TECTONICS, VOL. 24, TC6005, doi:10.1029/2004TC001650, 2005
Rapid forearc uplift and subsidence caused
by impinging bathymetric features:
Examples from the New Hebrides and Solomon arcs
F. W. Taylor,1 Paul Mann,1 M. G. Bevis,2 R. L. Edwards,3 Hai Cheng,3
Kirsten B. Cutler,3,4 S. C. Gray,5 G. S. Burr,6 J. W. Beck,6
David A. Phillips,7 G. Cabioch,8 and J. Recy8
Received 15 March 2004; revised 14 October 2004; accepted 22 June 2005; published 15 November 2005.
[1] Isotopically dated corals from the central New
Hebrides and New Georgia Island Group, Solomon
Islands, indicate that both forearcs underwent rapid
late Quaternary subsidence that was abruptly
replaced by hundreds of meters of uplift at rates
up to 8 mm/yr, while total plate convergence was
only a few kilometers. Two mechanisms that might
account for these rapid reversals in vertical motion
include (1) a ‘‘displacement’’ mechanism in which
the forearc is displaced upward by the volume of an
object passing beneath on the subducting plate (as
the object moves deeper and vacates the base of the
forearc, the forearc subsides to near its original
position) and (2) a ‘‘crustal shortening’’ mechanism
in which the forearc thickens and uplifts because of
horizontal shortening when a large object impinges
on the forearc and abruptly increases interplate
coupling on the shallow end of the main thrust
zone. Rapid subsidence follows when the impinging
object is broken or otherwise decoupled, shallow
interplate coupling becomes weak, and the uplifted
forearc extends and subsides. The displacement
mechanism surely plays a role on timescales over
which plates converge tens of kilometers, but it fails to
explain the geographic pattern, short time frame, and
abruptness of the change from subsidence to uplift that
we observe. The crustal shortening mechanism is
preferred because it allows the observed abrupt uplift
1
Institute for Geophysics, Jackson School of Geosciences, University of
Texas at Austin, Austin, Texas, USA.
2
Department of Civil and Environmental Engineering and Geodetic
Science, Ohio State University, Columbus, Ohio, USA.
3
Minnesota Isotope Laboratory, Department of Geology and Geophysics,
University of Minnesota, Minneapolis, Minnesota, USA.
4
Now at U.S. Department of State, Office of Senior Coordinator for
Nuclear Safety, Washington, D. C.
5
Department of Marine and Environmental Studies, University of
San Diego, San Diego, California, USA.
6
NSF Accelerator Facility, Physics Department, University of Arizona,
Tucson, Arizona, USA.
7
UNAVCO, Boulder, Colorado, USA.
8
Institut de Recherche pour le Développement, Nouméa, Vanuatu.
Copyright 2005 by the American Geophysical Union.
0278-7407/05/2004TC001650$12.00
when an object impinges on a forearc and causes
locking of a shallow segment of the interplate thrust
zone. Citation: Taylor, F. W., et al. (2005), Rapid forearc uplift
and subsidence caused by impinging bathymetric features:
Examples from the New Hebrides and Solomon arcs,
Tectonics, 24, TC6005, doi:10.1029/2004TC001650.
1. Introduction
[2] When prominent bathymetric features subduct at
convergent plate margins they influence interplate coupling
and forearc tectonic deformation roughly in proportion to
their size although other factors such as sediment thickness
may affect the process [Cloos, 1992; Geist et al., 1993;
Scholz and Small, 1997; Dominguez et al., 1998; Mann et
al., 1998; Fisher et al., 1998; Gardner et al., 1992; Gardner
et al., 2001; Meffre and Crawford, 2001; Bilek et al., 2003].
Subduction of large buoyant ridges or plateaus may produce
several km of surface uplift due to isostatic adjustments
(Figure 1). Resistance to subduction by such large bodies
may lead to arc polarity reversal as inferred for Ontong-Java
Plateau collision with the Solomons arc [e.g., Kroenke,
1984; Phinney, 1997]. Smaller subducting features must
cause proportionately less forearc isostatic displacement
depending on their volume [e.g., Meffre and Crawford,
2001]. Other processes such as underplating of the forearc
by sediments or pieces of oceanic crust and tectonic erosion
also result in vertical displacement and tectonic evolution of
forearcs on the million-year timescale. However, our results
from the central New Hebrides (CNH) and New Georgia
Group (NGG), Solomon Islands, forearcs indicate that other
mechanisms may be responsible for very abrupt, rapid uplift
when seamounts and ridges on a converging plate impinge
on forearcs and cause very strong coupling across the
shallow end of the interplate thrust zone where interplate
slip normally occurs with little resistance [e.g., Pacheco et
al., 1993] (Figure 1).
[3] The d’Entrecasteaux Ridge is a bathymetrically
prominent extinct Oligocene island arc [Weissel et al.,
1982; Maillet et al., 1983; Baker et al., 1994] that is
crossing the trench outer rise and subducting beneath the
CNH arc (Figures 1, 2, and 3). A reef grew on one of the
arc volcanoes after it crossed the trench outer rise and
began to resubmerge to form Bougainville Guyot on the
southern part of the d’Entrecasteaux Ridge. Accelerating
subsidence drowned the reef by about 300 ka and carried
TC6005
1 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
it to 1000 m below sea level (BSL) where it now indents
the forearc wedge (Figure 4). At the NGG, western Solomons arc, the young Woodlark Basin has highstanding
seamounts and ridges just offshore of the San Cristobal
Trench. For example, the distal slope of Coleman seamount has begun to impinge on the forearc [Taylor and
Exon, 1987; Mann et al., 1998]. Large areas, both across
and along the CNH and NGG forearcs underwent
hundreds of meters of rapid late Quaternary subsidence
followed by an abrupt change to rapid uplift. At the CNH
and NGG, uplift is centered on forearc islands that begin
only 20 to 30 km from the trench. Tectonic deformation at
the NGG and CNH is very similar in terms of geographic
distribution of uplift and the 100 – 200 ka cycles of
subsidence and uplift [Taylor, 1992; Taylor et al., 1987;
F. W. Taylor et al., 1995; Mann et al., 1998] (Figure 1).
This implies similar processes despite differences between
many aspects of the two settings including age and buoyancy of the downgoing lithosphere.
[4] There is considerable scientific interest in processes
affecting the shallow end of the interplate thrust zone, but
land rarely exists over this part of the arc so that direct
measurements can be made. Our observations reveal a
surprisingly abrupt change from rapid subsidence to rapid
uplift in the extreme trenchward part of both forearcs
TC6005
during a brief interval of time marked by only a few
kilometers of plate convergence. The limited amount of
subduction and abrupt transition between rapid forearc
subsidence and uplift cannot be explained by previously
proposed tectonic models of simple volumetric displacement of a subducted bathymetric high, forearc underplating, or tectonic erosion at the base of the forearc. Instead,
we propose that the rapid subsidence and uplift are related
to impingement of bathymetric features that cause abrupt,
strong interplate coupling at the shallow end of the interplate thrust zone that persists for thousands of years while
at greater depths the fault continues to slip. Uplift results
from crustal shortening and thickening of the upper plate
above the locked zone to accommodate plate convergence
(Figure 1) whether interplate slip at greater depths proceeds
by coseismic stick slip or creep. Eventually, an impinging
feature is either subducted or dismembered so that strong,
shallow coupling with the forearc abruptly diminishes or
terminates in rapid forearc subsidence because strong force
ceases to be applied across the interplate boundary. Such a
mechanism is important as a cause for deformation at
convergent plate boundaries. Many seamounts exist and
even when they subduct at island arc systems where
interplate coupling is weak, locally strong interplate cou-
Figure 1. (a) Ridges, seamounts, and plateaus which often
approach and impinge on forearcs with significant tectonic
consequences. At the CNH and New Georgia Islands, two
quite different mechanisms may explain how impinging
bathymetric features cause the tectonic deformation that we
have observed: (b) Displacement model. Upon impinging
on the forearc, a seamount slips beneath the upper plate
displaces the forearc upward in proportion to its volume. A
forearc bulge may propagate through the arc followed by
subsidence after the downgoing object has passed. The
cycle of uplift and subsidence may require hundreds of
thousands of years due to the convergence rate and
dependence on displacement of the forearc by the presence
of the downgoing object. (c) Compression model. If the
forearc is relatively strong, then significant compressive
force may be applied against the forearc. This may cause
shortening of the thinner trenchward wedge of the forearc
resulting in rapid uplift and even backthrusting of a
segment of the arc [e.g., F. W. Taylor et al., 1995].
Because this mechanism depends on force transmitted
through the arc, deformation of the entire forearc can
occur during an interval in which total plate convergence
is only a few kilometers or less. (d) Vertical tectonic
effects of the compressional model are quite different than
those of the simple displacement model. Curve I shows
that the onset of uplift caused by sudden strong coupling
at the upper part of the interplate thrust zone is quite
abrupt. Cessation of uplift and subsidence may also be
abrupt because the impinging object does not have to be
physically removed from the interplate zone for interplate
coupling to cease. Curve II shows that in contrast, both
uplift and subsidence caused by upward displacement by
the volume of a downgoing object are prolonged.
2 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 2. Regional setting of (box A) the New Georgia Group of the western Solomons arc and (box b)
the Central New Hebrides arc. These arc systems lie between the Australian and Pacific plates (AUS/
PAC) which are converging ENE-WSW at 75– 100 mm/yr in this region [DeMets et al., 1990, 1994;
Beavan et al., 2002]. At both arc systems the Australian plate surface has rugged bathymetric features
that locally resist subduction. The Woodlark Basin has rough topography on its margins and central area,
and the part subducting at the New Georgia Group is only 1 –3 Ma [Crook and Taylor, 1994]. The
d’Entrecasteaux Ridge System underthrusts the central New Hebrides arc. There is evidence of
significant back-arc compression at both arcs that is believed to be related to compressive force
transmitted across the arcs from the rugged bathymetric features on the Australian plate impinging on the
forearcs [F. W. Taylor et al., 1995; Phinney et al., 1999].
pling may be induced [e.g., Cloos, 1992; Scholz and Small,
1997].
[5] Uplifted coral reefs provide precise rates of vertical
deformation for the NGG and CNH islands via their
relationship with sea level [e.g., Chappell, 1974; Bloom et
al., 1974; Edwards et al., 1988; Taylor, 1992]. Our previous
work in the CNH used vertical displacements measured
using living and ancient coral reefs and GPS measurements
to establish the relationships between Australian plate
convergence with the CNH, seismicity, and vertical and
horizontal deformation for latest Quaternary time. Our work
in the NGG was less extensive than for the CNH, but
sufficient to recognize a tectonic process very similar to that
of the CNH is occurring as detailed by Mann et al. [1998].
Here we summarize NGG neotectonics based on the work
by Mann et al. [1998] and add new data that more precisely
constrain the NGG tectonic history. New 14C dates for coral
samples drilled at depths down to 55 m below sea level
(BSL) and new dates on surface samples from both the
CNH and NGG allow us to much more precisely constrain
3 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 3. Tectonic features and GPS vectors summarized for the New Hebrides arc [after F. W. Taylor et
al., 1995; Pelletier et al., 1998]. GPS convergence rates between the New Hebrides arc and Australian
plate are 100 mm/yr for the southern part of the New Hebrides [F. W. Taylor et al., 1995; Calmant et al.,
1995, 2003] and somewhat exceed the NUVEL-1A Australia-Pacific rates [DeMets et al., 1994].
However, central New Hebrides convergence rates are only about half the rates expected from
NUVEL-1A. The d’Entrecasteaux Ridge and Bougainville Guyot resist subduction so strongly that a
250-km-long central New Hebrides segment is being shoved eastward onto the North Fiji Basin [F. W.
Taylor et al., 1995; Pelletier et al., 1998; Calmant et al., 2003; Lagabrielle et al., 2003]. This crustal
shortening is accommodated by cross-arc strike-slip faults and by back-arc reverse faults that emerge in
the North Fiji Basin east of Maewo and Pentecost islands [Collot et al., 1985; Greene et al., 1988].
Both the strike-slip and back-arc reverse faults have ruptured with Ms > 7.5 earthquakes in 1994 and
1999 [Calmant et al., 1997, 2003; Pelletier et al., 2000; Regnier et al., 2003].
the timing, geography, and rates of vertical tectonism over
the past few hundred thousand years (Tables 1 and 2). We
discuss the CNH and NGG together in one paper because
both tectonic histories represent a similar process and one
that may be more common than previously suspected.
2. Plate Convergence at the Central New
Hebrides and Subduction of the
d’Entrecasteaux Ridge System
2.1. Relative Plate Motions at the
Central New Hebrides Arc
[6] The New Hebrides arc is part of a microplate separated from the Pacific plate by seafloor spreading centers in
the North Fiji Basin (Figures 2 and 3). The northern
boundary of the New Hebrides microplate is a fault zone
linking North Fiji Basin spreading centers to the New
Hebrides arc north of the CNH [Pelletier et al., 1998]. Global
plate motion models and regional GPS analyses indicate that
net Australian-Pacific plate motion at the CNH is 84–
87 mm/yr in an ENE direction [DeMets et al., 1990, 1994;
Beavan et al., 2002]. However, convergence rates of 110 to
130 mm/yr were estimated based on the motion of New
Hebrides outer rise islands [Dubois et al., 1977; F. W. Taylor
et al., 1994]. These paleoconvergence rates compare well
with rates of 100 to 120 mm/yr measured by GPS at several
southern New Hebrides islands [F. W. Taylor et al., 1995;
Calmant et al., 1995, 2003] (Figure 3). However, we do
not know precisely why convergence rates exceed the
predicted Australian-Pacific relative velocity of 79 –
82 mm/yr for those islands (Figure 3).
[7] In contrast to GPS rates in the southern New Hebrides
and regional plate motion rates, GPS measurements indicate
convergence rates of only 30– 50 mm/yr at numerous
sites on the CNH. On the basis of the possibility that part
of the convergence between the CNH and Australian plate
4 of 23
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
TC6005
Figure 4. Cross section of the CNH through the Bougainville Guyot. Bougainville Guyot has
penetrated the edge of the CNH forearc and displaced the plate boundary about 10 km east of its
normal position. Focal mechanisms on the interplate thrust zone are for one major foreshock and the
main shock of the 11 August 1965 earthquake sequence. The focal mechanisms beneath Bougainville
Guyot are for the largest aftershocks on 13 August 1965. The foreshock and other clear thrust-type
events define the interplate seismogenic zone that ruptured during larger historical interplate events.
Following the 1965 interplate thrust earthquake, the main aftershock on 13 August was not a thrust and
occurred near the Bougainville Guyot. This aftershock consisted of two successive subevents that
occurred at depths of about 10 and 30 km within the downgoing plate as shown beneath Bougainville
Guyot [Ebel, 1980]. These focal mechanisms suggest a very steeply dipping fault plane that could be
related to bending of the downgoing plate or to stresses imposed by extensional force related to
impingement of the Bougainville Guyot resisting slab-pull forces. The alternative interpretation would
be two nearly parallel thrust events.
is concealed as elastic strain within the upper plate, we
infer that on the millennial timescale, the mean convergence rate could be greater than 40 mm/yr, but probably
does not exceed 70 mm/yr. The anomalously low
Australian-CNH convergence rates indicate that the whole
CNH arc segment is moving eastward at a rate of 10–
60 mm/yr relative to the rest of the arc and that this motion is
accommodated in the back arc by thrust faulting [Collot et
al., 1985; F. W. Taylor et al., 1995; Pelletier et al., 1998].
Strike-slip earthquake focal mechanisms occur along fault
zones separating the CNH from the rest of the arc and
thrusting focal mechanisms occur at the boundary between
the CNH and North Fiji Basin [e.g., F. W. Taylor et al.,
Table 1. New
230
1995]. A recent example of this eastward motion of the CNH
is a Ms = 7.5 event just east of Ambrym Island in 1999 that
has a thrust focal mechanism dipping 30 to the west
[Regnier et al., 2003; Lagabrielle et al., 2003] consistent
with eastward motion of the CNH. Therefore the CNH is
a small microplate separated from the rest of the New
Hebrides by faults [Pelletier et al., 1998] (Figure 3).
2.2. Impingement and Subduction of the
d’Entrecasteaux Ridge and Bougainville Guyot
[8] The d’Entrecasteaux Ridge intersects the CNH arc
between southern Malakula Island and northern Espiritu
Santo Island and is a major recognized factor in the
Th/234U Dates From Southern Espiritu Santo Island, New Hebrides Arca
Sample
Coral Species
Elevation, m
10-254
10-290
11-145
TM-GA
TM-GB
unknown
Faviidae sp.
Acroporid sp.
Lobophyllia sp.
Lobophyllia sp.
58
63.7
75.3
+36
+35
238
U, ppb
3561
2957
3346
2369
2189
±
±
±
±
±
5
4
5
3
5
232
Th, ppt
792 ± 18
2299 ± 25
1906 ± 30
1100 ± 3
35 ± 1
234
Ub (Measured)
61.1
59.9
58.4
77.3
73.6
a
±
±
±
±
±
1.1
1.1
1.2
1.1
2.3
230
Th/238U (Activity)
1.0028
1.0286
1.0554
0.9375
0.9495
Errors are 2s.
Here d234U = ([234U/238U]activity 1) 1000.
c
Here l230 = 9.1577 10 6 yr 1, l234 = 2.8263 10 6 yr 1, l238 = 1.55125 10 10 yr 1.
d
Here d234Uinitial was calculated based on 230Th age (T), i.e., d234Uinitial = d234Umeasured el234xT.
b
5 of 23
±
±
±
±
±
0.0059
0.0049
0.0084
0.0022
0.0030
230
Th Age,c ka
288.4 ± 8.1
333 ± 10
412 ± 35
210.9 ± 1.7
222.1 ± 2.9
234
Ud (Initial)
137.9 ± 4.0
153.4 ± 5.1
188 ± 20
140.4 ± 2.1
137.9 ± 4.5
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
TC6005
Table 2. New Radiocarbon Dates for Coral Samples From Borehole 3, Central Rendova Island, New Georgia Group, Solomon Islands
Samplea
RDV-3-0 m
RDV-3-2 m
RDV-3-5 m
RDV-3-6 m
RDV-3 12 m
RDV-3- 15 m
RDV-3-30 m
RDV-3 -33 m
RDV-3-54.0 m
RDV-3 -56.4 m
RDV-3-56.9 m
RDV-3-58 m
Elevation,b
m
+1.75
0.25
3.25
4.25
10.25
13.25
28.25
31.25
52.25
54.65
55.15
56.25
d13C
0.0
0.9
0.7
0.0
3.1
1.6
0.0
2.0
3.3
3.3
1.6
0.8
Fraction Modern
Carbon (F)
0.8938
0.7088
0.6652
0.5959
0.4880
0.4265
0.3056
0.3063
0.1191
0.1455
0.0984
0.1069
±
±
±
±
±
±
±
±
±
±
±
±
0.0057
0.0053
0.0037
0.0036
0.0037
0.0028
0.0032
0.0061
0.0014
0.0021
0.0023
0.0016
14
C Agec (Raw),
years B.P.
Error
1s years
902
2764
3275
4158
5763
6845
9523
9500
17,092
15,480
18,620
17,960
50
60
44
48
59
51
83
160
94
110
180
120
Reservoir Corrected Error 1s, Calibrated 14C Age Range,e
C Age,d years B.P.
years
cal years B.P. 2s
14
437
2299
2810
3693
5298
6380
9058
9035
16,627
15,015
18,155
17,495
51
61
45
49
60
52
84
160
95
110
180
120
542 – 325
2431 – 2151
3059 – 2783
4217 – 3875
6274 – 6237
7425 – 7212
10,465 – 9920
10,577 – 9633
20,486 – 19,177
18,577 – 17,402
22,381 – 20,808
21,528 – 20,125
a
Sample number includes depth below surface in borehole.
Meters are above (positive) or below (negative) present sea level.
c
Uncorrected ages are reported using the conventional 5568 year half-life and corrected with measured d13C values.
d
Reservoir correction is 465 ± 10 (1s) based on analyses of live coral specimens from Solomon Islands [Schmidt et al., 2004].
e
The final ages are calibrated using the INTCAL program (4.01) [Stuiver et al., 1998].
b
unusual tectonic evolution, structure, and morphology of
the CNH including backthrusting of the 250-km-long
CNH arc segment [Taylor et al., 1980; F. W. Taylor et
al., 1995; Collot et al., 1985; Greene and Collot, 1994;
Fisher, 1986; Fisher et al., 1986, 1991a, 1991b; Calmant
et al., 1995, 2003; Pelletier et al., 1998; Meffre and
Crawford, 2001]. The d’Entrecasteaux Ridge trends nearly
east-west, is about 100 km wide, and consists of two
parallel linear features: a more-or-less continuous northern
d’Entrecasteaux Ridge and a southern d’Entrecasteaux
seamount chain. The northern ridge is about 40 km wide
and rises 3 km higher than the adjacent seafloor immediately to the north [Monzier et al., 1984]. The seamount
chain includes the living reef-capped Sabine Bank about
50 km west of the plate boundary which has just begun to
submerge and accumulate a reef cap as it follows the
trajectory of the Australian plate across the New Hebrides
trench outer rise. The Bougainville Guyot is farther along
the trajectory with its cap of reefs that drowned as it
descended from sea level to 1000 m BSL. Bougainville
Guyot now indents the edge of the arc by about 10 km
[Collot et al., 1985; Fisher et al., 1991a; Greene and
Collot, 1994; Taylor et al., 1994]. The upper part of the
southeast flank of the Bougainville Guyot slopes on the
order of 25, typical of an atoll forereef slope. At its base,
the Bougainville Guyot is about 50 km in diameter and its
lower flanks slope 5. Including its lower slopes, the
guyot has an area of 2500 km2 and its summit plateau
has an area of 400 km2. The northern part of the CNH has
begun to be underthrust by the distal flank of the Torres
Plateau (Figure 3). The Torres Plateau appears to have
impinged recently at the CNH arc, but may contribute to
interplate coupling along the northern part of the CNH.
[9] We assume that the d’Entrecasteaux Ridge extends
with the Australian plate beneath the CNH and that the
subducted ridge had bathymetric relief similar to the part that
we can see in bathymetry. A d’Entrecasteaux Ridge seamount probably has arrived at the CNH every few hundred
thousand years, given the 50 km spacing between Bougainville Guyot and Sabine Bank and the convergence rate [e.g.,
Collot and Fisher, 1989]. The tectonic effects of impinging
seamounts should occur as distinct events in contrast to
ridges which continuously interact with the forearc.
3. Coral Reefs as Recorders of Vertical
Tectonic History
[10] Given the age and present altitude of a coral and
the paleosea level relative to present sea level at which it
lived, we can infer its vertical displacement and vertical
velocity [Taylor et al., 1980, 1985, 1987, 1990; Taylor,
1992; Mann et al., 1998]. Fossil corals can be precisely
dated by thermal ionization mass spectrometry (TIMS)
230
Th/234U [Edwards et al., 1988] and by 14C accelerator
mass spectrometry (AMS) dating. At the CNH we also
refer to older, less precise alpha spectrometry (AS)
230
Th/234U ages from earlier papers. Errors are two sigmas
for all 230Th/234U ages and for 14C ages calibrated to
calendar years by INTCAL 98 [Stuiver et al., 1998]
(Tables 1 and 2).
[11] Paleosea level history enables more precise estimates
of vertical displacement from fossil reefs because it constrains the original heights of reefs before vertical tectonic
displacement. We use a late Quaternary paleosea level
history based on the work by Chappell et al. [1996] and
Cutler et al. [2003] back to about 130 ka and extend the
history to 400 ka using the Specmap oxygen isotope curve
that approximates sea level history [Imbrie et al., 1984].
For the last interglacial marine isotope stage (MIS) 5e
(130 ka), we assume that sea level was 6 m above
present sea level (ASL) [Veeh, 1966; Bloom et al., 1974]
and that MIS 5c (105 ka) and 5a (80 ka) paleosea
levels for this region reached a maximum of 15 m BSL
[e.g., Chappell and Shackleton, 1986; Chappell et al.,
1996; Fleming et al., 1998; Peltier, 1998]. Fortunately,
vertical motions in the CNH and NGG are much larger
than uncertainties in sea level history. Refer to the explanation for site 1 on Espiritu Santo Island in Section 5.1
where we explain in detail the logic and calculation of
6 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 5. Map of total Holocene coastal emergence (since 6 ka) [Taylor et al., 1987]. Cross-arc faults
and the back-arc reverse fault that allow the CNH to shift eastward are depicted [Taylor et al., 1980,
1987]. Emerged reefs show that the entire western edge of the CNH has uplifted, with the axis of
maximum uplift running NNW-SSE through Espiritu Santo and western Malakula islands about 30–
50 km from the convergent plate boundary. Also depicted by bold black lines are the boundaries
between four arc segments that have behaved semi-independently with respect to rupture zones and
tilt directions [Taylor et al., 1980, 1987]. However, the amount of differential motion is no more than
tens of meters. The cross-arc faults at the northern and southern limits of the central New Hebrides
that allow the entire segment to move eastward are far more significant. The eastward motion is
accommodated by reverse faulting on a back-arc fault just to the east of and parallel to Maewo and
Pentecost islands. The inset shows sites 1 –4 on southern Espiritu Santo Island where we obtained
coral reef evidence of the subsidence that preceded the ongoing rapid uplift (Figure 6).
subsidence and uplift history based on reef altitudes and
ages.
4. Vertical Tectonic Motions of the Central
New Hebrides Island Arc Based on Previous
Work
[12] Several cycles of subsidence and uplift since 1 – 2 Ma
were established previously for the CNH forearc and backarc islands [e.g., Mitchell, 1966; Mallick and Neef, 1974;
Mallick and Greenbaum, 1977; Carney and Macfarlane,
1982; Carney, 1986; Macfarlane et al., 1988; Greene et
al., 1988; Katz, 1988]. In particular, Macfarlane et al.
[1988] recognized a ‘‘brief marine incursion’’ of at least
several hundred meters ASL at a biostratigraphic age of
300 – 400 ka on southern Espiritu Santo Island that
matches the subsidence of several hundred meters that
we describe in the vertical tectonic history from coral
reefs.
[13] Reef morphostratigraphy and numerous isotopic
coral ages across southern and all of eastern Espiritu Santo
Island and northern Malakula Island consistently date an
uplifted 0 – 130 ka late Quaternary reef sequence [Neef and
Veeh, 1977; Jouannic et al., 1980, 1982; Taylor et al., 1980,
1987, 1990; Gilpin, 1982; Urmos, 1985]. Maximum Holocene forearc uplift with rates 5 mm/yr is centered unusually close to the trench on the topographic axes of western
7 of 23
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 6
TC6005
8 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
Espiritu Santo Island and the northwestern part of Malakula
Island (Figure 5) [Jouannic et al., 1980; Taylor et al., 1987,
1990; Taylor, 1992; Cabioch and Ayliffe, 2001]. The preHolocene reefs indicate a geographic distribution of uplift
identical to the Holocene reef and reveal four semiindependent segments composing the CNH forearc. These
arc segments include southern and northern Malakula Island
and southern and northern Espiritu Santo Island (Figure 5)
and are distinguished by having distinctly different late
Quaternary and paleoseismic uplift histories [Taylor et al.,
1980, 1987, 1990; Greene et al., 1988].
5. The 230Th/234U Coral Dates and
Stratigraphy for Southern Espiritu Santo and
Nearby Islands
[14] We drilled through the Holocene reef at site 1 on
southern Espiritu Santo Island and at site 3a on a small
offshore island (Figure 5) and dated both subsurface and
surface samples by TIMS 230Th/234U [Cabioch et al.,
1998, 2003; Cabioch and Ayliffe, 2001; Cutler et al.,
2003] (Table 1). The surface samples come from erosional
or nondepositional ‘‘windows’’ and the drill cores sampled
coral limestone underlying the Holocene part of the 0 –
130 ka coral reef sequence at sites 1 and 3a (Figures 5 and 6).
The TIMS 230Th/234U ages led to reevaluation of samples
previously dated by AS 230Th/234U at sites 3b and 4
(Figure 5) and their implications for the vertical tectonic
history of the rest of Espiritu Santo and Malakula islands.
5.1. Reefs and 230Th/234U Ages at Sites 1 – 4,
Southern Espiritu Santo and Offshore Islands
[15] At site 1 (Figure 5) on southwestern Espiritu Santo
Island, the reef flat of a 6 ka mid-Holocene reef is at 35 m
ASL. About 2 m of its emergence is due to hydroisostatic
effects [e.g., Cabioch et al., 1989; Miyata et al., 1990;
Peltier, 1998], but tectonic uplift at 5.5 mm/yr is respon-
TC6005
sible for the other 33 m. However, immediately behind 35 m
ASL Holocene reef flat is an emerged sea cliff that rises to
45 m ASL. Two fossil corals from this cliff gave TIMS
230
Th/234U ages of 210.9 ± 1.7 ka and 222.1 ± 2.9 ka
(Table 1, TM-GA and TM-GB). Yet, farther uphill is a
sequence of reefs with increasing ages back to 130 ka at
400 m ASL. Similar sequences of 0 – 130 ka reefs also
occur 10 km east at island site 2, on southeastern Espiritu
Santo Island (site 3b), and at site 4 on Malo Island a few km
offshore of southeast Espiritu Santo Island (Figures 5 and
6) [Neef and Veeh, 1977; Jouannic et al., 1980, 1982;
Gilpin, 1982; Urmos, 1985]. Because the 130 ka reef at
site 1 is now 400 m ASL and paleosea level at 130 ka
was 6 m ASL, the 130 ka reef must have uplifted
394 m. However, we stated above that 33 m of the
Holocene uplift occurred from 6 ka to present. Everything
in the vicinity had to undergo this 33 m of uplift including
the 130 ka reef and the 216 ka reef. Therefore before
6 ka, the 130 ka reef uplifted 394 –33 = 361 m from 130
to 6 ka at a mean rate of 2.9 mm/yr (Figure 6a).
[16] The isotopic ages and heights of the fossil reefs at
sites 1 and 2 combined with paleosea level history allow us
to reconstruct the sequence of vertical motions (Figures 6a
and 6b). The reef at 35– 45 m ASL at site 1 corresponds to
the interglacial sea level highstand at 216 ka (MIS 7c) in
oxygen isotope records [Imbrie et al., 1984; Martinson et
al., 1987] when sea level was near present [e.g., Taylor
and Bloom, 1977; Harmon et al., 1983]. Yet, because the
216 ka reef at site 1 preexisted the 130 ka reef now
located inland at 400 m ASL, the 216 ka reef had to
undergo all of the 394 m of uplift that raised the 130 ka
reef. Because the 216 ka reef is now at 35 –45 m ASL, it
must have been 350 m BSL when the 130 ka reef grew
in order to have undergone the 400 m of uplift and now be
only 45 m ASL. An interpretation of the 216 ka reef as
having grown during a low sea level would account for no
more than 100– 150 m of the 350 m difference between
the present elevations of the 130 and 216 ka reefs.
Figure 6. Vertical tectonic histories at four sites on southern Espiritu Santo and adjacent islands. (a) The Holocene reef at
site 1 (Tasmaloum, southwest Espiritu Santo Island) which reaches to 35 m ASL and has uplifted 5– 6 mm/yr. Inland from
site 1, the 130 ka reef is 400 m ASL. No reefs occur above 400 m ASL inland from this reef [Gilpin, 1982]. Behind the
Holocene reef flat at 35– 45 m ASL (in black) is a reef containing surprisingly well preserved fossil corals that gave
230
Th/234U ages of 216 ka (Table 1). At least 350 m of subsidence at a mean rate of 4 mm/yr from 220 to 130 ka
followed by uplift of 400 m at a mean rate of 3 mm/yr is required to account for this sequence. (b) Site 2 (Araki Island),
10 km east of site 1, which has a Holocene reef to 28 m ASL. Coral ages indicate that the flat summit of the island formed
105 ka (MIS 5c) and that 130 ka (MIS 5e) and older corals are buried beneath this reef [Urmos, 1985]. This suggests
that subsidence continued past 130 ka and uplift began 105 ka. (c) Site 3a (Urelapa Island) which is an offshore island and
site 3a which is a former island that became connected with Espiritu Santo Island when the intervening seafloor emerged.
At site 3a we drilled completely through the Holocene and latest Pleistocene reef and sampled 22 ka corals at 55 m BSL
that grew during the last glacial maximum when sea level was 120 m BSL. From 55 to 70 m BSL we sampled much older
coral limestone that gave TIMS ages of 288 and 333 ka (Table 1). On the flanks of the steep hill composing site 3b, well
below its peak, we had previously found corals of 170 ka although the summit at 220 m morphologically correlates to
the 130 ka reef (MIS 5e) [Jouannic et al., 1980, 1982; Gilpin, 1982]. The older corals underlying the Holocene coral
limestone require that this part of Espiritu Santo Island must have subsided at least 290 m between 333 and 130 ka and then
uplifted 220 m at a mean rate of 1.9 mm/yr. (d) Farther east site 4. A well-preserved reef within the Holocene – 130 ka
reef sequence gave 230Th/234U ages of 223 + 44/ 30 ka and 175 ± 20 ka. We interpret this to represent a 216 ka (MIS 7c)
highstand reef that records subsidence during the interval between deposition of the 216 ka and 130 ka reefs and subsequent
uplift since 130 ka at a mean rate of 0.7 mm/yr.
9 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
Figure 7. Comparison of the vertical tectonic histories of
the four sites on and near southern Espiritu Santo Island
(Figure 6) plotted at same scale. The scale at the right shows
the relative magnitudes of the vertical motions and not
absolute altitude. Overall, the four sites indicate a very
similar history with the amplitude of vertical deformation
increasing from east to west (site 4 to 1) consistent with other
uplift data (Figure 5). We conclude that the change from
subsidence to uplift most likely occurred at 130 ka or later.
Figure 6a simply starts with the Holocene reef altitude and
graphically shows the slope of the change in height with time
required to get it from its paleosea level of 2 m ASL to 35 m
ASL. The next reef also has a line connecting its paleosea
level to its present altitude, but this reef has to include
exactly the uplift that the Holocene reef underwent. By this
method we construct the graph step by step back in time until
the 216 ka reef which underwent all of the 130 ka to
present uplift and before that, 350 m of subsidence, in
order to arrive at its present height of 45 m ASL.
[17] Fossil corals from holes drilled at island site 3a
(Figures 5 and 6c) and surface samples gave isotopic ages
for a vertical sequence of last glacial maximum to modern
(21 ka to modern) reef extending from 55 m BSL to 20 m
ASL [Jouannic et al., 1980, 1982; Cabioch et al., 2003;
Cutler et al., 2003]. This reef sequence indicates a mean
uplift rate of 3 mm/yr since 21 ka [Cabioch et al.,
2003]. However, deeper than 55 m BSL, this reef is
underlain by fossil corals at 58 m, 64 m, and 75 m BSL
that gave 230Th/234U TIMS ages of 288.4 ± 8.1 ka, 333 ±
10 ka, and 412 ± 35 ka, respectively (Table 1, samples
10-254; 10-290, and 11-145). Site 3b is a flat-topped hill
on Espiritu Santo Island with reef terraces up to the
summit at 220 m ASL. Corals gave AS 230Th/234U ages
of 37 ± 4 and 38 ± 4 ka at 41 m ASL and 149 ± 24 ka
TC6005
near the summit at 220 m ASL [Jouannic et al., 1980,
1982; Gilpin, 1982]. However, two corals from an intermediate terrace level at 160 m ASL gave AS 230Th/234U
ages of 176 ± 32 and 160 ± 24 ka [Jouannic et al., 1980,
1982]. These dates indicate that the Holocene to 149 ±
24 ka reefs (assume 130 ka given the large error) are
underlain by older coral reefs roughly corresponding to
interglacial high sea levels of MIS 7c.
[18] The 333 –412 ka corals drilled from 55– 75 m BSL
at site 3a (Figures 5 and 6c) correspond to interglacial MIS
9 to 11. Subsidence of at least 150 –270 m is required for
300– 400 ka reefs to be at 55– 75 m BSL now and yet find
the nearby 130 ka reef at 220 m ASL (Figure 6c) because
the 300– 400 ka reefs had to uplift >200 m along with the
130 ka reef. The AS 230Th/234U coral ages of 176 ± 32 and
160 ± 24 ka coral ages from beneath the Holocene –130 ka
reefs at site 3b support subsidence and uplift consistent with
the reefs drilled at island site 3a (Figure 6c).
[19] Site 4 is on the western end of an island off
southeastern Espiritu Santo Island has uplifted most slowly
of the four sites (Figures 5 and 6d). AS 230Th/234U coral
ages from site 4 (Figures 5 and 6d) date Holocene to 134 ±
10 ka reef terraces from sea level up to 100 m ASL
(Figures 5 and 6d) [Neef and Veeh, 1977; Jouannic et al.,
1980, 1982; Gilpin, 1982]. Yet, within this sequence at 38 m
ASL a small reef terrace gave AS 230Th/234U ages of 223 +
88/ 60 ka and 170 ± 30 ka [Jouannic et al., 1980, 1982;
Gilpin, 1982]. If the 223 +88/ 60 ka and 170 ± 30 ka reefs
that occur at 38 m ASL grew during an interglacial high sea
level, then 50– 60 m of subsidence occurred between the
times of their deposition and the last interglacial (MIS 5e) at
130 ka (Figure 6d).
5.2. Subsidence as an Explanation for Reef
Morphostratigraphy on Espiritu Santo and
Malakula Islands
[20] Hundreds of meters of subsidence followed by
similar amounts of uplift accounts for the mesa-like morphology of the emerged eastern plateau limestones of
Espiritu Santo Island and island site 4 as well as sites 2
and 3b, which are flat-topped atoll-like structures. Figure 7
combines the late Quaternary vertical tectonic histories of
the four southern Espiritu Santo Island areas into one figure
plotted at the same scale. The change from subsidence to
rapid uplift occurred within the time interval of 170 to
100 ka. At site 4, the 10 m separation between the last
interglacial (130 ka) reef and next lower terraces suggests
that subsidence continued until 100 ka. Similarly, at site 2
the summit age is 105 ka and at a lower level is a reef of
about 143+14 to 154+20 ka. Whether this older reef is really
a 130 ka reef or something older, it indicates that subsidence continued after deposition of the 130 ka reef.
[21] Whether or not the drowned reefs off the eastern
shores of Espiritu Santo and Malakula islands [Katz, 1988]
drowned during the most recent episode or a previous one,
they provide evidence for the concept that hundreds of
meters of subsidence occurred prior to the ongoing phase of
uplift. Likewise, the presence of coral limestone to heights
of 800 m on Espiritu Santo and Malakula islands provide
10 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 8. Idealized sequence of forearc vertical motions caused by impinging bathymetric features.
(a) Bathymetric feature approaching the forearc. The peripheral slope of the seamount begins to
subduct and raise the edge of the forearc. (b) Upper slopes of the impinging object that have lifted the
edge of the arc several kilometers from its normal depth of 4 km. Compression of the forearc has
begun with deformation and uplift of the thin leading edge of forearc. (c) Upper flanks of the
impinging edifice that fully engage the hard rocks of the forearc across a broad front tens of kilometers
long with vertical dimensions of up to 3 –4 km. The leading edge of the plate is shortening and
uplifting at rates exceeding 5 mm/yr. Compressive force is transmitted across the arc causing some
forearc convergence to be accommodated on back-arc thrusts. (d) Seamount or other object that is
somehow ‘‘digested’’ by the subduction zone as indicated by the rapid subsidence rates that we have
documented in both the CNH and NGG forearcs. This may be accomplished by rupture propagation
through the seamount or ridge or by some other mechanism, but the rapid subsidence over 100 to
200 ka suggests that collapse occurs on timescales similar to uplift.
evidence of previous uplift that was not removed by the
most recent subsidence event that we consider here. The
buried 200– 400 ka reefs at several sites on Espiritu Santo
Island confirm the widespread occurrence of older reefs at
low levels and can be explained only by rapid regional
subsidence followed by abrupt change to uplift portrayed in
Figures 6 and 7.
[22] It is likely that all of Espiritu Santo and northern
Malakula islands shared the subsidence as completely as the
subsequent uplift from 130 ka to present. There is no
evidence that the four CNH arc segments have moved
hundreds of meters vertically relative to one another. The
continuity of the 130 ka summit plateau reefs from
southeastern to northeastern Espiritu Santo Island [Gilpin,
1982] indicate that northern and southern Espiritu Santo
Island have moved almost synchronously. Northern Malakula Island tilts in a slightly different direction, but its
130 ka to Holocene uplift history is similar. Stratigraphy
on northern Malakula Island indicates subsidence preceding
the ongoing uplift [Mitchell, 1971]. Much of southern Malakula Island also shares with northern Malakula and Espiritu
Santo islands a history of an earlier uplift, subsidence, and the
present accelerating uplift [Taylor et al., 1980].
[23] The Holocene reefs of back-arc Pentecost and Maewo
islands indicate Holocene uplift rates ranging from 0 – 1 mm/
yr (Figure 5). However, this ongoing uplift began very late in
11 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
Pleistocene time based on the absence of a sequence of coral
reef terraces ranging from Holocene to 130 ka that should
exist from the Holocene reef to tens of meters ASL if
continuous uplift extended back to 130 ka. On the basis
of this evidence, and the supposition that uplift of Pentecost
and Maewo islands is related to eastward thrusting of the
entire CNH, that backthrusting began long after the initial
impingement of the Bougainville Guyot and after uplift of the
western part of the forearc had begun. This is consistent with
the estimate of Lagabrielle et al. [2003] that the back-arc
thrust fault began to move in the time range of 15– 37 ka.
6. Interpretation of the Effects of
Bougainville Guyot Impingement at the CNH
[24] We propose a direct cause-and-effect relationship
between (1) Bougainville Guyot impingement on the forearc, (2) cessation of subsidence of Espiritu Santo and
Malakula islands by 130 ka, (3) initiation of uplift of
Espiritu Santo and Malakula islands at130 ka, (4) eastward displacement of the CNH, and (5) renewed uplift of
the back-arc Pentecost and Maewo islands in response to
backthrusting of the CNH.
6.1. Impingement of the Bougainville Guyot and
Eastward Motion of the Central New Hebrides
[25] Given mean Quaternary convergence rates of
100 mm/yr, the peripheral slopes of the Bougainville
Guyot began subducting between Espiritu Santo and
Malakula islands by 150 ka (Figure 8a). Resistance to
subduction gradually increased as steeper slopes of the
Bougainville Guyot slope began to subduct (Figure 8b).
By 100 ka and 5 – 10 km from its present location, the
steep upper wall of the Bougainville Guyot began to
impinge on the inner trench wall of the forearc. For the
first 75 kyr after Bougainville Guyot first impinged, only
the forearc may have deformed. The area of contact and
the force applied to the CNH arc by the Bougainville
Guyot increased exponentially, adding significantly to the
nearly continuous force exerted west of Espiritu Santo
Island by the subducting d’Entrecasteaux Ridge and Torres
Plateau. The horizontal force applied to the arc by the
d’Entrecasteaux Ridge and Bougainville Guyot grew so
great that a 250-km-long section of the arc began to be
displaced eastward (Figure 8c) [F. W. Taylor et al., 1995].
Uplift of the back-arc Maewo and Pentecost islands by
15 –37 ka resulted from reverse faulting accommodating
eastward CNH motion [Collot et al., 1985; F. W. Taylor et
al., 1995; Pelletier et al., 2000; Calmant et al., 2003;
Lagabrielle et al., 2003]. In the future, Bougainville Guyot
will cease to cause strong coupling and the forearc uplift
will collapse as it has in the past (Figure 8d).
6.2. Role of Seismic Rupture in the Bougainville Guyot
Impingement and Subduction Process
[26] Interplate seismic ruptures mark increments of
plate motion along the intermittently locked interplate
thrust between the Australian plate and the CNH forearc.
TC6005
Large interplate thrust events of Ms 7.0– 7.75 have
accompanied interplate slip on the order of one to a few
meters [Taylor et al., 1990]. These events and their
aftershocks define the rupture zone geometry from depths
of 15– 50 km [Chinn and Isacks, 1983] (Figure 4). We
had previously assumed that uplift near the times of these
events was directly related to the elastic strain cycle
associated with the intermittently locked rupture zone
[Taylor et al., 1980, 1990].
[27] Here we propose a different relationship in which
most of the coseismic uplift centered in western Espiritu
Santo and northwestern Malakula islands is only indirectly
related to elastic strain accumulation and release on the
main thrust zone at 15 to 50 km depth. Instead, the key
process is the advance of the Bougainville Guyot and other
parts of the DR by several meters, which happens to occur
at the time of seismic rupture. Uplift is produced because
the upper part of the interplate thrust zone slips less or much
less than the amount of slip on the deeper seismogenic zone.
Impingement of the d’Entrecasteaux Ridge and the Bougainville Guyot creates very strong coupling between the
Australian plate and forearc at depths from seafloor to
15 km. The interplate thrust zone to 15 km deep is
locked and does not rupture or slip significantly. Instead,
the outer forearc responds to several meters of coseismic
interplate slip by shortening and ‘‘upbowing’’ which we
see as island uplift unusually close to the trench. This
explanation is consistent with observations that uplift of
southern Espiritu Santo and northern Malakula islands
occurred during the hours after the 11 August 1965 main
shock rather than during the main event [Taylor et al., 1980].
We also noted the unexpected absence of evidence for
interseismic subsidence in the zones of uplift associated
with the 11 August 1965 earthquakes [Taylor et al., 1987].
However, the uplift mechanism proposed in this paper does
not require significant interseismic subsidence prior to uplift
if the uplift that we are discussing is not significantly related
to accumulation and release of elastic strain.
[28] We propose that the August 1965 earthquakes and
associated deformation are typical events in the process by
which the d’Entrecasteaux Ridge and Bougainville Guyot
have advanced on the CNH forearc during a series of
interplate ruptures on the 15 to 50 km deep interplate thrust
zone. Repetition of deformation similar to that of August
1965 has progressively shortened the CNH trenchward
forearc and caused hundreds of meters of net forearc uplift
across Espiritu Santo and Malakula islands (Figure 8).
Elastic deformation related to the interplate seismogenic
zone must occur, but does not require accumulation of any
vertical deformation.
6.3. Fate of Subducting Seamounts and Their Tectonic
Effects at the Central New Hebrides Arc
[29] Force applied by the impinging Bougainville Guyot
to the CNH is at least partly self-limiting and its tectonic
effects may terminate as abruptly as the onset of uplift
began. Termination of strong coupling could follow shearing off the Bougainville Guyot or some other mechanism to
drastically reduce interplate coupling. If this occurs, the
12 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
Figure 9. Eastward motion of the central New Hebrides
has geometrical implications for the interplate thrust zone.
GPS measurements show Australian – New Hebrides convergence of >100 mm/yr farther south at Efate and Tanna
islands but only 40 mm/yr at CNH sites [F. W. Taylor et
al., 1995; Calmant et al., 1995, 2003]. If we assume that
30 mm/yr of convergence is hidden as elastic strain, then
the longer-term mean convergence rate is 70 mm/yr. In
this case the central New Hebrides would be moving
eastward 30 mm/yr. Suppose that while the central New
Hebrides retreats eastward by 3 km the Australian plate and
the Bougainville Guyot continues subducting along its
present trajectory which remains fixed. Over 100 kyr the
Bougainville Guyot would penetrate another 7 km into the
forearc, but it also would subside 1 km. If the entire
central New Hebrides arc moves eastward and the
Australian plate profile remains fixed and does not bulge
eastward, then a gap would try to open along the interplate
thrust zone between the Australian plate and the arc. A
gap cannot open so the upper plate must sag to stay in
contact with the Australian plate. Thus eastward retreat of
the central New Hebrides may reduce stress across the
interplate thrust zone leading to easier subduction of the
Bougainville Guyot and d’Entrecasteaux Ridge and may
lead to decoupling of the Bougainville Guyot and collapse
of the uplifted forearc.
Bougainville Guyot will cease pushing against the forearc,
the fully locked uppermost interplate thrust zone may
decouple and begin to slip, and the presently uplifted
forearc may rapidly subside as it did prior to the present
uplift.
TC6005
[30] For this and other reasons, arc polarity reversal is
unlikely. For example, resistance to back-arc thrusting is
likely to increase as the length of the back-arc thrust faults
increase and the back arc rides onto the North Fiji Basin. To
initiate back-arc subduction, hinge faults must form near the
ends of the thrust so that a flap of North Fiji Basin
lithosphere could descend into the asthenosphere beneath
the back arc. This is conceivable, but additional geometric
changes associated with Bougainville Guyot impingement
zone that make this unlikely.
[31] The CNH is moving eastward relative to the cross
section or trajectory of the downgoing Australian plate
(Figure 9) that is anchored in the asthenosphere. Eastward
motion of the CNH relative to the Australian plate trajectory
reduces interplate coupling as a gap tries to open between
the Australian plate and the CNH lithosphere. This potential
gap must be filled either by the Australian plate profile
bulging upward and eastward or by the CNH forearc
subsiding to maintain contact with the Australian plate.
For example, suppose that while the Australian plate advanced 10 km along its trajectory, the CNH retreated 5 km.
The western edge of the forearc would have to subside to
remain in contact with the top of the Australian plate. The
Bougainville Guyot also would descend from its present
depth of 1 km to 2 km as it moved 10 km farther east
along the Australian plate trajectory, but only 7 km toward
the retreating CNH.
[32] However, at present, the two plates remain strongly
coupled both at the upper end of the interplate thrust zone
where the Bougainville Guyot is impinging and at depth
where the interplate thrust zone has ruptured repeatedly.
This suggests that total eastward thrusting of the CNH has
been minimal and that the Australian plate– CNH interface
has thus far been able to deform to fill the gap opening
between the two plates and keep them firmly coupled.
Minimal eastward transport of the CNH is consistent with
the observation that backthrusting and uplift at Pentecost
and Maewo islands began much later than uplift of Espiritu
Santo and Malakula islands.
[33] If backthrusting of the CNH does not lead to arc
polarity reversal, then eventually, the Bougainville Guyot
must be subducted or obducted (Figure 8). It is possible that
(1) the main thrust zone may propagate through the Bougainville Guyot and decapitate it so that its top is accreted to
the upper plate; (2) the Bougainville Guyot may break up
progressively; or (3) Bougainville Guyot may plow a furrow
through the base of the upper plate and return to the
asthenosphere more-or-less intact. If the Bougainville
Guyot is subducted farther beneath the forearc then it
may leave a forearc reentrant as off northern Malakula
Island that has been proposed as the site of subduction of
a seamount in the past [Fisher et al., 1986].
[34] Collapse of the forearc during the interval from
400 – 130 ka at mean rates up to 3 mm/yr across
southern Espiritu Santo Island provides clues as to how
the forearc processes impinging topographic features. We
suspect that this collapse and the preceding uplift of the
forearc may correspond to impingement and subduction of a
pre-Bougainville Guyot seamount [e.g., Collot and Fisher,
13 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 10. Map showing plate tectonic setting and Holocene uplift rates in the New Georgia Group,
western Solomons arc [after Mann et al., 1998]. Nearly all islands are uplifting, but only those areas
having 5 m of Holocene emergence are contoured. Maximum uplift has occurred at Rendova and
Tetepare islands in an elongate zone parallel to the plate boundary. Paleobathymetric data from
microfossils show that a major pulse of uplift at Tetepare and southern Rendova began after 270 ka,
followed by an erosional unconformity near sea level, subsidence to several hundred meters depth and
finally the ongoing uplift [Mann et al., 1998]. Note the relationship of Coleman Seamount to the trench
and the location of the drilling site on Rendova Island.
1989]. The rapid subsidence suggests an abrupt decrease in
coupling between the downgoing plate and the CNH.
Shearing off of a paleoseamount is a mechanism that might
allow such a sudden decoupling of the forearc if the
seamount is the key element locking the shallow part of
the interplate thrust zone (Figure 8d) [e.g., Cloos, 1992].
The eastward retreat of the CNH from impinging objects
also may contribute to abrupt decoupling by reducing stress
across the interplate thrust zone (Figure 9).
7. Vertical Tectonic History of the New
Georgia Group Forearc, Western Solomons
Arc
[35] Neotectonic studies of the NGG, western Solomon
Islands arc, document a history of subsidence and uplift that
is very similar to that of the CNH in terms of geography,
rates, timescales, and amplitude of vertical motion [Mann et
al., 1998] (Figures 10, 11, and 12). However, the change
from subsidence to uplift occurred more recently than at
CNH. Most of the details of the NGG uplift history are
presented in a previous paper [Mann et al., 1998]. Here we
summarize previous work and present a number of new
results.
7.1. Subduction of the Woodlark Spreading System
at the New Georgia Group Forearc
[36] The NGG segment of the Solomons arc is being
underthrust by 1 – 3 Ma Woodlark Basin lithosphere generated at the active Woodlark seafloor spreading system
[Taylor and Exon, 1987; Crook and Taylor, 1994; B. Taylor
et al., 1995]. Woodlark lithosphere subducting at NGG is
part of the Australian plate and converges 97– 100 mm/yr
toward N75E relative to the Pacific plate [DeMets et al.,
1990, 1994; Beavan et al., 2002]. Of the total AustralianPacific relative motion, some unknown fraction is accommodated by convergence across the North Solomons Trench
[e.g., Phinney, 1997; Phinney et al., 1999]. However, GPS
14 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 11. Reconstructed uplift history for central Rendova Island, New Georgia Group. The Holocene
reef (6 ka) reaches 20 m ASL and has uplifted at 3 mm/yr at this location. Drilling penetrated the Last
Glacial Maximum shoreline (21 ka) at a depth of 55 m BSL showing that the 3 mm/yr uplift rate
extends back to at least 21 ka (Table 2). Beneath this reef is very altered older coral limestone having
features typical of subaerial exposure. There are reefs uphill from the 20 m ASL Holocene reef surface,
but they also are extremely weathered and are much older than the 40 to130 ka time range. Therefore
uplift that has raised the Holocene reef to 20 m ASL and the 21 ka reef to 55 m BSL began recently,
probably after 50 ka. Prior to the uplift, subsidence is required to account for the older reefs beneath the
21 ka reef and to explain the absence of 40 –130 ka reefs up hill from the Holocene reef. Prior to that
subsidence, an earlier period of uplift raised the older reefs that we now find at 50– 300 m altitude. This
interpretation is consistent with the paleowater depth data of Mann et al. [1998] from Rendova and
Tetepare islands.
measurements indicate that at least 80 mm/yr of this
convergence is accommodated by subduction at the NGG
(D. A. Phillips et al., manuscript in preparation, 2005).
[37] Woodlark seafloor is topographically rugged with
the distal slopes of two young volcanic seamounts presently
slipping beneath the plate boundary [Crook and Taylor,
1994]. The closest of these to the trench, Coleman seamount, rises 3 km above the seafloor from a depth of
4 km. Mann et al. [1998] concluded that interaction of
these volcanic edifices with the forearc is probably
responsible for the ongoing pulse of rapid forearc uplift.
We assume that other rugged topographic features have
subducted beneath the NGG islands.
[38] Another major difference between subduction at
the NGG and CNH is that there are no known large
interplate thrust events and relatively little shallow interplate seismicity in the NGG forearc from Gizo Island
eastward past Rendova and Tetepare islands (Figure 10).
This observation has led to speculation that subduction
has proceeded by aseismic creep because the downgoing
lithosphere is too young and warm for seismic rupture
[Cooper and Taylor, 1987; Shinohara et al., 2003].
However, it is possible that the low seismic energy
release in this area reflects locking of the interplate thrust
zone and storing of elastic energy that will be released
later.
7.2. Vertical Tectonics of the New Georgia Group
Forearc Islands
[39] Microfossil biostratigraphy and paleobathymetry for
the rapidly uplifting trenchward edge of the forearc occupied
by Tetepare and southern Rendova islands show that from
1 Ma until some time after 270 ka, the islands rose from
abyssal depths to wave base [Mann et al., 1998]. Then uplift
ceased and these same islands reversed direction and subsided to a water depth of 500 m BSL. Renewed uplift
elevated the islands tens of meters above sea level as
recorded by subaerial mid-Holocene reefs (Figures 10 and
11). The intense deformation on Tetepare Island and southern Rendova Island resulted in folding of Late Pleistocene
sedimentary sequences underlying the coral reefs [Mann et
al., 1998].
[40] Previous work had established that the NGG islands
have undergone hundreds of meters of subsidence prior to
the ongoing uplift [Stoddard, 1969; Mann et al., 1998].
15 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
TC6005
Figure 12. Comparison of vertical deformation profiles across the CNH and NGG arc segments. Two
quite different settings display very similar vertical tectonic responses to forearc impingement by rugged
topography on the downgoing plate. The profiles were selected across islands where uplifted reefs record
uplift rates. The profiles were adjusted so that the maximum uplift peaks match. The maximum uplift at
NGG is 10 km closer to the trench than at the CNH. The similarity of these profiles suggests that the
deformation mechanism at NGG and CNH is similar even if the NGG convergence is by creep instead of
a stick-slip process.
Nearly all NGG island coastlines have uplifted in Holocene
time with rates up to 8 mm/yr on parts of Tetepare Island
(Figure 10). Most of New Georgia and Vangunu islands
uplifted more slowly with maximum rates of 1.2 mm/yr.
On New Georgia Island, the maximum heights of late
Quaternary coral limestone associated with the ongoing
uplift require that the ongoing uplift began in latest
Pleistocene time (Figure 10). For example, at one site
on southern New Georgia Island where the Holocene reef
reaches 8 m ASL the maximum height of older coral
limestone is only 30 m ASL [Mann et al., 1998].
Therefore 25% of the total reef emergence has occurred
just since mid-Holocene time at 6 ka. The change from
subsidence to uplift has been very recent.
[41] Our new data are from 1999 drilling through the
Holocene reef on Rendova Island during which we recovered Last Glacial Maximum corals (19.2 to 22.35 ka;
Table 2) at 52 to 56 m BSL (Figures 10 and 11). A much
older coral reef underneath the 21 ka corals in the drill hole
is too altered for isotopic dating, but is consistent with
subsidence prior to the ongoing uplift. Because sea level
was 120 m BSL at 21 ka, corals of this age found at 55 m
BSL have uplifted 65 m at a mean rate of 3 mm/yr. This
rate matches the mid-Holocene to present uplift rate of
3 mm/yr based on the nearby mid-Holocene reef flat that
reaches 20 m ASL. Because the 3 mm/yr uplift rate extends
back to 21 ka the initiation of uplift is likely to have been
even more abrupt than otherwise.
[42] The lack of young emerged reefs inland from the top
of the 20 m ASL Holocene reef flat at our Rendova drill site
(Figures 10 and 11) further constrains the amount of uplift
prior to 21 ka (Figure 10) and strengthens the case for a
recent initiation of uplift. For example, paleosea level
history shows a high sea level reaching 56–59 m BSL at
55 ka [Chappell et al., 1996]. To be exposed at central
Rendova Island, a 55 ka reef must be higher than the
Holocene level of 20 m ASL and must, therefore have
uplifted from 59 m BSL to perhaps 25 m ASL (84 m total)
at a mean rate of 1.53 mm/yr for the reef to be higher than
the Holocene reef. However, we just showed that the mean
uplift rate has been 3 mm/yr since at least 21 ka which
requires that the 21 ka reef and everything older, including
corals that grew at 55 ka, were uplifted 63 m since 21 ka.
Therefore only 21 m of uplift between 55 ka and 21 ka
would be required for the 55 ka reef to be exposed inland of
the mid-Holocene reef. Nevertheless, no 55 ka or any
other MIS 3 or MIS 5 reefs are subaerially exposed on
Rendova Island. This indicates that the initiation of uplift
prior to 21 ka was quite abrupt and that uplift may have
begun even later than 50 ka but before 21 ka.
[43] In summary, our results show a remarkable sequence of changes since 270 ka: uplift, subsidence, and,
finally, abrupt transition from subsidence to rapid ongoing
uplift of the NGG forearc. With this many changes in
such a short time, it is not surprising that we find that the
ongoing uplift began by 50 ka or later (Figure 11). Plate
convergence during the entire sequence of uplift-subsidence-uplift since 300 ka has been 30 km, but only
20 km normal to the arc trend. If the latest change from
subsidence to occurred as early as 50 ka, then Australian
plate convergence during all the uplift has been 5 km.
During the 63 m of uplift since 21 ka at the Rendova drill
site, there has been 2.1 km of convergence of which
1.4 km was in the direction normal to the arc trend. As
16 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
at the CNH, GPS measurements suggest that elastic strain
is accumulating and that the Holocene uplift pattern is
consistent with the shallow end of the interplate thrust
zone being fully locked for many thousands of years
(D. A. Phillips et al., manuscript in preparation, 2005).
Taylor et al. [2003] presented paleouplift evidence of
several distinct 1 – 2 m uplifts at Rendova Island over
the past 2,000 years. This does not prove that seismic
ruptures occurred, but it does suggest that interplate slip
may proceed episodically and thus generally in the style
interplate rupture at the CNH. We conclude that the NGG
forearc uplift process is similar to that of the CNH
forearc.
8. Elastic Modeling of the Interplate Thrust
Zone at the Central New Hebrides Arc
[44] A great deal is known about the seismicity and
paleoseismicity of the CNH. GPS measurements provide
quantitative data regarding ongoing deformation. We used
earthquake hypocenters to infer the geometry of the CNH
interplate thrust zone for elastic models. Less is known
about the NGG interplate thrust zone, but there are enough
features in common that we suggest that the model applies
to both areas. The purpose of our model is that it simulates
uplift of the trenchward edge of the upper plate where we
suggest that impinging seamounts have caused full or partial
locking of the shallow end of the interplate thrust zone.
Regardless of the precise geometry of the interplate thrust
zone and the mode of interplate slip, this principle applies
wherever the shallow end of an interplate thrust zone is
locked or slips much less than the convergence rate. The
result in either case is that the upper plate has to shorten to
accommodate the convergence.
[45] We have conceptualized the tectonic processes at the
CNH and NGG forearcs and integrated plate convergence,
interplate seismic rupture, and arc deformation. In previous
studies we modeled the CNH subduction zone assuming the
full Australia-Pacific convergence rate of about 100 mm/yr
and with a conventional locked interplate thrust zone at
depths of 15– 50 km that intermittently ruptured [Taylor et
al., 1980, 1987; F. W. Taylor et al., 1995]. The amount of
interplate slip accounted for by historical large thrust events
was much less than the convergence rate, so that we had to
invoke a significant proportion of aseismic slip [Taylor et
al., 1987, 1990]. Since then, GPS measurements have
shown that the convergence rate is only 40 mm/yr, much
less than the 100 mm/yr previously assumed [F. W. Taylor et
al., 1995; Calmant et al., 1995, 2003], so that little if any
aseismic slip is required. We propose that subduction of the
d’Entrecasteaux Ridge and Bougainville Guyot so strongly
resist underthrusting that the uppermost part of the interplate
thrust zone is permanently locked. This requires a new
modeling approach.
[46] The main CNH thrust zone at 15 – 50 km depth
intermittently ruptures and slips on the order of several
meters per event (Figure 4). This ‘‘stick-slip’’ behavior of
the plate interface leads to a cycle of sustained interseismic
TC6005
strain accumulation in the upper plate and occasional
coseismic release of elastic strain energy. However, for
vertical deformation to accumulate over thousands of years,
producing hundreds of meters of topographic change, there
must be some (nonelastic) deformation in the upper plate
that is not recovered on the interseismic timescale. We
propose that this occurs when the upper part of the plate
interface (perhaps 0 – 15 km) is locked permanently (or at
least for thousands of years) in response to the ‘‘collision’’
of inbound topography with the plate boundary (Figure 13).
We suggest that further down the plate interface, immediately below this permanently locked zone of the CNH, is a
zone (perhaps 15– 50 km) that sticks and slips in more
conventional fashion. That is, the anomalous segment of the
plate boundary associated with impinging topographic features consists of both stick-stick and stick-slip zones. It is
the permanently locked (or stick-stick) zone that leads to the
rapid growth of topography in the upper plate, since plastic
failure of the upper plate is the mechanism that accommodates plate convergence when part of the plate interface is
never allowed to slip.
[47] It may seem strange that we model the deformation
associated with the permanently locked zone (PLZ) using an
elastic formalism. However, we propose that as plate
convergence proceeds, secular straining of the upper plate
is essentially an elastic process. However, as the PLZ never
breaks and the stresses in the upper plate inevitably exceed
the yield stress, the failure of this plate occurs by sequences
of small earthquakes on many faults and flexures located in
various parts of the upper plate. That is, the nonelastic
deformation of the upper plate is achieved through seismic
flow driven by stresses that accumulate through elastic
compression of the upper plate. Accordingly, we postulate
that the largest seismic deformations occur in that part of the
plate where plate convergence in the presence of a PLZ
causes the highest rate of elastic straining. We use secular
rate of elastic deformation associated with the PLZ as a
proxy for the longer-term deformation associated with the
growth of topography over thousands of years.
[48] We base our new model on the well-known back-slip
formalism of Savage [1983]. This model assumes that for a
typical subduction zone undergoing stick-slip behavior,
elastic strain accumulation in the upper plate during the
interseismic phase of the earthquake cycle will be very
similar to that associated with a dislocation coincident with
the locked portion of the plate interface but with the
opposite sense of shearing. In effect this back slip represents
the perturbation associated with locking of the plate boundary in that it exactly cancels the slip that would occur if the
plate boundary freely slipped. We modify this approach by
dividing the locked zone into a permanently locked zone
(PLZ) and a temporarily locked zone (TLZ), and keeping
track of their separate contributions to the displacement of
the leading edge of the upper plate. The net deformation is
assumed to represent the interseismic displacement field. If
the PLZ ruptures, the straining it produced would be
recovered, so only the surface velocity field associated with
back slip on the PLZ is used to estimate the growth of
topography. Again we assume that the straining associated
17 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
with this displacement is eventually released by small
earthquakes on faults that pervade the upper plate, and this
ensures the uplifts associated with locking of the PLZ are
not eventually recovered.
[49] We defined the interplate thrust zone profile based
on the relocated hypocenters of large interplate thrust events
in the CNH [Pascal et al., 1978; Ebel, 1980; Chinn and
Isacks, 1983] (Figure 13a). These earthquakes, the GPS
sites in the CNH (Figures 13b and 13c), and coral uplift data
TC6005
(Figure 13d) are projected into a profile trending N70 E
through southern Espiritu Santo Island (Figure 13e), perpendicular to the arc trend and parallel to relative plate
motion [e.g., Calmant et al., 2003]. Epicenters of aftershocks of the August 1965 CNH interplate thrust sequence
indicate that the zone from 15 to 50 km ruptured [e.g.,
Isacks et al., 1981], consistent with the analysis of maximum depth of New Hebrides rupture zones of Pacheco et
al. [1993]. We assume a convergence rate of the Australian
Figure 13
18 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
plate of 70 mm/yr, which requires that approximately 30–
50 mm/yr of convergence is accommodated by back-arc
thrusting [Pelletier et al., 1998, 2000].
[50] We implemented the back-slip model of Savage
[1983] using the formulas derived by Singh and Rani
[1993] for a 2-D elastic deformation (one which is infinitely
long in the strike direction). We applied a simpler version of
this methodology with stick slip at 15 – 50 km and an
uncoupled zone at 0 – 15 km depth in our previous analysis
of GPS data for the CNH [F. W. Taylor et al., 1995]. We
have adapted the original code that assumes a planar fault to
simulate a more realistic curve of the downgoing Australian
plate– New Hebrides arc interface by dividing the fault into
many small increments and adding them together. We have
held the shallow end of the fault from the seafloor to 15-km
depth locked through interseismic and coseismic cycles on
the deeper stick-slip zone (Figure 13a).
[51] When the elastic model runs, horizontal and vertical
elastic strain accumulates in a pattern dictated by an
interplate thrust zone that is continuously locked from
the seafloor at the plate boundary to a depth of 50 km
(Figures 13a and 13b). Interseismic strain produces eastward horizontal motion and subsidence of all of Espiritu
Santo and Malakula islands (Figure 13b). Horizontal strain
rates are consistent with measured GPS rates at four
stations from the western to eastern coasts of Espiritu
Santo Island. Model subsidence rates are on the order of
10 mm/yr across Espiritu Santo and Malakula islands.
[52] The next version of the model (Figure 13c) represents the coseismic vertical deformation that would occur if
the zone from 15 – 50 km depth on the interplate zone
ruptures and the shallow zone from seafloor – 15 km depth
remains fully locked. The upper curve represents uplift as a
percentage of the coseismic slip on the rupture zone at 15–
50 km. The lower curve shows the net vertical deformation
rate in mm/yr from running the model through numerous
interseismic and coseismic cycles on geological timescales
with a 70 mm/yr convergence rate. Interseismic subsidence
is much less than coseismic uplift in the case in which the
shallow end of the interplate thrust zone is always locked
and only the deeper zone is stick slip. The result is rapid
uplift of the forearc with maximum mean uplift rates on the
TC6005
order of 10 mm/yr comparing well with reef uplift rates. The
distribution of net uplift corresponds almost perfectly with
the actual distribution of uplift based on the total emergence
of the Holocene coral reefs and with topography of Espiritu
Santo and Malakula islands [Taylor et al., 1987]. With
only minor adjustments, this model applies equally well to
the rapid uplift of the trenchward edge of the NGG upper
plate.
9. Discussion
[53] Coral-based late Quaternary tectonic histories indicate that both the CNH and NGG forearcs underwent
hundreds of meters of rapid subsidence that ended with
the abrupt onset of rapid uplift (Figure 12). The short time
frames over which these events occurred differ from the
timescale of isostatic adjustments imposed by subducting
bathymetric features. The transition from subsidence to
uplift at the CNH and NGG required no more than a few
tens of thousands of years and perhaps much less.
[54] Rapid uplift at CNH and NGG indicates that impinging bathymetric features can create strong interplate
coupling or even complete locking of the shallow end of the
interplate thrust zone, which normally slips aseismically
[e.g., Pacheco et al., 1993; Oleskevich et al., 1999]. We
have proposed that slip on the shallow interplate thrust zone
ceased or became much less than the interplate convergence
rate not only near Bougainville Guyot, but also along most
of the CNH forearc. It appears that a similar process is
occurring at the NGG. At both the CNH and NGG, the outer
forearcs consist of relatively competent sedimentary and
crystalline rocks [e.g., Collot et al., 1992; Dunkley, 1986]
that would resist deformation. When strong interplate coupling occurs at the shallowest depths beneath the thin upper
plate, the upper plate can deform and shorten more easily
with greater uplift than when the strong coupling begins at
greater depths. Our elastic model provides a plausible
mechanism whereby full or at least partial locking of the
shallowest 15 km of the interplate thrust zone would
account for uplift at the trenchward part of the upper plate
at both the CNH and NGG forearcs.
Figure 13. Elastic model of the main interplate thrust zone compared with Holocene uplift and topographic profiles across
southern Espiritu Santo Island. (a) Elastic model. We keep the shallower end of the interplate thrust zone locked
(permanently locked zone, PLZ) from the seafloor to a depth of 15 km throughout multiple cycles of elastic strain
accumulation and release on the deeper temporarily locked zone (TLZ) from a depth of 15– 50 km. The locations of five
GPS sites are shown (C, Cape Lisburn; T, Tasmaloum; R, Ratard; S, Santo; and M, Maewo). (b) Rates of interseismic
horizontal and vertical deformation assuming that both the PLZ and TLZ are locked with plate convergence rate
(PCON) of 70 mm/yr. The horizontal strain rates from the model are consistent with GPS measurements of horizontal strain
at the five GPS sites along this profile. The component of deformation related to the TLZ will be removed when the TLZ
ruptures. Deformation associated with the PLZ will remain. (c) Vertical motion. The upper curve shows the modeled vertical
surface deformation for one rupture of the TLZ (with the PLZ remaining locked) in percent of coseismic slip. Again, part of
this deformation is recovered during the subsequent period of interseismic strain as shown in the panel immediately below.
The lower curve shows the mean uplift rate in mm/yr over multiple ruptures of the TLZ assuming the PLZ is always
locked. (d) Mean uplift rates since the mid-Holocene based on emerged coral reef terraces. We do not include subsidence
because we do not have accurate rates, but we do know that the central Aoba Basin is actively subsiding [Taylor et al., 1987].
The model net uplift rates are about half the measured uplift rates. Using a convergence rate of <70 mm/yr would reduce the
model uplift rate. (e) A topographic profile across the CNH reflecting the net vertical deformation on the 106 year timescale.
19 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
[55] The CNH and NGG have large differences in the age
and thermal state of the subducting lithosphere. However,
Oleskevich et al. [1999] suggest that interplate coupling
begins when the thrust zone reaches 100C and ends when
it reaches 350C. In the case of very young lithosphere,
100C and seismogenic coupling should begin at shallower
depths and end at relatively shallow depths as temperatures
rapidly rise along the interplate thrust zone to 350C. An
unusually narrow seismogenic zone due to young subducting lithosphere does not necessarily preclude strong interplate coupling and would be consistent with the series of
several 1 –2 m uplift events over the past 2000 years at the
NGG Rendova Island [Taylor et al., 2003]. Our proposed
uplift mechanisms would operate regardless of whether the
fault segment of the interplate thrust deeper than 15 km
moves by creep or by stick slip.
[56] Although seamounts such as Bougainville Guyot
may be sheared off allowing forearc collapse, larger more
continuous objects such as the main d’Entrecasteaux Ridge
may cause intense deformation upon first impinging on the
forearc. The leading parts of the ridge may plow a furrow in
the base of the upper plate that the trailing ridge may follow.
After this initial deformation, further effects are probably
related to variations in ridge topography. If the ridge impact
zone migrates along the arc trend, then resistance to
subduction will continue [e.g., Collot and Fisher, 1992].
However, because of the timing of the ongoing CNH uplift,
we suggest that impingement of the Bougainville Guyot is
the key event leading to locking of the shallow end of the
interplate thrust zone for tens of thousands of years. A
seafloor feature impinging on the NGG forearc may also
have initiated abrupt locking of that interplate thrust zone.
That feature may be Coleman Seamount, but the relationship is not as convincing as for Bougainville Guyot.
[57] Australian plate convergence with both the CNH and
NGG is accommodated in part by some amount of back-arc
convergence (Figures 2 and 5). The CNH is being shoved
eastward toward the North Fiji Basin [Collot et al., 1985;
F. W. Taylor et al., 1995; Lagabrielle et al., 2003] and the
NGG arc segment resists subduction so strongly that convergence continues at the North Solomons trench [Phinney
et al., 1999]. If resistance to subduction of the Bougainville
Guyot and d’Entrecasteaux Ridge continues to increase,
then Australian plate subduction may cease at least locally
and initiate a process of arc-polarity reversal. Even though
most impinging objects such as the Bougainville Guyot are
eventually subducted or decapitated, they cause extensive
tectonic modification to the arcs. Cross-arc faults and backarc thrusts, once formed, are zones of weakness that may be
exploited in the future when new objects impinge on the
forearcs.
[58] There is evidence of previous similar cycles of uplift
and collapse of the forearc, perhaps as earlier objects
impinged and then were ‘‘digested’’ by the subduction
zones. Both the CNH and NGG have undergone vertical
oscillations throughout the Quaternary Period that may have
been related to previous episodes of seamount or ridge
impingement [e.g., Dunkley, 1986; Macfarlane et al., 1988].
No detached seamounts are known in the CNH or NGG
TC6005
forearcs, but decapitation may occur at significant depth in
the subduction zone.
[59] Espiritu Santo Island’s western belt reaches nearly
2000 m ASL only about 40 km from the plate boundary
where a forearc is typically about 3000 –4000 m BSL. The
ongoing uplift phase has imposed 400 m of the total
2000 m of elevation ASL, but the late Quaternary uplift
distribution closely mimics topography. Similar island
altitudes and remnants of ancient limestone at altitudes of
hundreds of meters show that earlier uplifts preceded the
most recent events with which we deal here. The similar
geographic distribution of net Quaternary deformation and
late Quaternary deformation indicates a close relationship
between the shorter-term and the longer-term vertical oscillations and structural evolution of the arc. Upward isostatic
displacement by the volume of the d’Entrecasteaux Ridge
and topographically prominent features of the Woodlark
Basin are certainly responsible for much of the longer-term
uplift. However, the total uplift correlates poorly to the area
of d’Entrecasteaux Ridge impingement. Western Espiritu
Santo Island represents the net effects of perhaps many
cycles of uplift and subsidence such as we have documented. There may be permanent modifications to the
forearc in addition to the upward volumetric displacement
of the d’Entrecasteaux Ridge system. Cloos [1992] proposed that subducting seamounts may be sheared off and
thus contribute to thickening of the upper plate and longterm uplift over millions of years.
[60] The CNH and NGG examples suggest that similar
rapid deformation of trenchward parts of forearcs may occur
at other locations where large bathymetric features impinge
and cause locking of the shallow end of the interplate thrust
zone that persists for hundreds to tens of thousands of years.
Deformation related to seamount subduction in Costa Rica
appears very similar to the cases in the New Hebrides and
Solomons arcs [Gardner et al., 1992, 2001; Fisher et al.,
1998]. Onset of uplift since 200 ka has resulted in
Holocene uplift rates up to 6 mm/yr in association with
subduction of seamounts on the Cocos Plate [Gardner et al.,
2001]. The rates of convergence, timing of uplift, and
geographic distribution of uplift appear comparable to the
CNH and NGG settings.
10. Conclusions
[61] The late Quaternary cycles of uplift and subsidence
at CNH and NGG occur over one or two hundred thousand
years while the subducting plate advances only 10– 20 km
relative to the forearc. The latest uplifts at NGG and CNH
have occurred while convergence has been only a few km.
These uplifts can be explained by our elastic model that
assumes ‘‘permanent’’ locking of the shallow end of the
interplate thrust zone. It is not possible to produce the uplift
and subsidence by a displacement mechanism that requires
the emplacement and removal of seamounts or ridges
beneath the arc fast enough to match the timing of the
vertical tectonic cycles. The rapid subsidence that we have
described represents a sudden loss of coupling across the
shallow end of the interplate thrust zone that allows exten-
20 of 23
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
sion and collapse of the uplifted forearc bulge. Catastrophic
failure of seamounts as described by Cloos [1992] is one
mechanism that would enable this sudden decrease of
interplate coupling.
[62] When bathymetric features on subducting seafloor
impinge on forearcs there are significant tectonic consequences for seismicity and the interplate slip process and for
the evolution of arc morphology and structure regardless of
the age or thermal state of the subducting lithosphere. As
Scholz and Small [1997] concluded, we infer that locally
strong coupling might be possible at arcs such as Tonga and
the Marianas where large objects on the subducting plate
impinge on the edge of the forearc. Strong, shallow interplate coupling may result from seamount impingement and
lead to uplift of the outermost forearc. The mechanism by
which this is accomplished at CNH and NGG is simulated
by an elastic model in which interplate slip ceases for an
extended period of time on the shallow end of the interplate
thrust zone. If there are structural weaknesses in the back
arc, then the horizontal forces transmitted to the back arc
may activate those weak zones and cause back-arc thrusting
and deformation in addition to deformation in the zone of
forearc shortening above the main thrust zone. These
preexisting weaknesses in the back arc such as we find at
the CNH may represent previous intervals of strong compression of the forearc by impinging features or inherited
fault zones related to previous subduction zones such as at
the North Solomons Trench.
[63] Seafloor topographic features in the intermediate size
range between oceanic plateaus and insignificant bumps are
abundant on the part of the Australian plate approaching
Solomons and New Hebrides arcs. With objects such as the
d’Entrecasteaux Ridge, the Loyalty Ridge, Woodlark Rise,
Torres Plateau, and New Caledonia approaching the arc, we
can assume that the 500 km or so of Australian plate already
subducted beneath the New Hebrides must have carried
many such objects and that impingement of these objects
has done much to modify these arc systems. In the case of
seamounts and ridges, it is unlikely that arc polarity will
TC6005
reverse because retreat of the upper plate will tend to
decouple the two plates as the objects on the downgoing
plate descend along the interplate thrust zone. However,
more buoyant features such as New Caledonia may not
follow the preceding Australian oceanic lithosphere into the
mantle and arc polarity reversal may occur.
[64] The rapid reversals in vertical motions we describe
have only been recognized in three forearc localities: Central
New Hebrides, New Georgia Group, and Costa Rica. Failure
to recognize this process in other active forearc settings may
indicate the lack of deposition or preservation of dateable
geological marker horizons that can be directly related to
paleosea level. The occurrence of unrecrystallized corals that
have yielded high-precision radiometric ages has been
critical in unraveling this complex process in all three of
the above areas. We expect that additional cases of rapid
vertical deformation will be discovered and investigated as
new neotectonic techniques are applied in other locations.
[65] Acknowledgments. This paper is dedicated to Charles Pakoa,
formerly the chief surveyor of Espiritu Santo Island, a collaborator on many
of our projects, and a great friend, who passed away early in 2003. This
paper draws on research conducted in the New Hebrides and Solomons arcs
since the mid-1970s and is the result of the contributions and goodwill not
only of government officials from both nations but scientists and residents
of external origin. We are particularly grateful to the scores of people living
in towns and villages that made us welcome in their homes and on their
land throughout the Republic of Vanuatu and the Solomon Islands. A few
who contributed most directly to our success include Edwin Tae, Chief
Kalman of Malakula Island, Jean-Louis Laurent, Douglas Charley, David
Nakedau, Yvan Join, Jean-Claude Willy, Claude Reichenfeld, Michel
Lardy, the Wilson family and friends on Maewo Island, Chief Thomas
Liu and his village on northern Pentecost Island, and many others. In the
Solomon Islands, we received invaluable aid from Donn Tolia, Nicholas
Biliki, Watson Sakuru, Thomas Toba, Michael Rahe, Allison Papabatu,
Bobby Kelley, Whitlam Utu, Richard Bedford, and Lloyd Hodge. Narsili
Pule, Permanent Secretary of Parliament, western Solomon Islands, contributed valuable political and logistic support to the success of this and
other projects. Aspects of this research were supported by NSF grants EAR7919912, EAR-8507983, EAR-8721013, EAR-8721811, EAR-8904987,
ATM-8922114, EAR-9420278, OCE-9810724, and OCE-0202549 to Taylor; NSF grant EAR-9019985 to Mann; NSF grants EAR-9730699 and
EAR-0115488 to Burr; NSF grants EAR-9712037 and EAR-9809459 to
Edwards; and NSF grant EAR-8721811 to Bevis. UTIG contribution 1800.
References
Baker, P. E., M. Coltorti, L. Briqueu, T. Hasenaka,
E. Condliffe, and A. J. Crawford (1994), Petrology and composition of the volcanic basement of
Bougainville Guyot, Site 831, Proc. Ocean Drill.
Program Sci. Results, 134, 363 – 373.
Beavan, J., P. Tregoning, M. Bevis, T. Kato, and
C. Meertens (2002), Motion and rigidity of the
Pacific Plate and implications for plate boundary
deformation, J. Geophys. Res., 107(B10), 2261,
doi:10.1029/2001JB000282.
Bilek, S. L., S. Y. Schwartz, and H. R. DeShon (2003),
Control of seafloor roughness on earthquake rupture behavior, Geology, 31, 455 – 458.
Bloom, A. L., W. S. Broecker, J. Chappell, R. K.
Matthews, and K. J. Mesolella (1974), Quaternary
sea level fluctuations on a tectonic coast: New
230
Th/234U dates from the Huon Peninsula, New
Guinea, Quat. Res., 4, 185 – 205.
Cabioch, G., and L. K. Ayliffe (2001), Raised coral
terraces at Malakula, Vanuatu, Southwest Pacific,
indicate high sea level during marine isotope stage
3, Quat. Res., 56, 357 – 365.
Cabioch, G., B. A. Thomassin, and J. F. Lecolle (1989),
Age d’emersion des recifs frangeants holocenes
autour de la ‘‘Grande Terre’’ de Nouvelle-Caledonie
(SO Pacifique): Nouvelle interpretation de la courbe
des niveaux marins depuis 8000 ans B.P. (Age of
emersion of the Holocene fringing reefs around the
mainland of New Caledonia, Southwest Pacific:
New interpretation of the sea-level curve since
8,000 years B. P.), C. R. Acad. Sci., 308, 419 – 425.
Cabioch, G., F. W. Taylor, J. Recy, R. L. Edwards, S. C.
Gray, G. Faure, G. S. Burr, and T. Correge (1998),
Environmental and tectonic influence on growth
and internal structure of a fringing reef at Tasmaloum (SW Espiritu Santo, New Hebrides island arc,
SW Pacific), in Reefs and Carbonate Platforms in
the Pacific and Indian Oceans, edited by G. F.
Camoin and P. J. Davies, Spec. Publ. Int. Assoc.
Sedimentol., 26, 267 – 277.
Cabioch, G., K. A. Banks-Cutler, W. J. Beck, G. S.
Burr, T. Corrège, R. L. Edwards, and F. W. Taylor
(2003), Continuous reef growth during the last
23 kyr BP in a tectonically active zone (Vanuatu,
south west Pacific), Quat. Sci. Rev., 22, 1771 – 1786.
Calmant, S., P. Lebellegard, F. W. Taylor, M. Bevis,
D. Maillard, J. Recy, and J. Bonneau (1995),
Geodetic measurements of convergence across
21 of 23
the New Hebrides subduction zone, Geophys.
Res. Lett., 22, 2573 – 2576.
Calmant, S., B. Pelletier, R. Pillet, M. Regnier,
P. Lebellegard, D. Maillard, F. W. Taylor,
M. Bevis, and J. Recy (1997), Aseismic and coseismic motions in GPS series related to the Ms
7. 3 July 13, 1994, Malekula earthquake, central
New Hebrides subduction zone, Geophys. Res. Lett.,
24, 3077 – 3080.
Calmant, S., B. Pelletier, P. Lebellegard, M. Bevis,
F. W. Taylor, and D. A. Phillips (2003), New
insights on the tectonics along the New Hebrides subduction zone based on GPS results,
J. Geophys. Res., 108(B6), 2319, doi:10.1029/
2001JB000644.
Carney, J. N. (1986), Geology of Maewo, regional report, Vanuatu Dep. of Geol., Mines and Rural Water
Suppl., Port Vila.
Carney, J. N., and A. Macfarlane (1982), Geological
evidence bearing on the Miocene to Recent structural evolution of the New Hebrides arc, Tectonophysics, 87, 147 – 175.
Chappell, J. (1974), Geology of coral terraces, Huon
Peninsula, New Guinea: A study of Quaternary tec-
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
tonic movements and sea-level changes, Geol. Soc.
Am. Bull., 85, 553 – 570.
Chappell, J., and N. Shackleton (1986), Oxygen isotopes and sea level, Nature, 324, 137 – 140.
Chappell, J., A. Omura, T. Esat, M. McCulloch,
J. Pandolfi, Y. Ota, and B. Pillans (1996), Reconciliation of late Quaternary sea levels derived from
coral terraces at Huon Peninsula with deep sea
oxygen isotope records, Earth Planet. Sci. Lett.,
141, 227 – 236.
Chinn, D. S., and B. L. Isacks (1983), Accurate source
depths and focal mechanisms of shallow earthquakes in western South America and in the New
Hebrides island arc, Tectonics, 2, 529 – 563.
Cloos, M. (1992), Thrust-type subduction zone earthquakes and seamount asperities: A physical model
for seismic rupture, Geology, 20, 601 – 604.
Collot, J. Y., and M. A. Fisher (1989), Formation of
forearc basins by collision between seamounts and
accretionary wedges: An example from the New
Hebrides subduction zone, Geology, 17, 930 – 933.
Collot, J. Y., and M. A. Fisher (1992), The d’Entrecasteaux zone-New Hebrides Island arc collision zone:
An overview, Proc. Ocean Drill. Program Sci. Initial Rep., 134, 363 – 373.
Collot, J. Y., J. Daniel, and R. V. Burne (1985), Recent
tectonics associated with the subduction/collision of
the d’Entrecasteaux Zone in the central New Hebrides: Structures and processes in subduction zones,
Tectonophysics, 112, 325 – 356.
Collot, J. Y., S. Lallemand, B. Pelletier, J.-P. Eissen,
G. Glacon, M. A. Fisher, H. G. Greene, J. Boulin,
J. Daniel, and M. Monzier (1992), Geology of the
d’Entrecasteaux-New Hebrides arc collision zone:
Results from a deep submersible survey, Tectonophysics, 212, 213 – 241.
Cooper, P., and B. Taylor (1987), The spatial distribution of earthquakes, focal mechanisms, and subducted lithosphere in the Solomon Islands, in
Marine Geology, Geophysics and Geochemistry of
the Woodlark Basin-Solomon Islands, Earth Sci.
Ser., vol. 7, edited by B. Taylor, and N. F. Exon,
pp. 67 – 88, Circum-Pac. Counc. Energy and Mineral. Resour., Houston, Tex.
Crook, K. A. W., and B. Taylor (1994), Structure and
Quaternary tectonic history of the Woodlark triple
junction region, Solomon Islands, Mar. Geophys.
Res., 16, 65 – 89.
Cutler, K. B., R. L. Edwards, F. W. Taylor, H. Cheng,
J. Adkins, C. D. Gallup, P. M. Cutler, G. S. Burr,
and A. L. Bloom (2003), Rapid sea-level fall and
deep-ocean temperature change since the last interglacial period, Earth Planet. Science Lett., 206,
253 – 271.
DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein
(1990), Current plate motions, Geophys. J. Int.,
101, 425 – 478.
DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein
(1994), Effect of recent revisions to the geomagnetic reversal time scale on estimates of current
plate motions, Geophys. Res. Lett, 21, 2191 – 2194.
Dominguez, S., S. E. Lallemand, J. Malavieille, and
R. von Huene (1998), Upper plate deformation
associated with seamount subduction, Tectonophysics, 293, 207 – 224.
Dubois, J., J. Launay, J. Marshall, and J. Recy (1977),
New Hebrides Trench: Subduction rate from associated lithospheric bulge, Can. J. Earth Sci., 14,
250 – 255.
Dunkley, P. N. (1986), Geology of the New Georgia
Group, Solomon Islands, 83 pp., Br. Geol. Surv.
Overseas Dir., Keyworth, U.K.
Ebel, J. E. (1980), Source processes of the 1965 New
Hebrides Islands earthquakes inferred from teleseismic waveforms, Geophys. J. R. Astron. Soc., 63,
381 – 403.
Edwards, R. L., F. W. Taylor, and G. J. Wasserburg
(1988), Dating earthquakes with high-precision
thorium-230 ages of very young corals, Earth Planet. Sci. Lett., 90, 371 – 381.
Fisher, D. M., T. W. Gardner, J. S. Marshall, P. B. Sak,
and M. Protti (1998), Effect of subducting sea-floor
roughness on fore-arc kinematics, Pacific Coast,
Costa Rica, Geology, 26, 467 – 470.
Fisher, M. A. (1986), Tectonic processes at the collision of the d’Entrecasteaux zone and the New
Hebrides Island arc, J. Geophys. Res., 91,
10,470 – 10,486.
Fisher, M. A., J. Y. Collot, and G. L. Smith (1986),
Possible causes for structural variation where the
New Hebrides island arc and the D’Entrecasteaux
zone collide, Geology, 14, 951 – 954.
Fisher, M. A., J. Y. Collot, and E. L. Geist (1991a), The
collision zone between the north d’Entrecasteaux
ridge and the New Hebrides island arc: 2. Structure
from multichannel seismic data, J. Geophys. Res.,
96, 4479 – 4495.
Fisher, M. A., J. Y. Collot, and E. L. Geist (1991b),
Structure of the collision zone between Bougainville Guyot and the accretionary wedge of the
New Hebrides island arc, southwest Pacific, Tectonics, 10, 887 – 903.
Fleming, K., P. Johnston, D. Zwartz, Y. Yokoyama,
K. Lambeck, and J. Chappell (1998), Refining
the eustatic sea-level curve since the last glacial
maximum using far- and intermediate-field sites,
Earth Planet. Science Lett., 163, 327 – 342.
Gardner, T. W., D. Verdonck, N. M. Pinter, R. L.
Slingerland, K. P. Furlong, T. F. Bullard, and
S. G. Wells (1992), Quaternary uplift astride
the aseismic Cocos Ridge, Pacific Coast, Costa
Rica, Geol. Soc. Am. Bull., 104, 219 – 232.
Gardner, T., J. Marshall, D. Merritts, B. Bee,
R. Burgette, E. Burton, J. Cooke, N. Kehrwald,
M. Protti, D. Fisher, and P. B. Sak (2001), Holocene
forearc block rotation in response to seamount
subduction, southeastern Peninsula de Nicoya,
Costa Rica, Geology, 29, 151 – 154.
Geist, E. L., M. A. Fisher, and D. W. Scholl (1993),
Large-scale deformation associated with ridge subduction, Geophys. J. Int., 115, 344 – 366.
Gilpin, L. M. (1982), Tectonic geomorphology of Santo
Island, Vanuatu (New Hebrides), M.S. thesis, 147
pp., Cornell Univ., Ithaca, New York.
Greene, H. G., and J. Y. Collot (1994), Ridge-arc collision: Timing and deformation determined by Leg
134 drilling, central New Hebrides island arc, Proc.
Ocean Drill. Prog. Sci. Results, 134, 609 – 621.
Greene, H. G., A. Macfarlane, D. P. Johnson, and A. J.
Crawford (1988), Structure and tectonics of the central New Hebrides arc, in Geology and Offshore
Resources of Pacific Island Arcs-Vanuatu Region,
Earth Sci. Ser., vol. 8, edited by H. G. Greene and
F. L. Wong, pp. 377 – 412, Circum-Pac. Counc. Energy and Mineral. Resour., Houston, Tex.
Harmon, R. S., R. M. Mitterer, N. Kriausakul, L. S.
Land, H. P. Schwarcz, P. Garrett, G. J. Larson,
H. L. Vacher, and M. Rowe (1983), U-series and
amino-acid racemization geochronology of Bermuda:
Implications for eustatic sea-level fluctuation over
the past 250,000 years, Palaeogeogr. Palaeoclimatol. Palaeoecol., 44, 41 – 70.
Imbrie, J., J. D. Hays, D. G. Martinson, A. McIntyre,
A. Mix, J. J. Morley, N. G. Pisias, W. Prell, and
N. J. Shackleton (1984), The orbital theory of
Pleistocene climate: Support from a revised chronology of the marine 180 record, in Milankovitch
and climate: Understanding the Response to Astronomical Forcing, edited by A. Berger, J. Imbrie,
J. Hays, G. Kukla, and B. Saltzman, pp. 269 –
305, Springer, New York.
Isacks, B. L., R. K. Cardwell, J. L. Chatelain,
M. Barazangi, J. M. Marthelot, D. Chinn, and
R. Louat (1981), Seismicity and tectonics of the
central New Hebrides Island Arc, in Earthquake
Prediction: An International Review, Maurice Ewing Ser., vol. 4, edited by D. W. Simpson, and P. G.
Richards, pp. 93 – 116, AGU, Washington, D. C.
Jouannic, C., F. W. Taylor, A. L. Bloom, and M. Bernat
(1980), Late Quaternary uplift history from
emerged reef terraces on Santo and Malekula Islands, central New Hebrides Arc, Tech. Bull. 3,
pp. 91 – 108, UNDP/CCOP, Econ. and Soc. Comm.
for Asia and the Pacific, Suva, Fiji.
22 of 23
TC6005
Jouannic, C., F. W. Taylor, and A. L. Bloom (1982), Sur
la surrection et la deformation d’un arc jeune: L’arc
des Nouvelles Hebrides, Trav. Doc. ORSTOM, 147,
223 – 246.
Katz, H. R. (1988), Offshore Geology of Vanuatu: Previous work, in Geology and Offshore Resources of
Pacific Island Arcs: Vanuatu Region, Earth Sci.
Ser.,vol. 8, edited by H. G. Greene and F. L. Wong,
pp. 93 – 122, Circum-Pac. Counc.Energy and Min.
Resour., Houston, Tex.
Kroenke, L. W. (1984), Cenozoic tectonic development
of the suthwest Pacific, CCOP/SOPAC Tech. Bull.
6, 122 pp.
Lagabrielle, Y., B. Pelletier, G. Cabioch, M. Régnier,
and S. Calmant (2003), Coseismic and long-term
vertical displacement due to back arc shortening,
central Vanuatu: Offshore and onshore data following the Mw 7.5, 26 November 1999 Ambrym earthquake, J. Geophys. Res., 108(B11), 2519,
doi:10.1029/2002JB002083.
Macfarlane, A., J. N. Carney, A. J. Crawford, and G. H.
Greene (1988), Vanuatu-A review of the onshore
geology, in Geology and Offshore Resources of Pacific Island Arcs-Vanuatu Region, Earth Sci. Ser.,
vol. 8, edited by H. G. Greene and F. L. Wong, pp.
45 – 91, Circum-Pac. Council Energy and Min. Resour., Houston, Tex.
Maillet, P., M. Monzier, M. Selo, and D. Storzer (1983),
The D’Entrecasteaux Zone (Southwest Pacific): A
petrological and geochronological reappraisal, Mar.
Geol., 53, 179 – 197.
Mallick, D. I. J., and D. Greenbaum (1977), Geology of
southern Santo, New Hebrides Geol. Surv. Reg.
Rep., 84 pp., Br. Serv., Port Vila, Vanuatu.
Mallick, D. I. J., and G. Neef (1974), Geology of Pentecost, New Hebrides Geol. Surv. Reg. Rep., 103
pp., Br. Serv., Port Vila, Vanuatu.
Mann, P., F. W. Taylor, M. B. Lagoe, and A. Quarles
(1998), Accelerating late Quaternary uplift of the
New Georgia Island Group (Solomon island arc)
in response to subduction of the recently active
Woodlark spreading center and Coleman Seamount,
Tectonophysics, 295, 259 – 306.
Martinson, D. G., N. G. Pisias, J. D. Hays, J. Imbrie,
T. C. Moore Jr., and N. J. Shackleton (1987), Age
dating and the orbital theory of the ice ages:
Development of a high-resolution 0 to 300,000 –
year chronostratigraphy, Quat. Res., 27, 1 – 29.
Meffre, S., and A. J. Crawford (2001), Collision tectonics in the New Hebrides arc (Vanuatu), Island
Arc, 10, 33 – 50.
Mitchell, A. H. G. (1966), Geology of south Malekula,
Rep. No. 3, 42 pp., Geol. Surv. New Hebrides Condominium, Port Vila, Vanuatu.
Mitchell, A. H. G. (1971), Geology of northern Malekula, New Hebrides Geol. Surv. Regional Report,
56 pp., Geol. Surv. New Hebrides Condominium,
Port Vila, Vanuatu.
Miyata, T., Y. Maeda, E. Matsumoto, Y. Matsushima,
P. Rodda, A. Sugimura, and H. Kayanne (1990),
Evidence for a Holocene high sea-level stand, Vanua
Levu, Fiji, Quat. Res., 33, 352 – 359.
Monzier, M., J.-Y. Collot, and J. Daniel (1984), Carte
bathymetrique des parties centrale et meridionale de
l’arc insulaire des Nouvelles-Hebrides, map, ORSTOM, Bondy, France.
Neef, G., and H. H. Veeh (1977), Uranium series ages
and late Quaternary uplift in the New Hebrides,
Nature, 269, 682 – 683.
Oleskevich, D. A., R. D. Hyndman, and K. Wang
(1999), The updip and downdip limits to great subduction earthquakes: Thermal and structural models
of Cascadia, South Alaska, SW Japan, and Chile,
J. Geophys. Res., 104, 14,965 – 14,991.
Pacheco, J., L. Sykes, and C. Scholz (1993), Nature of
seismic coupling along simple plate boundaries of
the subduction type, J. Geophys. Res., 98, 14,133 –
14,159.
Pascal, G., B. L. Isacks, M. Barazangi, and J. Dubois
(1978), Precise relocations of earthquakes and seismotectonics of the New Hebrides island arc, J. Geophys. Res., 83, 4957 – 4973.
TC6005
TAYLOR ET AL.: RAPID FOREARC VERTICAL MOTION
Pelletier, B., S. Calmant, and R. Pillet (1998), Current
tectonics of the Tonga-New Hebrides region, Earth
Planet. Sci. Lett., 164, 263 – 276.
Pelletier, B., et al. (2000), Le Seisme d’Ambrym-Pentecote (Vanuatu) du 26 novembre 1999 (Mw 7,5):
Donnees preliminaires sur la seismicite, le tsunami
et les deplacements associes (The Ambrym-Pentecost (Vanuatu) earthquake on November 26, 1999
(magnitude 7. 5), C. R. Acad. Sci., 331, 21 – 28.
Peltier, W. R. (1998), Postglacial variations in the level
of the sea: Implications for climate dynamics and
solid-Earth geophysics, Rev. Geophys., 36, 603 –
689.
Phinney, E. J. (1997), Sequence stratigraphy, structure,
and tectonics of the southern Ontong Java Plateau
and Malaita accretionary prism, Solomon Islands,
M.A. thesis, 128 pp., Univ. of Tex. at Austin,
Austin.
Phinney, E. J., W. P. Mann, M. F. Coffin, and T. H.
Shipley (1999), Sequence stratigraphy, structure,
and tectonic history of the southwestern Ontong
Java Plateau adjacent to the North Solomon Trench
and Solomon islands arc, J Geophys. Res., 104,
20,449 – 20,466.
Regnier, M., S. Calmant, B. Pelletier, Y. Lagabrielle,
and G. Cabioch (2003), The Mw 7.5 1999 Ambrym
earthquake, Vanuatu: A back arc intraplate thrust
event, Tectonics, 22(4), 1034, doi:10.1029/
2002TC001422.
Savage, J. C. (1983), A dislocation model of strain
accumulation and release at a subduction zone,
J. Geophys. Res., 88, 4984 – 4996.
Schmidt, A., G. S. Burr, F. W. Taylor, J. O’Malley, and
J. W. Beck (2004), A semiannual radiocarbon
record of a modern coral from the Solomon Islands,
Nucl. Instrum. Methods Phys. Res., Sect. B, 223 –
224, 420 – 427.
Scholz, C. H., and C. Small (1997), The effect of seamount subduction on seismic coupling, Geology,
25, 487 – 490.
Shinohara, M., K. Suyehiro, and T. Murayama (2003),
Microearthquake seismicity in relation to double
convergence around the Solomon Islands arc by
ocean-bottom seismometer observation, Geophys.
J. Int., 153, 691 – 698.
Singh, S. J., and S. Rani (1993), Crustal deformation
associated with two-dimensional thrust faulting,
J. Phys. Earth, 41, 87 – 101.
Stoddard, D. R. (1969), Geomorphology of Solomon
Islands coral reefs, Philos. Trans. R. Soc. London,
Ser. B, 255, 383 – 402.
Stuiver, M., P. J. Reimer, E. Bard, J. W. Beck, G. S.
Burr, K. A. Hughen, B. Kromer, G. McCormac,
J. van der Plicht, and M. Spurk (1998),
IN TC AL9 8 r ad i oca rb on age ca l i bra t i on,
24,000— cal BP, Radiocarbon, 40, 1041 – 1083.
Taylor, B., and N. F. Exon (1987), An investigation of
ridge subduction in the Woodlark-Solomons region:
Introduction and overview, in Marine Geology,
Geophysics and Geochemistry of the Woodlark Basin-Solomon Islands, Earth Sci. Ser., vol. 7, edited
by B. Taylor and N. F. Exon, pp. 1 – 24, CircumPac. Council Energy and Min. Resour., Houston,
Tex.
Taylor, B., A. Goodliffe, F. Martinez, and R. Hey
(1995), Continental rifting and initial sea-floor
spreading in the Woodlark basin, Nature, 374,
534 – 537.
Taylor, F. W. (1992), Quaternary vertical tectonics of
the central New Hebrides island arc, Proc. Ocean
Drill. Program Initial Rep., 134, 33 – 42.
Taylor, F. W., and A. L. Bloom (1977), Coral reefs on
tectonic blocks, Tonga island arc, paper presented at
International Coral Reef Symposium, Rosenstiel
Sch. of Mar. and Atmos. Sci., Univ. of Miami,
Miami, Fla.
Taylor, F. W., B. L. Isacks, C. Jouannic, A. L. Bloom,
and J. Dubois (1980), Coseismic and Quaternary
vertical tectonic movements, Santo and Malekula
islands, New Hebrides island arc, J. Geophys.
Res., 85, 5367 – 5381.
Taylor, F. W., C. Jouannic, and A. L. Bloom (1985),
Quaternary uplift of the Torres Islands, northern
New Hebrides frontal arc: Comparison with Santo
and Malekula islands, central New Hebrides frontal
arc, J. Geol., 93, 419 – 438.
Taylor, F. W., C. Frohlich, J. Lecolle, and M. R.
Strecker (1987), Analysis of partially emerged
corals and reef terraces in the central Vanuatu
Arc: Comparison of contemporary coseismic
and nonseismic with Quaternary vertical movements, J. Geophys. Res., 92, 4905 – 4933.
Taylor, F. W., R. L. Edwards, G. J. Wasserburg, and
C. A. Frohlich (1990), Seismic recurrence intervals
and timing of aseismic subduction inferred from
emerged corals and reefs of the central Vanuatu
(New Hebrides) frontal arc, J. Geophys. Res., 95,
393 – 408.
Taylor, F. W., T. M. Quinn, C. D. Gallup, and R. L.
Edwards (1994), Quaternary plate convergence
rates at the New Hebrides island arc from the chronostratigraphy of Bougainville Guyot (Site 831),
Proc. Ocean Drill. Prog. Sci. Results, 134, 47 – 57.
Taylor, F. W., M. G. Bevis, B. E. Schutz, D. Kuang,
J. Recy, S. Calmant, D. Charley, M. Regnier,
B. Perin, M. Jackson, and C. Reichenfeld (1995),
Geodetic measurements of convergence at the
New Hebrides island arc indicate arc fragmenta-
23 of 23
TC6005
tion due to an impinging aseismic ridge, Geology,
23, 1011 – 1014.
Taylor, F. W., C. A. Frohlich, W. P. Mann, G. S. Burr,
W. Beck, R. L. Edwards, and D. A. Phillips (2003),
Episodic forearc uplift related to subduction of the
Woodlark Basin, western Solomons Arc: Intermittent aseismic slip or multi-century earthquake recurrence intervals?, Eos Trans. AGU, 84(46), Fall
Meet. Suppl. Abstract T42C-04.
Urmos, J. (1985), Oxygen isotopes, sea levels, and uplift of reef terraces, Araki Island, Vanuatu, M.S.
thesis, 123 pp., Cornell Univ., Ithaca, N. Y.
Veeh, H. H. (1966), Th230/U238 and U234/U238 ages of
Pleistocene high sea level stand, J. Geophys. Res.,
71, 3379 – 3386.
Weissel, J. K., B. Taylor, and G. D. Karner (1982), The
opening of the Woodlark Basin, subduction of the
Woodlark spreading system, and the evolution of
northern Melanesia since mid-Pliocene time, Tectonophysics, 87, 253 – 277.
J. W. Beck and G. S. Burr, NSF Accelerator Facility,
Physics Department, University of Arizona, 1118 E.
Fourth Street, Tucson, AZ 85721, USA. (wbeck@
physics.arizona.edu; [email protected])
M. G. Bevis, Department of Civil and Environmental Engineering and Geodetic Science, Ohio State
University, 470 Hitchcock Hall, 2070 Neil Avenue,
Columbus, OH 43210, USA. ([email protected])
G. Cabioch and J. Recy, Institut de Recherche pour
le Développement, BP A5, Nouméa cedex, Vanuatu.
([email protected])
H. Cheng and R. L. Edwards, Minnesota Isotope
Laboratory, Department of Geology and Geophysics,
University of Minnesota, 310 Pillsbury Drive, SE,
Minneapolis, MN 55455-0219, USA. (cheng021@umn.
edu; [email protected])
K. B. Cutler, U.S. Department of State, Office of
Senior Coordinator for Nuclear Safety, 2001 C Street
NW, Washington, DC 20523, USA.
S. C. Gray, Department of Marine and Environmental Studies, University of San Diego, 5998 Alcalá
Park, San Diego, CA 92110-2492, USA. (sgray@
sandiego.edu)
P. Mann and F. W. Taylor, Institute for Geophysics,
Jackson School of Geosciences, University of Texas at
Austin, 4412 Spicewood Springs Road, Austin, TX
78759-8500, USA. ([email protected]; fred@ig.
utexas.edu)
D. A. Phillips, UNAVCO, 6350 Nautilus Drive,
Boulder, CO 80301, USA. ([email protected])