345 CHARACTERISTICS OF SEVERAL TSUNAMIGENIC

Transcription

345 CHARACTERISTICS OF SEVERAL TSUNAMIGENIC
CHARACTERISTICS OF SEVERAL TSUNAMIGENIC SUBMARINE
LANDSLIDES
H.J. LEE, R.E. KAYEN, J.V. GARDNER
U.S. Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025 USA
J. LOCAT
Laval University, Québec, Canada
In: J.Locat and J. Mienert,Eds., Submarine Mass Movements and their Consequences, Kluwer, 2003
Abstract
Submarine landslides are becoming recognized as a potential source of damaging local
tsunamis. However, there are presently few documented case studies of landslide events
that have caused historic tsunamis or likely caused prehistoric tsunamis. We present
three case studies of submarine landslide environments off the west coast of the United
States, including Alaska. Each environment has been imaged using multibeam
technology allowing excellent resolution of the morphology of the seafloor. Based on
this imagery and the historic record, we document the character of these environments
and the resulting tsunamis. In the case of one of the failures, we present a model of the
motion of the landslide and the size of tsunami that this motion would have produced.
Kewwords: submarine landslides, tsunamis, multibeam imagery, Alaska, Puget Sound,
Los Angeles
1. Introduction
Following the occurrence of the catastrophic 1998 Papua New Guinea tsunami, the
scientific community developed a renewed interest in submarine landslides as a
mechanism for generating large and potentially deadly tsunamis. Although the relative
contribution of earthquake and landslide sources is not clear (Geist, 2000), most
scientists agree that the very high (10- to 15-m) local runup of this tsunami is anomalous
for the magnitude 7.0 earthquake that preceded it. Accordingly, they suggest that it
must be explained by a hybrid landslide/earthquake, rather than simply the earthquake
acting alone (Tappin et al., 1999). Subsequent investigations have indeed located what
appear to be the remains of a giant landslide off the coast of Papua New Guinea, lending
credence to their argument (Tappin et al., 2001). This renewed interest in submarine
landslides has led to a continuing search for large submarine landslide features and
explanations for other anomalous tsunami characteristics. Note that landslide-generated
tsunamis are primarily destructive locally, in contrast with seismogenic tsunamis that
commonly cause damage at great distances. Also note that the height of a landslidegenerated tsunami increases with the thickness and volume of the landslide and
decreases with the water depth of the landslide head. The tsunami height also varies
with landslide velocity: landslide velocities near the tsunami velocity, which is related
to the water depth, produce the highest tsunamis (Ward, 2001).
345
346
Lee et al.
Another recent advancement has been the development and use of multibeam
echosounder systems (MBES) for seafloor mapping of bathymetry. As opposed to
earlier techniques that acquired bathymetry along tracklines, multibeam systems provide
complete bathymetric coverage over an area. These data can be visualized in a variety
of ways including shaded relief, producing imagery that resembles aerial photography.
When MBES systems are used to map seafloor landslide features, they allow us to
discern morphologic details that could not be resolved with earlier techniques.
Accordingly, we are acquiring a much improved appreciation of the characteristics of
submarine landslides, including those that cause tsunamis.
The present paper presents recently obtained MBES imagery of three US margin
submarine landslide environments: Resurrection Bay, Alaska, Commencement Bay,
Washington, and the Los Angeles margin, California. In the first two examples, deadly
tsunamis occurred, most likely as a result of seafloor slope failure. We present these
examples to illustrate the characteristics of failures known to have caused tsunamis.
The third example shows large, well defined landslide features near a major
metropolitan area. We consider whether such landslides might have been tsunamigenic
and whether a new landslide of similar size adjacent to this failure might be a hazard to
coastal development.
2. Methods
A general discussion of MBES technology is given by Hughes Clarke et al. (1996). In
principle, a multibeam echosounder differs from the traditional single-beam
echosounder in that the MBES records the return echoes with a contiguous series of
narrow receive apertures that are precisely aimed at angles that together typically cover
a swath of 60˚ to more than 150˚ of the seafloor. Each received signal is adjusted for
spherical spreading, attenuation in the water column, the ship’s roll, pitch, yaw, heave,
and the sound-speed structure of the water column. These systems are capable of
producing geodetic-quality bathymetry with spatial accuracies of better than 1 m and
vertical accuracies of 0.5% or better of the water depth. Multibeam data from
Resurrection Bay, Alaska and Commencement Bay, Washington were obtained using
high-resolution MBES systems aboard the National Oceanographic and Atmospheric
Administration (NOAA) ship Rainier in 2001. The area off Los Angeles is a composite
of MBES data collected by the US Geological Survey in 1997 and 1999 (Gardner and
Mayer, 1998; Gardner et al., 1999).
3. Landslide-induced tsunamis in Seward, Alaska, USA
The great Alaska earthquake of 1964 caused death and destruction at a number of
locations in south-central Alaska. The community of Seward, situated at the head of
Resurrection Bay along the east side of the Kenai Peninsula was particularly hard hit
(Lemke, 1967, Hampton et al., 1993). Much of the damage in Seward was clearly
related to coastal and seafloor slope failure and the generation of local tsunamis.
Between 35 and 45 seconds after the strong shaking began, an approximately 1-km-long
section of the waterfront, containing three large docks, many oil tanks, and a railroad
Characteristics of several tsunamigenic submarine landslides
347
yard, began sliding seaward as a result of large-scale offshore landsliding. Pipes from
the oil tanks ruptured, the tanks overturned, and the spilled oil erupted into flame. The
landslide displacements caused a “mound of water” to form in the bay (Lemke, 1967).
Waves radiated in all directions from the mound, having a height of about 10 m as they
struck a point south of Seward. As the wave hit Seward it brought back burning oil,
destroyed an 80-car freight train in the railway yard and caused 40 railway cars loaded
with oil to burst into flame in a chain reaction. These waves and damage that occurred
almost immediately after the earthquake were related to local coastal and marine failure.
About 30 minutes later, the seismogenic tsunami, roughly 10 m in height, arrived from
outside Resurrection Bay, struck Seward and caused significant additional damage. In
all, 13 persons died in Seward, most of them as a result of wave action.
Fig. 1 shows multibeam bathymetry presented in map and oblique views of the head of
Resurrection Bay. The imagery shows the front of the major fjord-head delta of the
Resurrection River extending completely across the northern end of the Bay. Within
the area mapped in Fig. 1a, the fronts of three smaller fan deltas produced by tributary
creeks ( Lowell Creek, Fourth of July Creek, and Spruce Creek) can also be identified.
These extend out into the bay, producing relatively flat areas of land and steep slopes
immediately offshore. The town of Seward was built on one of these fan-deltas (Lowell
Creek), and it was the face of this feature that failed catastrophically in 1964, removing
significant parts of the town’s infrastructure. The remains of this failure appear as
blocky debris extending offshore for about 500 m (Fig. 1). Some of the blocks are 10 to
15 m high. A profile across the subaerial part of the fan delta and extending offshore
(Fig. 2, location shown in Fig. 1a) shows that the delta is perched on the side of
Resurrection Bay. The area immediately offshore Seward has a slope of about 25°
decreasing to about 5° along the deeper parts of the profile. The other two small fan
deltas (Fourth of July Creek and Spruce Creek) do not show comparable blocky debris
although a deep trough south of the Fourth of July Creek fan-delta may be related to a
flow failure. The face of the major Resurrection River delta is covered with chutes and
gullies related to numerous and perhaps ongoing failure of the fjord-head-delta front.
The local tsunamis created in 1964 must have been related in part to the failure of the
Lowell Creek fan-delta and the production of the blocky debris. Other failures
throughout Resurrection Bay may have contributed to the tsunamis.
4. Landslide-induced tsunamis in Commencement Bay, Washington, USA
The Puyallup River flows into Commencement Bay near Tacoma, Washingon, carrying
much of the meltwater from the glaciers of Mount Rainier, including large quantities of
glacial silt. The rapidly prograding delta has advanced an average of 2 km/103 years
over the last 6000 years as indicated by dating of a mid-Holocene mudflow deposit
(Dragovich et al., 1994). Such rapid accumulation rates at the delta front lead to high
348
Lee et al.
Fig. 1. Shaded relief multibeam imagery of the head of Resurrection Bay, Alaska, showing map (a) and
oblique (b) views. The oblique view has a vertical exaggeration of 3; the base of the oblique image is 2500m
across.
Characteristics of several tsunamigenic submarine landslides
349
200
Seward Profile
Tacoma Profile
Palos Verdes Profile
Elevation (m)
0
-200
-400
-600
-800
-1000
0
1000
2000
3000
4000
5000
6000
Distance (m)
Figure 2. Profiles through landslide terrain in Resurrection Bay (Fig. 1a for location), Commencement Bay
(Fig. 3a for location) and Palos Verdes margin (Fig. 4a for location)
excess pore water pressures and increase the likelihood of failure. At least two failures
of the delta front occurred in historic time (Gardner et al., 2001) and included visible
displacements at the shoreline. Both failures were under aseismic conditions. The first
occurred in the southwestern margin of the Bay in 1894 (Tacoma Daily News, 1894)
and resulted in a tsunami that caused the loss of railway track, roadway, and several
buildings and led to two deaths. According to the Tacoma Daily News, "About two
acres of land fell into the water and caused a wave variously estimated from 10 to 20
feet (3 to 6 m) to ride down the Sound like a whirlwind." In a competing newspaper, a
policeman stated that a 10 to 15 foot (3 to 5 m) wave rushed through Old Town. The
second failure occurred in 1992 immediately off the mouth of the Puyallup River.
During the magnitude 6.8 Nisqually earthquake of 2001, there were no tsunamis
generated anywhere in Puget Sound.
A multibeam bathymetric survey of Commencement Bay (Fig. 3) was conducted in
2001 shortly after the Nisqually earthquake (Gardner et al., 2001). A number of failure
features were identified, some of which may have been related to the earthquake. In
addition, large sediment lobes were found off the mouth of the Puyallup River which
may be related to the 1992 failure. Most significantly, a 2-km-long channel was
mapped in the southwestern portion of the Bay that almost certainly corresponds to the
1894 failure. The head of the channel (identified as “1894 landslide” in Fig. 3a) is 450
m wide and shows little remaining failure material. The landslide debris must have
fully mobilized, flowed down the mapped channel, and dispersed along the floor of
350
Lee et al.
Puget Sound. A profile of the channel (Fig. 2) shows a steady gentle slope (2° to 5°)
throughout its length.
5. Potential landslide-induced tsunami near Palos Verdes, California, USA
The existence of submarine landslide terrain immediately south of the Palos Verdes
Peninsula has been known for 50 years (Emery and Terry, 1956), but the extent and
detailed morphology of the features was unknown until multibeam imagery became
available recently (Gardner and Mayer, 1998; Gardner et al., 1999, Bohannon and
Gardner, in press). This imagery (Fig. 4) shows a large variety of landslide deposits and
gullies extending along most of the steep (10° to 15°) basin slope that lies about 5 km
south of the Palos Verdes Peninsula. Two of the deposits appear particularly sharp,
fresh and ungullied and are identified as Palos Verdes Landslides 1 and 2 (Fig. 4).
Blocky landslide debris that extends to the south of the base of the slope for about 10
km shows that the landslides had considerable momentum when they reached the basin
floor and that they did not fullly mobilize into dispersed flows. Using radiocarbon dates
from piston-core samples obtained near the distal toe of the landslide deposits, Normark
et al. (in press) determined that the age of Palos Verdes Landslide 2 is 7,500 years. The
age of Palos Verdes Landslide 1 is unknown although the freshness of the feature, and
the lack of gullies in its headwall, indicates that it is also relatively recent. Given the
scale and apparent momentum of Palos Verdes Landslides 1 and 2, in comparison with
the scale of the known tsunamigenic landslide features described above for Resurrection
Bay and Commencement Bay, it is likely that both caused tsunamis.
Locat et al. (in press) developed a simplified model for the motion of Palos Verdes
Landside 1 (profile shown in Fig. 2), relying upon available geometric and physical
properties information. This model showed that the initial velocity of the landslide may
have exceeded 40 m/s and that most of the event was completed within 5 minutes.
Applying Murty’s (1979) model, Locat et al. (in press) estimated that the landslide
likely produced at least an 8-m high tsunami at its source, depending upon several
poorly constrained input parameters. The tsunami height decreases with distance from
the source area as a result of radial spreading and then increases as the tsunami shoals
toward shore.
Given the extent of landslide features on the Palos Verdes margin, a reoccurrence of
large-scale slope failure at some time in the future appears likely. There is a risk of
landslide-induced tsunamis, and these could impact both the nearby Ports of Los
Angeles and Long Beach and adjacent coastal communities.
Characteristics of several tsunamigenic submarine landslides
351
Fig.3. Shaded relief multibeam imagery of the head of Commencement Bay, Washington, showing map (a)
and oblique (b) views. The oblique view has a vertical exaggeration of 3; the base of the oblique image is
900m across.
352
Lee et al.
Figure 4. Shaded relief, multibeam imagery of the Palos Verdes margin, near Los Angeles, southern
California, showing map (a) and oblique (b) views. The oblique view had a vertical exaggeration of 3; the
base of the oblique image is 600-m across.
Characteristics of several tsunamigenic submarine landslides
353
6. Conclusions
Multibeam imagery of Resurrection Bay, Alaska, and Commencement Bay,
Washington, shows that tsunamigenic landslides can have varied morphologies. Such a
finding is reasonable given that the scale of landslide-induced tsunamis varies with
initial volume, thickness, water depth, and velocity, not necessarily the character of the
landslide in its depositional phase. In Resurrection Bay, the earthquake-induced
landslide that was at least partially responsible for devastating tsunamis has left a
deposit of blocky debris along the fan-delta slope immediately off the town of Seward.
These blocks extend up to 500 m offshore along a 5° to 25° slope. In Commencement
Bay, a large failure occurred under aseismic conditions and generated a tsunami that
damaged coastal structures and threatened life. This failure appeared to nearly
completely mobilize, removing most material from the source area and flowing down
channels. This failure occurred on a more gentle slope (2° to 5°) and likely resulted
from underconsolidated sediment deposited at the head of a rapidly prograding delta.
Along the Palos Verdes margin, large-scale failures of the basin slope produce partial
mobilization and the development of a blocky debris field that extends up to 10-km
beyond the base of slope. Such a morphology indicates high speed movement of the
failed blocks and the likelihood for tsunami generation.
7. Acknowledgements
We would like to acknowledge support from the USGS Coastal and Marine Geology
Program, the National Oceanic and Atmospheric Administration, the Orange County
Sanitation Districts and the Office of Naval Research. Eric Geist and Monty Hampton
contributed helpful reviews.
8. References
Bohannon, R.G. and Gardner, J.V., in press. Submarine landslides of San Pedro Sea Valley, southwest Los
Angeles Basin. Marine Geology.
Dragovich, J.D., Pringle, P.T., and Walsh, T.J., 1994. Extent and geometry of the mid-Holocene Osceola
mudflow in the Puget lowland—implications for Holocene sedimentation and paleogeography.
Washington Geology, 22: 3-26.
Emery, K.O. and Terry, R.D., 1956. A submarine slope off southern California. Journal of Geology, 64: 271280.
Gardner, J.V., and Mayer, L.A., 1998. Cruise Report RV Ocean Alert C5uise4 A2-98-SC, Mapping the
southern California continental margin. US Geological Survey Open-File Report 98-475.
Gardner, J.V., Hughes Clarke, J.E., and Mayer, L.A., 1999. Multibeam mapping of the Long Beach,
California, continental shelf; Cruise report: RV Coastal Surveyor cruise C-1-99-SC: US Geological
Survey Open-File Report 98-360.
Gardner, J.V., Van den Ameele, E. J., Gelfenbaum, G., Barnhardt, W., Lee, H.J., and Palmer, S., 2001.
Mapping southern Puget Sound delta fronts after 2001 earthquake. Eos, Transactions, American
Geophysical Union, 82: 485, 488-489.
Geist, E.L., 2000. Origin of the 17 July 1998 Papua New Guinea tsunami: earthquake or landslide?
Seismological Research Letters, 71:344-351.
Hampton, M.A., Lemke, R.W., and Coulter, 1993. Submarine landslides that had a significant impact on man
and his activities: Seward and Valdez, Alaska. in Schwab. W.C., Lee, H.J., and Twichell, D.C.(Eds.),
Submarine Landslides: Selected Studies in the U.S. EEZ: USGS Bulletin 2002, p. 123-142.
354
Lee et al.
Hughes Clarke, J.E., Mayer, L.A., and Wells, D.E., 1996. Shallow-water imaging multibeam sonars: a new
tool for investigation seafloor processes in the coastal zone and on the continental shelf. Marine
Geophysical Research, 18: 607-629.
Lemke, R.W., 1967. Effects of the earthquake of March 27, 1964, at Seward, Alaska. US Geological Survey
Professional Paper 542-E, 43 pp.
Locat, J., Locat, P. Lee, H.J., and Imran, J., in press. Nunerical modeling of the mobility of the Palos Verdes
debris avalanche, California, and its implication for the generation of tsunamis. Marine Geology.
Murty, T.S., 1979. Submarine slide-generated water waves in Kitimat Inlet, British Columbia. Journal of
Geophysical Rsearch, 84: 7777-7779.
Normark, W.R., McGann, M., and Sliter, R., in press. Age of Palos Verdes submarine debris avalanche,
southern California. Marine Geology.
Tacoma Daily News, 1894. The big cave-in. Thurs., Nov. 29, 1894, 1.
Tappin, D.R. and others, 1999. Sediment slump likely caused 1998 Papua New Guinea tsunami. Eos,
Transactions, American Geophysical Union, 80: 329. 334, 340.
Tappin, D.R., Watts, P., McMurtry, G.M., Lafoy, Y., and Matsumota, 2001. The Sissano, Papua New Guinea
tsunami of July 1998 — offshore evidence on the source mechanism. Marine Geology, 175: 1-23.
Ward, S.N., 2001. Landslide tsunami. Journal of Geophysical Research. 106: 11201-11215.