A Dozen Tsunamis from 2004 to 2014: Lessons and Revelations

Abstract:

The past 10 years have witnessed two truly catastrophic tsunamis (2004 and
2011), but also a dozen other ones, some significantly
damaging. Among the scientific lessons learnt from the Sumatra event, the most
important is the failure of the concept of maximum earthquake predictable
from simple tectonic parameters, resulting in the precautionary suggestion that
all subduction zones may conceivably support a mega earthquake, as strikingly
illustrated by the 2011 Tohoku event. A critical analysis of tsunami warning
strategies and population response in the past decade reveals
considerable progress in the far field (with only two or three
past-Sumatra casualties), but a relatively random record in the near field,
where successes (Solomon Islands 2007; Bengkulu 2007) are overshadowed by
outright failures, notably during the 2010 Mentawai "tsunami earthquake".
We review remaining challenges and progress towards their resolution, both
scientific (such as the real-time identification of "tsunami earthquakes" and
the understanding of wave sequencing in the far field), and societal (including
a post-mortem of the 2011 Fukushima disaster).
The exceptional size of the 2004 Sumatra earthquake (and to some
extent of following ones) has allowed the coupling of their tsunamis to
both the solid Earth and the atmosphere in non-traditional ways, offering the
possibility of detection by a remarkable variety of "incompetent" instruments
ranging from hydrophones to magnetometers and ionospheric probes. Beyond
the mere intellectual interest of these observations, they open the door
to possible applications towards a range of complementary methods of tsunami
warning.
Finally, the deep 2013 Okhotsk Sea earthquake has generated a
small but recordable tsunami, whose millimetric amplitude is correctly
predicted by a number of methodologies, suggesting that deep events may
contribute to tsunami hazard, if even larger-sized deep earthquakes could
occur.