Tide gauge measurements and analysis of the Indian Ocean tsunami on the Pacific coast of South Ameri

1
Tide gauge measurements and analysis of the
Indian Ocean tsunami on the Pacific coast of
South America

• A.B. Rabinovich1,2 and R.E. Thomson1

1 Institute of Ocean Sciences, Sidney, B.C.
Canada 2 P.P. Shirshov Institute of Oceanology,
Moscow, Russia
2
Tsunami of the December 26, 2004 recorded in the
World Ocean
Maximum tsunami amplitudes computed by V.V. Titov
(PMEL/NOAA)
Titov et al., Science, 2005
3
MOST model of the December 26, 2005 tsunami
Numerical model by V.V. Titov (PMEL/NOAA)
4
Location of tide gauge stations on the coast of
South America
5
2004 Sumatra tsunami Observed wave parameters
6
The 2004 Indian Ocean tsunami records on the
Pacific coast of South America
(4) Train structure

• Long ringing (gt2 days)

(5) Maximum waves in 8-30 hrs after the
first arrival (in the second or third
train)
(2) Slow energy decay
(3) Unclear first arrival
7
The 2004 tsunami as recorded in the Indian Ocean

• Relatively short ringing
• (1-1.5 days)

(3) Maximum waves are in the very
beginning
(4) Fast amplitude decay
(2) First arrival is abrupt and clear
8
North Pacific Ocean
Aleutian and Kuril Islands
Pacific coast of Canada
9
Indian Ocean (Sumatra) tsunami of December 26,
2004
Recorded tsunami wave heights along the coast of
South America
10
Tsunami and background spectra
The spectra of tsunamis from different
earthquakes are similar at the same location but
are quite different for the same event for nearby
locations Omori, 1902, and many others
11
Reconstructed source functions
Northern group of stations
12
Reconstructed source functions
Southern group of stations
13
Averaged source function
While individual tsunami spectra are
significantly different for different stations,
the reconstructed source function is supposed
to be independent of local topographic effects
and be related to the actual spectral
characteristics of the source. Good agreement of
this function calculated for various sites and
different regions supports this assumption.
14
Reconstructed source functions based on Canadian
records (BC coast)
15
Wavelet plots of the 2004 Indian Ocean tsunami
(1)
Time 25/12 18h 28/12 18h
Periods 2.5h 0.20h
16
Wavelet plots of the 2004 Indian Ocean tsunami
(2)
Time 25/12 18h 28/12 18h
Periods 2.5h 0.20h
17
Wavelet analysis of the Indian Ocean records
Time 25/12 12h 28/12 12h Periods
2.5h 0.20h
(Cocos1 1.67h 0.13h)
18
Conclusions

• The 2004 Sumatra tsunami was the first
  global-scale tsunami in the instrumental era
  and it was observed throughout the World Ocean,
  including the North Pacific, North Atlantic and
  Antarctic Oceans.

2. This tsunami was clearly measured by tide
gauges along the entire Pacific coast of South
America with wave heights ranging from a few cm
(Punta Corona) to 72 cm (Arica).
19
3. The resonant characteristics of the shelf and
coastline strongly determine the observed
features of tsunami waves, in particular wave
heights and periods maximum waves were observed
at sites having eigen (natural) periods from 30
to 60 minutes.
4. The recorded tsunami waves are characterized
by long ringing (gt2days), slow energy decay, and
clear train wave structure with maximum waves
related mainly to the second or third train (8-30
hours after the first arrival) from this point
of view these records are quite different from
those observed in the Indian Ocean but similar to
those observed in the North Pacific and North
Atlantic.
20
5. The observed spectra of tsunami waves had
peaks significantly different for different
stations in contrast, these peaks were in good
agreement with background peaks at the same
sites. Significant tsunami oscillations were
observed at stations with local resonant periods
of 30-50 min, apparently corresponding to the
periods of arriving tsunami waves.
6. In contrast to individual tsunami spectra, the
reconstructed source function is found to be
very consistent and almost independent of local
topographic effects and is apparently related to
the spectral characteristics of the source.