Downloaded from geology.gsapubs.org on March 15, 2012
Geology
Prehistoric earthquakes on the Caribbean−South American plate boundary,
Central Range fault, Trinidad
Carol S. Prentice, John C. Weber, Christopher J. Crosby and Daniel Ragona
Geology 2010;38;675-678
doi: 10.1130/G30927.1
Email alerting services
click www.gsapubs.org/cgi/alerts to receive free e-mail alerts when new
articles cite this article
Subscribe
click www.gsapubs.org/subscriptions/ to subscribe to Geology
Permission request
click http://www.geosociety.org/pubs/copyrt.htm#gsa to contact GSA
Copyright not claimed on content prepared wholly by U.S. government employees within scope of
their employment. Individual scientists are hereby granted permission, without fees or further
requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequent
works and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms
to further education and science. This file may not be posted to any Web site, but authors may post
the abstracts only of their articles on their own or their organization's Web site providing the posting
includes a reference to the article's full citation. GSA provides this and other forums for the
presentation of diverse opinions and positions by scientists worldwide, regardless of their race,
citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflect
official positions of the Society.
Notes
© 2010 Geological Society of America
Downloaded from geology.gsapubs.org on March 15, 2012
Prehistoric earthquakes on the Caribbean–South American plate
boundary, Central Range fault, Trinidad
Carol S. Prentice1, John C. Weber2, Christopher J. Crosby1*, and Daniel Ragona3†
1
U.S. Geological Survey, 345 Middlefield Road, MS 977, Menlo Park, California 94025, USA
Department of Geology, Grand Valley State University, Allendale, Michigan 49401, USA
3
Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography, University of California–San Diego, La Jolla,
California 92121, USA
2
ABSTRACT
Recent geodetic studies suggest that the Central Range fault is
the principal plate-boundary structure accommodating strike-slip
motion between the Caribbean and South American plates. Our study
shows that the fault forms a topographically prominent lineament in
central Trinidad. Results from a paleoseismic investigation at a site
where Holocene sediments have been deposited across the Central
Range fault indicate that it ruptured the ground surface most recently
between 2710 and 550 yr B.P. If the geodetic slip rate of 9–15 mm/yr
is representative of Holocene slip rates, our paleoseismic data suggest
that at least 4.9 m of potential slip may have accumulated on the fault
and could be released during a future large earthquake (M > 7).
INTRODUCTION
Trinidad is located within the actively deforming Caribbean–South
American plate-boundary zone (Fig. 1A). Global positioning system
(GPS)–based geodetic studies show that the current motion of the Caribbean plate is 20 ± 3 mm/yr directed N86° ± 2°E with respect to South
America (Weber et al., 2001; Perez et al., 2001). West of Trinidad, in
Venezuela, the principal plate-boundary fault is the El Pilar fault, a rightlateral strike-slip fault that has produced multiple moderate to large earthquakes in the last several hundred years (Mendoza, 2000). However,
recent GPS-based geodetic studies in Trinidad indicate that the eastward
projection of the El Pilar fault in northern Trinidad is currently accommodating no right-lateral motion within the error of the measurements
(Saleh et al., 2004; Weber et al., 2010) (Fig. 1B) and that a significant part
of the plate-boundary motion is accommodated on the Central Range fault
(CRF) (Saleh et al., 2004; Weber et al., 2010) (Fig. 2).
Trinidad’s geology is characterized by a 10-km-thick passive margin
and foreland basin sedimentary section (metamorphosed and exhumed in
the north) with strike-slip faults that cut older fold-and-thrust belt structures (e.g., Pindell et al., 1998) (Fig. 2C). Bedrock faults in the vicinity
of the CRF were initially mapped by Kugler (1959) but not recognized
as being active structures. Weber et al. (1999) proposed an active fault
B
NA
20˚N
SF
PRT
EPGF H
A
A
N
SCALE
1km 3 km 5 km
Linear ridge
LA
Caribbean plate
15˚
Gulf of Paria
pull-apart
MB
Port of Spain
Central
Range
fault
Samlalsingh
Samla
Sam
lalsin
la
sin
i gh
hs
sit
site
it
ite
Right-laterally
offset drainages
Saddle
Trinidad
10˚
SA
70˚W
EPF WSF
CRF
Venezuela
65˚
Linear drainage
Linear ridge
San Fernando
Scarp
Pointe-áPierre
Saddle
GE
RAN
LT
FAU
Navet
avett D
av
Dam
Vegetation lineament
L
TRA
CEN Linear drainage
B
Saddle
Shutter ridge
Saddle
Linear drainages
10 mm/yr
C
Figure 1. A: Tectonic setting of Trinidad. Caribbean plate moves
~20 mm/yr eastward relative to South America (SA). CRF—Central
Range fault; EPF—El Pilar fault; EPGF—Enriquillo–Plantain Garden
fault; H—Hispaniola; LA—Lesser Antilles; MB— Maricaibo Block;
NA—North America; PRT—Puerto Rico Trench; SF—Septentrional
fault; WSF—Warm Springs fault (after Babb and Mann, 1999). Cooler
colors indicate lower topography and bathymetry. B: Map of Trinidad showing CRF (solid red line). Heavy arrows with error ellipses
show GPS-to-GPS motions (Weber et al., 2001). Thin arrows show
triangulation-to-GPS motions. Both show that the CRF is associated
with a strong gradient in the velocity field (Saleh et al., 2004; Weber et
al., 2010). Black rectangle shows area of Figure 2A. Dashed red line
shows location of the eastward projection of the El Pilar fault (EPF),
inferred as a buried structure by Kugler (1959), which takes up no horizontal plate-boundary motion within the error of GPS measurements.
*Current address: San Diego Supercomputer Center, University of California–
San Diego, MC 0505, 9500 Gilman Drive, La Jolla, California 92093-0505, USA
†
Current address: BP America, 501 Westlake Park Blvd., Houston, Texas
77079, USA
NCF
Oceanic crust
Caribbean
lithosphere
N
EMB
AF EP?
Fold-and-thrust belt cut
by strike-slip faults
CRF Continental crust
So. Am. lithosphere
100 km
S
Figure 2. A: Hillshade image derived from digital elevation model for
western Trinidad showing geomorphic features indicative of Quaternary activity. Red line is Central Range fault (CRF), dashed where
uncertain. B: Oblique aerial photograph showing linear valley associated with CRF. View is eastward from point above view icon shown
in A. White arrows show location of CRF. C: Lithospheric-scale schematic cross section through Caribbean–South American plate-boundary zone showing active CRF and thick sedimentary section coupled
to deeper shear (modified from Teyssier et al., 2002). AF—Arima fault,
a normal fault in bedrock (Kugler, 1959); CRF—Central Range fault;
EMB—exhumed metamorphic belt; EP?—eastward projection of El
Pilar fault, inferred by Kugler (1959); NCF—North Coast fault.
© 2010 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].
GEOLOGY,
August
2010
Geology,
August
2010;
v. 38; no. 8; p. 675–678; doi: 10.1130/G30927.1; 4 figures; Data Repository item 2010189.
675
Downloaded from geology.gsapubs.org on March 15, 2012
through central Trinidad carrying a significant part of the current plate
motion, which they referred to as the CRF. Babb and Mann (1999) interpreted that Neogene motion accommodated on the El Pilar fault in Venezuela steps to the south across the Gulf of Paria to the offshore Warm Springs
fault, creating a Neogene basin in the Gulf of Paria, and continues eastward
through central Trinidad. The CRF continues offshore east of the island as
a Quaternary structure (Soto et al., 2007) (Fig. 2). Russo et al. (1993) and
Teyssier et al. (2002) discuss how active strike-slip faults could couple the
thick sedimentary section to deeper plate-boundary shear (Fig. 2C). The
results of our study show that surface rupture has occurred on the CRF in
the late Holocene, most recently between 2710 and 550 yr B.P.
FAULT GEOMORPHOLOGY
We conducted field reconnaissance along the 25-km-long section
of the CRF between Navet Dam and Pointe-a-Pierre in western Trinidad
(Fig. 2), as well as aerial reconnaissance along the entire 50 km length
of the fault onshore (Crosby et al., 2009). Analysis of 1:20,000-scale stereo aerial photographs showed multiple geomorphic features commonly
associated with Quaternary strike-slip faults (Fig. 2) (Crosby et al., 2009).
Between Navet Dam and Manzanilla Bay on the east coast, dense vegetation and limited access prevented detailed study of this section of the fault.
However, geomorphic features such as scarps and linear drainages near
Manzanilla Bay strongly suggest that the fault continues northeast from
Navet Dam across the entire island. Analysis of offshore three-dimensional seismic data shows that the CRF continues for at least 60 km east of
Manzanilla Bay (Soto et al., 2007).
Significant anthropogenic modification of the landscape coupled
with the tropical climate in Trinidad creates a challenging environment for
the study of Holocene faulting. In many instances, geomorphic features
we identified on 1982 aerial photography have been modified by development and agriculture and are no longer recognizable in the field. Rapid
erosion and deposition due to the tropical climate and deforestation also
mask small-scale geomorphic features typically associated with Holocene
surface faulting. Thus our mapping of geomorphic features, while suggestive, is not sufficient to conclude that the CRF is a Holocene fault.
Therefore, we excavated trenches across the fault to determine the age of
the most recent surface rupture on the CRF in western Trinidad.
EXCAVATIONS ACROSS THE CENTRAL RANGE FAULT
We excavated ten trenches at the Samlalsingh site near the town of
Bonne Aventure (Figs. 2 and 3). This site is an agricultural field where
a small south-flowing stream has deposited Holocene fluvial sediments
B
B
Shutter
ridge
Sh
hut
utte
ttteer rri
idg
ge
3
Terrace 2
Trench 1
Terrace 1
am
Stre
CRF
C
CR
RF 2
1
7
10
0 4 6
9 5
Trench 6
Trench 4
A
Trench 9A
Trench 3 Trench 9
Trench 5
Trench 10
S
Trench 2
Trench 7
Unmapped
20 m
CRF
Explanation
Trench, fault
Trench, no fault
Modern channel
Terrace riser
Lower terrace 1
Higher terrace 2
Figure 3. A: Map of Samlalsingh site from tape and compass survey;
note that south is up, to match aerial view in B. Stream flows southeast across study site. B: Oblique aerial view southward across
Samlalsingh site. Area of photo is nearly coincident with map extent.
676
across the fault (Fig. 3). No small scarps or other expression of the active
CRF are present at the site, indicating that the most recent surface ruptures
have been buried by fluvial deposits or obscured by agricultural activities.
We excavated three trenches exposing the sediments across most of the
valley (Fig. 3, trenches 1–3; complete logs of all trenches at this site are
available in Crosby et al., 2009). Most of the trenches exposed deeply
weathered and highly deformed bedrock overlain by 0.5–3 m of fluvial
channel gravel and silty overbank deposits. We located the CRF in trench
3 and excavated additional trenches across the fault southwest of trench
3 to find exposures that would allow us to place constraints on the age of
the most recent surface ruptures. We exposed the fault in five of our ten
trenches (Fig. 3). Four of the other trenches did not cross the fault; the
remaining trench (trench 5) was excavated across the fault, but historical
sediments buried the fault in this location (Crosby et al., 2009).
Radiocarbon analysis of detrital charcoal samples collected from the
excavations shows that the fluvial gravel and overbank deposits are late
Holocene. Five radiocarbon samples (6 through 10, Table DR1 in the GSA
Data Repository1) collected from unit 30 exposed in trenches 9, 10, and
3 give indistinguishable, dendrochronologically calibrated, 2σ calendar
age ranges of 4830–5290 yr B.P., 4880–5290 yr B.P., 4970–5300 yr B.P.,
4970–5300, and 4880–5320 yr B.P. This unit is faulted in trenches 9, 10,
and 3 (Figs. 3 and 4) (Crosby et al., 2009), showing that the CRF has been
active after the mid-Holocene.
Relations among faults and sedimentary units in trench 9 allow us to
narrow the age of the most recent fault displacement. Trench 9 exposed a
strath surface cut into weathered bedrock (units 50 and 60) and overlain
by 1.5–2.0 m of fluvial gravel (unit 30) and silty overbank deposits (unit
20) with a well-developed soil A horizon (unit 10) (Fig. 4). Three fault
strands, F1, F2, and F3, cut the units in the eastern wall. F1 juxtaposes
bedrock units 50 and 60, but does not offset unit 30. F2 and F3 displace
unit 30 and the lower part of unit 20, and do not cut into the upper part of
unit 20 or unit 10. F2 and F3 form a small graben (Figs. 4A and 4B), most
likely the result of a local extensional step or bend in the CRF. The graben
has been filled in with upper unit 20 and unit 10 deposits. Both of these
units thicken substantially across the graben, suggesting their deposition
was affected by the graben. We see no evidence of deformation of either
unit 10 or upper unit 20, and interpret the deepening of the unit 10-20
contact over the fault zone as being due to thickening of the units within
the graben and not as a deformational feature. Gravel similar to unit 30 is
exposed in the graben at the base of the trench, and we interpret this gravel
as unit 30 offset across F2 and F3. This correlation provides a measurement of vertical separation across these faults of ~1.7–1.8 m. The change
in thickness and grain size of unit 30 across the faults is due to strike-slip
motion juxtaposing different facies of this unit.
We interpret the relations to indicate that the most recent surfacerupturing earthquake at this location occurred after deposition of unit 30
and during the deposition of unit 20, but prior to development of unit 10.
Unit 20 is composed of massive, very fine-grained silty clay that we interpret to have accumulated slowly during multiple flood events. The fault
displacement that created the graben had a significant vertical component
of displacement, creating free faces at the ground surface at the time of the
earthquake (Nelson, 1992). With time, the free faces eroded back, becoming less planar, and sediments of upper unit 20 were deposited against
them, eventually burying the fault scarps that bound the graben. The contact between the eroded free face and the postearthquake sediments is a
depositional contact, even though the units on the upthrown side of the
fault have been faulted. Both boundaries of the lower part of the graben
1
GSA Data Repository item 2010189, Table DR1 (calculated dates from
C analysis of angular charcoal fragments, Central Range Fault, Trinidad), is
available online at www.geosociety.org/pubs/ft2010.htm, or on request from
[email protected] or Documents Secretary, GSA, P.O. Box 9140, Boulder,
CO 80301, USA.
14
GEOLOGY, August 2010
Downloaded from geology.gsapubs.org on March 15, 2012
A
N18W
1: 520–650 yr B.P.
TRENCH 9: EAST WALL
0m
10
0.5
20
20
1.0
30
30
1.5
Bench
B
20
Unit 10
2.0
Unit 20
60
Unit 30
60
Unit 30
2.5
Eroded free face
30
Gra
b
Un en
it 2
0
F310
50
F3
Eroded free face
2: 550–670 yr B.P.
F2
9Unit 60
F2
5: 4660–5310 yr B.P.
4: 2360–2740 yr B.P.
7
6
8
F1
8: 4970–5300 yr B.P. (99%)
5
10: 4880–5320 yr B.P.
3
2
4
1
1 METER
N18W
SAMLALSINGH TRENCH 9 EXPLANATION
UNIT #
DESCRIPTION
10 Massive dark gray silty clay—A horizon
20 Massive sandy, silty clay
30 Pebbles, cobbles, and boulders in sand matrix
50 Silver-gray claystone. Bedrock.
60 Gray claystone. Brasso Shale. Bedrock.
0
0m
0.5
10
20
1.0
30
30
20
Bench
1.5
Bedrock shear fabric
Fault
20
Charcoal sample
2.0
60
50
Buried free face (eroded fault scarp)
60
2.5
Contact, dashed where gradational
F2
Approximate location of event horizon
F1
3
4
5
6
F4
F5 F3
TRENCH 9: WEST WALL
7
8
9
10
1 METER
11
Figure 4. Trench 9. A: Logs of trench 9, showing faulted Holocene units and samples 1, 2, 4, 5, 8, and 10 (Table DR1 in the GSA Data Repository1). Ages are 2σ calibrated ranges. Fault F1 offsets Miocene units only; faults F2 and F3 offset unit 30 and lower unit 20. Faults F4 and F5
may record an earlier event. B: Photograph showing graben in eastern wall. Grid above bench is 1 m × 0.5 m.
are steep and planar, and exhibit a shear fabric, while in the upper part of
the graben the boundaries are eroded and laid back (Fig. 4). We therefore
interpret the contacts between units 20 and 60 in the upper part of the graben as depositional contacts, and infer that the sediments within the upper
graben, including samples 1 and 2, postdate the earthquake. The contacts
between units 20 and 60 in the lower part of the graben are fault contacts,
and lower unit 20 (including samples 3, 4, and 5) predates the earthquake
(Table DR1; Fig. 4A).
Three detrital charcoal samples collected from lower unit 20 exposed
in trenches 9 (Fig. 4) and 9A (Crosby et al., 2009) were dated. One of
these (sample 5, Table DR1; Fig. 4A) yielded an age of 4660–5310 yr
B.P. and is most likely reworked from the gravel unit. The remaining two
samples (3 and 4, Table DR1) yielded indistinguishable, dendrochronologically calibrated, 2σ age ranges of 2350–2710 yr B.P. (sample 3; see
Crosby et al., 2009, for location in trench 9A) and 2360–2740 yr B.P.
(sample 4; see Fig. 4A for location). Samples 1 and 2 (Table DR1; Fig. 4A)
were collected from unit 10, the unfaulted soil unit exposed in trench 9.
These samples yield indistinguishable, dendrochronologically calibrated,
2σ age ranges of 520–650 and 550–670 yr B.P. Detrital charcoal forms
an unknown amount of time prior to the deposition of the sediment into
which it is incorporated, and in some cases (e.g., sample 5) can be significantly older than the time of deposition. However, we note that the radiocarbon ages for units 30, 20, and 10 are in correct stratigraphic order and
GEOLOGY, August 2010
that multiple samples from the same unit are indistinguishable (with the
one exception of sample 5). Because the tropical environment makes survival of charcoal through multiple phases of reworking relatively rare, and
because there is little indication in the age results that reworked charcoal
is a significant problem at this site, we assume that the ages of the detrital
charcoal are close to the age of deposition of the sedimentary units.
The two radiocarbon ages (3 and 4 on Table DR1) from unit 20 predate the most recent surface rupture of the CRF, and show that the most
recent event occurred after 2710 yr B.P. The minimum limiting age for
the most recent earthquake is given by samples 1 and 2 (Table DR1) collected from unit 10 in trench 9 (Fig. 4A). The ages of these samples are
520–650 yr B.P. for sample 1, and 550–670 yr B.P. for sample 2. Because
these sediments were deposited sometime after the most recent event, this
event must have occurred prior to 550 yr B.P., assuming that the ages of
the charcoal samples are close to the time of deposition of unit 10.
There is evidence that is suggestive of at least one earlier Holocene
event exposed in the western wall of trench 9 (Fig. 4A). Two additional
faults appear on this wall near the bottom of the graben, F4 and F5. Both
faults break unit 30 within the graben, but do not break unit 20, suggesting
they may have ruptured during an earlier episode of motion than faults
F2 and F3, prior to the deposition of unit 20. The evidence for this earlier earthquake is weak, and only appears on one of the two trench walls,
implying that F4 and F5 merge with F2 and F3 between the two trench
677
Downloaded from geology.gsapubs.org on March 15, 2012
walls. Therefore, we note that while these relations suggest the possibility
of two events since gravel unit 30 was deposited ca. 5000 yr B.P., additional evidence is needed to test this scenario. Because samples 3 and 4
collected from lower unit 20 were deposited after the earlier event in this
scenario, this earthquake, if it occurred, would have occurred prior to
2360 yr B.P. (Table DR1).
DISCUSSION AND CONCLUSIONS
Our paleoseismic investigations demonstrate that the CRF is a
Holocene fault that has produced at least one earthquake large enough
to rupture the ground surface within the past 2710 yr. We conclude that
this fault is capable of producing similar earthquakes again, and therefore
constitutes a significant seismic hazard for Trinidad. Our data suggest that
the most recent earthquake is prehistoric and occurred between 550 and
2710 yr B.P. This is consistent with the historical record, which does not
show a significant earthquake felt in central Trinidad that is likely to have
originated on the CRF since European settlement in the sixteenth century
(Robson, 1964). No Holocene geologic slip rate data are available for the
onshore CRF. However, if the geodetic rate of 9–15 mm/yr (Weber et al.,
2010) is typical of the last several thousand years, then a lapsed time of
>550 yr suggests that strain energy equivalent to over 4.9 m of slip is currently available for seismic release, corresponding to an earthquake of M
> 7 (Wells and Coppersmith, 1994).
A locked CRF that periodically ruptures to produce large earthquakes is consistent with our findings from a field search for aseismic
creep-related offset cultural features. We examined fault-crossing, linear
roads and fences in the Plaissance Park housing development near Pointeá-Pierre and saw no evidence for creep. We also found no evidence for
creep-related offsets across Navet Dam or nearby, nor across any of the
roads examined between Navet Dam and Pointe-á-Pierre. In addition,
the trenches at the Samlalsingh site showed no fault displacement of the
youngest sediments. However, we note that repeat measurements of carefully established alignment arrays are needed to be certain there is no fault
creep associated with the CRF.
The one and possibly two surface-rupturing prehistoric earthquakes
that we interpret from our trenches at the Samlalsingh excavation site represent a minimum number of earthquakes in the past 5000 yr. Although
we do not see evidence for additional events, it is possible that surface
ruptures occurred that were not recorded in the stratigraphic section. Excavations in a location with a more complete Holocene section might provide evidence for additional earthquakes. We note that the geodetically
determined slip rate of 12 ± 3 mm/yr is inconsistent with only one or two
large, surface-rupturing earthquakes: At this slip rate, the fault would slip
45–75 m in 5000 yr, requiring more than two earthquakes. Either these
additional earthquakes are not preserved in the stratigraphic sections
exposed in our excavations, or the geodetic slip rate is significantly higher
than the longer-term, late Holocene geologic slip rate. Despite uncertainties regarding the fault’s slip rate and recurrence interval, it is clear that the
CRF represents a significant seismic hazard to Trinidad and that the thick
sedimentary section must be sufficiently strong to store and release elastic
strain that results from plate-boundary shearing.
ACKNOWLEDGMENTS
We thank the Trinidad and Tobago Ministry of Energy and Energy Industries for financial support. We also thank Samlalsingh residents and landowners
Lindrum Subedar, Ferooz Boodram, Basdeo Ball, and Mahadeo Swoknanan for
access to their farms and for field assistance. We thank Alan Nelson, Paul Mann,
678
Maritia Tuttle, and James Pindell for helpful reviews. Thanks to Anthony Rodriguez and Martha Roldan for assistance with figures.
REFERENCES CITED
Babb, S., and Mann, P., 1999, Structural and sedimentary development of a
Neogene transpressional plate boundary between the Caribbean and South
American plates in Trinidad and the Gulf of Paria, in Mann, P., ed., Caribbean basins: Sedimentary basins of the world, volume 4: Amsterdam, Netherlands, Elsevier Science B.V., p. 495–557.
Crosby, C.J., Prentice, C.S., Weber, J.C., and Ragona, D., 2009, Logs of paleoseismic excavations across the Central Range fault, Trinidad: U.S. Geological Survey Open-File Report 2009-1228, Plates 1 and 2.
Kugler, H.G., 1959, Geological map and sections of Trinidad: Port of Spain, Petroleum Association of Trinidad, scale 1:100,000. 1 sheet.
Mendoza, C., 2000, Rupture history of the 1997 Cariaco, Venezuela, earthquake
from teleseismic P waves: Geophysical Research Letters, v. 27, no. 10,
p. 1555–1558, doi: 10.1029/1999GL011278.
Nelson, A.R., 1992, Lithofacies analysis of colluvial sediments; an aid in interpreting the Recent history of Quaternary normal faults in the Basin and
Range Province, Western United States: Journal of Sedimentary Research,
v. 62, no. 4, p. 607–621.
Perez, O., Bilham, R., Bendick, R., Velandia, J., Moncayo, C., Hoyer, M., and Kozuch, M., 2001, Velocity field across the southern Caribbean plate boundary
and estimates of Caribbean/South American plate motion using GPS geodesy 1994–2000: Geophysical Research Letters, v. 28, no. 15, p. 2987–2990,
doi: 10.1029/2001GL013183.
Pindell, J.L., Higgs, R., and Dewey, J., 1998, Cenozoic palinspastic reconstruction, paleogeographic evolution, and hydrocarbon setting of the northern
margin of South America, in Pindell, J.L., and Drake, C., eds., Paleogeographic evolution and non-glacial eustasy, northern South America: SEPM
(Society for Sedimentary Geology) Special Publication 58, p. 45–85.
Robson, G.R., 1964, An earthquake catalog for the Caribbean, 1530–1960: Bulletin of the Seismological Society of America, v. 54, p. 785–832.
Russo, R., Speed, R.C., Okal, E., Shepherd, J., and Rowley, K., 1993, Seismicity
and tectonics of the southeastern Caribbean: Journal of Geophysical Research, v. 98, no. B8, 14,299–14,319.
Saleh, J., Edwards, K., Barbaste, J., Balkaransingh, S., Grant, D., Weber, J., and
Leong, T., 2004, On some improvements in the geodetic framework of Trinidad and Tobago: Geodesy Review, v. 37, p. 294.
Soto, M.D., Mann, P., Escalona, A., and Wood, L.J., 2007, Late Holocene strikeslip offset of a subsurface channel interpreted from three-dimensional
seismic data, eastern offshore Trinidad: Geology, v. 35, p. 859–862, doi:
10.1130/G23738A.1.
Teyssier, C., Tikoff, B., and Weber, J., 2002, Attachment between brittle and
ductile crust at wrenching plate boundaries: European Geosciences Union
Stephan Mueller Special Publication Series, v. 1, p. 75–91.
Weber, J., Saleh, J., Balkaransingh, S., Ambeh, W., Dixon, T., and Webb, F., 1999,
1901–1995 Geodetic displacements in Trinidad: New constraints on local
earthquake hazards and Caribbean–South American neotectonics [abs.]:
Eos (Transactions, American Geophysical Union), v. 80, p. F.
Weber, J., Dixon, T., DeMets, C., Ambeh, W., Jansma, P., Mattioli, G., Saleh,
J., Sella, G., Bilham, R., and Perez, O., 2001, A GPS estimate of relative
motion between the Caribbean and South American plates, and geologic
implications for Trinidad and Venezuela: Geology, v. 29, p. 75–78, doi:
10.1130/0091-7613(2001)029<0075:GEORMB>2.0.CO;2.
Weber, J.C., Saleh, J., Balkaransingh, S., Dixon, T., Ambeh, W., Leong, T.,
Rodriguez, A., and Miller, K., 2010, Triangulation-to-GPS and GPS-toGPS geodesy in Trinidad, West Indies: Neotectonics, seismic risk, and
geologic implications: Marine and Petroleum Geology, doi: 10.1016/j
.marpetgeo.2009.07.010 (in press).
Wells, D.L., and Coppersmith, K.J., 1994, New empirical relationships among
magnitude, rupture length, rupture width, rupture area, and surface displacement: Bulletin of the Seismological Society of America, v. 84, p. 974–1002.
Manuscript received 2 December 2009
Revised manuscript received 19 February 2010
Manuscript accepted 21 February 2010
Printed in USA
GEOLOGY, August 2010