BASEMENT SEISMICITY BENEATH the ANDEAN PRECORDILLERA THIN&Hyphen

TEC'rO••, VOL. 12, NO. 1, PAGES 63-76, FEBRUARY 1993

BASEMENT SEISMICITY BENEATH THE by subductionof the oceanic Nazca plate beneath the ANDEAN PRECORDILLERA THIN- continentalSouth American plate. The modernAndes are due SKINNED THRUST BELT AND to Neogene tectonics characterized by along-strike IMPLICATIONS FOR CRUSTAL AND segmentationof majorfeatures of thesubducted and overriding LITHOSPHERIC BEHAVIOR plates [Barazangiand Isacks, 1976; Jordan et al., 1983] (Figures1 and2). In addition,the along-strikefeatures of the Robert Smalley,Jr., 1 Jos6Pujol, 1 MarcRegnier, 2 two platesare correlated,both spatiallyand in their temporal Jer-MingChiu, 1 Jean-LucChatelain, 2,3Bryan L. Isacks,4 development.Between 15øS and 24øS the Altiplano-Puna MarioAraujo, 5 and N. Puebla 5 plateau, the EasternCordillera and SubandeanZone thrust belts,and an activevolcanic arc are located above a moderately Abstract.Data froma digitallyrecording seismic network dippingWadati-Benioff zone (WBZ). In contrast,a narrow in SanJuan, Argentina, provide the first imagesof crustal thin-skinned thrust belt (lyecordillera), a wide zone of scalebasement faults beneath the lyecordillera. This seismicity basementuplifts (SierrasPampeanas), and an extinctare are is nearthe boundary between the lyecordillera (a thin-skinned locatedabove an intermediate-depthsubhorizontally subducting thrustbelt) and the SierrasPampeanas (a regionof thick- WBZ between 28øS and 33øS. Shutdownof volcanism, onset skinnedbasement deformation), two seismically active tectonic of Pampean and Precordilleran crustal deformation, and provincesof theAndean foreland. The seismicity data support developmentof flat subductionare thoughtto coincide at modelsfor thisregion in whichcrustal thickening, rather than 10-15Ma [Jordanet al., 1983]. The SierrasPampeanas and magmaticaddition or thermaluplift, plays the dominant the lYecordilleraare thoughtto have a tectonicrelationship mountainbuilding role. The lYecordillera seismicity occurs in similarto thatof theMesozoic Sevier and Laramide provinces three segmentsdistributed north to south.The southern of North America,which are alsothought to havedeveloped segmentis an areaof diffuseactivity extending across the overa regionof flat subduction[Dickenson and Snyder, 1978; Precordilleraand eastwardinto the SierrasPampeanas that Fieldingand Jordan, 1988]. showsno patternsin mapor crosssection. The northernand The Andeanorogen is seismicallyactive, and variations of centralsegments have well-defined dipping planes that define crustalintraplate seismicity in the overridingplate correlate crustalscale faults extending from 5 to 35 km depth.It is with the along-strikevariations of tectonicstyle (Figure2). clear from the relativefault geometriesthat the overlying Crustalseismicity is concentratedin a narrowband along the Precordillerais not simplyrelated to the basementactivity. easternmargin of theforeland in areasover normal subduction, Theseismicity here may result from reactivation of anancient while above flat subductiona wide area of the foreland, with suturebetween the lYecordillera and Pampeanas terranes or be thin- and thick-skinnedtectonic provinces, is highly active occurringin basementof unknownaffinity west of thesuture. seismically.The easternmostlimit of crustal seismicity Theseismicity provides the first constraints on basement fault coincideswith the easternmostWBZ activityin bothflat and geometries,and we present models integrating this information steepsubduction regions. The depthof crustalearthquakes also with the surfacegeology. These basement faults may have correlateswith tectonicstyle. In regionsof normalsubduction, been responsiblefor the 1944 Ms 7.4 earthquakethat seismicityoccurs in the upper and mid crust to depthsof destroyedthe city of SanJuan. The imagingof thesefaults 25 km [Cahill et al., 1992], while in areasof flat subduction suggeststhat seismic risk estimates for SanJuan made on the crustalevents occur at mid and lower crustaldepths to 40 km basisof surfacegeologic studies may be too low. ISmalleyand Isacks,1990; Su•ez et al., 1983].This variation may be due to regionaldifferences in lithosphericor crustal INTRODUCTION stress,strain rate, crustalgeotherm, or rheology,both across andalong the orogen. The Andesprovide a uniquelaboratory to studymountain belt formation in a subduction-driven but noncollisional GEOLOOY environment.The presentorogen began in theJurassic, driven Precordillera. The lYecordillera is a thin-skinned thrust belt of Paleozoicelastics divided into three structural subprovinces: 1 Centerfor Earthquake Research and Information, Memphis Western,Central, and Eastern(Figure 3) [Baldis,1975; Ortiz StateUniversity, Memphis, TN. and Zambrano, 1981; Baldis et al., 1982]. The Central and 2 InstitutFranqais de RechercheScientifique pour le Westernprovinces form the forelandthrust belt foundon the D6veloppementen CoopdrationNoumea, New Caledonia, cratonicside of compressionalorogens. These mountain belts South West Pacific. are interpretedas the resultof a crustalshortening across an 3 Nowat Institut Franqais deRecherche Scientifique pour le orogenin which crustalthickening associated with A-type D6veloppementen Coop6ration,Domaine Universitaire, subductionoccurs along a majorintracontinental thrust [Bally, Grenoble,France. 1981]. Crustalmaterial from the orogenproper is thrust 4 Institutefor the Studyof the Continents,Cornell cratonwardover relatively undeformedforeland basement, University,Ithaca, New York. pushingthe forelandbasin sediments into a thin-skinnedthrust 5 InstitutoNacional de Prevenci6nSfsmica, San Juan, belt. This thrustbelt is typicallythe cratonwardmostlimit of Argentina. deformation associatedwith an orogenic system. The developmentof thesethrust belts generally follows a set of Copyfight1993 by the AmeficanGeophysical Union. rules, derived from geologic studies [Chappie, 1978; Dahlstrom,1970; Davis et al., 1983]and physical models of Paper number92TC01108. deformation[Panian and Pilant, 1990] describing the spatial 0278-7407/93/92TC-01108510.00 andtemporal evolution of their structures. 64 Smalleyet al.: BasementSeismicity in AndeanPrecordillera

75 W 7O 65

Fig. 1. Map of westernSouth America illustrating upper plate tectonic features, volcanic arc, and contours of theWadati-Benioff zone (WBZ). Along-strikesegmentation of the upperplate tectonic provinces correlateswith the along-strike segmentation of the WBZ [Jordanet al., 1983].The distribution of Neogene volcanoes(open circles) [Isacks, 1988] shows an active magmatic arc over the steep WBZ segmentand the absenceof anactive arc over the flat segment. The drainage divide, which defines the internally drained Puna/Altiplanoplateau and then follows the crest of theAndes southward between the Principal and Frontal cordilleras,is also shown. The inset shows the major topographic features of theoceanic Nazca plate (Juan Fernandezand Nazca ridges). 20

o o

75 W 70 65

Fig. 2. Shallowintraplate seismicity (solid circles) and damaging historic earthquakes (open circles) [Zamarbideand Castano,1978] of the Andeanback-arc. Epicenters are selectedfrom Preliminary Determinationof Epicenters(1963, 1983-1990)and International Seismological Center (1964-1982) catalogs.Selected events were located with 15or moreP arrivals,are shallower than 50 km,and are east of the 100km WBZ contourensuring they are in SouthAmerican lithosphere and are notinterplate or WBZ events.A selectionof focal mechanisms[Chinn and Isacks,1983; Dziewonski et al., 1987; Kadinsky-Cade, 1985;Stauder, 1973] is alsoshown. The level of seismicityin thePie dePalo area (31.5øS, 67.5øW) is so intensethat only focalmechanisms with accuratesource depths from syntheticmodeling are shown. Tectonicprovinces and inset are as in Figure1. 66 Smalleyet al.: BasementSeismicity in AndeanPrecordillera

894 .2

[] 1977 6.7 7A

Okm 50km Salinas i i

69 W 68 67 Fig.3. Geology,mapped and inferred faults (solid and dotted bold lines), seismic station sites (small triangles:solid for stations, open for arrays of 8 to 11stations with apertures of 1 m to 1 km),damaging historicalseismicity (dated open squares, 1944 location from Kadinsky-Cade [1985]), the 1944earthquake intensityVIII isoseismal(large triangle outline) [Castellanos, 1945], and two LANDSAT lineaments in SanJuan (heavy gray lines). The LANDSAT lineament near La Lajaincludes the fault suffering offset in 1944.Mountain range patterns are light gray for sedimentaryrocks (Precordillera) and dark gray for basementblocks (Sierras Pampeanas); intermontane valleys are in white.The Eastern Precordillera is the chainof mountainseast of the MatagusanosValley (SierrasChica de Zonda,Villicum, Morado, and extensionsnorth and south). West of thePrecordillera are the Calingasta and Iglesia valleys.

The CentralPrecordillera is a typicalforeland thin-skinned Pampeanas,a province of Precambrianmetamorphic basement thrust belt. The Eastern Precordillera was fu'st described as the block uplifts [Caminoset al., 1982; Cingolaniand Varela, easternmostpart of this thrust belt [Orfiz and Zambrano, 1975]. Such deformationcratonward of the thin-skinnedthrust 1981], but it appearsto break two rules of thin-skinned belt is unusualand found only in regions overlying thrusting.First, the EasternPrecordillera is thrustwestward, subhorizontalsubducfion, a correlation also thought to holdin away from the craton.Second, the basalunit of the Eastern extinctorogens [Dickenson and Snyder, 1978]. The Pampeanas Precordillerais over 2 km down-sectionwith respectto the mountainsare generallynorth-south trending blocks with a basalunit of theCentral and Western precordilleras, suggesting distinctmorphology consisting of a steepfault-bounded front the d6collementcuts downsectionin the transportdirection side and a gentlydipping back side.The fault is typically [Baldis and Bordonaro, 1984]. Previouscross sections of the moderatelydipping, thrust or reversetype, and has large EasternPrecordillera, based on geologic[Insfituto Nacional de amounts(up to severalkilometers) of structuralrelief [Jordan Prevenci6nSfsmica (INPRES), 1982; Whitney, 1991] or and Allmendinger,1986]. The back sideis a late Paleozoic geologicand low-resolutionseismicity and shallowseismic erosionalbasement surface [Caminos, 1979; Rassmuss,1916; reflection studies[Allmendinger et al., 1990; Fielding and Dalla Salda and Varela, 1984; Toselli et al., 1985]. Wide, Jordan,1988], show the observed east-dipping basal thrusts of shallowbasins of unmetamorphosed,relatively undeformed the EasternPrecordillera as upwardcontinuations of inferred Carboniferousand youngersediments separate the blocks east-dippingPampean basement faults. [Salfityand Gorustovich, 1983]. SierrasPampeanas. East of thePrecordillera are the Sierras Boundaryregion. The Bermejoand Tulum valleys separate Smalleyet al.: BasementSeismicity in AndeanPrec•rdillera 67 the two provinces(Figure 3). As one crossesbetween the SEISMICITY provincesthe natureof the basementis a majorunanswered question.Unfortunately, basement in theboundary region is Seismicityin San Juanoccurs in two zones: in the crust buried by Paleozoicrocks of the Precordilleraor young and in thesubhorizontal Nazca plate at about100 km depth sedimentsof the Bermejoand Tulum valleys.Three small [Barazangiand Isacks, 1976; Smalley and Isacks, 1990; Pujol fault-boundedoutcrops of Pampeanbasement follow an et al., 1991](Figure 4). The crustalseismicity also occurs in inferredbasement fault extending southwest from Pie de Palo two zones: in the SierrasPampeanas basement block uplift and comewithin 10 km of the EasternPrecordillera (Cerros Sierra Pie de Palo [Regnieret al., 1992] and in the thin- Barbosa,Valdivia, and Salinas;Figure 3), and no basement outcrops are found farther west [Zambrano, 1969]. No Previousstudies have shown seismicity in thePrecordillera occurrences of the Paleozoic Precordillera section are known is concentratedat mid to lower crustaldepths of 10 km to eastof theEastern Preeordillera. A melangeassociated with a 40 km, placing the activity in the basementbeneath the Devonianstrike-slip suturing of thetwo terranes is exposed sedimentsof the thrustbelt ISmalleyand Isacks,1990]. Our alongthe eastside of the EasternPrecordillera [Ramos et al., preliminarylocations agree with this result, but our improved 1986;Ortiz andZambrano, 1981]. If thebasement changes final locationshave a depthrange of 5 km to 35 km. The from Pampeanto an unknownand probably allochthonous earlierstudies resolved the overallmap and depth range of basementunder the Prec•rdillera, the change occ• westof the seismicityhere but were unableto resolvestructures in the threebasement outcrops described. The locationof a basement seismicitypattern. These results were neverthelessused to suturewith respect to the melangeis problematicas thethin- buildmodels in whichthe Eastern Precordillera is notpart of skinnedthrust belt sediments are decoupled from the basement the thin-skinnedthrust belt but resultsfrom underlying across the basal d•collement. Pampeandeformation. In thesemodels, the east dipping basal

w E 0km

Okm 200 ]an

Fig.4. Selectedshallow (solid circles) and intermediate-depth (open circles) earthquakes from this study locatedusing Joint Hypocentral Determination [Pujol et al., 1991]and a laterallyhomogeneous layered crustal/uppermanfie velocity model. Cross sections show the seismicity has a bimodaldepth distribution (crustaland Wadati-Benioff zone). Map showsthat crustal seismicity distribution is alsobimodal and associatedwith the Pampeanrange Sierra Pie de Palo and the easternside of the Precordillera.In the Precordilleraall A, B, andC qualityevents are shown (.-220). For Pie de Palo only A eventsplus B andC eventsfrom a 1-monthwindow around an ML 5.3 eventare shown(-720). A smallsubset of A andB intermediate-depthevents (-500) is alsoshown. 68 Smalleyet al.: BasementSeismicity in AndeanPrecordillera

thrustof the EasternPrecordillera is interpretedas a simple 1.5 continuationof a Pampeanstyle, east dipping crustalscale o basementfault to the surface [Allmendinger et al., 1990; Fieldingand Jordan, 1988]. 1.0"' ø o o% DATA COLLEC'rION AND ANALYSIS

A digitally recordingportable seismic network (PANDA . øOo•.• [Chiuet al., 1991]) wasoperated in SanJuan, Argentina, from 0.0- o August 1987 to May 1988 and recorded nearly 20,000 o earthquakes. The network consisted of 26 sites in a -0.5 - o 100 km by 150 km area includingparts of the Precordillera o o and Sierras Pampeanasand the boundarybetween them o o o (Figure3). Three of the sitescontained small aperture arrays o for a total of 48 three-componentstations. Stations sent data by frequencymodulation telemetry to a centralsite whereit I : I * I ' I was digitized at 100 samplesper second.Earthquakes were -40 0 40 80 identified and savedautomatically using an event detection algorithmbased on coincidenceof triggersat specifiedgroups distance (km) of stations.Because recording and digitizing are performed at a Fig.5. Crosssection (azimuth N105øE) of JointHypocentral centralsite, there were no interstationtiming problems and the Determinationstation corrections for P (solidcircles) and S data were availablein real time. To guaranteerecovery of (opencircles) showing general decrease from west to east(the absolute time, we recorded an analog Inter-Range outliers, one P and one S, on the east are located on InstrumentationGroup (IRIG) timecode from a satelliteclock. unconsolidatedsandy sediments south of SierraPie de Palo, Each station had six channels for use with two three- Figure 3). componentsensors and couldbe operatedwith (1) identical sensors(4.5 Hz, Mark Products © L-28 geophones) in a dual gainmode to obtainwide dynamic range (~90 dB), (2) identical 0.01 s; i.e., the first samplecontaining the arrival can be sensorsat the samegain to form small-aperturearrays, or (3) clearlyand unambiguouslyidentified. For emergentarrivals two different sensors(e.g., a geophoneand force balance andsmall events with poorsignal to noiseratios, the problem accelerometer).Half the stationsare capableof repeatingdata becomesidentification of the arrival,as theprecision remains from one other station,and severalrepeat-only stations were 0.01 s. SuchP arrivalscan typicallybe identifiedwithin 1/2 capableof repeatinga two-stationdata stream. This ability to cycleor 0.05 s at thepredominant frequencies for localevents. repeatthe telemeteredsignals meant network design was not Similar considerationsapply to identificationof S buried in constrainedby topographyand could be optimizedfor coverage largeamplitude P codaor mixedwith convertedphases, which of the seismictarget areas. limitsidentification to withinone cycle or 0.1 s. Preliminaryprocessing included event classification and Initial locationswere found with HYPOINVERSE [Klein, quality gradingfollowed by locationusing automatically 1978]and a laterallyhomogeneous three-layer crustal velocity determinedP and S times[Smalley et al., 1989].Events were model (Table 1). As the geology has large-scalelateral gradedon thebasis of thenumber of well-recordedP arrivals, heterogeneitiesacross the network(metamorphic basement on countingmultistation arrays as one station.Events with over the east and sedimentaryrocks on the west) and the set of 19 well-recordedP arrivalsreceived quality grades of A; 11 to stations recording each event is not homogeneous 18, B; 7 to 10, C; and less than 7, D. Events for this study (approximatelyhalf the selectedevents are "C" quality and were selectedbased on the preliminaryprocessing and were recordedonly by stationsin the Precordillera)the station reprocessedto verify timesand phasecharacteristics of the subsetused to locatean eventstrongly influenced the result. automaticallydetermined arrivals and to pick additional We madea simpletest usingevents recorded by the whole arrivals. ImpulsiveP and S arrivals were easily measuredto networkby comparinglocations obtained using all thestations

TABLE 1. Three-LayerOver Half-Space Model For CrustalVelocity

Depth,km Velocity,km/s

0.0 5.88 10.0 6.20 32.0 7.30 45.0 8.10

Three-layerover half-space model for SanJuan area from theInstituto Nacional de PrevencfonSfsmica (INPRES).Data used to generate model could not resolve velocity differences between the Precordillera and SierrasPampaneas provinces resulting in a singlevelocity model for both. The velocity of thetop layer is basedon shallow refraction data in the Tullurn and Bermejo valleys [Bollinger and Langer, 1988]. Velocities in thebottom two layers are based on regional earthquake studies [Volponi, 1968]. Smalleyet al.: BasementSeismicity in AndeanPrecordillera 69

to locationsobtained using only stationsin the Precordillera. seismicitysuggests the velocityeffect predominates and this is The mostimportant variation was a consistentdepth difference our preferredinterpretation. From here on we will consider of about 5 km, with the full network locations having only the final JHD locations. shallowerhypocenters. To addressthis problem and retain as manywell located events as possible, events with greater than DISCUSSION 10 P and 10 $ arrivals were selectedfor relocationby joint hypocentraldetermination (JHD) [Pujolet al., 1991]and the Our data provide the first view of seismicity in the wholedata set was relocated using the JHD stationcorrections. Precordillerawith sufficient resolutionto image fault-like The JHD locationsare uniformlyshifted a few kilometerswest features (Figures6 and 7). The activity is concentrated and shallower than the HYPOINVERSE locations. Most beneaththe Eastern Precordillera, the Matagusanos Valley, and importantly,however, effects of inhomogeneousstation sets the easternmostCentral Precordillera.Almost no activity is are removed.The JHD stationcorrections have a clear pattern found westwardin the Central and Westernprecordilleras or that correlateswith the changein surfacegeology across the eastwardin the Tulum Valley. Seismicityin the Precordillera network(Figure 5). This correlationis alsofound in station is alsonot evenlydistributed north-south, as northof 3 IøS it correctionsobtained using events from only the WBZ or only is relatively aseismic.The sharpnessof the terminationof SierraPie de Palo[Pujol et al., 1991].The stationcorrections activityat 31øS, which is sharedby the Pie de Palo andWBZ indicate that arrivals on the west side of the network are activity, is an importantobservation although its significance delayedwith respect to thoseon theeast. Such a delaycan be is poorly understoodISmalley and Isacks, 1987, 1990]. The causedby a lower averagevelocity to the west or, for crustalseismicity has a roughlynorth-south striking pattern intermediate-depthevents, a thickeningof the crustto the extendingfrom approximately31 øS to near32.25øS where the west.The agreementof stationcorrections obtained from the terminationof activity is poorly defined.This seismicityis twocrustal source areas with thosefrom the intermediate-depth also clearly divided into segments:northern, central, and southern. The northernsegment, between Sierra de Moradoand La Laja, forms a northeast-southweststriking linear pattern (Figure6). In crosssection we seea narrow,linear, dipping feature over a wide range of azimuths.The distributionof hypocentersis thereforealmost pencil-like. If the distribution 31 is interpretedas a line in a plane,the shallowestapparent dip S is the true dip of the plane, and the cross-sectiontrend is normal to the strike of the plane. Alternatively, if the seismicityis a lineardistribution not relatedto a plane,the crosssection showing the shallowestdip givesthe plungeof the line andthe projectiondirection is normalto the trend.As

N S O!•nb 601an

,, 0 k,O 0

/ • o 32 o \ o oio

7.2 o o o: o

69 W 68

Fig. 6. (a) Map and (b) one-to-onecross section of locationsfor northernsegment (solid circles). A cross sectionstriking S4øE has the shallowestdip (31øS) and is interpretedas a N86øE strikingfault plane dippingsouth. (The strikeof the crosssection is the strikeof the verticalplane onto which the data are projected).Map featuresare as in Figure3, the numbersby thefocal mechanism solutions are eventdepths in kilometers.Solid circles on the mapindicate events selected for displayin thecross section. In thecross sectionthere is an additionaldivision of eventsrepresented by solidand patterned circles. The solidcircles indicateevents used in the leastsquares fit to a plane,and the patternedcircles are the remainingevents selectedin themap view thatwere removed before the fit wasmme. 70 Sma!leyet al.: BasementSeismicity in AndeanPrecordillera

faultsare generallythought to be two-dimensionalwe have selectedthe planarinterpretation. A leastsquares fit of the hypocentemto a planegives a strikeof N86øEand dip of 31øS for thebest fit plane.This agreeswith the strikeand dip found in theplanar interpretation of thecross sections. 31 The centralsegment (Figure 7) is the triangleof activity $ beneath Sierras Villicum and Chica de Zonda. In cross section (Figure7b) we seea northwestdipping plane plus a clusterof activityin the footwall. A leastsquares fit to a plane,after removingthe footwall activity, gives a strikeof N45øEand dip of N35øW for the best fit plane. The clear gap in activity between15 and 19 km depthmay be relatedto thesediment- basementcontact, estimated to be deeperthan 8 to 10 km beneaththe MatagusanosValley [Fieldingand Jordan, 1988], or to the d6eollement beneath the Central and Western precordillerasestimated at 15 km depth[Allmendinger et al., 1990]. Additional structurein the Central segmentis also

o suggestedin the crosssection and map, but the smallnumber •)o • of eventslimits resolution.For example,two cross-section o featuressuggest back thrustingin the hangingwall block. Theyproject upwards, opposite the dip of themain fault, from o oG about 10 km and 15 km depth.The activity in the footwall narrows to a minimum width, with a roughly vertical 32 distribution, when the cross section is rotated 15ø counterclockwise(Figure 7c), suggestinga vertical fault plane øio o with a strikeof N30øE. Figure 7 alsoshows that the eastern o o• boundaryof the seismicityis not one straightline but two parallelsegments offset by a step.This step is foundwhere the o seismicitypattern crosses beneath the EasternPrecordillera at the north end of Sierra Chica de Zonda. Associated with this 69 W 68 stepare changesin the strikeof the EasternPrecordillera and

NW $B • SE 01anb 60Inn 0kin(• 60Inn 01ma

me leo • o ß ß ß

ß ß***: -, ,- * 'ooo % ß ß • oo ß q)e ß ß ß oo eo ß ß ) ß

ß ! ß

Fig.7. (a) Mapand (b,c) cross sections of locationsfor central segment (solid circles). A crosssection Co) strikingN45øW has the steepest dip (35øNW) and a goodplanar distribution indicating a fault plane striking N45øEand dipping NW. Thesurface intercept of theproposed fault (long southwest to northeastdashed line),and the cross section strike (northwest to southeastdashed line through center segmen0 are shown on themap. The projection of theends of theLANDSAT lineament near La Lajaare shown in thecross sectionby thesmall triangles. Map and earthquake symbols ale as in Figure6. A tighteningof theactivity in thefootwall block into a lesswell-defined vertical plane is seenwhen the cross section is rotatedto a strikeof N60øW(c) indicatinga verticalplane that strikes N30øE. Smalleyet al.: BasementSeismicity in AndeanPrecordillera 71

the minimumdepth of the earthquakes.To the northeastof 1944, similarto the multifaultruptures proposed for the 1977 thisstep the shallowestseismicity is about10 km deep,while Cauceteearthquake that occurred nearby beneath Sierra Pie de to the southwestseismicity continues upward to about5 km Palo in the Sierras Pampeanas[INPRES, 1977; Volponi, depth.Finally, the surface intercept of theplane defined by the 1979; Kadinsky-Cade,1985] or the 1988 TennantCreek, seismicity overlaps a LANDSAT lineament related to Australia,earthquake [Bowman, 1991]. Although the depth of agriculturalexploitation of a line of springs,the Mtdanosde the 1944 earthquakeis poorlycontrolled, it probablyocc• Oro, in the southwesternTulum Valley (Figure7a). in the basement(the International SeismologicalCenter The southernsegment is thepoorly defined area of sparse reportsa depth of 50 km) becauseno significantsurface seismicityin the southlacking clear patterns in mapor cross- faulting occurred.Six centimetersof coseismicdip-slip sectionviews. Seismicity in the Tulum Valley southwestof movement,which increased to 30 cm over a periodof weeks, Pie de Palois mostlyeast of theline of basementoutcrops and wasobserved on a 6 km long fault in soft sedimentsnear La the inferred basement fault that both strike southwest from the Laja[Castellanos, 1945; Groeber, 1944] (Figure 3). Theoffset westernside of Pie de Palo. This seismicityis thereforein on thisfault was east side up, similar to theoffset on thebasal Pampeanbasement and not considered part of thePrecordillera thrustsof the Eastern Precordillera.This is oppositethe activity. directionexpected from thrustmovement on the northwest The northernand centralsegments clearly define distinct dippingbasement fault defined by theseismicity and supports faultsthat may be usedto estimatemaximum magnitudes, or the interpretationof Castellanos[1945] that surface faulting characteristicevents, for this area. The Ms 7.4 San Juan was a secondaryfeature. The planarfeatures imaged in the earthquakeof 1944 (location 31.6ø+0.4øS,68.5ø+0.6øW basement,whose areas are in the rangeassociated with M 7+ [Kadinsky-Cade,1985], which produced a maximumintensity earthquakes,and any maskingof surfacefaulting by the of IX in San Juan [InstitutoNacional de Prevenci6nSfsmica overlyingsediments reinforces the idea that seismic risk in San (INPRES),1977]), may have occurred on oneor moreof these Juanmay be higher than that estimated from surface geologic fault-likefeatures. The largeuncertainty in thelocation of the informationalone [INPRES, 1982]. 1944 earthquakeprecludes associating the eventwith any While there is clearly a spatial relation between the specificfault. The areaof theplane in thecentral segment, or basementseismicity and the Eastern Precordillera, the the centraland northernsegments combined, is within the structuresimplied by the seismicitydo not correlatewith rangeof areasfor Ms 7.0-8.0events [Kanamori and Anderson, featuresof the surfacegeology such as segmentationof the 1975].These planes may have ruptured in a multipleevent in EasternPrecordillera into individualranges. The strikes,dips,

t pro-sutureSierra Pampeanaa baseanent terrano of / Precmbfianme•hic basement.x i,-,-,-,-,-,-[,-,-,-,-,-,-,-,-,-,-,-, ,-,-,-,-,-,-,-,-,',',, ,,,--,,,,,,,

P••ic •vo• •d ol•r) fo•l•d bes••

ta• of •n• •

fa•t • •• pr•••di• f•l•d b• t•e. mapview rosssection view afterDevonian strike-slip suturing of Truncationo_f . . proto-Precordilleraforeland basin I preexistingtoreiano withSierras Pampeanas basement [ basinby .suture •e•ran•. traceof strike-slip • x I-• CE fault

Z %%%%%%%%%%% ii i i i i L,&d.gn"•l q•

•% % ß ß ß % % % % ß %1 %%%%%%%%%%%1 inßr out

Fig.8. Mapand cross section views showing the proto-Precordillera andSierras Pampeanas terranes before andafter a Devonianstrike-slip suturing of thePrecordillera terrane, composed of anearlyPaleozoic foreland basin,to thewestern margin of SouthAmerica [Ramos et al., 1986].Note the trunCation of theforeland basinby thesuture zone. (Actual sense of strike-slipoffset is unknown.) 72 Smalleyet al.: BasementSeismicity in AndeanPrecordillera andsurface projections of the seismicallydefined faults also do cratonwardgrowth of thethin-skinned thrust belt is interrupted not match the faults and structures of the Eastern Precordillera. by the wall of Pampeanbasement at the ancientstrike-slip In addition,immediately north of the terminationof seismicity suture zone where the foreland basin is truncated. The wall near3 IøS thedepth to basementin theBermejo Valley deepens forms a buttresspushing up the west-vergingEastern dramatically to the north or northeast [Jordan and Precordillera,which may also be upliftedand rotated as a unit Allmendinger,1986]. We will discussthe importanceof this atopthe hanging-wallblock. The thin-skinnedthrusting is in more detail below. thereforesignificantly modified by truncationof theforeland The strikesof the seismicallydefined planes in the central basinalong the suture boundary. Except for theeffects due to segmentare also subparallel to the westernside of Pie de Palo the anomaloustermination of the foreland basin, however,the and to the inferred fault and line of small fault-bounded Precordillerais a normal foreland thrustbelt driven cratonward basementoutcrops stalking southwest from Pie de Palo. The by the orogenproper. Last, as the sedimentarycover in the seismicitybeneath the EasternPrecordillera and Matagusanos TulumValley is relativelythin, 1-2 km [unpublishedseismic Valley mayrepresent compressional reactivation, in a roughly refractiondata, Argentine national oil company,Yacimientos east-west direction, of faults along a north/northeast- PetrolfferosFiscales, (YPF)], and thereare no majorsurface south/southwesttrending, originally strike-slip,suture zone featuresin theEastern Precordillera or TulumValley associated betweenthe SierrasPampeanas and Precordilleraterranes with projectionsof the basementfaults to the surface,the (Figure 8). This geometrywould also allow strike-slip offseton thesefaults must be relativelysmall, especially in movements,but suchdeformation has not beenrecognized in comparisonto thelarge offsets found in theSierras Pampeanas the Eastern Precordillera. province.The northeast portion of thecentral segment, where The incongruenceof the basementfaults and the surface the seismicityis east of the EasternPrecordillera, may geologic structuressuggests a nonuniqueclass of models representthis case. relatingthe two deformations(Figure 9). The intrinsicthree- If the fault is on the foreland basin side of the suture and dimensionalitymakes it difficultto illustratethe full geometry sufficientlyclose to it, basementshortening will carrythe in two-dimensionalcross section, so we will present our sedimentson thehanging wall sideeastward into the abutment modelsin three-dimensionalperspective view. In thesemodels of Pampeanbasement at thesuture. In thiscase, shortening in an inferredsharp change in the depthto basement,related to the sedimentary section associated with the Eastern the suturebetween the Precordilleraand SierrasPampeanas Precordillerais a directresult of thebasement shortening. The basements,truncates the foreland basin of the Precordillera southwestportion of the central segment,where the (Figure8). The relativelocation of the basementfaults with easternmostlimit of seismicitycrosses westward beneath the respectto the sutureand to the Precordillerasedimentary EasternPrecordillera, may representthis case.In thisarea a sectiondetermine if thesedimentary section is passivelycarried bandof deformationsubparallel to thestrike of theedge of the on the hangingwall block of the basementfault or if the seismicityis found acrossthe EasternPrecordillera in sedimentarysection participates in thesame deformation as the LANDSATimages and may represent surficial structure related basement. to the basement deformation. If the fault defined by the seismicitypasses beneath the In both models,back thrustsin the hangingwall block sedimentarysection, basement shortening will not directly could exaggerate,or be responsiblefor, the Eastern shorten the sediments above the fault. In this case, normal Precordillera,outcropping with or as the basalfaults of the

N Eastdipping Eastem Continuationof plane Precordillera basal definedby seismicity thrust fault to the surface

Okm suture

,%%%%%%%%

40 km NW dipping planedefined basement by seismicity Precordillera

Fig.9. Schematicthree-dimensional perspective view illustrating tectonic model for relationshipbetween basementfaults and surficialgeology. Note changein depthof the shallowestseismicity as the plane definedby theseismicity crosses beneath the Eastern Precordillera. Smalleyetal.: Basement Seismicity inAndean Precordillera 73

EasternPrecordillera. East-dipping faults observed along the events in this study are all small, less than ML = 4, so the eastern side of the Eastern Precordillera, such as the fault fault areas and source volumes involved are also small. The showingdisplacement associated with the 1944earthquake, earthquakesmay thereforebe samplingsmall-scale variations maybe samedip bending moment thrusts [Whitney, 1991] or of the deviatoricstress field from the regionalstress field. The a second set of back thrusts related to the basement buttress. orientationof the deviatoricstress field at shortwavelengths The upliftand rotation of sedimentsin the hangingwall may can vary widely from the regional stressfield becauseof alsoexplain the apparent down section change in stratigraphic roughnesson the surfacesof the major faults or the fractal level of the d6collement between the Central and Eastern natureof fault systemsproducing a mix of focal mechanism precordilleras. types. While seismicitycannot identify the type of basement, Along-striketectonic variations. Simple subduction models comparingthe Precordillera activity with thatof thePampean and the associatedtectonics are typically two-dimensional, rangeSierra Pie de Palo immediately to theeast shows that the modelingupper plate deformation as a seriesof lineartectonic rateof earthquakeoccurrence and the seismicity pattern of the provincesperpendicular to the convergencedirection with no two areasare very different.One explanationis that the along-strikestructures. While this is valid at large scales, activity level is relatedto the time sincethe last large or manysubduction systems have along-strike and therefore three- characteristicevent. In the EasternPrecordillera, the last large dimensional variations. The Precordillera and Sierras event was the 1944 Ms 7.4 San Juan earthquake,while Pampeanasterranes are examples of such regional-scale beneath Pie de Palo it was the 1977 Ms 7.4 Caucete variations. Within these terranes,which are two-dimensional earthquake.Although the times involved here are much longer on a regionalscale, we find along-strikevariations on the scale thanthose associated with normalaftershock activity, ongoing of individualmountain blocks and ranges. activitybeneath Pie de Palo (Regnieret al., 1992) is very In San Juan a sharpeast-west striking linear featurecuts similarto the first monthof the 1977 eventaftershock activity across the Sierras Pampeanas-Precordilleraboundary near [Bollingerand Langer, 1988] and to theactivity during 1980, 31.1øS.On Pie de Palo, it is definedby steeplydipping faults 1984,and 1985ISmalley and Isacks, 1990]. The seismicityis thatstrike east-west across the northernboundary of the range. thereforeremarkably persistent and, like the immediate These faults are downthrown on the north and show evidence aftershocks,may continueto representthe redistribution of of Quaternary activity [Bastias and Weidmann, 1983; stress from the 1977 event onto a network of smaller faults. Zambrano,1969]. Inferredand mapped continuations of these Neitherthe activity before the Caucete earthquake nor thatof faultsare foundin the Sierrade Valle F6rtil, in otherPampean theother Pampean ranges is knownwell enoughto determine ranges to the east, and in the Precordillerato the west if theactivity level and pattern associated with Pie dePalo is [Zambrano,1969]. An along-strikechange in the structureof the normalbackground activity in the SierrasPampeanas. In the EasternPrecordillera, from thrustsin the south to folds in the Eastern Precordillera, the time since the last large the north, also occurs near 31.1øS where the southernend of earthquakeis greaterthan that for Pie de Palo,and stresses Sierra de Motado bifurcates. The Central and Western transferredto theregion surrounding the mainshock fault may precordillerasalso have many faults with evidenceof Holocene have been relieved by past seismicity.The stressesmay activity southof 31.1øS, but few Holocenefaults are foundto concentratenear the main fault again, which is therefore the north [Bastias and Weidmann, 1983]. Several unusual imagedin the seismicity.Activity associatedwith the closed basins are also found in the Central Precordillera north Precordillera,both before and for the first 30 yearsafter the of 31.1øS,and fartherwest a major structuralhigh separates 1944earthquake, is alsotoo poorly known to determinethe the Calingastaand Iglesia valleys at about3 IøS. activitypattern as a functionof timein theearthquake cycle. The across-strikefeature is alsocoincident with the major No seismicity pattern is visible using Preliminary changein basementdepth across the southernmargin of the Determinationof Epicentersand International Seismological BermejoValley. UnpublishedYPF seismicreflection data Centerdata because too few eventsare reported. Another major suggesta depthto basementin the footwallblock beneath differencebetween the two regionsis the seismicdepth Sierra de Valle F6rtil of 8kin or 17km [Jordan and distribution.Beneath Pie dePalo, seismicity is concentratedin Allmendinger, 1986]. The difference depends upon a pronouncedsubhorizontal midcrustal layer near 25 kmdepth, interpretationof a 9 km thick packageof reflectorsat the andthe maximum depth is 30 km (Regnieret al., 1992).This bottom of the seismic section. A minimum structural relief of midcrustalconcentration is absentin the Precordilleraactivity, 11 km thereforeexists between the peaksof SierrasPie de and the maximumdepth is 35 km. The very different Palo and Valle F6rtil and the footwall basement block under seismicitypatterns suggest that the two areashave distinct Sierra de Valle F6rtil. The basementsurface in the Bermejo basementswith differentcrustal rheology. Valley thereforedips almost opposite the inferredsouthward The focalmechanisms presented here are the first single- dip of theseismically defined fault in thenorthern segment. eventmechanisms from the Precordillera(Figures 6 and7). The basementblock in the Bermejo Valley mustplay an Althoughthere are strike-slip and normal mechanisms, most importantbut poorlyunderstood role in theregional tectonics. of the mechanismsare thrusts,as expectedfor regionaleast- It is triangularshaped and knownto be boundedby crustal- westcompression. In the central segment, focal mechanisms scale faults on the east, where it is the footwall block beneath of eventsshallower than about20 km and deeperthan about Sierra de Valle F6rtil, and on the south, where it terminatesin 30 km indicatethrust faulting, while eventsbetween 20 km vertical east-weststriking faults crossingthe north end of and30 km depthindicate normal faulting. We wereunable to SierraPie de Palo. A crustal-scalethrust fault is also thought obtain focal mechanismsolutions for events deeper than to existalong the westside [Allmendinger et al., 1990].The 30 km, but the first motionsat the closeststations, those valley may be flooredby a crustalblock being pushedor plottingin thecenter of thefocal sphere, are compressions, rotatedout the bottomof the crustdue to the regionaleast- suggestinga return to thrustfaulting deeper than 30 km. The westshortening and the weightof sedimentsin the Bermejo 74 Smalleyet al.: BasementSeismicity in AndeanPrecordillera

Valley. Southof the BermejoValley, shorteningis expressed of crustalscale basement faults beneath the eastern part of the in the uplifted mountainblock Sierra Pie de Palo. In the Precordilleratectonic province of the Andeanforeland. The valleyson either side of Pie de Palo, basementis shallow, geometryof thesefaults does not correlatewith the surface typically only 1 to 2 km [Gray de Cerdan, 1969; YPF, geologyand showsthat a simplestructural relationship does unpublishedrefraction and reflection dam]. not exist between the basement deformation and deformation in Beneath and west of the network a roughly east-west the overlyingsediments. Improved locations (due to network orientedlinear concentrationof WBZ seismicity(Figures 3 density,use of three-componentsensors, and application of the and4) coincideswith an along-strikecontinuation of the Juan JHD method)allowed us to developnew modelsfor the Fernandezridge of theoceanic Nazca plate ISmalley and Isacks, anomalous structures of the Eastern Precordillera and the 1987]. The WBZ here is subhorizontal, and subductionof relationof theEastern Precordillera deformation to underlying buoyantaseismic ridges has been proposed as a possiblecause basementdeformation expressed in the seismicity.The two for such fiat subduction[Cross and Pilger, 1979; Pilger, end-membermodels are (1) the deformationsare unrelatedand 1981]. As the seismicityfollows an along-strikeprojection of their spatial overlap is a coincidenceand (2) the Eastern the ridge,it supportsthe proposalthat the ridge continuesin Precordilleraand the basement seismicity are relatedthrough thesubducted plate. Neogenereactivation of a Devonianstrike-slip suture between The intermediate-depthand crustal seismicity also exhibit the SierrasPampeanas and Precordillera.Figure 10 is an poorlydefined southern boundaries near 32.25øS and share the orogen-widecross section combining the seismicityresults sharpboundary near 3 IøS thatis coincidentwith thetransverse with other establishedfeatures of the Andeanorogen. The featuresof theupper plate geology described above. The crustal basementfaults may alsohave been responsible for the 1944 seismicityis concentratedbeneath Pie de Palo,just eastof an San Juanearthquake and may continueto posesignificant intensenest of intermediate-depthseismicity beneath La Laja. seismic risk in the San Juan area. Lack of surface effects from The intriguingspatial correlations between the large-scale largeevents in the basementsuggests that the overallseismic mountainblock structures and crustal seismicity of theupper risk in San Juan may be higher than that estimatedfrom plate with the intermediate-depthseismicity suggestsan surfacegeologic studies alone. The dataalso suggest complex interaction of a subducted extension of the Juan Fernandez structuralrelationships between previously unidentified surface ridgewith theupper plate through a lithosphericscale structure geologicfeatures in the San Juanarea and the Precordillera in the upperplate that is aseismicbetween 35 km and95 km. basementfaulting.

CONCLUSIONS Acknowledgments.We are indebtedto J. C. Castano,and R. Recio of the Instituto Nacional de Prevenci6n Sfsmica The high resolution afforded by a temporaryseismic (INPRES)in SanJuan, Argentina, for scientificand logistical networkin SanJuan, Argentina, images a segmentedpattern support.We also thank our engineerG. Steiner (CERI),

0km ,n•n ,n ,n ,n ,n ,n ,n •, :.'r.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'."•.•,'./..'..'.'••F• , ß ß ß %• ß ß %•% ß ß ß '"'"":"•::•...... i•....:•...•:...•...... ,,• :..:.. :.. :.•,:.. :.. :.•-•. ,. :..:......

100kin

300 km 200km 400km 600km Fig. 10. One-to-onecartoon showing relationship of fault systemimaged with our data to Precordillera- SierrasPampeanas suture and formation of anomalousstructures of theEastern Precordillera. The major terranesof theAndes at 31øS(Chilenit, the Precordillera and the Sierras Pampeanas), their boundaries, and importantgeographic features are identified. Smalleyet al.: BasementSeismicity in AndeanPrecordillera 75 techniciansJ. Bollwerk(CERI), andF. Bondoux(ORSTOM), KatherineKadinsky-Cade helped improve the paper. This work and field personnelJ. Vlasity(CERI), D. Vlasity(CERI), was supportedby NationalScience Foundation grants EAR- T. Cahill (Cornell), C. Avila (INPRES), and M.C. Reta 8608301and EAR-8804925, the Institut Franqais de Recherche (INPRES).We aregrateful to Mike Ellis, Rick Allmendinger, Scientifique pour le Drveloppement en Cooprration andTerry Jordan for manyinteresting discussions and useful (ORSTOM), and the Stateof TennesseeCenters of Excellence comments.Constructive reviews by David James and Program.

REFERENC•

Allmendinger,R., D. Figueroa,E. Snyder,J. Cross,T.A., and R.H. Pilger, Jr., Controlsof Ortiz, A., and J.J. Zambrano, La provincia Beer, C. Mpodozis, and B.L. Isacks, subduction geometry and consequent geo16gicaPrecordillera oriental, Actas del Forelandshortening and crustalbalancing locationof magmaticarcs, Geol. Soc.Am., VIII Congr. Geol. Argent., 1, pp. 59-74, in the Andes at 30 ø latitude, Tectonics, 9, Boulder,Abstr. Programs, I1, 407, 1979. AsociacionGeologies Argentina, Buenos 789-809, 1990. DaMstrom,C.D.A., Structuralgeology in the Aires, Argentina,1981. Baldis, B., and O. Bordonaro,Ctrnbrico y eastern margin of the Canadian Rocky PanJan, J., and W. Pilant, A possible Ordov•cicode la Sierra Chica de Zonda y Mountains, Bull Can. Pet. Geol., 18, 332- explanationfor forelandthrust propagation, Cerro Pedemal, provincia de San Juan, 406, 1970. J. Geophys.Res., 95, 8607-8616,1990. Grnesis del margert continental en la Dalla Salda, L.H., and R. Varela, E1 Pilger, R.H., Jr., Plate reconstructions, Precordillera, Actas del IX Congr. Geol. metamorfismo en el tercio sur de la Sierra aseismicridges, and low-anglesubduction Argent., IV, pp. 190-207, Asociacirn Pie de Palo, San Juan,Asoc. Geol. Argent. beneath the Andes, Geol. Soc. Am. Bull., Geolr•ica Argentina,Buenos Aires, Rev., 34, 68-93, 1984. 92,448-456, 1981. Argentina,1984. Davis, D., J. Suppe, and F.A. Dahlen, Puj•l,J., J.-M.Chiu, R. SinalleyJr., M. Baldis,B.A., Acercade la estructuraprofunds Mechanics of fold-and-thrust belts and t•egnier, B.L. Isacks,J.-L. Chatelain,J. en la Precordillera Central, Rev. Geol. accretionarywedges, J. Geophys.Res., 88, Vlasity, D. Vlasity, J.C. Castano,and N. Mineral.,XXXIII(l/2), 13-17, 1975. 1153-1172, 1983. Puebla, Lateral velocity variationsin the Baldis, B.A., M.S. Bereft, O. Bordonaro,and Dickenson,W.R. and W.S. Snyder,Laramide Andean foreland in Argentina determined A. Vaca, Sintesis evolutiva de la folding associatedwith basementblock with the JHD method, Bull. Seismol. Soc. Precordillera Argentina, Actas del IX faulting in the western United States, in Am., 81, 2441-2457, 1991. CongresoLatinoamericano deGeolog[a, IV, Plate Tectonicsof the Laramide Orogeny, Ramos, V.A., T.E. Jordan, R.W. pp. 399-445,Organizado pot el Servicio editedby V. MatthewsTII, Mem. Geol. Soc. Allmendinger,C. Mpodozis,S.M. Kay,J.M. Geolergica Nacional, Buenos Aires, Am., 151,355-366, 1978. Cortrs, andM. Palma,Paleozoic terranes of Dziewonski, A.M., G. Ekstrom, J.E. Franzen, the central Argentine-Chilean Andes, Argentina,1982. Tectonics,5, 855-880, 1986. BaBy, A.W., Thoughtson the tectonicsof and T.H. Woodhouse, Centtold-moment tensor solutionsfor July-September1986, Rassmuss,J., Rasgosgeo16gicos generales de foldedbelts, in Thrustand Nappe Tectonics, las SierrasPampeanas, Dir. Gral. Minas. pp. 13-32,Geol. Soc. Am., Boulder, 1981. Phys. Earth Planet. Inter.,46, 305-315, 1987. Geol. Hidrol., 13, 78-89, 1916. Barazangi, M., and B.L. Isacks, Spatial Regnier,M., J.-L. Chatelain,R. Smalley,Jr., distributionof earthquakesand subduction Fielding, E.J., and T.E. Jordan, Active deformationat the boundarybetween the J.-M. Chiu, B.L. Isacks, M. Araujo, of the Nazcaplate beneath South America, Seismotectonicsof the Sierra Pie de Palo, a Precordillera and Sierras Pampeanas, Geology,4, 686-692, 1976. basement block uplift in the Andean Bastias,H.E., andN.E. Weidmann,Maps de Argentina, and comparisonwith ancient Rocky Mountain deformation,in Rocky foreland, Argentina, Bull. Seismol. Soc. fallamientomodemo de la provinciade San Mountain Foreland and the Cordilleran Am., in press,1992. Juan,scale 1:500,000, Inst. Nacl. de Prev. Salfity, J.A., and S. Gorustovich, Sismica,San Juan, Argentina, 1983. Thrust Belt, editedby W.J. Perry and C.J. Schmidt,pp. 143-163, Geol. Soc. of Am., Paleogeograffade la cuenca del grupo Bollinger, G.A., and C.J. Langer, Boulder, 1988. Paganzo(Paleozoico Superior), Asoc. Geol. Developmentof a velocity model for Argent.Rev., 38, 437-453, 1983. locatingaftershocks in the SierraPie de Gray de Cerdan, N.A., Bosquejo Sinalley, R., Jr., J.M. Chiu, J. Pujol, A.C. Palo regionof westernArgentina, U.S. geomorfo16gicodel Bols6n de Tulum, San Johnston, G.C. Steiner, J. Bollwerk, B.L. Geol. Surv.Bull., 1795, 1988. Juan-Argentina, Bol. Estud. Geogr. Isacks, T. Cabill, D. Whitman, M. Regnier, Mendoza, 17, 215-250, 1969. J.-L. Chatelain, F. Bondoux, J.C. Castano, Bowman, J. R., Geodetic evidence for Groeber, P., Movimientos tect6nicos and N. Puebla,Automatic determination of conjugatefaulting during the 1988Termant contempor•eos,Univ. Nac. La Plata Notas. Creek, Australia, earthquake sequence, Mus. Geol., 9, 263-375, 1944. hypocenters:Results from the PANDA Geophys.J. lnt., 107, 47-56, 1991. digital network deploymentin San Juan, Instituto Nacional de Prevenci6n Sismica Argentina,paper presented at 61st Annual Cahill, T., B.L. hacks, D. Whitman, J.-L (INPRES), Zonificacirn Sismica de la Meeting,Seismological Society of America Chatelain, A. Perez, and J.-M. Chiu, Repdblica Argentina, Publ. Tdc., 5, San Seismicityand tectonics in Jujuyprovince, EasternSection, Oct. 11-12,1989. Juan,Argentina, 1977. Smalley, R., Jr., and B.L. Isacks, northwesternArgentina, Tectonics, ill, Instituto Nacional de Prevenci6n Sismica Seismotectonics of thin and thick-skinned 944-959, 1992. (INPRES), Microzonificacirn Sismica del deformation in the Andean foreland from Caminos, R., Cordillera Frontal, Segundo Valle de Tulum - Provinciade San Juan, local network data: Evidence for a Simposiode Geolog;aRegional Argentina, Infotree Trcnico General, tres tomos, seismogeniclower crust,J. Geophys.Res., pp. 397-454, Acad. Nac. Ciencias, InstitutoNational de Prevenci6nSismica, 95, 12487-12498,1990. C6rdoba,Argentina, 1979. San Juan, 1982. Sinalley,R.F., Jr., and B.L. Isacks,A high Caminos,R., C.A. Cingolani,F. Hervr, andE. hacks, B.L., Uplift of the central Andean resolutionlocal networkstudy of the Nazca Linares,Geochronology of the pre-Andean Plateau and bending of the Bolivian Plate Wadati-Benioff zone under western metamorphismand magmatismin the Orocline,J. Geophys.Res., 93, 3211-3231, Argentina, J. Geophys.Res., 92, 13903- Andean Cordillera between latitudes 30 ø 1988. 13912, 1987. and 36øS,Earth $ci. Rev., 18, 253-283, Jordan,T.E., and R.W. Allmendinger,The Stauder, W., Mechanism and spatial 1982. SierrasPampeanas of Argentina:A modem distribution of Chilean earthquakeswith Castellanos,A., E1 terremorode San Juan,in analogue of Rocky Mountain foreland relationto subductionof the oceanicplate, CuatroLecciones Sobre Terremows, pp. 79- deformation, Am. J. Sci., 286, 737-764, J. Geophys.Res., 78, 5033-5061,1973. 242, AsociacirnCultural de Conferencias 1986. Sutrez, G., P. Molnar, and B.C. Burchfiel, de Rosario,Rosario, Argentina, 1945. Jordan,T.E., B.L. Isacks,R. W. Allmendinger, Seismicity,fault plane solutions,depth of Chapple,W.M., Mechanicsof thin-skinned J.A. Brewer, V.A. Ramos,and C.J. Ando, faultingand activetectonics of the Andes fold-and-thrustbelts, Geol. Soc. Am. Bull., Andeantectonics related to geometryof of Peru, Ecuador, and southernColumbia, 89, 1189-1198, 1978. subductedNazca plate, Geol. Soc. Am. J. Geophys.Res., 88, 10403-10428,1983. Chinn, D.S., and B.L. Isacks,Accurate source Bull., 94, 341-361. 1983. Toselli, A.J., J.R. de Toselli,and F. Acefiolaza, depthsand focal mechanismsof shallow Kadinsky-Cade,K.• Seismotectonicsof the Milonitas de bajo gradode la megafractura earthquakesin westernSouth America and Chile margin and the 1977 Caucete de SierrasPampeanas en la Quebradade in the New HebridesIsland Are, Tectonics, earthquakeof western Argentina, Ph.D. La Rioja, Sierrade Velasco,Argentina, 4th 2, 529-563, 1983. dissertation, Comell Univ., Ithaca, N.Y., Geol.Congr. Chile, 159-171, 1985. Chiu, J.M., G.C. Steiner,R. Smalley,Jr., and 1985. Volponi,F.S., Los terremowsde Mendozadel A.C. Johnston,the PANDA seismicarray - Kanamori, H., and D.L. Anderson, 21 de Octubrede 1968 y la estructurade la a simple,working system,Bull. Seismol. Theoretical basis of some empirical corteza terrestre,Acta Cuyana Ing., XII, Soc.Am., 81, 1000-1014, 1991. relationsin seisinology,Bull. Seismol.Soc. 95-110, 1968. Cingolani,C.A., andR. Varela,Geocronologia Am., 65, 1073-1095,1975. Volponi,F.S., Reportof the Argentina-United Rb-Sr de las rocasigneas y metamrrficas Klein, F.W., HypocenterLocation Program States binational symposiumabout the de Jassierras Chics y Grandede C6rdoba, HYPOINVERSE, part 1, Users guide to Caucete earthquakeof November 23rd, Rep. Argentina,Congr. Ibero Am. Geol. versions 1, 2, 3, and 4, U.S. Geol. Surv. 1977,44 pp., Univ. Nac. de SanJuan, Inst. Econ. Mem. 2nd, 33-91, 1975. OpenFile Rep. 78-694, 1978. Sistool.Zonda, San Juan, Argentina, 1979. 76 Smalleyet al.: BasementSeismicity in AndeanPrecordillera

Whitney, R. A., Faulting and tectonicsin the de Prey. Sfsniica,San Juan, Argentina, B. L. Isacks, Institute for the Study of the foreland basin fold/thrust belt of the San 1969. Continents,Cornell University,Ithaca, NY Juan province, Argentina, and a 14853. comparisonto the Yakimafold/thrust belt of M. Regnier,Institut Francaisde Recherche the northwestern United States, Ph.D. M. Araujo and N. Puebla,Instituto Nacional Scientifique pour le Developpementen dissertation,Univ. of Nev., Reno, 1991. de Prevenci6n Sismica, 5400 San Juan, Coop6ration, BPA 5, Noumea, New Argentina. Zamarbide, J.L., and J.C. Castano, Caledonia,South West Pacific. Determinaci6n de los coeficientes s{smicos J.-L. Chatelain,Institut Francaisde Recherche Scientifique pour le Developpementen zonalespara la RepdblicaArgentina, Publ. Coop6ration,Domaine UniversitaireBP Tdc. 6, Inst. Nac. de Prev. Sismica, San 53X, 38041 GrenobleCEDEX, France. Juan, 1978. J. M. Chiu,J. Pujol,and R. Smalley,Jr., Center Zambrano, J.J., Mapa geo16gico de la for EarthquakeResearch and Information, (ReceivedJuly 19, 1991; provinciade San Juany alrededores,scale MemphisState University,Memphis, TN revisedMay 4, 1992; 1:500,000, Mal• M-S J3697(93), Inst. Nac. 38152. acceptedMay 11, 1992.)