Content uploaded by Daisuke Kobayashi
Author content
All content in this area was uploaded by Daisuke Kobayashi on Feb 06, 2019
Content may be subject to copyright.
Content uploaded by Daisuke Kobayashi
Author content
All content in this area was uploaded by Daisuke Kobayashi on Feb 06, 2019
Content may be subject to copyright.
RESEARCH LETTER
10.1002/2014GC005234
Kinematics of the western Caribbean: Collision of the Cocos
Ridge and upper plate deformation
Daisuke Kobayashi
1,2
, Peter LaFemina
1
, Halld
or Geirsson
1
, Eric Chichaco
3
, Antonio A. Abrego
4
,
Hector Mora
5
, and Eduardo Camacho
3
1
Department of Geosciences, The Pennsylvania State University, University Park, Pennsylvania, USA,
2
Now at Department
of Geological Sciences, University of Idaho, Moscow, Idaho, USA,
3
Instituto de Geociencias, Universidad de Panam
a, Pan-
ama City, Panama,
4
Engineering Division, Panama Canal Authority, West Corozal, Panama City, Panama,
5
Servicio Geo-
l
ogico Colombiano, Bogot
a D. C., Colombia
Abstract Subduction of the Cocos plate and collision of the Cocos Ridge have profound effects on the
kinematics of the western Caribbean, including crustal shortening, segmentation of the overriding plate,
and tectonic escape of the Central American fore arc (CAFA). Tectonic models of the Panama Region (PR)
have ranged from a rigid block to a deforming plate boundary zone. Recent expansion of GPS networks in
Panama, Costa Rica, and Colombia makes it possible to constrain the kinematics of the PR. We present an
improved kinematic block model for the western Caribbean, using this improved GPS network to test a suite
of tectonic models describing the kinematics of this region. The best fit model predicts an Euler vector for
the counterclockwise rotation of the CAFA relative to the Caribbean plate at 89.10W, 7.74N, 1.193Ma
21
,
which is expressed as northwest-directed relative block rates of 11.3 61.0–16.5 61.1 mm a
21
from north-
ern Costa Rica to Guatemala. This model also predicts high coupling along the Nicoya and Osa segments of
the Middle American subduction zone. Our models demonstrate that the PR acts as a single tectonic block,
the Panama block, with a predicted Euler vector of 107.65W, 26.50N, 0.133Ma
21
. This rotation manifests
as northeast migration of the Panama block at rates of 6.9 64.0–7.8 64.8 mm a
21
from southern Costa Rica
to eastern Panama. We interpret this motion as tectonic escape from Cocos Ridge collision, redirected by
collision with the North Andes block, which migrates to the northwest at 12.2 61.2 mm a
21
.
1. Introduction
Subduction and collision of bathymetric highs can have profound effects on the kinematics and geomor-
phic development of overriding plates, affecting crustal accretion, transport of fore-arc terranes, and seismic
and volcanic hazards. In the New Hebrides subduction zone, the D’Entrecasteaux Ridge, located on the sub-
ducting Australian plate, collides with the West North Fiji Basin, resulting in back-arc thrusting [Calmant
et al., 2003], arc fragmentation [Taylor et al., 1995], and rapid arc rotation relative to the Australian plate
[Wallace et al., 2005, 2009]. The Carnegie Ridge, an aseismic ridge formed above the Galapagos hotspot and
located on the Nazca plate, collides into and underthrusts the North Andes block along the Colombia-
Ecuador trench. Carnegie Ridge collision is assumed to contribute to northeastward motion of the North
Andes block, possibly by tectonic escape [e.g., Gutscher et al., 1999; Trenkamp et al., 2002; Egbue and Kellogg,
2010]. Ridge subduction and collision results in strong mechanical coupling on the plate boundary [Vogt
et al., 1976; Cloos, 1993; Scholz and Small, 1997] and horizontal forces [van Benthem and Govers, 2010] that
are accommodated by deformation within the overriding plate.
Along the western Caribbean plate boundary, the Cocos plate (CO) subducts along the Middle America
Trench (MAT) at rates of 67 mm a
21
in Guatemala to 80 mm a
21
in southern Costa Rica [DeMets et al., 2010],
resulting in M
W
>7 earthquakes and a Holocene volcanic arc stretching from central Costa Rica to Guate-
mala (Figure 1). In southern Central America, collision of the Cocos Ridge, a 2 km high aseismic ridge stand-
ing on >20 km thick oceanic crust [Walther, 2003], is the largest geodynamic force acting on the Caribbean
plate (CA) [van Benthem and Govers, 2010], driving crustal shortening directly inboard [Fisher et al., 2004],
and tectonic escape of the Central American fore arc (CAFA) northwest of the ridge axis [LaFemina et al.,
2009]. Cocos Ridge collision has resulted in extensive upper-plate deformation and tectonic modification of
the western Caribbean from the Pliocene to present [MacMillan et al., 2004; LaFemina et al., 2009]. Geodetic
studies indicate interseismic elastic strain accumulation along the MAT and northwestward motion of the
Key Points:
Collision of the Cocos Ridge is
responsible for the upper plate
kinematics
The Panama Region acts as a single
tectonic block, migrating northeast
Interseismic coupling is high along
the Nicoya and Osa segments of the
trench
Supporting Information:
ReadMe
kobayashietal_supp
Correspondence to:
D. Kobayashi,
daisuke@vandals.uidaho.edu
Citation:
Kobayashi, D., P. LaFemina, H.
Geirsson, E. Chichaco, A. A. Abrego, H.
Mora, and E. Camacho (2014),
Kinematics of the western Caribbean:
Collision of the Cocos Ridge and upper
plate deformation, Geochem. Geophys.
Geosyst.,15, 1671–1683, doi:10.1002/
2014GC005234.
Received 8 JAN 2014
Accepted 15 APR 2014
Accepted article online 19 APR 2014
Published online 29 MAY 2014
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1671
Geochemistry, Geophysics, Geosystems
PUBLICATIONS
CAFA relative to the CA at 8–16 mm a
21
from northern Costa Rica to Guatemala [Correa-Mora et al., 2009;
LaFemina et al., 2009, and references therein]. Shallow, destructive M
w
6.5 strike-slip earthquakes along
the Central American volcanic arc accommodate CAFA-CA relative motion [La Femina et al., 2002; Corti
et al., 2005]. The Central Costa Rica Deformed Belt (CCRDB; Figure 1), a seismically active zone of distributed
deformation, has been proposed to mark the southeastern margin of the CAFA [Marshall et al., 2000; Lewis
et al., 2008].
Directly inboard of the Cocos Ridge, marine deposits record Quaternary uplift of the outer fore arc at rates of
6.5 mm a
21
and 2.1–7.7 mm a
21
on the Osa and Burica Peninsulas, respectively [Sak et al., 2009; Morell et al.,
2011]. The inner fore arc Fila Coste~
na Thrust Belt (Figure 1), an imbricate thrust system running subparallel to
the MAT, has a minimum cumulative displacement of 36 km since the middle Pliocene [Fisher et al., 2004;
Sitchler et al., 2007]. The Cordillera de Talamanca, a 3800 m high extinct volcanic arc, coincides with the
width of the Cocos Ridge (Figure 1), which has been implicated with back-arc migration of the volcanic arc in
central Costa Rica at <2Ma[Marshall et al., 2003]. Crustal shortening is also indicated by back-arc thrusting
and M
w
>7 earthquakes along the North Panama Deformed Belt (NPDB), a seismically active fold and thrust
belt [Adamek et al., 1988; Silver et al., 1990; Camacho et al., 2010] and incipient subduction zone [Camacho
et al., 2010], running from central Costa Rica to the Maracaibo subduction zone, Colombia (Figure 1).
The Panama Region (PR) lies north and east of Cocos Ridge and is bounded by the Cocos (CO), Nazca (NZ),
and CA plates, and the CAFA, Choco (CH), and North Andes (ND) blocks (Figure 1). Adamek et al. [1988]
-7000
5000
0 100 200
km
20 mm a
(m)
94˚ W 92˚ W 90˚ W 88˚ W 86˚ W 84˚ W 82˚ W 80˚ W 78˚ W 76˚ W
2˚ N
4˚ N
6˚ N
8˚ N
10˚ N
12˚ N
14˚ N
16˚ N
Colombia
Panama
Nicaragua
El Salvador
Guatemala
Honduras
80
74
67
37
NORTH
AMERICAN
PLATE
CARIBBEAN
PLATE
CAFA
NAZCA
PLATE
PANAMA
REGION
NORTH
ANDES
BLOCK
CHOCO
BLOCK
60 mm a
Middle America Trenc h
CHOCO
BLOCK
Cocos Ridge
NPDB
SPDB
UF
60
COCOS
PLATE
COCOS
PLATE
NAZCA
PLATE
NORTH
AMERICAN
PLATE
Z
FUA
Colombia Trench
Maracaibo SZ
NP
CCRDB
OP
FC
CT
BP
Panama Fracture Zone
CFZ
BFZ
Figure 1. GPS-derived velocity field (black vectors) in a Caribbean-fixed reference frame. Velocity uncertainty ellipses are removed for
clarity; see supporting information Figure S1 and Tables S1 and S2 for velocity uncertainties. Note vector scale in legend. Azimuth and rela-
tive rate in mm a
21
of the Cocos and Nazca plates relative to the Caribbean plate, open arrows. Holocene volcanoes, red triangles. Plate
boundaries, heavy black lines. National boundaries, thin black lines. CAFA, Central American fore arc; NPDB, North Panama Deformed Belt;
SPDB, South Panama Deformed Belt; UF, Unguia Fault Zone; AUFZ, Atrato-Uraba Fault Zone; Maracaibo SZ, Maracaibo Subduction Zone;
BFZ, Balboa Fracture Zone; CFZ, Coiba Fracture Zone. Inset: geologic features in Costa Rica, including; CCRDB, Central Costa Rica Deformed
Belt; FC, Fila Coste~
na Thrust Belt; CT, Cordillera de Talamanca; NP, Nicoya Peninsula; OP, Osa Peninsula; BP, Burica Peninsula. Black box indi-
cates an area of data used for profile shown in Figure 4. Topographic and bathymetric data from Amante and Eakins [2009].
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1672
suggested that a diffuse seismic zone marked the eastern boundary of the PR. Nazca-CA relative motion is
highly oblique to the margin and is dominated by oblique convergence along the western segment of the
system (i.e., where the Panama, Balboa, and Coiba fracture zones converge on the margin; Figure 1) and
left-lateral shear in the central and eastern segments, along the South Panama Deformed Belt (SPDB) at 37
mm a
21
[DeMets et al., 2010] (Figure 1). Along the Colombia Trench, the Nazca plate subducts obliquely
(20)[Trenkamp et al., 2002] under the Choco block from 7Nto3
N latitude and the North Andes block
south of 3N (Figure 1). The NZ-ND convergence rate is 52 mm a
21
at 2N in the MORVEL56 plate model
[Argus et al., 2011]. Caribbean-PR convergence is accommodated along the NPDB.
The Panama Region has been described by several tectonic models based on geologic, seismic, and geo-
detic data. Silver et al. [1990] suggest internal deformation of the isthmus along strike-slip faults, while
others suggest either the motion of a discrete tectonic block called the Panama block (PB) [e.g., Adamek
et al., 1988, Vergara Mu~
noz, 1988], or interaction of smaller tectonic blocks [Mann and Corrigan, 1990; Tren-
kamp et al., 2002; Rockwell et al., 2010]. The effect of Cocos Ridge collision on the kinematics of the Panama
Region, itself colliding with northwestern South America since Miocene time [e.g., Kellogg and Vega, 1995;
Farris et al., 2011], has not been studied or documented. We present a new GPS-derived surface velocity
field with unprecedented coverage in Central America (i.e., the western Caribbean) and northwestern South
America, and use these data to test these tectonic models. We invert the three-dimensional GPS velocities
and earthquake slip vectors for M
w
>6 plate and block boundary earthquakes in a three-dimensional kine-
matic block model [e.g., McCaffrey, 2002] to simultaneously solve for the Euler vectors that best describe
GPS site velocities and the magnitude and pattern of interseismic coupling on block bounding faults. We
construct and test various model geometries to find the best fit CAFA-PR boundary, to improve estimates of
CAFA motion, to reveal the tectonic nature (i.e., one block versus multiple blocks) of the PR, and to detail
coupling patterns along the Nicoya and Osa segments of the MAT.
2. GPS Velocity Field and Earthquake Slip Vector Data
We present a new 154 site, GPS-derived interseismic surface velocity field in Central America and northwest-
ern South America based on measurements spanning the period 1993–2012 (Figure 1; Table S1). We utilized
new GPS data collected at sites, where data existed from previous studies [e.g., Lundgren et al., 1993, 1999;
Trenkamp et al., 2002; Norabuena et al., 2004; Turner et al., 2007; LaFemina et al., 2009; Mora et al., 2011], and
from episodic and continuous GPS sites installed as part of this study. These GPS data sets were processed
together with GIPSY-OASIS II version 5.1 in precise point positioning mode, using standard clock and satel-
lite ephemeris data from the Jet Propulsion Laboratory [Zumberge et al., 1997] for consistent analysis. Daily
positions were estimated in the International Terrestrial Reference Frame (ITRF) 2005 [Altamimi et al., 2007]
and time series generated. The weighted root-mean squared error (WRMS) of a linear fit to the time series
was calculated to determine velocity uncertainties, following Geirsson et al. [2006]. The resultant velocities
were transformed to a stable Caribbean reference frame using the ITRF05-Caribbean Euler vector from
MORVEL [DeMets et al., 2010] (Table S2). We removed episodic sites with time series shorter than 3 years
and continuous sites shorter than 1 year. We also removed sites on volcanoes that may be influenced by
volcanic deformation. In addition to our data set, we used velocity data and their uncertainties from pub-
lished velocity fields for El Salvador and Honduras [Correa-Mora et al., 2009] and Guatemala [Lyon-Caen
et al., 2006]. These additional velocity data were transformed and rotated to a stable Caribbean reference
frame for consistency with our analysis (Table S2). The velocity data of Lyon-Caen et al. [2006] were first
transformed into ITRF05.
The horizontal velocity field in a CA reference frame [DeMets et al., 2010] (Table S2) indicates the major
tectonic signals for the western CA (Figures 1 and S1): (1) the MAT-CAFA system is dominated by margin
parallel motion from central Costa Rica to Guatemala at rates up to 19 mm a
21
; (2) convergence-parallel
rates on the Nicoya and Osa Peninsulas range between 12–27 mm a
21
and 24–37 mm a
21
, respectively;
(3) site velocities in central Costa Rica and western Panama indicate a fanning pattern of motion away
from the Cocos Ridge axis; (4) northeast-directed motion in central Panama at rates of 6–13 mm a
21
; and
(5) sites located in northwestern South America are moving northwest at 17 mm a
21
. The vertical veloc-
ity field demonstrates: (1) outer fore-arc subsidence along the Nicoya and Osa Peninsulas at rates of up to
24 mm a
21
and 11 mm a
21
, respectively and (2) subsidence in the Managua graben (Figure S1b).
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1673
Earthquake focal mechanism solu-
tions and slip vectors of interplate
earthquakes indicate the geometry
of and direction of relative plate
motions on plate boundary faults.
We used centroid moment tensor
solutions for M
W
>6 earthquakes
with depths less than 60 km from
the Global CMT Catalog and for the
time period 1976–2011 [Dziewonski
et al., 1981] (Figure 2; Table S3). We
selected events and nodal planes
consistent with the location and
nature of known plate and block
boundaries to obtain slip vector
data that further constrained block
kinematics in our model inversion.
3. Kinematic Block
Modeling
We use kinematic block modeling
and invert the three-dimensional
velocity field and earthquake slip vectors (Tables S1–S3), to investigate the kinematics of the western Carib-
bean, including the estimation of Euler vectors for proposed tectonic blocks, and the magnitude and spatial
variability in coupling on block bounding faults. We use the modeling code DEFNODE and the methods of
McCaffrey [2002]. The modeled block boundaries are discrete boundaries defined using the results of pub-
lished geologic mapping, seismic tomography, seismic reflection, and refraction profiles, and earthquake
relocation studies. In DEFNODE, the block boundaries are approximated by a three-dimensional mesh
defined by surface nodes and the downdip geometry of the fault systems (see discussion of model resolu-
tion below).
We test nine block model geometries (Models 1–9) to investigate the different tectonic and kinematic mod-
els of the western Caribbean [e.g., Adamek et al., 1988; Silver et al., 1990; Trenkamp et al., 2002; McCaffrey,
2002; Turner et al., 2007; LaFemina et al., 2009; Rockwell et al., 2010] (Figure 3) and compare long-term geo-
logic and decadal geodetic estimates of plate boundary deformation. Models 1–5 investigate the location
and nature of the CAFA-PR boundary (Figures 3a–3d), while Models 6–9 test different models specific to the
PR, whereby the isthmus is transversed by proposed active or mapped faults (Figures 3e–3h). Models 4 and
5 (Figure 3d) have a discrete CCRDB block; the CCRDB is modeled as a rigid block in Model 4 and as a
deforming zone in Model 5. Model 9 (Figure 3h) reproduces boundaries within the PR proposed by Rockwell
et al. [2010] on the basis of paleoseismic fault studies in the Panama Canal Zone.
The spatial limitation of land-based geodetic networks makes it difficult to constrain the pattern and magni-
tude of interseismic elastic deformation (i.e., coupling) near a subduction zone trench. Several paradigms
have been proposed for modeling interseismic coupling on subduction zone plate interfaces and specifi-
cally the seismogenic zone. One model suggests that the shallow interface from the trench to the updip
limit of seismogenesis is a region of free slip [e.g., Bevis and Martel, 2001]. However, McCaffrey et al. [2000]
and McCaffrey [2002] suggest that the estimated strain rate at these depths would be anomalously high if
the shallow interface is not fully coupled. Wang et al. [2003] predict an exponentially decaying downdip
trend in coupling at depths deeper than the updip limit of the seismogenic zone, based on thermal models
of the Cascadia margin [Hyndman and Wang, 1993].
The proximity of the Nicoya (50 km) and Osa (20 km) Peninsulas to the Middle America Trench allows us
to test different models for the magnitude and spatial variability of interseismic coupling at the updip limit
of the seismogenic zone. Coupling (u) here is the ratio of slip deficit to the total relative plate motion. In
DEFNODE, one can impose various coupling constraints, three of which we test along the MAT interface in
0 100 200
0
20
40
60
80
0
20
40
60
80
Depth (km)
MW 6.0
7.0
8.0
km
2˚ N
4˚ N
6˚ N
8˚ N
10˚ N
12˚ N
14˚ N
16˚ N
94˚ W 92˚ W 90˚ W 88˚ W 86˚ W 84˚ W 82˚ W 80˚ W 78˚ W 76˚ W
Figure 2. Earthquake focal mechanisms for Central America and western Colombia
for the time period 1976–2011. We selected M
W
>6 events shallower than 60 km and
consistent with the location and nature of known plate and block boundaries. Focal
mechanisms are scaled by magnitude and color-coded by depth range (see legend).
Data are from the Global CMT project [Dziewonski et al., 1981].
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1674
PR
CAFA
b
8˚ N
12˚ N
ND
NZ
CA
CAFA
CO
a
PR
CAFA
c
PR
CAFA
CCRDB
d
PR
CAFA
e
PR
CAFA
f
PR
CAFA
g
CAFA
h
88˚ W 84˚ W 80˚ W 76˚ W
200 km CH
UF
CD
FC
SAF
8˚ N
12˚ N
8˚ N
12˚ N
8˚ N
12˚ N
88˚ W 84˚ W 80˚ W 76˚ W
PANC
Figure 3. Block boundaries for Models 1–9. Thick gray lines indicate boundaries of the Central American fore arc (CAFA) and the Panama Region (PR). Thin gray lines are other block
boundaries. Thick black lines indicate boundaries that segment the PR. The Central Costa Rica Deformed Belt (CCRDB) is indicated by the gray shaded area. CO, Cocos plate; CA, Carib-
bean plate; NZ, Nazca plate; CH, Choco block; ND, North Andes block; UF, Unguia Fault Zone; PANC, Panama-Caribbean block. See text for model descriptions.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1675
this study. The first pattern imposes no constraints and independently determines ufor each node (here-
after Type 0 or no constraint). This option allows for the estimation of detailed coupling patterns in high-
resolution areas, namely the Nicoya and Osa segments. We tested our GPS network resolution and found
that the network can resolve coupling patches up to 30 km (along strike) by 50 km (downdip) and up to
20 km offshore on the Osa segment, and patches up to 30 km (along strike) by 40 km (downdip) up to
30 km offshore on the Nicoya segment of the MAT (Figure S2). The second constraint type imposes a linear
decrease in coupling downdip (hereafter Type 1 or linear decrease), which forces uat a node to be less
than or equal to the node directly updip of it. In our modeling using the Type 1 constraint, each fault plane
is fully coupled at the surface nodes (e.g., at the Middle America Trench), and udecreases to zero at or
before the bottom-most node. This modeling constraint is assigned for all block-bounding faults (e.g., the
CAFA-CA boundary fault or the NPDB) for which coupling is estimated in all model runs, except for the MAT
(see above). The last model constraint tested is a constraint where a downdip decrease in ubetween
depths Z
1
and Z
2
(Z
1
<Z
2
) is prescribed by an exponential function (hereafter Type 2 or exponential
decrease) as proposed by Wang et al. [2003]. Full coupling is assigned from the surface to a depth Z
1
with
no coupling below Z
2
in our models; Z
1
and Z
2
are estimated in the inversion.
For each model, we test the three coupling constraints on the MAT and solve for the Euler vectors of the
modeled blocks and the magnitude and spatial variability of interseismic coupling on block boundaries,
minimizing the misfit between the model and data (i.e., the three-dimensional GPS velocity field and earth-
quake slip vectors). Values of the chi-square statistic are compared to determine a model of better fit (Table
S4). We perform an Ftest [Stein and Gordon, 1984] to determine if the better fit is statistically significant.
The probability of a significant improvement, not due to mere chance, is given with the Ftest statistics and
the Fprobability density function (Table S5).
4. Results and Discussion
We tested nine models to find the best fit CAFA-PR boundary to improve estimates of CAFA motion, to
reveal the tectonic nature of the PR, and to detail coupling patterns along the Nicoya and Osa segments of
the MAT. The resultant Euler vectors (Table 1) and estimated magnitudes and patterns of coupling improve
our understanding of the kinematics of the western Caribbean. Below we describe the results from our best
fit models and how they reflect on the kinematics of the western Caribbean.
4.1. Modeling Results
All of the models tested consistently predict: (1) counterclockwise rotation of the CAFA relative to the CA
about an Euler pole located 500–600 km to its south (Figures 4a and S3–S28), and (2) moderate (60%)
coupling to a depth of at least 30 km along the Nicoya and Osa segments of the MAT (Figures 5a and S3–
S28). All the models that have a nonsegmented PR (Models 2–5; Figures 3a–3d) predict counterclockwise
Table 1. Estimated Euler Vectors Relative to the Caribbean Plate
Block Longitude Latitude x(Ma
21
) Azimuth
a
e
maxa
e
mina
CAFA
This study, the best fit model 289.10 7.74 1.193 60.122 12.73 0.68 0.33
Alvarado et al. [2010] 291.2 4.5
LaFemina et al. [2009] 290.15 3.90 0.469
Turner et al. [2007] 288.4 8.9 1.957
McCaffrey [2002] 285 10.5 0.04
PR
This study, the best fit model 2107.65 26.50 0.133 60.265 118.9 32.32 3.05
Argus et al. [2011] 2171.75 20.24 0.100
Bird [2003] 2172 0 0.10
ND
This study, the best fit model 258.77 31.10 20.233 60.163 212.95 30.78 1.30
Argus et al. [2011] 270.82 41.3 20.198
Bird [2003]
b
270.74 41.04 20.197
CH 295.91 265.11 0.405 60.832 22.01 101.07 19.27
a
Azimuth, e
max
, and e
min
refer to the azimuth of the major axis of the error ellipse and its maximum and minimum uncertainties in
degrees.
b
Calculated using the South America-Caribbean angular velocity by DeMets et al. [2010].
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1676
rotation of the PB relative to the CA (Figures 4a and S6–S16). Segmented PR models (Models 6–9; Figures
3e–3h) do not significantly improve model fit (Tables S4 and S5). For all the block configurations, models
with the Type 2 coupling constraint (i.e., exponential decrease) have significantly poorer model fits com-
pared to those with Type 0 or 1 constraint (see Table S4).
The best fit model is Model 2 (Figure 3b) with Type 1 coupling constraint, whose reduced v
2
is 8.94 (Table
S4). This model indicates that the western Caribbean can be defined as two tectonic blocks, the CAFA and
PR (hereafter named the Panama block (PB); Figure 4a). The boundary between the CAFA and PB was found
to be the western limit of the CCRDB, with a relative rate of 15 64.2 mm a
21
at an azimuth of
N73.5W 68.3(Figure 4b). This indicates extension between CAFA and PB that is accommodated within the
CCRDB by shallow, low-magnitude normal and strike-slip earthquakes indicating northwest-southeast
extension [Marshall et al., 2000; Lewis et al., 2008].
The second and third best fit models have a comparable model fit. The second best fit model is Model 9
(Figure 3h) with Type 1 coupling constraint (Model 9-1; v
2
59.53; Figure S27; Table S4), in which the Pan-
ama isthmus is segmented following Rockwell et al. [2010]. Slip rates along the faults transecting the PB in
Model 9-1 are not well constrained geodetically due to the sparse data set. The third best fit model is
Model 2 (Figure 3b) with Type 0 coupling constraint (v
2
59.74; Figure S6; Table S4). This model has the
same block boundaries as the best fit model, but no coupling constraints are imposed on the MAT. The
results for the third best fit model are in good agreement with the best fit model.
4.2. Kinematics of the Central American Fore Arc
In our best fit model, the CAFA rotation is defined by an Euler pole located at 89.10W, 7.74N and at a rate
of 1.193 60.122Ma
21
(Figure 4a). Table 1 shows a comparison of CAFA Euler vectors with other studies
using GPS data. The differences in the Euler vectors possibly result from inversion method and an extent of
the CAFA in each model. McCaffrey [2002], LaFemina et al. [2009], and this study adopt a simultaneous inver-
sion for Euler vector and elastic strain accumulation, while the others solve only for the Euler vector. This
study has the longest modeled CAFA extending from northern Costa Rica to Guatemala.
The Euler vector of the CAFA in our best fit model is expressed as modeled fore-arc rates of 11.3 61.0–
16.5 61.1 mm a
21
from northern Costa Rica to Guatemala, respectively (Figure 4a; Table S6). For compari-
son, our second and third best models result in fore-arc rates of 12.1–16.9 mm a
21
and 10.6–16.5 mm a
21
,
respectively. These results constrain the magnitude of dextral shear strain accommodated by shallow
destructive earthquakes along the CA-CAFA boundary [La Femina et al., 2002; Corti et al., 2005] and there-
fore the potential seismic moment of future earthquakes along the Central American volcanic arc.
4.3. Kinematics of the Panama Block
The Euler vector for motion of the PB relative to the CA has a pole located at 107.65W, 26.50N, and a rota-
tion rate of 0.133 60.265Ma
21
(counterclockwise; Table 1). This Euler vector is significantly different from
those estimated in other studies based mainly on geologic and seismologic observations (Table 1) [Bird,
2003; Argus et al., 2011]. The rotation about this pole is expressed as northeast motion at rates of 6.9 mm
a
21
to 7.8 mm a
21
(Figure 4a; Table S6). Although the predicted Euler vector for the PB is poorly con-
strained (Table 1), the modeled horizontal velocities clearly capture the overall trend of the observed veloc-
ity field, constraining relative plate and block motions and indicating negligible internal deformation of the
PB (Figures 4a and S3). The northeast motion of the PB results in oblique convergence with the CA along
the NPDB, convergence with the Choco and North Andes blocks, and sinistral shear and extension along
the SPDB.
Panama block-CA convergence in central Panama is 7.5 64.4 mm a
21
at an azimuth of N39E 626(Figure
4a; Table S6). A Wadati-Benioff zone within the subducting Caribbean plate has been hypothesized and
traced to 80 km depth (assuming a dipping slab of 170 km length) along the eastern segment of the
NPDB [Camacho et al., 2010]. If the Wadati-Benioff zone indicates the leading edge of the subducting CA,
and we assume a constant convergence rate, onset of CA subduction occurred at least 22–23 Ma. This indi-
cates that underthrusting of the CA predates Cocos Ridge collision, hypothesized at <3Ma[MacMillan et al.,
2004], but is synchronous with Panama’s collision with South America starting at 23–25 Ma [Farris et al.,
2011]. In the second best fit model (Model 9-1), the central and eastern segments of the NPDB form the
northern boundary of an elongated block, the PANC block (Figure 3h). The convergence rate (14.0 678.8
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1677
mm a
21
) and azimuth (N35E 647) between the block and the CA are poorly constrained (Figure S27).
Nevertheless, this rate would result in onset of CA subduction at 12 Ma.
We tested proposed tectonic models of the PR that have suggested segmentation of the isthmus. Our Mod-
els 6–9 (Figures 3e–3h and S17–S28) provide some constraints on several of the major faults proposed for
the region (e.g., the Sona-Azuero and Panama Canal fault systems); however, our estimated rates and azi-
muths are poorly constrained, and the sense of slip for most faults is the reverse of that found in geologic
studies. Further expansion of the geodetic network, with a focus on investigating the major regional faults,
may improve our understanding of the rates across these faults.
4.4. Earthquake Cycle Deformation
Our new estimates of coupling on the CO-CAFA plate boundary have implications for interseismic elastic
strain accumulation and therefore earthquake hazards. Along the Nicoya segment of the MAT, we estimate
a CO-CAFA relative rate of 72.0 61.0 mm a
21
at an azimuth of N28.1E 60.7(Figure 4b; Table S6) and pre-
dict an 2500 km
2
zone of moderate (40–60%) coupling, likely reflecting a smaller region of 100% coupling,
located beneath central Nicoya Peninsula (Figure 5a). The location of interseismic coupling is colocated
with the location and rupture area of historical earthquakes (Figure 5a). Earthquakes in 1900 (M
S
7.1)
[Pacheco and Sykes, 1992], 1950 (M
S
7.7) [Abe, 1979], 1978 (M
W
6.9) [Protti et al., 2001], and 2012 (M
W
7.6)
[Protti et al., 2013; Yue et al., 2013] were located in this region of high coupling (Figure 5a). Assuming that
the full convergence rate was accumulated as elastic strain, 3.6 m of slip deficit accumulated prior to the
1950 event (Figure 5b). If the 1950 event fully released this slip deficit, then this segment accumulated 2.0
m of slip deficit by the 1978 event. Protti et al. [2001] estimate only 15% of potential slip for the 1978 event
was released, which left 1.7 m of slip deficit (Figure 5b). Since 1978, 2.4 m of slip deficit has accumulated
for a total of 4.1 m. Preliminary analysis of the 5 September 2012 Nicoya earthquake indicates a maximum
of 3.0 m slip [Protti et al., 2013; Yue et al., 2013] (Figure 5b), but with considerable postseismic deformation
[Geirsson et al., 2013].
CAFA CA
CO
NZ
ND
PB
NA
CH
2˚ N
4˚ N
6˚ N
8˚ N
10˚ N
12˚ N
14˚ N
16˚ N
94˚ W 92˚ W 90˚ W 88˚ W 86˚ W 84˚ W 82˚ W 80˚ W 78˚ W 76˚ W 74˚ W
a
20 mm a relative to:
Block Rate
CA
PR
CAFA
Site Velocity
20 mm a relative to CA
Data
Model
0200
km
CAFA
b
CA
CO
NZ
PB
CAFA
ND
11
7
15
75
73
59
30
13
12
N
10 mm a
80
Figure 4. (a) Observed (black) and modeled (blue) velocities relative to CA in the best fit model. Note vector scale in legend. Arrows indi-
cate azimuth and rate of block rate relative to the CA (pink), PB (green), and CAFA (white). Relative block rates and the Euler pole for the
CAFA (red star) are shown with 1runcertainty ellipses. Block boundaries are shown as thick gray lines. (b) Inset: velocity triangle diagram.
Relative velocities of blocks adjacent to each other are given in mm a
21
. For the CAFA and PB motions, velocities in northern Costa Rica
and central Panama, respectively, are used.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1678
Along the Osa segment of the MAT, we estimate a CO-PB convergence rate of 73.1 63.8 mm a
21
at an azi-
muth of N22.3E 62.5(Figure 4b; Table S6) and predict a broad zone of 50–75% coupling in this region
(Figure 5a), which likely reflects smoothing of fully coupled patches and is consistent in area with coupling
on the underthrusting Cocos Ridge axis. Large (M >7) earthquakes have occurred in 1904, 1941, and 1983
on the Osa segment, where the Cocos Ridge collides and underthrusts the PB (Figure 5a). The 1941 M
S
7.5
and 1983 M
W
7.4 earthquakes were located in the same epicentral area northeast of Osa Peninsula [Adamek
et al., 1987] (Figure 5a). If CO-PB convergence is accommodated only as elastic strain, 3.1 m of slip deficit
accumulated between the 1941 and 1983 events (Figure 5c). Adamek et al. [1987] estimated 0.6 m of slip for
the 1983 event, indicating a remaining 2.5 m of slip deficit. Approximately 2.2 m of slip deficit has accrued
since 1983 for a total of 4.7 m, a larger slip deficit than that accumulated when the 1983 event occurred
(Figure 5c). This implies either that there is the potential for a larger magnitude earthquake on the segment
or that a portion of accumulated slip deficit was released as afterslip following the 1983 event.
Another possibility is that the estimated slip for the 1983 earthquake is an accurate representation of the
amount of slip deficit prior to the 1983 event. In this case, only 14 mm a
21
of slip deficit accumulated
between 1941 and 1983, and the remaining 59 mm a
21
was either released as afterslip or accommodated
by plastic deformation, such as crustal shortening, or a combination of both. The Fila Coste~
na Thrust Belt
has accommodated up to 40 mm a
21
of cumulative slip since middle Pliocene time [Fisher et al., 2004; Sitch-
ler et al., 2007]. The four sites along the Pacific coast of the Osa and Burica Peninsulas have large model
residuals showing a fanning-out pattern from the Cocos Ridge axis (Figure S29). These residuals may indi-
cate nonmodeled plastic deformation within the Fila Coste~
na Thrust Belt, creeping on faults of this system,
Middle America Trench
OP
NP
BP
1904
1950
1900
1978
72
73
CCRDB
NPDB
60
1983
1941
60
1991
CO
CAFA
PB
CO-C AFA
CO-CAFA
CO-PB
CO-PB
2012
2012
1822
0 50 100
km
0.0 0.5 1.0
Phi 60 mm a
a
7˚ N
8˚ N
9˚ N
10˚ N
11˚ N
86˚ W 85˚ W 84˚ W 83˚ W 82˚ W
bc
1900
Ms 7.1
2012
Mw 7.6
1950
Ms 7.7
1978
Mw6.9
3.6
2.0
4.1
Slipde cit (m)
Yearand event
72 mm a
2012
3.1
Slipde cit (m)
Yearand event
1941
Ms 7.5
1983
Mw7.4
73 mm a
2.5
4.6
0.6
14 mm a
60
Figure 5. (a) Estimated interseismic coupling (phi) along the Nicoya and Osa segments of the MAT as predicted by the best fit model with
no coupling constraints. Open arrows show velocity of the Cocos plate relative to the overriding CAFA and PB blocks. Gray, white, and
pink dots indicate epicenters of historical earthquakes along the Nicoya and Osa segments of the MAT and western NPDB [Abe, 1979; Ada-
mek et al., 1987; Protti et al., 2001; Pacheco and Sykes, 1992; National Earthquake Information Center, 2012]. Dashed loops, rupture area for
selected events. Color of dashed loops is the same as the corresponding epicenter. Inset: Estimated slip deficit on the (b) Nicoya and (c)
Osa segments of the MAT.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1679
or nonmodeled geodynamic effects
of Cocos Ridge collision [e.g., LaFe-
mina et al., 2009; van Benthem and
Govers, 2010]. Although our best fit
model does not require elastic
strain accumulation across the Fila
Coste~
na, our data are not inconsis-
tent with the expected pattern of
deformation across this system
(Figure 6). If the full residual slip
deficit (59 mm a
21
) of the CO-PB
motion is accommodated by plastic
deformation across the Fila Coste~
na
and inboard of the Cocos Ridge,
then the minimum slip deficit
accrued on the Osa segment since
the 1983 earthquake is 0.4 m, 67%
of the slip released during the 1983
event (Figure 5c). Assuming a peri-
odic and slip predictable earthquake model, and strain partitioning across the plate boundary thrust and
Fila Coste~
na Thrust Belt, this indicates that the next M >7 earthquake may not occur until 2026 on the
Osa segment.
Panama block-CA convergence along the western NPDB is 6.9 64.0 mm a
21
at an azimuth of N40E 626
(Figure 4a; Table S6), constraining the magnitude of back-arc thrusting. Large-magnitude earthquakes that
have occurred on this segment accommodate this relative motion. The 1822 M
s
7.6 and 1991 M
w
7.7 El
Limon earthquakes ruptured the same segments [Su
arez et al., 1995, and references therein], suggesting a
169 year recurrence interval. However, our estimated PB-CA convergence rate indicates that the 1822 M
S
7.6 earthquake did not fully release the accumulated slip deficit. If 100% of convergence is accommodated
on the NPDB, and if earthquakes occur here periodically, then the recurrence interval for large 1991-type
0 100 200
0
20
40
OP FC CT
SW NE
0.0 1.0
Phi
0
60
Depth (km)
CO CA
PB
Rate (mm a )
20
40
Distance from trench (km)
MAT
NPDB
Figure 6. Velocity profile and generalized cross section for southern Costa Rica. Red
dots show plate-motion parallel component of observed site velocity with 1rerror
bars. Best fit model profile, black line. Model with the Fila Coste~
na (FC) fault included,
dashed line. Vertical dashed lines indicate the location of the MAT and the NPDB. Geo-
logic cross section modified from Fisher et al. [2004]. Estimated coupling pattern (phi)
for the CO-PB interface is shown. Location of modeled d
ecollement of FC and the PB-
CA interface (i.e., NPDB) are in red.
0 100200
0 200
km
Block Rate
20 mm a relative to CA
CAFA
CA
CO
NZ ND
NA
CH
Cocos Ridge Axis
2˚ N
4˚ N
6˚ N
8˚ N
10˚ N
12˚ N
14˚ N
16˚ N
94˚ W 92˚ W 90˚ W 88˚ W 86˚ W 84˚ W 82˚ W 80˚ W 78˚ W 76˚ W
NZ
CH
CAFA
N
Figure 7. Proposed driving mechanism for motion of the Central America fore arc (CAFA) and Panama block (PB). Red vectors are relative
blocks rates from the best fit model. Motion of the Cocos Ridge and the ND and CH blocks, open white arrows. The CAFA and PB rotations
suggested by the best fit model, orange arrows. Theoretical rotation of the PB due to the Cocos Ridge collision, dashed blue arrow. Axis of
Cocos Ridge, dashed blue line.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1680
earthquakes would be 220–830 years. In the second best fit model (Model 9-1), the CCRDB block converges
with the CA at a rate of 15.8 69.1 mm a
21
at an azimuth of N32E 638along the western segment of the
NPDB. This rate translates into a recurrence interval of 100–360 years for a 1991-type earthquake.
4.5. Escape Tectonics in the Western Caribbean
The motion of the PB and CAFA away from the Cocos Ridge is indicative of tectonic escape [e.g., LaFemina
et al., 2009]. However, the eastern end of the region does not rotate clockwise toward the obliquely con-
verging NZ as predicted for a rigid indenter [Molnar and Tapponnier, 1975]; rather, the motion is counter-
clockwise toward the CA (Figure 7). Our best fit model predicts convergence of the ND block relative to the
CA at 12.2 61.2 mm a
21
at an azimuth of N56.4W 65.2(Figure 4a; Table S6). The motion of the CH block
relative to the CA is predicted to be 43.2 661.6 mm a
21
at N81.9W 616.5although it is poorly constrained
(i.e., there is only one GPS site on this block; Figure 4a; Table S6). In our models, the ND-CA and ND-CH rela-
tive motion is accommodated on the Atrato-Uraba fault zone (Figure 1). Here we propose that the current
motion of the PB is driven by tectonic escape from Cocos Ridge collision with a rotation symmetrical to the
CAFA; however, the CH and ND blocks drive the PB to the northwest inhibiting clockwise rotation of the PB,
resulting in the northeast migration of the block and its continued collision with northwestern South Amer-
ica (Figure 7).
5. Conclusions
Our kinematic block modeling demonstrates a major effect of Cocos Ridge collision on deformation and
development of the western Caribbean. Northwest motion of the CAFA at 11–17 mm a
21
is accommodated
by shallow destructive earthquakes along the Central America volcanic arc. Earthquakes within the CCRDB
accommodate relative extension between the CAFA and PB estimated here to be 15 mm a
21
. Ridge colli-
sion causes uplift in the outer fore arc and shortening across the Fila Coste~
na inboard of the Cocos Ridge,
potentially at 50% of the CO-PB convergence rate. A pressing implication of our results is the rapid accumu-
lation of interseismic strain along the Osa segment of the Middle America Trench for the next large, 1983-
type earthquake in southern Costa Rica. Northeast motion of the Panama block is interpreted as tectonic
escape from Cocos Ridge collision, redirected by northwestward forcing by and collision with the Choco
and North Andes blocks.
References
Abe, K. (1979), Size of great earthquakes of 1837–1974 inferred from tsunami data, J. Geophys. Res.,84(B4), 1561–1568, doi:10.1029/
JB084iB04p01561.
Adamek, S., F. Tajima, and D. Wiens (1987), Seismic rupture associated with subduction of the Cocos ridge, Tectonics,6, 757–774, doi:
10.1029/TC006i006p00757.
Adamek, S., C. Frohlich, and W. D. Pennington (1988), Seismicity of the Caribbean-Nazca boundary: Constraints on microplate tecton ics of
the Panama region, J. Geophys. Res.,93(B3), 2053–2075, doi:10.1029/JB093iB03p02053.
Altamimi, Z., X. Collilieux, J. Legrand, B. Garayt, and C. Boucher (2007), ITRF2005: A new release of the International Terrestrial Reference
Frame based on time series of station positions and Earth Orientation Parameters, J. Geophys. Res.,112, B09401, doi:10.1029/
2007JB004949.
Alvarado, D., et al. (2010) Forearc motion and deformation between El Salvador and Nicaragua: GPS, seismic, structural, and paleomagnetic
observations, Lithosphere,3(1), 3–21, doi:10.1130/L108.1.
Amante, C., and B. W. Eakins (2009), ETOPO1 1 arc-minute global relief model: Procedures, data sources and analysis, NOAA Tech. Memo.
NESDIS NGDC 24, National Geophysical Data Center, pp. 19, Boulder, Colorado.
Argus, D. F., R. G. Gordon, and C. DeMets (2011), Geologically current motion of 56 plates relative to the no-net-rotation reference frame,
Geochem. Geophys. Geosyst.,12, Q11001, doi:10.1029/2011GC003751.
Bevis, M., and S. J. Martel (2001), Oblique plate convergence and interseismic strain accumulation, Geochem. Geophys. Geosyst.,2(8), 1033,
doi:10.1029/2000GC000125.
Bird, P. (2003), An updated digital model of plate boundaries, Geochem. Geophys. Geosyst.,4(3), 1027, doi:10.1029/2001GC000252.
Calmant, S., B. Pelletier, P. Lebellegard, M. Bevis, F. W. Taylor, and D. A. Phillips (2003), New insights on the tectonics along the New Hebri-
des subduction zone based on GPS results, J. Geophys. Res.,108(B6), 2319, doi:10.1029/2001JB000644.
Camacho, E., W. Hutton, and J. F. Pacheco (2010), A new look at evidence for a Wadati-Benioff zone and active convergence at the North
Panama Deformed Belt, Bull. Seismol. Soc. Am.,100(1), 343–348, doi:10.1785/0120090204.
Cloos, M. (1993), Lithospheric buoyancy and collisional orogenesis: Subduction of oceanic plateaus, continental margins, island arcs,
spreading ridges and seamounts, Geol. Soc. Am. Bull.,105, 715–737, doi:10.1130/0016-7606(1993)105<0715:LBACOS>2.3.CO;2.
Correa-Mora, F., C. DeMets, D. Alvarado, H. L. Turner, G. Mattioli, D. Hernandez, C. Pullinger, M. Rodriguez, and C. Tenorio (2009), GPS-
derived coupling estimates for the Central America subduction zone and volcanic arc faults: El Salvador, Honduras and Nicaragua, Geo-
phys. J. Int.,179(3), 1279–1291, doi:10.1111/j.1365-246X.2009.04371.x.
Corti, G., E. Carminati, F. Mazzarini, and M. O. Garcia (2005), Active strike-slip faulting in El Salvador, Central America, Geology,33, 989–992,
doi:10.1130/G21992.1.
Acknowledgments
This study was supported by NSF
award (EAR-0955560) to P.L. We thank
reviewer J. Freymueller for his
constructive comments on our
manuscript, R. McCaffrey for providing
DEFNODE, and C. Ammon and D.
Fisher for useful suggestions. We
acknowledge the many people who
helped with GPS data collection. This
material is based on services provided
by the UNAVCO Facility with support
from the National Science Foundation
(NSF) and National Aeronautics and
Space Administration (NASA) under
NSF Cooperative Agreement EAR-
0735156. The figures were produced
using the Generic Mapping Tools
(GMT) software.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1681
DeMets, C., R. G. Gordon, and D. F. Argus (2010), Geologically current plate motions, Geophys. J. Int.,181, 1–80, doi:10.1111/j.1365-
246X.2009.04491.x.
Dziewonski, A. M., T.-A. Chou, and J. H. Woodhouse (1981), Determination of earthquake source parameters from waveform data for stud-
ies of global and regional seismicity, J. Geophys. Res.,86(B4), 2825–2852, doi:10.1029/JB086iB04p02825.
Egbue, O., and J. Kellogg (2010), Pleistocene to present North Andean ‘‘escape’’, Tectonophysics,489, 248–257, doi:10.1016/
j.tecto.2010.04.021.
Farris, D. W., et al. (2011), Fracturing of the Panamanian Isthmus during initial collision with South America, Geology,39(11), 1007–1010,
doi:10.1130/G32237.1.
Fisher, D. M., T. W. Gardner, P. B. Sak, J. D. Sanchez, K. Murphy, and P. Vannucchi (2004), Active thrusting in the inner fore arc of an erosive
convergent margin, Pacific coast, Costa Rica, Tectonics,23, TC2007, doi:10.1029/2002TC001464.
Geirsson, H., T.
Arnad
ottir, C. V€
olksen, W. Jiang, E. Sturkell, T. Villemin, P. Einarsson, F. Sigmundsson, and R. Stef
ansson (2006), Current plate
movements across the Mid-Atlantic Ridge determined from 5 years of continuous GPS measurements in Iceland, J. Geophys. Res.,111,
B09407, doi:10.1029/2005JB003717.
Geirsson, H., P. LaFemina, C. DeMets, G. S. Mattioli, and D. A. Hernandez (2013), Co-seismic deformation of the August 27, 2012 Mw 7.3 El
Salvador and September 5, 2012 Mw 7.6 Costa Rica Earthquakes, Abstract T51A-03 presented at Meeting of the Americas, AGU, Cancun,
Mexico, 14–17 May.
Gutscher, M.-A., J. Malavieille, S. Lallemand, and J.-Y. Collot (1999), Tectonic segmentation of the North Andean margin: Impact of the Car-
negie Ridge collision, Earth Planet. Sci. Lett.,168(3–4), 255–270, doi:10.1016/S0012-821X(99)00060-6.
Hyndman, R. D., and K. Wang (1993), Thermal constraints on the zone of major thrust earthquake failure: The Cascadia subduction zone, J.
Geophys. Res.,98(B2), 2039–2060, doi:10.1029/92JB02279.
Kellogg, J. N., and V. Vega (1995), Tectonic development of Panama, Costa Rica, and the Colombian Andes: Constraints from Global Posi-
tioning System geodetic studies and gravity, in Geologic and Tectonic Development of the Caribbean Plate Boundary in Southern Central
America, Spec. Pap. Geol. Soc. Am., 295, edited by P. Mann, pp. 75–90, Boulder, CO.
La Femina, P. C., T. H. Dixon, and W. Strauch (2002), Bookshelf faulting in Nicaragua, Geology,30, 751–754, doi:10.1130/0091-
7613(2002)030<0751:BFIN>2.0.CO;2.
LaFemina, P., T. H. Dixon, R. Govers, E. Norabuena, H. Turner, A. Saballos, G. Mattioli, M. Protti, and W. Strauch (2009), Fore-arc motion and
Cocos Ridge collision in Central America, Geochem. Geophys. Geosyst.,10, Q05S14, doi:10.1029/2008GC002181.
Lewis, J. C., A. C. Boozer, A. L
opez, and W. Montero (2008), Collision versus sliver transport in the hanging wall at the Middle America sub-
duction zone: Constraints from background seismicity in central Costa Rica, Geochem. Geophys. Geosyst.,9, Q07S06, doi:10.1029/
2007GC001711.
Lundgren, P. R., S. Kornreich Wolf, M. Protti, and K. J. Hurst (1993), GPS measurements of crustal deformation associated with the 22 April
1991, Valle de la Estrella, Costa Rica Earthquake, Geophys. Res. Lett.,20(5), 407–410, doi:10.1029/9 3GL00294.
Lundgren, P., M. Protti, A. Donnellan, M. Heflin, E. Hernandez, and D. Jefferson (1999), Seismic cycle and plate margin deformation in Costa
Rica: GPS observations from 1994 to 1997, J. Geophys. Res.,104(B12), 28,915–28,926, doi:10.1029/1999JB900283.
Lyon-Caen, H., et al. (2006), Kinematics of the North American-Caribbean-Cocos plates in Central America from new GPS measurements
across the Polochic-Motagua fault system, Geophys. Res. Lett.,33, L19309, doi:10.1029/2006GL027694.
MacMillan, I., P. B. Gans, and G. Alvarado (2004), Middle Miocene to present plate tectonic history of the southern Central American Vol-
canic Arc, Tectonophysics,392, 325–348, doi:10.1016/j.tecto.2004.04.014.
Mann, P., and J. Corrigan (1990), Model for late Neogene deformation in Panama, Geology,18(6), 558–562, doi:10.1130/0091–
7613(1990)018<0558:MFLNDI>2.3.CO;2.
Marshall, J. S., D. M. Fisher, and T. W. Gardner (2000), Central Costa Rica deformed belt: Kinematics of diffuse faulting across the western
Panama Block, Tectonics,19(3), 468–497, doi:10.1029/1999TC001136.
Marshall, J. S., B. D. Idleman, T. W. Gardner, and D. M. Fisher (2003), Landscape evolution within a retreating volcanic arc, Costa Rica, Central
America, Geology,31(5), 419–422, doi:10.1130/0091-7613(2003)031<0419:LEWARV>2.0.CO;2.
McCaffrey, R. (2002), Crustal block rotations and plate coupling, in Plate Boundary Zones, Geodyn. Ser., vol. 30, edited by S. Stein and J. Frey-
mueller, pp. 101–122, AGU, Washington, D. C.
McCaffrey, R., P. C. Zwick, Y. Bock, L. Prawirodirdjo, J. F. Genrich, C. W. Stevens, S. S. O. Puntodewo, and C. Subarya (2000), Strain partitioning
during oblique plate convergence in northern Sumatra: Geodetic and seismologic constraints and numerical modeling, J. Geophys. Res.,
105(B12), 28,363–28,376, doi:10.1029/1999JB900362.
Molnar, P., and P. Tapponnier (1975), Cenozoic tectonics of Asia: Effects of a continental collision, Science,189, 419–426, doi:10.1126/
science.189.4201.419.
Mora, H., N. Acero, G. Mart
ınez, J. Ram
ırez, and S. Cardozo (2011), Proyecto GEORED, Informe de actividades 2011, Reporte Interno, Sevicio
Geol
ogico Colombiano, Bogot
a.
Morell, K. D., D. M. Fisher, T. W. Gardner, P. La Femina, D. Davidson, and A. Teletzke (2011), Quaternary outer fore-arc deformation and uplift
inboard of the Panama Triple Junction, Burica Peninsula, J. Geophys. Res.,116, B05402, doi:10.1029/2010JB007979.
National Earthquake Information Center (NEIC) (2012), United States Geological Survey, Denver, CO [Available at http://earthquake.usgs.
gov/earthquakes/eqinthenews/2012/usc000cfsd/.]
Norabuena, E., et al. (2004), Geodetic and seismic constraints on someseismogenic zone processes in Costa Rica, J. Geophys. Res.,109,
B11403, doi:10.1029/2003JB002931.
Pacheco, J. F., and L. R. Sykes (1992), Seismic moment catalog of large, shallow earthquakes, 1900–1989, Bull. Seismol. Soc. Am.,82, 1306–
1349.
Protti, M., F. G€
uendel, and E. Malavassi (2001), Evaluaci
on del Potencial S
ısmico de la Pen
ınsula de Nicoya, Editorial Fundacion UNA, p. 143,
Heredia, Costa Rica.
Protti, M., V. Gonz
alez, A. V. Newman, T. H. Dixon, S. Y. Schwartz, J. S. Marshall, L. Feng, J. I. Walter, R. Malservisi, and S. E. Owen (2013), Nic-
oya earthquake rupture anticipated by geodetic measurement of the locked plate interface, Nat. Geosci.,7(2), 117–121, doi:10.1038/
ngeo2038.
Rockwell, T. K., R. A. Bennett, E. Gath, and P. Franceschi (2010), Unhinging an indenter: A new tectonic model for the internal deformation
of Panama, Tectonics,29, TC4027, doi:10.1029/2009TC002571.
Sak, P. B., D. M. Fisher, T. W. Gardner, J. S. Marshall, and P. C. LaFemina (2009), Rough crust subduction, forearc kinematics, and Quaternary
uplift rates, Costa Rican segment of the Middle American Trench, Geol. Soc. Am. Bull.,121(7/8), 992–1012, doi:10.1130/B26237.1.
Scholz, C. H., and C. Small (1997), The effect of seamount subduction on seismic coupling, Geology,25(6), 487–490, doi:10.1130/0091-
7613(1997)025<0487:TEOSSO>2.3.CO;2.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1682
Silver, E. A., D. L. Reed, J. E. Tagudin, and D. J. Heil (1990), Implications of the North and South Panama thrust belts for the origin of the Pan-
ama orocline, Tectonics,9(2), 261–281, doi:10.1029/TC009i002p00261.
Sitchler, J. C., D. M. Fisher, T. W. Gardner, and J. M. Protti (2007), Constraints on inner forearc deformation from balanced cross sections, Fila
Coste~
na thrust belt, Costa Rica, Tectonics,26, TC6012, doi:10.1029/2006TC001949.
Stein, S., and R. G. Gordon (1984), Statistical tests of additional plate boundaries from plate motion inversions, Earth Planet. Sci. Lett.,69(2),
401–412, doi:10.1016/0012-821X(84)90198-5.
Su
arez, G., M. Pardo, J. Dom
ıguez, L. Ponce, W. Montero, I. Boschini, and W. Rojas (1995), The Lim
on, Costa Rica earthquake of April 22,
1991: Back arc thrusting and collisional tectonics in a subduction environment, Tectonics,14(2), 518–530, doi:10.1029/94TC02546.
Taylor, F. W., et al. (1995), Geodetic measurements of convergence at the New Hebrides island arc indicate arc fragmentation caused by an
impinging aseismic ridge, Geology,23, 1011–1014, doi:10.1130/0091-7613(1995)023<1011:GMOCAT>2.3.CO;2.
Trenkamp, R., J. N. Kellogg, J. T. Freymueller, and H. P. Mora (2002), Wide plate margin deformation, southern Central America and north-
western South America, CASA GPS observations, J. South Am. Earth Sci.,15(2), 157–171, doi:10.1016/S0895-9811(02)00018-4.
Turner, H. L. I., P. LaFemina, A. Saballos, G. S. Mattioli, P. E. Jansma, and T. Dixon (2007), Kinematics of the Nicaraguan forearc from GPS
geodesy, Geophys. Res. Lett.,34, L02302, doi:10.1029/2006GL027586.
van Benthem, S., and R. Govers (2010), The Caribbean plate: Pulled, pushed, or dragged?, J. Geophys. Res.,115, B10409, doi:10.1029/
2009JB006950.
Vergara Mu~
noz, A. (1988), Tectonic patterns of the Panama Block deduced from seismicity, gravitational data and earthquake mechanisms:
Implications to the seismic hazard, Tectonophysics,154(3–4), 253–267, doi:10.1016/0040-1951(88)90107-2.
Vogt, P. R., A. Lowrie, D. R. Bracey, and R. N. Hey (1976), Subduction of aseismic oceanic ridges: Effects on shape, seismicity, and other char-
acteristics of consuming plate boundaries, Spec. Pap. Geol. Soc. Am.,172, 1–59, doi:10.1130/SPE172-p1.
Wallace, L. M., R. McCaffrey, J. Beavan, and S. Ellis (2005), Rapid microplate rotations and back-arc rifting at the transition between collision
and subduction, Geology,33(11), 857–860, doi:10.1130/G21834.1.
Wallace, L. M., S. Ellis, and P. Mann (2009), Collisional model for rapid fore-arc block rotations, arc curvature, and episodic back-arc rifting in
subduction settings, Geochem. Geophys. Geosyst.,10, Q05001, doi:10.1029/2008GC002220.
Walther, C. H. E. (2003), The crustal structure of the Cocos ridge off Costa Rica, J. Geophys. Res.,108(B3), 2136, doi:10.1029/2001JB000888.
Wang, K., R. Wells, S. Mazzotti, R. D. Hyndman, and T. Sagiya (2003), A revised dislocation model of interseismic deformation of the Casca-
dia subduction zone, J. Geophys. Res.,108(B1), 2026, doi:10.1029/2001JB001227.
Yue, H., T. Lay, S. Y. Schwartz, L. Rivera, M. Protti, T. H. Dixon, S. Owen, and A. V. Newman (2013), The 5 September 2012 Nicoya, Costa Rica
Mw 7.6 earthquake rupture process from joint inversion of high-rate GPS, strong-motion, and teleseismic P wave data and its relation-
ship to adjacent plate boundary interface properties, J. Geophys. Res.,118, 5453–5466, doi:10.1002/jgrb.50379.
Zumberge, J. F., M. B. Heflin, D. C. Jefferson, M. M. Watkins, and F. H. Webb (1997), Precise point positioning for the efficient and robust
analysis of GPS data from large networks, J. Geophys. Res.,102(B3), 5005–5017, doi:10.1029/96JB03860.
Geochemistry, Geophysics, Geosystems 10.1002/2014GC005234
KOBAYASHI ET AL. V
C2014. American Geophysical Union. All Rights Reserved. 1683
