State of Stress in the Conterminous United States

Home , Jack Oliver (scientist), Mary Lou Zoback, Ramapo Fault, Rio Grande Rift

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

USGS Staff -- Published Research US Geological Survey

1980

Mary Lou Zoback U.S. Geological Survey, [email protected]

Mark D. Zoback U.S. Geological Survey, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/usgsstaffpub

Part of the Earth Sciences Commons

Zoback, Mary Lou and Zoback, Mark D., "State of Stress in the Conterminous United States" (1980). USGS Staff -- Published Research. 453. https://digitalcommons.unl.edu/usgsstaffpub/453

This Article is brought to you for free and open access by the US Geological Survey at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in USGS Staff -- Published Research by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. JOURNALOF GEOPHYSICALRESEARCH, VOL. 85, NO. BII, PAGES6113-6156, NOVEMBER 10, 1980

State of Stress in the Conterminous United States

MARY LOU ZOBACK AND MARK ZOBACK

U.S. GeologicalSurvey, Reston, Virginia 22092

Inferringprincipal stress directions from geologicdata, focal mechanisms, and in situstress measurements,we haveprepared a map of principalhorizontal stress orientations for the conterminousUnited States.Stress provinces with lineardimensions which range between 100 and 2000km weredefined on thebasis of thedirections and relative magnitude of principalstresses. Within a givenprovince, stress orientationsappear quite uniform (usually within the estimated range of accuracyof thedifferent methods usedto determinestress). Available data on the transitionin stressdirection betweenthe different stressprovinces indicate that these transitions can be abrupt,occurring over <75 km in places.In the westernUnited States, a regionof activetectonism characterized byhigh levels of seismicityand generallyhigh heat flow, the stress pattern is complex,but numerous stress provinces can be well delineated. Despiterelative tectonic quiescence in the eastern and central United States, a major variation in principal stressorientation is apparentbetween the Atlantic Coastand midcontinentareas. Most of the eastern UnitedStates is markedby predominantlycompressional tectonism (combined thrust and strike slip faulting),whereas much of the regionwest of the southernGreat Plains is characterizedby predominantlyextensional tectonism (combined normal and strike slip faulting). Deformation along the San Andreasfault and in partsof theSierra Nevada is nearly pure strike slip. Exceptions to this general pat- terninclude areas of compressionaltectonics in the westernUnited States (the PacificNorthwest, the ColoradoPlateau interior, and the Big Bend segment of theSan Andreas fault) and the normal growth faultingalong the Gulf CoastalPlain. Sources of stressare constrained not only by theorientation and relativemagnitude of the stresseswithin a givenprovince but alsoby the mannerof transitionof the stressfield from one province to another.Much of the modempattern of stressin the westernUnited Statescan be attributedto presenttransform motion and residual thermal and dynamic effects of Ter- tiarysubduction along the western edge of theNorth American plate. Abrupt stress transitions around activelyextending regions in the westernUnited Statesprobably reflect shallow sources of stressand anomalouslythin lithosphere.Large areas characterized by a uniformstress field in the centraland east- ernUnited States suggest broad scale plate tectonic forces. In themidcontinent region, both ridge push andasthenospheric viscous drag resistance to lithospheric motion can explain the NE-SW compression in thecold, thick lithosphere of the craton, although drag-induced stress directions (resistance to absolute platemotion) correlate better with the data than do the ridge push directions. Asthenospheric counter- flowmodels do not apply in thisregion, as predicted stress orientations are about 90 ø off.A regionof compressionoriented approximately perpendicular to the continental margin and Appalachian fold belt is definedby thestress data along the Atlantic Coast. This NW-SE compression is in directcontrast with previousmodels predicting extension perpendicular topassive continental margins due to lateraldensity contrastsat the continental-oceanic crust interface. Ridge push forces, while capable of producinga componentof compressionacross the coastal area, do not explainthe observedorientation of thestress. Two speculativemechanisms are suggestedto explain the observedorientations: (1) rotationof stress(or strain)due to anisotropicAppalachian basement structure and (2) flexuraleffects associated with erosion and isostaticrebound of the Appalachians.

INTRODUCTION Data are presentedin terms of the orientation and relative Detailedknowledge of the patternof intraplatestress pro- magnitude of maximum and minimum horizontal stressesand videsimportant constraints on modelsof globaltectonic proc- a verticalstress. We haveassumed that oneprincipal stress is essesand the mechanismof plate motions.On a finer scale, vertical, and the horizontalcomponents of the in situ stress boththe patternof stressand variations in the patternmust be fieldrepresent principal stresses. Evidence supporting this as- known to understand intraplate volcanism and tectonism sumptionincludes observations of the nearlyvertical attitude properly.In addition,in regionsof relativetectonic quies- of dikesexposed over significant depth intervals at a givenlo- cence,knowledge of the in situstress field is requiredfor de- cality[R. B. Johnson,1961] as well asthe absenceof exposed lineationof potentialseismic hazards associated with pre- dikesthat appearnot to havebeen eraplaced vertically (D. existing zones of weaknessin the crust. Pollard,oral Communication,1980). In addition,measure- This studyrepresents an attemptto map the modernstress mentsof the completestress tensor in deepmines [McGarr field (_primarilyQuaternary in the westernUnited Statesand and Gay, 1978] and the near-horizontal orientations of the Tertiary and youngerin the East)in the conterminousUnited greatmajority of stressaxes inferred from earthquakefocal States. Principal stress orientations have been determined mechanismsalso indicate that regionally the principal stresses from geologicobservations, earthquake focal mechanisms, are horizontal and vertical. and in situ stressmeasurements. An attempthas also been The majorcontribution of the presentstudy is the inclusion madeto categorizebroad regions not only by the orientations of new geologicdata on the orientationof the stressfield, of the principalstresses but alsoby their relativemagnitudes, largelyin the westernUnited States.In addition,we have at- as inferredfrom currentlyactive tectonism. temptedto go backto the originalreferences and checkthe reliability of previouslycompiled data, particularlyfor focal Thispaper is not subjectto U.S. copyright.Published in 1980by mechanisms.We haveexcluded poorly constrained points and the AmericanGeophysical Union. haverelied, whenever possible, on averagesof a numberof so- Paper number 80B 1093. 6113 6114 ZOBACKAND ZOBACK.'STATE OF STRESSIN CONTERMINOUSUNITED STATES

lutions in a given locality. The quality criteria used are de- break can be identified, the maximum vertical and strike slip tailed below. offsets(along the same segmentof the break) can be used to determinethe net horizontalcomponents of slip. PRINCIPAL STRESS ORIENTATION INDICATORS Large-scalegrooves and parallel slickensideson an exposed Some discussionis warranted of the different techniques fault surface indicate the relative direction of motion of two used for determining principal stressorientations, especially crustal blocks. Unfortunately, well-preservedfault scarpsin the major assumptionsassociated with eachtechnique as well competentrocks are rarely exposed.In the Basin and Range as the inherent difficultiesand uncertainties.Incorporation of province, a region of abundant Quaternary faulting, most the geologicstress indicators is our main contributionto pre- young scarps occur in alluvium, basinward of the main vious compilationsof stressdata. ranges.However, surfaceexposures of fault scarpsare known and are sometimesexposed several kilometers (occasionally GeologicData tens of kilometers)along strike.Measurement of groovesand Geologicinformation on principal stressorientations can be slickensidesat a numberof sitesalong a fault providesa mean divided into two main categories:observations of fault slip horizontal direction of slip for a major crustal block. Con- and the alinements of young (<5 m.y.B.P., predominantly sistencyof the horizontalcomponent of slip despitechanges in Quaternary)volcanic feeders. For the westernUnited States, strike of as much as 90ø [M. L. Zoback, 1978; Pavlis and many of thesedata came from M. L. Zoback [1979], Thomp- Smith, 1980]indicates that thesegrooves and slickensidesac- son and Zoback [1979] and M. L. Zoback and M.D. Zoback curately record the major block motion. One sourceof error [1980]. In the east,geologic data on young faulting compiled in this kind of studyis the possibilityof a local downdipcom- by Prowell [1980] was especiallyuseful. ponent of motion of gravitationallyunstable blocks that slide Detailed information on fault slip can be usedto determine down into the valley. This sourceof error could result in mea- the net horizontal componentof motion on oblique slip nor- sured slip directionsthat are spuriouslyrotated toward the mal and thrust faults. Measurementsof fault slip direction valley. yield the net directionof horizontal shorteningin the caseof Some of the geologicslip data reported here were derived oblique slip thrust faults, and the net horizontalextension di- from detailed studiesinvolving a number of localities on a rection on oblique slip normal faults. The occurrence of given fault surface,whereas other data representonly a single oblique slip suggestsslip on favorably oriented preexisting measurement(see Table 1 for details).The standarddeviation planesof weakness(i.e., faults). We assumehere that the ac- of the mean horizontal componentof slip in the detailed stud- tual horizontal slip direction reflectsthe regional stressfield. iesis ñ10ø-15'?, which is probablya reasonableestimate of Available strain data for the western United States indicate the accuracyof all the fault slip data despitethe fact that indi- that the regionalstress and strain axes,in general,do coincide vidual measurementsgenerally have a precisionof :1:2ø. [e.g., Prescottet al., 1979],and in detail, in situ stressand re- An additional method of determining fault slip utilizes regional strainmeasurements at the Nevada Test Site yield con- cent observationsof coreholes offset by motion on preexisting sistentprincipal axes [S. W. Smith and R. Kind, 1972; Haim- fault surfaces. These core hole offsets are considered to be re- son et al., 1974].Thus in a normal faulting regime, suchas in liable stressindicators when (1) fault motion occursyears af- the Basin and Range and the Guff Coast provinces,the hori- ter excavation,and (2) the senseof motion is not causedby zontal componentof slip (or opening)is inferred to be region- the creation of a stress-freesurface and thus is gravitationally ally in the directionof the leastprincipal stress.Similarly, in a controlled(see the work of Schiifer[1979] in the Appalachian thrust faulting regime the net horizontal slip shouldbe in the Fold Belt). direction of the greatestprincipal stress. Where detailed information on the direction of slip is una- The actual directionof slip on a preexistingfault plane oc- variable,an approximatemethod for determiningstress orien- curs in the direction of maximum resolved shear stress on that tation relieson the trend of youngfaults and the senseof only plane when the ratio of shearto normal stressexceeds the fric- the predominanttype of offset(generally vertical, either nor- tional strength.To predictaccurately the directionof slip on a mal or reverse).This method has been used for the Atlantic fault with a given orientation, both the orientation of the re- CoastalPlain and Gulf Coastareas. The assumptionof purely gional principal stressesand their relative magnitudesmust be dip slip faulting for active Gulf Coast normal faults is prob- known [e.g., Wallace, 1951]. Thus it is impossibleto infer ably valid. For the Atlantic Coastal Plain this method was uniquelyprincipal stress orientations from observationsof slip usedbecause it providedthe only availabledata; however,the on a singlefault. However, if slip observationsare made for a possibleerrors are recognizedas potentially being quite large family of faults influenced by a single regional stressfield, (perhapsas large as 20ø-30ø;see Wallace[195'1] and Raleigh then these observations can be inverted to determine the ori- et al. [ 1972]). entation of the principal stresses[e.g., Angelier, 1979; Carey, The secondtype of geologicindicator of principal stress 1979]. Unfortunately, such detailed studies have not been entation is linear volcanic feeders, such as dikes and cinder done in the United States.The fault slip data included here conealinements, where the cindercone alinement presumably constrainonly the approximateprincipal stressorientations. reflectsthe geometryof an underlying fissureor dike. E. M. Two methods of determining fault slip geologicallyare Anderson[1951] and Odd [1956]have concludedon theoretical from measurementsof (1) historicoffsets and (2) groovesand grounds that dike intrusion should follow planes per- slickensideson exposedfault scarps.Of these two methods, pendicular to the axis of least principal stresswithin a rock data from the first are often the most diffcult to obtain and mass,resulting in magma-fracing[Yoder, 1976] analogousto the least reliable. Surfacescarps of earthquakesare rarely a hydraulic fracturing. Clearly, there can be no static shear singletrace but rather a zone of breaks,so a completepicture stressacross a magma-filledcrack, and numerousexamples of slip requiresknowledge of the total dip slip and strike slip exist of dike swarms that crosscut earlier structures and main- componentsof motion for all breaks. However, if a major tain a constanttrend in rockscontaining abundant preexisting ZOBACKAND ZOBACK:STATE OF STRESSIN CONTERMINOUSUNITED STATES 6115 fractures and faults [e.g., Christiansenand McKee, 1978, p. however,can only be shownto be of Tertiary or Quaternary 294]. Koons[1945], in a careful studyof late Quaternarybasalt age. fields north of the Grand Canyon, saw through the possible Focal Mechanisms confusion of parallel cinder cone alignments and nearby faults: PressureP and tension T axes derived from earthquake Detailed examination of exposed bedrock near these aligned focal mechanismsare one of the most commonly used in- groupsshows that, thoughthe linesof conesparallel fault trends dicatorsof tectonicstress. However, principal stressorienta- the cones do not lie on observed fault lines ... cones tend to occur tions obtainedfrom fault plane solutionsare inherentlythe in distinctlines, parallel to, but not associatedwith, observedsur- least reliable indicators of stress orientation because the P and face fractures.This parallelism may have resultedfrom stresses in existence at the time of deformation. T axes cannot be equated with certainty to the greatestand least principal stressdirections. McKenzie [1969] demon- In addition, Nakamura [1977] and Nakamura et al. [1978] stratedthat for the generalcase of triaxial stressthe only re- convincinglydemonstrated the correspondencebetween lin- strictionon the greatestprincipal stressimposed by the fault ear volcanic feeders and regional stressas determined from plane solution is that this stressdirection must lie in the same studiesof active faults, earthquakefocal mechanisms,and the quadrant as the P axis but could, in fact, be nearly normal to directionof convergencealong major plate boundaries.Thus the P direction.McKenzie's objections were basedlargely on in regionsof active volcanism,linear volcanicfeeders can be the fact that most shallowcrustal earthquakes occur on pre- used to determine principal stressorientations. Lachenbruch existingfaults rather than by the fracturingof intact homoge- and Sass[1978] also have suggestedthat the axisof maximum neousmaterial and that shearstresses at shallowdepths are elongationof calderasshould correspond to the regional least much too smallto causefailure. Raleighet al. [1972],however, principal stressdirection. In the presentstudy, however,we pointed out that the strengthof intact homogeneousrock is have usedonly dike trends and cinder cone alinementswhich suchthat new faults will begenerated where faults exist at un- can be determined with greater precision. favorableorientations. Experimentally derived faulting rela- Stress directions inferred from cinder cone alinements are tions reportedby Raleigh et al. suggestthat only preexisting generallybased on linear alinementsof four or more vents. planeslying at • 10ø-50ø to S,, the greatestprincipal stress di- Some of these linear zones are quite impressive,extend for rection, would slip (i.e., when S, is between 40ø-80 ø to the nearly 20 km, and contain as many as 16 individual eruptive normal to the fault plane). Therefore they concludedthat the centers(e.g., the Don Carlos Hills in northeasternNew Mex- P axis (taken as lying at 45ø to the fault plane) could be in er- ico, site NM-22, Table 1). Dike trends are based on a visual ror by no more than 35o-40ø when sliding on a preexisting regional average in which the longestand most continuous fault producesthe earthquake.Furthermore, they suggested dikes generallyare given the greatestweight. Again, as in the that if the nodal plane correspondingto the fault is known, case of the fault slip data, individual measurementsmay be the P axis (correspondingto the greatestprincipal stress) made with an accuracyof +_1 ø-2ø. However, variationsin any shouldbe plotted at 60ø of the normal of the fault plane and given area suggestthat a more reasonableassessment of the at 30ø to the slip direction. In this casethe orientation of the reliability of the stressorientation determination is _+5ø-10ø. greatestprincipal stresswould be in error by no more than Measured ellipticity of wells resultingfrom caving of walls 20 ø . (breakouts)has also been proposedas a principal stressin- However, selectionof the actual fault plane is frequently dicator. Cox [1970] and Babcock[1978] have reported consis- difficult and the P and T axes standardlyreported are at 45 ø tent orientations of elongationswithin individual wells and to the nodal planes. Thus it appearsthat the best method of between wells distributed over an area of more than 3 x 105 analyzing focal mechanism data is to consider a number of km2 on the high plains of Alberta, Canada. In one 200 x 250 earthquakesoccurring on differentfaults in a particular area km region, nine wells showedonly a 20ø variation in mean and then to rely on averageP and T directions.In this way the predominantelongation orientations (azimuths: 130ø , 135ø , errors due to slip on preexisting planes of various trends 140ø , 142ø , 142ø , 143ø , 146ø , 146ø , 150ø). These consistent should hopefully tend to cancel, and the correct average P orientationsare independentof lithologicand age boundaries and T directions will be obtained. M. L. Zoback and M.D. as well as dip of the strata. Bell and Gough[1979] have sug- Zoback [ 1980] comparedextension directions on major faults gestedthat the breakoutsresponsible for theseelongations are in the northernBasin and Range, determinedusing fault slip causedby concentrationof stressesat the walls of wells with data, with the T axesof focal mechanisms.Despite large scat- the direction of elongation coinciding with the least hori- ter in the focal mechanismdata (standarddeviation, _+25ø), zontal principal stressorientation. Utilizing experimentally the mean directions and the geologicallydetermined exten- derived stressmagnitudes required to produce spalling, they sion directionsfall within 3 ø of one another;this result sug- suggestthat at least one of the horizontalstresses must exceed geststhat the T axesare fairly reliable indicatorsof principal the lithostat indicating a compressional(strike slip or thrust stress/straindirections. Both R. B. Smith [1977] and Eaton faulting) stressregime. [1979] have also noted that regionally good correlation be- Finally,some discussion is waiTanted regarding the age of tween earthquakeP and T axesand nearby in situ stressdata the geologicstress field indicatorsutilized in our study. In the throughout the United States. However, as outlined in the westernUnited States,only featuresyounger than 5 m.y. and previousparagraph, stressdirections derived from individual generally younger than 3 m.y. were included. (The ages of focal mechanisms could be in error as much as 35o-40 ø. We specificsites are listedunder 'comments'on Table 1.) Consid- have attemptedto rely, wheneverpossible, on averageP and erable stressfield uniformity is found over this time period in T directionsfrom a number of solutionsin any given area. severalareas. In the easternUnited States,two groupsof data The best seismiccoverage is in California, where detailed were used:the data in New England representpostglacial and studieshave been made of focal mechanismsalong segments Holocenefaulting; the tectonismin the Atlantic Coastal Plain, of the San Andreas fault. From these studies the mean P and 6116 ZOBACK AND ZOBACK.' STATE OF STRESSIN CONTERMINOUS UNITED STATES

Z Z Z Z Z Z Z Z Z •. Z • Z Z Z • • • • •- •

o o o o o o o o o o

Z Z • Z Z • Z Z Z Z Z Z Z Z Z Z Z • Z ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUSUNITED STATES 6117

Z Z Z • • • Z Z Z Z Z • Z Z Z Z Z Z Z Z Z • Z Z 6118 ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUS UNITED STATES

o o • Z Z Z Z Z Z Z Z Z Z Z • Z Z ZZ Z

4: 4: 4: < 4: :• •: z z z z o 6 o o o ZOBACKAND ZOBACK: STATE OF STRESSIN CONTERMINOUSUNITED STATES 6119

•- Z •- Z Z Z Z Z Z •- Z

Z Z Z Z Z Z • Z Z ZZ ZZ Z Z Z Z Z Z Z • Z 6120 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Z • • •

o ml r-- • o Z Z Z Z Z Z Z Z Z Z Z ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUS UNITED STATES 6121

r.J.,r.J., d r.J.,r.J., r.J., r.J., r.J., r.J., a 6 • • a a r.J.,r.J., 6 r.J.,r.J., 6 6 a 6 r.J.,6• 6 z z z z z • z z • z z z z z z z z z z z z z z z z z z o • o o o • • • • • • • • • • • 6122 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

• Z Z • Z Z Z Z Z Z Z Z Z Z r.,o r.,o Z ZZZ Z Z

I I I I I I

Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z , ZOBACKAND ZOBACK.'STATE OF STRESSIN CONTERMINOUSUNITED STATES 6123

o o o o o o o o - • • • r-- r-- r-- - ,• • - - • Z Z Z • • Z Z Z Z Z Z Z Z Z Z Z Z Z • • • Z • Z

z z z z z z z z z z z z z z z z z o o o 6124 ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUSUNITED STATES

0 • •

o o

Z Z Z Z Z Z Z Z Z Z Z Z Z Z • Z Z Z ZOBACK AND ZOBACK.'STATE OF STRESSIN CONTERMINOUSUNITED STATES 6125

Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z ZZ 6126 ZOBACKAND ZOBACK:STATE OF STRESSIN CONTERMINOUSUNITED STATES

o o • • o • • • o o o •

Z Z Z • • • Z Z • • • Z Z • Z Z Z Z ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6127

[- • [- • Z 6128 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

T axeswere consideredfor the presentcompilation. Elsewhere typically cannot be used more than several tens of meters in the United States,coverage is largely from single event so- from free surfacesand that they seemto be extremely sensi- lutions and, lessfrequently, from microearthquakecomposite tive to local inhomogeneitiesin the rock. McGarr and Gay solutions. In the central and eastern United States, most of the [1978] gave an excellent review of these techniquesand dis- focal mechanisms used [from Herrmann, 1979] are con- cussedmany of the inherent experimentaldifficulties. strained by both body wave and surfacewave solutions. Stress measurements made within several tens of meters of As discussedearlier, we have assumedthat two of the prin- the surface have not been included in this study. Such mea- cipal stressdirections are horizontal and that the third is verti- surementsare particularly susceptibleto the effectsof weath- cal. Thus only horizontal stressaxes, as inferred from the fault ering, erosion, and deglaciation.Also, nearby surface frac- plane solutions,that have plungesof lessthan 20ø were con- tures and joints can apparently act to decouplesurface rocks sidered. A few rare exceptionsof T axes with plungesup to from the tectonic stressfield [Haimson, 1978a; M.D. Zoback 30ø were included where the consistencyof the stressorienta- and J. C. Roller, 1979]. tion with surroundingdata justified their inclusion.Informa- STRESS MAP OF THE CONTERMINOUS UNITED STATES tion on the plungesof P and T axes is listed in Table 1. Compilations of fault plane solutions for the western Utili?.ing the three types of principal stressindicators dis- United Statesby R. B. Smith and A. G. Lindh [1978], in the cussed above, we have assembledtwo maps of horizontal Rio Grande rift by Sanford et al. [1979] and for the eastern stress orientations for the conterminous United States. The United Statesby Sbar and Sykes [1977] and Herrmann [1979] data are numbered by states (Table 1) and presentedon a were the primary data sourcesin the presentstudy. However, map of physiographicprovinces of the United Statesin Plates we attempted, whenever possible,to return to the original ref- 1 and 2. Figure 1 names the physiographicprovinces of the erences and reject poorly constrained mechanisms. Further United States [after Fenneman, 1946] for reference. In Plate 1 details of particular focal mechanismsare discussedin the sec- we showthe least principal horizontal stressdirection and the tions on individual stressprovinces. type of stressindicator used in the determination;sites where more than one type of stressindicator was used are designated In Situ Stress Measurements by multiple symbols. A descriptionof each site is listed in The third stressindicator used in this study was the direct Table 1. We choseto presentonly the least principal horizon- determinationsof both the orientation and magnitude of tec- tal stressin Plate 1 becauseof the consistencyof this principal tonic stressat depth. We primarily consideredmeasurements axis throughout a broad region of the western United States made by the hydraulicfracturing ('hydrofrac') technique, cur- that includes the Basin and Range province, the Sierra Ne- rently the only method of measuringstress at large distances vada, and the area along the San Andreasfault. In regionsof from free surfaces.The technique consistsof hydraulically active thrust or reversefaulting, the least principal horizontal isolating a sectionof a well or borehole (by means of inflat- stressshown is actually the intermediate principal stress able rubber packers) and pressurizingthe isolated sectionun- Plate 2 presentsthe stressdata in a manner to depict the rela- til a tensile fracture is induced at the well bore. If the borehole tive magnitude of the horizontal and vertical stresseswith the parallels one principal stress(usually the lithostat), a vertical stressprovinces we have defined to indicate regions in which fracture will form at the azimuth of the greatest horizontal the relative magnitude and orientation of the stressfield is principal stress,which generally can be determinedto better fairly uniform. than +_10ø. However, repeated measurementsat different As shown in Plate 2, the United States can be divided into depth intervalswithin a given boreholegenerally have a range stressprovinces which correspond,in a general way, with the of + 150, probably a more accurate estimate of the reliability physiographicprovinces. In the following sectionsthe individ- of the method. ual stressprovinces are discussedin detail. First, a few impor- From the pressure-timehistory of hydraulic fracture forma- tant general observationsthat can be made from Plates 1 and tion and extension,the magnitudesof both horizontal princi- 2: pal stressescan be computed. Assessingthe accuracy of the 1. In regionswhere more than one type of stressindicator stressmagnitudes determined from any given measurement was available a good correspondenceexists between the re- can be quite difficult and requiresdetailed knowledgeof the sults of the different methods, generally well within the esti- pressure-timedata and well conditions.Typically, the least mated accuracyof individual determinations. horizontal principal stresscan be determinedwith greater ac- 2. The intraplate stressfield for that part of the North curacy than the greatestbecause determination of the greatest American plate represented by the conterminous United principal horizontal stressrequires an assumptionof linear States is not uniform. Variations in the relative magnitude elasticity around the well bore. The hydrofrac technique was and orientation of the principal stresseshave wavelengths describedin detail by Haimson and Fairhurst [1970] and inter- which range from 100 to 3000 km. The westernUnited States, pretation of hydrofrac data was discussedat length by M.D. a site of active tectonismcharacterized by a high level of seis- Zoback et al. [1977, 1980b]and McGarr and Gay [1978]. Com- micity and generally high heat flow, is marked by a complex pilations of hydrofrac stress measurements by Haimson pattern of stress.Even in the easternUnited States,despite its [1977b] and McGarr and Gay [1978] were used extensivelyin relative tectonic quiescence,major variations in principal this study. stressorientations appear to exist. Several other widely used stressmeasurement methods can 3. Broad regionsof the crust (with linear dimensionsas be broadly characterized as stressrelief techniques. These large as 3000 km in the midcontinentarea) can be charac- methods are all basically passiveand involve measurementof terizedby a relativelyuniform stressfield (within +10ø-20ø). the strain or displacementthat occurswhen the ambient stress 4. Large changesin orientationof the principal stressfield field is relieved by techniquessuch as overcoring.The most (up to 90ø) can occurlaterally over shortdistances (

i 130' ,•o' , ;s' , ,04 ,os' ,•o' ss' so' ss' so' 70'

CANADIAN SHIELD/SUPERIOR PROVINCE MOUNTAINS

NORTHERN ROCKY

KLAi•ATH MIDDLE% MTNS

ROCKY

COLORADO PLATEAU

.. 300 SCALE OULF COASTAL PLAIN o 100 2•0 300MI

100, 200 , 300ß KM

\ •'.\.

\. 120' 116' 106' 1004 S 6' SO' 85' 80' 76' I

Fig. 1. Index map with namesof physiographicprovinces shown on Plates I and 2.

principal stressfield or a rapid changein the relative magnitudesof the principal stresscan be representedin the follow- tude of principal stresses. ing manner:

STRESS PROVINCES: WESTERN UNITED STATES S ! > S2 > S 3 The westernUnited Statesrepresents a broad area of active or tectonism,as evidencedby the high level of seismicity(Figure 2) and the generallyhigh heat flow (Figure 6). Plate 2 in- SN-S> SE-w > Sv dicatesour interpretationsof the various stressprovinces in that region. Where the data coverageis somewhatsparse, we Crosson[1972] has suggestedthat the occurrenceof both havealso relied on the seismicitymap (Figure 2) aswell as on strike slip and thrust earthquakesmay result from a more Howard et al.'s [1978] map of young faults in the United rapid increasewith depth of the verticalstress relative to the States(along with their accompanyingdiscussion) to delineate horizontal stresses.Thus thrust faulting (Sv, least principal the stressprovinces. As might be expected,in areasof good stress)would be favoredat shallowdepths, whereas strike slip data coveragea good generalcorrespondence is apparent be- faulting (Sv, intermediateprincipal stress) would be expected tweenthe provincesdefined by Howardet al., on the basisof at greaterdepths. patternsand stylesof Cenozoicfaulting, and the provincesde- Two distinctareas of relatively active tectonismwithin the fined by stressmagnitudes and orientations. Pacific Northwest provincedemonstrate contrasting styles of deformationresulting from the N-S compression.The Puget PacificNorthwest Sound-OlympicPeninsula area is the siteof severallarge his- The PacificNorthwest stress province includes the only ac- torical earthquakes (including both upper crustal and tive andesitic volcanic chain within the conterminous United mantles60- to 70-kin depthsvents) and also of numerous States (the Cascades);it also includesthe Pacific coastal Quaternary scarpsshowing strike slip and reversefaulting. rangesin Washingtonand Oregonand the 'back arc' region, Farther east in the central Columbia Plateau, a largely aseis- the Columbia Plateau. The following discussiondraws heav- mic region,folding along east-westtrending axes that began ily from summariesof the regionalstresses and tectonicset- in late Tertiarytime havecontinued throughout the Quater- ting of the PacificNorthwest by Crosson[1972], Davis [1977], nary [cf. Davis, 1977, for references].Inclusion of the Cas- and R. B. Smith [ 1977, 1978]. cades andesitic volcanic chain in the Pacific Northwest com- The least principal horizontal stressdirection throughout pressionalstress province is consistentwith observationsof a this provinceaverages about east-west;the consistentstress stateof compressionwithin island arcsbehind which there is axis,however, appears to be alined with the greatestprincipal no active extension[Nakamura et al., 1978].This compression stressS!, a horizontalN-S compression.Focal mechanismsin- is generallyparallel to the directionof convergencewhich, as dieate both strike slip and thrust faulting; a predominanceof discussedbelow, is not the case in the Pacific Northwest. thrust mechanismssuggests that the vertical stressS, may, in The predominanceof relativeN-S compressionwithin the general,be the least principalstress S3. The relative magni- PacificNorthwest distinguishesthis stressprovince from the 6130 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

5O"

•DENVER

/-os

EPICENTER MAP OF WESTERN UNITED STATES

Fig. 2. Seismicitymap of westernUnited States[from R. B. Smith, 1978].Data set primarily includesinstrumentally locatedearthquakes from 1950 to 1972,but someearlier earthquakesare also incorporated.Events are not differentiated by magnitude.In California, minimum magnitudeearthquakes plotted are M ~ 1, and for rest of the western United States,M-• 3. Outlines of physiographicprovinces have been added. Dashed line in the southernpart of northern Basin and Range marks approximateeastward extent of pure strike slip fault plane solutionsin that region.

Basin and Range provinceto the south,which is characterized San Andreas by WNW-ESE extensionaltectonics. An approximatelyeast- west zone at about 44.5øN latitude (roughly the northernmost The most seismicallyactive region within the United States extent of basin-range-typenormal faulting) separatesthese is the California coast area, which includes the San Andreas two provinces.R. B. Smith [1978] identified this east-westzone fault system.The San Andreas and subsidiarysubparallel as a relatively aseismicintraplate boundarythat may extend faultsmark a major fight-lateraltransform plate boundarybe- eastwardeither to the edgeof the aseismicSnake River plain tween the Pacific and North American plates. or to a zone of earthquakestrending northwestthrough the Leastprincipal horizontal stresses, which are relativelyuni- Idaho batholith (Figure 2). Blackwell[ 1978]has shownthat an form throughoutthe province,trend E-W to WNW-ESE, but east-westtrending thermal energy boundary coincideswith the regionalpattern of stresscan be complicatedlocally by an this stressboundary. often complex geometry of interacting faults [e.g., Pavoni, ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6131

1979; Ellsworth and Marks, 1980]. The generally strike slip late Miocene time the Mojave block was deformedprimarily styleof deformationindicates that this horizontal stressdirec- by normal faulting along NW trending faults [Dibblee, 1967]. tion doesin fact correspondto S3. The relative magnitudesof Current tectonismin this block, as evidencedby recent seis- the stresses are thus micity, faulting, geodeticdata, and generallysubdued topography, indicates primarily fight-lateral strike slip motion on S• > S2 > S3 theseNW trending faults and justifiesits inclusionas part of or the San Andreas province. S-N-s > Sv > S-E-W The Mojave block is separatedfrom the Sierra Nevada and southwesternGreat Basin to the north by the ENE trending Exceptionsto this general style of strike slip deformation Garlock fault. The Garlock fault has been described as a ma- occur throughoutthe province on faults striking at large an- jor left-lateral fault conjugateto the San Andreassystem [M. glesto the NW-SE San Andreastrend. Motion on thesefaults L. Hill and T. W. Dibblee, 1953]and also as a major continen- appearsconsistent with the regionalstress field, i.e., thrustor tal transform accommodating extension in the Basin and reverse faulting on approximately east-westtrending faults Range to the north in relation to the Mojave block; increasing and normal displacements on approximately north-south fault offsetson the Garlock to the west supportthe transform faults. The consistencyof the P and T axes for slip on these fault interpretation [Davis and Burchfiel, 1973]. Furthermore, faults with widely varying trendsreinforces the use of average westward shifting of the block to the north of the Garlock P and T directionsas reliable indicatorsof the principal stress fault due to this extensionhas probably contributed to the directions. westward bending or deflection of the San Andreas fault The most obvious example of a change in deformational where the two faults meet [Davis and Burchfiel,1973]. Limited style along the San Andreas occurswhere the trend of the information on extension directions on normal faults north of fault markedlychanges in the Big Bend area of southernCali- the Mojave block, however, indicate that the horizontal dis- fornia. The post-Miocene to present development of the placementmay not parallel the Garlock fault as would be re- TransverseRanges is dramatic evidenceof crustal shortening quired for true transform-stylefaulting. in that area [Jahns, 1973]. There, the fault system trends nearly east-west,and the primary mode of deformation is Sierra Nevada folding and thrust or reversefaulting. This large-scalethrust The Sierra Nevada marks a transition from the primarily and reversefaulting requiresa vertical leastprincipal stressas strike slip deformation along the San Andreas fault to the ex- opposedto the regionalhorizontal orientation of S3.This ap- tensionaltectonics of the Basin and Range province. Accord- parentexchange (or rotation) of the principalstresses can eas- ingly, severallines of evidenceon active tectonicsof the Sierra ily be explainedif in that area the leastprincipal horizontal indicate both strike slip and normal faulting. Examples of stressand the vertical stressare approximatelyequal in mag- pure strike slip deformationinclude the Kern Canyon fault, a nitude. major (--•80 km long) N-S trending fight-lateral fault active Compressionalfeatures outside the Big Bend area include throughout late Tertiary time [Moore and duBray, 1978] and Neogeneanticlines subparallel to the San Andreasfault [e.g., the microfaulting discussedbelow. However, the extensional Harding, 1976; Wilcox et al., 1973; Blake et al., 1978] and tectonicsof the Basin and Range province,including normal thrust type focal plane mechanismsin the Coast Ranges(see faulting and high heat flow, extends50-100 km into the east- CA-9 and CA-10, Table 1). These featuresprovide evidence ern Sierra;Lake Tahoe, for example,is a down-droppedbasin that the San Andreas province is not in pure shear,as a com- like thoseto the eastin Nevada. Also, major earthquakeswith ponentof compressionis actingnormal to the fault. both strike slip (1966 Truckee earthquake,CA-31) and nor- Crustal shorteningin the Big Bend area can be thought of mal (1975 Oroville earthquake,CA-32) faulting mechanisms as resultingfrom the large left-steppingshift in the San An- have been recorded in the northern Sierra. dreas fault. Similarly, near en echelonoffsets between active Least principal horizontal stressdirections trend WNW- strandsof the fault, a changein deformationalstyle can be ex- ESE to E-W in the northern Sierra but appeargradually to ro- pected. The geometry of these offsetsrequires local crustal tate until they are trending approximately NW-SE in the shorteningin the caseof left-steppingoffsets and local crustal southernSierra. Stressorientations from different types of in- extensionin the caseof right-steppingoffsets. Hill [1977] has dicatorsagree fairly well within the Sierra Nevada block; fur- presenteda model to explainthe styleof deformationand ex- thermore, least horizontal stressorientations agree well with tension within right-steppingoffsets. This model was tested those in the surroundingSan Andreas and Basin and Range with data from two localities along the San Andreas fault provinces. [Weaver and Hill, 1979]. Active extensionin one of these lo- The existenceof contemporaneousstrike slip and normal calities(Salton Trough) is stronglysuggested by normal fault faulting in the Sierra Nevada suggeststhat the magnitudesof scarps[Sharp, 1976],high heat flow, and a few normal fault the vertical and greatest principal horizontal stressesare focal mechanisms within the zone of offset. Weaver and Hill nearly equal and can readfly exchange.The principal stress also reported that the T axes for fault plane solutionsfrom field probably has the form earthquakesalong the offsetfault strandsand within the offset volume are nearly invariant. This result suggeststhat the di- S• = S2 >> S3 rection of S3 is not significantlyperturbed from its regional or orientation in the vicinity of the offsets.The occurrenceof SN•E = Sv >> Sw•w normal faulting, however,suggests that locally, $1 is vertical rather than horizontal. Thus local variations in the relative magnitudesof SSSEand Also included in the San Andreas subprovinceis the left- Sv and/or local faulting trends may determine whether strike lateral, Garlock fault and the Mojave block. During middle to slip or normal faulting occurs. 6132 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Many of the least principal stressdirections in the Sierra Seismicityextends throughout the province (see Figure 2); Nevada block (CA 19-22, 24-25, and 28-30, Table 1) come however, it is most concentrated along the margins of the from a unique studyby Lockwoodand Moore [1979] of offsets northern Basin and Range, in the N-S trending Nevada seis- on near-vertical strike slip microfaults. These microfaults are mic zone on the west and the Intermountain Seismic Belt on oriented in two conjugatesets, a north to northeaststriking set the east [Sbar et al., 1972]. A roughly north trending belt of that sho•s fight-lateraloffset, and an eastto northeaststriking historicalearthquakes of magnitude >7 occurswithin the Ne- set that showsleft-lateral offset.The age of this microfaulting vada SeismicBelt [Ryall et al., 1966]. However, as $1emmons is poorly established;the faults formed after solidificationof [1967] pointed out, this pattern of modem seismicityis not a the youngestplutons in the Sierra Nevada (79 m.y.B.P.) and good indicator of tectonic activity in the recent past because are known to cut a late Miocene volcanic dike in one area. A Quaternary scarps are widespread throughout the interior maximum horizontal extensional-straindirection (assumedto northern Basin and Range. be alined regionally with the least principal stress)was taken Least principal stressorientations are generallyWNW-ESE as the bisector of the obtuse angle of the average microfault throughout the Basin and Range-Rio Grande rift province. trends.A pure shear constantvolume analysisbased on a de- Exceptionsto this general pattern occur in southernNevada, tailed study of microfault offsetsin one locality indicatesthat where the data seemto indicate a more NW-SE least principal the true maximum extensional-straindirection differs by only stressdirection, and along the Wasatch fault in Utah, where 6 ø from the bisector. modem crustalextension appears to be more in an E-W direc- For the bisector analysis used by Lockwood and Moore tion. Possibleexplanations of theseregional variations are dis- [1979] to be valid, the fault setsmust be true conjugates,and cussed later in the source of stress section. the strain must be small (<10%) and distributedon both sets. The primary mode of deformation within the Basin and Absenceof offseton somemicrofaults and filling by vein min- Range-Rio Grande riff is normal faulting with a consistent eralssuch as quartz and epidotesuggest a tensileorigin for the approximatelyWNW direction of opening,indicating a stress microfaults, many of which were reactivatedin shear after a field of the form change in regional stressorientation. In addition, both the S• > S2 >> S3 aerial photo lineation analysisand the microfault orientations reportedby Lockwoodand Moore [1979]indicate that the east or to northeast striking (left-lateral) set predominatesand sug- Sv •, SNNE >> SWNw gestthat the strain may be very unevenly divided between the two sets. Thus the trends of the microfaults, while indicative A detailed discussionof the relative magnitudesof theseprin- of the regional stressfield, may not be very preciselyoriented cipal stresses,constrained by an analysisof slip directionsand with respectto that stressfield (perhaps only within +25ø). pattern of faulting in northern Nevada, has recentlybeen pre- However, parallelismbetween the least principal stressorien- sented by M. L. Zoback and M.D. Zoback [1980]. tations inferred from these microfaults and those in the sur- A consistentdirection of extensionthroughout much of the rounding San Andreas and Basin and Range provinces sug- Basin and Range-Rio Grande rift province may be inferred gests that these faults are young features recording from the relatively uniform pattern of horizontal dis- deformation in a stress field similar to the modem one. placements,despite the complex geometryof the fault blocks. A necessaryresult of this uniform extensionis oblique slip Basinand Range-Rio GrandeRiff normal faulting (i.e., both strike slip and dip slip components The Basin and Range province is a region of active crustal of motion) on faults that are not perpendicularto the least spreadingcharacterized by high regional elevation, thin crust principal stressdirection. Examples of this obliqueslip styleof (average thickness-30 km), and high heat flow [Thompson faulting are abundant throughout the province. As has been and Burke, 1974]. The Basin and Range-Rio Grande rift noted by many workers [e.g., Slemmons,1967; Wright, 1976], stressprovince discussed here extendsfrom the Sierra Nevada the geometryof the pattern of displacementsrequires purely eastward to well within the Colorado Plateau physiographic dip slip motion on approximately NNE trending faults, a province and around the south margin of the Colorado componentof right-lateral slip on faults trending N to NW, Plateau and northward into the Rio Grande rift proper in and a componet of left-lateral slip on faults trending NE to New Mexico. The southern Basin and Range has been in- ENE. Thus modem components of fight-lateral strike slip cluded with the currently more tectonically active northern shouldbe expectedon the generallyNW trendingfaults in the Basin and Range on the basis of generally consistentleast southern Basin and Range (in contrast to nearly purely dip principal stressorientations and regionallyhigh heat flow, the slip motion expectedon the NNE striking faults that charac- mean value of which is similar to the northern Basin and terize much of the northern Basin and Range). Scantydata on Rangemean of 2.1 heat flow units(HFU) •cal/cm2/s = 41.9 fault slip in the southern Basin and Range and Rio Grande mW/m 2) [Lachenbruchand Sass,1977, U.S. GeologicalSur- rift (AZ-7; MX-I; NM-6, 16, 27; Table 1) confirm this ex- vey, unpublished data, 1979]. Thus, although a large differ- pectation. ence in strain rate between the two regions has persisted It is important to point out that theseoblique slip faults are throughout Quaternary time and possiblylonger (on the basis basicallynormal, not strike slip, faults. The oblique slip arises of young fault scarpsmapped by Howard et al. [1978]), the re- as the major crustal blocks attempt to extend in roughly the gional stressfield appearsto be uniform. The Rio Grande rift least principal stressdirection regardlessof the trend of the is often referred to as a distinct continentalrift [e.g., Bridwell, fault. 1976];however, similaritiesin geophysicalcharacteristics and Although most of the Basin and Range provinceis charac- tectonic style [cf. Cordell, 1978; Thompsonand Zoback, 1979] terized by normal faulting, the stresstransition (from strike warrant its inclusion in the same stressprovince as the Basin slip deformation along the San Andreas fault) evident in the and Range. Sierra Nevada extends into the westernmost northern Basin ZOBACKAND ZOBACK: STATE OF STRESSIN CONTERMINOUSUNITED STATES 6133

and Range, the Walker Lane-Las Vegas shear zone region, The Basin and Range-Rio Grande rift extensional stress where examplesof normal, pure strike slip, and oblique slip field appearsto continue 100-200 km into the plateau proper faults can be found [cf. Slemmonset al. [1979] for a detailed (cf. Plate 2), consistent,as mentionedabove, with high heat discussionof the rate and style of deformation in this region]. flow, normal faulting, and recentvolcanism along the plateau Earthquake focal mechanismsin this region indicate consis- margins [Thompsonand Zoback, 1979]. The 90ø change in tent T axes regardlessof the style of faulting [e.g., Hamilton stressorientation between the Basin and Range Colorado and Healy, 1969]. This zone of stresstransition, as inferred Plateau interior provinces occurs locally over a lateral dis- from the distributionof purely strikeslip fault plane solutions, tance of <50 km (cf. UT-8, 9, 10, 11, 12, 13 and NM-12, 13, extends roughly 200-250 km eastward into the Basin and 14, 15). Data in central Arizona along with a reverse fault Range (seeFigure 2). The relative magnitudesof the principal focal mechanism at 35.16øN, 112.15øW ([Brumbaugh, 1980] stresseswithin this region, as inferred from the coexistenceof not included on the maps becauseof the late data at which we normal and strike slip faulting, are becameaware of it) suggesta broaderstress transition, up to 150 km wide, in which all three stressesare approximately S1'• S2>> S3 equal in magnitude. This is suggestedby the occurrenceof Sv = S•E >> SWNW both WNW and NNE trending dikes and cinder cone alinements (AZ-7, 8, 10), as well as focal mechanismsfor two Thus, althoughthe relative magnitudesof Sv and SNNE can in- events70 km apart which imply both normal (AZ-14) and re- terchangelocally, the S3direction (WNW) remainsinvariant. verse[Brumbaugh, 1980] faulting on NW trendingplanes. The The available data indicate a rather abrupt transition in observed contrastingstyles of deformation in this 150 km stressorientation between the Rio Grande rift province and wide zone can be explainedby minor local variationsin mag- the southernGreat Plains, occurringover <75 km in one area nitude of one of the principal stressesand, also, possiblyby (compare NM-26 and NM-27). The Basin and Range-Rio the influence of local, preexistingstructural grain. Grande rift stressfield continueswell into the physiographic Colorado Plateauconsistent with high heat flow, faulting, and Snake River Plain-Yellowstone recent volcanismalong the plateau margins [Thompsonand Zoback, 1979]. The nature of the stresstransition between The Snake River plain is a downwarpedvolcanic plain thesetwo provincesis discussedin detail below. formedduring the past 17 m.y. by a steadynortheastward mi- gration of silicic volcanismand subsequentsemicontinuous Colorado Plateau Interior outpouringsof basaltalong its older (western)reaches [Arm- stronget al., 1975].The site of modern silici½volcanism on the The Colorado Plateau stressprovince is distinctly smaller plain is the Yellowstonecaldera. The SnakeRiver plain-Yel- than the Colorado Plateau physiographicprovince (see Plate lowstonetrend has been consideredby someworkers to be a 2). The stateof stresswithin the plateauinterior, discussed in propagatinglithospheric crack (possiblyalined with a pre- detail by Thompsonand Zoback [1979],is briefly summarized existingzone of weakness)[e.g., R. B. Smith et al., 1974b;Ea- below. ton et al., 1975]and by othersto mark an activehot spottrace The Colorado Plateau interior is characterizedby roughly [e.g., Morgan, 1972; Suppeet al., 1975]. An inversionof rela- NNE-SSW least principal horizontal stress direction, per- tive plate motions to a mantle referenceframe indicatesthat pendicularto the generalWNW-ESE leastprincipal stress di- the absolutemotion of the North Americanplate for the last rection in the surroundingBasin and Range-Rio Grande rift 10 m.y. is well representedby the SnakeRiver plain-Yellow- province.The data alsoappear to supporta radial distribution stonetrend [Minsterand Jordan, 1978]. of least principal horizontal stressorientations perpendicular The Snake River plain is largely aseismicwith the ex- to the plateau margins.The preferredinterpretation of a uni- ceptionof someshallow (<6 kin) earthquakesat Yellowstone. form NNE least principal horizontal stresswithin the Colo- The only data availableon principalstress orientations on the rado Plateau interior comes from data along the northern plain (with the exceptionof a compositefocal mechanism edge of the plateau acrosswhich there are no known major within the caldera) are geologicand come from numerous geologic,topographic, or geophysicalcontrasts. young (<1 m.y.B.P.) feedervents which cut acrossthe plain. The absenceof major faulting or seismicitywithin the As Weaver et al. [1979] noted, these indicators suggesta plateauinterior couldbe interpretedto indicategenerally low roughlyNE-SW leastprincipal stress direction, parallel to the differential stresses.Apparently consistent with this inter- axisof the SnakeRiver plain, not perpendicular,as suggested pretation,in situ stressmeasurements (at •0.5 km depth) in by Hamilton and Myers [1966].This extensionalong the axis the PiceanceBasin (CO-3) in northwesternColorado indicate of the plain requirestransform style of faultingalong the mar- that all three principalstresses are approximatelyequal to the gins of the plain, particularly to the north. lithostat [Bredehoeft et al., 1976]; alternatively, these low The enormous volume of volcanism on the Snake River deviatoric stressesmay simply be due to the shallow depth plain and related normal faulting indicate extensionaltecton- and rather weak rock within the basin. The occurrence of ics.Thus the stressfield on the SnakeRiver plain can be char- both strike slip and thrust focal mechanismsclearly demon- acterized by strates the existenceof high horizontal stresseslocally ex- ceeding the lithostat and indicates a compressionalstress re- S1>S2>S3 gime of the form or

S v > S--NW> S_NE or The direction of the least principal stresson the Snake River SWNW> SNNE •' Sv plain appearsoblique to that in the Basinand Range province 6134 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES directly to the south,where the least principal stressdirection ary of the province (MT-4, 5) are also somewhat con- trends WNW-ESE to E-W. This distinction in trend and its tradictory. The northern boundary remains undefined by the unique Late Cenozoicvolcanic history are the justificationfor presentdata. making the Snake River plain a separatestress province. In general, the stressfield for the Northern Rocky Moun- The area immediately surroundingthe Yellowstonecaldera tain stressprovince appearsto be characterizedby NE-SW to is marked by a high level of seismicity(see Figure 2). Detailed E-W least principal stressorientations, consistentwith the focal mechanism studiesin the area north of the caldera by prominent N to NW trending normal faulting. The occur- Trimble and Smith [1975],R. B. Smith et al. [1977],and Pitt et rence of both strike slip and normal focal mechanismswith lo- al. [1979] further documentan approximately90 ø rotation of cally consistentT axesled Freidlineet al. [1976] to suggestthat the least principal horizontal stressdirection between the the greatest and intermediate principal stresses(Sv and Snake River plain and the area to the north first noted by SNoStoNW_SE) may be approximately equal in magnitude. The Freidline et al. [1976]. This rotation occursover a lateral dis- generally E-W to NE-SW extensionin the Northern Rocky tance of only 100 kin. Mountains appears to representa northward continuation of basin-range structure interrupted by the Snake River plain Hebgen Lake-Centennial Valley and with possiblya slight clockwiserotation of the extension As evidencedby focal mechanismsand recent faulting, the direction. Hebgen Lake-Centennial Valley stressprovince is characterized by N-S extension.The south boundary of this province Southern Great Plains near Yellowstone, as discussedabove, is marked by an abrupt 90ø rotation of the least principal stressdirection that occurs The southern Great Plains area is characterizedby a very over a lateral distance of <100 kin. The northward extent of uniform state of stressin which the least principal horizontal this province is undefined, and available focal mechanism stressdirection to NNE-SSW. The data come from two pri- data, discussedin detail by Freidline et al. [1976], suggesta mary sources:(1) alinementof post-5m.y. volcanicfeeders in broad zone of transition in which T axes again rotate 90 ø the Raton-Clayton volcanic field of northern New Mexico (from N-S to E-W) from southernto northern Montana. We and (2) fractureorientations obtained from hydraulicallyfrac- have drawn a tentative northern boundary for the province on tured wells in the Permian basin of western Texas [Zemanek the basisof changesin the trends of faulting (from approxi- et al., 1970].An earthquakefocal mechanismand in situ stress mately E-W to NW-SE) to the north. The provinceis alsoten- measurements in eastern Colorado show consistent orienta- tatively extended to the west to include a composite focal tions. Despite the similar least principal horizontal stressori- mechanismin the Idaho batholith (ID-5) that indicatesE-W entationsin the southernGreat Plains and in the plateau inte- trending normal faulting. rior, the probable extensional regime (as inferred from earthquakefocal mechanismsand basalticvolcanism) distin- Northern Rocky Mountains guishesthe southernGreat Plainsfrom the ColoradoPlateau The Northern Rocky Mountains stressprovince is charac- interior and Midcontinent provinceswhich are dominated by terized by normal and strike slip faulting along generallyNW- compressionalstress regimes. The least principal horizontal SE trends that probably began in early Miocene time and stress directions to the southern Great Plains province are continuesto the present, as evidencedby Quaternary scarps generallyoblique to thosein the Midcontinentstress province; [Freidline et al., 1976; Howard et al., 1978; Reynolds, 1979]. however, in west-central Texas the data suggesta counter- Above average heat flow [Lachenbruchand Sass, 1977], the clockwise rotation of the stress field from west to east to direc- occurrenceof hot springs[e.g., Muffer, 1979, Map 1], high re- tions more consonant with midcontinent stress orientations gional elevation [e.g., Reynolds,1979], and moderately high (plates 1 and 2). The easternmostsite of this group of data seismicity[e.g., R. B. Smith and M. L. Sbar, 1974] are all sug- (TX-15) is a normal fault focal mechanism,suggesting that gestiveof active extensionaltectonism in this province and the boundary with the compression-dominatedMidcontinent distinguishesit from the Pacific Northwest province to the stressprovince must lie further to the east. west, which appears to be dominated by N-S compressional Two important characteristicsof the stressfield in the tectonics. southernGreat Plains provinceare (1) its uniformity (•+15 ø, The exact boundaries of the Northern Rocky Mountains which is within the range of estimatedaccuracy of the differ- and, in fact, even the existenceof such a province are difficult ent stressindicators) over a region with a north-south extent to define from the available stress data. As discussedabove, of at least 1100 km and (2) an abrupt transitionin stressorien- the small, poorly defined Hebgen Lake-Centennial Valley tation in relation to the Rio Grande rift that occurs locally stressprovince boundsthe Northern Rocky Mountains prov- over a lateral distanceof <50 kin. This •90 ø changein least ince to the south. The positions of the eastern and western principal horizontal stressorientation along the Rio Grande boundariesare based on the extent of young NW trending rift-Great Plains boundary correspondsto a decreasein heat normal faulting shownon Howard et al.'s [1978] map and are flow and a crustaland lithosphericthickening under the Great consistent with boundaries of extensional tectonism deline- Plains in relation to the Rio Grande rift [Thompsonand Zo- ated by Reynolds[1979]. Three in situ stressmeasurements in back, 1979].In addition, seismicreflection profiling acrossthe northern Idaho along the western boundary are completely Rio Grande rift's easternboundary [Brownet al., 1979]has re- contradictoryboth in orientationsand relative magnitudesof vealed a high-angle,narrow, linear zone definedby a lack of principal stress;two are independent strain relief measure- coherent reflections which extends from near the surface to ments in mines 3 km apart (ID-7) which are inconsistentwith the base of the crust. Physically, this sharp boundary may one another, and both disagreewith a nearby hydrofrac mea- mark a zone of lateral 'decoupling'within the crustfacilitating surement(ID-6). Focal mechanismsnear the easternbound- the abrupt changein stressorientation. ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6135

STRESS PROVINCES: South Carolina, earthquakeof 1886 and the 1811-1812 New CENTRAL AND EASTERN UNITED STATES Madrid, Missouri, earthquakesdominate the seismichistory of the eastern United States, but there are numerous other Despite suggestionsthat the in situ stressfield is uniform areas of concentratedseismic activity. Several of the larger- throughout the eastern United States [e.g., Hairnson, 1977b], scalepatterns and trendsin easternseismicity have been sum- recent data demonstratethe existenceof distinct stressprov- marized in detail by Sykes[1978]. These include the northwest inces(Plate 2). As in the West, earthquakefocal mechanism, trendingseismic zone in South Carolina and Georgia [Bolli- geologic, and in situ stressdata available for the eastern nger, 1973] and a similarly oriented zone in central Virginia United Statesgenerally yield consistentresults. Along the At- [Bollinger,1975], the NW-SE trending Boston-Ottawaseismic lantic coast,the primary data set available is composedof the zone [Diment et al., 1972],the Ramapo fault systemin north- trends and sensesof offset of Tertiary or younger faults. As eastern New Jersey [Aggarwal and Sykes, 1978], the Grand discussedpreviously, these data yield only approximatestress Banks area, the Lower St. Lawrence valley [Leblanc et al., directions.A more significantproblem concerningthese in- 1973],and a diffuseearthquake trend along the Appalachian dicators, however, is that chance exposuresmay reveal sec- fold belt [Woollard, 1969]. Sykes [1978] points out that the ondary features, whereas the primary structure can remain Boston-Ottawaseismic zone may not be continuousbecause unrecognized.In addition, the general paucity of stressfield of the paucity of activity in southern Vermont and New indicators in the East makes definition of the boundaries be- Hampshire. The boundary separatingthe Midcontinent and tween the stressprovinces difficult and may tend to give un- Atlantic Coast stressprovinces is shownin Figure 3 by the warranted significanceto isolateddata points. shadedline. In the southernAppalachians more activity oc- Although the central and easternUnited Statesis generally curs west of the boundary, while in the northern Appala- thought of as a tectonicallystable region, numeroushistorical chians the NW-SE trending zone of earthquakesassociated earthquakeshave occurredthere (Figure 3). The Charleston, with the Ottawa-Bonnechere graben [Sykes, 1978] end

Fig. 3. Seismicitymap of easternNorth Americafrom historicaland instrumentaldata for the period1928-1971 from York and Oliver [1976].Shaded line is stressprovince boundary between the Midcontinentand Atlantic Coast provinces from Plate 2. 6136 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES abruptly at the stressprovince boundary. Earthquakes extend- interpreted as indicating that a major decollementunderlies ing from New Jerseyto New England seemto define a north- thin-skinned deformation in the miogeosynclinalsequence westly trend of activity located east of the stressprovince along the westernedge of the belt (Appalachian Plateau and boundary. Valley Ridge). This decollement is interpreted to continue Various sourcesand mechanismshave been proposedto ex- eastward beneath Paleozoic and Precambrian rocks pre- plain the distributionof intraplateearthquakes in the central viously believed to be rooted forming the mobile core of the and easternUnited States.Some of theseare preexistingzones Appalachian orogen. Stressorientations within the Appala- of crustalweakness that correlatewith zonesof alkalic mag- chianPlateau along the New York-Pennsylvaniaborder (NY- matism and major offshorefracture zones [Sykes, 1978], reac- 1, 2, PA-3) are parallel to orientationsfarther to the north and tivation of a late Precambrian-early Paleozoiccontinental rift west in the Midcontinent province; these localitieslie within in the New Madrid area [M.D. Zoback et al., 1980a], reacti- detached rock. The orientations at these three sites are also vation of Triassic basin bounding normal faults (with high- parallel to the stress orientation inferred from the Attica angle reverse-typemotion) along the easternseaboard [Jaco- earthquake(NY-3), which is probablydeeper than the thrusts been, 1972; Mixon and Newell, 1977; Aggarwal and Sykes, now known to have detached the rocks there. Thus the stress 1978; Behrendtet al., 1980; Wentworthand Mergner-Keefer, data seemsto imply that the Appalachianthrust sheet(s)is 1980; Prowell, 1980], and stressconcentrations in the vicinity mechanicallycoupled to the underlyingbasement where, pre- of igneousintrusions [McKeown, 1978;Kane, 1976]. sumably,the major earthquakesoccur. The high level of intraplateseismicity in the Northern Miss- Midcontinent isippi Embayment might appear to warrant delineationof a The Midcontinent stressprovince, a large and tectonically distinct stressprovince. Earthquakeswithin the embayment stable region, extendseastward to the Appalachians(and in- that occur along a 100 km long northeast-southwesttrending dudes the Adirondack uplift), southwardto the Gulf Coastal seismiczone in northeastArkansas (All-1, 2) indicate pre- Plain, and westward to the southern Great Plains province. dominately strike slip faulting and approximatelyeast-west The available data suggestthat this stressprovince can be ex- compression.These earthquakesare occurringon a major tended northward into Canada. Throughoutthis broad prov- basementfault associatedwith a late Precambrianrift system ince, a uniform NE-SW compressivestress field exists,largely that is favorably oriented to the regional stressfield [M.D. defined by hydraulic-fracturing work of B.C. Haimson [cf. Zoback et al., 1980a].Earthquake focal mechanismsfrom the Haimson, 1977b] and earthquake focal mechanismsof R. B. roughly north-southtrending, relatively diffuse zone of seis- Herrmann [cf. Herrmann, 1979]. micity near New Madrid, Missouri (MO-1, 2), however, in- The easternboundary of the Midcontinent stressprovince is dicate thrusting on northwest trending faults and NE-SW poorly defined but appearsto coincidein the south approxi- compression.This rapid and localized apparent change in mately with the eastern edge of the Blue Ridge province, a stress orientation is observed in both regional surface and major topographicbreak. To the north, where the Appala- body wave studies[Herrmann, 1979] and local studiesof mi- chian fold belt narrows,the stressprovince boundary takes a croearthquakes[O'Connell et al., 1980].However, the validity more northerly coursecrossing the Valley and Ridge province of the apparentcomplex pattern of stressorientations is ques- and approximatelycoincides with the westernedge of the Ap- tionable, as it may be due merely to the uncertaintyinherent palachian fold belt. As stressorientations in the Valley and in inferring stressorientations from focal mechanismsfor Ridge and Blue Ridge provincesare scatteredand locally con- earthquakesoccurring on preexistingfaults. Thus as the mean tradictory (Plates 1 and 2) the eastern•200 km of the Mid- orientation of the stressfield in the embayment area is consis- continent stressprovinces may representa transitionzone be- tent with the surroundingregion and the earthquakesare ap- tween the Midcontinent and Atlantic Coast provinces.The parently associatedwith specific;tructures, the Northern Mis- scatterin stressorientations in this region may reflect the in- sissippiEmbayment is included as part, albeit an active part, herent uncertainties of the different methods used to deter- of the Midcontinent stressprovince. mine stressor may represent a broad zone of transition in Atlantic Coast which the horizontal stressesare approximatelyequal in magnitude and the apparent stressorientation is more controlled Northwest-southeast to west-northwest-east-southeast com- by local inhomogeneties. pression characterizes the Atlantic Coast stress province, Tectonicsin the Midcontinent stressprovince are compres- which includesthe Atlantic CoastalPlain, the Piedmontprov- sional, and earthquakesthroughout the province typically ince of the southernAppalachians, and the entire Appala- show componentsof both strike slip and reversemovement. chian fold belt in the northeast. In marked contrast to earlier Thus studies suggestingnortheast compressionfor this area [see Sbar and Sykes,1973], the generallynorthwest compression is documentedby offsetcore holes (CT-1), hydrofracmeasure- SNE >> SNW > Sv ments (SC-1, MD-2), earthquakefocal mechanisms(NJ-1, SC-6, VA-1), and numerous late Cenozoic fault offsets.Be- The only exceptionto this pattern was a normal fault earth- causeof generallypoor, limited exposuresand small magni- quake at the edge of the Ozark uplift in southern Missouri tude vertical offsetsthe faults are difficult to recognize,and (MO-3), but in this casethe orientationof the least principal possiblelateral offsetsof the subhorizontalcoastal plain sedi- horizontal stress field is consistent with the northwestward di- mentscan rarely be establishedor eliminated.Components of rection observedin the remainder of the province (Plates 1 strike slip offsethave been postulatedfor two major reacti- and 2). vated (now reverse)northeast trending fault zones.A com- Recent seismicreflection profiling in the southernAppala- ponent of right-lateral offsethas been inferred along the Staf- chians [Cook et al., 1979; Harris and Bayer, 1979] have been ford fault zone (VA-2) from an analysisof subsidiaryreverse ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6137 and normal faults within the fault zone [Mixon and Newell, is vertical, the orientation of the least principal stressis per- 1977], whereas significantleft-lateral offset has been docu- pendicular to the continental margin, and the magnitude of mented in Paleozoic rocks along the Cenozoic Belair fault the least principal stressis 60% of the vertical stress[cf. Hub- zone [Prowelland O'Connoi',1978], although this offsetcan bert and l•illis, 1957;M.D. Zoback et al., 1978]. It should be not be recognizedin the overlyingstrata. Lacking data on lat- emphasizedthat the stateof stressin this provinceis probably eral offsetsand given the large uncertaintiesin stressorienta- only the result of sediment loading and not tectonic forces. tions inferred from only the trend and senseof vertical offset The state of stresswithin the bedrock underlying Gulf Coast of a fault, the precisehorizontal principal stressorientation re- sediments remains unknown. mains poorly defined. However, the excellentcorrespondence SOURCES OF STRESS between stress orientations inferred from the trend of reverse faulting along the Stafford (VA-2) and Brandywine (MD-I) In this sectionwe investigatesome of the major sourcesof and a nearby in situ stressmeasurement (MD-2) imply that lithosphericstress in termsof the patternsand relative magni- the predominantsense of modernmotion on thesefault zones tudes of the stressesthat characterizethe provinces defined is reverse. Further substantiation of the Atlantic Coast stress above. The stateof stressin the earth's lithosphereis the result provinceand its continuationinto New England is provided of superpositionof forcesderived from many different proc- by a recent compilation of 18 focal mechanismsin the area esses. It is often possible, however, to identify the pre- east of 73øW longitude and north of 43øN latitude that in- dominant force or forcesresponsible for the pattern of stress dieate largely reversefaulting on N to NE strikingplanes im- field within a given region. The broad patterns of stressare plying E-W to NW-SE compression([Graham and Chiburis, best explained by plate tectonic forces(see Richardsonet al. 1980] thesedata have not been included in this study because [1979] for a comprehensivereview both tectonic and non- of the late date at which they becameavailable). tectonic sourcesof stress). Earthquakes in the Atlantic Coast province often have componentsof both thrust and strike slip motion [cf. A ggar- Sourcesof Stressin the WesternUnited States wal and Sykes, 1978; Graham and Chiburis, 1980]. Such The most important sourceof stressaffecting the western oblique motion is expectedbecause the earthquakesseem to United States is current and past plate motions along the occur on favorably oriented preexistingstructures. For mod- westernplate boundary. The pattern of stressin many of the em earthquakes associatedwith northeast striking normal stressprovinces in this region can be attributed directly either faults which bound Triassic basins the current sense of motion to presenttransform motion betweenthe Pacific and North (reverse)on the fault is exactly oppositeto the motion that America plates or residual thermal and dynamic effects of created the faults. For the Atlantic Coast stress province pastsubduction. The northwesterlymotion of the Pacificplate therefore with respectto North America resultsin fight-lateral transform motion along the northwesttrending San Andreas and S ! > S 2 > S3 Fairweather fault systems, N-S crustal shorterning in the S•w > S•. > Sv TransverseRanges of Southern California, and distributed fight-lateral deformation superimposedon the extensionoc- An early earthquake focal mechanismstudy near Charles- curring in the Basinand Range province.The Sierra Nevada, ton, South Carolina, suggesteddip slip motion on a near-ver- characterizedby both normal and strike slip faulting, appears tical, northwest striking fault plane [Tarr, 1977]. For that to representa transition zone between the San Andreas and event the P and T axes both strike in the northeast-southeast Basin and Range stressprovinces. direction and plunge -45 ø. Because the horizontal com- North-south compressioncharacterizes the Pacific North- ponentsof both the P and T axeshave the sameazimuth (and west stressprovince, a region including an active andesitic the relative magnitudeof the stresscomponent in the null di- volcanic chain and the back arc area. This direction of com- rectionis unknown),the directionof leasthorizontal compres- pressionis roughly parallel to the trend of the offshoretrench, sion could not be determined. More recent focal mechanism and it is about 45ø oblique to the northeast-southwestdirec- data indicate reversemotion on both NE and NW trending tion of convergencebetween the Juan de Fuca and North faults and thus suggestthat the horizontal stressesmay be ap- American plates.Whereas in other back arc areasunder com- proximately equal (A. Tarr, written communication,1980). pressionthe maximumcompressive stress orientation is in the Northeasttrending high-angle reverse faults with Tertiary off- direction of plate convergence[Nakarnura et al., 1978],this is set, recently discoveredby seismic reflection profiling near not the casein the Pacific Northwest province. Crosson[1972] Charleston by Behrendtet al. [1980] are indicative of NW-SE has suggestedthat decouplingbetween the subductingJuan compression. de Fuca plate and the overridingNorth American plate may be consistentwith estimatesof a relatively slow rate of con- Gulf Coast vergenceand the absenceof an inclined seismicBenioff zone Active listrie growth faulting characterizesthe Gulf Coast beneathwestern Washington and Oregon.If thisd6coupling stressprovince. The stressdirections used to define this prov- has occurred,then the stressfield may be controlled more by ince were taken from the averagelocal orientation of recently the larger-scaletransform motion between the Pacific and active faults at each site, from the fault map by Howard et al. North American plates.An observationin possiblesupport of [1978], and from recognizedlate Tertiary normal faulting. this conceptis the rather limited extent of the Juan de Fuca The axis of most currently active faulting runs near and sea- plate (extendingfrom approximately40øN to 51øN) in rela- ward of the presentshoreline. The northern boundary of this tionship to the extensiveSan Andreasand Fairweather fight- province is constrainedonly by a strike slip earthquake in lateral transform fault systemswhich form most of the re- Mississippi(MS-3). The state of stressthroughout this prov- mainder of westernplate boundary of North America. An- ince is apparently quite uniform: the greatestprincipal stress other hypothesisto explain the N-S crustal shorteningin the 6138 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Pacific Northwest province was proposed by R. B. Smith until about 6 m.y.B.P. [Eberly and Stanley,1978; Scarborough [1977]. Within the context of a subplatemodel of the entire and Sha•qullah, 1979]; whereas in the northern Basin and westernCordillera he suggestedthat the N-S shorteningin the Range, basin-rangefaulting was initiated post 10 m.y.B.P. Pacific Northwest may be due to compressionbetween a N-S and in places post-7 m.y.B.P. under a stressfield oriented extendingNorthern Rocky Mountain subplate(with a NW similarily to the modern one (M. L. Zoback et al., 1980]. The trending western boundary) and the Juan de Fuca plate. •,45 ø clockwisechange in stressorientation occurred between The available stressdata, however,though admittedly sparse, 14-7 m.y.B.P. (possiblyaround 10 m.y.B.P.) as constrained suggestgenerally east-west extension in the Northern Rocky both locally and by regional differencesin faulting trends [M. Mountain provincewith only a small area of north-southex- L. Zoback et al., 1980]. tensionin the HebgenLake-Centennial Valley stressprovince M. L. Zoback and G. A. Thompson[1978] attributed the (discussedbelow). clockwisechange in stressorientation to superpositionof dex- The origin of present-dayextensional tectonism in the Ba- tral shear related to the development of the San Andreas sh and Range and Rio Grande rift provinceshas been the transform on the Miocene back arc extensional stress field. subjectof numberousinvestigations (see Stewart [1978] for a Since transform faulting may have begun as early as 29 m.y. summary of the various proposedtheories). Models for the B.P., the changein stressorientations may mark the timing of origin of the modern state of stressmust take into account the significantor increasedcoupling of right-lateral shearbetween late Cenozoic evolution of the western plate boundary. Sub- the Pacific and North American plates. This increasedcou- duction under the western United Statesin middle Tertiary pling may have resultedfrom one or a combinationof the fol- time destroyed the subductingFarallon plate more rapidly lowing effects:(1) lengtheningof the San Andreas transform than it was being created,resulting in a ridge-trenchencoun- to a 'critical' length [seeChristiansen and McKee, 1978],(2) an ter and developmentof the San Andreastransform accommo- acceleration of relative motion between the Pacific and North dating the relative motion between the Pacific and North American plates at approximately 10 m.y.B.P. [Atwater and American plates [Atwater, 1970; Atwater and Molnar, 1973]. Molnar, 1973], and (3) greater normal stressand hence in- Initial ridge-trenchencounters occurred between 20-30 m.y. creasedfriction, along the San Andreaspossibly due to small B.P., possibly-•29 m.y.B.P., and probably closelycoincided changein the pole of rotation. with the initiation of right-lateral transform faulting. Since its Distributed right-lateral shear related to the San Andreas inception,the right-lateraltransform boundary has continued transform,superimposed on fundamentalextension within the to lengthen at the expenseof the subductionzone. Basin and Range has been proposedto explain the modern Extensional tectonism in the Basin and Range and Rio stateof stressin the Basin and Range [e.g., R. B. Smith, 1977, Grande rift was initiated in both an intra-arc and back arc set- 1978; Eaton et al., 1978, Stewart, 1978; M. L. Zoback and G. ting [e.g., Eaton, 1979;Elston and Bornhorst,1979]; true basin- A. Thompson,1978; Eaton, 1979, 1980].Continued extensional range structure,the fault-controlled sedimentarybasins and tectonismin a 'back transform' settingmay be related to de- topographicforms resemblingthose seen today, can be con- velopment of a gradually enlarging 'slab-free' region beneath sidered a unique late-stageepisode of extensionaltectonism the part of the continental block adjacentto the San Andreas [M. L. Zoback et al., 1980].The generalNNE trend of ranges transform, while subduction continues to the north and south in the northern Basin and Range, roughly perpendicularto of the transform [Dickinsonand Snyder, 1979]. Dickinson and the modern least principal stressdirection, suggestsdevelop- Snyder attribute basin-range extensionaltectonism and bi- ment of this region in a stressfield similar to the modern one. modal volcanismto upwelling of hot asthenospherereplacing This NNE trend of faulting contrastsmarkedly with the gen- the volume of mantle formerly occupiedby the subducted eral NW trend of the structuralgrain in the southernBasin slab of lithosphere[cf. Stewart, 1978;Best and Hamblin, 1978]. and Range. The Rio Grande rift and the individual ranges Second-ordereffects can be invoked to explain regional within it trend roughlyN-S and are thoughtby someto follow variationsof the stressfield within the Basin and Range stress a zone of weaknessresulting from superimposedearlier defor- province. The regional variation from WNW to NW exten- mations [e.g., Chapin and Seager, 1975]. sion on normal, oblique slip, and strike slip faults in the west- A low level of modern seismicityas well as geomorphicevi- ern Great Basin (northern Basin and Range) to -•E-W di- dence(including the paucityof Quaternaryscarps) argue that rected, pure extensionin the Great Basin has been attributed much of the deformationresponsible for present-dayranges to an eastwart diminishing effect of distributed shear related and basinsin the southernBasin and Range occurredearlier to the transformplate boundaryon basin-rangeextension [R. than that in the northern Basin and Range. The general NW B. Smith, 1977, 1978; Best and Hamblin, 1978; Eaton et al., trend of the structuralgrain in the southernBasin and Range 1978; Eaton, 1979; Davis, 1979]. The apparent WNW exten- suggeststhat the region developedlargely under a stressfield sion documentedby the presentstudy in the Rio Grande rift, oriented differentlyfrom the modernone [Eaton, 1979].Dike which lies still farther east, remainsunexplained by this hy- trends and extension directions inferred from fault trends and pothesis. related stratal tilt directionssuggest a SW-NE to WSW-ENE A secondmajor regional variation in least principal stress direction throughout the western United States in Miocene orientation within the Basin and Range (and possibly the time (•,20-10 m.y.B.P.) [11,1.L. Zoback and G. A. Thompson, Sierra Nevada) occurs in a broad east-westtrending band 1978;M. L. Zoback et al., 1980].This leastprincipal stressdi- transectingsouthern Nevada and Utah. In this region the least rection is roughly perpendicularto the trend of the rangesin principal stressdirections trend more northwestwardand rep- the southernBasin and Range and is oriented -•45 ø different resent a clockwise rotation relative to the well-established from the modern WSW-ENE least principal stressdirection. WNW-ESE to E-W extensionin the region to the north. This Regional studiesin the southern Basin and Range indicate east-westtrending band that crossesthe southernGreat Basin that major basin-rangefaulting, responsiblefor the modern coincideswith an east-westtrending zone of seismicityand a ranges, occurred 13-10 m.y.B.P. and probably continued up major decreasein elevationfrom the Great Basin (mean ele- ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6139 vation, • 1700 m) to the Arizona-New Mexico (southern)Ba- suggeststhat the sourceof the compressionwithin the plateau sin and Range (mean elevation,-750 m) that is accompanied interior may be related to dynamic (drag?) forces resulting by a • 100 regal increasein gravity of crustalorigin [Eaton et from the regional extension. The stressdifferences arising al., 1978].Presumably, this east-westtrending zone actsas the from lateral densitycontrasts along the plateau marginsmay boundary betweenthe actively extendingregion to the north representa second-ordereffect superimposedon a more re- and the currently tectonically quiescentsouthern Basin and gional stress field. R. B. Smith [1977] proposed that the Range. Thus this zone can be interpretedas a broadband of roughly E-W compressionwithin the Colorado Plateau is due left-lateral transform-style shearing consistent with focal to the buttresseffect betweenthe westerndeforming Cordil- mechanismsand geologicevidence [Stewart, 1978] in that re- lera and the easternstable interior. Such an interpretation is gion. Locally, this left-lateral shearing(or drag resistanceto valid only if a fixed boundaryexists for the westerndeforming this shearing) could cause a clockwiserotation of the stress zone; if, however, this boundary is consideredfree, then ex- field, compatiblewith the more northwestwardleast principal tensionin both the Rio Grande rift and Basinand the Range stressorientations observedin that region. need not necessarilyaffect the plateau stressfield. The E-W to NE-SW extension in the Northern Rocky A probably extensionalstress regime in the southernGreat Mountain province may representa northward continuation Plains distinguishesthat province from the Colorado Plateau of basin-rangeextension across the Snake River plain. Mod- interior (which appears to be dominated by compressional em leastprincipal horizontal stressorientations for the Snake tectonism)despite similar least principal stressdirections in River plain are NW-SE, parallel to the axis of the plain, not the two regions.The sourceof the apparentNNE-SSW exten- perpendicularto it. Profound migrating volcanismalong the sion in the Southern Great Plains is obscure and is in marked Snake River plain has been attributed to a mantle hot spot contrastto the roughly E-W compressionresponsible for de- [Morgan, 1972;Suppe et al., 1975]and to a propagatinglithos- velopment of the Rocky Mountains in Laramide time (•80- pheric crack possiblyalined with a preexistingzone of weak- 40 m.y.B.P.). Furthermore,analysis of the 22-25 m.y. Spanish ness[e.g., R. B. Smith et al., 1974b;Eaton et al., 1975]. In favor Peak radial dike systemin south-centralColorado (37.63øN, of the hot spotmodel an inversionof relative plate motionsto 104.88øW) suggestsa strike slip regime with a N8øW least a mantle reference frame indicate that the absolute motion of principal horizontal stressorientation [Muller and Pollard, the North American plate for the last 10 m.y. is well repre- 1977]; this orientation is 20o--40ø oblique to inferred least sentedby the Snake River plain-Yellowstonetrend [Minster principal horizontal stressdirections for past 5 m.y. cinder- and Jordan, 1978]. cone alinementsin the Raton-Clayton volcanic field .directly The N-S extensionin the Hebgen-Lake-Centennial Valley to the south in New Mexico. province appearsto representa local anomaly in the regional Sourcesof Stressin the Central pattern of E-W or NE-SW extensionin the Snake River plain and Eastern United States to the south and the Northern Rocky Mountains to the north. R. B. Smith and M. L. Sbar [1974] first recognizedthis local To accountfor the stressfield in this broad midplate region, deviation in the stressfield and suggestedthat it resultedfrom we considertwo large scalesources of stress;'asthenospheric a generally N-S relative separation of the Northern Rocky drag' and ridge push. Asthenosphericdrag is viscousresis- Mountains and the Great Basin in their subplate schemefor tance to motion of the thick continental lithosphere. Most the western United States.The abrupt southernboundary of simply, if the asthenospherebeneath the continentsis station- this provincein the Yellowstonearea, however,suggests that ary, a basalshear stress is inducedby the motion of the conti- the N-S extensionmay be a local, possiblysecond-order effect nental lithospherein the directionopposite to the directionof related to the actual mechanismor cause for the profound absoluteplate motion [Richardsonet al., 1976]. A somewhat propagating volcanic trend along the eastern Snake River related sourceof stressis asthenosphericcounterflow; numer- plain. ous theoreticalmodels have been developedin which the as- The Colorado Plateau interior appears characterized by thenospherebeneath the continentsflows in directionsdeter- WNW-ESE compression,i.e., compressionin the direction of mined by the total massflux: that which is producedat ridges extension in the surrounding Basin and Range and Rio and consumed at subduction zones [Harper, 1978; Chase, Grande rift provinces.The sourceof this WNW directedcom- 1979; Hager and O'Connell, 1979]. In this case the astheno- pressionwithin the plateau interior remains equivocal. As- spherehas a directionof motion not necessarilyrelated to the thenospheric'drag' (discussedbelow) can be ruled out, as the absoluteplate motion direction,and the resultingstress in the direction of absolutevelocity of the plateau is SW, nearly per- lithosphereshould be in the directionof asthenosphericflow. pendicular to the WNW direction of maximum horizontal The absoluteand counterflowvelocity directionsare nearly compression.Thompson and Zoback [1979] have suggested parallel for the Pacific plate, which, becauseof its size and that the overall state of compressionin the Colorado Plateau speed,is the prime determinantof the large-scaleflow pattern; interior may be relatedto its thick 'keel' of mantle lithosphere, however,for other platesthe directionscan differ up to 90ø . the edgesof which are subjectedto a static 'ridge push' force Broad-scalelateral densityinhomogenities give rise to lith- (also discussedbelow) resultingfrom densitycontrasts by sur- osphericstresses such as the ridge push force [e.g.Frank, 1972; rounding, less dense asthenosphereunder the Basin and Artyushkov,1973]. These forcesarise becauselateral changes Range-Rio Grande rift. However,shallow (crustal level) den- in the thicknessof the crust and lithosphere(and changesin sity contrastsbetween the plateau and surroundingregions crustal density) result in buoyanceforces oriented normal to predict extensionalstress differences in the upper crust along the interfacedefining the lateral densitychange. As only the the margins of the plateau. vertical component of this bouyance force is compensated The parallelism betweenthe direction of greatestprincipal isostatically,a net horizontal force is left. For example, ridge stresson the plateau and the directionof regionalextension in push resultsfrom both the elevation of the midocean ridge the surroundingBasin and Range-Rio Grande rift province and the coolingand thickeningof the lithosphereaway from 6140 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES the ridge; the magnitude of the stressderived from is ridge ern United States. The asthenospherecounterflow models push force is estimatedat severalhundred bars [Hales, 1969; mentioned previously result in basal shear stressdirections Frank, 1972; McKenzie, 1972; Artyushkov, 1973]. Other ex- (lithosphericcompressive stress directions) indicated by the amples of lateral density inhomogeneitiesinclude lateral vari- large arrows in Figure 4 [after Chase, 1979]. ations in crustal densitiesand thickness(e.g. ocean-continen- As can be seenin Figure 4, the roughly NE-SW compres- tal crust interface [Bott and Dean, 1972]) as well as lateral sion in the Midcontinent stressprovince is generally consis- contrastsin lithospherethickness. tent with either ridge push or asthenosphericviscous drag re- To examine sources of stress in the central and eastern sistance to absolute plate motion. Examination of the United States,Figure 4 comparesthe direction of maximum histograminset in Figure 4, however,indicates that drag resis- horizontal compressionwith computeddirections of (1) abso- tance directions correlate better with the stress data than do lute motion of the North American plate and (2) ridge push the ridge push directions,suggesting that drag may be the pri- (from the relative motion of North America and Europe), us- mary sourceof stressin the Midcontinent province. ing polesof rotation from Minster and Jordan [1978].(The ori- A potential problem with the asthenosphericdrag hypothe- entation of the fundamentally E-W compressionderived from sisis that the magnitudeof the basal stress(based on plate ve- ridge push would have been modified only slightlyif the pole locity and asthenosphericviscosity) is typically interpretedto of relative motion between North America and Africa had be quite low [e.g., Chappleand Tullis, 1977;Richardson et aL, been used.) As can be seen,the relative and absolutevelocity 1976; Chase, 1979]. However, the cold thick lithosphere be- azimuths differ by only -10 ø throughout the central and east- neath the Midcontinent region and the lack, or poor develop-

115ø I10 ø 105 ø I00 ø 95 ø 90 ø 85 ø 80 ø 75 ø 70 ø 65 ø MII•-CONTINENT RtlDGE PUSH ATLANT,C\ DIRECTION COAST RIDGE PUSH I • I 50ø t DIRECTION• "ABSOLUTE" VELOCITY •o ...... DIRECTION DIRECT ION

m -80 -40 0 40 ø•o' ' '_,•o...... •,o •5 ø DEGREES • DEGREES

';5ø

ji •,0 ø CONTI

,;tOo

35 ø

35ø ERN/ / ' GRE• , Ig I 50 ø

EXPLAN;TION GREATESTPRINCIPAL STRESS DIRECTION ...•'•'• CALCULATED"ABSOLUTE" VELOCITY o, ,qo •,oo •o,oM,LES .....ß'"" CALCULATED"RIDGE PUSH" DIRECTION ,o ,o,o 2,oo3,oo ,(,LOMETERS e,•,*""•DRAG FROM "COUNTER FLOW" MODEL •5 ø

105 ø I00 ø 95 ø 90 ø 85 ø 80 ø 75 ø

Fig. 4. Comparisonof greatesthorizontal principal stress directions (from Plate 2) with predictedmaximum compressive stressdirections derived from ridge push (dotted lines) and drag resistanceto 'absolute'plate velocity (dashed lines) calculatedfrom polesof rotation of Minsterand Jordan[1978] (see text). Large arrowsshow direction of net asthenospheric counterflowfrom modelsof Chase[1979]; drag-induced compressive stress in the lithospherewould have the same orientation as the flow. Heavy solid lines mark stressprovince boundaries, as in Plate 2. Short dashedlines mark edge of the continentalmargin. Inset histogramsshow angular deviation of stressdata from predictedridge push and absolutevelocity directions.Positive deviation (to the right) indicatesa clockwiseorientation of the actual data in relation to the calculated direction,negative deviation indicatesa counterclockwiseorientation of the data in relation to the calculateddirection. ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6141 ment, of a low-velocity zone [Herrin, 1972; Chapmanand Pol- compressionalstress along the Atlantic Coast province [See- lock, t 977] imply that the baseof the lithospheremay extend ber and ,4rmbruster,1980]. Such a mechanismpredicts exten- into more viscousmantle material. This implies that the resis- sional stressesin the main elevatedcore of the Appalachians. tance to lithosphericmotion may be fairly large. If estimates However, the only availablestress data in that region [Schafer, of flow stressesbased on grain size analysisof xenoliths are 1979] indicates Recent thrusting, implying a compressional correct, the magnitude of shear stressesin the lower litho- state of stress. spherecould exceed tOObars [Mercier, 1980]. Slowly accumulating flextural stressesin responseto ero- Compressive stress directions predicted from astheno- sion-inducedisostatic rebound of the Appalachiansis a some- sphericcounterflow models discussed previously (large arrows what speculativemechanism that would generatecompressive Figure 4) are nearly 90ø different than observedmaximum stressperpendicular to the topographictrend of the Appala- compressivestress directions in the Midcontinent region. This chians.An analogyto this processis unloadingdue to melting large discrepancyimplies that as modeled, asthenospheric of an ice sheet[see Stein et al., 1979].Unloading resultsin ex- counterflowis not an important processcontrolling the intra- tensional flextural stresses in the area where the load is re- plate stressfield. Perhapsasthenospheric flow gets channeled moved(Appalachians) and compressionalstresses in the adja- around regionsof thick continental lithosphere,and current cent area (coastalplain). Thus in a general sense,erosion in models assuminga uniform lithospherethickness worldwide the Appalachiansmay induce the type of compressionalstress are too simplistic. It is also possiblethat the flow occurs at field observed in the central and southern Atlantic Coast suchgreat depths as to not affect the lithosphericstress field. province. A reasonableaverage erosion rate for the Appala- In the Atlantic Coast province, maximum compressionap- chiansis about3 x 10-2 mm/yr [Hack, 1979],and throughout pearsto be orientedNW-SE, perpendicularto the trend of the the Cenozoic,isostatic rebound has apparently kept the mean Appalachianfold belt and the continentalmargin. As can be elevation of the mountains nearly constant [Hack, 1979]. seen Figure 4 and the inset histogram,both ridge push and Flexure produced by isostaticrebound will result in an accu- drag resistanceto absolute velocity are in poor agreement mulation of about tOO-150bars of compressionalstress in the with the observed maximum horizontal compressionin the upper 10 km of the lithosphereevery t0 m.y. [seeStein et al., Atlantic Coast province, whereas compressivestresses pre- 1979]. This does not seem to be an unreasonablerate of stress dicted from the asthenosphericcounterflow model match the accumulationin light of the low recurrencerate inferred for observedorientations quite well. However, as this counterflow great earthquakesin the Atlantic Coast provinceon the basis is directedfrom the northwest,it appearsunlikely that it could of the small magnitude of post 100 m.y. cumulative fault off- exert a significanteffect on the lithosphericstress field in the sets (D.C. Prowell, written communication, 1980). Diffi- Atlantic Coast province without having any apparent influ- culties with the erosion-inducedflexture hypothesisare (1) ence on the stressfield in the Midcontinent region. predictedextensional deformation within the reboundingAp- The fact that the orientation of the maximum compressive palachiansis not observedand (2) the correlationbetween to- stressin the Atlantic Coast provinceis roughly perpendicular pography and the predicted areas of compressivestress in to the trend of the Appalachian fold belt and the continental New England is not good. However, the effectsof glacial load- margin would seem to suggestsome sort of causal relation- ing in the New England area may complicatethe stressfield in ship. However, lateral densitycontrasts associated with crustal that region. structure at a passivecontinental margin predicts extension A secondspeculative mechanism that mightresult in the perpen•icular to the continental margin rather than the ob- maximum compressivestress direction being perpendicularto served,•ompression [e.g., Bott and Dean, 1972]. Lithospheric the Appalachian fold belt is a reorientation of the stressfield flexture resulting from sediment load also predicts extension by highly anisotropicbasement structure (N.H. Sleep, oral on the continental margin [Walcott, 1972; Watts and Ryan, communication, 1980). In other words, a compressivestress 1976; Turcotteet al., 1977]. Compressivestresses derived from field resulting from ridge push or asthenosphericdrag might major lithosphericthickening along the edgeof the continent be oblique to the fold belt (Figure 4), but the direction of also appear unlikely; available data suggestlittle or no con- maximum compressionis rotated by the anisotropicstructure trast in thicknessbetween continental and oceaniclithosphere perpendicularto the compliant direction. in this region. While no seismicinvestigations are available It may seemfruitless to try to identify a single,or dominant, for the Atlantic Coastal Plain or the Appalachian region, sur- sourceof stressfield in the Atlantic Coast province.The stress face wave studiesof the Gulf Coastal Plain indicate a possible field may reflect superpositonof erosion and sedimentation lithosphere thickness between 80 and 145 km [Biswas and induced flexture, deglaciation, ridge push, asthenospheric Knopoff,1974]. Age of the oldestoceanic lithosphere 'm the drag, and other sources.However, available data on faulting central Atlantic is generally believed to be 180-190 m.y. in the Atlantic Coastal Plain (D.C. Prowell, written commu- [Watts and Ryan, 1976]. Surfacewave studiesin the western nication, 1980) indicate that directions,senses, and rates of Pacific suggesta thicknessof 140 +_35 km for oceanic litho- motionon faultsin thisregion have not changed markedly in spherewith a mean age of 150 m.y. [Leedset al., 1975]. the last t t 0 m.y., thussuggesting that a similarstress field has As mentionedpreviously, seismic reflection profiling in the acted throughout that entire time period. southernAppalachians have been interpreted as indicating a In the Gulf Coastal Plain the seaward extension caused by major decollement (detachment) beneath the Appalachians active normal faulting appears consistentwith the state of extending eastward from the leading (western) edge of the crustalstress at a passivecontinental margin predicted by lat- fold and thrust belt to beneath the Piedmont [Cook et al., eral density variations such as those describedby Bott and 1979;Harris and Bayer, 1979]. Southeastwarddirected gravi- Dean [1972]. However, as the active faults are listric growth tational backslidingalong this detachment(and its postulated faults within the sedimentarysection, it is possiblethat the eastwardextension beneath the Coastal Plain) has been sug- faulting merely reflectsthe effect of sedimentloading and not gestedas a mechanismto explain the observedorientation of the stateof lithosphericstress. Since the sourceof stressin the 6142 ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUS UNITED STATES

Atlantic Coast province remains poorly understood,the state The geologicstress indicators consideredhere also involve of stresswithin the Gulf Coastbasement might be dominated difficulties. Dike trends and cinder cone alinements are not al- by a stressfield similar to that in the Atlantic Coast province, ways linear and can be influencedby favorablyoriented pre- the Midcontinent province,or it may be dominatedby lithos- existingfracture and joint sets.To usefault slip data correctly, pheric flexural effectsresulting from the sedimentaryload. major block movementshould be consideredbecause anomalous small-scalemovements may occur between major blocks DISCUSSION and thus yield erroneousresults. The use of offset-corehole Stress Field Indicators data, such as those of Schiifer [1979], can be questionedbe- cause the magnitudes and rates of movement are so large The regional consistencyof the in situ stressdata presented (centimetersof motion in severalyears) as to be possiblynon- above implies that they generally reflect the pattern of stress tectonic in origin. in the crust. At a given locality the good correspondenceof the variousmethods used to determinestress orientations suggests that the assumptionsrequired to determine principal Stress Provinces stressorientations by the differentmethods are basicallyvalid. On the basisof the orientationsand relative magnitudesof At the Nevada Test Site (NV-3), for example,all three meth- principal stresses(determined from the currentstyle of tecton- ods were utilized: hydraulic fracturing and overcoring at ism) we have divided the United Statesinto stressprovinces. depth, geologicindicators, and strike slip and normal faulting Figure 5 showsa generalizedversion of the stressmap on (as well as oblique slip) earthquakefocal mechanismsyield which the stressprovinces are indicated; a summary of the consistentresults. Furthermore, the generallygood correspon- principal stressorientations and currentstyles of tectonismfor dence of stress orientations inferred from focal mechanisms the different stressprovinces is given in Table 2. It is quite which commonly sample the depth range 5-15 km and from likely that thesestress provinces, particularly in the East, will geologicand in situ stressindicators which generally sample be refined as new data accumulate.The presentboundaries the upper ---2 km implies a relatively uniform stressfield appear to represent the most general interpretation of the throughoutthe upper crust (<15 kin). This observationlends stressdata consistentwith available information. Although confidenceto extrapolationof the near-surfacestress field (<_2 more complexpatterns could be drawn, presentdata coverage kin) to depthscomparable to thoseof major earthquakesand doesnot appear to warrant them. also suggeststhat in situ stressmeasurements can be used to In general, in the tectonically active westernUnited States fill important gaps in the current data set. the stresspattern is complex, and numerous distinct stress The least reliable regional tectonic stressindicator seemsto provincescan be well delineated.These stressprovinces gen- be the overcoringmeasurements made in mines. For example, erafly correlatewell with the physiographicprovinces. In the in the Coeur d'Alene districtin Idaho (ID-7), markedly differ- central and easternUnited States,it is apparenteven from the ent stressorientations and magnitudes were obtained from relatively sparse data coverage that the state of intraplate two measurementsin mines <3 km apart at nearly the same stressis not uniform despitethe relativetectonic quiescence of depth. It is unclear whether thesemethods are, at times, unre- that region.The available data appear to defineseveral major liable because of perturbations in the stressfield due to min- areas of consistentprincipal stressorientation. ing operationsor becausethe mines are typically in areas of Within the stressprovinces defined in this studythe orienta- complex geologicstructure and history. tion of the least principal horizontal stressfield is generally Hydraulic fracturing seemsto be a reliable stressmeasure- uniform (to +_---15ø, within the estimatedaccuracy of the dif- ment techniquewhen done at depthssufficient to avoid the ef- ferent methodsused to determine stressorientations). In the fects of topography [Haimson, 1979] and near-surfacefractur- Midcontin.entstress province, the orientationand possibly ing [M.D. Zoback et al., 1980b].Care must also be taken with also the relative magnitude of the principal stressesappear this techniqueto insurethat a preexistingfracture or joint was uniform over a broad region with linear dimensionsup to not opened rather than a hydraulic fracture at the well bore. 2000 kin. The smallestareas characterized by a distinctstress The consistent11 setsof hydrofracmeasurements in western field are in the western United States. Both the Rio Grande Texas [Zemanek et al., 1970] demonstratethe usefulnessof the Rift and the Hebgen Lake-Centennial Valley stressprovinces hydrofrac techniqueto define regional stresses. are •-150-200 km wide and exhibit differentlyoriented stress Despitethe fairly large uncertaintiesinherent in usingfocal fieldsand different stylesof tectonismrelative to surrounding plane mechanismsfor determiningstress orientations, a gen- areas. eraBy good agreementexists between orientationsdetermined In addition, broad areas up to 1000 km wide that cross from focal mechanismsand other techniques.Focal mecha- stressprovince boundariesare characterizedby a relatively nism data are probablymost reliable where earthquakesoccur uniform least principal stressdirection. In thoseareas the dif- on fault planesof varying trend in a given area, so that the av- ferent stressprovinces are distinguishedon the basisof the rel- erage of P or T directionscan be used [cf. M. L. Zoback and ative magnitudesof the principal stressesand the stylesof de- M.D. Zoback, 1980].Where smallearthquakes occur in areas formation. The most obvious example of this distinction is in of sparsenetwork coverage,large errors are possible.When- the westernUnited States,where the least principal stressdi- ever possible, the use of surface waves to constrain focal rection in the San Andreas, Sierra Nevada, and Basin and mechanismsseems to be advantageous;an earthquakein the Range-Rio Grande rift provincesis fairly uniformly west- Attica, New York area (NY-3) is a good example. The stress northwest. If this distinction is broadened to include areas direction resulting from the surfacewave mechanismof Herr- with similarly oriented principal stressaxes (NNE, WNW, mann [1979] is more consistentwith nearby stressmeasure- and vertical) but with different relative magnitudesof hori- ments than that constrained by body waves alone [Fletcher zontal stresses,then both the Colorado Plateau interior and and Sykes, 1977]. the southern Great Plains can also be included. Principal / / / / / 6144 ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUS UNITED STATES

TABLE 2. Summary of StressProvinces

Primary Least Mode(s) of Principal Faulting Horizontal (Secondary Stress StressProvince Mode)* Direction Comments Pacific Northwest T + SS E-W both thrust and strike slip focal mechanisms;late Tertiary to Quaternary folding along E-axesin easternpart of province San Andreas SS (T) E-W to WNW predominantly strike slip deformation exceptin Big Bendregion, where a large componentof thrust and reverse faulting exist;local normal faulting betweenright-stepping en echelonoffsets of fault Sierra Nevada N + SS WNW to NW region of stresstransition from strike slip deformation in San Andreasto extensionin Basin and Range province Basinand Range-Rio N (SS) WNW, locally regionof activecrustal spreading by Grande rift E-W and distributed normal and oblique slip NW normal faulting; westermostpart of provinceexhibits both purely strikeslip and purely normal faulting Colorado Plateau interior SS + T NNE very low level of tectonicactivity, near-hydrostaticstress (?) Snake River plain- N NE to ENE smallestdistinct stress province, Yellowstone characterizedby a leastprincipal stress directionparallel, not perpendicular, to axis of plain HegbenLake-Centennial N N small anomalousregion of N-S extension; Valley abrupt stresstransition (over <100 km) to Snake River plain at Yellowstone Northern Rocky N + SS NE to E-W abundant late Tertiary normal faulting; Mountains moderate level of modem seismic activity Southern Great Plains N NNE very uniform stressorientation over broad region (800-12007.kin); abrupt 90ø changein leastprincipal stressto Rio Grande rift (<50 kin) Midcontinent T + SS NW stable interior of the United States; broad region of uniform NE-SW compressivestress field; few earthquakesin region showcomponents of thrust and strike slip motion Atlantic Coast T NE compressiongenerally perpendicular to continentalmargin and axisof Appala- chian fold belt; faulting in Charleston, S.C., area may be more complex Gulf Coast N WNW to N activelisttic normal faultingresulting from sedimentloading defines this province; faults strike subparallelto continentalmargin *N, normal faulting; SS, strike slip faulting;and T, thrust faulting. stressaxes in the Pacific Northwest, however, appear some- There appear to be at least two ways in which the stress what oblique to thosein areasto the south. transition can occur: (1) by actual rotation of the stressfield, Available data on the transitions in orientation between dif- as reported in the Hebgen Lake-Centennial Valley to Snake ferent stressprovinces indicate that these transitionscan be River plain-Yellowstone transition [Freidline et al., 1976; R. abrupt and occur over <75 km, as in the Rio Grande rift and B. Smith et al., 1977;Pitt et al., 1979]or (2) by a changein the the southern Great Plains areas. Data from around the Colo- relative magnitudesof the principal stresses,as seemsto occur rado Plateau marginssuggest the existenceof a zone up to 150 in the Basin and Range-Rio Grande rift to Colorado Plateau km wide in which the two horizontal principal stressesare ap- interior and the Rio Grande rift to southern Great Plains proximatelyequal in magnitude,and the directionof the least transitions. principal horizontalstress alternates between the two. Locally, In contrastto these abrupt stresstransitions in the tectoni- in central Arizona the occurrence of normal and reverse focal cally active western United States, in the eastern United mechanismsfor similarily oriented fault planes suggeststhat Statesa broad zone (~200 km) of stresstransition may sepa- all three principal stressesare approximatelyequal in magni- rate the region of predominantly NE-SW compressionalong tude. the Atlantic Coast from the predominantlyNE-SW compres- ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6145

sion characterizingthe Midcontinent region. While included oriented stressaxes in the Puget Sound area, Washington, to a here as part of the Midcontinent stressprovince, the Blue more rapid increasewith depth of the vertical stressrelative to Ridge and Valley and Ridge provincesare characterizedby the horizontal stresses. scatteredand locally contradictorystress orientations (Plates 1 and 2). The scatter may reflect the inherent uncertaintiesof Comparison[rffith Heat Flow Data the different methods used to determine stressor may repre- Figure 6 showsthe map of least principal horizontal stress sent a broad zone of transition in which the horizontal stresses directionsoverlying the heat flow map of the United Statesby are approximatelyequal in magnitudeand the apparentstress Lachenbruchand Sass [1977]. The map illustrates that the field is more controlled by local inhomogeneitieswithin the general level of heat flow is much higher in the West than in crust. the East, that in the West the pattern of heat flow is much The relative magnitudesof the principal stresseswere con- more complex and marked by large regional variations (the strainedon the basisof information from the measuredmag- areasof both highestand lowestheat flow in the United States nitudesof in situ stressand from the current style of deforma- are in the West), and that abrupt transitionssometimes occur tion within each region. In the westernUnited Statesthe high between the major heat flow provinces(cf. Lachenbruchand level of seismicityand broad zone of active faulting indicate Sass [1977, 1978] and Blackwell [1978] for discussionof the generallylarge stressdifferences. In the easternUnited States heat flow provincesin the western United States). As men- the seismicity is more locally concentrated, and whether the tioned above, similar features (a complex pattern, major re- regions of high seismicityoccur within localized zones of gional variations,and abrupt transitions)also characterize the weakness or result from some mechanism producing local modern stressfield in the westernUnited States.In particular, stress concentrations is unknown. a generally good correlation exists between the stressdirec- Two main styles of deformation can be broadly distin- tions and heat flow data in the actively extending Basin and guishedon the basisof the relative magnitudesof the princi- Range-RioGrande Rift provinceand alongits margins.This pal stresses:(1) a predominantlyextensional mode in which broad area of crustal rifting is characterizedby a mean heat the least principal stressis near horizontal and remains in- flow of 88 mW/m 2 (2.1 HFU), well above the continental variant and (2) a predominantly compressionalmode in meanof 63 mW/m 2 (1.5 HFU) [Lachenbruchand Sass,1978]. which the greatestprincipal stressis horizontal and remains A fairly good correlation exists between regions of high invariant. One end-member of the extensionalmode is pure heat flow and the lateral extent of crustal extension, most no- normal faulting, whereas pure thrust faulting is the corre- tably along the Colorado Plateau margins, where the high spondingend-member of the compressionalmode. Pure strike heat flow and associatedrecent faulting and Quaternary vol- slip faulting, such as occursalong much of the San Andreas canism extend well inward of the plateau physiographic fault (with the exceptionof the Big Bend area), representsthe boundary [Thompsonand Zoback, 1979].An area of the Colo- other end-member in both modes. rado Plateau interior of relatively uniform average heat flow The available stressdata often indicate a mixed style of (63-67 mW/m :, 1.5-1.6 HFU) can be definedthat resembles faulting within a given stressprovince (Plate 2, Table 2). Re- the Colorado Plateau interior defined on the basis of stress gions characterized by extensional tectonics (normal and data [Reiter et el., 1979; Thompsonand Zoback, 1979]. The strike slip faulting) are found in the westernUnited Statesand correlation breaks down along the Rio Grande rift-southern include: the Sierra Nevada, the Basin and Range-Rio Grande Great Plains boundary in northern New Mexico, where a 90 ø rift province,the northern Rocky Mountains,the Snake River change in orientation stresstransition is well controlled by plain, and local regions along the San Andreas fault near data from Pliocene to Quaternary volcanic fields in both the right-stepping en echelon offsets. Normal, growth faulting Rio Grande rift and the southern Great Plains. As might be within the sedimentarysection in the Gulf Coastal Plain prob- expectedfrom the recent volcanismon the Great Plains, the ably resultsfrom a purely extensionalstress regime. Regions regionof high heat flow extendsthrough that area, and the ac- of compressionaltectonics (strike slip and thrust faulting) are tual heat flow transition lies to the east. However, the avail- located in the eastern and central as well as the western able stressdata suggestan extensional stressregime in the United Statesand include the PacificNorthwest, the Big Bend southern Great Plains with --•NNE-SSW direction of exten- area of the San Andreas, the Colorado Plateau interior, the sion, as opposedto the much more active WNW-ESE exten- Atlantic Coast province, and the Midcontinent provinces. sion in the Rio Grande rift. The occurrenceof a mixed style of faulting within a stress New data on heat flow in the Mojave block (San Andreas province with consistentlyoriented stressaxes can be inter- stressprovince) are quite uniform and reveal a mean heat pretedin severalways. Two of theprincipal stresses may be flow of 67mW/m: (1.6 HFU) [Lachenbruchet el., 1978].The approximatelyequal in magnitudeand minorregional varia- heat flow rises sharply to the east along a north-northwest tionsin their relative valuesmight determinethe style of fault- trending boundary that coincideswith the easternlimit of ac- ing; this is suggestedby earthquakesalong much of the west- tive seismicityand a changefrom predominantlystrike slip to ern margin of the northern Basin and Range [e.g., Hamilton normal faulting within the Basin and Range province. and Healy, 1969]. Alternately, while comparisonof shallow Lachenbruch and Sass [1978] have developed thermo- (geologic data and in situ stressmeasurements) and deep mechanical models compatible with observedextension rates (focal mechanisms)stress indicators within a stressprovince to explain the high surfaceheat flow in the Basin and Range generallysuggests that the orientationsof the principal stress province. These models, which require convection (either axes may remain unchangedthroughout the upper crust, it is solid state or by magmatic intrusion) in the crust and up- possiblethat the relative magnitudesof the principal stresses permost mantle, are consistentwith gravity data which re- may vary markedly with depth. Crosson[1972] attributed the quire that an influx of mass must accompanythe horizontal occurrenceof shallow thrust events,deeper strike slip events, extensionin this province [Thompsonand Burke, 1974]. Such and very deep (_>60-kmdepth) normal faults with similarily shallow level thermal sourcesare consistentwith abrupt heat 6146 ZOBACKAND ZOBACK'. STATE OF STRESSIN CONTERMINOUSUNITED STATES

/ 4- / 1 /

ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6!49

flow transitionsalong the margins of the Basin and Range- Much of the complex pattern of stressin the western Unitecl Rio Grande rift province. The Sierra Nevada to Basin and Statescan be attributed to presenttransform motion between Range heat flow transitionhas beenfound to occurover <20 the Pacific and North American plates and residual effectsof km in places[Sass et aL, 1971;A. H. Lachenbruch,oral com- past subductionalong the westernedge of the North Ameri- munication, 1979]. Although data on the eastern Basin and can plate. Dynamic effectsrelated to the gradual cessationof Range-Colorado Plateau heat flow transition are sparse,the subductionand developmentof transform faulting may have available data limit this transitionto <75 km [Thompsonand resultedin the anomalouslyhot upper mantle underlying much Zoback, 1979]. The Colorado Plateau-Rio Grande rift heat of the westernUnited States.This hot upper mantle replaced flowtransition isalso abrup[ and occurs over <•50 km [Reiter the volume of mantle previously occupied by a subducted et aL, 1975; Thompsonand Zoback, 1979]. lithosphericslab [e.g., Dickinsonand Snyder, 1979]. The mod- These abrupt heat flow transitions are consistentwith the em state of stressin the Basin and Range province is appar- abrupt stresstransition found along the margins of actively ently largely the result of superpositionof dextral shearon a extending regionsand indicate shallow sources(crust or up- back arc extensional stress field initiated in Miocene time. permostmantle) for both the extensionalstresses and the heat. Anomalous patterns in the western United States, notably The excellentcorrelation in both the pattern of the heat flow WNW-ESE compressionwithin the Colorado Plateau interior and stressdata and the abrupt transitionsbetween provinces and NNE-SSW extension within the southern Great Plains suggeststhat the stressesrelated to rifting are intimately area, may representcomplex second-ordereffects related to linked to the thermal processes. the active kinematics of the surroundingregions. Regional variations in heat flow with the eastern United The observed pattern of stressin the central and eastern Statesare generally small. These variationsare believedto be United Statescan be used to constrainproposed mechanisms nontectonic in origin and to arise from variations in radio- of intraplate stress.The roughly NE-SW compressionin the active heat generation[Birch et aL, 1968;Diment et aL, 1972]. Midcontinent stressprovince is generally consistentwith ei- However, one thermal anomaly of possibletectonic origin in ther ridge push or asthenosphericviscous drag resistanceto the East has been identified by Swanberget al. [1979] in the plate motion. However, drag resistancedirections correlate New Madrid (Northern Mississippi Embayment) on the better with the stressdata than do ridge push directions,sug- basis of slightly elevated heat flow values and bottom hole gestingthat drag may be the primary sourceof stressin the temperature measurements.The absolute values of the heat Midcontinent province. Calculated magnitudesof shear stress flow (55-67 mW/m:, 1.3-1.6 HFU) are not, however,suffi- at the baseof the lithospherederived from a drag mechanism ciently above the midcontinentalmean to qualify as a major are typically only a few bars [e.g., Chappeland Tullis, 1977]; heat flow anomaly. Swanberget al. favored the interpretation however, accumulating evidence from xenolith studies in- of a small convectiveheat flow component, resulting from dicatesthat stressesin the upper mantle may be of the order deep groundwatercirculation along upper crustalfractures as- of a hundred bars [Mercier, 1980]. Also, a poorly developed sociatedwith the active faulting in that region to explain the low-velocity zone [Herrin, 1972] and a thick cold lithosphere local thermal anomaly. Although silica geothermometrydata under cratonic regionsmay explain why viscousdrag resis- suggesta possiblethermal anomaly in the Charleston, South tance may be an important mechanismfor slowmoving conti- Carolina, area [Swanbergand Morgan, 1978], heat flow mea- nents. sured in a 790-m-deep well near the epicenter of the 1866 Asthenospheric counterflow models, whose effects have Charlestonearthquake was 1.3 HFU (Sassand Ziagos, 1977], been suggestedto be 2-3 timesthat of simpledrag resistance similar to measurementsmade in surroundingareas. to plate motion [cf. Chase, 1979; Hager and O'Connell, 1979], do not apply;predicted stress orientations are nearly 90ø from CONCLUDING REMARKS thoseobserved in the Midcontinent province.Perhaps major Using a variety of techniquesto determine principal stress variations in lithospheric thicknessact to 'channel' this coun- orientations,regional patterns of the in situ stressfield of the terflow and present models which assumea worldwide uni- conterminous United States have been defined. Within a formly thick lithosphereare too simplistic. given region, stressorientations inferred from geologicobser- Sourcesof stressin the Atlantic Coast region remain unre- vations, earthquake sourcemechanisms, and direct measure- solved. Calculated ridge push and drag resistancedirections ment of in situ stressare generallyconsistent (within the accu- are 300-80 ø oblique to inferred stressorientations and in- racy ot the techniques)despite the different depth intervals dicate that if ridge pushis an importantforce providingcom- sampled.This implies a relatively uniform upper crustalstress pressionacross the Atlantic Coast region, additional effects field which can be investigatedby making measurementsat must be superimposedto explain the observed direction of relatively shallow depths. Stress provinces with consistent maximum horizontal compression.The orientation of this principal stressorientations and relative magnitudescan be compression,approximately perpendicularto both the conti- delineated;these stress provinces have linear dimensionsthat nental margin and the Appalachian Fold Belt, suggestsa range from 100 to 2000 km. causal relationship. Lateral density contrastsassociated with With the exception of normal 'growth' faulting within the crustalstructure at a passivecontinental margin predict exten- sedimentarysection on the Gulf Coastal Plain the central and sion perpendicularto the continentalmargin rather than the easternUnited Statesare dominatedby compressionaltecton- observed compression.Lithospheric flexure resulting from ism, generally NW-SE compressionalong the Atlantic Coast sediment loading also predicts extension on the continental and •NE-SW compressionwithin the midcontinent area. margin. Stressesderived from major lithosphericthickening Whereas broad regionsof the westernUnited Statesare char- alongthe edgeof the continentalso appear unlikely; presently acterizedby extensionaland strikeslip styletectonism, the Pa- available seismicdata suggestlittle contrast in thickness be- cific Northwest, the Colorado Plateau interior, and the Big tween the coastalplain lithosphereand the adjacent 180-190 Bend area along the San Andreas fault are dominated by m.y. old oceaniclithosphere. Lithospheric structure under the compressionaldeformation. Appalachian Fold Belt, however,remains unexamined. Postu- 6150 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

lated southeastwarddirected gravity backslidingof the Ap- REFERENCES palachiansalong a major detachment,while capableof ex- Abou-Sayed, A. S., C. E. Brechtel,and R. J. Clifton, In situ stressde- plaining the observedorientation of compressivestress within terminationby hydrofracturing:A fracturemechanics approach, J. the Atlantic Coast province, is inconsistentwith evidenceof Geophys.Res., 83(B6), 2851-2862, 1978. compressionaltectonism (Recent thrust faulting) in the main Aggarwal, Y. P., and L. R. Sykes,Earthquakes, faults, and nuclear power plants in southernNew York, Science,200, 425-429, 1978. elevated core of the Appalachians.Rotation of stress(or Aggarwal,Y. P., J.P. Yang, and E. Cranswick,Seismological investi- strain) by anisotropicbasement structure and flexural effects gationin the Adirondacksand environs(abstract), Geol. $oc. Amer. in responseto erosionand deglaciationare suggestedas pos- Abstr. Program, 9, 234, 1977. sible influences on the stress field. Aggson,J., Report on in situ determination of stresses,Mather Mine, Ishpenning, Michigan, Prog. Rep. DMRC 10006, U.S. Bur. of Transitionsbetween stressprovinces also help constrain Mines, Denver, Colo., 1972. sourcesof stress.The abrupt transitionsobserved in the orien- Aggson, J. R., and V. E. Hooker, In-situ rock stressdetermination: tation or magnitude of in situ stressesin the western United Techniques and applications,in UndergroundMining Handbook, Statesimply a shallow(crustal or uppermostmantle) source Societyof Mining Engineersof AIME, New York, in press,1980. for thesestresses. These abrupt transitionsare found only in Akers, J.P., J. C. Shorty, and P. R. Stevens,Hydrogeology of the Ce- nozoic igneous rocks, Navajo and Hopi Indian reservations,Ari- areasof active extensionaltectonics (Basin and Range, Rio zona, New Mexico, and Utah, U.S. Geol. $urv. Prof Pap., 521-D, Granderift, SnakeRiver plain). The accompanyinghigh heat 18 pp., 1971. flow in theseareas also shows abrupt transitions and thus sim- Anderson, C. A., E. A. Scholz, and J. D. Strobell, Jr., Geology and ilarly impliessources in the crustor uppermostmantle. The ore depositsof the Bagdad area, Yavapai County, Arizona, U.S. excellent correlation of heat flow with stress orientations in Geol. Surv. Prof. Pap., 278, 103 pp., 1955. Anderson, E. M., The Dynamicsof Faulting and Dyke Formation With theseregions indicates that the stressesrelated to rifting are Applicationsto Britain, 2nd ed., 206 pp., Oliver and Boyd, Edin- intimately linked to the thermal processes.In the relatively burgh, 1951. quiescenteastern United Statesa possiblybroad region of Angelier, J., Determination of the mean principal directions of stresstransition suggestsstresses derived from 'broad-deep stressesfor a given fault population, Tectonophysics,56, T17-26, 1979. sources.Both asthenosphericdrag resistanceto plate motion Arabasz,W. J., R. B. Smith, and W. B. Richins,Earthquake studies and transmitted ridge push reflect such sources.'In terms of along the Wasatch Front, Utah: Network monitoring, seismicity, broad-deepsources of stressthe apparent complexityof the and seismichazards, in EarthquakeStudies in Utah, 1850 to 1978, stressfield in the transitionzone is not unexpected;it may be edited by W. J. Arabasz,R. B. Smith, and W. D. Richins,pp. 253- controlledin part by local preexistingstructure. 286, University of Utah, Salt Lake City, 1979. Armstrong,R. L., W. P. Leeman,and H. E. Malde, K-Ar dating, A final noteshould be addedon the time scaleof changesof Quaternaryand Neogenevolcanic rocks of the SnakeRiver plain, regional stressfields. As we appeal to broad scale plate tec- Idaho, Amer. J. Sci., 275, 225-251, 1975. tonic forcesto explain much of the regionalstress patterns, it Artyushkov,E. V., Stressesin the lithospherecaused by crustalthick- is likely that stressfields derived from theseforces could per- nessinhomogeneities, J. Geophys.Res., 78, 7675-7708, 1973. Atwater, T., Implications of plate tectonicsfor the Cenozoic tectonic sistfor a period of time similar to the lengthsof time between evolution of western North America, Geol. $oc. Amer. Bull., 81, major plate reorganizations.In the Atlantic Coastregion the 3513-3535, 1970. consistentstyle of observedfault offsetsalong this passive Atwater, T., and P. Molnar, Relative motion of the Pacific and North marginsuggests a similarstress field (both orientationand rel- American platesdeduced from sea-floorspreading in the Atlantic, ative magnitude)for the past --110 m.y. [D.C. Prowell, writ- Indian, and SouthPacific oceans, Proceedings of the Conferenceon Tectonic Problems of the San Andreas Fault System, Stanford ten communication,1980; Wentworthand Mergner-Keefer, Univ. Publ. Geol. $ci., 13, 136-148, 1973. 1980].In areasof activetectonism adjacent to plate bounda- B'abcock,E. A., Measurementof subsurfacefractures from dipmeter ries, suchas the westernUnited States,the stressfield may logs, Amer. Ass. Petrol. Geol. Bull., 62, 1111-1126, 1978. changeas often aschanges in relativemotions or plate bound- Babenroth,D. L., and A. N. Strahler,Geomorphology and structure aries occur. In the Basin and Range province,evidence exists of the East Kaibab monocline, Arizona and Utah, Geol. Soc. Amer. Bull., 56, 107-150, 1945. of-45 ø clockwisechange in the leastprincipal stressorienta- Bache,T. C., andD. G. Lambe.rt,Earthquake ground motion from tion since middle Miocene time [M. L. Zoback and G. A. the 1975 PocatelloValley earthquake,Rep. $$$-R-78-3507, 43 pp., Thompson,1978]; detailed data in severalareas suggestthat Syst., Sci. and Software, La Jolla, Calif., 1977. this changeoccurred between -14 and 7 m.y.B.P. The gen- Bailey, J.P., Seismicityand contemporarytectonics of the Hegben Lake-Centennial Valley, Montana area, M.S. thesis,Univ. of Utah, eral trend of the largely post-10 m.y. rangesin the northern Salt Lake City, 1976. Basin and Range, roughly perpendicularto the modern least Baldwin, B., and W. R. Muehlberger, Geologic studiesof Union principal stressdirection, suggests that this changemay have County, New Mexico, N. Mex. Bur. Mines Miner. Resour.Bull., 63, occurredrapidly (>4 m.y.?). On a similar time scale,tecton- 171 pp., 1959. ism in the ColumbiaPlateau region (within the PacificNorth- Balk, R., Geologic map and sectionsof Tres Hermanas Mountains, west) changedfrom a major episodeof-E-W extensioncoin- Geol.Map 16, N. Mex. Bur. of Mines Miner. Resour.,Socorro, 1962. Bateman,P. C., Willard D. Johnsonand the strike-slipcomponent of oiding with eruption of the Columbia River flood basalts 14- fault movementin the Owens Valley, California, earthquakeof 17 m.y.B.P. to Plioceneto Quaternary tectonismdominated 1872, Bull. Seismol.Soc. Amer., 51, 483-493, 1971. by N-S compression. Bell, J. S., and D. I. Gough, Northeast-southwestcompressive stress Acknowledgments. We are indebted to the numerous workers in Alberta: Evidence from oil wells, Earth Planet. $ci. Lett., 45, whoseefforts over the yearshave producedthe data usedin this com- 475-482, 1979. pilation.The manuscriptwas greatly improved by criticalreviews and Behrendt,J.C., R. M. Hamilton,it. D. Ackerman,V. J. Henry, and comments by Art McGarr, Randy Richardson, Bill Brace, Hans K. C. Bayer, Cenozoic reactivationof faulting in the vicinity of Swolfs,George Thompson, Bob Smith,Russ Wheller, Marc Sbar,and Charleston,S.C., earthquakezone, submittedto Geology,1980. Art Tarr. We thank John Sassfor allowing us to use the heat flow Best, M. G., and W. K. Hamblin, Origin of the northern Basin and map, Bob Smithfor the map of westernUnited Statesseismicity, and Range province: Implications from the geology of its eastern James E. York and Jack Oliver for the earthem United States seismic- boundary, Cenozoic Tectonicsand Regional Geophysicsof the ity map. Western Cordillera, Geol. Soc. Amer. Mem., 152, 313-340, 1978. ZOBACK AND ZOBACK: STATE OF STRESSIN CONTERMINOUS UNITED STATES 6151

Bickel, D. L., and D. R. Dolinar, In situ determination of stresses, southernAppalachians; COCORP seismic-reflectionprofiling of LakeshoreMine, Casa Grande, Arizona, Prog. Rep., DMRC 10018, the Blue Ridge and Piedmont,Geology, 7, 563-567, 1979. U.S. Bur. of Mines, Denver, Colo., 1976. Cooper,J. B., and E. C. John,Geology and groundwater occurrence Birch, F., R. F. Roy, and E. R. Decker, Heat flow and thermal history in southeasternMcKinlay County, New Mexico, N. Mex. State in New England and New York, in Studiesof AppalachianGeology: Eng. Tech.Rep., 35, 108 pp., 1968. Northern and Maritime, edited by E. Zen et al., pp. 437-451, John Copeland,C. W., Faultsin Tertiary rocksof southwesternAlabama, Wiley, New York, 1968. Cretaceousand Tertiary Faultsin SouthwesternAlabama, pp. 16- Biswas,N. N., and L. Knopoff, The structureof the upper mantle un- 57, Guideb.Annu. Field Trip 14th,Ala. Geol.Soc., University, 1976. der the United States from the dispersionof Rayleigh waves, Cordell,L., Regionalgeophysical setting of the Rio Granderift, Geol. Geophys.J. Roy. Astron. Soc., 36, 515-539, 1974. Soc. Amer. Bull., 89, 1073-1090, 1978. Blackwell, D. D., Heat flow and energy loss in the western United Couch,R., G. Thrasher,and K. Keeling,The DeschutesValley earth- States,Cenozoic Tectonicsand Regional Geophysicsof the West- quake of April 12, 1976, Ore Bin, 38, 151-161, 1976. ern Cordillera, Geol. Soc. Amer. Mere., 152, 175-208, 1978. Cox, J. W., The high resolutiondipmeter reveals dip-related borehole Blake, M. C., Jr., R. H. Campbell, T. W. Dibblee, Jr., D. G. Howell, and formation characteristics,paper presentedat the 1l th Annual T. H. Nilsen, W. R. Normark, J. C. Vedder, and E. A. Silver, Neo- LoggingSymposium, Soc. Prof. Well Log Anal., Los Angeles, gene basin formation in relation to plate-tectonicevolution of San Calif., 1970. Andreasfault system,California, Amer. Ass.Petrol. Geol. Bull., 62, Crosson,R. S., Small earthquakes,structure, and tectonicsof the Pu- 344-372, 1978. get Sound region (Washington),Bull. Seismol.Soc. Amer., 62, Block, J. W., R. C. Clement, L. R. Lew, and J. de Boer, Recent thrust 1133-1171, 1972. faulting in southeasternConnecticut, Geology, 7, 79-82, 1979. Crosson,R. S., andJ.-W. Lin, A noteon the Mt. Rainierearthquake Bollinger,G. A., A microearthquakesurvey of the central Virginia of April 20, 1974,Bull. Seismol.Soc. Amer., 65, 549-552, 1975. seismiczone, Earthquake Notes, 46, 3-13, 1975. Damesand Moore, Summaryreport on the in-progressseismic mon- Bollinger, G. A., Seismicityof the southeasternUnited States, Bull. itoringprogram at the North Anna Site,January 21, 1974through Seismol. Soc. Amer., 63, 1785-1808, 1973. May 1, 1976,submitted to the Nuclear RegulatoryCommission, Bolt, B., C. Lomnitz, and T. V. McEvilly, Seismologicalevidence on Washington, D.C., 1976. the tectonics of central and northern California and the Mendocino Daniels,R. B., and E. E. Gamble,Atlantic Coastal Plains Geological escarpments,Bull. Seismol. Soc. Amer., 58, 1725-1767, 1968. AssociationField Trip Guidebook, vol. 13,36 pp., CarolinaGeologi- Bott, M. H. P., and D. S. Dean, Stresssystems at young continental cal Society,Raleigh, N. C., 1972. margins, Nature, 235, 23-25, 1972. Davis, G. A., Tectonic evolution of the Pacific Northwest: Pre- Bredehoeft,J. D., R. G. Wolff, W. S. Keys, and E. Shuter,Hydraulic cambrianto present,Subappendix 2R C of preliminarysafety anal- fracturingto determinethe regionalin situ stressfield, PiceanceBa- ysis report, Amendment23, Nuclear ProjectNo. 1, p. i-2Rc-46, sin, Colorado, Geol. Soc. Amer. Bull., 87, 250-258, 1976. Wash. Public Power Supply Syst., 1977. Bridwell, R. J., Lithosphericthinning and the late Cenozoicthermal Davis, G. A., Problemsof intraplateextensional tectonics, western and tectonicregime of the northern Rio Grande Rift, N. Mex. Geol. United States,with specialemphasis on the Great Basin, in 1979 Soc. Field Conf. Cruideb.,27, 283-292, 1976. Basinand RangeSymposium, edited by G. W. Newman and H. D. Brown, L. D., P. A. Krumhansl, C. E. Chapin, A. R. Sanford, F. A. Goode,pp. 41-54, RockyMountain Association of Geologistsand Cook, S. Kaufman, J. E. Oliver, and F. S. Schilt, COCORP seismic Utah Geological Association,Denver, Colo., 1979. reflectionstudies of the Rio Grande Rift, in Rio GrandeRiff: Tec- Davis, G. A., and B.C. Burchfiel, Garlock fault: An intracontinental tonicsand Magmatism,edited by R. E. Riecker, pp. 169-184, AGU, transformstructure, southern California, Geol. Soc.Amer. Bull, 84, Washington, D.C., 1979. 1407-1422, 1973. Brumbaugh,D. S., Analysisof the Williams, Arizona earthquakeof Dewey,J. W., W. M. Dillinger, J. Taggart,and S. T. Algermissen,A November 4, 1971, Bull. Seismol.Soc. Amer., 70, 885-891, 1980. techniquefor seismiczoning: Analysis of earthquakelocations and Bufe, C. G., S. M. Marks, F. W. Lester, R. S. Ludwin, and M. C. mechanismsin northernUtah, Wyoming,Idaho, Montana,in Pro- Stickney,Seismicity of the Geysers-ClearLake geothermalarea, ceedingsof the InternationalConference on Microzonationfor Safer California, U.S. Geol. Surv. Prof. Pap., in press,1980. Construction,Research and Application,vol. 2, pp. 879-895, Wash- Byerly, P., The Montana earthquake of June 28, 1925, G.M.C.T., ington University, St. Louis, Mo., 1972. Bull. Seismol. Soc. Amer., 16, 209-263, 1926. Dibblee,T. W., Jr., Arealgeology of thewestern Mojave Desert, Cali- Carey, E., Computation of principal stressaxes associatedwith ob- fornia, U.S. Geol.Surv. Prof Pap., 522, 153 pp., 1967. servedmotions on fault planes,Rev. Geol.Dyn. Geogr.Phys., 21, Dickinson,W. R., and W. S. Snyder,Geometry of subductedslabs re- 57-66, 1979. lated to San Andreas transform, J. Geol., 87, 609-627, 1979. Carr, W. J., Summary of tectonic and structural evidence for stress orientationat the NevadaTest Site,Open File Rep. 74-176,53 pp., Diment, W. H., T. C. Urban, and F. A. Revetta,Some geophysical U.S. Geol. Surv., Reston, Va., 1974. anomaliesin the easternUnited States,in The Nature of the Solid Chan, S.S. M., and T. J. Crocker,A casestudy of in situ rock defor- Earth, edited by F. C. Robertson,pp. 544-572, McGraw-Hill, New York, 1972. mation behavior in the Silver Summit Mine, Coeur D'Alene mining district, in Proceedingsof the 7th Canadian Rock Mechanics Donnelly, M. F., Geologyof the Sierra del Pinacatevolcanic field, Symposium,Edmonton, 1971, pp. 135-160,Mines Branch,Depart- northern Sonora, Mexico, and southernArizona, U.S.A., Ph.D. ment of Energy, Mines and Resources,Ottawa, 1972. thesis,722 pp., Stanford Univ., Stanford, Calif., 1974. Chapin, C. E., and W. R. Seager,Evolution of the Rio Grande rift in DuffSeld,W. A., Late Cenozoicring faultingand volcanismin the the Socorroand LasCruces area, in N. Mex. Geol.Soc. Field Conf CosoRange area of California, Geology,3, 335-338, 1975. Guideb., 26, 297-321, 1975. Dumas, D. B., H. J. Dorman, and G. V. Latham, A reevaluation of Chapman,D. S., and H. N. Pollock,Regional geotherms and litho- the August 16, 1931 Texas earthquake,Bull. Seismol.Soc. Amer., sphericthickness, Geology, 5, 265-268, t977. 70, 1171-1180, 1980. Chapple,W. M., and T. E. Tullis, Evaluationof the forcesthat drive Eaton,G. P., A platetectonic model for late Cenozoiccrustal spread- the plates,J. Geophys.Res., 82(14), 1967-1984,1977. ing in the westernUnited States,in Rio GrandeRiff: Tectonicsand Chase, C. G., Asthenosphericcounterflow: A kinematic model, Magmatism,edited by R. E. Riecker,pp. 7-32, AGU, Washington, Geophys.J. Roy. Astron. Soc., 56, 1-18, 1979. D.C., 1979. Christiansen,R. L., and E. H. McKee, Late Cenozoic volcanic and Eaton, G. P., Geophysicaland geologicalcharacteristics of the crust tectonic evolution of the Great Basin and Columbia Intermountain of the Basinand Rangeprovince, in ContinentalTectonics• National regions,Cenozoic Tectonics and RegionalGeophysics of the West- ResearchCouncil Studies in Geophysics,edited by B.C. Burchfiel,L. ern Cordillera, Geol. Soc.Amer. Mem., 152, 283-312, 1978. T. Silver,and J. E. Oliver, pp. 96-113, National Academyof Sci- Colton, H. S., The basaltic cinder cones and lava flows of the San ences,Washington, D.C., 1980. FranciscoMountain volcanic field, report,58 pp., Mus. of N. Ariz., Eaton,G. P., R. L. Christiansen,H. M. Iyer, A.M. Pitt, D. R. Mabey, Flagstaff, Ariz., 1967. H. R. Blank,Jr., I. Zietz, andM. E. Gettings,Magma beneath Yel- Cook, F. A., D. S. Albaugh, L. D. Brown, S. Kaufman, J. E. Oliver, lowstoneNational Park, Science,188, 787-796, 1975. and R. D. Hatcher, Jr., Thin-skinnedtectonics in the crystalline Eaton, G. P., R. R. Wahl, H. J. Prostka,D. R. Mabey, and M.D. 6152 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Kleinhopf, Regional gravity and tectonicpatterns: Their relation to Monogr. Ser., vol. 20, edited by J. C. Heacock, pp. 576-592, AGU, late Cenozoicepeirogeny and lateral spreadingin the westernCor- Washington, D.C., 1977b. dillera, CenozoicTectonics and Regional Geophysicsof the West- Haimson, B.C., Near-surfaceand deeperhydrofracturing stress mea- ern Cordillera, Geol. Soc. Amer. Mere., 152, 51-91, 1978. surementsin the Waterloo quartzite (abstract),Eos Trans.AGU, 59, Eberhart-Phillips, D. M., M. L. Sbar, R. M. Richardson, and R. B. 327-328, 1978a. Herrmann, Analysisof the 4 February 1976Chino Valley, Arizona Haimson,B.C., Additional stressmeasurements in the Michigan Ba- earthquake (abstract), Eos Trans. AGU, 60, 894, 1979. sin (abstract),Eos Trans.AGU, 59(12), 1209, 1978b. Eberly, L. D., and T. B. Stanley, Jr., Cenozoicstratigraphy and geo- Haimson, B.C., New hydrofracturingmeasurements in the Sierra Ne- logic history of southwesternArizona, Geol. Soc. Arner. Bull., 89, vada Mountains and the relationship between shallow stressesand 921-940, 1978. surfacetopography, paper presentedat the 20th U.S. Symposium Ellsworth, W. L., and S. M. Marks, Seismicityof the Livermore Val- on Rock Mechanics, U.S. Natl. Comm. for Rock Mech., Austin, ley, California region, 1969-1979, Open File Rep. 80-515, 42 pp., Tex., 1979. U.S. Geol. Surv., Reston, Va., 1980. Haimson,B. C., and C. Fairhurst,In-situ stressdetermination at great Elston, W. E., and T. J. Bornhorst, The Rio Grande rift in context of depth by means of hydraulic fracturing, in Rock MechanicsmThe- regional post-40 m.y. volcanic and tectonic events,in Rio Grande ory and Practice,Proceedings, 11th Symposiumon Rock Mechanics, Riff: Tectonicsand Magmatism,edited by R. E. Riecker,pp. 416- Berkeley,1'969, chap. 28, edited by W. H. Somerton,pp. 559-584, 438, AGU, Washington, D.C., 1979. Societyof Mining Engineersof AIME, New York, 1970. Fenneman, N.M., Physical divisions of the United States, map, Haimson,B.C., and C. F. Lee, Stressmeasurements in underground 1:7,000,000 scale, U.S. Geol. Surv., Reston, Va., 1946. nuclear slant design,in Proceedingsof 1979 Rapid Excavationand Finnell, T. L., Geologicmap of the ChediskiPeak quadrangle,Nav- TunnelingConference, Atlanta, Georgia,Society of Mining Engi- ajo County, Arizona, Quad. Map GQ-544, U.S. Geol. Surv., Res- neers of AIME, Littleton, Colo., 1979. ton, Va., 1966. Haimson, B.C., J. Lacomb,A. H. Jones,and S. J. Green, Deep stress Fischer, F. G., P. J. Papanek, and R. M. Hamilton, The Massachu- measurements in tuff at the Nevada Test Site, in Advances in Rock setts Mountain earthquake of 5 August 1971 and its aftershocks, Mechanics,vol. IIa, pp. 557-561, National Academy of Sciences, Nevada Test Site, Rep. USGS-474-149, 16 pp., U.S. Geol. Surv., Washington, D.C., 1974. Reston, Va., 1972. (Available from U.S. Department of Commerce, Hales, A. L., Gravitational sliding and continental drift, Earth Planet. National Technical Information Service,Springfield, Va.) Sci. Lett., 6, 31-34, 1969. Fletcher, J. B., and L. R. Sykes, Earthquakesrelated to hydraulic Hamilton, R. M., Aftershocksof the Borrego Mt. earthquakefrom mining and natural seismicactivity in westernNew York State, J. April 12 to June 12, 1968, The Borrego Mountain Earthquake of Geophys.Res., 82, 3767-3780, 1977. April 9, 1978, U.S. Geol. Surv. Prof. Pap., 787, 31-54, 1972. Frank, F. C., Plate tectonics,the analogy with glacier flow and isos- Hamilton, R. M., and J. H. Healy, Aftershocksof the Benham nuclear tasy, in Flow and Fracture of Rocks, Geophys.Monogr. Ser., vol. 16, explosion, Bull. Seismol.Soc. Amer., 59, 2271-2281, 1969. edited by H. C. Heard et al., pp. 285-292, AGU, Washington, D. Hamilton, W., and W. B. Myers, Cenozoic tectonicsof the western C., 1972. United States,Rev. Geophys.Space Phys., 4, 509-549, 1966. Freidline, R. A., R. B. Smith, and D. D. Blackwell, Seismicity and Harding, T. P., Tectonic significanceand hydrocarbontrapping con- contemporarytectonics of the Helena, Montana area, Bull. Seisrnol. sequencesof sequential folding synchronouswith San Andreas Soc. Amer., 66, 81-95, 1976. faulting, San Joaquin Valley, California, Amer. Ass. Petrol. Geol. Gawthrop, W. H., Seismicity and tectonicsof the central California Bull., 60, 356-378, 1976. coastal region, M.S. thesis,76 pp., Univ. of Colo., Boulder, 1977. Harper, J. F., Asthenosphereflow and plate motions,Geophys. J. Roy. Graham, T., and E. F. Chiburis, Fault plane solutionsand the state of Astron. Soc., 55, 87-110, 1978. stressin New England, Earthquake Notes, 31, in press, 1980. Harris, L. D., and K. C. Bayer, Sequentialdevelopment of the Ap- Griggs,R. L., Geologyand ground-waterresources of the easternpart palachianorogen above a masterdecollement--A hypothesis,Geol- of Colfax County, New Mexico, N. Mex. Bur. Mines Min. Resour. ogy, 7, 568-572, 1979. Ground Water Rep., 1, 180 pp., 1948. Healy, J. H., D. T. Griggs, W. W. Rubey, and C. B. Raleigh, The Grose, L. T., Late Quaternary tectoniccontrols of occurrenceof geo- Denver earthquakes,Science, 161, 1301-1310, 1968. thermal systemsin Gerlach-Hualapai Flat area, northwesternNe- Herrin, E., A comparativestudy of upper mantle models:Canadian vada, Colo. Sch. Mines Quart., 73(3), 11-14, 1978. Shield and Basin and Range province, in The Nature of the Solid Gumper, F. J., and C. Scholz, Microseismicityand tectonicsof Ne- Earth, edited by F. C. Robertson,pp. 216-234, McGraw-Hill, New vada Seismic Zone, Bull. Seismol. Soc. Amer., 61, 1413-1432, 1971. York, 1972. Hack, J. C., Sedimentationand volcanismin the Hopi Buttes, Ari- Herrmann, R. B., Surface wave focal mechanisms for eastern North zona, Geol. Soc. Amer. Bull., 53, 335-372, 1942. American earthquakeswith tectonic implications, J. Geophys.Res., Hack, J. T., Rock control and tectonismmTheirimportance in shap- 84(B7), 3543-3552, 1979. ing the Appalachian Highlands, U.S. Geol. Surv. Prof Pap., 1126- Herrmann, R. B., and J. A. Canas, Focal mechanism studies in the B, 1979. New Madrid seismic zone, Bull. Seismol. Soc. Amer., 68, 1095- Hager, B. H., and R. J. O'Connell, Kinematic modelsof large-scale 1102, 1978. flow in the earth's mantle, J. Geophys.Res., 84, 1031-1048, 1979. Hill, D. P., A model for earthquake swarms,J. Geophys.Res., 82, Haimson, B.C., Earthquakerelated stressesat Rangely, Colorado, in 1347-1352, 1977. New Horizons in Rock Mechanics:Proceedings, 14th Symposiumon Hill, M. L., and T. W. Dibblee, Jr., San Andreas, Garlock, and Big Rock Mechanics,edited by H. R. Hardy, Jr., and R. Stefanko, pp. Pine faults, CaliforniamA study of the character,history, and tec- 689-708, American Society of Civil Engineers,New York, 1973. tonic significanceof their displacements,Geol. Soc. Amer. Bull., 64, Haimson, B. C., A simple method for estimating in situ stressesat 443-458, 1953. great depths, Field Testing and Instru,x•entationof Rock, Spec. Hill, R. L., and D. J. Beeby, Surfacefaulting associatedwith the 5.2 Tech. Publ., 554, pp. 156-182, Amer. Soc. for Test. Mater., Phila- magnitude Galway Lake earthquake of May 31, 1975: Mojave delphia, Pa., 1974a. Desert, San Bernardino County, Calif., Geol. Soc. Amer. Bull., 88, Haimson, B.C., Stress measurements in faults and their vicinities, 1378-1384, 1977. semiannualreport, Grant 14-08-0001-6118,U.S. Geol. Surv., Res- Hoffer, J. M., The Potrillo basalt field, south-central New Mexico, Ce- ton, Va., 1974b. nozoic Volcanism in SouthwesternNew Mexico, Spec.Publ. 5, pp. Haimson, B.C., Preexcavationdeep-hole stressmeasurements for de- 89-92, N. Mex. Geol. Soc., Socorro, 1976. sign of undergroundchambers--Case histories,in Proceedingsof Hooker, V. E., D. L. Bickel, and J. R. Aggson, In situ determination the 1976 Rapid Excavation and TunnelingConference, edited by R. of stressin mountainoustopography, U.S. Bur. Mines Rep. Invest., J. Robins and R. J. Conlon, pp. 699-714, Societyof Mining Engi- 7654, 19 pp., 1972. neers of AIME, New York, 1976. Hooker, V. E., J. R. Aggson,and D. L. Bickel, Improvementsin the Haimson, B. C., A stressmeasurement in West Virginia and the state three-componentborehole deformation gage and overcoringtech- of the stressin southernAppalachians (abstract),Eos Trans. A G U, niques, U.S. Bur. Mines Rep. Invest., 7894, 29 pp., 1974. 58, 493, 1977a. Horner, R. B., A. E. Stevens,H. S. Hasegawa,and G. LeBlanc, The Haimson, B. C., Crustal stressin the continental United States as de- Maniwaki, Quebec,earthquake of July 12, 1975 (abstract),Earth- rived from hydrofracturingtests, in The Earth's Crust, Geophys. quake Notes, 46, 48, 1975. ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6153

Howard, K. A., J. M. Aaron, E. E. Brabb, M. R. Brock, H. D. Gower, Arabasz, R. B. Smith, and W. D. Richins, pp. 383-394, University S. J. Hunt, D. J. Milton, W. R. Muehlberger,J. K. Nakata, G. Plaf- of Utah, Salt Lake City, 1979. ker, D.C. Prowell R. E. Wallace, and I. J. Witkind (Compilers), Langston,C. A., and R. Butler, Focal mechanismof the August 1, Preliminarymap of youngfaults in the United Statesas a guide to 1975, Oroville, earthquake, Bull. Seisrnol.Soc. ,•rner., 66, 1111- possiblefault activity, Misc. Field Stud. Map MF-916, U.S. Geol. 1120, 1976. Surv., Reston, Va., 1978. Leblanc, G., and G. Buchbinder, Secondmicroearthquake survey of Howell, D. E., and A. J. W. Zupan, Evidencefor post-Cretaceoustec- the St. Lawrence Valley near La Malbaie, Quebec, Can. J. Earth tonic activity in the Westfield Creek area north of Cherow, South Sci., 14, 2778-2789, 1977. Carolina, Geol. Notes, •8(4), 98-105, 1974. Leblanc,G., A. E. Stevens,and R. J. Wetmiller, A microearthquake Hubbert, M. K., and D. G. Willis, Mechanicsof hydraulicfracturing, surveyof the St. Lawrence Valley near La Malbaie, Quebec, Can. J. •4IME Trans., 210, 153-168, 1957. Earth Sci., 10, 42-53, 1973. Huff, L. C., E. Santos, and R. G. Raabe, Mineral resourcesof the Leeds, A. R., L. Knopoff, and E.G. Kausel, Variations of upper SycamoreCanyon primitive area, Arizona, U.S. Geol. Surv. Bull., mantle structure under the Pacific Ocean, Science, 186, 141-143,

1230-F, 19 pp., 1966. _ 1975. Hunt, C. B., Igneousgeology and structureof the Mount Taylor vol- Lockwood, J.P., and J. G. Moore, Regional extensionof the Sierra canic field, New Mexico, U.S. Geol. Surv. Prof. Pap., 189-B, 1938. Nevada, California, on conjugatemicrofault sets, J. Geophys.Res., Inden, R. F., and A. J. W. Zupan, Normal faulting of upper coastal 84, 6041-6049, 1979. plain sediments,Ideal Kaolin Mine, Langley,South Carolina, Geol. Luedke,R. G., and R. L. Smith, Map showingdistribution, composi- Notes, 19(4), 160-165, 1975. tion, and age of late Cenozoicvolcanic centers in Arizona and New Jacobeen,F. H., Jr., Seismicevidence for high-anglereverse faulting Mexico, Misc. Invest. Map 1-1091, U.S. Geol. Surv., Reston, Va., in the coastalplain of Prince Georgesand Charles County, Mary- 1978a. land, Inf. Circ. 13, 21 pp., Md. Geol. Surv., Baltimore, 1972. Luedke,R. G., and R. L. Smith, Map showingdistribution, composi- Jahns,R. H., Tectonicevolution of the TransverseRanges province as tion, and age of late Cenozoic volcanic centersin Colorado, Utah related to the San Andreasfault system,Proceedings of the Confer- and southernWyoming, Misc. Invest. Map 1-1091-B, U.S. Geol. ence on Tectonic Problemsof the San Andreas Fault System,Stan- Surv., Reston, Va., 1978b. ford Univ. Publ. Univ. Ser. Geol. Sci., 13, 149-170, 1973. Malone, S. D., G. H. Rothe, and S. W. Smith, Details of micro- Jicha, H. L., Jr., Geology and mineral resourcesof Mesa del Oro earthquakeswarms in the Columbia Basin,Washington, Bull. Seis- quadrangle,Socorro and Valencia counties,New Mexico, N. Mex. rnol. Soc. ,•rner., 65, 855-864, 1975. Bur. Mines Miner. Resour.Bull., 56, 67 pp., 1958. McDowell, F. W., K-At ages of igneous rocks from the western Johnson, C. E., and D. M. Hadley, Tectonic implications of the United States, Isochron/West, 2, 1-16, 1971. Brawley earthquake swarm, Imperial Valley, California, January McEvilly, T. V., Preliminaryseismic data, June-July 1966,section of 1975, Bull. Seismol. Soc. •4mer., 66, 1133-1144, 1976. Parkfieldearthquakes of June27-29, 1966,Monterey and San Luis Johnson,R. B., Patterns and origin of radial dike swarms associated Obispo counties, California--Preliminary report, Bull. Seisrnol. with West Spanish Peak and Dike Mountain, south-centralColo- Soc. ,•rner., 56, 967-971, 1966. rado, Geol. Soc. •4mer. Bull., 72, 579-590, 1961. McGarr, A., and N. C. Gay, State of stressin the earth's crust, ,•nn. Kanamori, H., and G. Fuis, Variation of P-wave velocity before and Rev. Earth Planet. Sci., 6, 405-436, 1978. after the Galway Lake earthquake(M = 5.2) and the Goat Moun- McKenzie, D. P., The relationshipbetween fault plane solutionsfor tain earthquakes(M -- 4.7, 4.7), 1975, in the Mojave Desert, Cali- earthquakesand the directionsof the principalstresses, Bull. Seis- fornia, Bull. Seismol. Soc. •4mer. 66, 2017-2037, 1976. rnol. Soc. ,•rner., 59, 591-601, 1969. Kane, M. F., Correlation of major eastern U.S. earthquake centers McKenzie, D. P., Platetectonics, in TheNature of the SolidEarth, ed- with mafic/ultramafic masses(abstract), Eos Trans.•4GU, 57, 963, ited by E. C. Robertson,pp. 323-360, McGraw-Hill, New York, 1976. 1972. Kelley, V. C., Albuquerque--Its mountains,valley, water, and volca- McKeown, F. H., Hypothesis:Many earthquakesin the central and noes,Scenic Trips Geol.Past 9, 101 pp., N. Mex. Bur. Mines Miner. southeasternUnited Statesare causallyrelated to mafic intrusive Resour., Socorro, 1969. bodies, J. Res. U.S. Geol. Surv., 6, 41-50, 1978. Kemp, J. F., and W. C. Knight, Leucite Hills of Wyoming, Geol. Soc. McKinlay, P. F., Geologyof Costilla and Latir Peak quadrangles, •4mer. Bull., 14, 305-336, 1903. Taos County, New Mexico, N. Mex. Bur. Mines Miner. Resour. Keys, W. S., Boreholegeophysics in igneousand metamorphicrocks, Bull., 42, 32 pp., 1956. paper presentedat the 20th Annual Symposium,Soc. Prof. Well Mercier, J.-C. C., Magnitudeof the continentallithospheric stresses Log Anal., 1979. inferred from rheomorphicpetrology, J. Geophys.Res., 85, this is- Koons, E. D., Geology of the Unikaret Plateau, northern Arizona, sue, 1980. Geol. Soc. •4mer. Bull., 56, 151-180, 1945. Merrill, R. H., In situ determinationof stressby relief techniques,in Kumamoto, L., Microearthquake survey in the Gerlach-Fly Ranch Stateof Stressin the Earth'sCrust, edited by W. R. Judd,pp. 343- area of northwesternNevada, Colo. Sch. Mines Quart., 73(3), 45- 369, Elsevier, New York, 1964. 64, 1978. Merrill, R. K., and T. L. P•w•, Late Cenozoicgeology of the White Kuntz, M. A., Geologic map of the Arco-Big Southern Butte area, Mountains,Arizona, Tech.Spec. Pap., 1, 65 pp., Ariz. Bur. of Geol. Butte, Blaine, and Bingham counties,Idaho, OpenFile Rep. 78-302, and Miner. Technol., Tucson, 1977. U.S. Geol. Surv., Reston, Va., 1978. Minster, J. B., and T. H. Jordan, Present-day plate motions, J. Lachenbruch, A. H., and J. H. Sass, Heat flow in the United States Geophys.Res., 83(B11), 5331-5354, 1978. and the thermal regime of the crust, in The Earth's Crust, Geophys. Mixon, R. B., and W. L. Newell, Stafford fault system:Structures Monogr. Ser., vol. 20, editedby J. G. Heacock,pp. 626-675, AGU, documentingCretaceous and Tertiary deformation along the Fall Washington, D.C., 1977. Line in northeasternVirginia, Geology,5, 437-440, 1977. Lachenbruch,A. H., and J. H. Sass,Models of an extending litho- Moench,R. H., and J. S. Schlee,Geology and uraniumdeposits of the sphere and heat flow in the Basin and Range province, Cenozoic Laguna district, New Mexico, U.S. Geol. Surv. Prof. Pap., 519, 117 Tectonics and Regional Geophysics of the Western Cordillera, pp., 1967. Geol. Soc. ,•rner. Mern., 152, 209-250, 1978. Moore, J. G., and E. duBray, Mapped offseton the fight-lateralKern Lachenbruch,A. H., and J. H. Sass,Heat flow and energeticsof the Canyon fault, southernSierra Nevada, California, Geology,6, 205- San Andreas fault zone, J. Geophys.Res., 85, this issue, 1980. 208, 1978. Lachenbruch, A. H., J. H. Sass,and S. P. Galanis, Jr., New heat flow Moore, R. B., and E. W. Wolfe, Geologic map of the eastern San results from southern California (abstract), Eos Trans. ,•GU, 59, Franciscovolcanic field, Arizona, Misc. Invest. Ser. Map 1-953, 1051, 1978. U.S. Geol. Surv., Reston, Va., 1976. Lambert, W., Notes on the Late Cenozoicgeology of the Taos-Questa Morgan, W. J., Plate motions and deep mantle convection,Studies in area, New Mexico, N. Mex. Geol. Soc. Field Conf. Guideb., 17, 43- Earth and SpaceSciences, Geol. Soc.,•rner. Mern., 132, 7-22, 1972. 50, 1966. Muffler, L. J.P., Geology of the Frenchie Creek quadrangle,north- Langer, C. J., G. R. Keller, and R. B. Smith, A study of the after- central Nevada, U.S. Geol. Surv. Bull., 1179, 99 pp., 1964. shocksof the October 1, 1972, M -- 4.7, Heber City, Utah, earth- Muffler, L. J.P. (Ed.), Assessmentof geothermalresources of the quake,in EarthquakeStudies in Utah, 1850-1978,edited by W. J. United States--1978,U.S. Geol.Surv. Circ., 790, 163 pp., 1979. 6154 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Muller, O. H., and D. D. Pollard, The state of stressnear Spanish as an indicatorof plate tectonicdriving forces,J. Geophys.Res., 81, Peaks, Colorado, determined from a dike pattern, Pure Appl. 1847-1856, 1976. Geophys.,115, 69-86, 1977. Richardson,R. M., S.C. Solomon,and N.H. Sleep,Tectonic stressin Nakamura, K., Volcanoesas possibleindicators of tectonicstress ori- the plates, Rev. Geophys.Space Phys., 17, 981-1019, 1979. entation-Principle and proposal,J. Volcanol.Geotherm. Res., 2, 1- Richins,W. D., Earthquake swarmnear Denio, Nevada, February to 16, 1977. April, 1973,ARPA Tech.Rep., Univ. of Nev., Reno, 1974. Nakamura, K., K. H. Jacob, and J. N. Davies, Volcanoes as possible Richins,W. D., The Hansel Valley, Utah, earthquakesequence of indicators of tectonic stress orientation--Aleutians and Alaska, November, 1976, in Earthquake Studiesin Utah, 1850 to 1978, ed- Pure AppL Geophys.,115, 87-112, 1978. ited by W. J. Arabasz,R. B. Smith, and W. D. Richins,pp. 409- Natali, S., D. Gish, and M. L. Sbar, A microearthquakesurvey of 421, University of Utah, Salt Lake City, 1979. northeasternSonora, Mexico (abstrac0, Eos Trans.A GU, 60, 894, Roegiers,J. C., and C. Fairhurst, The deep stressprobe--A tool for 1979. stressdetermination (abstract), in New Horizonsin Rock Mechanics; Nelson,W. A., Geologyand mineral resourcesof Albemarle County, Proceedings,14th Symposiumon Rock Mechanics,edited by H. R. Va. Div. Geol. Bull., 77, 92 pp., 1962. Hardy, Jr., and R. Stefanko,p. 755-760, American Societyof Civil Obert, L., In situ determinationof stressin rock, Miner. Eng., 14(8), Engineers,New York, 1973. 51-58, 1962. Rogers,A.M., and W. H. K. Lee, Seismicstudy of earthquakesin the O'Connell, D., C. Bufe, and M.D. Zoback, Microearthquakesand Lake Mead, Nevada-Arizona region, Bull. Seismol.Soc. Amer., 66, faultingin the areaof New Madrid-ReelfootLake, Tennessee,U.S. 1657-1681, 1976. Geol. Surv. Prof Pap., in press,1980. Romney, C., Seismicwaves from the Dixie Valley-Fairview Peak Od6, H., A note concerningthe mechanismof artificial and natural (Nevada) earthquakes,Bull. Seismol.Soc. Amer., 47, 301-319, 1957. hydraulic fracture systems,Colo. Sch. Mines Quart., 51, 19-29, Russell,B. J., A structuralbreak and kinematicsof faulting in the 1956. White Mountains, California (abstract), Geol. Soc. Amer. Abstr. Oliver, J., T. Johnson,and J. Dorman, Postglacialfaulting and seis- Programs, 9, 491, 1977. micity in New York and Quebec, Can. J. Earth $ci., 7, 579-590, Ryall, A., The Hebgen Lake, Montana, earthquakeof August 18, 1970. 1959: P waves, Bull. Seismol. Soc. Amer., 52, 235-271, 1962. Overbey,W. K., Jr., and R. L. Rough, Surfacestudies predict orienta- Ryall, A., and S. D. Malone, Earthquakedistribution and mechanism tion of inducedformation fractures,Prod. Mon., 32(8), 16-19, 1968. of faulting in the Rainbow Mountain-Dixie Valley-Fairview Peak Parsons,R. C., and H. D. Dahl, A study of the causesof roof instabil- area, central Nevada, J. Geophys.Res., 76, 7241-7248, 1971. ity in the Pittsburghcoal seam,in Proceedingsof the 7th Canadian Ryall, A., and K. Priestley,Seismicity, secular strain, and maximum Rock MechanicsSymposium, Edmonton, Mines Branch, Depart- magnitude in the Excelsior Mountains area, western Nevada and ment of Energy, Mines and Resources,Ottawa, 1972. easternCalifornia, Geol. Soc. Amer. Bull., 86, 1585-1592, 1975. Pavlis, T. L., and R. B. Smith, Slip vectorson faults near Salt Lake Ryall, A., D. B. Slemmons,and L. D. Gedney, Seismicity,tectonism, City, Utah, from Quaternary displacementsand seismicity,Bull. and surfacefaulting in the westernUnited Statesduring historic $eismol. $oc. Amer. in press,1980. time, Bull. Seismol. Soc. Amer., 56, 1105-1136, 1966. Pavoni,N., A studyof earthquakefocal mechanismsin the Hollister Sanford,A.R., K. H. Olsen, and L. H. Jaksha, Seismicity ofthe Rio area, southernCoast Ranges,California, Tectonophysics,52, 363- Grande rift, in Rio GrandeRiff: Tectonicsand Magmatism,edited 371, 1979. by R. E. Riecker,pp. 145-168,AGU, Washington,D.C., 1979. Pechmann,J. C., Tectonic implicationsof small earthquakesin the Sass,J. H., and J.P. Ziagos, Heat flow from a coreholenear Charles- central TransverseRanges, California, U.S. Geol. $urv. Prof Pap., ton, SouthCarolina, U.S. Geol.Surv. Prof Pap., 1028-H, 115-117, in press, 1980. 1977. Pitt, A.M., and D. W. Steeples,Microearthquakes in the Mono Lake- Sass,J. H., A. H. Lachenbruch, R. J. Munroe, G. W. Greene, and northern Owens Valley, California, region from September28 to T. H. Moses, Jr., Heat flow in the western United States, J. October 18, 1970, Bull. $eismol. $oc. Amer., 65, 835-844, 1975. Geophys.Res., 76, 6376-6413, 1971. Pitt, A.M., C. S. Weaver, and W. Spence,The Yellowstone Park Sbar, M. L., and L. R. Sykes,Contemporary compressive stress and earthquakeof June30, 1975,Bull. $eismol.$oc. Amer., 69, 187-205, seismicityin easternNorth America:An exampleof intraplatetec- 1979. tonics, Geol. Soc. •4mer. Bull., 84, 1861-1882, 1973. Power, D. V., C. L. Schuster,R. Hay, and J. Twombly, Detection of Sbar, M. L., and L. R. Sykes,Seismicity and lithosphericstress in hydraulicfracture orientation and dimensionsin casedwells. J. Pe- New York and adjacent areas, J. Geophys.Res., 82, 5771-5786, trol. Tech., 28, 1116-1124, 1976. 1977. Prescott,W. H., J. C. Savage,and W. T. Kinoshita, Strain accumula- Sbar, M. L., M. Barazangi,J. Dorman, C. H. Scholz,and R. B. Smith, tion rates in the western United States between 1970 and 1978, J. Tectonicsof the Intermountain SeismicBelt, western United States: Geophys.Res., 84, 5423-5435, 1979. Microearthquakeseismicity and compositefault plane solutions, Prowell, D.C., Index of Cretaceousand Cenozoic faults in the eastern Geol. Soc. Amer. Bull., 83, 13-28, 1972. U.S., Misc. Field Invest. Map, U.S. Geol. Surv., Reston, Va., in Scarborough,R. B., and M. Sha•qullah, Cenozoicsediments and tec- press, 1980. tonics, Arizona Basin and Range province(abstract), Geol. Soc. Prowell, D.C., and B. J. O'Connor, Belair fault zone: Evidence of Amer. •4bstr.Programs, 11, 511, 1979. Tertiary fault displacementin eastern Georgia, Geology,6, 681- Schiifer,K., Recentthrusting in the Appalachians,Nature, 280, 223- 684, 1978. 226, 1979. Prowell, D.C., B. J. O'Connor, and M. Rubin, Preliminary evidence Schaff,C. S., The 1968 Adel, Oregon,earthquake swarm, M.S. thesis, for Holocene movement along the Belair fault zone near Augusta, Univ. of Nev., Reno, 1976. Georgia, U.S. OpenFile Rep. 75-680, 16 pp., U.S. Geol. Surv., Res- Schultz,A. R., and W. Cross,Potash-bearing rocks of the Leucite ton, Va., 1975. Hills, SweetwaterCounty, Wyoming, U.S. Geol.Surv. Bull., 512, 39 Raleigh, C. B., J. H. Healy, and J. D. Bredehoeft,Faulting and crustal pp., 1912. stress at Rangely, Colorado, in Flow and Fracture of Rocks, Scott,D. H., and N.J. Trask, Geologyof the Lunar Crater volcanic Geophys.Monogr. $er., vol. 16, editedby H. C. Heard, et al., pp. field, Nye County, Nevada, U.S. Geol.Surv. Prof Pap., 599-L 22 275-284, AGU, Washington, D.C., 1972. pp., 1971. Reiter, M., C. L. Edwards, H. Hartman, and C. Weideman, Terres- Seeber,L., and J. G. Armbruster, The 1886 Charleston,South Caro- trial heat flow alongthe Rio Grande rift, New Mexico and southern lina, earthquakeand the Appalachiandetachment, submitted to J. Colorado, Geol. $oc. Amer. Bull., 86, 811-818, 1975. Geophys.Res., 1980. Reiter, M., A. J. Mansure, and C. Shearer,Geothermal characteristics Sharp, R. V., Surfacerupturing in Imperial Valley during the earth- of the ColoradoPlateau, Tectonophysics,in press, 1980. quake swarmof January-February,1975, Bull. Seismol.Soc. Amer., Reynolds,M. W., Characterand extentof basin-rangefaulting, west- 66, 1145-1154, 1976. ern Montana and east-central Idaho, in 1979 Basin and Range Sym- Shoemaker,E. M., R. L. Squires,and M. J. Abrams,Bright Angel and posium,pp. 185-193, Rocky Mountain Associationof Geologists Mesa Butte fault systemsof northern Arizona, Cenozoic Tectonics and Utah GeologistsAssociation, Denver, Colo., 1979. and Regional Geophysicsof the Western Cordillera, Geol. Soc. Richardson, R. M., S.C. Solomon, and N.H. Sleep, Intraplate stress •4mer. Mere., 152, 341-367, 1978. ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES 6155

Skinner, E. H., G. G. Waddell, and J.P. Conway, In situ determina- Heat flow in the Upper Mississippi Embayment (abstract), Eos tion of rock behavior by overcorestress relief method, physical Trans. A GU, 60, 310, 1979. propertymeasurements and initial deformationmethod, U.S. Bur. Sykes,L. R., Intraplate seismicity,reactivation of preexistingzones of Mines Rep. Invest., 7962, 30 pp., 1974. weakness,alkaline magmatism,and other tectonismpostdating Slemmons,D. B., Plioceneand Quaternarycrustal movements of the continentalfragmentation, Rev. Phys.Space Phys., 16(4), 621-688, Basin-and-Rangeprovince, USA, J. Geosci.,10, 91-101, 1967. 1978. Slemmons, D. B., D. Van Wormer, E. J. Bell, and M. L. Silberman, Tart, A. C., Recent seismicitynear Charleston, South Carolina, and Recent crustal movements in the Sierra Nevada-Walker Lane re- its relationshipto the August31, 1886,earthquake, U.S. Geol. Surv. gion of California-Nevada,I, Rate and style of deformation,Tec- Prof. Pap., 1028-D, 43-57, 1977. tonophysics,52, 561-570, 1979. Thaden, R. E., S. Merrin, and O. B. Raup, Geologic map of the Smith, C. T., Geology of the Little Black Peak quadrangle,Socorro Grants SE quadrangle,Valencia County, New Mexico, Geol. Quad. and Lincoln counties,New Mexico, N. Mex. Geol. $oc. Field Conf. Map GQ-682, U.S. Geol. Surv., Reston,Va., 1967. Guideb., 15, 92-99, 1964. Thompson,G. A., and D. B. Burke, Rate and direction of spreading Smith, M. B., Effect of fracture azimuth on production with appli- in Dixie Valley, Basin and Range province, Nevada, Geol. Soc. cationto the Wattenberggas field, paperpresented at 54th Annual Amer. Bull., 84, 627-632, 1973. Fall Technical Conference, Soc. of Petrol. Eng., Amer. Inst. of Thompson,G. A., and D. B. Burke, Regional geophysicsof the Basin Mech. Eng., Las Vegas, Sept. 23-26, 1979. and Range province,Ann. Rev. Earth Planet. $ci.,2, 213-228, 1974. Smith, R. B., Intraplate tectonicsof the western North American Thompson,G. A., and M. L. Zoback, Regional geophysicsof the Col- plate, Tectonophysics,37, 323-336, 1977. orado Plateau, Tectonophysics,61, 149-181, 1979. Smith, R. B., Seismicity,crustal structure, and intraplate tectonicsof Tobin, D. G., and L. R. Sykes,Seismicity and tectonicsof the north- the interior of the western Cordillera, Cenozoic Tectonics and Re- eastern Pacific Ocean, J. Geophys.Res., 73, 3821-3845, 1968. gional Geophysicsof the Western Cordillera, Geol. Soc. Amer. Trexler, D. T., E. J. Bell, and G. R. Roguemore,Evaluation of linea- Mem., 152, 111-114, 1978. ment analysisas an exploration technique for geothermal energy, Smith, R. B., and A. G. Lindh, Fault-plane solutionsof the western westernand central Nevada, final report, contractEY-76-S-08-067, United States: A compilation, Cenozoic Tectonics and Regional U.S. Dep. of Energy, Washington,D.C., 1978. Geophysicsof the Western Cordillera, Geol. Soc. Amer. Mem., 152, Trimble, A. B., and R. B. Smith, Seismicityand contemporarytecton- 107-109, 1978. ics of the HegbenLake-Yellowstone Park region,J. Geophys.Res., Smith, R. B., and M. L. Sbar, Contemporarytectonics and seismicity 80, 733-741, 1975. of the western United Stateswith emphasison the intermountain Tsai, Y.-B., and K. Aki, Source mechanism of the Truckee, Califor- seismic belt, Geol. Soc. Amer. Bull., 85, 1205-1218, 1974. nia, earthquake of September 12, 1966, Bull. Seismol.Soc. Amer., Smith, R. B., P. Winklet, J. Anderson, and C. H. Scholz, Source 60, 1199-1208, 1966. mechanismsof microearthquakesassociated with underground Turcotte, D. L., J. L. Ahem, and J. M. Bird, The state of stressat con- mines in Utah, Bull. Seismol. Soc. Amer., 64, 1295-1317, 1974a. tinent margins, Tectonophysics,42, 1-28, 1977. Smi.th, R. B., R. T. Shuey,R. O. Freidline, R. M. Otis, and L. B. Al- Untermann, G. E., and B. R. Untermann, Geologyof Uintah County, ley, Yellowstonehot spot:New magneticand seismicevidence, Ge- Utah Geol.Mineral Surv. Bull., 72, 112 pp., 1964. ology, 2, 451-455, 1974b. Vernon, R. O., Geology of Citrus and Levy counties,Florida, Fla. Smith, R. B., R. T. Shuey, J. R. Pelton, and J.P. Bailey, Yellowstone Geol. Surv. Geol. Bull., 33, 256 pp., 1951. hot spot:Contemporary tectonics and crustalproperties from earth- von Schonfeldt,H. A., R. O. Kehle, and K. E. Gray, Mapping of quake and aeromagneticdata, J. Geophys.Res., 82, 3665-3676, stressfield in the upper earth'scrust of the U.S., final technicalre- 1977. port, grant 14-08-0001-1222,78 pp., U.S. Geol. Surv., Reston, Va., Smith, R. L., R. A. Bailey, and C. S. Ross,Geologic map of the Jemez 1973. Mountains, New Mexico, Misc. Geol.Invest. Map 1-571, U.S. Geol. Voss,J. A., andR. B. Herrmann,A surfacewave s,tudy of theJune 16, Surv., Reston, Va., 1970. 1978Texas earthquake, Earthquake Notes, 51, 3-14, 1980. Smith, S. W., Skagit Valley earthquakesof 1974-1975 (abstract),Eos Walcott,R. I., Gravity flexureand the growthof sedimentarybasins Trans. AGU, 57, 90, 1976. at a continentedg6, Geol. Soc. Amer. Bull., 83, 1845-1848,1972. Smith, S. W., and R. Kind, Observationsof regional strain variations, Wallace,R. E., Geometryof shearingstress and relationto faulting,J. J. Geophys.Res., 77, 4976-4980, 1972. Geol., 59, 118-130, 1951. Speed, R. C., and A. H. Cogbill, Candelaria and other left-oblique Wallace, R. E., Strain pattern representedby scarpsformed during slip faultsof the Candelariaregion, Nevada, Geol.$oc. Amer. Bull., the earthquakesof October2, 1915,Pleasant Valley, Nevada, Tec- 90, 149-163, 1979. tonophysics,52, 599, 1979. Stauder, W., and O. W. Nuttli, Seismic studies:South central Illinois Wanek, A. A., Reconnaissancegeologic map of partsof Harding, San earthquakeof November 9, 1968, Bull. $eismol.Soc. Amer., 60(3), Miguel,and Mora counties,New Mexico,Oil GasInvest. Map 0M- 973-981, 1970. 208, U.S. Geol. Surv., Reston, Va., 1962. Stein, S., N.H. Sleep, R. J. Geller, S.-C. Wang, and G. C. Kroeger, Watts,A. B.,and W. B. F. Ryan,Flexure of thelithosphere and conti- Earthquakesalong the passivemargin of easternCanada, Geophys. nental margins,Tectonophysics, 36, 25-44, 1976. Res. Lett., 6, 537-540, 1979. Weaver, C. S., and D. P. Hill, Earthquakeswarms and local crustal Stevenson,P. R., Microearthquakesof Flathead Lake, Montana: A spreadingalong major strike-slipfaults in California, Pure Appl. study using automatic earthquakeprocessing, Bull. Seismol.$oc. Geophys.,117, 51-64, 1979. Amer., 66, 61-80, 1976. Weaver, C. S., A.M. Pitt, and D. P. Hill, Crustalspreading direction Stewart, J. H., Basin-rangestructure in westernNorth America: A re- of the SnakeRiver plain-Yellowstone system (abstract), Eos Trans. view, Cenozoic Tectonicsand Regional Geophysicsof the Western AGU, 60, 946, 1979. Cordillera, Geol. Soc. Amer. Mem., 152, 1-31, 1978. Weber,R. H., Geologyof the Carrizozoquadrangle, New Mexico,N. Stierman, D. J., and W. C. Ellsworth, Aftershocksof the February 21, Mex. Geol. $oc. Field Conf. Guideb.,15, 100-109, 1964. 1973 Point Mugu, California, earthquake, Bull. $eismol. $oc. Wentworth, C. M., and M. Mergner-Keefer, Regeneratefaults of Amer., 66(6), 1931-1952, 1976. small Cenozoicdisplacement: Probable earthquake sources in the Strubhar, M. K., J. L. Fitch, and E. E. Glenn, Jr., Multiple vertical southeasternU.S., Studies Related to the Charleston, South Caro- fracturesfrom an inclined wellbore--A field experiment,J. Petrol. lina, Earthquakeof 1886--Early MesozoicHistory of the Southern Tech., 2 7, 641-647, 1975. AtlanticCoastal Plain and ContinentalShelf, U.S. Geol.Surv. Prof. Suppe,J., C. Powell, and R. Berry, Regionaltopography, seismicity, Pap., in press, 1980. Quaternary volcanism,and the present-daytectonics of the western Whitcomb,J. H., C. R. Allen, J. D. Gatmany, and J. A. Hileman, San United States, Amer. J. $ci., 275-A, 397-436, 1975. Fernando earthquakeseries, 1971: Focal mechanismsand tectonics, Swanberg,C. A., and P. Morgan, The linear relation betweentemper- Rev. Geophys.Space Phys., 11(3), 369-730, 1973. aturesbased on the silica contentof groundwaterand regionalheat White, W. A., Post-Cretaceousfaulting in Virginia and North Caro- flow: A new heat flow map of the United States, Pure Appl. lina, Geol. Soc. Amer. Bull., 63, 745-748, 1952. Geophys.,117, 227-241, 1978. Wilcox,R. E., T. P. Harding,and D. R. Seely,Basic wrench tectonics, Swanberg,C. A., B. J. Mitchell, R. L. Lohse, and D. D. Blackwell, Amer. Ass. Petrol. Geol. Bull., 57, 74-96, 1973. 6156 ZOBACK AND ZOBACK: STATE OF STRESS IN CONTERMINOUS UNITED STATES

Winchester, D. E., Geology of Alamosa Creek valley, Socorro surementsin central California using the hydraulic fracturing tech- County, New Mexico, with specialreference to the occurrenceof oil nique, Pure •lppl. Geophys.,115, 135-152, 1977. and gas, U.S. Geol. Surv. Bull., 716-,4, 1-15, 1920. Zoback, M.D., J. H. Healy, J. C. Roller, G. S. Gohn, and B. B. Hig- Wood, G. H., Jr., S. A. Northrup, and R. L. Griggs, Geology and gins, Normal faulting and in situ stressin the South Carolina stratigraphyof Koehler and Mount Laughlinquadrangles and parts coastalplain near Charleston, Geology,6, 147-152, 1978. of Abbott and Springerquadrangles, eastern Colfax County, New Zoback, M.D., R. M. Hamilton, A. J. Crone, D. P. Russ, F. A. Mexico, Oil Gas Invest. Map OM-141, U.S. Geol. Surv., Reston, McKeown, and S. R. Brockman, Recurrent intraplate tectonismin Va., 1953. the New Madrid seismic zone, Science, 209, 971-976, 1980a. Woodward, L. A., Rate of crustal extension acrossthe Rio Grande; rift Zoback, M.D., Tsukahara, H., and S. Hickman, Stressmeasurements near Albuquerque, New Mexico, Geology,5, 269-272, 1977. in the vicinityof the San Andreasfault: Implicationsfor the magni- Woodworth, J. B., Postglacialfaults of easternNew York, Geology, tude of shearstress at depth, J. Geophys.Res., 85, this issue,1980b. 12, 5-28, 1907. Zoback, M. L., Mid-Miocene rifting in north-centralNevada: A de- Woollard, G. P., Tectonic activity in North America as indicated by tailed study of late Cenozoic deformation in the northern Basin and earthquakes,in The Earth's Crust and Upper Mantle, Geophys. Range, Ph.D. thesis,247 pp., StanfordUniv., Stanford,Calif., 1978. Monogr. Ser., vol. 13, edited by P. J. Hart, pp. 125-133, AGU, Zoback,M. L., Stateof stressin the southernBasin and Rangeprov- Washington, D.C., 1969. ince and Colorado Plateau (abstract), Eos Trans. •IGU, 60, 311, Wright, H. E., Jr., Tertiary and Quaternarygeology of the Lower Rio 1979. Puerco area, New Mexico, Geol. Soc. •lmer. Bull., 57, 383-456, Zoback, M. L., and G. A. Thompson,Basin and Range rifting in 1946. northern Nevada: Clues from a mid-Miocene rift and its sub- Wright, L., Late Cenozoicfault patternsand stressfields in the Great sequentoffsets, Geology, 6, 111-116, 1978. Basin and westwarddisplacement of the Sierra Nevada block, Geol- Zoback,M. L., and M.D. Zoback, Faulting patternsin north-central ogy, 4, 489-494, 1976. Nevada and strengthof the crust, J. Geophys.Res., 85, 275-284, Wright, L. A., J. K. Otton, and B. W. Troxel, Turtleback surfacesof 1980. Death Valley viewed as phenomenaof extensionaltectonics, Geol- Zoback, M. L., R. E. Anderson,and G. A. Thompson,Cenozoic evo- ogy, 2, 53-54, 1974. lution of the state of stressand style of tectonismin the western Yoder, H. S., Jr., Generationof BasalticMagma, 265 pp., National United States,Phil. Trans.Roy. Soc. London,in press,1980. Academy of Sciences,Washington, D.C., 1976. Zupan, A. J. W., and W. H. Abbot, Clasticdikes: Evidence for post- York, J. E., and J. E. Oliver, Cretaceous and Cenozoic faults in east- Eocene(?)tectonics in the upper coastalplain of South Carolina, ern North America, Geol. Soc. •lmer. Bull., 87, 1105-1114, 1976. Geol. Notes, 19(1), 14-23, 1975. Zemanek, J., E. E. Glenn, L. J. Norton, and R. L. Caldwell, Forma- tion evaluation by inspectionwith the borehole televiewer, Geo- (Received November 29, 1979; physics,35(2), 254-269, 1970. revisedJuly 21, 1980; Zoback, M.D., J. H. Healy, and J. C. Roller, Preliminary stressmea- acceptedAugust 7, 1980.)