lable at ScienceDirect

Journal of Structural Geology 58 (2014) 8e21

Contents lists avai

Journal of Structural Geology

journal homepage: www.elsevier .com/locate/ jsg

Evolution of the stress and strain fields in the Eastern Cordillera,Colombia

Obi Egbue a,c, James Kellogg a,*, Hector Aguirre b, Carolina Torres a,c

aDepartment of Earth and Ocean Sciences, University of South Carolina, Columbia, SC, USAb Equion Energy, Bogota, ColombiacBP, Houston, TX, USA

a r t i c l e i n f o

Article history:Received 17 December 2012Received in revised form28 August 2013Accepted 17 October 2013Available online 30 October 2013

Keywords:Stress mapGlobal Positioning SystemBorehole breakoutsEscapeFracturesEastern Cordillera

* Corresponding author.E-mail address: kellogg@sc.edu (J. Kellogg).

0191-8141/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.jsg.2013.10.004

a b s t r a c t

This work integrates stress data from Global Positioning System measurements and earthquake focalmechanism solutions, with new borehole breakout and natural fracture system data to better understandthe complex interactions between the major tectonic plates in northwestern South America and toexamine how the stress regime in the Eastern Cordillera and the Llanos foothills in Colombia has evolvedthrough time. The dataset was used to generate an integrated stress map of the northern Andes and topropose a model for stress evolution in the Eastern Cordillera. In the Cordillera, the primary present-daymaximum principal stress direction is WNWeESE to NWeSE, and is in the direction of maximumshortening in the mountain range. There is also a secondary maximum principal stress direction that is EeW to ENEeWSW, which is associated with the northeastward “escape” of the North Andean block,relative to stable South America. In the Cupiagua hydrocarbon field, located in the Llanos foothills, thedominant NNEeSSW fractures are produced by the Panama arceNorth Andes collision and range-normalcompression. However, less well developed asymmetrical fractures oriented EeW to WSWeENE andNNWeSSE are also present, and may be related to pre-folding stresses in the foreland basin of the CentralCordillera or to present-day shear associated with the northeastward “escape” of the north Andean block.Our study results suggest that an important driver for orogenic deformation and changes in the stressfield at obliquely convergent subduction zone boundaries is the arrival of thickened crust, such as islandarcs and aseismic ridges, at the trench.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Present-day stress in the northern Andes (Fig. 1) is a result of theconvergence of the South American, Caribbean, and Nazca platesand the Panama microplate. Global Positioning System (GPS)measurements (Trenkamp et al., 2002) show that the North Andeanblock is escaping toward the northeast relative to the stable SouthAmerican plate along a transpressive system of faults located alongthe Merida Andes, Eastern Cordillera, and Ecuadorian Andes(Fig. 2). Simultaneously, in the Eastern Cordillera of Colombia, theon-going Panama ArceSouth America collision is driving NWeSEpermanent shortening and rapid uplift of the range. Low-anglesubduction of the buoyant Caribbean plate is also driving perma-nent NWeSE shortening in the Merida Andes and Eastern Cordil-lera. Knowledge of the evolving stress regime in the EasternCordillera is key to understanding the geodynamic processes at

All rights reserved.

work within the region, as well as the evolution of structural trapsand fracture systems for hydrocarbons. The present-day geologyshows the cumulative results of all the deformational events thathave affected the region. To better understand both the neostressand paleostress regimes in the Eastern Cordillera, we have analyzeddata from a variety of stress indicators. The present-day stressregime can be inferred from stress-induced borehole breakouts, aswell as GPS vectors, and earthquake focal mechanism solutions. Thepaleostress is inferred from natural fracture systems and theorientation of structural features in the Llanos foothills. It can alsobe inferred from geologically datedmeasurements of north Andeannortheastward “escape”, the timing of the Panama arc collision, andage data for the uplift of the Eastern Cordillera. This study suggeststhat the range-normal compressive stress field responsible fordeforming rocks and producing the structural traps and fracturesystems in the Eastern Cordillera in the last 12 Ma has evolved toinclude a range-parallel shear component in the last 2 Ma.

Paleostress analysis is an important tool in oil and gas explo-ration, development, and reservoir simulation, constraining thedevelopment of subsurface fracture systems models. Subsurface

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 9

fracture systems in hydrocarbon reservoirs usually result frompaleostress that is unrelated to the present-day stress (Engelder,1987). However, the neostress regime can strongly impact thepermeability of fractures that formed in a different, older stressregime. It is important therefore to understand how the stressregime has evolved through time.

The North Andean margin provides an excellent Present-dayexample of an evolving stress field associated with obliqueconvergence and arc collision at a subduction zone. The strainassociated with oblique convergence at subduction zones is oftenpartitioned between trench-normal contraction on thrust faultsand trench-parallel strain on strike-slip faults (Fitch, 1972; Jarred,1986; McCaffrey, 1994). In general, fore arc slivers form over theregion of interplate coupling and are driven along strike by thebasal shear (Platt, 1993; McCaffrey et al., 2000). A volcanic arc canhelp the partitioning process by localizing the margin-parallelshear strain in a zone of thermally induced lithospheric weakness(Beck, 1983). In the North Andean margin, however, the MioceneePresent collisionwith the Panama arc has increased coupling on theplate boundary, inverted the back arc basin, and displaced themargin-parallel shear strain to the back arc. In this study wecombine GPS measurements and earthquake focal mechanismswith geologic measurements of deformation and new boreholebreakout and well fracture data to develop a model for the evolu-tion of the strain field in the Eastern Cordillera over the last 14 Ma.Our study results suggest that the main driver for orogenic defor-mation and changes in the stress field was the arrival of thickenedcrust at the Ecuador and Colombia trenches, in particular, thearrival of the Panama arc at 12e15 Ma and the Carnegie Ridge at2 Ma.

2. Present-day stress field data

2.1. Global Positioning System

GPS studies are an important method for studying the kine-matics of plate boundary zones, and intraplate deformation. Thevelocity vectors used for this study (Fig. 1) are from the Central andSouth American (CASA) GPS project (Trenkamp et al., 2002) whichmeasured plate motions and crustal deformation of the four majorplates in the region. The main results from the study can be sum-marized as follows. In northern Colombia the Caribbean plate isconverging with stable South America in an EeSE direction at a rateof 20 � 2 mm/a (Fig. 2). Panama is actively colliding with north-western South America at a rate of approximately 25 mm/a. Thelarge eastward components of the GPS vectors in northernColombia (Fig. 1, CART, MONT, RION, MZAL) suggest transfer ofmotion from the Panama microplate. The oceanic Nazca plate andthe aseismic Carnegie Ridge are rapidly subducting at the EcuadoreColombia trench at a rate of 58 � 2 mm/a relative to stable SouthAmerica. Approximately 50% of the NazcaeSouth America conver-gence is locked at the subduction interface, producing elastic strainin the overriding South American plate. GPS measurements alsoshow that the northern Andes is escaping northeastward at a rate of6 � 2 mm/a relative to stable South America (Fig. 1, e.g., BOGO,PASTO, BUCM, MERI, VDUP, MZAL).

2.2. Borehole breakouts

Borehole breakouts indicate the horizontal stress field at inter-mediate depths; usually less than 5 km. Borehole breakouts areconsistently oriented elongations in the wellbore cross-section thatform as a result of differential stress applied to an initially circularborehole. Borehole breakouts were first observed by Cox (1970) inthe Wapiabi shales of west-central Alberta. However, it was not

until later in the early 1980’s that theywere attributed to the in-situstress field (Gough and Bell, 1981; Hickman et al., 1982; Plumb,1982; Cox, 1983). A borehole breakout occurs when the stressesaround the borehole exceed that required to cause compressivefailure of the borehole wall (Zoback et al., 1985). The developmentof intersecting conjugate shear planes causes spalling of the bore-hole wall, resulting in the enlargement of thewell bore. In a verticalborehole, the wall shear stress is greatest in the direction of theminimum horizontal stress (Sh). The long axes of borehole break-outs therefore, are oriented approximately perpendicular to themaximumhorizontal stress direction (SHmax) (Plumb and Hickman,1985).

The borehole breakout data presented in this work (Fig. 1) isfrom four-arm caliper, high resolution dipmeter logs, and wascollected from 60 industry wells in the Eastern Cordillera andsurrounding areas. Each well was assigned a quality ranking basedon standard deviation, number of breakouts, and cumulative lengthof broken out intervals using the rating scheme of Zoback andZoback (1989). The breakout symbols plotted on the map repre-sent the location and direction of SHmax of each of the wells. Thebars show the direction of the mean azimuth of SHmax and thelength of the bars corresponds to the quality ranking of thebreakout data, which ranges from A to D. An “A” ranking representsa well with many breakouts which have tightly clustered elonga-tion azimuths. A “D” ranking represents a well with very fewbreakouts, multiple azimuthal populations, and/or a large standarddeviation for the mean azimuth. In general, borehole breakoutorientations are bimodal, in the sense that data between 180� and360� are equivalent to those between 0� and 180�. Boreholebreakouts in the region show significant scatter, but we proposethat the data fall into three major subsets by region: northern,central, and southern. The northern subset is made up of wellslocated above 6� N. Though a number of these wells show variableSHmax orientation, the majority have EeW trending SHmax. Thecentral subset is made up of wells located between 2� N and 6� N.Most of the wells in this subset have WNWeESE trending SHmax.Wells in the southern segment (0�e2� N) have a NWeSE and lesscommonly NNEeSSW trending SHmax. It is important to point outthat there are no clear patterns of azimuthal variations with depthin the individual wells or in the data subsets except for wells in theLlanos basinwhere the data shows a change in breakout orientationat about 7800 ft (Torres Fleischmann,1992). Below this depth, thereare two main breakout population clusters. These correspond toSHmax orientations of 110� and 070�.

2.3. Earthquake focal mechanism solutions

Some of the stress data in Fig. 1 are derived from singleearthquake focal mechanism solutions (FMS) from the WorldStress Map database (Reinecker et al., 2005). An earthquake focalmechanism solution is constructed using the radiation pattern ofseismic waves generated by an earthquake. The solution containstwo perpendicular planes that represent the fault plane and theauxiliary plane, and divide the volume into compressional andextensional quadrants. The orientation of the compressional (P),intermediate (B), and extensional (T) axes are determined by theorientations of these planes. It is important to point out, that themoment tensor axes of earthquake focal mechanisms are not equalto the principal stress axes. The only conclusion that can be madeis that the maximum principal stress lies within the dilatationalquadrant of the focal mechanism (McKenzie, 1969). The principalstrain axes of earthquake focal mechanism solutions (P-axis, B-axis, and T-axis) are used as proxies for the orientation of theprincipal stresses axes s1, s2, and s3. To account for the un-certainties in this first-order approximation, the quality of the

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2110

Fig. 2. Schematic 3-D tectonic model for the northern Andes showing principal plate boundaries (after Egbue and Kellogg, 2010). Cross-section through 2� S latitude shows thesubducting Nazca plate and the detached North Andean block. Arrows show Present-day GPS-determined relative velocities of major plates (white) and the North Andes (black)relative to stable South America (Trenkamp et al., 2002).

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 11

stress orientations derived from FMS in the World Stress Mapdatabase is limited to a C-quality regardless of the size of theearthquake and how well the focal mechanism is constrained. Thisquality ranking assumes a deviation of up to �25�. Focal mecha-nism solutions located in the vicinity of a plate boundary and withkinematics similar to the plate boundary, are omitted and flaggedas possible Plate Boundary Events (PBE) indicating that the P-, B-,and T-axes of the focal mechanism solutions might predominantlyreflect the geometry and kinematics of the plate boundary, ratherthan the orientation of the regional stress field. The accuracy ofstress derivations from the FMS presented in this work are limitedby the fault-plane ambiguity and the coefficient of friction (Barthet al., 2008).

SHmax orientations determined from FMS shown in Fig. 1 appearto fall into two main subsets: showing NWeSE and ENEeWSWmaximum compressive stress directions (Ego et al., 1996; Corredor,2003; Cortés and Angelier, 2005). The NWeSE compression hasbeen associated with CaribbeaneSouth American convergence andPanamaeSouth America collision (Kellogg and Vega, 1995; Taboadaet al., 2000; Trenkamp et al., 2002; Corredor, 2003; Colmenaresand Zoback, 2003; Acosta et al., 2004), and the ENEeWSWcompression has been ascribed to NazcaeSouth America conver-gence and Carnegie Ridge subduction at the ColombiaeEcuadortrench (Ego et al., 1996; Gutscher et al., 1999; Trenkamp et al., 2002;Corredor, 2003; Colmenares and Zoback, 2003; Egbue and Kellogg,2010).

Fig. 1. Stress map of the northern Andes, showing GPS vectors (Trenkamp et al., 2002), andand Torres Fleischmann, 1992) and earthquake focal mechanisms (Reinecker et al., 2005).lengths of the stress bars indicate data quality.

3. Paleostress field data

The paleostress field in an orogenic belt can be inferred from thegeologic record, including fault and fold orientations, age-constrained fault displacements, rock deformation studies, andnatural fracture orientations. Here we summarize structuralmodels for the Eastern Cordillera and present new natural fracturedata.

3.1. Structure of the Eastern Cordillera

The Eastern Cordillera has been generally interpreted as a wideMesozoic extensional basin that evolved into a foreland basin forthe Central Cordillera, and was subsequently inverted by Andeancompression and orogeny (Colletta et al., 1990; Dengo and Covey,1993; Cooper et al., 1995). Based on a retrodeformed cross-section through the Eastern Cordillera, Colletta et al. (1990) pro-posed that the Cordillera was formed by the inversion of two deepUpper JurassiceLower Cretaceous basins during Mio-Pliocenetimes. In this model, the structure is controlled by low anglethrusts and reactivated normal faults that constitute frontal ramps.They estimated a total shortening of at least 105 km with adecollement at a depth of about 20 km. Toro et al. (2004) obtainedsimilar results along two partial volume-balanced transects of theCordillera. Dengo and Covey (1993) proposed that basement-detached (“thin-skinned”) shortening was followed by uplift on

inferred directions of maximum horizontal stress from borehole breakouts (This studyGPS velocity vectors are relative to stable South America at 95% confidence level. The

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2112

high-angle basement involved reverse faults (“thick-skinned”)deformation and estimated approximately 150 km of shortening(40%) from a regional retrodeformable cross-section. Cooper et al.(1995) proposed that deformation in the Llanos foothills was pre-dominantly inversion of pre-existing extensional faults bybasement-involved “thick skinned” listric reverse faults. Theyestimated only 68 km of shortening.

Egbue and Kellogg (2012) studied the three-dimensionalstructural evolution of the Piedemonte block in the NEeSWtrending central Llanos foothills, the frontal thrust zone along theeastern flank of the Eastern Cordillera (Fig. 3). NWeSE directedcompression has formed a duplex zone containing as many as fiveeast-verging stacked thrust sheets. Using 2-D and 3-D seismicreflection data, surface geology, and well data, Egbue and Kellogg

Fig. 3. Geologic map of the Llanos foothills on the eastern flank of the Eastern Cordillera shthe Piedemonte block and the Cupiagua and Cusiana fields (BP Exploration Company Colom

(2012) modeled approximately 4 km (9%) NWeSE shortening inthe Late Oligocene (25 Ma) followed by 13e19 km (30e40%) NWe

SE shortening in the last 10 million years.Based on paleostress indicators (fault and fold orientations),

Cortés et al. (2005) present evidence of ENEeWSW directedcompression and shear stresses in the present day western flankand center of the Eastern Cordillera during the MaastrichtianeLatePaleocene, prior to the uplift of the Eastern Cordillera. Geologicindicators of ENEeWSW compression and right-lateral shear in theEastern Cordillera in the last 2 Ma are largely obscured by theongoing NWeSE compression. The timing of this northeastward“escape” is inferred from geological observations in Ecuador,southern Colombia, and Venezuela (compiled in Egbue and Kellogg,2010). The ongoing ENEeWSW compression is also evidenced by

owing locations of several major hydrocarbon features in the Llanos foothills, includingbia). Insets show the locations of Figs. 5 and 11.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 13

present-day GPS measurements, earthquake focal mechanisms,and bore-hole breakout data.

3.2. Natural fractures

Natural fractures are typically present in all rocks to varyingdegrees. In fold and thrust belt regions, the rocks are even moreintensely fractured. These fractures are developed and/or modifiedduring deformation in well consolidated rocks. Most of the tight,well-cemented reservoirs, like those in the flanks of the EasternCordillera, have natural fractures which are important features thatcontrol fluid-reservoir dynamics. Subsequent stress regimes canstrongly affect the permeability of ancient fractures that formed ina different, older stress regime. The propagation paths for fracturesin a geologic setting are strongly influenced by the state of stressand are only weakly dependent on the rock fabric (Olson andPollard, 1989). Vertical fractures propagate normal to the leastprincipal stress, thus, they follow the path of the stress field presentat the time of their propagation (Engelder and Geiser, 1980). Byexamining a map of fracture-set orientation, we can infer the dif-ferential stress that was active during their propagation.

In the Cupiagua hydrocarbon field (Fig. 3) on the eastern flank ofthe Eastern Cordillera, cores and image logs (Ultrasonic BoreholeImager-UBI and Fullbore Formation MicroImager-FMI) show that

Fig. 4. Stratigraphic summary diagram for the Llanos foothills showing typical gamma ray lrocks, and regional seals. On the left is a schematic cross section for the structure of the Cu

natural fractures are present at different stratigraphic intervalsincluding themain reservoirs. Fractures are amodifier of reservoirs,enhancing permeability when they occur open and connected, andreducing permeability when they are closed. In the reservoir in-tervals, fractures with a low acoustic amplitude and high transittime (i.e. dark on UBI images) are dominant, and are interpreted asbeing open or partially open at the borehole wall (Adams et al.,1999). A high transit time suggests an aperture at the boreholewall and a mud or clay fill which gives rise to low acousticimpedance. However, clay smeared fractures are rare in cores fromCupiagua; hence the fractures are interpreted as being filled withdrilling mud. Fractures with high acoustic amplitude are rare. UBIimages show acoustic amplitude contrast which is a response tochanges in acoustic impedance at the borehole wall, which inclastic sequences is related mainly to changes in matrix porosityand lithology. Within the Cupiagua field, open fractures commonlyoccur in proximity to cataclastic faults suggesting that fracturepropagation has a genetic relationship with major deformationepisodes. The mean fracture density varies within the reservoirunits, and ranges between 0.11e0.62 fractures per foot (Adamset al., 1999). A stratigraphic summary diagram for the Llanos foot-hills (Fig. 4) shows the key lithostratigraphic units/formations,reservoir units, unconformities, and detachment levels. In general,the Barco formation is more intensely fractured than the Mirador

og response, the key lithostratigraphic units/formations, proven reservoir rocks, sourcepiagua field. Stratigraphic information is from BP Exploration Company Colombia.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2114

and Guadalupe Formations. It is important to point out that thevariations in fracture density may not be solely a result of geologicalfactors. It is also dependent on variations in UBI image quality.

Based on log images (UBI and FMI) from wells in the Cupiaguafield, we recognize three main families of fractures in the Cupiaguafield (Fig. 5). The data was classified, correlated, and plotted as rosediagrams to group fracture events into clusters/families. A fracturedip vs depth plot from UBI for the wells (Fig. 6) shows that most ofthe natural fractures present in the reservoir between 12,000 and16,000 ft are steeply dipping to subvertical (40e90�) suggestingthat the maximum principal stress directions are subhorizontal.Only one anomalous well shows a wide dispersion of fracture dip

Fig. 5. Fracture strike rose diagram plot for the northern, central, and southern Cupiagua fie

angles below 15,000 ft. Stearns (1967) observed that infold struc-tures in the western United States, conjugate and tensile fracturesform sets that are oriented in varying directions relative to beddingand the fold hinge line (Fig. 7). In Cupiagua, a NWeSE trendingfracture set is widespread throughout the field, and is dominant inthe southern portion of the field (Fig. 5). These fractures are alignedparallel to the present-day regional stress maximum. An EeW toWNWeESE family of fractures is approximately orthogonal to theaxis of the structure. This family of fractures is dominant in thenorthern segment of the field, but these fractures also occur in thecentral and southern segments. This fracture set corresponds toStearns’ fracture type 1 (Fig. 7a). A third NNEeSSW family of

ld. The Cupiagua hydrocarbon field is located in Fig. 3. Black dots show well locations.

Fig. 6. Fracture dip/depth plot from UBI for Cupiagua field wells shown in Fig. 5. Different symbols represent different wells. Note that most of the natural fractures present in thereservoir between 12,000 and 16,000 ft depths are steeply dipping to subvertical (40e90�).

Fig. 7. Schematic diagrams showing (a) fractures types associated with folding, afterStearns (1967) and (b) an example of an asymmetrical fracture unrelated to folding.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 15

fractures is common in the central portion of the field (Fig. 5), and isalso present in the southern segment. These type 2 fractures runparallel to the axis of the Cupiagua structure and occur in both thecrest and flank regions.

4. Discussion

4.1. Present-day stress

The regional stress map (Fig. 1) shows that the primarypresent-day maximum principal stress direction in the EasternCordillera is WNWeESE to NWeSE. This direction is roughlyorthogonal to the mountain range and in the direction ofmaximum shortening. Colmenares and Zoback (2003) proposedthat the negatively buoyant Caribbean slab may be driving therotation of SHmax to the ESE in the direction of relative subductionunder the northern Andes. The WNWeESE stress direction is alsothe PanamaeNorth Andes relative convergence vector when cor-rected for the 6 mm/a northeastward “escape” of the North Andes,and the magnitude of the shortening required to build the EasternCordillera (68e150 km, 3.1 Structure of the Eastern Cordillera),suggests that most of the continuing shortening and uplift of therange is associated with the ongoing PanamaeNorth Andes colli-sion. Close to the Panama block indenter on the western side ofthe northern Andes, the collision-related stresses are particularlystrong. For example, south of MZAL, the 1999 Armenia earthquake(Fig. 1, focal mechanism at 4.4� N, 284.3� E) may have beenlocalized left-lateral slip on the NNE-trending Romeral fault sys-tem related to the PanamaeNorth Andes collision (Pulido, 2003).We interpret this as Panama collision-produced shear on thebroad western flank of the northeastward “escaping” North Andesblock.

A secondary present-day maximum principal stress direction inthe Eastern Cordillera is EeW to ENEeWSW. This is particularlypronounced in the northern termination of the Eastern Cordillera,the Bocono fault system in Venezuela, and in southern Colombiaand Ecuador. We propose that this stress direction is associatedwith the subduction of the Carnegie Ridge and the northeastward“escape” of the North Andes relative to stable South America. Thestress is particularly noticeable at the northern end of the Eastern

Fig. 8. Eastern Cordillera uplift, shortening, and escape history (a) Uplift history of the Eastern Cordillera. Estimated paleoelevation of the Eastern Cordillera (black squares witherror bars after Wijninga, 1996; Gregory-Wodzicki, 2000) and tectonic uplift (red lines). Tectonic uplift is estimated from present day total uplift (10 km) for erosion rates of 0.5 and1.0 mm/a. b) Estimated shortening rates for the Eastern Cordillera since PanamaeSouth America collision began and northeastward “escape” rates for the North Andean block(Egbue and Kellogg, 2010). Shortening rates proportional to tectonic uplift rates are estimated for 60 km and 120 km shortening. The present-day PanamaeSouth Americaconvergence rate is approximately 30 mm/yr (Trenkamp et al., 2002). (For interpretation of the references to color in this figure legend, the reader is referred to the web version ofthis article.)

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2116

Cordillera where a left-stepping restraining bend produces per-manent compressional deformation.

4.2. Paleostress

During the Late Jurassic and Cretaceous periods, the area of thepresent-day Eastern Cordillera was a rift basin with approximatelyWNWeESE oriented extension (minimum principal stress) (Ojeda,1996; Sarmiento-Rojas et al., 2006). Based on low-temperaturethermochronometry and seismic reflection data from the middleMagdalena Valley and field paleostress indicators, Parra et al.(2012) and Cortés et al. (2005) claim evidence for Paleocene toEocene ENEeWSW contraction on the western flank of the presentEastern Cordillera.

4.2.1. Oligocene stressesBy Oligocene time, uplift and shortening of the Central

Cordillera led to uplift in the foreland basin (now the westernfoothills of the Eastern Cordillera) (Horton et al., 2010; Morenoet al., 2011). Apatite fission-track and zircon (UeTh)/He 30 (ZHe)ages support an Oligocene age for the onset of deformation in thecentral axis of the Eastern Cordillera (Mora et al., 2010a; Sayloret al., 2012). Egbue and Kellogg (2012) estimate Late Oligoceneshortening as 20% of the total Cenozoic shortening in the easternfoothills of the Eastern Cordillera. Hossack (1999) reported thepresence of asymmetrical fracture sets (Fig. 7b) that appear topredate folding in the Cupiagua structure. This was done byremoving those fractures that are related to folding from thefractures dataset. Prior to the subtraction, the fracture data wasrotated so that the bedding is unfolded to the horizontal. Type 1

Fig. 9. Schematic block diagram showing (a) early compressive fracture generation and (b) late stage shear fractures superimposed on earlier fracture system.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 17

and type 2 fracture sets were removed from all the fracture data, aprocess which is similar to backstripping. Two less well developedfracture sets remain that are orthogonal: a WSWeENE and aNNWeSSE trend. Since these fractures remain after the beddingdips on the different parts of the fold are unfolded back to hori-zontal, they are interpreted as predating folding of the Cupiaguastructure. Orthogonal fracture sets of this type are not uncommon,and are typically found in undeformed foreland areas in front ofthrust belts. Some examples are the fracture sets of the Appala-chian Plateau (Engelder and Geiser, 1980), the Alberta foreland(Babcock, 1974), and the European Platform from the Alps to as faras southern England (Bevan and Hancock, 1986). In all cases, amain fracture set is usually orthogonal to the thrust belt and asecond set lies at right angles to the main set. Younes and Engelder(1999) have shown that the earliest fractures in the Appalachianplateau developed prior to a major phase of layer parallel short-ening. Continued shortening and detachment sliding of the coverrocks developed a new set of fractures that are slightly rotatedcompared to the earlier set. If we assume that the orthogonalfractures in the Cupiagua field are analogous to the Appalachianfracture sets, then it is possible that they could have formedduring an earlier phase of layer-parallel shortening (Hossack,1999). On the other hand, this EeW to ENEeWSW stress orien-tation is the same as that observed in the present-day stress fieldassociated with the northeastward “escape” of the northernAndes, and the two may therefore be confused.

4.2.2. Andean stresses (Miocene to Present)Paleobotanical data (Fig. 8a; Wijninga, 1996; Gregory-Wodzicki,

2000) and fission-track ages (Mora, 2007; Mora et al., 2010b)indicate that most of the uplift in the central and eastern flank ofthe Eastern Cordillera occurred in the last 10 Ma. Tectonic uplift(Fig. 8a) is estimated from present-day total uplift and paleo-elevation data. Elevation is shown in black with present meanelevation just over 3 km and estimated paleoelevation based onpaleobotanical and geomorphological data (black squares with er-ror bars after Wijninga, 1996; Gregory-Wodzicki, 2000). Tectonicuplift (red line in online version) is estimated using present daytotal structural relief (10 km) and assuming erosion rates of 0.5 and1.0 mm/a. Neogene exhumation history from apatite fission trackdata and low-T thermochronology (Mora, 2007; Parra et al., 2009)supports the continued rapid tectonic uplift proposed for the last3 Ma. Using apatite fission-track data, Mora et al. (2010a,b) esti-mated exhumation rates of 1e1.5 mm/a for compressional struc-tures in the eastern foothills of the Eastern Cordillera over the last3 Ma. Wittmann et al. (2011) used cosmogenic nuclide-basedmeasurements to estimate denudation rates of 0.49e1.2 mm/a forthe Ecuadorian Andes. Note that the uplift curve for an erosion rateof 0.5 mm/a shows tectonic uplift in the Eastern Cordilleradistributed over at least 10 Ma.

The shortening rates for the Eastern Cordillera (Fig. 8b; Egbueand Kellogg, 2012) since 12 Ma (the initiation of the PanamaeSouth America collision) are estimated from the tectonic uplift rates

Fig. 10. Model for fracture development in the Cupiagua field (after Hossack, 1999) (a) fracture sets orthogonal and perpendicular to the early thrust belt. (b) ENEeWSW superposedfracture set related to propagating thrust front or present-day shear.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2118

(Fig. 8a) and constrained to fit the present GPS measured rate,12 � 4 mm/a (Trenkamp et al., 2002). Shortening rates (v) areassumed to be proportional to tectonic uplift rates (u): v ¼ ku,where k is a dimensionless proportionality constant. Shorteningrates are shown for a total shortening of 120 km in the last 12 Ma(0.8 � 150 km; Dengo and Covey, 1993; Egbue and Kellogg, 2012).For 120 km shortening, k ¼ 18. Shortening rates are also shown for60 km shortening, the minimum published estimate for the EasternCordillera (0.8 � 68 km; Cooper et al., 1995), where k ¼ 9. North-eastward escape rates for the North Andean block (Fig. 8b) are fromthe compilation by Egbue and Kellogg (2010). The present-dayPanamaeSouth America convergence rate is approximately30 mm/yr (Trenkamp et al., 2002). Note that range-parallel shear“escape” has become an important factor in the last 2 Ma, butrange-normal shortening rates exceed the shear rates, and totalshortening is 3e6 times as great as the shear displacement in thelast 12 Ma.

About 12e15 Ma the Panama arc began colliding with thenorthern Andes (Coates et al., 2004; Montes et al., 2012), and theEastern Cordillera became a major sediment source (Gregory-Wodzicki, 2000), although some sediments were being suppliedto the Eastern Foothills area as early as the Oligocene (Mora et al.,2010a; Saylor et al., 2012). During the Andean stage (Miocene to

Present) of deformation (Fig. 9a), thrusting and folding in the re-gion generated most of the Eastern Cordillera and foothills struc-tures orthogonal to the main stress and developing fold-relatedconjugate fractures. Two sets of fractures, symmetrically orientatedto the fold hinges of the Cupiagua structure, form a strike-paralleltype 2 fracture set and a strike-normal type 1 fracture set. A rota-tion of the shortening direction to a direction normal to the Cusiana(and Cupiagua structures (Fig. 3)) probably occurred as the thrustfront advanced into the Llanos foothills (Hossack, 1999) (Fig. 10).Subsequent thrusting and folding generated the type 1 and type 2fractures which were then superposed onto the earlier orthogonalset. The orientation of structures in the Piedemonte hydrocarbonblock located northeast of the Cupiagua field, as seen on 3-Dseismic reflection depth slices (Fig. 11), shows evidence of thislate stage thrusting. The structures are oriented perpendicular tothe NWeSE present-day maximum principal stress direction in theEastern Cordillera.

4.2.3. Early Pleistocene range-parallel shearBy Early Pleistocene (Figs. 8 and 9b), in addition to the

compressional stress regime, a NEeSW strike-slip component wasintroduced as the northern Andes began to “escape”. About 2 Ma,the aseismic Carnegie ridge, which was formed by the Galapagos

Fig. 11. 3-D seismic reflection depth slices through the Piedemonte block showing the orientation of major structural features at (a) 2270 m, (b) 4300 m, and (c) 5280 m.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 19

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e2120

hotspot, arrived at the ColombiaeEcuador trench (Lonsdale andKlitgord, 1978; Pedoja, 2003; Cantalamessa and Di Celma, 2004),and initiated the northeastward “escape” of the northern Andes(Egbue and Kellogg, 2010). This resulted in the development ofasymmetrical shear fractures oriented EeW to WSWeENE andNNWeSSE similar to the fractures interpreted by Hossack (1999) aspredating the folding. Presently, in the Eastern Cordillera theescape rate is almost as great as the rate of shortening due to thecollision with Panama (Trenkamp et al., 2002; Egbue and Kellogg,2012). However, since compressive strain results in higherstresses than shear strain, range normal compression still domi-nates the stress field in the Eastern Cordillera.

5. Conclusions

Borehole breakouts and seismic focal mechanisms indicate twopresent-day principal stress directions in the Eastern Cordillera;WNWeESE to NWeSE, aligned with the GPS measured PanamaeNorth Andes collision, and EeW to WSWeENE, aligned with thenortheastward “escape” of the North Andes. The Panama collisionhas produced most of the folding, shortening, and uplift of theEastern Cordillera. The dominant fracture systems observed in theCupiagua fold structures on the eastern flank of the EasternCordillera are produced by the collision and range-normalcompression. Less well developed asymmetrical shear fracturesoriented EeW toWSWeENE and NNWeSSE may be related to pre-folding stresses in the foreland basin of the Central Cordillera(future Eastern Cordillera) or to present-day shear associated withthe northeastward “escape” of the north Andean block. However,range-normal compression still dominates the stress field in theEastern Cordillera. Our study suggests that the critical driver fororogenic deformation at transpressive plate boundaries is theincreased coupling associated with the subduction of aseismicridges and island arc collisions. Both range-normal shortening andrange-parallel shear displacement must be considered when esti-mating the earthquake hazard for the Eastern Cordillera and thepopulous Bogota area.

Acknowledgments

The authors would like to thank Editor Cees Passchier, JoelSaylor, and an anonymous reviewer, for very constructive reviews.We thank BP Colombia and Equion Energy Colombia for making thefracture data from Cupiagua field and the 3-D seismic reflectiondata available, as well as providing partial support for the firstauthor’s research at the University of South Carolina. The CASAproject was supported by NSF Grants EAR-8617485, EAR-8904657,and NASA. The borehole breakout data was supplied by Ecopetrol.Figs. 1 and 3 were created with Generic Mapping Tool (GMT)(Wessel and Smith, 1998).

References

Acosta, J., Lonergan, L., Coward, M.P., 2004. Oblique transpression in the westernthrust front of the Eastern Cordillera. J. South Am. Earth Sci. 17, 181e194. http://dx.doi.org/10.1016/j.jsames.2004.06.002.

Adams, J., Styles, A., Garcia-Carballido, C., 1999. Fracture Characterization from UBIImages, Cupiagua Field: Pilot Study. Internal Document. BP Exploration,Colombia, p. 125.

Babcock, E., 1974. Jointing in central Alberta. Can. J. Earth Sci. 11, 1181e1186.Barth, A., Reinecker, J., Heidbach, O., 2008. Stress Derivation from Earthquake Focal

Mechanisms. World Stress Map Project e Guidelines, p. 12.Beck, M.E., 1983. On the mechanism of tectonic transport in zones of oblique

subduction. Tectonophysics 93, 1e11.Bevan, T.G., Hancock, P.L., 1986. A late Cenozoic regional microfracture system in

southern England and northern France. J. Geol. Soc. Lond. 143, 355e362.Cantalamessa, G., Di Celma, C., 2004. Origin and chronology of Pleistocene ma-

rine terraces of Isla dela Plata and of flat, gently dipping surfaces of the

southern coast of Cabo San Lorenzo (Manabí, Ecuador). J. South Am. EarthSci. 16, 633e648.

Coates, A.G., Collins, L.S., Aubry, M.-P., Berggren, W.A., 2004. The geology of theDarien, Panama, and the late Miocene-Pliocene collision of the Panama arc withnorthwestern South America. Geol. Soc. Am. Bull. 116, 1327e1344. http://dx.doi.org/10.1130/B25275.1.

Colletta, B., Hebrard, F., Letouzey, J., Werner, P., Rudkiewickz, J., 1990. Tectonic styleand crustal structure of the Eastern Cordillera (Colombia) from a balanced crosssection. In: Letouzey, J. (Ed.), Petroleum and Tectonics in Mobile Belts. EditionsTechnip, Paris, pp. 81e100.

Colmenares, L., Zoback, M.D., 2003. Stress field and seismotectonics of northernSouth America. Geology 31 (8), 721e724.

Cooper, M.A., Addison, F.T., Álvarez, R., Coral, M., Graham, R.H., Hayward, S.H.,Martinez, J., Naar, J., Peñas, R., Pulham, A.J., Taborda, A., 1995. Basin develop-ment and tectonic history of the Llanos Basin, Eastern Cordillera, and MiddleMagdalena Valley: Colombia. Am. Assoc. Pet. Geol. Bull. 79 (10), 1421e1443.

Corredor, F., 2003. Seismic strain rates and distributed continental deformation inthe northern Andes and three-dimensional seismotectonics of northwesternSouth America. Tectonophysics 372 (3e4), 147e166.

Cortés, M., Angelier, J., 2005. Current states of stress in the northern Andes asindicated by focal mechanisms of earthquakes. Tectonophysics 403, 29e58.

Cortés, M., Angelier, J., Colleta, B., 2005. Paleostress evolution of the northern Andes(Eastern Cordillera of Colombia): implications on plate kinematics of the SouthCaribbean region. Tectonics 24, TC1008. http://dx.doi.org/10.1029/2003TC001551.

Cox, J.W., 1970. The high resolution dipmeter reveals dip-related borehole andformation characteristics. In: Transactions, 11th Annual Logging Symposium,Los Angeles, pp. D1eD25.

Cox, J.W., 1983. Long axis orientation in elongated boreholes and correlation withrock stress data. In: 24th Annual Logging Symposium, Calgary, pp. 1e17.

Dengo, C., Covey, M., 1993. Structure of the Eastern Cordillera of Colombia: impli-cations for trap styles and regional tectonics. Am. Assoc. Pet. Geol. Bull. 77,1315e1337.

Egbue, O., Kellogg, J., 2010. Pleistocene to present North Andean “escape”. Tecto-nophysics 489, 248e257 http://dx.doi.org/10.1016/j.tecto.2010.04.021.

Egbue, O., Kellogg, J., 2012. Three-dimensional structural evolution and kinematicsof the Piedemonte Llanero, Central Llanos foothills, Eastern Cordillera,Colombia. J. South Am. Earth Sci.. http://dx.doi.org/10.1016/j.jsames.2012.04.012.

Ego, F., Sebriera, M., Lavenuc, A., Yepes, H., Egues, A., 1996. Quaternary state of stressin the northern Andes and the restraining bend model for the EcuadorianAndes. Tectonophysics 259 (1e3), 101e116.

Engelder, T., 1987. Joints and shear fractures in rock. In: Atkinson, B.K. (Ed.), FractureMechanics of Rock. Academic Press, London, pp. 27e69.

Engelder, T., Geiser, P., 1980. On the use of regional joint sets as trajectories ofpaleostress fields during the development of the Appalachian plateau, NewYork. J. Geophys. Res. 85, 6319e6341.

Fitch, T.J., 1972. Plate convergence, transcurrent faults, and internal deformationadjacent to southeast Asia and the western Pacific. J. Geophys. Res. 84, 4432e4460.

Gough, D.I., Bell, J.S., 1981. Stress orientations from oil-well fractures in Alberta andTexas. Can. J. Earth Sci. 18, 638e645.

Gregory-Wodzicki, K.M., 2000. Andean paleoelevation estimates: a review andcritique. Geol. Soc. Am. Bull. 112, 1091e1105.

Gutscher, M.-A., Malavieille, J., Lallemand, S., Collot, J.-Y., 1999. Tectonic segmen-tation of the North Andean margin: impact of the Carnegie Ridge collision.Earth Planet. Sci. Lett. 168, 255e270.

Hickman, S.H., Healy, J.J., Zoback, M.D., Svitek, J.F., Bretcher, J.E., 1982. In situ stress,borehole elongation, and natural fracture distribution, at depth in central NewYork State [abs.]. Eos (Trans., Am. Geophys. Union) 63, 1118.

Horton, B.K., Saylor, J.E., Nie, J., Mora, A., Parra, M., Reyes-Harker, A., Stockli, D.F.,2010. Linking sedimentation in the northern Andes to basement configuration,Mesozoic extension, and Cenozoic shortening: evidence from detrital zircon U-Pb ages, Eastern Cordillera, Colombia. Geol. Soc. Am. Bull. 122 (9/10), 1423e1442,. http://dx.doi.org/10.1130/B30118.1.

Hossack, J., 1999. An Analysis of the Geometry, Orientation, and Structural Devel-opment of the Schlumberger UBI Fracture Sets in Cupiagua and Cusiana Wells.Internal Document. BP Exploration, Colombia, p. 33.

Jarred, R.D., 1986. Relations among subduction parameters. Rev. Geophys. 24,217e284.

Kellogg, J.N., Vega, V., 1995. Tectonic development of Panama, Costa Rica, and theColombian Andes: constraints from Global Positioning System geodetic studiesand gravity. In: Mann, P. (Ed.), Geologic and Tectonic Development of theCaribbean Plate Boundary in southern Central America, Geological Society ofAmerica Special Paper 295, pp. 75e90.

Lonsdale, P., Klitgord, K.D., 1978. Structure and tectonic history of the easternPanama Basin. Geol. Soc. Am. Bull. 89, 981e999.

McCaffrey, R., 1994. Global variability in subduction thrust zone-forearc systems.Pure Appl. Geophys. 141, 173e224.

McCaffrey, R., Zwick, P.C., Bock, Y., Prawirodirdjo, L., Genrich, J.F., Stevens, C.W.,Puntodewo, S.S.O., Subarya, C., 2000. Strain partitioning during oblique plateconvergence in northern Sumatra: geodetic and seismologic constraints andnumerical modeling. J. Geophys. Res. 105 (B12), 28,363e28,376.

McKenzie, D.P., 1969. The relation between fault plane solutions for earthquakesand the directions of the principal stress. Bull. Seismol. Soc. Am. 59, 591e601.

O. Egbue et al. / Journal of Structural Geology 58 (2014) 8e21 21

Montes, C., Bayona, G., Cardona, A., Buchs, D.M., Silva, C.A., Morón, S., Hoyos, N.,Ramírez, D.A., Jaramillo, C.A., Valencia, V., 2012. Arc-continent collision andorocline formation: closing of the Central American seaway. J. Geophys. Res. 117,B04105. http://dx.doi.org/10.1029/2011JB008959.

Mora, A., 2007. Inversion Tectonics and Exhumation Processes in the EasternCordillera of Colombia (Ph.D. thesis). University of Potsdam, Potsdam, 133 p.

Mora, A., Horton, B.K., Mesa, A., Rubiano, J., Ketcham, R.A., Parra, M., Blanco, V.,Garcia, D., Stockli, D.F., 2010a. Migration of Cenozoic deformation in the EasternCordillera of Colombia interpreted from fission track results and structural re-lationships: implications for petroleum systems. AAPG Bull. 94 (10), 1543e1580.http://dx.doi.org/10.1306/01051009111.

Mora, A., Parra, M., Strecker, M.R., Sobel, E.R., Zeilinger, G., Jaramillo, C., Da Silva, S.,Blanco, M., 2010b. The eastern foothills of the Eastern Cordillera of Colombia: anexample of multiple factors controlling structural styles and active tectonics.Geol. Soc. Am. Bull. 112, 1846e1864. http://dx.doi.org/10.1130/B30033.1.

Moreno, C., Horton, B., Caballero, V., Mora, A., Parra, M., Sierra, J., 2011. Depositionaland provenance record of the Paleogene transition from foreland to hinterlandbasin evolution during Andean orogenesis, northern Middle Magdalena ValleyBasin, Colombia. J. South Am. Earth Sci. 32 (3), 246e263.

Ojeda, G., 1996. Present Day Retrodeformed Structure of the Eastern CordilleraCretaceous Basin, Colombia (M.S. thesis). University of South Carolina,Columbia, 57 p.

Olson, J., Pollard, D.D., 1989. Inferring paleostresses from natural fracture patterns: anew method. Geology 17, 345e348.

Parra, M., Mora, A., Sobel, E.R., Strecker, M.R., González, R., 2009. Episodic orogenicfront migration in the northern Andes: constraints from low-temperaturethermochronology in the Eastern Cordillera, Colombia. Tectonics 28, TC4004.http://dx.doi.org/10.1029/2008TC002423.

Parra, M., Mora, A., Lopez, C., Rojas, L.E., Horton, B.K., 2012. Detecting earliestshortening and deformation advance in thrust belt hinterlands: example fromthe Colombian Andes. Geology 40 (2), 175e178. http://dx.doi.org/10.1130/G32519.1.

Pedoja, K., 2003. Les terrasses marines de la marge Nord Andine (Equateur et NordPe0rou): Relations avec le contexte ge0odynamique (Ph.D. thesis). Univ. Paris VI,Paris, p. 413.

Platt, J.P., 1993. Mechanics of oblique convergence. J. Geophys. Res. 98, 16,239e16,256.

Plumb, R.A., 1982. Breakouts in the geothermal well, Auburn, N.Y [abs.]. Eos (Trans.,Am. Geophys. Union) 63, 1118.

Plumb, R.A., Hickman, S.H., 1985. Stress-induced borehole enlargement: a com-parison between the four-arm dipmeter and the borehole televiewer in theAuburn geothermal well. J. Geophys. Res. 90 (B7), 5513e5521. http://dx.doi.org/10.1029/JB090iB07p05513.

Pulido, N., 2003. Seismotectonics of the northern Andes (Colombia) and thedevelopment of seismic networks. Bull. Int. Inst. Seismol. Earthq. Eng., 69e76.Special Edition.

Reinecker, J., Heidbach, O., Tingay, M., Sperner, B., Müller, B., 2005. The 2005 Releaseof the World Stress Map online at: www.world-stress-map.org.

Sarmiento-Rojas, L.F., Van Wess, J.D., Cloetingh, S., 2006. Mesozoic transtensionalbasin history of the Eastern Cordillera, Colombian Andes: inferences fromtectonic models. J. South Am. Earth Sci. 21 (4), 383e411. http://dx.doi.org/10.1016/j.jsames.2006.07.003.

Saylor, J., Horton, B., Stockli, D., Mora, A., Corredor, J., 2012. Structural and ther-mochronological evidence for Paleogene basement-involved shortening in theaxial Eastern Cordillera, Colombia. J. South Am. Earth Sci. 39, 202e215.

Stearns, D.W., 1967. Certain aspects of fracture in naturally deformed rocks. In:Riecker, R.E. (Ed.), National Science Foundation Advanced Science Seminar inRock Mechanics: Special Report: Air Force Cambridge Research Lab, Bedford,Massachusetts, pp. 97e118.

Taboada, A., Rivera, L.A., Fuenzalida, A., Cisternas, A., Hervé, P., Harmen, B., Olaya, J.,Rivera, C., 2000. Geodynamics of the northern Andes: subductions and intra-continental deformation (Colombia). Tectonics 19 (5), 787e813.

Toro, J., Roure, F., Bordas-Le Floch, N., Le Cornec-Lance, S., Sassi, W., 2004. Thermaland kinematic evolution of the Eastern Cordillera fold and thrust belt,Colombia. In: Swennen, R., Roure, F., Granath, J.W. (Eds.), Deformation, FluidFlow, and Reservoir Appraisal in Foreland Fold and Thrust Belts, AAPG HedbergSeries, no. 1, pp. 79e115.

Torres Fleischmann, C., 1992. In-situ Stress in the North Andes from BoreholeBreakout Analysis (MS. thesis). University of South Carolina, Columbia, 153 p.

Trenkamp, R., Kellogg, J.N., Freymuller, J.T., Mora, H.P., 2002. Wide plate margindeformation, southern Central America and northwestern South America, CASAGPS observations. J. South Am. Earth Sci. 15 (2), 157e171. http://dx.doi.org/10.1016/S0895-9811(02)00018-4.

Wessel, P., Smith, W.H.F., 1998. New, improved version of the generic mapping toolsreleased [abs.]. Eos (Trans., Am. Geophys. Union) 79, 579.

Wittmann, H., von Blanckenburg, F., Guyot, J.L., Laraque, A., Bernal, C., Kubik, P.W.,2011. Sediment production and transport from in situ-produced cosmogenic10Be and river loads in the Napo River basin, an upper Amazon tributary ofEcuador and Peru. J. South Am. Earth Sci. 31 (1), 45e53. http://dx.doi.org/10.1016/j.jsames.2010.09.004.

Wijninga, V.M., 1996. Palynology and Paleobotany of Neogene Sediments from theHigh Plain of Bogotá (Colombia): Evolution of the Andean Flora from anEcological Perspective (PhD thesis). University of Amsterdam, Amsterdam,p. 370.

Younes, A., Engelder, T., 1999. Fringe cracks: key structures for the interpretation ofprogressive Alleghanian deformation of the Appalachian Plateau. Geol. Soc. Am.Bull. 111, 219e239.

Zoback, M.D., Moos, D., Mastin, L.G., Anderson, R.N., 1985. Well bore breakouts andin situ stress. J. Geophys. Res. 90, 5523e5530.

Zoback, M.L., Zoback, M.D., 1989. Tectonic stress field of the continental U.S. In:Pakiser, L., Mooney, W. (Eds.), Geophysical Framework of the Continental UnitedStates. Geol. Soc. Am. Mem. 172, 523e539.