis from the Mexican
coast, particularly the Campeche escarpment. Tsunamis generated by
earthquakes do not appear to impact the Gulf of Mexico coast.
5. Several approaches to quantitatively calculate the probability of
tsunamis impacting the U.S. East and Gulf Coasts have been
developed, but their accuracy depends in large part on the available
of observations of size distribution, recurrence interval, and
geotechnical parameters of the sea floor.