storm’s circulation pattern, and the historical record
includes storms such as 1979 typhoon Tip in the western
Pacific which had a diameter of 2,200 km (Dunnavan
and Diercks, 1980). This is larger than the Gulf of
Mexico, the Caribbean, or the Mediterranean; it is difficult
to imagine sustaining a significantly larger storm. There
has been discussion that sea surface temperatures exceeding 36
C might generate ‘hypercanes’, storms much stronger than any observed historically (Parks Camp and
Montgomery, 2001). However, no one has yet suggested
that such conditions occurred within the past several
million years.
Tsunamis may be generated by several different natural
phenomena, including earthquakes, undersea landslides,
subaerial landslides that reach the sea, undersea volcanic
eruptions or flank collapses, and meteoroid impacts.
Although tsunami hazard analyses focus principally on
earthquake-induced tsunamis, the largest tsunami waves
are generated by other causes. For example, two of the
highest historical tsunami waves were generated by
a subaerial landslide (wave runup to 520 m elevation,
Lituya Bay, Alaska, 1958; see Mader and Gittings, 2002)
and a volcanic eruption ($36 m waves, Krakatau, Indonesia, 1883; see Simkin and Fiske, 1983).
Evaluating megablock occurrence may help quantify
hazards and occurrence rates for undersea landslides. This
is important because several investigations conclude that
giant-to-moderate (100–10 km
3
) submarine slope failures
may cause exceptionally large tsunamis (Ward and Day,
2001). Sometimes undersea landslides accompany or are
triggered by earthquakes and produce tsunamis much
larger and more devastating than those attributable to the
earthquake alone (e.g., the 18 November 1929 Grand
Banks earthquake; see Fine et al., 2005; and the 17 July
1998 New Guinea earthquake; see Heinrich et al., 2001;
Satake and Tanioka, 2003). In many coastal locations even
a cursory evaluation of offshore bathymetry demonstrates
that undersea landslides have occurred.
Dating the emplacement of megablocks
During the past 400 ky, sea level has only been at or above
its present only during interglacials. Interglacial
highstands typically occur at intervals of about 100,000
years and have durations of $10,000 years. The two most
recent highstands are the past 7,000 years and 120–
130 ka; thus these are the most plausible times of emplacement for large megablocks presently situated at elevations
near or above their sources. Generally the more recent
interval is likely because exposed coral limestone dissolves with rainfall, destroying megablocks over time.
During subsequent highstands, sea level returns to near
the level of paleomegablocks and they are destroyed by
bioerosion. And of course, any that were deposited during
lower sea levels are now underwater or had sea level pass
over them, thereby eroding them and then burying them
under new reef growth. When megablocks contain wellpreserved corals,
230
Th dating can determine when they
lived and thus constrain an earliest credible time for
emplacement.
Where megablocks occur as part of a well-preserved
stratigraphic sequence emplaced by a storm or tsunami,
this may contain mineral or organic material datable using
U-series or
14 C methods. Where megablocks contain identifiable coral species indicating that the blocks are
overturned, it may be possible to use
10
Be or
26
Al to estimate how long the upper surface has been exposed (e. g.,
see Nishizumi et al., 1993).
In some environments where erratic megablocks are situated on a raised coral-reef terrace, they are perched on
pedestals that have formed beneath the blocks as the adjacent terrace limestone is dissolved by rainfall solution.
Matsukura et al. (2007) describe examples of such
perched megablocks found on Kikaijima Island in the
Ryukyus; they estimate that the dissolution rate of the
unsheltered terrace surface is $205 mm/ky. As many pedestal heights are $1.0–1.5 m, they conclude that the
megablocks were deposited on a reef flat which emerged
and began experiencing dissolution about 6,000 years
ago. Although it might seem that measuring pedestal
heights would provide an accurate way of dating
megablock emplacement, at present this is a subject for
additional research as reported dissolution rates vary by
two orders of magnitude or more and depend on numerous
factors including amount of rainfall, whether or not the
block is exposed to both rainfall and sea spray, average
temperature, and porosity and composition of the terrace
surface.
How fast the current? How high the wave?
Wherever large erratic megablocks occur, they provoke
the question of how strong a current was required to displace them. If they are situated on land, how high was
the storm or tsunami wave that moved them?
A physics-based analysis provides some approximate
answers to these questions. In classical hydrodynamics,
an object of cross sectional area A in a fluid of density
r w streaming at velocity V experiences a drag force of
1
2 C d r w AV
2 , where C d is a dimensionless parameter known
as the drag coefficient which is $1/2 for objects such as
smooth spheres. To displace a megablock with dimension
h b and density r b , the drag force approximately equals its
buoyant weight; thus as its cross section is $h b
2 and its
volume is $h b
3 ,
1
2 C d r w h
2
b V
2
> gðr b À r w Þh
3
b
(1)
and
V >
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2g
C d
r b
r w
À 1
h b
s
(2)
If a wave impacts a shoreline, potential energy stored at
the wave crest may be converted to kinetic energy, producing a current as the wave breaks and flows onshore. Estimating the height h w of the wave necessary to displace
MEGABLOCKS
681
includes storms such as 1979 typhoon Tip in the western
Pacific which had a diameter of 2,200 km (Dunnavan
and Diercks, 1980). This is larger than the Gulf of
Mexico, the Caribbean, or the Mediterranean; it is difficult
to imagine sustaining a significantly larger storm. There
has been discussion that sea surface temperatures exceeding 36
C might generate ‘hypercanes’, storms much stronger than any observed historically (Parks Camp and
Montgomery, 2001). However, no one has yet suggested
that such conditions occurred within the past several
million years.
Tsunamis may be generated by several different natural
phenomena, including earthquakes, undersea landslides,
subaerial landslides that reach the sea, undersea volcanic
eruptions or flank collapses, and meteoroid impacts.
Although tsunami hazard analyses focus principally on
earthquake-induced tsunamis, the largest tsunami waves
are generated by other causes. For example, two of the
highest historical tsunami waves were generated by
a subaerial landslide (wave runup to 520 m elevation,
Lituya Bay, Alaska, 1958; see Mader and Gittings, 2002)
and a volcanic eruption ($36 m waves, Krakatau, Indonesia, 1883; see Simkin and Fiske, 1983).
Evaluating megablock occurrence may help quantify
hazards and occurrence rates for undersea landslides. This
is important because several investigations conclude that
giant-to-moderate (100–10 km
3
) submarine slope failures
may cause exceptionally large tsunamis (Ward and Day,
2001). Sometimes undersea landslides accompany or are
triggered by earthquakes and produce tsunamis much
larger and more devastating than those attributable to the
earthquake alone (e.g., the 18 November 1929 Grand
Banks earthquake; see Fine et al., 2005; and the 17 July
1998 New Guinea earthquake; see Heinrich et al., 2001;
Satake and Tanioka, 2003). In many coastal locations even
a cursory evaluation of offshore bathymetry demonstrates
that undersea landslides have occurred.
Dating the emplacement of megablocks
During the past 400 ky, sea level has only been at or above
its present only during interglacials. Interglacial
highstands typically occur at intervals of about 100,000
years and have durations of $10,000 years. The two most
recent highstands are the past 7,000 years and 120–
130 ka; thus these are the most plausible times of emplacement for large megablocks presently situated at elevations
near or above their sources. Generally the more recent
interval is likely because exposed coral limestone dissolves with rainfall, destroying megablocks over time.
During subsequent highstands, sea level returns to near
the level of paleomegablocks and they are destroyed by
bioerosion. And of course, any that were deposited during
lower sea levels are now underwater or had sea level pass
over them, thereby eroding them and then burying them
under new reef growth. When megablocks contain wellpreserved corals,
230
Th dating can determine when they
lived and thus constrain an earliest credible time for
emplacement.
Where megablocks occur as part of a well-preserved
stratigraphic sequence emplaced by a storm or tsunami,
this may contain mineral or organic material datable using
U-series or
14 C methods. Where megablocks contain identifiable coral species indicating that the blocks are
overturned, it may be possible to use
10
Be or
26
Al to estimate how long the upper surface has been exposed (e. g.,
see Nishizumi et al., 1993).
In some environments where erratic megablocks are situated on a raised coral-reef terrace, they are perched on
pedestals that have formed beneath the blocks as the adjacent terrace limestone is dissolved by rainfall solution.
Matsukura et al. (2007) describe examples of such
perched megablocks found on Kikaijima Island in the
Ryukyus; they estimate that the dissolution rate of the
unsheltered terrace surface is $205 mm/ky. As many pedestal heights are $1.0–1.5 m, they conclude that the
megablocks were deposited on a reef flat which emerged
and began experiencing dissolution about 6,000 years
ago. Although it might seem that measuring pedestal
heights would provide an accurate way of dating
megablock emplacement, at present this is a subject for
additional research as reported dissolution rates vary by
two orders of magnitude or more and depend on numerous
factors including amount of rainfall, whether or not the
block is exposed to both rainfall and sea spray, average
temperature, and porosity and composition of the terrace
surface.
How fast the current? How high the wave?
Wherever large erratic megablocks occur, they provoke
the question of how strong a current was required to displace them. If they are situated on land, how high was
the storm or tsunami wave that moved them?
A physics-based analysis provides some approximate
answers to these questions. In classical hydrodynamics,
an object of cross sectional area A in a fluid of density
r w streaming at velocity V experiences a drag force of
1
2 C d r w AV
2 , where C d is a dimensionless parameter known
as the drag coefficient which is $1/2 for objects such as
smooth spheres. To displace a megablock with dimension
h b and density r b , the drag force approximately equals its
buoyant weight; thus as its cross section is $h b
2 and its
volume is $h b
3 ,
1
2 C d r w h
2
b V
2
> gðr b À r w Þh
3
b
(1)
and
V >
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2g
C d
r b
r w
À 1
h b
s
(2)
If a wave impacts a shoreline, potential energy stored at
the wave crest may be converted to kinetic energy, producing a current as the wave breaks and flows onshore. Estimating the height h w of the wave necessary to displace
MEGABLOCKS
681