Geophys J Int (2018) 0 1ndash0 doi 101093gjiggy230

GJI Geodynamics and tectonics

On the interaction of the North Andes plate with the Caribbean andSouth American plates in northwestern South America from GPSgeodesy and seismic data

Omar J Perez1 Steven G Wesnousky2 Roberto De La Rosa1 Julio Marquez1

Redescal Uzcategui1 Christian Quintero1 Luis Liberal3 Hector Mora-Paez4 andWalter Szeliga5

1Department of Earth Sciences Simon Bolıvar University Caracas 1080A Venezuela E-mail ojperezusbve2Center for Neotectonic Studies University of Nevada Reno NV 89557 USA3Department of Geodesy Central University of Venezuela Caracas 1050 Venezuela4Sptial Geodetic Research Group Colombian Geological Survey Bogota 111321 Colombia5Department of Geological Sciences Central Washington University Ellensburg WA 98926 USA

Accepted 2018 June 8 Received 2018 March 27 in original form 2017 April 12

S U M M A R YWe examine the hypocentral distribution of seismicity and a series of geodetic velocity vectorsobtained from Global Positioning System observations between 1994 and 2015 both off-shore and mainland northwestern South America (66Wndash77W 8Nndash14N) Our analysisthat includes a kinematic block modelling shows that east of the CaribbeanndashSouth AmericanndashNorth Andes platesprime triple junction at sim68W 107N right-lateral easterly oriented shearmotion (sim196 plusmn 20 mm yrminus1) between the Caribbean and South America plates is split alongtwo easterly striking right-lateral strike-slip subparallel fault zones the San Sebastian faultthat runs off-shore the Venezuelan coast and slips about 170 plusmn 05 mm yrminus1 and the La Victoriafault located on-shore to the south which is accumulating strain equivalent to 26 plusmn 04 mmyrminus1 West of the triple junction relative right-lateral motion between the Caribbean and SouthAmerican plates is mostly divided between the Morrocoy and Bocono fault systems that strikenorthwest and southwest from the triple junction respectively and bound the intervening NorthAndes plate that shows an easterly oriented geodetic slip of 150 plusmn 10 mm yrminus1 relative to theSouth American plate Slip on the Morrocoy fault is right-lateral and transtensional Motionacross the Bocono fault is also right-lateral but instead transpressional divided between sim9and 11 mm yrminus1 of right-slip on the Bocono fault and 2ndash5 mm yrminus1 of convergence acrossadjacent and subparallel thrust faults Farther west of the triple junction sim800 km away innorthern Colombia the Caribbean plate subducts to the southeast beneath the North Andesplate at a geodetically estimated rate of sim5ndash7 mm yrminus1

Key words Plate motions Space geodetic surveys Earthquake hazards Seismicity andtectonics Dynamics seismotectonics Kinematics of crustal and mantle deformation

I N T RO D U C T I O N

The purpose of this paper is to combine Global Positioning System(GPS) geodesy-derived velocity vectors and seismological data toassess the kinematics of the relative motion between the CaribbeanSouth American and North Andes plates in the northwestern tipof South America (Fig 1) a highly populated region that includesthe Venezuelan capital city of Caracas and other major towns inColombia including a population of sim16 million people (INE 2011DANE 2016)

GPS geodetic investigations have shown that the Caribbean plateslips easterly at a rate of sim196 plusmn 2 mm yrminus1(Perez et al 2001Weber et al 2001) relative to the South American plate Under-standing of how this motion is distributed is complicated by thepresence of the CaribbeanndashNorth AndeanndashSouth American triplejunction (Fig 1) proposed by Perez et al 1997a Perez amp Mendoza1998 and Bird 2003 and for this reason has remained somewhatelusive This type of junction formed by three strike-slip faults isunstable and it is required that the faults have the same sense ofrelative motion (McKenzie amp Morgan 1969) To the west of thetriple junction complexity arises because relative motion between

Ccopy The Author(s) 2018 Published by Oxford University Press on behalf of The Royal Astronomical Society This is an Open Accessarticle distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby40) whichpermits unrestricted reuse distribution and reproduction in any medium provided the original work is properly cited 1

2 OJ Perez et al

Figure 1 Tectonic setting of the southern Caribbean and adjacent plates (compiled from Silver et al 1975 Perez amp Aggarwal 1981 Kellogg amp Vega 1995Perez et al 1997b Trenkamp et al 2002 Bird 2003 Gorney et al 2007 Magnani et al 2009 Weber et al 2009 Camacho et al 2010 Escalona amp Mann2011 Kroehler et al 2011) Main plate boundaries are shown by thick lines Oceanic segments with teeth indicate subduction with teeth on overthrust plateThe North Andes plate (Trenkamp et al 2002 Bird 2003) is addressed as a block by several authors (Pennington 1981 1995 Kellogg et al 1985) Platemotions (arrows) are relative to stable South America (DeMets et al 1990 Perez et al 2001 Weber et al 2001 Trenkamp et al 2002 Perez et al 2011)

Figure 2 Historical large (Mw 7+) earthquake ruptures (thick lines) in northndashcentral Venezuela since AD 1640 Numerals indicate year of occurrence Blackdots are events of Mw 6+ Triangles are GPS sites visited at various epochs since 1994 f fault Focal mechanisms (lower hemisphere equal area projection)are for events with Mw 6+ since 1950 (Rial 1978 Perez et al 1997a Choy 2001 Global CMT Catalog) They show the dextral dominant nature of the BoconoMorrocoy and San Sebastian faults

the Caribbean and South American plates is separated by the NorthAndes plate and divided as we will show in this paper between themajor northwest striking Morrocoy and southwest striking Boconofault systems respectively (Fig 1 Perez et al 1997a Perez amp Men-doza 1998 Bird 2003) with some motion possibly taken up by otherfaults such as the Oca and Santa Marta faults (Symithe et al 2015)Only along the San SebastianndashEl Pilar fault system does motionoccur directly between the Caribbean and South American platesYet even here the fault system accommodating the motion and theGPS velocity field reflecting that motion is distributed across an 80ndash120 km wide zone (Perez et al 2001 Jouanne et al 2011 Reinozaet al 2015) including other faults such as the La Victoria fault

A number of large earthquakes have occurred on the mainbranches of the triple junction since 1640 (Fig 2) These includea sequence of two events in 1812 (Mw 74 and 71) that took placewithin about 1 hr and broke across the triple junction to rupture por-tions of both the Bocono and San Sebastian faults (Fielder 1961Altez 1998 Choy et al 2010) In 1900 an Mw 76 shock (Pacheco ampSykes 1992) broke the eastern segment of the San Sebastian fault asrecognized from both intensity observations (Centeno-Grau 1900Sievers 1905 Choy et al 2010) and marine geophysics studies(Colon et al 2015) A number of moderate (Mw sim6+) yet destruc-tive dextral strike-slip shocks are documented (Fig 2) along theBocono fault in 1950 and 1975 as well as the San Sebastian fault

GPS and plate kinematics in south Caribbean 3

Figure 3 Epicentral locations (circles) of seismicity (Mw ge 25) reported by the Venezuelan National Seismological Array (FUNVISIS 2016) in northndashcentralVenezuela and southndashcentral Caribbean from 2003 to 2012 Small-sized circles are microearthquakes of Mw lt 45 Red arrows are GPS velocity vectorsrelative to South America either compiled or derived in this study Ellipses are 2-D errors plusmn 1σ

in 1967 and 2009 the Morrocoy fault in 1989 (Mw 62) and the LaVictoria fault in 1641 and 1878 (Perez et al 1997b) The damagingeffects of the latter events are described in detail by Centeno-Grau(1940) Grases (1980) and Ernst (1878) Focal mechanisms of smallearthquakes along La Victoria fault also show dextral motion alongstrike (Perez et al 1997b) Microseismicity (Mw lt 4) has been re-ported on each of the major branches of the triple junction (Fig 3eg Perez et al 1997b Perez amp Mendoza 1998) No significantearthquake activity is reported along La Tortuga fault (Fig 2) oralong the Oca and Santa Marta faults (Fig 1) since the colonytimes The known seismicity record along the Caribbean subduc-tion in northern Colombia (Fig 1) is limited to a small quantity ofteleseismically located events and microseismicity recorded since1993

To check and quantify the style and sense of motion of the variousactive faults that define the branches of the triple junction and ulti-mately better define the tectonics of northwestern South Americawe here conduct an integrated analysis of a series of new GPS obser-vations collected in northndashcentral and northwestern Venezuela andColombia during the period from 1999 to 2015 and seismologicaldata including well located sets of microearthquakes reported by theColombian (1993ndash2011) and Venezuelan (2003ndash2012) broad-bandnational seismological arrays operated by the Colombian GeologicalSurvey (Colombian Geological Survey 2016) and the VenezuelanFoundation for Seismological Research (FUNVISIS 2016) respec-tively

In the next sections we describe the GPS data acquisition andanalysis methods and the velocity vectors and field subsequentlyobtained Next we present details of the models that were used toemulate the velocity field across the San Sebastian and La Victoriafaults in northndashcentral Venezuela Then we analyse the seismicityand the velocity field observed in western Venezuela which en-compasses the Bocono and Morrocoy faults and northern Colom-bia where the Caribbean plate is being subducted to the southndasheastbeneath the North Andes plate (Fig 1) along the South CaribbeanMarginal fault (Toto amp Kellogg 1992 Perez et al 1997b Trenkampet al 2002) Then we performed a kinematic block modelling forthe southern Caribbean plate boundary to further analyse and quan-tify the regional GPS velocity field and rate and sense of motionalong the major active faults that mark the plate boundary For the

analysis we used the modelling approach and associated softwarelsquoBlocksrsquo of Meade amp Loveless (2009) Finally the implications ofthe geodetic findings for regional tectonics and for the quantificationof seismic hazards in the region are discussed

G P S M E T H O D S A N D A NA LY S I S

GPS data in Venezuela and adjacent regions were obtained usingdual frequency receivers during a series of 5ndash15 d and 12ndash24 hrsurveying sessions between 1994 and March 2015 During thiswhole period the monument CANOA located deep within stableSouth America and Maracaibo (MARA Fig 1) as well as USB0and USB1 (replacing USB0 in 2006 Fig 2) were occupied 7 10 7and 7 times respectively since 1999 sites CASI and ALTA in 19992000 and 2012 CHUR and AMUA in 1994 1999 and 2014 ROQUand CALA in 1994 1999 and 2000 ARUB in 1999 and 2012ROSA and MONA in 2000 and 2014 We also used data collectedsince 2012 at sites CN38 and CN40 which belong to the COCOnetpermanent GPS array as well as data collected in 1994 and 1999 byPerez et al (2001) and Perez et al (2011) to re-calculate severalgeodetic vectors reported in those papers using newly acquired datain 2014 at those sites

In general all Venezuelan sites were occupied simultaneouslywith CANOA MARA (Fig 1) and USB0 (or USB1 Fig 2) or all ofthem We used 15 or 30 s sampling intervals and an elevation maskof 15 Sitersquos positions for each epoch were obtained using simul-taneous observations at 8ndash11 permanent GPS stations of the Inter-national GNSS Service located in the Americas and the Caribbeanandor directly relative to CANOA as a fixed reference point to checkfor consistency We used the Bernese GPS v 50 (Dach et al 2007)software package to process the data following the procedures de-scribed by these and other authors (eg Jouanne et al 2011 Reinozaet al 2015) Additionally the Space Geodesy Research Group thatruns the GeoRED Project in Colombia (Mora-Paez et al 2016ab)provided us with the velocity vectors they obtained for six GPS sitesin Colombia territory that they processed using the GIPSY-OASISII software v 63 The corresponding velocity vectors at all sitesused in this study calculated in ITRF2008 Epoch 2010 (Altamimiet al 2011) are given numerically in Table 1 The velocities werethen calculated in the South America reference frame using the pole

4 OJ Perez et al

Table 1 Velocities in the ITRF2008 reference frame showing east and north components with one sigma error

Site ID Long (E) Lat (N) Ve (mm yrminus1) σ e (mm yrminus1) Vn (mm yrminus1) σ n (mm yrminus1) Ref

ALPA minus72918 11528 87 28 137 09 1ALTA minus66178 9866 minus51 05 92 05 2AMUA minus70188 11753 97 06 127 05 3ARAY minus63711 10655 59 25 129 23 4ARUB minus70013 12574 130 13 111 12 2AVES minus63618 15667 146 27 149 18 4BAPA minus74660 5470 28 04 146 07 5BARB minus59620 13250 129 18 143 16 4BARI minus70381 8657 minus49 04 120 03 6BARU minus75590 10258 114 08 103 05 1BATE minus71110 9170 19 04 150 03 6BLAN minus64599 11822 107 25 150 18 4BUCM minus73182 7117 30 23 135 08 1 7CALA minus67449 9061 minus37 25 105 12 8CANO minus63861 8568 minus56 09 115 06 2 8CAPI minus72430 5350 minus30 05 115 06 5CART minus75500 10360 122 09 105 09 1 7CARU minus63244 10675 74 31 120 25 4CASI minus66960 9926 minus42 05 120 05 2CHUR minus69541 10818 94 07 139 05 8CN35 minus81363 13374 173 minus27 07 06 1CN37 minus75263 10793 90 17 136 21 1CN38 minus71988 12222 115 06 141 04 1CN40 minus68960 12180 125 07 118 07 8COCH minus63994 10782 132 2 126 2 4DALI minus72314 8267 44 06 115 05 6FABI minus61658 10098 minus19 12 126 11 4FRAI minus70799 8850 minus14 04 134 03 6FUND minus71863 7783 minus10 06 103 05 6GOYA minus69708 10995 91 07 137 05 3GREO minus61640 12220 134 05 150 05 8GRIT minus71998 8156 03 03 109 02 6GUAC minus71333 7515 minus46 03 104 03 6GUAY minus72342 8530 67 06 114 05 6JUAN minus63369 10357 minus31 3 120 26 4LUCI minus60970 14090 122 18 141 16 4MARA minus71624 10674 88 08 128 08 2MARG minus6436 11042 138 15 137 25 4MONA minus70460 9550 minus10 06 126 05 2MONT minus75681 8895 113 06 111 04 1 7MZAL minus75470 5030 123 06 179 04 1 7PREG minus71775 8029 minus14 03 109 03 6QUIB minus69783 9914 33 32 119 13 3RION minus75430 6176 51 03 147 02 1 7ROSA minus69774 10166 31 05 130 05 2RQUE minus66678 11953 140 24 134 18 3STOD minus70630 8860 minus29 04 127 03 3SANA minus81720 12580 123 04 70 04 1 7TORO minus71126 9064 minus12 04 105 03 3TORT minus65227 10963 133 2 133 17 3TRIN minus61400 10680 124 14 134 12 9USB0 minus66792 10409 minus11 05 120 05 2USB1 minus66883 10411 minus14 05 121 05 2UWAS minus72390 6450 03 04 130 06 5VBUV minus73860 5530 39 05 156 07 5VDPR minus73248 10436 85 04 153 04 1VPOL minus74861 10794 77 08 166 05 1

1 Mora-Paez et al (2016a) 2 this study 3 Re-evaluated from Perez et al (2001) 4 Perez et al (2001) 5 Mora-Paez et al (2016b) 6 Re-evaluated fromPerez et al (2011) Trenkamp et al (2002) 8 D Mencin unpublished data 2015 9 Estimated from Weber et al (2001)

of rotation between ITRF2008 and South America provided by Al-tamimi et al (2012) They are graphically shown in Figs 3 4 and8 Finally we incorporated seven velocity vectors in the Colombianregion expressed in the South America reference frame originally

provided by Trenkamp et al (2002) which were re-calculated inthe ITRF2008 frame (Mora-Paez et al 2016a)

GPS and plate kinematics in south Caribbean 5

Table 2 Euler pole locations and rotation rates relative to the South Amer-ican plate for the best fitting models described in the text

Blockplate Latitude () Longitude ()Rotation rate (deg

Myrminus1)

(lowast) Caribbean 619 plusmn 20 2843 plusmn 20 0229 plusmn 0015Model 4Block A 252 plusmn 93 2807 plusmn 23 0575 plusmn 0296Block B 51 plusmn 55 1122 plusmn 13 0403 plusmn 0135Block C minus63 plusmn 66 1144 plusmn 21 0339 plusmn 0499Model 5Block A 220 plusmn 62 2813 plusmn 16 0672 plusmn 0293Block B minus23 plusmn 10 1106 plusmn 02 0660 plusmn 0080Block C 130 plusmn 20 2947 plusmn 15 0657 plusmn 0541Block D 03 plusmn 14 1106 plusmn 02 0603 plusmn 0079Block E 213 plusmn 357 1154 plusmn 50 0341 plusmn 0322

(lowast) From Perez et al 2001

R E S U LT S

Slip-partitioning associated with the San Sebastian and LaVictoria fault zones

In this section we combine GPS data with simple elastic strainaccumulation models and seismic evidence to constrain slip rate es-timates along these two major dextral easterly striking fault zonesThe two faults (Fig 3) form the CaribbeanndashSouth American mainplate boundary east of sim68W Our goal is to quantify how thesim196 mm yrminus1 in the N85E direction of CaribbeanndashSouth Amer-ican rate of relative motion indicated by the corresponding pole ofrotation (Table 2) derived from GPS geodesy (Perez et al 2001)is distributed between these fault zones N85E is also the strikedirection of the San Sebastian fault

Fig 3 shows the seismicity (moment magnitude Mw ge 26 cir-cles) reported by FUNVISIS (2016) for the period 2003ndash2012 to-gether with the GPS-derived velocity vectors (arrows) relative tostable South America in northndashcentral Venezuela About 985 percent of the seismic events reported are microearthquakes in the Mw

range of 26ndash44 Only minor activity is observed in the inner re-gions of the Caribbean though a series of seismological stationsare located in Venezuelan islands and along the northern coast ofthe country (see httpwwwfunvisisgobve) FUNVISIS catalogshows a fairly constant rate of listing for events with Mw ge 26during the time span under consideration

East of sim68W the San Sebastian and La Victoria faults co-exist These fault zones are nearly parallel (Fig 3) making anangle of sim5ndash10 and are separated by a distance that ranges fromsim30ndash50 km Recent marine geophysics studies show that at someplaces the San Sebastian fault is multibranched and the fault zonereaches a width of 10ndash12 km (Colon et al 2015) As found outby Perez et al (1997b) from a different set of earlier earthquakesmost of the seismic activity in the region straddles both fault zonesoccurs in general at focal depths shallower than sim14 km and focalmechanisms commonly show dextral motion on easterly orientedvertical fault planes (Figs 2 and 4) These observations show thatthese two fault zones accommodate most of the relative motionbetween the Caribbean and South American plates in the region

Fig 5 is a stacked profile showing the N85E velocity componentof GPS sites located to the north and south of the San Sebastianand La Victoria faults The stations used in the stack are marked inFig 4 with blue stars Many benchmarks were vandalized since thefirst occupations in 1999 and thus the number of data points remainsrelatively few Nonetheless the velocities of USB0 and USB1 are

very well constrained since all together they have been occupied 14times since 1999 and their location between the two faults providesstrong constraints on the slip partitioning we next describe

The field (Fig 5) across these faults is confined to a shear zonelt120 km wide and most of the surface deformation is containedwithin a 70ndash80 km wide zone encompassing the San Sebastianand La Victoria faults The dashed line in Fig 5 is the elasticmodelled surface velocities (Okada 1985 Dixon et al 1998) for asingle N85E oriented vertical fault taken up the full CaribbeanndashSouth American relative motion (sim196 mm yrminus1) The solid lineis in contrast calculated for a system of two subparallel verticalfaults (ie The La Victoria and San Sebastian faults) separated by adistance of 45 km in both cases with dextral shear below a lockingdepth of 14 Km The two-fault model provides a good fit for theobserved data and corresponds to a dextral shear of 26 plusmn 04 mmyrminus1 for La Victoria fault and 170 plusmn 05 mm yrminus1 for San Sebastianfault below a locking depth of 14 plusmn 3 km

The modelling result does not preclude the possibility of someportion of slip modelled on the San Sebastian fault being accom-modated on other faults to the north such as the easterly striking LaTortuga fault (Fig 2) One could for example allow up to sim3 mmyrminus1 of slip along La Tortuga fault by decreasing a similar amountof modelled slip on the San Sebastian fault without violating theobservations We hesitate to argue for this latter possibility becauseboth the seismic record and marine geophysics studies (Colon et al2015) do not support the existence of the La Tortuga fault alongthe entire northndashcentral coast of Venezuela Similarly it is possiblethat the plate boundary may include a third fault farther north alongthe easterly striking section of the South Caribbean Deformed Beltin the off-shore area (Fig 1) with a small and probably negligi-ble amount of slip on that portion of the structure (Symithe et al2015) expressed by an actively deforming accretionary prism (Gor-ney et al 2007 Clark et al 2008 Magnani et al 2009 Escalonaamp Mann 2011 Kroehler et al 2011) Finally based on GPS in-vestigations there is general agreement (Perez et al 2001 Clarket al 2008 Jouanne et al 2011 Reinoza et al 2015 Symithe et al2015) in that east of sim655W in eastern Venezuela (Figs 1 and 2)the easterly striking El Pilar fault zone is the main strand of theCaribbeanndashSouth America plate boundary taking up most of therelative motion between the two plates In our interpretation boththe San Sebastian and La Victoria faults merge around 655W withthe El Pilar fault zone so that their motion is absorbed into the singleEl Pilar strand

The interaction of the North Andes plate with theCaribbean and the South American plates in westernVenezuela and northern Colombia

Numerous authors (eg Dewey 1972 Pennington 1981 Kellogg ampBonini 1982 Kellogg 1984 Toto amp Kellogg 1992 van der Hilst ampMann 1994 Malave amp Suarez 1995 Perez et al 1997b Perez ampMendoza 1998 Trenkamp et al 2002 Perez et al 2011 Bernal-Olaya et al 2015 Chiarabba et al 2015 Syracuse et al 2016Mora-Paez et al 2016b) have studied the seismicity and tectonicsof northwestern South America There is a general agreement thatalong the northwestern tip of the continent there exists a northeast-trending slab of Caribbean plate that is subducting along the SouthCaribbean Marginal fault to the southeast (Figs 1 and 6) beneath theNorth Andes plate (Toto amp Kellogg 1992 Trenkamp et al 2002Bird 2003) Here we shall re-visit the seismicity of the region usingteleseismic locations (mb ge 45) reported in the revised catalogues

GPS and plate kinematics in south Caribbean 3

Figure 3 Epicentral locations (circles) of seismicity (Mw ge 25) reported by the Venezuelan National Seismological Array (FUNVISIS 2016) in northndashcentralVenezuela and southndashcentral Caribbean from 2003 to 2012 Small-sized circles are microearthquakes of Mw lt 45 Red arrows are GPS velocity vectorsrelative to South America either compiled or derived in this study Ellipses are 2-D errors plusmn 1σ

in 1967 and 2009 the Morrocoy fault in 1989 (Mw 62) and the LaVictoria fault in 1641 and 1878 (Perez et al 1997b) The damagingeffects of the latter events are described in detail by Centeno-Grau(1940) Grases (1980) and Ernst (1878) Focal mechanisms of smallearthquakes along La Victoria fault also show dextral motion alongstrike (Perez et al 1997b) Microseismicity (Mw lt 4) has been re-ported on each of the major branches of the triple junction (Fig 3eg Perez et al 1997b Perez amp Mendoza 1998) No significantearthquake activity is reported along La Tortuga fault (Fig 2) oralong the Oca and Santa Marta faults (Fig 1) since the colonytimes The known seismicity record along the Caribbean subduc-tion in northern Colombia (Fig 1) is limited to a small quantity ofteleseismically located events and microseismicity recorded since1993

To check and quantify the style and sense of motion of the variousactive faults that define the branches of the triple junction and ulti-mately better define the tectonics of northwestern South Americawe here conduct an integrated analysis of a series of new GPS obser-vations collected in northndashcentral and northwestern Venezuela andColombia during the period from 1999 to 2015 and seismologicaldata including well located sets of microearthquakes reported by theColombian (1993ndash2011) and Venezuelan (2003ndash2012) broad-bandnational seismological arrays operated by the Colombian GeologicalSurvey (Colombian Geological Survey 2016) and the VenezuelanFoundation for Seismological Research (FUNVISIS 2016) respec-tively

In the next sections we describe the GPS data acquisition andanalysis methods and the velocity vectors and field subsequentlyobtained Next we present details of the models that were used toemulate the velocity field across the San Sebastian and La Victoriafaults in northndashcentral Venezuela Then we analyse the seismicityand the velocity field observed in western Venezuela which en-compasses the Bocono and Morrocoy faults and northern Colom-bia where the Caribbean plate is being subducted to the southndasheastbeneath the North Andes plate (Fig 1) along the South CaribbeanMarginal fault (Toto amp Kellogg 1992 Perez et al 1997b Trenkampet al 2002) Then we performed a kinematic block modelling forthe southern Caribbean plate boundary to further analyse and quan-tify the regional GPS velocity field and rate and sense of motionalong the major active faults that mark the plate boundary For the

analysis we used the modelling approach and associated softwarelsquoBlocksrsquo of Meade amp Loveless (2009) Finally the implications ofthe geodetic findings for regional tectonics and for the quantificationof seismic hazards in the region are discussed

G P S M E T H O D S A N D A NA LY S I S

GPS data in Venezuela and adjacent regions were obtained usingdual frequency receivers during a series of 5ndash15 d and 12ndash24 hrsurveying sessions between 1994 and March 2015 During thiswhole period the monument CANOA located deep within stableSouth America and Maracaibo (MARA Fig 1) as well as USB0and USB1 (replacing USB0 in 2006 Fig 2) were occupied 7 10 7and 7 times respectively since 1999 sites CASI and ALTA in 19992000 and 2012 CHUR and AMUA in 1994 1999 and 2014 ROQUand CALA in 1994 1999 and 2000 ARUB in 1999 and 2012ROSA and MONA in 2000 and 2014 We also used data collectedsince 2012 at sites CN38 and CN40 which belong to the COCOnetpermanent GPS array as well as data collected in 1994 and 1999 byPerez et al (2001) and Perez et al (2011) to re-calculate severalgeodetic vectors reported in those papers using newly acquired datain 2014 at those sites

In general all Venezuelan sites were occupied simultaneouslywith CANOA MARA (Fig 1) and USB0 (or USB1 Fig 2) or all ofthem We used 15 or 30 s sampling intervals and an elevation maskof 15 Sitersquos positions for each epoch were obtained using simul-taneous observations at 8ndash11 permanent GPS stations of the Inter-national GNSS Service located in the Americas and the Caribbeanandor directly relative to CANOA as a fixed reference point to checkfor consistency We used the Bernese GPS v 50 (Dach et al 2007)software package to process the data following the procedures de-scribed by these and other authors (eg Jouanne et al 2011 Reinozaet al 2015) Additionally the Space Geodesy Research Group thatruns the GeoRED Project in Colombia (Mora-Paez et al 2016ab)provided us with the velocity vectors they obtained for six GPS sitesin Colombia territory that they processed using the GIPSY-OASISII software v 63 The corresponding velocity vectors at all sitesused in this study calculated in ITRF2008 Epoch 2010 (Altamimiet al 2011) are given numerically in Table 1 The velocities werethen calculated in the South America reference frame using the pole

4 OJ Perez et al

Table 1 Velocities in the ITRF2008 reference frame showing east and north components with one sigma error

Site ID Long (E) Lat (N) Ve (mm yrminus1) σ e (mm yrminus1) Vn (mm yrminus1) σ n (mm yrminus1) Ref

ALPA minus72918 11528 87 28 137 09 1ALTA minus66178 9866 minus51 05 92 05 2AMUA minus70188 11753 97 06 127 05 3ARAY minus63711 10655 59 25 129 23 4ARUB minus70013 12574 130 13 111 12 2AVES minus63618 15667 146 27 149 18 4BAPA minus74660 5470 28 04 146 07 5BARB minus59620 13250 129 18 143 16 4BARI minus70381 8657 minus49 04 120 03 6BARU minus75590 10258 114 08 103 05 1BATE minus71110 9170 19 04 150 03 6BLAN minus64599 11822 107 25 150 18 4BUCM minus73182 7117 30 23 135 08 1 7CALA minus67449 9061 minus37 25 105 12 8CANO minus63861 8568 minus56 09 115 06 2 8CAPI minus72430 5350 minus30 05 115 06 5CART minus75500 10360 122 09 105 09 1 7CARU minus63244 10675 74 31 120 25 4CASI minus66960 9926 minus42 05 120 05 2CHUR minus69541 10818 94 07 139 05 8CN35 minus81363 13374 173 minus27 07 06 1CN37 minus75263 10793 90 17 136 21 1CN38 minus71988 12222 115 06 141 04 1CN40 minus68960 12180 125 07 118 07 8COCH minus63994 10782 132 2 126 2 4DALI minus72314 8267 44 06 115 05 6FABI minus61658 10098 minus19 12 126 11 4FRAI minus70799 8850 minus14 04 134 03 6FUND minus71863 7783 minus10 06 103 05 6GOYA minus69708 10995 91 07 137 05 3GREO minus61640 12220 134 05 150 05 8GRIT minus71998 8156 03 03 109 02 6GUAC minus71333 7515 minus46 03 104 03 6GUAY minus72342 8530 67 06 114 05 6JUAN minus63369 10357 minus31 3 120 26 4LUCI minus60970 14090 122 18 141 16 4MARA minus71624 10674 88 08 128 08 2MARG minus6436 11042 138 15 137 25 4MONA minus70460 9550 minus10 06 126 05 2MONT minus75681 8895 113 06 111 04 1 7MZAL minus75470 5030 123 06 179 04 1 7PREG minus71775 8029 minus14 03 109 03 6QUIB minus69783 9914 33 32 119 13 3RION minus75430 6176 51 03 147 02 1 7ROSA minus69774 10166 31 05 130 05 2RQUE minus66678 11953 140 24 134 18 3STOD minus70630 8860 minus29 04 127 03 3SANA minus81720 12580 123 04 70 04 1 7TORO minus71126 9064 minus12 04 105 03 3TORT minus65227 10963 133 2 133 17 3TRIN minus61400 10680 124 14 134 12 9USB0 minus66792 10409 minus11 05 120 05 2USB1 minus66883 10411 minus14 05 121 05 2UWAS minus72390 6450 03 04 130 06 5VBUV minus73860 5530 39 05 156 07 5VDPR minus73248 10436 85 04 153 04 1VPOL minus74861 10794 77 08 166 05 1

1 Mora-Paez et al (2016a) 2 this study 3 Re-evaluated from Perez et al (2001) 4 Perez et al (2001) 5 Mora-Paez et al (2016b) 6 Re-evaluated fromPerez et al (2011) Trenkamp et al (2002) 8 D Mencin unpublished data 2015 9 Estimated from Weber et al (2001)

of rotation between ITRF2008 and South America provided by Al-tamimi et al (2012) They are graphically shown in Figs 3 4 and8 Finally we incorporated seven velocity vectors in the Colombianregion expressed in the South America reference frame originally

provided by Trenkamp et al (2002) which were re-calculated inthe ITRF2008 frame (Mora-Paez et al 2016a)

GPS and plate kinematics in south Caribbean 5

Table 2 Euler pole locations and rotation rates relative to the South Amer-ican plate for the best fitting models described in the text

Blockplate Latitude () Longitude ()Rotation rate (deg

Myrminus1)

(lowast) Caribbean 619 plusmn 20 2843 plusmn 20 0229 plusmn 0015Model 4Block A 252 plusmn 93 2807 plusmn 23 0575 plusmn 0296Block B 51 plusmn 55 1122 plusmn 13 0403 plusmn 0135Block C minus63 plusmn 66 1144 plusmn 21 0339 plusmn 0499Model 5Block A 220 plusmn 62 2813 plusmn 16 0672 plusmn 0293Block B minus23 plusmn 10 1106 plusmn 02 0660 plusmn 0080Block C 130 plusmn 20 2947 plusmn 15 0657 plusmn 0541Block D 03 plusmn 14 1106 plusmn 02 0603 plusmn 0079Block E 213 plusmn 357 1154 plusmn 50 0341 plusmn 0322

(lowast) From Perez et al 2001

R E S U LT S

Slip-partitioning associated with the San Sebastian and LaVictoria fault zones

In this section we combine GPS data with simple elastic strainaccumulation models and seismic evidence to constrain slip rate es-timates along these two major dextral easterly striking fault zonesThe two faults (Fig 3) form the CaribbeanndashSouth American mainplate boundary east of sim68W Our goal is to quantify how thesim196 mm yrminus1 in the N85E direction of CaribbeanndashSouth Amer-ican rate of relative motion indicated by the corresponding pole ofrotation (Table 2) derived from GPS geodesy (Perez et al 2001)is distributed between these fault zones N85E is also the strikedirection of the San Sebastian fault

Fig 3 shows the seismicity (moment magnitude Mw ge 26 cir-cles) reported by FUNVISIS (2016) for the period 2003ndash2012 to-gether with the GPS-derived velocity vectors (arrows) relative tostable South America in northndashcentral Venezuela About 985 percent of the seismic events reported are microearthquakes in the Mw

range of 26ndash44 Only minor activity is observed in the inner re-gions of the Caribbean though a series of seismological stationsare located in Venezuelan islands and along the northern coast ofthe country (see httpwwwfunvisisgobve) FUNVISIS catalogshows a fairly constant rate of listing for events with Mw ge 26during the time span under consideration

East of sim68W the San Sebastian and La Victoria faults co-exist These fault zones are nearly parallel (Fig 3) making anangle of sim5ndash10 and are separated by a distance that ranges fromsim30ndash50 km Recent marine geophysics studies show that at someplaces the San Sebastian fault is multibranched and the fault zonereaches a width of 10ndash12 km (Colon et al 2015) As found outby Perez et al (1997b) from a different set of earlier earthquakesmost of the seismic activity in the region straddles both fault zonesoccurs in general at focal depths shallower than sim14 km and focalmechanisms commonly show dextral motion on easterly orientedvertical fault planes (Figs 2 and 4) These observations show thatthese two fault zones accommodate most of the relative motionbetween the Caribbean and South American plates in the region

Fig 5 is a stacked profile showing the N85E velocity componentof GPS sites located to the north and south of the San Sebastianand La Victoria faults The stations used in the stack are marked inFig 4 with blue stars Many benchmarks were vandalized since thefirst occupations in 1999 and thus the number of data points remainsrelatively few Nonetheless the velocities of USB0 and USB1 are

very well constrained since all together they have been occupied 14times since 1999 and their location between the two faults providesstrong constraints on the slip partitioning we next describe

The field (Fig 5) across these faults is confined to a shear zonelt120 km wide and most of the surface deformation is containedwithin a 70ndash80 km wide zone encompassing the San Sebastianand La Victoria faults The dashed line in Fig 5 is the elasticmodelled surface velocities (Okada 1985 Dixon et al 1998) for asingle N85E oriented vertical fault taken up the full CaribbeanndashSouth American relative motion (sim196 mm yrminus1) The solid lineis in contrast calculated for a system of two subparallel verticalfaults (ie The La Victoria and San Sebastian faults) separated by adistance of 45 km in both cases with dextral shear below a lockingdepth of 14 Km The two-fault model provides a good fit for theobserved data and corresponds to a dextral shear of 26 plusmn 04 mmyrminus1 for La Victoria fault and 170 plusmn 05 mm yrminus1 for San Sebastianfault below a locking depth of 14 plusmn 3 km

The modelling result does not preclude the possibility of someportion of slip modelled on the San Sebastian fault being accom-modated on other faults to the north such as the easterly striking LaTortuga fault (Fig 2) One could for example allow up to sim3 mmyrminus1 of slip along La Tortuga fault by decreasing a similar amountof modelled slip on the San Sebastian fault without violating theobservations We hesitate to argue for this latter possibility becauseboth the seismic record and marine geophysics studies (Colon et al2015) do not support the existence of the La Tortuga fault alongthe entire northndashcentral coast of Venezuela Similarly it is possiblethat the plate boundary may include a third fault farther north alongthe easterly striking section of the South Caribbean Deformed Beltin the off-shore area (Fig 1) with a small and probably negligi-ble amount of slip on that portion of the structure (Symithe et al2015) expressed by an actively deforming accretionary prism (Gor-ney et al 2007 Clark et al 2008 Magnani et al 2009 Escalonaamp Mann 2011 Kroehler et al 2011) Finally based on GPS in-vestigations there is general agreement (Perez et al 2001 Clarket al 2008 Jouanne et al 2011 Reinoza et al 2015 Symithe et al2015) in that east of sim655W in eastern Venezuela (Figs 1 and 2)the easterly striking El Pilar fault zone is the main strand of theCaribbeanndashSouth America plate boundary taking up most of therelative motion between the two plates In our interpretation boththe San Sebastian and La Victoria faults merge around 655W withthe El Pilar fault zone so that their motion is absorbed into the singleEl Pilar strand

The interaction of the North Andes plate with theCaribbean and the South American plates in westernVenezuela and northern Colombia

Numerous authors (eg Dewey 1972 Pennington 1981 Kellogg ampBonini 1982 Kellogg 1984 Toto amp Kellogg 1992 van der Hilst ampMann 1994 Malave amp Suarez 1995 Perez et al 1997b Perez ampMendoza 1998 Trenkamp et al 2002 Perez et al 2011 Bernal-Olaya et al 2015 Chiarabba et al 2015 Syracuse et al 2016Mora-Paez et al 2016b) have studied the seismicity and tectonicsof northwestern South America There is a general agreement thatalong the northwestern tip of the continent there exists a northeast-trending slab of Caribbean plate that is subducting along the SouthCaribbean Marginal fault to the southeast (Figs 1 and 6) beneath theNorth Andes plate (Toto amp Kellogg 1992 Trenkamp et al 2002Bird 2003) Here we shall re-visit the seismicity of the region usingteleseismic locations (mb ge 45) reported in the revised catalogues

4 OJ Perez et al

Table 1 Velocities in the ITRF2008 reference frame showing east and north components with one sigma error

Site ID Long (E) Lat (N) Ve (mm yrminus1) σ e (mm yrminus1) Vn (mm yrminus1) σ n (mm yrminus1) Ref

ALPA minus72918 11528 87 28 137 09 1ALTA minus66178 9866 minus51 05 92 05 2AMUA minus70188 11753 97 06 127 05 3ARAY minus63711 10655 59 25 129 23 4ARUB minus70013 12574 130 13 111 12 2AVES minus63618 15667 146 27 149 18 4BAPA minus74660 5470 28 04 146 07 5BARB minus59620 13250 129 18 143 16 4BARI minus70381 8657 minus49 04 120 03 6BARU minus75590 10258 114 08 103 05 1BATE minus71110 9170 19 04 150 03 6BLAN minus64599 11822 107 25 150 18 4BUCM minus73182 7117 30 23 135 08 1 7CALA minus67449 9061 minus37 25 105 12 8CANO minus63861 8568 minus56 09 115 06 2 8CAPI minus72430 5350 minus30 05 115 06 5CART minus75500 10360 122 09 105 09 1 7CARU minus63244 10675 74 31 120 25 4CASI minus66960 9926 minus42 05 120 05 2CHUR minus69541 10818 94 07 139 05 8CN35 minus81363 13374 173 minus27 07 06 1CN37 minus75263 10793 90 17 136 21 1CN38 minus71988 12222 115 06 141 04 1CN40 minus68960 12180 125 07 118 07 8COCH minus63994 10782 132 2 126 2 4DALI minus72314 8267 44 06 115 05 6FABI minus61658 10098 minus19 12 126 11 4FRAI minus70799 8850 minus14 04 134 03 6FUND minus71863 7783 minus10 06 103 05 6GOYA minus69708 10995 91 07 137 05 3GREO minus61640 12220 134 05 150 05 8GRIT minus71998 8156 03 03 109 02 6GUAC minus71333 7515 minus46 03 104 03 6GUAY minus72342 8530 67 06 114 05 6JUAN minus63369 10357 minus31 3 120 26 4LUCI minus60970 14090 122 18 141 16 4MARA minus71624 10674 88 08 128 08 2MARG minus6436 11042 138 15 137 25 4MONA minus70460 9550 minus10 06 126 05 2MONT minus75681 8895 113 06 111 04 1 7MZAL minus75470 5030 123 06 179 04 1 7PREG minus71775 8029 minus14 03 109 03 6QUIB minus69783 9914 33 32 119 13 3RION minus75430 6176 51 03 147 02 1 7ROSA minus69774 10166 31 05 130 05 2RQUE minus66678 11953 140 24 134 18 3STOD minus70630 8860 minus29 04 127 03 3SANA minus81720 12580 123 04 70 04 1 7TORO minus71126 9064 minus12 04 105 03 3TORT minus65227 10963 133 2 133 17 3TRIN minus61400 10680 124 14 134 12 9USB0 minus66792 10409 minus11 05 120 05 2USB1 minus66883 10411 minus14 05 121 05 2UWAS minus72390 6450 03 04 130 06 5VBUV minus73860 5530 39 05 156 07 5VDPR minus73248 10436 85 04 153 04 1VPOL minus74861 10794 77 08 166 05 1

1 Mora-Paez et al (2016a) 2 this study 3 Re-evaluated from Perez et al (2001) 4 Perez et al (2001) 5 Mora-Paez et al (2016b) 6 Re-evaluated fromPerez et al (2011) Trenkamp et al (2002) 8 D Mencin unpublished data 2015 9 Estimated from Weber et al (2001)

of rotation between ITRF2008 and South America provided by Al-tamimi et al (2012) They are graphically shown in Figs 3 4 and8 Finally we incorporated seven velocity vectors in the Colombianregion expressed in the South America reference frame originally

provided by Trenkamp et al (2002) which were re-calculated inthe ITRF2008 frame (Mora-Paez et al 2016a)

GPS and plate kinematics in south Caribbean 5

Table 2 Euler pole locations and rotation rates relative to the South Amer-ican plate for the best fitting models described in the text

Blockplate Latitude () Longitude ()Rotation rate (deg

Myrminus1)

(lowast) Caribbean 619 plusmn 20 2843 plusmn 20 0229 plusmn 0015Model 4Block A 252 plusmn 93 2807 plusmn 23 0575 plusmn 0296Block B 51 plusmn 55 1122 plusmn 13 0403 plusmn 0135Block C minus63 plusmn 66 1144 plusmn 21 0339 plusmn 0499Model 5Block A 220 plusmn 62 2813 plusmn 16 0672 plusmn 0293Block B minus23 plusmn 10 1106 plusmn 02 0660 plusmn 0080Block C 130 plusmn 20 2947 plusmn 15 0657 plusmn 0541Block D 03 plusmn 14 1106 plusmn 02 0603 plusmn 0079Block E 213 plusmn 357 1154 plusmn 50 0341 plusmn 0322

(lowast) From Perez et al 2001

R E S U LT S

Slip-partitioning associated with the San Sebastian and LaVictoria fault zones

In this section we combine GPS data with simple elastic strainaccumulation models and seismic evidence to constrain slip rate es-timates along these two major dextral easterly striking fault zonesThe two faults (Fig 3) form the CaribbeanndashSouth American mainplate boundary east of sim68W Our goal is to quantify how thesim196 mm yrminus1 in the N85E direction of CaribbeanndashSouth Amer-ican rate of relative motion indicated by the corresponding pole ofrotation (Table 2) derived from GPS geodesy (Perez et al 2001)is distributed between these fault zones N85E is also the strikedirection of the San Sebastian fault

Fig 3 shows the seismicity (moment magnitude Mw ge 26 cir-cles) reported by FUNVISIS (2016) for the period 2003ndash2012 to-gether with the GPS-derived velocity vectors (arrows) relative tostable South America in northndashcentral Venezuela About 985 percent of the seismic events reported are microearthquakes in the Mw

range of 26ndash44 Only minor activity is observed in the inner re-gions of the Caribbean though a series of seismological stationsare located in Venezuelan islands and along the northern coast ofthe country (see httpwwwfunvisisgobve) FUNVISIS catalogshows a fairly constant rate of listing for events with Mw ge 26during the time span under consideration

East of sim68W the San Sebastian and La Victoria faults co-exist These fault zones are nearly parallel (Fig 3) making anangle of sim5ndash10 and are separated by a distance that ranges fromsim30ndash50 km Recent marine geophysics studies show that at someplaces the San Sebastian fault is multibranched and the fault zonereaches a width of 10ndash12 km (Colon et al 2015) As found outby Perez et al (1997b) from a different set of earlier earthquakesmost of the seismic activity in the region straddles both fault zonesoccurs in general at focal depths shallower than sim14 km and focalmechanisms commonly show dextral motion on easterly orientedvertical fault planes (Figs 2 and 4) These observations show thatthese two fault zones accommodate most of the relative motionbetween the Caribbean and South American plates in the region

Fig 5 is a stacked profile showing the N85E velocity componentof GPS sites located to the north and south of the San Sebastianand La Victoria faults The stations used in the stack are marked inFig 4 with blue stars Many benchmarks were vandalized since thefirst occupations in 1999 and thus the number of data points remainsrelatively few Nonetheless the velocities of USB0 and USB1 are

very well constrained since all together they have been occupied 14times since 1999 and their location between the two faults providesstrong constraints on the slip partitioning we next describe

The field (Fig 5) across these faults is confined to a shear zonelt120 km wide and most of the surface deformation is containedwithin a 70ndash80 km wide zone encompassing the San Sebastianand La Victoria faults The dashed line in Fig 5 is the elasticmodelled surface velocities (Okada 1985 Dixon et al 1998) for asingle N85E oriented vertical fault taken up the full CaribbeanndashSouth American relative motion (sim196 mm yrminus1) The solid lineis in contrast calculated for a system of two subparallel verticalfaults (ie The La Victoria and San Sebastian faults) separated by adistance of 45 km in both cases with dextral shear below a lockingdepth of 14 Km The two-fault model provides a good fit for theobserved data and corresponds to a dextral shear of 26 plusmn 04 mmyrminus1 for La Victoria fault and 170 plusmn 05 mm yrminus1 for San Sebastianfault below a locking depth of 14 plusmn 3 km

The modelling result does not preclude the possibility of someportion of slip modelled on the San Sebastian fault being accom-modated on other faults to the north such as the easterly striking LaTortuga fault (Fig 2) One could for example allow up to sim3 mmyrminus1 of slip along La Tortuga fault by decreasing a similar amountof modelled slip on the San Sebastian fault without violating theobservations We hesitate to argue for this latter possibility becauseboth the seismic record and marine geophysics studies (Colon et al2015) do not support the existence of the La Tortuga fault alongthe entire northndashcentral coast of Venezuela Similarly it is possiblethat the plate boundary may include a third fault farther north alongthe easterly striking section of the South Caribbean Deformed Beltin the off-shore area (Fig 1) with a small and probably negligi-ble amount of slip on that portion of the structure (Symithe et al2015) expressed by an actively deforming accretionary prism (Gor-ney et al 2007 Clark et al 2008 Magnani et al 2009 Escalonaamp Mann 2011 Kroehler et al 2011) Finally based on GPS in-vestigations there is general agreement (Perez et al 2001 Clarket al 2008 Jouanne et al 2011 Reinoza et al 2015 Symithe et al2015) in that east of sim655W in eastern Venezuela (Figs 1 and 2)the easterly striking El Pilar fault zone is the main strand of theCaribbeanndashSouth America plate boundary taking up most of therelative motion between the two plates In our interpretation boththe San Sebastian and La Victoria faults merge around 655W withthe El Pilar fault zone so that their motion is absorbed into the singleEl Pilar strand

The interaction of the North Andes plate with theCaribbean and the South American plates in westernVenezuela and northern Colombia

Numerous authors (eg Dewey 1972 Pennington 1981 Kellogg ampBonini 1982 Kellogg 1984 Toto amp Kellogg 1992 van der Hilst ampMann 1994 Malave amp Suarez 1995 Perez et al 1997b Perez ampMendoza 1998 Trenkamp et al 2002 Perez et al 2011 Bernal-Olaya et al 2015 Chiarabba et al 2015 Syracuse et al 2016Mora-Paez et al 2016b) have studied the seismicity and tectonicsof northwestern South America There is a general agreement thatalong the northwestern tip of the continent there exists a northeast-trending slab of Caribbean plate that is subducting along the SouthCaribbean Marginal fault to the southeast (Figs 1 and 6) beneath theNorth Andes plate (Toto amp Kellogg 1992 Trenkamp et al 2002Bird 2003) Here we shall re-visit the seismicity of the region usingteleseismic locations (mb ge 45) reported in the revised catalogues

GPS and plate kinematics in south Caribbean 5

Table 2 Euler pole locations and rotation rates relative to the South Amer-ican plate for the best fitting models described in the text

Blockplate Latitude () Longitude ()Rotation rate (deg

Myrminus1)

(lowast) Caribbean 619 plusmn 20 2843 plusmn 20 0229 plusmn 0015Model 4Block A 252 plusmn 93 2807 plusmn 23 0575 plusmn 0296Block B 51 plusmn 55 1122 plusmn 13 0403 plusmn 0135Block C minus63 plusmn 66 1144 plusmn 21 0339 plusmn 0499Model 5Block A 220 plusmn 62 2813 plusmn 16 0672 plusmn 0293Block B minus23 plusmn 10 1106 plusmn 02 0660 plusmn 0080Block C 130 plusmn 20 2947 plusmn 15 0657 plusmn 0541Block D 03 plusmn 14 1106 plusmn 02 0603 plusmn 0079Block E 213 plusmn 357 1154 plusmn 50 0341 plusmn 0322

(lowast) From Perez et al 2001

R E S U LT S

Slip-partitioning associated with the San Sebastian and LaVictoria fault zones

In this section we combine GPS data with simple elastic strainaccumulation models and seismic evidence to constrain slip rate es-timates along these two major dextral easterly striking fault zonesThe two faults (Fig 3) form the CaribbeanndashSouth American mainplate boundary east of sim68W Our goal is to quantify how thesim196 mm yrminus1 in the N85E direction of CaribbeanndashSouth Amer-ican rate of relative motion indicated by the corresponding pole ofrotation (Table 2) derived from GPS geodesy (Perez et al 2001)is distributed between these fault zones N85E is also the strikedirection of the San Sebastian fault

Fig 3 shows the seismicity (moment magnitude Mw ge 26 cir-cles) reported by FUNVISIS (2016) for the period 2003ndash2012 to-gether with the GPS-derived velocity vectors (arrows) relative tostable South America in northndashcentral Venezuela About 985 percent of the seismic events reported are microearthquakes in the Mw

range of 26ndash44 Only minor activity is observed in the inner re-gions of the Caribbean though a series of seismological stationsare located in Venezuelan islands and along the northern coast ofthe country (see httpwwwfunvisisgobve) FUNVISIS catalogshows a fairly constant rate of listing for events with Mw ge 26during the time span under consideration

East of sim68W the San Sebastian and La Victoria faults co-exist These fault zones are nearly parallel (Fig 3) making anangle of sim5ndash10 and are separated by a distance that ranges fromsim30ndash50 km Recent marine geophysics studies show that at someplaces the San Sebastian fault is multibranched and the fault zonereaches a width of 10ndash12 km (Colon et al 2015) As found outby Perez et al (1997b) from a different set of earlier earthquakesmost of the seismic activity in the region straddles both fault zonesoccurs in general at focal depths shallower than sim14 km and focalmechanisms commonly show dextral motion on easterly orientedvertical fault planes (Figs 2 and 4) These observations show thatthese two fault zones accommodate most of the relative motionbetween the Caribbean and South American plates in the region

Fig 5 is a stacked profile showing the N85E velocity componentof GPS sites located to the north and south of the San Sebastianand La Victoria faults The stations used in the stack are marked inFig 4 with blue stars Many benchmarks were vandalized since thefirst occupations in 1999 and thus the number of data points remainsrelatively few Nonetheless the velocities of USB0 and USB1 are

very well constrained since all together they have been occupied 14times since 1999 and their location between the two faults providesstrong constraints on the slip partitioning we next describe

The field (Fig 5) across these faults is confined to a shear zonelt120 km wide and most of the surface deformation is containedwithin a 70ndash80 km wide zone encompassing the San Sebastianand La Victoria faults The dashed line in Fig 5 is the elasticmodelled surface velocities (Okada 1985 Dixon et al 1998) for asingle N85E oriented vertical fault taken up the full CaribbeanndashSouth American relative motion (sim196 mm yrminus1) The solid lineis in contrast calculated for a system of two subparallel verticalfaults (ie The La Victoria and San Sebastian faults) separated by adistance of 45 km in both cases with dextral shear below a lockingdepth of 14 Km The two-fault model provides a good fit for theobserved data and corresponds to a dextral shear of 26 plusmn 04 mmyrminus1 for La Victoria fault and 170 plusmn 05 mm yrminus1 for San Sebastianfault below a locking depth of 14 plusmn 3 km

The modelling result does not preclude the possibility of someportion of slip modelled on the San Sebastian fault being accom-modated on other faults to the north such as the easterly striking LaTortuga fault (Fig 2) One could for example allow up to sim3 mmyrminus1 of slip along La Tortuga fault by decreasing a similar amountof modelled slip on the San Sebastian fault without violating theobservations We hesitate to argue for this latter possibility becauseboth the seismic record and marine geophysics studies (Colon et al2015) do not support the existence of the La Tortuga fault alongthe entire northndashcentral coast of Venezuela Similarly it is possiblethat the plate boundary may include a third fault farther north alongthe easterly striking section of the South Caribbean Deformed Beltin the off-shore area (Fig 1) with a small and probably negligi-ble amount of slip on that portion of the structure (Symithe et al2015) expressed by an actively deforming accretionary prism (Gor-ney et al 2007 Clark et al 2008 Magnani et al 2009 Escalonaamp Mann 2011 Kroehler et al 2011) Finally based on GPS in-vestigations there is general agreement (Perez et al 2001 Clarket al 2008 Jouanne et al 2011 Reinoza et al 2015 Symithe et al2015) in that east of sim655W in eastern Venezuela (Figs 1 and 2)the easterly striking El Pilar fault zone is the main strand of theCaribbeanndashSouth America plate boundary taking up most of therelative motion between the two plates In our interpretation boththe San Sebastian and La Victoria faults merge around 655W withthe El Pilar fault zone so that their motion is absorbed into the singleEl Pilar strand

The interaction of the North Andes plate with theCaribbean and the South American plates in westernVenezuela and northern Colombia

Numerous authors (eg Dewey 1972 Pennington 1981 Kellogg ampBonini 1982 Kellogg 1984 Toto amp Kellogg 1992 van der Hilst ampMann 1994 Malave amp Suarez 1995 Perez et al 1997b Perez ampMendoza 1998 Trenkamp et al 2002 Perez et al 2011 Bernal-Olaya et al 2015 Chiarabba et al 2015 Syracuse et al 2016Mora-Paez et al 2016b) have studied the seismicity and tectonicsof northwestern South America There is a general agreement thatalong the northwestern tip of the continent there exists a northeast-trending slab of Caribbean plate that is subducting along the SouthCaribbean Marginal fault to the southeast (Figs 1 and 6) beneath theNorth Andes plate (Toto amp Kellogg 1992 Trenkamp et al 2002Bird 2003) Here we shall re-visit the seismicity of the region usingteleseismic locations (mb ge 45) reported in the revised catalogues

6 OJ Perez et al

Figure 4 Representative focal mechanisms of microearthquakes reported by several authors in northndashcentral Venezuela (Perez et al 1997ab Audemard et al2005) and GPS velocity vectors Focal Mechanisms straddling the Bocono fault show dextral slip and reverse faulting Those along the Morrocoy fault showa combination of dextral slip and normal faulting San Sebastian fault shows dextral motion La Victoria fault shows dextral slip along strike and southeastdextral motion (right-stepping offset) at places where it is intercepted by southeast striking dextral slip faults GPS vectors marked with a blue star are used toconstruct Fig 5

Figure 5 Velocity field in northndashcentral Venezuela and southndashcentralCaribbean Dots are the N85E component of the observed velocity vec-tors at sites marked with a red star in Fig 4 located to the north and south ofthe San Sebastian and La Victoria faults plusmn 1σ uncertainties (bars) Brokenline is the modelled surface velocities due to 20 mm yrminus1 of dextral shearbelow a locking depth of 14 km on a single N85E trending vertical faultSolid line is the corresponding modelling for two subparallel faults located47 km apart namely La Victoria and San Sebastian faults with a slip rateof 26 and 17 mm yrminus1 respectively

of the ISC for the period 1980ndash2013 as well as microearthquakelocations (25 lt mb lt 45) reported in Colombian territory bythe Colombian Geological Survey (Colombian Geological Survey2016) for the period 1993ndash2015 and by FUNVISIS (2016) for theperiod 2003ndash2012 in Venezuelan territory followed by the analysisof GPS-derived velocity vectors and their tectonic implications

The epicentral locations are plotted as a function of increas-ing depth in Fig 6 The distribution of seismicity defines an in-clined northeast striking slab dipping southeasterly in northwesternVenezuela and northern Colombia The top of the interpreted slabis defined by the series of depth contours (dotted lines) shown inthe figure These contours illustrate the northeast strike and south-east dip of the Benniof zone that defines the subduction of theCaribbean slab along the South Caribbean Marginal fault beneath

South America Three northwestndashsoutheast vertical cross-sections(1ndash3) are shown and described in Fig 7 and its caption Each ofthem contains the earthquake hypocentres (dots) located within therectangles 1ndash3 shown in Fig 6 The horizontal projection of thedip of the subducted slab as manifest by the seismicity (Fig 6) isoblique to the calculated slip vector (Perez et al 2001 Weber et al2001) of the Caribbean plate relative to South America at 12N77W (broken arrow in Figs 6 and 8) It is also observed in Fig 6that seismicity generally terminates and does not occur northeast ofthe Morrocoy fault zone an observation that led Perez et al (1997b)to conclude that this fault zone marks the northeastern end of theNorth Andes plate

Fig 8 shows a series of GPS velocity vectors (arrows) projectedrelative to the South American plate reported by several authors(Trenkamp et al 2002 Perez et al 2011 Mora-Paez et al 2016abthis study) at sites within the northernmost regions of the NorthAndes plate and adjacent Caribbean and South American areasThe GPS vectors at sites located in the northern inner regions ofthe North Andes plate (marked with blue stars in Fig 8) show anaverage magnitude of 150 plusmn 10 mm yrminus1 and are easterly oriented(average Az 80 plusmn 6) which for reference is the motion of siteMARA The 196 mm yrminus1 relative motion of the Caribbean platewith respect to the South American plate shown by a dashed arrowat two places in Fig 8 is calculated from the corresponding pole ofrotation provided by Perez et al (2001) To shift the reference frameto the North Andes plate we subtract the MARA vector from all otherplotted vectors in Fig 8 The result is shown in Fig 9 and alongwith Fig 8 serves to illustrate the rate and sense of displacementsoccurring at the boundaries of the North Andes plate

It is observed with the vectors in Fig 9 that the Caribbean plateis subducting at rate of sim5ndash7 mm yrminus1 beneath the North Andesplate and that motion along each the Morrocoy and Bocono faultsboundaries is right-lateral As well there is an obliquity of motionalong the Bocono and Morrocoy boundaries that predicts each tohave a component of transpression and transtension respectivelyThus the Morrocoy fault can be interpreted as a continental rift apartboundary as pointed out by Bird (2003) at the same time beinga dextral shear zone that transforms the Caribbean subduction innorthwestern South America into pure eastndashwest strike-slip motion

GPS and plate kinematics in south Caribbean 7

Figure 6 Seismicity in the northwestern tip of South America Epicentres are discriminated only by their focal depth (h) as indicated in the figure Cataloguesused ISC (2013) revised for Mw gt 45 from 1980 to 2013 FUNVISIS (2016) from 2003 to 2012 and Colombian Geological Survey (2016) from 1993 to2011 for 25 lt Mw lt 45 f fault Boxes 1ndash3 comprise the hypocentres that are used to define the Benioff zone and subducted Caribbean slab in those regionsshown in Fig 7 Broken arrows show calculated slip rate (Perez et al 2001 Weber et al 2001) of the Caribbean plate relative to South America at thoselocations Dotted sinuous lines are depth contours to the top surface of the inclined Caribbean slab whose southern tip appears to collide at depth with theeasterly subducted Nazca slab beneath Bucaramanga (B Syracuse et al 2016)

along the easterly oriented San Sebastian fault in northndashcentralVenezuela

Similarly it is observed in Fig 9 that South America is movingabout 15 mm yrminus1 at an azimuth of 260 relative to the North Andesplate Assuming the relative motion is mostly accommodated bydeformation along trend of the Bocono fault system the GPS vectorsand model in Fig 9 indicate about 9ndash11 mm yrminus1 of dextral motionand 2ndash5 mm yrminus1 of contraction across the fault consistent with thegeodetic results reported by Perez et al (2011 see their figs 3 and 4)and Symithe et al (2015) The interaction between the North Andesand South American plates along the northeast striking Bocono faultzone in the Venezuelan Andes has been studied using seismic andgeomorphological data as well as GPS-derived velocity vectors (egPerez et al 1997b 2011 Perez amp Mendoza 1998 Wesnousky et al2012 Guzman et al 2013 Symithe et al 2015) Our additionalGPS vectors confirm their main results which indicate that due tothe obliqueness of the Bocono fault relative to the easterly orientedslip of the North Andean plate relative to South America this slipis partitioned into 9ndash11 mm yrminus1 of dextral motion in the far fieldof the Bocono fault and a compressive regime perpendicular to theAndes that results in 2ndash5 mm yrminus1 of horizontal shortening acrossthe ranges and seismically active thrust faulting along the foothills(Wesnousky et al 2012)

Several faults with slip rates estimated on the order 2 mm yrminus1

or less occur within the northern tip of the North Andes plate andmay also accommodate a small fraction of the CaribbeanndashSouthAmerica relative motion (Audemard 1996 Audemard amp Audemard2002 Symithe et al 2015) The Oca and Santa Marta faults (Figs 8and 9) are the longest and most prominent of these slower slip-ping faults (Audemard amp Audemard 2002 Audemard et al 2005Symithe et al 2015) These faults have not shown the occurrence

of significant earthquake activity since the colony times but palaeo-seismicity studies show they have generated earthquakes in recent(Quaternary) times (Audemard 1996 Idarraga-Garcıa amp Romero2010) In the next section we perform a kinematic block motionanalysis aimed to better quantify the rate of motion taking placealong each of the main active faults of the southern Caribbean plateboundary with emphasis in our study area that is the northwesterntip of South America including the Morrocoy Oca and Santa Martafaults

K I N E M AT I C B L O C K M O D E L L I N G

We use GPS observations in conjunction with the linear block theoryand associated code lsquoBlocksrsquo presented in Meade amp Loveless (2009)to estimate slip rates on the major active faults and partitioning ofdeformation occurring along the southern Caribbean plate Thekinematic block modelling approach is used to study regional GPSvelocity fields and quantify the rates of motions along locked orpartially locked active faults that form a plate or block boundary(eg McCaffrey 2002 Meade et al 2002 Reilinger et al 2006Loveless amp Meade 2010 Saria et al 2014 Symithe et al 2015) Itcan also be used to confirm or reject a proposed fault boundary

To perform the modelling we assume for analysis the plate bound-ary geometry provided by Symithe et al (2015) for the Lesser An-tilles the Trinidad region and northeastern Venezuela (Fig 10)The geometry is based on their analysis of the seismicity alongthe Lesser Antilles the results and interpretations of Weber et al(2009) for the Trinidad and surrounding areas and Perez et al(2001) analysis showing the El Pilar fault as the main strand ofthe Caribbean plate boundary in northeastern Venezuela West of65W the main South American block boundaries faults tested are(Fig 10) the San Sebastian (S) La Victoria (V) Bocono (B) Oca

8 OJ Perez et al

Figure 7 Northwestndashsoutheast vertical cross-sections of earthquake hypocentres whose epicentres are located within the corresponding rectangles 1 in thenorth to 3 in the south shown in Fig 6 These cross-sections clearly define the Benioff zone and Caribbean slab subducting to the southeast beneath the NorthAndes plate in northwestern Venezuela and northern Colombia Arrows in cross-section 1 show sense of motion relative to South America Maracaibo (M)Lake and Perija Peninsula are also shown in Fig 6 for reference Faults SCMf South Caribbean Marginal Of Oca SMf Santa Marta Bf Bocono C-Lcoast line In the north (cross-sections 1 and 2) the shallower section of the slab dips sim10 above a depth of 40 km and 45ndash50 below that depth In thesouth (cross-section 3) it flattens and the shallower and deeper sections dip sim5 and sim35 respectively These results were used in our kinematic block motionmodelling

(O) Morrocoy (M) Santa Marta (SM) East Andean (E) SouthCaribbean Marginal (SCMF) faults and the eastern section of theSouth Caribbean Deformed Belt (CDB in Fig 10) east of sim72WThis belt has been proposed as an accretionary wedge along whichthe Venezuelan basin is obliquely subducted beneath South Amer-ica (Jordan 1975 Silver et al 1975 Talwani et al 1977 Gorneyet al 2007 Magnani et al 2009 Escalona amp Mann 2011 Kroehleret al 2011) We also examine the effect of the Oca Morrocoy andCaribbean Deformed Belt faults by assembling and modelling a se-ries of block geometries in which they are systematically includedand excluded In these cases it is assumed consistent with geologythat the Santa Marta fault must have a left-lateral component of mo-tion (Paris et al 2000 Idarraga-Garcıa amp Romero 2010) and thatmotion on the Oca fault must be rather small (lt2 mm yrminus1) such asindicated by palaeoseismicity studies (Audemard 1996 Audemardet al 2000) and the block motion analysis of Symithe et al (2015)

The set-up of the geometric configuration of the various faultsegments used to define each plate or block boundary to be usedin all models run was done in the following way (1) We used alocking depth of 14 km for all strike-slip faults in agreement withthe seismogenic depth reported along the El Pilar San SebastianLa Victoria and Bocono faults (all faults are shown in Fig 1) inseismological studies (eg Perez amp Aggarwal 1981 Perez et al1997ab) (2) In the case of a series of subduction segments alongthe southern Lesser Antilles and Panama (Fig 10) we followedSymithe et al (2015) and used a constant dip of 16 and a constantlocking depth of 40 km (3) For the subduction segments along theSouthern Caribbean Marginal fault (SCM in Fig 10) we used alocking depth of 40 km and a dip of 10 except along its southwestsegment where we used 5 all according to the seismicity profilesshown in Fig 7 The same locking depth (40 km) and a dip of 13

obtained by Schmitz et al (2008) is used for the east section of the

GPS and plate kinematics in south Caribbean 9

Figure 8 GPS velocity vectors (red arrows) relative to the South American plate in northwestern South America Ellipses are plusmn 1σ uncertainties Blue starsmark the velocity vectors used to estimate the relative slip of the North Andes plate relative to the South American plate in the northwestern tip of SouthAmerica B Bucaramanga

Figure 9 Estimated motion (open arrows) of the Caribbean and South American plates relative to the North Andes plate Black arrows are the correspondingvelocity vectors at each GPS site Site Mara and other sites in the inner regions of the North Andes plate show little or no motion in the North Andes referenceframe B Bucaramanga

South Caribbean Deformed Belt east of 72W (CDB in Fig 10)(4) All our analyses are done in the South American (fix) referenceframe Following Manaker et al (2008) we constrain the angularvelocity of the Caribbean plate in our case relative to the SouthAmerican plate using the Perez et al (2001) pole of rotation givenin Table 2

Fig 10 shows a series of five tested models (Models 1ndash5) InModels 1ndash4 we simultaneously solve for block rotation rates kine-matically consistent fully coupled fault slip rates and effectiveelastic coupling coefficients whereas in Model 5 we impose andfix a slip rate on two slow slipping faults namely the Oca faultand the east-striking section of the Caribbean Deformed Belt eastof 72W as we later describe We evaluate each model fit by ex-amining the residual (observed-modelled) velocities estimated by

the block model analysis while remaining consistent with the pre-viously mentioned geological constraints on the sense and rate offault slip of the Santa Marta and Oca faults We follow the style ofLoveless amp Meade (2010) in presentation of the results in the figurethe estimated fault slip rates predicted from each modelling of blockor plate bounding segments are annotated along each segment Boldblack lines indicate the fault traces and the corresponding labelsgive strike (top) and fault-normal (bottom) slip rates and uncertain-ties Positive signs are for left-lateral and convergent fault motionsand negative signs for dextral slip and tensional faults Blue andmagenta arrows are the observed and residual (observed-modelled)velocities respectively Capital letters in the figure provide faultnames The strengths and weaknesses of each model are discussedin the following paragraphs

10 OJ Perez et al

Figure 10 Estimated fault slip rates relative to South America on plate and block bounding segments of southern Caribbean and northern South America forthe five models discussed in the text Bold black lines indicate the fault traces The corresponding labels give strike (top) and fault-normal (bottom) slip ratesand uncertainties Negative signs indicate dextral and tensile faults positive signs left-lateral and convergent faults Blue and magenta arrows are the observedand residual velocity vectors respectively Faults BS BajosndashEl Soldado P El Pilar S San Sebastian V La Victoria M Morrocoy B Bocono SM SantaMarta E East Andean SCMF South Caribbean Marginal CDB South Caribbean Deformed Belt T Trinidad Model 2 provides the names of several islandsmentioned in the text Ar Aruba Cu Curacao Lr Los Roques Lt La Tortuga Models 1ndash3 are rejected on the basis of geological and geodetic constraintsModel 4 is the block model that best fits the GPS data and geological constraints we used Model 5 is a hybrid model that incorporates slower slipping faults(O Oca and CDB Caribbean Deformed Belt) fixing their slip rates as explained in the text Letters within circles in Models 4 and 5 indicate the blocks thatconfigurate each model whose calculated poles of rotation are given in Table 2

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

GPS and plate kinematics in south Caribbean 7

Figure 6 Seismicity in the northwestern tip of South America Epicentres are discriminated only by their focal depth (h) as indicated in the figure Cataloguesused ISC (2013) revised for Mw gt 45 from 1980 to 2013 FUNVISIS (2016) from 2003 to 2012 and Colombian Geological Survey (2016) from 1993 to2011 for 25 lt Mw lt 45 f fault Boxes 1ndash3 comprise the hypocentres that are used to define the Benioff zone and subducted Caribbean slab in those regionsshown in Fig 7 Broken arrows show calculated slip rate (Perez et al 2001 Weber et al 2001) of the Caribbean plate relative to South America at thoselocations Dotted sinuous lines are depth contours to the top surface of the inclined Caribbean slab whose southern tip appears to collide at depth with theeasterly subducted Nazca slab beneath Bucaramanga (B Syracuse et al 2016)

along the easterly oriented San Sebastian fault in northndashcentralVenezuela

Similarly it is observed in Fig 9 that South America is movingabout 15 mm yrminus1 at an azimuth of 260 relative to the North Andesplate Assuming the relative motion is mostly accommodated bydeformation along trend of the Bocono fault system the GPS vectorsand model in Fig 9 indicate about 9ndash11 mm yrminus1 of dextral motionand 2ndash5 mm yrminus1 of contraction across the fault consistent with thegeodetic results reported by Perez et al (2011 see their figs 3 and 4)and Symithe et al (2015) The interaction between the North Andesand South American plates along the northeast striking Bocono faultzone in the Venezuelan Andes has been studied using seismic andgeomorphological data as well as GPS-derived velocity vectors (egPerez et al 1997b 2011 Perez amp Mendoza 1998 Wesnousky et al2012 Guzman et al 2013 Symithe et al 2015) Our additionalGPS vectors confirm their main results which indicate that due tothe obliqueness of the Bocono fault relative to the easterly orientedslip of the North Andean plate relative to South America this slipis partitioned into 9ndash11 mm yrminus1 of dextral motion in the far fieldof the Bocono fault and a compressive regime perpendicular to theAndes that results in 2ndash5 mm yrminus1 of horizontal shortening acrossthe ranges and seismically active thrust faulting along the foothills(Wesnousky et al 2012)

Several faults with slip rates estimated on the order 2 mm yrminus1

or less occur within the northern tip of the North Andes plate andmay also accommodate a small fraction of the CaribbeanndashSouthAmerica relative motion (Audemard 1996 Audemard amp Audemard2002 Symithe et al 2015) The Oca and Santa Marta faults (Figs 8and 9) are the longest and most prominent of these slower slip-ping faults (Audemard amp Audemard 2002 Audemard et al 2005Symithe et al 2015) These faults have not shown the occurrence

of significant earthquake activity since the colony times but palaeo-seismicity studies show they have generated earthquakes in recent(Quaternary) times (Audemard 1996 Idarraga-Garcıa amp Romero2010) In the next section we perform a kinematic block motionanalysis aimed to better quantify the rate of motion taking placealong each of the main active faults of the southern Caribbean plateboundary with emphasis in our study area that is the northwesterntip of South America including the Morrocoy Oca and Santa Martafaults

K I N E M AT I C B L O C K M O D E L L I N G

We use GPS observations in conjunction with the linear block theoryand associated code lsquoBlocksrsquo presented in Meade amp Loveless (2009)to estimate slip rates on the major active faults and partitioning ofdeformation occurring along the southern Caribbean plate Thekinematic block modelling approach is used to study regional GPSvelocity fields and quantify the rates of motions along locked orpartially locked active faults that form a plate or block boundary(eg McCaffrey 2002 Meade et al 2002 Reilinger et al 2006Loveless amp Meade 2010 Saria et al 2014 Symithe et al 2015) Itcan also be used to confirm or reject a proposed fault boundary

To perform the modelling we assume for analysis the plate bound-ary geometry provided by Symithe et al (2015) for the Lesser An-tilles the Trinidad region and northeastern Venezuela (Fig 10)The geometry is based on their analysis of the seismicity alongthe Lesser Antilles the results and interpretations of Weber et al(2009) for the Trinidad and surrounding areas and Perez et al(2001) analysis showing the El Pilar fault as the main strand ofthe Caribbean plate boundary in northeastern Venezuela West of65W the main South American block boundaries faults tested are(Fig 10) the San Sebastian (S) La Victoria (V) Bocono (B) Oca

8 OJ Perez et al

Figure 7 Northwestndashsoutheast vertical cross-sections of earthquake hypocentres whose epicentres are located within the corresponding rectangles 1 in thenorth to 3 in the south shown in Fig 6 These cross-sections clearly define the Benioff zone and Caribbean slab subducting to the southeast beneath the NorthAndes plate in northwestern Venezuela and northern Colombia Arrows in cross-section 1 show sense of motion relative to South America Maracaibo (M)Lake and Perija Peninsula are also shown in Fig 6 for reference Faults SCMf South Caribbean Marginal Of Oca SMf Santa Marta Bf Bocono C-Lcoast line In the north (cross-sections 1 and 2) the shallower section of the slab dips sim10 above a depth of 40 km and 45ndash50 below that depth In thesouth (cross-section 3) it flattens and the shallower and deeper sections dip sim5 and sim35 respectively These results were used in our kinematic block motionmodelling

(O) Morrocoy (M) Santa Marta (SM) East Andean (E) SouthCaribbean Marginal (SCMF) faults and the eastern section of theSouth Caribbean Deformed Belt (CDB in Fig 10) east of sim72WThis belt has been proposed as an accretionary wedge along whichthe Venezuelan basin is obliquely subducted beneath South Amer-ica (Jordan 1975 Silver et al 1975 Talwani et al 1977 Gorneyet al 2007 Magnani et al 2009 Escalona amp Mann 2011 Kroehleret al 2011) We also examine the effect of the Oca Morrocoy andCaribbean Deformed Belt faults by assembling and modelling a se-ries of block geometries in which they are systematically includedand excluded In these cases it is assumed consistent with geologythat the Santa Marta fault must have a left-lateral component of mo-tion (Paris et al 2000 Idarraga-Garcıa amp Romero 2010) and thatmotion on the Oca fault must be rather small (lt2 mm yrminus1) such asindicated by palaeoseismicity studies (Audemard 1996 Audemardet al 2000) and the block motion analysis of Symithe et al (2015)

The set-up of the geometric configuration of the various faultsegments used to define each plate or block boundary to be usedin all models run was done in the following way (1) We used alocking depth of 14 km for all strike-slip faults in agreement withthe seismogenic depth reported along the El Pilar San SebastianLa Victoria and Bocono faults (all faults are shown in Fig 1) inseismological studies (eg Perez amp Aggarwal 1981 Perez et al1997ab) (2) In the case of a series of subduction segments alongthe southern Lesser Antilles and Panama (Fig 10) we followedSymithe et al (2015) and used a constant dip of 16 and a constantlocking depth of 40 km (3) For the subduction segments along theSouthern Caribbean Marginal fault (SCM in Fig 10) we used alocking depth of 40 km and a dip of 10 except along its southwestsegment where we used 5 all according to the seismicity profilesshown in Fig 7 The same locking depth (40 km) and a dip of 13

obtained by Schmitz et al (2008) is used for the east section of the

GPS and plate kinematics in south Caribbean 9

Figure 8 GPS velocity vectors (red arrows) relative to the South American plate in northwestern South America Ellipses are plusmn 1σ uncertainties Blue starsmark the velocity vectors used to estimate the relative slip of the North Andes plate relative to the South American plate in the northwestern tip of SouthAmerica B Bucaramanga

Figure 9 Estimated motion (open arrows) of the Caribbean and South American plates relative to the North Andes plate Black arrows are the correspondingvelocity vectors at each GPS site Site Mara and other sites in the inner regions of the North Andes plate show little or no motion in the North Andes referenceframe B Bucaramanga

South Caribbean Deformed Belt east of 72W (CDB in Fig 10)(4) All our analyses are done in the South American (fix) referenceframe Following Manaker et al (2008) we constrain the angularvelocity of the Caribbean plate in our case relative to the SouthAmerican plate using the Perez et al (2001) pole of rotation givenin Table 2

Fig 10 shows a series of five tested models (Models 1ndash5) InModels 1ndash4 we simultaneously solve for block rotation rates kine-matically consistent fully coupled fault slip rates and effectiveelastic coupling coefficients whereas in Model 5 we impose andfix a slip rate on two slow slipping faults namely the Oca faultand the east-striking section of the Caribbean Deformed Belt eastof 72W as we later describe We evaluate each model fit by ex-amining the residual (observed-modelled) velocities estimated by

the block model analysis while remaining consistent with the pre-viously mentioned geological constraints on the sense and rate offault slip of the Santa Marta and Oca faults We follow the style ofLoveless amp Meade (2010) in presentation of the results in the figurethe estimated fault slip rates predicted from each modelling of blockor plate bounding segments are annotated along each segment Boldblack lines indicate the fault traces and the corresponding labelsgive strike (top) and fault-normal (bottom) slip rates and uncertain-ties Positive signs are for left-lateral and convergent fault motionsand negative signs for dextral slip and tensional faults Blue andmagenta arrows are the observed and residual (observed-modelled)velocities respectively Capital letters in the figure provide faultnames The strengths and weaknesses of each model are discussedin the following paragraphs

10 OJ Perez et al

Figure 10 Estimated fault slip rates relative to South America on plate and block bounding segments of southern Caribbean and northern South America forthe five models discussed in the text Bold black lines indicate the fault traces The corresponding labels give strike (top) and fault-normal (bottom) slip ratesand uncertainties Negative signs indicate dextral and tensile faults positive signs left-lateral and convergent faults Blue and magenta arrows are the observedand residual velocity vectors respectively Faults BS BajosndashEl Soldado P El Pilar S San Sebastian V La Victoria M Morrocoy B Bocono SM SantaMarta E East Andean SCMF South Caribbean Marginal CDB South Caribbean Deformed Belt T Trinidad Model 2 provides the names of several islandsmentioned in the text Ar Aruba Cu Curacao Lr Los Roques Lt La Tortuga Models 1ndash3 are rejected on the basis of geological and geodetic constraintsModel 4 is the block model that best fits the GPS data and geological constraints we used Model 5 is a hybrid model that incorporates slower slipping faults(O Oca and CDB Caribbean Deformed Belt) fixing their slip rates as explained in the text Letters within circles in Models 4 and 5 indicate the blocks thatconfigurate each model whose calculated poles of rotation are given in Table 2

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

8 OJ Perez et al

Figure 7 Northwestndashsoutheast vertical cross-sections of earthquake hypocentres whose epicentres are located within the corresponding rectangles 1 in thenorth to 3 in the south shown in Fig 6 These cross-sections clearly define the Benioff zone and Caribbean slab subducting to the southeast beneath the NorthAndes plate in northwestern Venezuela and northern Colombia Arrows in cross-section 1 show sense of motion relative to South America Maracaibo (M)Lake and Perija Peninsula are also shown in Fig 6 for reference Faults SCMf South Caribbean Marginal Of Oca SMf Santa Marta Bf Bocono C-Lcoast line In the north (cross-sections 1 and 2) the shallower section of the slab dips sim10 above a depth of 40 km and 45ndash50 below that depth In thesouth (cross-section 3) it flattens and the shallower and deeper sections dip sim5 and sim35 respectively These results were used in our kinematic block motionmodelling

(O) Morrocoy (M) Santa Marta (SM) East Andean (E) SouthCaribbean Marginal (SCMF) faults and the eastern section of theSouth Caribbean Deformed Belt (CDB in Fig 10) east of sim72WThis belt has been proposed as an accretionary wedge along whichthe Venezuelan basin is obliquely subducted beneath South Amer-ica (Jordan 1975 Silver et al 1975 Talwani et al 1977 Gorneyet al 2007 Magnani et al 2009 Escalona amp Mann 2011 Kroehleret al 2011) We also examine the effect of the Oca Morrocoy andCaribbean Deformed Belt faults by assembling and modelling a se-ries of block geometries in which they are systematically includedand excluded In these cases it is assumed consistent with geologythat the Santa Marta fault must have a left-lateral component of mo-tion (Paris et al 2000 Idarraga-Garcıa amp Romero 2010) and thatmotion on the Oca fault must be rather small (lt2 mm yrminus1) such asindicated by palaeoseismicity studies (Audemard 1996 Audemardet al 2000) and the block motion analysis of Symithe et al (2015)

The set-up of the geometric configuration of the various faultsegments used to define each plate or block boundary to be usedin all models run was done in the following way (1) We used alocking depth of 14 km for all strike-slip faults in agreement withthe seismogenic depth reported along the El Pilar San SebastianLa Victoria and Bocono faults (all faults are shown in Fig 1) inseismological studies (eg Perez amp Aggarwal 1981 Perez et al1997ab) (2) In the case of a series of subduction segments alongthe southern Lesser Antilles and Panama (Fig 10) we followedSymithe et al (2015) and used a constant dip of 16 and a constantlocking depth of 40 km (3) For the subduction segments along theSouthern Caribbean Marginal fault (SCM in Fig 10) we used alocking depth of 40 km and a dip of 10 except along its southwestsegment where we used 5 all according to the seismicity profilesshown in Fig 7 The same locking depth (40 km) and a dip of 13

obtained by Schmitz et al (2008) is used for the east section of the

GPS and plate kinematics in south Caribbean 9

Figure 8 GPS velocity vectors (red arrows) relative to the South American plate in northwestern South America Ellipses are plusmn 1σ uncertainties Blue starsmark the velocity vectors used to estimate the relative slip of the North Andes plate relative to the South American plate in the northwestern tip of SouthAmerica B Bucaramanga

Figure 9 Estimated motion (open arrows) of the Caribbean and South American plates relative to the North Andes plate Black arrows are the correspondingvelocity vectors at each GPS site Site Mara and other sites in the inner regions of the North Andes plate show little or no motion in the North Andes referenceframe B Bucaramanga

South Caribbean Deformed Belt east of 72W (CDB in Fig 10)(4) All our analyses are done in the South American (fix) referenceframe Following Manaker et al (2008) we constrain the angularvelocity of the Caribbean plate in our case relative to the SouthAmerican plate using the Perez et al (2001) pole of rotation givenin Table 2

Fig 10 shows a series of five tested models (Models 1ndash5) InModels 1ndash4 we simultaneously solve for block rotation rates kine-matically consistent fully coupled fault slip rates and effectiveelastic coupling coefficients whereas in Model 5 we impose andfix a slip rate on two slow slipping faults namely the Oca faultand the east-striking section of the Caribbean Deformed Belt eastof 72W as we later describe We evaluate each model fit by ex-amining the residual (observed-modelled) velocities estimated by

the block model analysis while remaining consistent with the pre-viously mentioned geological constraints on the sense and rate offault slip of the Santa Marta and Oca faults We follow the style ofLoveless amp Meade (2010) in presentation of the results in the figurethe estimated fault slip rates predicted from each modelling of blockor plate bounding segments are annotated along each segment Boldblack lines indicate the fault traces and the corresponding labelsgive strike (top) and fault-normal (bottom) slip rates and uncertain-ties Positive signs are for left-lateral and convergent fault motionsand negative signs for dextral slip and tensional faults Blue andmagenta arrows are the observed and residual (observed-modelled)velocities respectively Capital letters in the figure provide faultnames The strengths and weaknesses of each model are discussedin the following paragraphs

10 OJ Perez et al

Figure 10 Estimated fault slip rates relative to South America on plate and block bounding segments of southern Caribbean and northern South America forthe five models discussed in the text Bold black lines indicate the fault traces The corresponding labels give strike (top) and fault-normal (bottom) slip ratesand uncertainties Negative signs indicate dextral and tensile faults positive signs left-lateral and convergent faults Blue and magenta arrows are the observedand residual velocity vectors respectively Faults BS BajosndashEl Soldado P El Pilar S San Sebastian V La Victoria M Morrocoy B Bocono SM SantaMarta E East Andean SCMF South Caribbean Marginal CDB South Caribbean Deformed Belt T Trinidad Model 2 provides the names of several islandsmentioned in the text Ar Aruba Cu Curacao Lr Los Roques Lt La Tortuga Models 1ndash3 are rejected on the basis of geological and geodetic constraintsModel 4 is the block model that best fits the GPS data and geological constraints we used Model 5 is a hybrid model that incorporates slower slipping faults(O Oca and CDB Caribbean Deformed Belt) fixing their slip rates as explained in the text Letters within circles in Models 4 and 5 indicate the blocks thatconfigurate each model whose calculated poles of rotation are given in Table 2

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

GPS and plate kinematics in south Caribbean 9

Figure 8 GPS velocity vectors (red arrows) relative to the South American plate in northwestern South America Ellipses are plusmn 1σ uncertainties Blue starsmark the velocity vectors used to estimate the relative slip of the North Andes plate relative to the South American plate in the northwestern tip of SouthAmerica B Bucaramanga

Figure 9 Estimated motion (open arrows) of the Caribbean and South American plates relative to the North Andes plate Black arrows are the correspondingvelocity vectors at each GPS site Site Mara and other sites in the inner regions of the North Andes plate show little or no motion in the North Andes referenceframe B Bucaramanga

South Caribbean Deformed Belt east of 72W (CDB in Fig 10)(4) All our analyses are done in the South American (fix) referenceframe Following Manaker et al (2008) we constrain the angularvelocity of the Caribbean plate in our case relative to the SouthAmerican plate using the Perez et al (2001) pole of rotation givenin Table 2

Fig 10 shows a series of five tested models (Models 1ndash5) InModels 1ndash4 we simultaneously solve for block rotation rates kine-matically consistent fully coupled fault slip rates and effectiveelastic coupling coefficients whereas in Model 5 we impose andfix a slip rate on two slow slipping faults namely the Oca faultand the east-striking section of the Caribbean Deformed Belt eastof 72W as we later describe We evaluate each model fit by ex-amining the residual (observed-modelled) velocities estimated by

the block model analysis while remaining consistent with the pre-viously mentioned geological constraints on the sense and rate offault slip of the Santa Marta and Oca faults We follow the style ofLoveless amp Meade (2010) in presentation of the results in the figurethe estimated fault slip rates predicted from each modelling of blockor plate bounding segments are annotated along each segment Boldblack lines indicate the fault traces and the corresponding labelsgive strike (top) and fault-normal (bottom) slip rates and uncertain-ties Positive signs are for left-lateral and convergent fault motionsand negative signs for dextral slip and tensional faults Blue andmagenta arrows are the observed and residual (observed-modelled)velocities respectively Capital letters in the figure provide faultnames The strengths and weaknesses of each model are discussedin the following paragraphs

10 OJ Perez et al

Figure 10 Estimated fault slip rates relative to South America on plate and block bounding segments of southern Caribbean and northern South America forthe five models discussed in the text Bold black lines indicate the fault traces The corresponding labels give strike (top) and fault-normal (bottom) slip ratesand uncertainties Negative signs indicate dextral and tensile faults positive signs left-lateral and convergent faults Blue and magenta arrows are the observedand residual velocity vectors respectively Faults BS BajosndashEl Soldado P El Pilar S San Sebastian V La Victoria M Morrocoy B Bocono SM SantaMarta E East Andean SCMF South Caribbean Marginal CDB South Caribbean Deformed Belt T Trinidad Model 2 provides the names of several islandsmentioned in the text Ar Aruba Cu Curacao Lr Los Roques Lt La Tortuga Models 1ndash3 are rejected on the basis of geological and geodetic constraintsModel 4 is the block model that best fits the GPS data and geological constraints we used Model 5 is a hybrid model that incorporates slower slipping faults(O Oca and CDB Caribbean Deformed Belt) fixing their slip rates as explained in the text Letters within circles in Models 4 and 5 indicate the blocks thatconfigurate each model whose calculated poles of rotation are given in Table 2

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

10 OJ Perez et al

Figure 10 Estimated fault slip rates relative to South America on plate and block bounding segments of southern Caribbean and northern South America forthe five models discussed in the text Bold black lines indicate the fault traces The corresponding labels give strike (top) and fault-normal (bottom) slip ratesand uncertainties Negative signs indicate dextral and tensile faults positive signs left-lateral and convergent faults Blue and magenta arrows are the observedand residual velocity vectors respectively Faults BS BajosndashEl Soldado P El Pilar S San Sebastian V La Victoria M Morrocoy B Bocono SM SantaMarta E East Andean SCMF South Caribbean Marginal CDB South Caribbean Deformed Belt T Trinidad Model 2 provides the names of several islandsmentioned in the text Ar Aruba Cu Curacao Lr Los Roques Lt La Tortuga Models 1ndash3 are rejected on the basis of geological and geodetic constraintsModel 4 is the block model that best fits the GPS data and geological constraints we used Model 5 is a hybrid model that incorporates slower slipping faults(O Oca and CDB Caribbean Deformed Belt) fixing their slip rates as explained in the text Letters within circles in Models 4 and 5 indicate the blocks thatconfigurate each model whose calculated poles of rotation are given in Table 2

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

GPS and plate kinematics in south Caribbean 11

Figure 10 continued

Model 1 is rejected because the calculated dextral slip rate alongthe Oca fault (100ndash106 mm yrminus1) is significantly larger than theslip rate obtained (lt2 mm yrminus1) from palaeoseismicity studies (Au-demard 1996 Audemard et al 2000) and the block motion analysisof Symithe et al (2015) Indeed changes in direction or magnitudeof velocity vectors across the Oca fault (Figs 8 and 9) are withinthe uncertainties of the data The large motion modelled for the Ocafault also leads to unacceptably large dextral slip (41 mm yrminus1)along the Santa Marta fault which is geologically recognized to bea left-lateral fault (Paris et al 2000 Idarraga-Garcıa amp Romero2010)

Model 2 is rejected because the combined dextral slip rate mod-elled along the San Sebastian (S)ndashLa Victoria (V) fault system isonly 132 mm yrminus1 and thus it is unable to reproduce the observedvelocities (18ndash20 mm yrminus1) on the islands of Aruba (Ar) Curacao

(Cu) Los Roques (Lr) and La Tortuga (Lt) where the residual ve-locities (magenta arrows in Fig 10) are very large in the rangeof 4ndash8 mm yrminus1 Also the model shows 34 mm yrminus1 of dextralmotion along the east-trending section of the Caribbean DeformedBelt which clearly violates the observed surface velocity field andsubsequent elastic model we derived (Fig 5) for the northndashcentralVenezuela and southndashcentral Caribbean regions

Model 3 which includes all faults under consideration is rejectedbecause as in the case of Model 1 contrary to geology the SantaMarta fault shows a large amount (41 mm yrminus1) of right-lateralslip and the calculated slip rate along the Oca fault (106ndash107 mmyrminus1) is significantly higher than the lt2 mm yrminus1 indicated bypalaeoseismicity studies (Audemard 1996 Audemard et al 2000Paris et al 2000 Idarraga-Garcıa amp Romero 2010)

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

12 OJ Perez et al

Models 4 and 5 in Fig 10 are the two block models that resultin velocity fields that best fit the GPS data and geological estimatesof fault slip rates In general velocity residuals (magenta arrowsFig 10) are within measurement uncertainties at most sites Thedifference between the two models is the presence in Model 5 of theeastern section of the Caribbean Deformed Belt (CDB in Fig 10)where we have fixed the total slip at 02 mm yrminus1 (from Symithe et al2015) along the east-striking section of the belt and the inclusionof the Oca fault with a fix dextral slip of 2 mm yrminus1 (Audemard1996 Audemard et al 2000 Symithe et al 2015) Recall that inModels 1ndash4 we simultaneously solve for block rotation rates andfault slip rates whereas in Model 5 we impose and fix a small sliprate on two slow slipping faults namely the Oca fault and the east-striking section of the Caribbean Deformed Belt east of 72W ThusModel 5 is a hybrid model that combines geological and geodeticdata and results taking into account slow slipping faults such asthe Oca and the Caribbean Deformed Belt east of 72W whereasModel 4 takes into consideration only the faults that take up most ofthe relative motion between the Caribbean and the South Americanplates that is the faults that conform the main seismically activeplate boundary between the two plates

Taking into account both Models 4 and 5 in Fig 10 the followingobservations are made (1) Subduction rates along nearly northndashsouth oriented segments of the Lesser Antilles subduction are 194ndash199 mm yrminus1 in agreement with the results reported by Symitheet al (2015) The segments which depart from the northndashsouth strikeshow a mechanically coherent strain partitioning that is dependenton segmentrsquos strike (2) The easterly oriented El Pilar (P) faultin northeastern Venezuela and the plate boundary fault zone goingacross Trinidad (T) and then off-shore to the east both show a dextralslip of 193ndash198 mm yrminus1 in agreement with the geodetic resultsoriginally reported by Perez et al (2001) Weber et al (2001 2009)and also in agreement with the results provided by the kinematicblock motion analysis of Symithe et al (2015) (3) The northwestoriented Los BajosndashEl Soldado (BS in Fig 10) submarine faultsystem located between Venezuela and Trinidad (T) shows dextralslip (137 mm yrminus1) and is transtensional (sim14 mm yrminus1) a typeof motion early suggested by Perez amp Aggarwal (1981) (4) Inboth Models 4 and 5 in Fig 10 the dextral relative plate motion innorthndashcentral Venezuela is partitioned between the San Sebastian(S 17 mm yrminus1) and La Victoria (V 2ndash23 mm yrminus1) faults inagreement with the results of the elastic modelling we presentedin an earlier section of the manuscript (5) Taking into accountboth models the Bocono (B) fault shows up to 102 mm yrminus1 ofdextral slip and a component of convergence in the order of 2ndash6 mm yrminus1 in agreement with the results we derived in the previoussection (6) The Santa Marta (SM) fault shows a left-lateral sliprate of 12ndash20 mm yrminus1 and a convergence rate of 66ndash80 mmyrminus1 depending on the model in agreement with early GPS resultsreported by Trenkamp et al (2002) (7) Subduction rate along theSouth Caribbean Marginal (SCM) fault segments ranges from sim5to sim7 mm yrminus1 (8) Both models indicate that the Morrocoy (M)fault is right-lateral (29ndash48 mm yrminus1) and transtensional (sim4 mmyrminus1) (9) The existence of the Morrocoy boundary is compulsory toexplain the observed velocity field in northwestern Venezuela andnorthern Colombia

D I S C U S S I O N A N D C O N C LU S I O N S

The observed velocity field and corresponding modellings for thesouthndashcentral Caribbean and northndashcentral Venezuela regions east

of the triple junction vertex between the North Andes Caribbean andSouth American plates (Figs 1 2 5 and 10) show that the sim196 mmyrminus1 of dextral relative motion of the Caribbean relative to SouthAmerica is split with sim17 mm yrminus1 taken up by San Sebastian faultand sim26 mm yrminus1 by La Victoria fault This latter rate of motion ishighly consistent with geological estimates (sim275 mm yrminus1) of theQuaternary rate of motion of La Victoria fault zone (Audemard et al2000) near the epicentre of the 1641 earthquake (Fig 2) where thefault is multibranched and is composed of several parallel strandsThis partitioning of slip between subparallel faults is similar tothe phenomenon occurring in the Hispaniola Island in northernCaribbean where the CaribbeanndashNorth America easterly left-lateralslip (18ndash20 mm yrminus1 DeMets et al 2010) is divided among severaleastward-trending faults that compose the plate boundary (Dixonet al 1998 Calais et al 2002 Symithe et al 2015) And in someregards it is the same phenomenon taking place in the southern SanAndreas fault system in California where several subparallel faultsaccommodate plate motions (eg Lindsey amp Fialko 2013)

The slip rate along San Sebastian fault is sim55 times larger thanthe one along La Victoria fault and this may explain the much lowerseismic moment productivity of the La Victoria fault compared toSan Sebastian fault in the last sim4 centuries (Perez et al 1997a)and qualifies the San Sebastian fault as the main strand of the plateboundary in the region If slipping at sim17 mm yrminus1 the sim100 kmlong western segment of this fault (Fig 2) has accumulated about34 m of potential slip since it last broke in 1812 (Audemard 2002Choy et al 2010) sufficient to be released during a large strike-slipearthquake of Mw gt 7 according to the scaling laws derived byHenry amp Das (2001) and Wesnousky (2008)

Northwest of the North AndesndashSouth AmericanndashCaribbeanplatesprime triple junction (Fig 1) the Morrocoy fault zone marks thenortheast boundary of the North Andes plate with the Caribbeanplate (Perez et al 1997b Bird 2003) Focal mechanisms along thisfault system show northwest oriented dextral slip with a small com-ponent of normal faulting about 100 km northwest of the triplejunction (Fig 2) and normal faulting with a strike-slip component(Audemard et al 2005) farther northwest around 12N (Fig 4)Right-lateral transtension is required across the Morrocoy fault be-cause easterly directed GPS velocity vectors (Fig 8) increase inmagnitude and in azimuth as one moves from the North Andesplate to the Caribbean plate As it was described in the previoussection from a kinematic block modelling we estimate the amountof dextral slip in sim3ndash48 mm yrminus1 along faultrsquos strike and about4 mm yrminus1 of normal slip in the direction perpendicular to the faultThese observations are consistent with the observed focal mecha-nisms of seismic events straddling the Morrocoy fault (Figs 2 and 4)and also with geological and geophysical studies carried out in theregion (Bellizzia et al 1976 Beltran 1993 Audemard et al 2000Gorney et al 2007 Escalona amp Mann 2011) which show the ex-istence of active northwest striking Quaternary faults with normaland dextral slip both off-shore and inland Fig 11(a) is a map thatcompiles the known traces (coloured sinuous lines) of faults activesince Tertiary in the regions adjacent to the Morrocoy (M) fault(Beltran 1993 Audemard et al 2000 Gorney et al 2007 Escalonaamp Mann 2011 Castillo et al 2017) Large portions of the Morrocoyboundary have not been seismically explored or the information isin the non-public oil industry domain Fig 11(b) shows a seismicreflection line collected during the BOLIVAR cruise in 2004 (Gor-ney et al 2007 Escalona amp Mann 2011) off-shore the ParaguanaPeninsula (PP Fig 11a) It shows normal faults cutting across the

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

GPS and plate kinematics in south Caribbean 13

Figure 11 (a) Known Tertiary and younger faults (coloured sinuous lines) in western Venezuela and the Leeward Antilles compiled from Beltran 1993 (blue)Gorney et al 2007 and Escalona amp Mann 2011 (red green and brown) and Castillo et al 2017 (purple) PP Paraguana Peninsula Main boundary faults areshown in black (b) Seismic line located southeast of Aruba (Fig 11a) with an interpreted graben structure (Gorney et al 2007 Escalona amp Mann 2011) thatmatches the proposed location of the Morrocoy fault

basement in an area that matches the proposed location of the Mor-rocoy fault Farther southeast the fault is apparently buried beneatha northwest trending thrust belt at the base of the Falcon slope

In the most northwestern regions of the South American conti-nent using teleseismic and microseismicity data we have improvedthe knowledge of the geometry concerning the northeast strikingsoutheast dipping subduction of the Caribbean plate beneath theNorth Andes plate (Figs 6ndash8) and have shown this subduction isoblique to the CaribbeanndashSouth America slip vector Also the mod-elled rate of southeast subduction of the Caribbean plate relative tothe North Andes plate (Fig 10) is rather small (5ndash7 mm yrminus1) whichexplains the low background seismic activity in the region Indeedthe Benioff zone in northwestern Venezuela is better defined onlywhen microearthquake locations are considered This small relativerate of subduction implies that 2000ndash3000 yr of strain accumulationare needed for a potential giant (Mw gt 85) shock with a co-seismic

slip of up to 15 m or more to break the entire 500ndash600 km longinterplate boundary a rupture length typical for events of this size(eg Thatcher 1990 Perez amp Scholz 1997) as it was the case duringthe 2010 Mw 88 Chilean earthquake (eg Vigny et al 2011 Wanget al 2012) At this time scale the seismic history of the regionis completely unknown A smaller yet strong event would need asmaller strain accumulation time For instance it would take around300ndash600 yr to accumulate 2ndash3 m of slip equivalent to the averageslip that would be expected in an Mw 75ndash78 subduction earthquakewith a rupture length on the order of 130 km (eg Henry amp Das2001 Wesnousky 2008)

Southwest of the triple junction the interaction between the NorthAndes and South American plates along the Bocono fault zone andVenezuelan Andes occurs in such a way that slip partitioning takesplace with 9ndash11 mm yrminus1 of dextral motion in the far field of thefault and 2ndash5 mm yrminus1 of shortening across the Venezuelan Andes

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

14 OJ Perez et al

Thus the slip partitioning exhibited by this region resembles that oc-curring in Central California where thrust faulting occurs on planesthat strike near parallel to the San Andreas fault as in the 1983Coalinga shock and related events (Stein amp King 1984 Ekstrom ampEngland 1989 Stein amp Ekstrom 1992) As along the central SanAndreas fault focal mechanisms reported for the area (eg Perezet al 1997b) show pure dextral motion for microseisms associatedwith the Bocono fault itself whereas many microearthquakes lo-cated well away from the fault and along the Andean foothills showpure thrust motion on planes dipping towards the Bocono fault

Several authors have proposed (eg Kellogg amp Bonini 19821985 Kellogg et al 1985 Kellogg amp Vega 1995 Trenkamp et al2002) that dextral motion along the Oriente East Andean andBocono faults in Ecuador Colombia and Venezuela respectively(Fig 1) is accommodating the northward component of motion ofthe North Andes plate that results from a squeezing of the platebetween the Nazca plate and the Oriente and East Andean faultsWhile this idea is acceptable and is consistent with our GPS datathe transtensional right-lateral motion we observe along the Morro-coy fault requires that the resulting eastward component of NorthAndean motion remains less than that of the Caribbean plate Fi-nally the GPS coverage for the southern Caribbean region is beingsignificantly improved thanks to the implementation by UNAVCOof a wide aperture GPS array the COCOnet GPS network cover-ing the whole Caribbean and adjacent regions Thus it is expectedthat refinements in the knowledge of the kinematics of the regionand subsequent quantification of seismic hazards will arise as newobservations are collected

A C K N OW L E D G E M E N T S

New GPS observations in Venezuela were funded by Simon BolıvarUniversity We thank the Colombian Geological Survey for pro-viding seismic data and a series of GPS velocity vectors of sitesin Colombian territory David Mencin provided us with severalunpublished velocity vectors and helped us in the geodetic dataprocessing UNAVCO provided data from sites that belong to theCOCOnet GPS array We are grateful to John P Loveless for his helpto use the software lsquoBlocksrsquo Our former students Yuleika MadrizFreddy Rondon Carlos Moncayo Mizael Bravo and the Physicaland Satellite Geodesy Laboratory of La Universidad del Zulia inMaracaibo Venezuela took care of the new data gathering Com-ments and suggestions made by Editor Juan C Afonso Eric Calaisand by an anonymous reviewer led to a significant improvement ofour work Roger Bilham provided us with GPS data he collected inAruba in 2012 In memoriam of LL who passed away during thepreparation of the manuscript

R E F E R E N C E SAltamimi Z Collileux X amp Metivier I 2011 ITRF2008 an improved

solution of the international terrestrial reference frame J Geod 85 45ndash473

Altamimi Z Metivier L amp Collilieux X 2012 ITRF2008 plate motionmodel J geophys Res 117 B07402 doi1010292011JB008930

Altez R 1998 Cronometrizacion extemporanea Los sismos del 26 deMarzo de 1812 en Caracas y en Merida Rev Geog Vla 39 297ndash326

Audemard FA 1996 Paleoseismicity studies on the Oca-Ancon fault sys-tem northwestern Venezuela Tectonophysics 259 67ndash80

Audemard FA 2002 Ruptura de los grandes sismos historicos venezolanosde los siglos XIX y XX revelados por la sismicidad instrumental contem-poranea in XI Congreso Venezolano de Geofısica Caracas Venezuelapp 8 17ndash20 Noviembre CD format

Audemard FA Machette M Cox J Dart R amp Haller K 2000 Map atscale 12000000 and database of Quaternary Faults in Venezuela and itsoffshore regions Open-File Rep 00ndash0018 US Geol Surv pp 78

Audemard FA Romero G Rendon H amp Cano V 2005 Quaternary faultkinematics and stress tensor along the southern Caribbean from fault-slipdata and focal mechanism solutions Earth-Sci Rev 69 181ndash233

Audemard FE amp Audemard FA 2002 Structure of the Merida AndesVenezuela relations with the South AmericandashCaribbean geodynamic in-teraction Tectonophysics 345 299ndash327

Bellizzia A Pimentel N amp Bajo R 1976 Structural Geology Map ofVenezuela Scale 1 500000 Min Minas e Hidrocarburos

Beltran C 1993 Mapa Neotectonico de Venezuela esc 12000000 Fun-dacion Venezolana de Investigaciones Sismologicas (FUNVISIS)

Bernal-Olaya R Mann P amp Vargas CA 2015 Earthquake tomographicseismic reflection and gravity evidence for a shallowly dipping subduc-tion zone beneath the Caribbean Margin of northwestern Colombia inPetroleum Geology and Potential of the Colombian Caribbean MarginVol 108 pp 247ndash270 eds Bartolini C amp Mann P AAPG Memoir

Bird P 2003 An updated digital model of plate boundaries GeochemGeophys Geosys 4 doi1010292001GC000252

Calais E Mazabraud Y de Lepinay BM Mann P Mattioli G ampJansma P 2002 Strain partitioning and fault slip rates in the northeasternCaribbean from GPS measurements Geophys Res Lett 29 3ndash1-3-4

Camacho E Hutton W amp Pacheco JF 2010 A new look at evidencefor a Wadati-Benioff Zone and active convergence at the north Panamadeformed belt Bull seism Soc Am 100 343ndash348

Castillo V Benkovics L Cobos C Demuro D amp Franco A 2017Perla field the largest discovery ever in Latin America in Giant Fieldsof the Decade 2000ndash2010 Vol 113 pp 141ndash152 eds Merrill RK ampSternbach CA AAPG Memoir

Centeno-Grau M 1900 El Terremoto de 1900 in La Linterna Magicain 1900 The reference is given as is by other authorsCenteno-Grau M

1940 Estudios Sismologicos 300 pp Litografıa del ComercioChiarabba C De Gori P Faccenna C Speranza F Seccia D Dionicio

V amp Prieto GA 2015 Subduction system and flat slab beneath theEastern Cordillera of Colombia Geochem Geophys Geosyst 17 16ndash27

Choy JE 2001 The El Tocuyo Venezuela earthquake of 3 August 1950focal parameters and tectonic implications Geofıs Int 40 285ndash292

Choy JE Palme C Guada C Morandi M amp Klarica S 2010 Macro-seismic Interpretation of the 1812 earthquake in Venezuela using intensityuncertainties and a priori fault-strike information Bull seism Soc Am100 241ndash255

Clark SA Zelt CA Magnani MB amp Levander A 2008 Characterizingthe CaribbeanndashSouth American plate boundary at 64 west using wide-angle seismic data J geophys Res 113 B07401

Colombian Geological Survey 2016 lsquoEarthquake locations National Seis-mological Array rsquo Available at wwwsgcgovco

Colon S et al 2015 The 1900 Mw 76 earthquake offshore northndashcentralVenezuela is La Tortuga or San Sebastian the source fault Mar PetGeol 67 498ndash511

Dach R Hugentobler U Fridez P amp Meindl M 2007 Bernese GPSSoftware Version 50 University of Bern Bern Open Publishing

DANE 2016 lsquoDepartamento Administrativo Nacional de Estadısticas deColombiarsquo Available at httpwwwdanegovco

DeMets C Gordon RG Argus F amp Stein S 1990 Current plate mo-tions Geophys J Int 101 425ndash478

DeMets C Gordon RG amp Argus DF 2010 Geologically current platemotions Geophys J Int 181 1ndash80

Dewey JW 1972 Seismicity and tectonics of western Venezuela Bullseism Soc Am 62 1711ndash1751

Dixon TH Farina F DeMets C Jansma P Mann P amp Calais E 1998Relative motion between the Caribbean and North American plates andrelated boundary zone of deformation from a decade of GPS observationsJ geophys Res 103 15 157ndash15 182

Ekstrom G amp England PC 1989 Seismic strain rates in regions of dis-tributed continental deformation J geophys Res 94 10 231ndash10 257

Ernst A 1878 Eartquake in Venezuela Nature 18

GPS and plate kinematics in south Caribbean 15

Escalona A amp Mann P 2011 Tectonics basin subsidence mechanismsand paleogeography of the Caribbean-South American plate boundaryzone Mar Pet Geol 28 8ndash39

Fielder G 1961 Areas afectadas por terremotos en Venezuela VenezGeol Conf Bol Geol 4 1791ndash1810

FUNVISIS 2016 lsquoVenezuelan Foundation for Seismological ResearchEarthquake Catalogsrsquo Available at httpwwwfunvisisgobve

Grases J 1980 Investigacion sobre los sismos destructores que han afec-tado el occidente y centro-norte de Venezuela Tech Rep INTEVEP p301

Gorney D Escalona A Mann P amp Magnani MB BOLIVAR StudyGroup 2007 Chronology of Cenozoic tectonic events in westernVenezuela and the Leeward Antilles based on integration of offshoreseismic reflexion and on-land geology AAPG Bull 91 653ndash684

Guzman O et al 2013 10Be dating of river terraces of Santo Domingo riveron Southeastern flank flank of the Merida Andes Venezuela tectonic andclimatic implications J South Am Earth Sci 48 85ndash96

Henry C amp Das S 2001 Aftershocks of large shallow earthquakes faultdimensions aftershock area expansion and scaling relations Geophys JInt 147 272ndash293

Idarraga-Garcıa J amp Romero J 2010 Neotectonic study of the SantaMarta Fault System western foothills of the Sierra Nevada de SantaMarta Colombia J South Am Earth Sci 29 849ndash860

Instituto Nacional de Estadısticas (INE) 2011 lsquoXIV Censo Nacional dePoblacion y Viviendarsquo Available at httpwwwinegovve

ISC 2013 lsquoInternational Seismological Centre on-line bulletin rsquo Available at httpwwwiscacuk

Jordan TH 1975 The present-day motion of the Caribbean plate J geo-phys Res 80 4433ndash4440

Jouanne F Audemard F Beck C Van Welden A Ollarves R amp ReinozaC 2011 Present-day deformation along the El Pilar Fault in easternVenezuela evidence of creep along a major transform boundary J Geo-dyn 51 398ndash410

Kellogg JN 1984 Cenozoic tectonic history of the Sierra de PerijaVenezuela-Colombia and adjacent basins Geol Soc Am Mem 162239ndash261

Kellogg JN amp Bonini WE 1982 Subduction of the Caribbean plate andbasement uplifts in the overriding South American plate Tectonics 1251ndash276

Kellogg JN amp Bonini WE 1985 Reply to comment on ldquoSubduction of theCaribbean plate and basement uplifts in the overriding South Americanplaterdquo by C Schubert Tectonics 4 785ndash790

Kellogg JN amp Vega V 1995 Tectonic development of Panama CostaRica and the Colombian Andes constraints from Global PositioningSystem geodetic studies and gravity Geol Soc Am Spec Pap 29575ndash90

Kellogg JN Ogujiofor IJ amp Kansaka DR 1985 Cenozoic tectonics ofthe Panama and North Andes block in Memorias VI Congreso Latino-Americano de Geologıa pp 34ndash49

Kroehler ME Mann P Escalona A amp Christeson GL 2011 LateCretaceous-Miocene diachronous onset of back thrusting along the SouthCaribbean deformed belt and its importance for understanding processesof arc collision and crustal growth Tectonics 30 TC6003

Lindsey EO amp Fialko Y 2013 Geodetic slip rates in the southern SanAndreas Fault system effects of elastic heterogeneity and fault geometryJ geophys Res 118 689ndash697

Loveless JP amp Meade BJ 2010 Geodetic imaging of plate motions sliprates and partitioning of deformation in Japan J geophys Res 115

Magnani MB Zelt CA Levander A amp Schmitz M 2009 Crustalstructure of the South AmericanndashCaribbean plate boundary at 67W fromcontrolled source seismic data J geophys Res 114 B02312

Malave G amp Suarez G 1995 Intermediate depth seismicity in northernColombia and western Venezuela and its relationship to Caribbean platesubduction Tectonophysics 14 617ndash628

Manaker DM et al et al 2008 Interseismic plate coupling and strainpartitioning in the northeastern Caribbean Geophys J Int 174 889ndash903

McCaffrey R 2002 Crustal block rotations and plate coupling in PlateBoundary Zones pp 101ndash122 eds Stein S amp Freymueller J AmericanGeophysical Union

Meade BJ et al 2002 Estimates of seismic potential in the Marmara Searegion from block models of secular deformation constrained by GlobalPositioning System measurements Bull seism Soc Am 92 208ndash215

Meade BJ amp Loveless JP 2009 Block modeling with connected fault-network geometries and a linear elastic coupling estimator in sphericalcoordinates Bull seism Soc Am 99 3124ndash3139

McKenzie DP amp Morgan WJ 1969 Evolution of triple junctions Nature224 125ndash133

Mora-Paez H Giraldo-Londono L Cardona-Piedrahita L Pelaez-Gaviria J amp Lizarazo S 2016a Velocidades geodesicas horizontalesde Colombia Grupo de Investigaciones Geodesicas Espaciales ProyectoGeoRED Servicio Geologico Colombiano Documento Interno Bogota

Mora-Paez H Mencin DJ Molnar P Diederix H Cardona-PiedrahitaL Pelaez-Gaviria J-R amp Corchuelo-Cuervo Y 2016b GPS velocitiesand the construction of the Eastern Cordillera of the Colombian AndesGeophys Res Lett 43 8407ndash8416

Okada Y 1985 Surface deformation due to shear and tensile faults in ahalf-space Bull seism Soc Am75 1135ndash1154

Pacheco JF amp Sykes LR 1992 Seismic moment catalog of large shallowearthquakes 1900 to 1989 Bull seism Soc Am 82 1306ndash1349

Paris G Machette MN Dart RL amp Haller KM 2000 Map andDatabase of Quaternary Faults and Folds in Colombia and its OffshoreRegions Open-File Rep 00ndash0284 US Geol Surv pp 61

Pennington WD 1981 Subduction of the eastern Panama basin and seis-motectonics of northwestern South America J geophys Res 80 10753ndash10 770

Perez OJ et al 2001 Velocity field across the southern Caribbean plateboundary and estimates of CaribbeanSouth American plate motion usingGPS geodesy 1994ndash2000 Geophys Res Lett 28 2987ndash2990

Perez OJ et al 2011 Campo de velocidades en el occidente de Venezuelacomponente lateral derecha asociada a la falla de Bocono y componenteconvergente perpendicular a los Andes Interciencia 36 39ndash44

Perez OJ amp Aggarwal YP 1981 Present-day tectonics of the southernCaribbean and northeastern Venezuela J geophys Res 86 791ndash804

Perez OJ amp Mendoza JS 1998 Sismicidad y tectonica en Venezuela yareas vecinas Fıs Tierra 10 87ndash110

Perez OJ amp Scholz CH 1997 Long-term seismic behavior of the focaland adjacent regions of great earthquakes during the time between twosuccessive shocks J geophys Res 102 8203ndash8216

Perez OJ Jaimes M amp Garciacaro E 1997b Microseismicity evidencefor subduction of the Caribbean plate beneath the South American platein northwestern Venezuela J geophys Res 102 17 875ndash17 882

Perez OJ Sanz C amp Lagos G 1997a Microseismicity tectonics andseismic potential in southern Caribbean and northern Venezuela J Seis-mol 1 15ndash28

Reilinger R et al 2006 GPS constraints on continental deformationin the Africa-Arabia-Eurasia continental collision zone and implica-tions for the dynamics of plate interactions J geophys Res 111doi1010292005JB004051

Reinoza C Jouanne F Audemard FA Schmitz M amp Beck C 2015Geodetic exploration of strain along the El Pilar fault in northeasternVenezuela J geophys Res 120 1993ndash2013

Rial JA 1978 The Caracas Venezuela earthquake of July 1967 a multiplesource event J geophys Res 83 5405ndash5414

Saria E Calais E Stamps D Delvaux D amp Hartnady C 2014 Present-day kinematics of the East African rift J geophys Res 119 3584ndash3600

Schmitz M et al 2008 Crustal thickness variations in Venezuela fromdeep seismic observations Tectonophysics 459 14ndash26

Sievers W 1905 Das Erdbeben in Venezuela von 29 Oktober 1900 Fest zurFeier des 70 Geburtstages von J J Rein Jahrb Veroffnet Geog Vereinzu Bonn 35ndash50

Silver EA Case JE amp MacGillvary HJ 1975 Geophysical study of theVenezuelan borderland Bull geol Soc Am 86 213ndash226

16 OJ Perez et al

Symithe S Calais E de Chabalier JB Robertson R amp Higgins M2015 Current block motions and strain accumulation on active faults inthe Caribbean J geophys Res 120 3748ndash3774

Syracuse EM Maceira M Prieto GA Zhang H amp Ammon CJ2016 Multiple plates subducting beneath Colombia as illuminated byseismicity and velocity from the joint inversion of seismic and gravitydata Earth planet Sci Lett 444 139ndash149

Stein R amp King G 1984 Seismic potencial revealed by surface folding1993 Coalinga California earthquake Science 224 869ndash872

Stein RS amp Ekstrom G 1992 Seismicity and geometry of a 110-km-longblind thrust fault 2 Synthesis of the 1982ndash1985 earthquake sequence Jgeophys Res 97 4865ndash4884

Talwani M Windish CC Stoffa PL Buhl P amp Houtz RE 1977Multichannel seismic study of the Venezuelan basin and the Curacaoridge in Island Arcs Deep Sea Trenches and Back-Arc Basins MauriceEwing Series pp 83ndash98 eds Talwani M amp Pitman WC III AmericanGeophysical Union

Thatcher W 1990 Order and diversity in the modes of Circum-Pacificearthquake occurrence J geophys Res 95 2609ndash2623

Toto EA amp Kellogg JN 1992 Structure of the Sinu-San Jacinto foldbeltmdashan active accretionary prism in northern Colombia J South AmEarth Sci 5 211ndash222

Trenkamp R Kellogg JN Freymueller JT amp Mora HP 2002 Wideplate margin deformation southern Central America and northwesternSouth America CASA GPS observations J South Am Earth Sci 15157ndash171

van der Hilst R amp Mann P 1994 Tectonic implications of tomographicimages of subducted lithosphere beneath northwestern South AmericaGeology 22 451ndash454

Vigny C et al 2011 The 2010 Mw 88 Maule megathrust earthquake ofCentral Chile monitored by GPS Science 332 1417ndash1421

Wang L et al 2012 Coseismic slip of the 2010 Mw 88 Great MauleChile earthquake quantified by the inversion of GRACE observationsEarth planet Sci Lett 335ndash336 167ndash179

Weber J et al 2001 GPS estimate of relative motion between the Caribbeanand South American plates and geological implications for Trinidad andVenezuela Geology 29 75ndash78

Weber JC Saleh J Balkaransingh S Dixon T Ambeh W Leong TRodriguez A amp Miller K 2009 Triangulation-to-GPS and GPS-to-GPSgeodesy in Trinidad West Indies neotectonics seismic risk and geologicimplications Mar Pet Geol 28 200ndash211

Wesnousky SG 2008 Displacement and geometrical characteristics ofearthquake surface ruptures issues and implications for seismic-hazardanalysis and the process of earthquake rupture Bull seism Soc Am 981609ndash1632

Wesnousky SG Araguren R Rengifo M Owen LA Caffee MWMurari MK amp Perez OJ 2012 Toward quantifying geomorphic ratesof crustal displacements landscape development and the age of glaciationin the Venezuelan Andes Geomorphology 141ndash142 99ndash113