JOURNAL OF GEOPHYSICAL
RESEARCH, VOL. tOO,NO. B9, PAGES 17,947-17,963, SEPTEMBER t0, 1995
Strength and rheology of the western U.S. Cordillera
Anthony R. Lowry and Robert B. Smith
Departmentof Geologyand Geophysics,Universityof Utah, Salt Lake City
Abstract. Effective elasticthicknessTedependsprimarilyon temperature,composition,and
stateof stressof the lithosphere.In this paper,we examinehigh-resolutionspectralestimatesof
Teandtheir relationships
to regionalheatflow, ageof thelithosphere,seismicproperties,stress
orientations,andearthquakefocal depthsof the westernU.S. Cordillera. The relationshipof Te
to heatflow indicates
thatductileflow accommodates
long-term(-106 to 108years)isostatic
responseat differentlevelsof the crustanduppermantle,dependingprincipallyon age (and,by
implication,bulk composition)of the lithosphere.Isostaticresponseis primarily controlledby
the uppermantlein Archeanlithosphereof the middleRockyMountains,whereasT• dependson
lower crustalflow in Early Proterozoiclithosphereof the ColoradoPlateau. The YellowstoneSnakeRiver Plain volcanicfield and significantlyextendedregionsin the Basin-Rangeand
northernRocky Mountainsare associated
with latestProterozoicagedlithosphereandindicate
middleto uppercrustalcontrolof long-termTe. We alsoshowthat azimuthalvariationsof Te
reflect deviatoricstressin the lithosphere.Teis foundempiricallyto approximatethe 95th percentilefocal depthof backgroundseismicity.The latterrelationshipis inconsistent
with brittleductilecontrolof focal depth,indicatingthat anotherrheologicaltransition(e.g., from stick-slip
to stableslidingfrictionalbehavior)is responsible.Tectonicandstructuralrelationshipsexpand
uponthehypothesis
thatthegeographic
distributionof tectonicfeaturesdependsfundamentally
on spatialvariationsin strengthof the lithosphere.Moreover,we find a spatialcorrelationof
the IntermountainSeismicBelt to a markedtransitionin Te,implying that forcesresponsiblefor
thisactiveseismiczoneare derivedfrom localbuoyancyanomaliesratherthanfrom current-day
plateboundaryinteractions.
Introduction
Late Cenozoic extensionand seismicity of the western U.S.
Cordillera has been, and continues to be, examined in the con-
ISB is a -100-200 km wide zone of intraplateseismicitywith
shallow focal depths, generally < 20 km deep, that coincides
with the easternmargin of late Cenozoicextensionaldeformation of the western United States [Smith and Sbar, 1974;
text of plate interactions [e.g., Atwater, 1970], hotspot or
plume dynamics [e.g., Smith and Sbar, 1974], and gravita- Smith and Arabasz, 1991]. Contemporarystrain acrossthe
tional potential energy stored during Late Cretaceous/early ISB inferred from cumulativemomentsof historicearthquakes
[Eddington et al., 1987] and terrestrial and satellite-based
Tertiary contraction [e.g., Wernicke et al., 1987]. Tectonic
geodetic
surveys [Martinez et al., 1994] comprisesabout half
models derived from these mechanisms have been used to exthe
total
-1 crn/yr Basin-Range opening rate indicated by
plain the observedtemporaland spatial distributionsof extensatellite
geodesy
and geologicconstraints[e.g., Minster and
sion and volcanism, but the relative importance of each processhasyet to be defined. It is expectedthat a workingmodel Jordan, 1987; DeMets et al., 1987; Dixon et al., 1995].
Effective elastic thickness is an integral function of temof Cordilleran extension should incorporate an understanding
perature,
composition, and state of stressof the lithosphere.
of both the three-dimensional rheological response of contiConsequently,
Tehas provided important insight into theolnental lithosphereand the relative importanceof forcesacting
ogy
and
state
of
stressfor oceanic lithosphere [e.g., Goetze
uponit. However, force balancesand rheologyof continental
and Evans, 1979; McNutt and Menard, 1982]. Rheological
lithosphereare poorly resolvedat this time.
An importantcomponentof rheologicalresponseis the re- studiesof continentallithospheregenerally have not incorposistance to bending by the Earth's elastic layer. Bending rated Te, however, partly becauseof the greatercomplexity of
strength of the lithosphere is characterizedby its flexural
rigidity, or equivalently, effective elastic thickness T e. A
maximum entropy-basedapproachto coherenceanalysisof
topographyand gravity fields [Lowry and Smith, 1994] maps
temperature and compositional variations in continents, and
partly becauseof an inability to resolveTe at the scaleof individual tectonic features. However, recent improvementsin the
resolution of Teestimates [Lowry and Smith, 1994] permit
of the relationshipof Teto theology
long-term (--106 to 108 years)Te at scalesof individual preliminary assessment
tectonostratigraphic
features. We apply this approachto the and state of stress. For example, the depth at which ductile
western U.S. Cordillera, focusing on the 1300-km-long, flow accommodatesisostaticresponsecan be seen to exhibit a
north-southtrending IntermountainSeismic Belt (ISB). The strong dependenceon age of the lithosphere. An apparentrelationship of azimuthal variationsin Te to lithosphericstress
is also explored, and the relationshipof Te to maximum seisCopyright1995 by the AmericanGeophysicalUnion.
mogenic depth in the crust confirms that depth of background
seismicity is not defined by the theological transition from
Papernumber 95JB00747
0 t 48-0227/95/95JB-00747505.00
frictional slip to dislocationcreep.
17,947
17,948
LOWRY AND SMITH: STRENGTH AND RHEOLOGY
The results of this analysis are insufficient to argue for a
particularmodel of extension. However, focusingof seismic
activity a!ong zones of transitionalstrengthsuggeststhat
forces responsiblefor contemporaryextensionand seismicity
of the easternBasin-Rangedo not originate at great distance,
so modernplate boundaryinteractionsare not likely to be directly involved. We suggestinsteadthat extensionalstresses
responsiblefor the ISB are derived from local variationsin
lithosphericbuoyancy (as defined by density averagedover
OF THE CORDILLERA
timates of flexural rigidity, implying some smoothingof the
rigidity distribution. Flexural rigidity was convertedto T, as-
suminga Young'smodulusof l0 TMPa andPoisson's
ratioof
0.25. Variations of density with depth were constrainedusing
publishedcrustalseismicvelocitydata from the Cordillera,detailed
in Table
1.
Elastic Thickness of the Western U.S. Cordillera
Previous investigationsof continental T, have concluded
that (1) greater T, generally correspondsto greater age of
lithosphericgenesis[Bechtelet at., 1990] and lower heat flow
Determination
of Elastic Thickness
[Lowry and Smith, 1994]; (2) significantextensionis limited
Effective elastic thicknessTe is a conceptualrepresentation to domainsof low T, [Lowry and Smith, 1994]' and (3) thin
of the flexural rigidity, or resistanceto bending, D, of the skin contractional deformation fronts generally occur 50 to
lithosphere. T, approximatesthe Earth's uppermost,elasti- 200 km toward the higher strengthside of transitionalzones
cally deforming layer as a perfectly elastic, thin plate, in
of T, [Lyon-Caenand Molnar, 1983]. Theseobservations
are
which case
reinforcedby resultspresentedin Plate 1, and by comparison
with Figure 1. Also, a striking correlationoccursbetween
epicentersdefining the ISB and the east to west transition
the thickness of the lithosphere).
12
(l-v2)
D.
ll/3
from high to low T, (Plate 1a).
A summaryof CordilleranT,, alongwith otherpertinentinformation, is given by physiographicprovince in Table 2.
The volcanic provinces, including the eastern and western
where v is Poisson'sratio and E is Young's modulusof the material [e.g., Turcotte and Schubert, 1982]. It is important to Snake River Plain and the margins of the Columbia Basin,
note that Te is not a physical length parameterbut an alterna- have the lowest T,, -6 km (Plate 1' Table 2). Notably, howtive expressionof bending strength. T, invariably underesti- ever, the interior of the ColumbiaBasin is much stronger,T,
19 km. Outside of the volcanic provinces,the next-lowest
mates the true thickness of the elastically deforming layer be-9 km, occurs in the northern Basin-Range and northern
cause it does not account for anelastic failure, and it can differ
from the true thicknessby more than 100% [e.g., McNutt and Rocky Mountains provinces,where significantlate Cenozoic
extension has occurred. Strongly extended domains, such as
Menard, 1982].
T, is estimatedfrom observationsof the Earth's responseto are describedby Wernicke [ 1991], exhibit the lowest T, in the
loading. Compensation of short-wavelength loads is dis- extensionalprovinces. There are pocketsof higher T, within
tributed by the strength of an elastic plate, whereas long- thesetwo provinces,however; one sucharea in the north-cenwavelength loads have a larger componentof local compensa- tral Basin-Rangecoincideswith an older isotopic domain detion. The degree to which compensationis localized versus scribedby Farmer and DePaolo [1983] (Plate la). T, of the
regionally distributed dependson the flexural rigidity. Thus Early ProterozoicColorado Plateau, -22 km, is significantly
higher than the Late Proterozoic-agedlithospherefurther west.
one can solve for the loads and the load response,hence
from comparison of topography or bathymetry to gravita- The highestTe, averaging30 km, occursin the Archeanmiddle
tional potential in the form of a free air, Bouguer or geoid Rocky Mountains.
anomaly field. A maximum entropy-basedanalysisof coherence of topography and Bouguer gravity, as described by Parameterization of Te
Lowry and Smith [1994], is used in this paper. This approach
enables systematic mapping of T, with resolution approachCorrelationsof T, to the variousmanifestations
of tectoning the scale of tectonostratigraphicfeatures mapped at the ism areintriguing,but we cangreatlyimproveourunderstandsurface. Comparisonof our estimatesof T, with heat flow, ge- ing of the rheologicalimplications
of T, with a simple,conologic information, and independentdeterminationsof elastic ceptualanalysisof yield strength.T, ultimatelydepends
on
thicknessindicatesthat coherenceanalysiscan be used to reli- the bendingstresses
maintainedby the lithosphere,
whichin
ably map T, at scalesof less than 100 km [Lowry and Smith, turndepend
onthefailureproperties
of rock. Depthdependent
(•)
E
1994].
failure is commonlyapproximated
via a yield strengthenve-
Coherenceanalysis was applied to the companiontopogra- lope[GoetzeandEvans,1979]definedby thelesserof thesusphy and completeBouguergravity data usedby Simpsonet al. tainabledifferentialstresses,
Ao, governedby frictionalresis[1986] to develop an isostatic residual gravity map of the tance and dislocationcreep. We adoptthe linear frictional
United States. Gravity data are griddedat a 4-km spacingfrom failure relations:
similarly distributed measurementsthat were processedto remove outliers [O'Hara and Lyons, 1983]. The topographic
(2)
data are on an identical grid generatedspecificallyfor tandem
signal processing [Simpson et al., 1986]. Flexural rigidity
was estimated for 200 km by 200 km and 400 km by 400 km
Ao=[(X/g2+
1_g)-2_
1]pgz
(1-•),
windowsof the data, with centersspacedat 50-km intervals.
Rigidity estimateswere then interpolatedto a 10-km spacing
using a minimum curvature algorithm. The close spacingre-
quiressignificant
overlapof datawindowsfor neighboring
es-
to describea compressional
stressregimeand
AO=[1(x/g2+
1- g)2]
pgz
(1-3,),
(3)
LOWRY AND SMITH' STRENGTH AND RHEOLOGY OF THE CORDILLERA
17,949
Table 1. Velocity andDensityValuesUsedto Constrainthe Inversion
EasternmostBasin and Rangea
Layer Depth to Top,
km
1
2
3
4
5
0
2
8
15
25
Western Middle Rocky Mountainsa
P Velocity,
Density,
Depth to Top,
P Velocity,
Density,
km/s
kg/m3
km
km/s
kg/m3
2280
2700
2610
2800
3050
0
3
16
40
4.6
5.9
6.7 •
7.9
3.4
6.0
5.5
6.5
7.4
Eastern Snake River Plain b
Layer Depthto Top,
km
1
2
3
4
5
6
0
2
5
11
19
42
Colorado
Density,
Depthto Top,
P Velocity,
Density,
km/s
kg/m3
km
km/s
kg/m3
2270
2560
2730
2810
2880
3240
0
2
26
41
East-CentralBasin-Range
d
1
2
3
4
0
2
13
27
Plateau c
P Velocity,
3.3
5.2
6.1
6.5
6.8
7.9
Layer Depth to Top,
km
2470
2680
2870
3230
P Velocity,
km/s
3.0
5.9
6.4
7.4
3.0
6.2
6.8
7.8
2210
2740
2880
3190
Northern Rocky Mountainse
Density,
kg/m3
2210
2680
2790
3070
Depth to Top,
P Velocity,
Density,
km
km/s
kg/m3
0
2
18
33
4.5
6.2
6.8
8.0
Yellowstone Plateau
2460
2730
2880
3280
Great Plainsg
,
Layer Depth to Top,
P Velocity,
km
km/s
Density,
Depth to Top,
P Velocity,
Density,
kg/m3
km
km/s
kg/m3
1
0
2.8
2170
2
1
5.7
2650
3
15
6.7
2860
4
43
7.9
3240
0
15
23
33
km
1
2
3
4
5
2360
2700
3060
3370
Western Snake River Plain i
Northwestern
Basin-Range
h
Layer Depth to Top,
3.8
6.0
7.4
8.2
P Velocity,
Density,
Depth to Top,
P Velocity,
Density,
km/s
kg/m3
km
km/s
kg/m3
0
2.8
2170
0
2.0
1
6
19
31
5.6
6.1
6.4
8.0
2630
2710
2780
3270
2
10
42
5.2
6.7
7.9
aBraile et al. [1974].
bBraileet al. [1982].
CRoller [1965].
dMuellerand Landisman[ 1971].
eSheriff and Stickney [1984] and Sparlin et al. [1982].
1960
2560
2850
3230
_-
fLehmanet al. [1982].
gMcCamy and Meyer [ 1964].
hBenzet al. [ 1990].
illill and Pakiser [ 1966].
17,950
LOWRY
AND
SMITH:
a. Cordilleran
STRENGTH
AND
RHEOLOGY
Seismicit
OF THE
CORDILLERA
b. QuaternaryNormal Faults
•
' '.
•
•.
m
Y •.
-
q' .•
t
L•r ' •:I•
[:
-.-'.T '•
•
' '•
•.•;
..
.-. •, •n• a
'..
.
',.'J '•.
-
-1 ..,,.'.. ' •'
•,&' -• I •''
-•"•1
.-. •.
'7
'%' ':'
- ,•..
•,,
'g •"-,.
• •.-•,•,-.' :' -.
'
•,- .'• , .•' 11.
. W..g,. ,.
t' •"
'?'
.1
.: .'.,.'
.
•... -". "'
,
.....
.•. •,
.: . .,..•½•',.• • '."•,•
•' ,-. •.. ,,..,..,,.., '..,,, .•
•½,
E&•
.
-•
.
,.'
;, -t ....
,,'.t'/
' '1•
t
,"•' fi' ,'•
r
•
.. /v ,
•
'
-..
Flexural Rigidity (Nm)
2x1021
4.0
5.7
lx1022
8.0
11.3
lx1023
16.0
22.6
8x1023
32.0
45.3
Elastic Thickness(km)
Scale (km)
Plate 1. Elasticthicknessof the westernU.S. Cordillera,with historicseismicityand Cenozoicnormal
faults. (a) Earthquake
epicenters,
Mr > 1, recorded
by Universityof Utah,Universityof Nevada-Reno,
and
U.S. GeologicalSurveyseismograph
networks.The whiteboxeslocateISB seismicity
examinedin detailin
Figures9, 10, and 11. Light grey linesare K/Ar isotopicagesof the ArcheanWyomingcraton[Condie,
1981]. Darkergreylinesare boundaries
betweengenetically
distinctlithospheric
blockslocatedvia geochemistryof magmas,from(1) and(2) Farmerand DePaolo[1983],(3) Leemahet al. [1992], (4) Manducaet
al. [1992],and(5) FleckandCriss [1985]. (b) Surfacetracesof normalfaultsexhibitinglate Quaternary
(<500 ka) surfacerupture[afterHecker, 1993;Smithand Arabasz,1991]. Thick white linesare the easternmostfaultswith significant,> I km, offset. Faultsare E, Bozeman;BV, Beaver;CM, CrawfordMtns; E,
Emigrant;
EBL,EastBearLake;EC, EastCache;GV, Gra•dValley;M, Madison;
S, Sevier;SV, StarValley;
T, Teton;WFZ, Wasatch
FaultZone. Boxesindicatelocations
of Teestimates
(largerboxesare400 km by
400 km windows;
smallerare200 km by 200 km windowsof data). Physiographic
provinces
areBR, BasinRange;CB, ColumbiaBasin;CP, ColoradoPlateau;MRM, middleRockyMountains;
NRM, northernRocky
Mountains; SRP, Snake River Plain.
for extension,
whereg is thecoefficient
of staticfriction,p is where• is strain
rate,A andn areempirically
derived
matedensityof overburden,g is acceleration
of gravity,z is depth, rial constants,R is the gas constant,T is temperature,and H*
2•= P/pgz,andP is porepressure
[e.g.,Sibson,1974]. Power is the activation energy of the material [e.g., Goetze and
law creepis describedby
Evans, 1979]. A moresophisticated
estimateof yield strength
might also incorporatecontributionsfrom the low-temperature ductile and semibrittle rheologicalregimes,but frictional
slip and ductile creep are generallysufficientfor flexural analysis [McNutt and Menard, 1982].
LOWRY
AND
SMITH'
STRENGTH
AND
RHEOLOGY
OF THE CORDILLERA
,,•
I
60
70
80
90
17,951
oO ,,•
I
I
ß
I
100
Heat Flow (mW m-2)
C>
Figure 1. Regional surfaceheat flow [after Blackwell and
S[eele, 1992].
Examples of yield strength envelopes for continental
crustalmaterialsare given in Figure 2a. Correspondingflow
law parametersare listed in Table 3. Envelopesare also shown
for a range of continentalgeothermalgradients(Figure 2b).
Note that, becausethe power law creeprelationapproaches
but
never equals zero differential stress,envelopesare truncatedat
a small differential stressvalue (in this paper, AO = 40 MPa)
chosensuch that the contributionto Te from greaterdepthsis
negligible [McNutt, 1984]. It is readily apparentthat yield
strengthprofiles can vary significantly with temperaturefield
and bulk compositionat depthsof ductile creep.
The state of stressrelative to rock strengthis also important. Flexural rigidity dependson plate bending moment M
via
M
(5)
K
where K is the curvature of the plate [e.g., Turcotte and
Schubert, 1982]. The momentin turn correspondsto the ver-
tically integratedfiber stresso/generated by bending,
weighted by distancefrom a neutral surface at depth zn(see
Figure 3a):
M=fo
rm
of(z)(Z-an)dz,(6)
where Tm is mechanicalthicknessof the plate (i.e., that sup-
o
o
17,952
LOWRY
AND
SMITH'
STRENGTH
AND
RHEOLOGY
OF THE CORDILLERA
0 Tm
=13.4km
go
Te=13.4km
2
4
Zn
a.
Dry granite
pOWer
lO
20-
25
,,,,i,,
-700
-600
Compression ,Ext,ensi,on
,,i,,,
,l,,,,ll,,ll,,,,i
-500
-400
i,,,
-300
-200
, , ,,
-100
0
,,
lOO
,
,
,
200
,,
,
12Compression _
Extension
14
-35b'"2'5b'"l'5b'' '-•6 '6''5b''' )}d'' •}d'' •50
300
Differential stressAo (MPa)
Differential stressAo (MPa)
0tTm=
13.4km
••
Te=
11.1
km
2
Do
10
•6
.
12
3o
35
'
-700
''
'
I'''
-600
C,omp,
res,sion
'
-500
'
''
'
'
'''
-400
''
-300
''
I'
''
-200
'l'
-100
'
''l'
143•C,ø'm'p'r,
es,
s!ø'n.
,.... ':'"',
'•'•'"""-i:"'•,,,,
.... ,...E.x,
te,
n,s(o,n
Extension
'
0
'
'1
''
100
''
I''
200
'
50
'
300
-250
Differential stressA• (MPa)
Figure 2.
-150
-50
0
50
150
250
350
DifferentialstressA6 (MPa)
O'
Tm=
13.4km,,
•
Te=4.1
km
'• •
2
Yield strength envelopes (a) for various crustal
4
rheologies,assuminga geothermalgradientof 20øCkm-• anda
strainrateof 10-•5s•; and (b) for variousgeothermal
gradients,
usinga dry-graniterheologyand 10-•5s-• strainrate.
c
ports deviatoric stresson geologic timescales). In a thin, perfectly elastic plate, the fiber stressvaries accordingto
12
g(Z--Zn).
(5)
Uornpression
, ....
, ....
14
....
-350
(7)
-250
-150
,..•..,
-50 0
....
50
Ex,
tension
, ........
150
250
350
Differential stressAo (MPa)
Figure 3. Fiber stressesin a thin bendingplate. Te depends
Superimposing a yield strength limit truncates the plate
bending stresses(Figure 3b) so that bending moment, and
consequentlythe effective elastic thickness,is reduced,i.e., Te
< Tm[McNutt and Menard, 1982]. Figure 4a illustratesthe effect of increasing plate curvature on Te. A uniform, regional
horizontal stress(henceforthin this paper referred to as a "tectonic" stress) will also have significant effect on Te [McNutt
on bending moment, i.e., integral of the shadedarea weighted
by distancefrom the neutral surfacez,. (a) For a perfectlyelastic plate, mechanicalthicknessTmequalselasticthicknessTe.
(b) If yield strength is limited by frictional slip and ductile
creep, sustainable bending moment, and hence Te, is decreased. (c) Bending moment is decreasedfurther by superimpositionof a large tectonicstress. Envelopesare for dry gran-
ite withtemperature
gradient25øCkm'• andstrainrate10-•5s-•.
Table 3. Material ParametersUsed in This Study
Rock
Density p,
Young's
PowerLaw
PowerLaw Activation
Thermal
Composition kg m-3 ModulusE,CoefficientA, Exponentn EnergyH*, ConductivityK,
Pa
MPa-" s4
kJ mol -•
W m4 øK4
Dry granite
Wet granite
Dry quartzite
2600
2600
2600
7x10•ø
7x10•ø
7x10•ø
2.5x10-9
2.0x10-4
3.2x10-s
3.4
1.9
1.9
139.
137.
123.
2.4-3.8
2.4-3.8
4.2-6.3
Diorite
2700
7x10 •ø
1.3x10 -3
2.4
219.
2.6-3.5
Dry olivine
3250
8x10•ø
6.3x104
3.5
533.
3.7-4.6
Power law propertiesare from Hansen and Carter [1982] for crustalcompositions
and from Kirby [1983] for olivine.
LOWRY AND SMITH: STRENGTH AND RHEOLOGY OF THE CORDILLERA
and Menard, 1982]. A superimposedtectonic stress(50 will
35
shiftthe referenceaxis,(sf(z) = 0, withinthe yield strength
10øC km-1
30
a.
envelope as indicated in Figure 3c. As the reference state of
stressmoves outward to either limb of the envelope, the sustainablebendingmoment (and hence Te) is decreased.Te as a
function of superimposedtectonic stressis shown in Figure
30øC
4b.
20
•15
17,953
'-10
m
5
Te and Tectonics
50øC km-1
Parameter
104
10-7
10-6
Sensitivity
of T e
i•-5
From inspection of equations (1) through (7), Te depends
upon the elastic propertiesof rock; the coefficient of friction
and pore fluid pressureat depthsof frictional failure; composition,
strainrate and temperatureat ductile depths;and the state
35
of stressdue to plate bending and tectonic forces. Many of
these properties are poorly constrained for most continental
lithosphere;that combined with the sheer number of variables
makes
analysisof the implicationsof Tea somewhatdaunting
20
task. However, Te is much more sensitiveto someparameters
than to others. Figure 5 displaysthe sensitivityof Te to reasonable ranges oi each of its variable parameters,as well as a
typical range of standarderror for estimatesof Tefrom the
Basin-Range(calculatedfrom the method describedby Lowry
!
,, !
and Smith [1994]). Virtually all of the sensitivitiesscalelog-700-600-500-400-300-200-100
0
100 200 300
arithmically, such that if a different reference state were choTectonicstress(MPa)
sen, the sensitivityrangeswould be displacedon the log plot,
but
lengthsof the bars would remain approximatelythe same.
Figure 4. Dependenceof elastic thicknesson state of stress
(The
exceptionsare for It and/•, which would showdecreased
for a dry-granitecomposition,strainrate of 10-15s-1, and various geothermalgradients. (a) Te versusfiber stress(expressed sensitivity ranges given a less extreme state of lithospheric
as plate bending curvature), assumingno far-field stress,and stress.) Thermal gradient, state of stress(due to plate bending
and tectonic loading), and power law compositionare the only
(b) Te versustectonicstress,givensmallcurvature,10-8m-1.
parameters whose sensitivity ranges are significant when
In-PlaneCurvature(m-i)
0•o,C,•,
,•,
f,,,
,•,•,,,•,o
,C,••,o,•!
,•,, ',,,,,
Rangeof Te
(for plausiblerangeof parameter)
10
.
Poisson's Ratio v -
Young'sModulusE, Pa -
FrictionCoefficientIt Pore Pressure Coefficient
•
-
PowerLaw CompositionStrain Rate •;, s-1 Thermal GradientdT/dz, øC/km -
.
.
. . . .ll
100
_•
_• _• : : _, ::
0.3•0.1
2x1011
"'"'"'"'":'"'•4xl
0.2"'::'""•
0.7
0.8""•0.2
Quartz
'"'"-'"'"'-'"• Olivine
10-18
'•'"'-'•"-:--•
10-14
50'--'"'--'"'"---'""•
10
Plate CurvatureK, m4 -
TectonicStress(50 Typical Error Bound
of a Te Estimate
Figure 5. Sensitivityof Te to its variableparameters.Rangesfor v andE arefor D = 1022N m; It and/•
sensitivities
usean extremeplatecurvature
K= 10-6m-l;all othercases
usedrygraniterheology
with(5o= 0,
K = 10-8m-1,dT/dz= 25øCkm-1, i;= 10-iss-1,/• = 0.37andIt = 0.65,exceptasotherwise
indicated.Error
boundsfor a typicalTeestimateare includedfor compari.son.
Powerlaw rheology,thermalgradient,and
stateof stresshave the most significanteffectson Te.
17,954
LOWRY
AND
SMITH'
STRENGTH
AND RHEOLOGY
OF THE CORDILLERA
comparedto the rangeof error in Te. Teerror rangesalso scale
logarithmically, and are more a reflection of resolutionthan of
random error: Te estimation assumesuniform and isotropic
rigidity within a data window, whereasactual strengthproperties can vary dramatically (e.g., Plate l a) and also vary with
direction. In general, a factor of 2 changecan be considered
significant relative to the resolution of T e estimates [e.g.,
[e.g., Fountain and Christensen,1989]. Stressmagnitudesat
depthsof interest are not known, but aspectsof the state of
stresscan be inferred from principalstressdirectionsgiven by
stress measurementsand stress indicators [e.g., Zoback and
Zoback, 1989], and from the maximum depths of frictional
failure observedin earthquakes.
Bechtel et al., 1990].
Heat
Flow
Parametersthat have insignificantinfluence on Te are fixed
for the remainderof this discussion. We adopt previousconvention [e.g., Brace and Kohlstedt, 1980] by using It = 0.65
and )• = 0.37. Somestudiessuggestthat dynamicshearstress
on major faults is a factor of 3 to 10 lower than predictedby
these parameters [Zoback et al., 1987; Bird and Kong, 1994],
but changing the linear friction coefficients to reflect these
studies would have no impact on this analysis. We assume
Sensitivity of T e to thermal gradient is reflected in the
strongcorrelationbetween areasof low Te(Plate 1) and high
heat flow (Figure 1). Regional heat flow Q, digitized from
Blackwell and Steele [1992], is plotted against measurements
of Tein Figure 6. Linear regressionof Q versusTe yields a
negativeslope to > 99% confidence,but with a correlationcostrainratesare depth-independent
andrangefrom 10-18to 10-14 efficient of-0.46 and very significantvariance. Some of the
s-1, as inferred from geodeticmeasurements
and cumulative scattercan be attributedto the poor resolutionof Te estimates
moments of historic earthquakes [Eddington et al., 1987;
as well as the poor distributionof, and influenceof hydrology
Dzurisin et al., 1990; Savage et al., 1992; Martinez et al.,
on, Q measurements,
but part of the scatteroccursbecauseTe
1994].
dependson otherfactorsbesidesjust temperature.The predicThe three most importantpropertiesin defining Te, temper- tive relationshipbetweenheat flow and Te, from equations(1)
ature, power law composition, and state of stress,can all be
through (7), is also shown in Figure 6 for severalcrustal and
constrainedto some extent from independentdata. Regional upper mantle rock types, using a geothermal gradient estiheat flow patterns [e.g., Blackwell et al., 1991] provide inmated from dT/dz = Q/r, where tc is thermal conductivity.
sight into the temperature field. Seismic velocity structure Clearly, the regression relationship is inconsistent with a
[e.g., Braile et al., 1989] can be used to infer composition power law correspondingto a single, uniform composition.
tto•
too-'
•" 90-
• 8o-'
70-
.
60
.
.
50
i
,
,
t0
Elasticthickness(km)
Wet Granite
Dry Granite
Snake River Plain &
PeripheralColumbiaBasin
ß NorthernRockyMountains
NorthernBasin-Range
Diorite
Olivine
[] Interior Columbia Basin
Middle Rocky Mountains
Colorado Plateau
Figure 6. Regionalheat flow, digitizedfrom Blackwell and Steele [1992], versusestimatesof Te.
Locationsof Te estimatesare indicatedby province. The predictiverange for variouscompositions,indicatedby light grey cross-hatch,
is basedon the rangeof thermalconductivity(Table3) andstrainratesranging from 10-14to 10-18s-1.
LOWRY
Compositional Control
AND
SMITH:
STRENGTH
AND RHEOLOGY
of Te
The particularpower law that definesTedependsprimarily
on thermal gradient and compositionallayering of the lithosphere. In general, higher thermal gradientsshouldmobilize
OF THE CORDILLERA
17,955
Mountainsis much less than crustalthickness,and averages
lessthan half the depthto the baseof the midcrust,generally
-15 to 20 km. Thus a middleto uppercrustalpowerlaw compositionis a reasonableassumption.ColoradoPlateauTe averages 22 km, and crustal thickness is 35 to 45 km, whereas
rocksin shallowerrheologicallayers. We havenotedthat Te depth to top of the lower crust is about 25 km. Hence, assumtends to be relatively consistent within physiographic ing moderatereductionof elastic thicknessby stress,a lower
provinces,but can vary dramaticallyfrom one provinceto an- crustalcontrolof Te would be mostprobable.
other (Plate l a; Table 2), and so one might reasonthat different provincesare characterizedby activation of different rheo- Stress Orientations and Anisotropy of Te
logical layers as a result of variations in mantle heat flow.
Then the relationshipof Te to O should be more consistent
Figure 4 representsthe reductionof Te by stresswhen horiwith the relationshippredictedfor a singleuniformcomposi- zontal componentsof the principal stresstensor, (Jhmax
and
tion if the observationsare broken out by province. This is
C•hmin,
are equal. One can infer, however,that if lJhmax
is signifindeedthe case(see Figure 6). TeversusO of the Colorado icantly greaterthan (Jhmin,
there shouldbe an azimuthalvariaPlateauhoversnear the quartzdiorite (-lower crustal)predic- tion in Te. Examples in Figures 7a, 7b, and 7c show the aztive relationship. The relationship in the middle Rocky imuthaldependence
of Tepredictedby applicationof equations
Mountains
is most consistent
with
lower
crustal
and mantle
(1) through (7) to examplesof extensional,strike-slip,and
rheologicalcontrol. Resultsfrom the Basin-Range,northern
Rocky Mountains, and the volcanic provinces mostly fall
within the range of upper to middle crustal (quartzo-feldspathic) rheology, except for the interior Columbia Basin,
which is more consistentwith a lower crustalrelationship.
There remainsan obviousproblem,however,in that each of
compressionaltectonic stressregimes.
terpreted by Blackwell and Steele [1992] is sensitiveto nearsurfaceprocessesthat are independentof mantleheat flow (for
comparisonto stressorientations. Figure 7d is an example of
observed azimuthal variation of Te from the eastern Snake
Observationsof lithosphericstate of stressare generally
limited to indicatorsof principal stressdirections,including
stress inversions of fault slip data and focal mechanisms
[Angelier, 1984; Gephart and Forsyth, 1984], well bore
breakouts[Goughand Bell, 1981], and volcanicvent alignthe physiographic
provincesspansnearly the samerangeof ments [Nakamura, 1977]. In situ measurements
suchas hyheatflow. If thermalgradientalonedetermines
the rheologi- draulicfracturing[Hickmanand Zoback,1983] and overcoring
cal layer that respondswith ductile flow, then heat flow should [McGarr and Gay, 1978] provide stressmagnitudes,but only
breakout according
to provinceaswell as Te. Therearea cou- in the upper few kilometersof the crust. However, the effect of
ple of reasonswhy it doesnot: (1) the regionalheat flow instresson isostatic responsecan be qualitatively evaluatedby
example,
the< 60 mW m-2anomaly
in thecentralBasin-Range River
(seeFigure 1) is thoughtto have a hydrologicorigin), and (2)
the defining rheology also dependson bulk compositionof
the layers.
Jordan [ 1978, 1981] notes that compositionaldifferentiation to depths of > 200 km is necessaryto accountfor the
buoyancy,petrologyand seismicvelocity structureof continental mantle. According to his hypothesis, cratonic lithosphereis stabilized by depletionof mantle basalt during extreme tectonomagmaticevents, resulting in reduced density,
viscosity, and thermal conductivity; Archean lithosphere
tends to be most stable becauseit has experiencedthe most
cycles of tectonomagmatic fractionation [Jordan, 1981].
Correlationof high Tewith greater lithosphericage (Table 2)
is consistentwith a bulk compositionalinfluence on rheology, and compositional stabilization also helps to explain
the apparentlong-term persistenceof lithosphericstrength
variations [Lowry and Smith, 1994]. However, whereas
Jordan's [1981] stabilizationhypothesisfocuseson the chemistry of mantle rocks, the rheology of many western U.S.
provincesis controlled by crustal compositions. Hence the
uniformity of Te within provinceshaving uniform ages but
varying heat flow would suggestthat tectonomagmaticfractionationmay increasethe relative stability of crustallayers
as well as the mantle.
The power law compositions
implicit in Figure 6 are supportedby comparisonof Tewith depth to seismicvelocity discontinuitiesdetailedin Table 1. For example,Te of muchof
the middle Rocky Mountains approachesor exceedsthe -40
km thicknessof the crust. Given that Te underestimates
Tm,
flexuralstrengthmustbe controlledby the uppermantle. Te in
the Basin-Range, Snake River Plain, and northern Rocky
Plain.
Formal
mathematical
inversion
for axes of a best
fitting ellipse yields a minor axis direction of N49øE+13ø.
Thus, assumingthe anisotropyreflects a state of extensional
deviatoric stress,the N49øE direction should correspondto
IJhmin.Volcanic vent alignmentsat the same location indicate
an extensionalstate of stresswith minimum compressiondirection N48øE [Zoback and Zoback, 1989].
Bechtel [1989] notes anisotropyof Te in both the BasinRange and East African Rift extensionalprovincesthat he attributesto mechanicalweakeningof the elasticlayer by faults.
However,we observeas muchas 50% reduction
of Tein the
minor axis direction, which is too large to accountfor by
alignment of faulting. Frictional slip can occur for a broad
range of fracture orientationswith relatively minor variation
in frictional yield strength [e.g., Forsyth, 1992], and we observe minor axis directions of Tein the middle Rocky
Mountainsand elsewherethat are not perpendicularto nearby
faults. Hence it is probably more accurate to associate
anisotropyof Te with stressorientationthan with fault orientation.
Figure 7 indicates a different pattern for each of the three
stressregimes,but the patternsare not diagnostic. For example, the bilobate pattern in the extensionalexample (Figure
7a) might be observed in a compressionalstress regime if
lJhmax--lJv approaches the maximum yield strength of the
lithosphere. It could also be seenin a region of strike-slipif
one horizontalprincipalstresswere near the lithostat,c•v. The
only diagnosticpattern is the cloverleaf pattern of Figure 7b,
predictedonly for strike-slip stress. Hence if azimuthal variation of Te is to be usedto infer principalstressorientation,independentinformation about the stressregime (focal mechanisms, for example) will be needed to determine whether the
17,956
LOWRY
AND
SMITH:
STRENGTH
AND RHEOLOGY
b. StrikeSlip
a. Extension
330of
--
•
300
ø/
OF THE CORDILLERA
30ø
330of
%
60 ø
270ø[
240ø
N
/
210 ø
--180 ø
120 ø
300ø
••....
--
30ø
240
ø
/120
ø
210o•
150 o
o
60
::•i:?:ii:i?::i?
.......
•:•:•:•:•:•i!ii!:!i::ii:?:::
.... }90ø
J
150ø
180 ø
d. Observed
c. Compression
330o/
•
•
90 270
ø[•
/
--
o/
300
•
330ø•
30ø
••::•::•::•::•
::•::•::•::•::•::•:•:•:.:•.:•
60ø
--
•
300
ø
30ø
> 60
ø
270[...............................................................................
•...........
}
\
210o•
--_ J
150
ø
240ø
N
•
210 ø
180 ø
--180 ø
/
•
150
ø
120ø
Figure 7. Examplesof azimuthalvariationof Te. Figures7a, 7b, and 7c were modeledusinga dry-granite
powerlaw rheology,strainrateof 10-•ss-• and20øCkm-• geothermal
gradient.Arrowsindicatehorizontal
principalstressdirections;smallerpatternsare tectonicstress(•0 in (blackfill) compression,
or (white fill)
extension. (a) Deviatoricextension,with (Jhmax----(•v
and(•hmin
= (•v + 80 MPa. (b) A strike-slipstressregime,
with (Jhmax
-- (Jv--80 MPa and(Jhmin
= (Jv+ 80 MPa. (c) Deviatoriccompression,
with(Jhmax
= (Jv--80 MPa and
(Jhmin
----(Jv' (d) An exampleof observed
azimuthalvariationfromtheeasternSnakeRiverPlain.
elongate axis correspondsto the maximum or minimum compressive stress.
The contemporarystate of stressin the Cordillera is extensional, with a possible component of strike-slip along the
westernmargin of the ColoradoPlateau [Zobackand Zoback,
1989; Bjarnason and Pechmann, 1989; Smith and Arabasz,
1991]. Minimum horizontal compressive stress directions
from various stress indicators are given in Figure 8a.
Minimum compressivestresshas been interpretedto be approximately NE in the easternSnake River Plain, E-W in the
easternBasin-Rangeand NNE-directedin the ColoradoPlateau
[Zoback and Zoback, 1989].
Given the extensional regime, minor axes of Te anisotropy
shouldcorrespondto the minimum horizontalcompressiondirection. Vectors representingthe minor axis direction of best
fit Teellipses are scaledby elliptical eccentricityand plotted
in Figure 8b. Stressdirectionsinferred from Teanisotropy are
in close agreement with stressindicator data from the northeast Basin-Range,where (Jhmin
is '•E-W, and the easternSnake
River Plain, where NE directed (Jhmin is observed.
Interestingly, we do not observe significant anisotropyof Te
in the interior of the Basin-Range. In the Colorado Plateau,
azimuthal variation of Tesuggestsan -ENE extension direction as comparedto the NE to NNE extensiondirectioninferred
from earthquake focal mechanisms [Wong and Humphrey,
1989; Zoback and Zoback, 1989].
The largest strength anisotropy in the study area is observed near the Wasatch fault zone, Utah, where minimum Te
directions are orthogonal to the Wasatch fault in the hanging
wall and parallel in the footwall. This apparentrotation of
stressfield is not well resolved by independentstressorientation data: severalof the (Jhmin
directionsindicatedby fault slip
and focal mechanisms
on the Wasatch
are consistent with the
E-W minimum Tedirectionwest of the fault, but the only support for -N-S (Jhmin
eastof the fault comesfrom a -500 m deep
hydraulic fracture experiment. However, the rotation is consistent with contemporary horizontal strain. Trilateration
measurements
from
1972 to 1990 indicate N85øE
uniaxial
ex-
tensionin the hanging wall block, and N20øE extensionin the
footwall block of the fault [Savage et al., 1992]. The observed
strength anisotropy is not a result of the extreme footwall
flexure of the Wasatch [e.g., Parry and Bruhn, 1987], as the direction of reduced strengthis orthogonalto that predicted for
footwall bending.
There are placeswhere stressdirectionsand anisotropyof Te
are decidedly inconsistent;Yellowstone and its surroundings
are an example. Focal mechanismsof the 1959 M s 7.5
HebgenLake, Montana earthquakeand its aftershockssuggest
-NNE extension, but T e is lowest WNW. Inversion of
Yellowstoneearthquakefocal mechanisms[Peyton, 1991] and
Global Positioning Satellite measurements[Meettens et al.,
1993] likewise indicate NE extension. Strain behavior at
LOWRY AND SMITH: STRENGTH AND RHEOLOGY
a. Minimum
stress orientation
from stress indicators
114 ø
I
113 ø
112 ø
I
I
111 ø
110 ø
I
I
45 ø--
Stress Orientation
QualityRanking
A
B
C
17,957
b. Minimum strengthorientation
from anisotropyof T
: 114
450_
I ø•
.•;1.2.•_•111•
1
•i'••
44o
(Part a.)
MRM
OF THE CORDILLERA
43 ø--
43 ø-
Te Ellipse
Eccentricity
•/a2_b
2
42 ø
a
1.0 0.5 .25
41 ø_
40 o-
BR
_
/
41o__
(Partb.)
39 ø_
CP
39 ø_
38 ø-
0
200
Scale
(km)
38 øl
•-
/
Figure 8. State of stressin the Basin-Rangetransitionto the ColoradoPlateauand Rocky Mountains. (a)
Orientations of minimum horizontal compressive stress, from the compilation of Zoback and Zoback
[1989]. (b) Minor axis directionsof best fitting ellipsesto polar plots of Te.
Yellowstone is very heterogeneouswith regard to both time
and space [e.g., Smith and Braile, 1994], and so the discrepancy may be due to differences in the timescales and spatial
wavelengths sampled by T e versus stress indicators.
Alternatively, it is possiblethat sensitivityof Te to strainrate
is amplified by rheologicalchannelingof crustal flow in some
regions. Our simplistic,one-dimensionalmodel of Te is insufficient to examine the latter possibility; a three-dimensional,
dynamical model of lithospheric strength behavior would be
required. Nevertheless,lithosphericstate of stressappearsto
influenceTe throughoutmuchof the Cordillera.
Earthquake
Focal
Depths
Many investigationsassociatefocal depths of large, M > 6,
earthquakeswith the transition from frictional slip (brittle) to
dislocation creep (ductile) deformation [e.g., Sibson, 1982;
Smith and Bruhn, 1984] or a transitionfrom velocity weakening (i.e., unstable) to velocity strengthening (i.e., stable)
frictional slip behavior [e.g., Tse and Rice, 1986]. Velocity
strengthening frictional slip and dislocation creep are endmembersof a single continuumof deformationbehavior, however [Tse and Rice, 1986], and each dependsfundamentallyon
temperature and composition. Maximum focal depths of
smaller, background seismicity may similarly depend on rheology, or they may dependinsteadon stateof stressif stressis
insufficient to fail the lithospherenear the rheological transition. The latter possibility seemsunlikely in the ISB, however, where background seismic depths approach the focal
depths of larger events. In any case, focal depths for both
large and small earthquakeswill dependon many of the same
parameters as effective elastic thickness, and so we expect
there shouldbe a relationshipbetweenearthquakefocii and Te.
Unfortunately, observing that relationship is problematic.
Maximum focal depths of backgroundseismicity are not well
constrainedin the ISB becauseof poor sampling in both time
and space. Focal depth information is available for only a
small fraction of the earthquake cycle, and regional seismograph networks do not provide reliable depths for most
smaller events except in special densifted study areas [Smith
and Bruhn, 1984; Smith and Arabasz, 1991]. In the ISB, focal
depths of small events are considered reliable if the vertical
hypocentral error calculated by the location algorithm is 2 km
or less and the distance
to the nearest seismometer
is less than
the focal depth [Arabasz et al., 1990]. Seismographspacing
is -35 km along the WasatchFront, and -20 km in the eastern
Snake River Plain and Yellowstone [Smith and Arabasz, 1991;
Jackson et al., 1993; Smith and Braile, 1994]. Elsewhere,
spacing is greater, typically exceeding 50 km. We carefully
selectedthe most reliable focal depthsfrom the ISB using the
above criteria; these are plotted with Te in Figures9, 10, and
11. With the exceptions of the aftershock sequencefor the
1983 M L 7.3 Borah Peak, Idaho, earthquake(Figure 1l a) and
continuing activity near the 1959 M s 7.5 Hebgen Lake,
Montana, event (Figure 1lb), maximum focal depths approximately coincidewith Te. It would be hazardousto assessa relationship when the data are spatially parameterized,however.
Not only is the earthquakedistributionpoorly sampled,but Te
and earthquakes are sensitive to rheological properties at
LOWRYAND SMITH: STRENGTHAND RHEOLOGYOF THE CORDILLERA
17,958
0
ß
E 20-
• 30'
Q 40-
a
-150
-100
-50
0
50
100
150
200
250
Distancefrom the Basin-Range-RockyMountainsboundary(km)
lO
1
•2o
401
60
..........
-150
. ....
-100
-50
0
. ...................
50
100
''''
150
200
250
Distancefromthe Basin-Range-Rocky
Mountainsboundary(km)
.
•.•..... ,.:•. ..¾•:.•..
.....
•.•.,•....,
....
g:.,.• • :i•.•
A
10•.
20•
301
..•"'""'•
.................
• ............
•....
...... .... :>½....
,•::•.-.;• ...... .• .:.-
5oi
C.
6o_.
•o'' '-•6o'
' '-•' ' ' • .... •o.... 1•' ' '1•' ' '2•' ' '25o
Distancefromthe BasinRange-RockyMountainsbounda• (km)
•
•
.....½•...--
•10
•,:
..
•;•
•...
• ....
•:
'•"--'--'--'-'-•'""•'••"'"•'••••••
...........
.`••``••••••••••1.......•....,•.......,.......,.....```.•`•.``.•..•.•.."
20
30
........
-150
, .....................
-100
-50
0
i
i
50
i
100
i
150
.
,
**'-"•:•:•'•:•*•-•....• :•:'•.•.'*'<: •*'<•:•
.....:*m*•--•.•.
,
200
250
-.•.e..... •.-:• :•
Distancefromthe BasinRange-RockyMountainsbounda• (km)
0
•$ •
g• •
113 ø
• •
.....
112'
111'
110'
•
•10
•20
30
e
......
-200
, ......
-150
-100
-50
0
50
.
100
150
200
Distancefromthe BasinRange-Roc• Mountainsbounda• (km)
o
510
- f
•'20
30
'
-200
''
i
-150
i
-100
i
-50
- -' 'i
0
.
,
,
i
50
....
i
........
100
150
200
Distancefromthe BasinRange-RockyMountainsboundary(km)
Figure 9. Focaldepthsplottedwith Tefor the Basin-Rangetransitionto the middle Rocky Mountains
physiographic
province.-SeePlate 1 for locationof studyarea.
completely different spatial scalesas well: whereasearthquakesrespondto the physicalstatewithin a relatively small
the depth of rheologicaltransitionfrom frictional to dislocation creep behavior. The latter relationshipis insensitiveto
power law composition;relations for both dry granite and
volume of rock, the absolutelimit of resolution for Te estimation occursat wavelengthsof--40 km. Thus comparisonof T• diorite are shown and are virtually indistinguishable. There
and earthquakefocal depthis liable to be valid only for gross are several notable relationships. The 95th percentile depth
of backgroundseismicity,calculatedfor 5-km-wide bins, apstatistical relationships.
Figure 12 showsfocal depthsplottedagainstT• at the epi- proximatelytracksT• (as mighthavebeendeducedfrom previcentral location,along with the predictedrelationshipof Te to ous figures). Regressionof the depthsof ISB background
LOWRY AND SMITH: STRENGTH AND RHEOLOGY OF THE CORDILLERA
o•
•
301 ....
-150
, ....
-100
,•
, ....
-50
, ....
0
, ....
50
17,959
e
, ....
100
, ....
150
, ....
200
250
Distancefromthe Basin-Range-ColoradoPlateauboundary(km)
.••
•....••.••.••
..........
3040
•0' ' '-{do" '-•6' ' ' 6 .... •0.... •bb' ' '•'5b' ' '2bb'' '250
Distancefromthe Basin-Range-Colorado
Plateauboundary(km)
•
•
•1o
...........
":I"
•'.-•••'--:
••'"
•' "•
'•',
C
.I..
•
111 ø
-100
-50
0
50
100
150
200
250
0o
300
Distancefromthe Basin-Range-Colorado
Plateauboundary(km) 0
•
•10
=
=20d ß
•
30 ''
-100
• ....
-50
• ....
0
• ....................
50
100
150
200
250
300
Distancefromthe Basin-Range-ColoradoPlateau boundary(km)
Figure 10. Focal depthsplotted with Te for the Basin-Range-ColoradoPlateau transition. See Plate 1 for
location of study area.
•-20]
0
40
'
80
120
•-20-•
/' • • ••
160
0
Distancefrom NW to SE (km)
50
U.
1 O0
150
200
Distancefrom NW to SE (km)
[............•...........•....••..•.••••••••••;:•...•..•.••.•
.....• .....•...l........••-•.••<<••••..•••q•
[...............
•••••••,,•.•••.•••.•••••
............
,W•.x••.,.•••::•
.. •.,..,
,•••..,...............,
•••: .••••
114"
113 ø
•10
..
0
112'
c.
' •0 ....
•b6 ' ' ' •d
' ' ' 266
DistancefromNW to SE (km)
111'
•10
110"
• •
0
•
d
,50;................
•00
•50
200
DistancefromNW to SE (km)
Figure 11. Focal depthsplottedwith Te for (a) the Borah Peak aftershocksequence,(b) southwestern
Yellowstone and Hebgen Lake, (c) central Yellowstone, and (d) northeasternYellowstone. See Plate 1 for
location of study area.
17,960
LOWRY
AND
SMITH:
STRENGTH
AND RHEOLOGY
40.]Range
ofbrittle-ductile
depth
30.1 ./
,,/
OF THE CORDILLERA
/ [ 40
/
1•20
•
• • 20q
• IJt JI] •- •/,•• ,g*4
•/ . •lbackground
195re%depthøf
Iseismicity I[
0
10
20
30
40
Elasti
cThickess
Te,
Large
Cordilleran
Ea•hq•kes:
• =Ms• 7.0;• =5.5• Ms• 6.9
ISB Seismicity:* = back•ound seismicity;:.:= aftershocks
ofM > 7 eqs
Figure 12. Focal depthsof large e•thqu•es &om the westernU.S. Cordillera [after Doser and Smith,
1989], ISB backgroundseismicity,and aftershocksof M > 7 eventsplotted as a •nction of effective elastic
thickness. The relationshipof T• to brittle-ductile depth is also shown for a r•ge of possiblestressstates
seismicity yields a positive correlation of depthsto Te to >
logical layers that compose the lithosphere. With careful
considerationof complementarydata sets,it shouldbe possirelation suggeststhat maximum focal depth dependson some ble to gather new clues regarding genesisof the lithosphere
from mappedTe (Plate l a), even in areaswherebasementrocks
of the sameparametersthat define Te.
Of the 5144 events shown, only a handful of the back- are concealedby sedimentaryor volcaniccover. The high Te
ground earthquakesand a few of the very largest earthquakes of the interior Columbia Basin is a good example (Plate l a).
(and their aftershocks)are deep enoughto associatewith brit- Given the anomalouslylow heat flow (Figure 1) and proximity
tle-ductile control, however. The deepest backgroundearth- of Cascadiasubduction,one could be temptedto interprethigh
quakes are almost certainly outliers, probably representing strengthas an artefactof thermal perturbationby cold subducthigh-strengthregions too small to be resolved by Teestima- ing lithosphere. However, upper mantle P velocity indicatesa
tion. Hence maximum focal depths of background seismicity steepeastwarddip of the modernJuande Fuca slab [Humphreys
are defined by some shallower theological transition, most and Dueker, 1994], and modelingof Cenozoicsubductionsugprobably the stick-slip to stable sliding transition argued for gests negligible thermal perturbation east of central
by Tse and Rice [1986]. Enough of the large earthquakesfall Washington state since 20 Ma [Severinghaus and Atwater,
into the range of possible brittle-ductile control to make the 1991]. The feature is near a known accretionaryterrane,howpossibility somewhat harder to dismiss. Many studies have ever: the Blue Mountains terrane, correspondingto lower
noted that some large mainshocks initiate at significantly strengthregions to the south and east, derives from Paleozoic
greater depth than their aftershocksor nearby backgroundac- island arc volcanism [e.g., White et al., 1992]. The Blue
tivity. Das [1982] arguedthat the increasein strain rate during Mountainsterraneis apparentlya small exotic heraldinga
large earthquakes,
approaching10-4tO 10-3s-1,hasthe effect Precambrian terrane to the northwest, currently obscuredby
of deepening the brittle-ductile transition, but while that ex- Columbia basalts. The postulatedPrecambrianorigin is supplains why ruptures might propagate to great depth, it does portedby high upper mantleP velocityto depthsof 200 km
not explain why large events should initiate there. Instead, [Humphreysand Dueker, 1994].
nucleation6f largerupturesbelowthe depthof regional-scale In another example, Farmer and DePaolo [1983] suggest
dislocation creep may result from stable slip heterogeneities that a discontinuity in the crustal contamination of granites
induced by localized, compositionallydefined strengthvaria- (the dashedgrey line (1) in Plate l a) correspondsto the westtions along the fault plane. This hypothesisis also consis- ern edge of the Paleozoic craton, composedof lithosphereattent with observationsthat large rupture eventsoften nucleate tenuated by latest Proterozoic rifting. However, the central
in or near high velocity zones [e.g., Lees and Malin, 1990; Basin-Range is distinguishedfrom lithosphereto the east by
Lees and Nicholson, 1993; Foxall et al., 1993].
higher œSr[Farmer and DePaolo, 1983], and Te(Plate la) indicates distinct lithospheric blocks as opposedto a gradational
Other Implications of Te
decreasein strengthbetweenthe ColoradoPlateauand the cenAn important result of this analysisis that effective elastic tral Basin-Range. The high œSrisotopic domain also exhibits
thicknessstrongly dependson bulk compositionof the rheo- relative high topography and greater crustal thicknessthan
99.5% confidence, with correlation coefficient 0.3.
That cor-
LOWRY AND SMITH: STRENGTH AND RHEOLOGY
the surroundingBasin-Range,and it is the Tertiary locusof an
unextendedbelt sandwichedby large detachmentsto east and
west [e.g., Axen et at., 1993]. Hence we propose that older
crustal material underlying the Paleozoic geoclinal sediments
between lines (1) and (2) of Plate l a is attached to pieces of
exotic lithosphere, presumably terranes that accreted prior to
the rifting event, that were then rafted away from the craton by
early Paleozoic extension.
Also, a large area surroundingthe active Yellowstone volcanic field
indicates
lower
crustal flow
rather
than the mantle
rheology typical of the remainder of the middle Rocky
Mountains. A recent study by Byrd et at. [1994] indicates
most of the 2 to 3 km offset on the Teton fault, 20 km south of
Yellowstone, has occurred since -2 Ma, coinciding with the
approximate time of arrival of the Yellowstone hotspot
[Smith and Braite, 1994]. This would imply that the sourceof
the volcanismmay also be responsiblefor mobilizationof the
lower crust, reducedTe and the onsetor accelerationof extension. However, this area also experiencedeffusive andesitic
volcanismduring the Laramideorogenyat -•50 Ma. Therefore,
while thermal perturbation and mass flux associated with
Yellowstone
volcanism
contribute
to the current extensional
OF THE CORDILLERA
17,961
mal faulting. An improved understandingof these relationshipsis achievedby parameterizationusing a yield strength
envelope. Comparisonswith regional heat flow, seismically
defined crustal properties, and principal stress orientations
confirms that T• is most sensitive to temperature,stressstate,
and composition at depths of ductile flow. Various tectonic
provinces accommodateisostatic responseat different levels
of the crustor upper mantle,dependingprincipallyon the age
of the lithosphere.Te in the Archeanmiddle Rocky Mountains
is controlled primarily by ductile flow in the mantle. The
Early Proterozoic-agedColorado Plateau exhibits lower crustal
control of long-term T•. Latest Proterozoic-agedlithosphere
that has experienced significant Cenozoic extension,including most of the Basin-Range,the variousvolcanicprovinces
(with the exception of the interior Columbia Basin), and the
northern Rocky Mountains, accommodatesflexure via middle
to upper crustalductile flow. Apparent lower crustalcontrol of
T• in the interiorColumbiaBasinandthe centralBasin-Range
likely corresponds to exotic terranes of older continental
lithosphere.
Azimuthal variation of T• exhibits a relationshipto princi-
pal horizontal stressorientation. Anisotropy of T• aligns
episode,the lithospherenear Yellowstoneprobablywas never
with stress orientations
as stable as cratonic lithosphereto the south and east.
Teprovides some insight into ISB extensionas well. There
is ample evidencethat a significantfraction of Cordillerantec-
northeast Basin-Range, eastern Snake River Plain, and
tonism is indirectly related to plate interactions,especially
subduction processes [e.g., Wernicke et at., 1987;
Severinghaus
and
Atwater,
1990; Axen et at.,
1993].
Nevertheless the spatial relationships of Cordilleran extension and ISB seismicityto Te (Plate 1) suggeststrongcontrol
by other factors. Extensional strength of the lithosphereis
controlled by the same parametersthat determine Te[e.g.,
inferred
from stress indicators
in the
Colorado Plateau, but is less consistent with stress orienta-
tions from the very dynamic Yellowstone volcanic system.
Azimuthal variation of Te also implies an-90 ø rotation of
stressorientation across the Wasatch Front. Comparison of
T• with maximum earthquakefocal depthsindicatesthat background seismic depths are too shallow for control by the brittle-ductile transition, but some of the largest (M > 7) earthquakes and their aftershocksinitiate within the range of possible brittle-ductile depths. Finally, qualitative relationships
between T• and the spatial distributionof seismicityand active
faulting suggest that seismicity and tectonism in the
Intermountain seismic belt are related more intimately to local
buoyancy than to interactionsat distant plate boundaries.
Kusznir and Park, 1987], so the lithosphereshouldbe weakest
approximatelywhere T•is lowest. It is expected[e.g., Chen
and Motnar, 1983] that much of the strain responseto a uniform regional stress(suchas an intraplatestressoriginatingat
a plate boundary)will be absorbedby the weakestportionsof
the lithosphere. However, there is relatively little seismicacAcknowledgments. This manuscriptwas greatlyimprovedby careful
tivity and contemporary strain where T• is lowest: the main reviews by E. Humphreys, C. Lowe, and A. G. Jones. Animated disISB seismicity (Plate 1a) is found instead in a zone of transi- cussionswith J. O. D. Byrd, R. L. Bruhn, R. W. Simpson,and J. M.
tion from low to high Te. Hence the forces responsiblefor
Bartley aided in crystallizationof someof the ideaspresented.We are
Cordilleran extension must originate locally, e.g., from
deeplyindebtedto D. Forsythand T. Bechtelfor their developmentof
nearby buoyancy distributions,rather than in the far field.
the coherencemethodandto M. Zuber for providingaccessto Bechtel's
originalcomputercode. This studyowesmuchto the dedicatedwork of
Zoback [1992] infers a local, buoyancy-related origin for
the University of Utah SeismographStations crew, particularly W.
extensional stressesfrom the relationship of topography to
Arabasz,J. C. Pechmann,and S. Nava. Pechmannalsogaveinvaluable
global distribution of stress. Artyushkov [1973] similarly
suggeststhat the largest intraplate stressesare likely to result advice on the quality of earthquakehypocenterlocations. Thanks also
to D. Blackwell for providing the regional heat flow data, to M. L.
from lateral buoyancy variations, i.e., changes in the vertiZoback for the stress orientation data, to J. Louie for the combined
cally integrated density column of the lithosphere. Much of
California-Nevadaearthquakecatalog,and to Rick Saltusfor the USGS
the ISB (Plate l a) correspondsto a major discontinuity in
topographyand Bouguergravity griddeddata sets. Supportwas prolithospheric buoyancy. Buoyancy variations alone are not vided by the National ScienceFoundationthroughgrantsEAR 89-04473
enough to explain the seismicityhowever: if they were, seis- and EAR 92-19694.
micity would be equally vigorous at passive continental margins. The important difference is that crustal ductile behavior
occurs throughout the ISB. Hence the concentrationof seis- References
micity in the ISB resultsfrom large stressesthat are accommodated by flux of low-viscosity crustal material acrossbuoy- Angelier, J., Tectonicanalysisof fault slip data sets,J. Geophys.Res.,
89, 5835-5848, 1984.
ancy gradients [e.g., Bird, 1991].
Arabasz, W. J., J. C. Pechmann, and E. D. Brown, Observational seis-
Conclusions
T• of the western U.S. Cordillera exhibits qualitative relationshipsto heat flow, lithosphericage, and Quaternarynor-
mology and the evaluation of earthquake hazards and risk in the
Wasatch Front area, Utah, USGS Prof Pap., 1500-D, 1990.
Armin, R. A., and L. Mayer, Subsidenceanalysis of the Cordilleran
miogeocline: Implications for timing of late Proterozoicrifting and
amountof extension,Geology,11,702-705, 1983.
17,962
LOWRYAND SMITH:STRENGTH
ANDRHEOLOGY
OFTHECORDILLERA
Artyushkov,
E. V., Stresses
in the lithosphere
caused
by crustalthicknessinhomogeneities,
J. Geophys.
Res.,78, 7675-7708, 1973.
Atwater,T., Implications
of platetectonics
for Cenozoic
tectonicevolution of westernNorth America, Geol. Soc.Am. Bull., 81, 3513-3536,
1970.
Axen, G. J., W. J. Taylor, and J. M. Bartley, Space-timepatternsand
tectonic controlsof Tertiary extensionand magmatismin the Great
Basin of the western United States, Geol. Soc. Am. Bull., 105, 57-76,
1993.
Bechtel,T. D., Mechanismsof isostaticcompensation
in East Africa and
North America, Ph.D. dissertation, 247 pp., Brown Univ.,
Providence, R. I., 1989.
Bechtel, T. D., D. W. Forsyth, V. L. Sharpton, and R. F. Grieve,
Variations in effective elastic thickness of the North American litho-
sphere,Nature, 343, 636-638, 1990.
Benz, H. M., R. B. Smith, and W. D. Mooney, Crustal structureof the
northwesternBasin-Range province from the 1986 Program for
Array SeismicStudiesof the ContinentalLithosphereseismicexper-
iment,J. Geophys.
Res.,95, 21,823-21,842,1990.
Bird, P., Lateral extrusionof lower crustfrom underhigh topography,in
the isostaticlimit, J. Geophys.Res.,96, 10,275-10,286,1991.
Bird, P., and X. Kong, Computersimulationsof California tectonics
confirmlow strengthof majorfaults,Geol.Soc.Am. Bull., 106, 159174, 1994.
crust formation and mantle evolution in the Proterozoic, Nature, 291,
193-196, 1981.
Dixon, T. H., S. Robaudo,and J. Lee, Constraintson presentday Basin
and Range deformationfrom spacegeodesy,Tectonics,in press,
1995.
Doser, D. I., and R. B. Smith, An assessmentof sourceparametersof
earthquakesin the Cordillera of the western United States,Bull.
Seismol.Soc. Am., 79, 1383-1409, 1989.
Dzurisin, D., J. C. Savage,and R. O. Fournier,Recentcrustalsubsidence at Yellowstone Caldera, Wyoming, Bull. Volcanol.,52, 247270, 1990.
Eddington,P. K., R. B. Smith, and C. Renggli, Kinematicsof BasinRange intraplateextension,in ContinentalExtensionalTectonics,
editedby M.P. Coward,J. F. Dewey, andP. L. Hancock,Geol. Soc.
Spec.Publ.London,28, 371-392, 1987.
Farmer, G. L., and D. J. DePaolo, Origin of Mesozoic and Tertiary
granite in the western United States and implicationsfor preMesozoic crustalstructure,1; Nd and Sr isotopicstudiesin the geocline of the northernGreat Basin,J. Geophys.Res.,88, 3379-3401,
1983.
Fleck, R. J., and R. E. Criss,Strontiumand oxygenisotopicvariationsin
Mesozoicand Tertiary plutonsof centralIdaho, Contrib. Mineral.
Petrol., 90, 291-308, 1985.
Forsyth, D. W., Finite extension and low-angle normal faulting,
Bjarnason,I. T., and J. C. Pechmann,Contemporary
tectonicsof the
Geology,20, 27-30, 1992.
WasatchFront region, Utah, from earthquakefocal mechanisms, Fountain,D. M., and N. I. Christensen,Compositionof the continental
Bull. Seisrnol.Soc. Am., 79,731-755, 1989.
crustanduppermantle: A review,in Geophysical
Frameworkof the
Blackwell, D. D., and J. L. Steele, Geothermal map of North America,
scale1:5,000,000, Geol. Soc. of Am., Boulder,Colo., 1992.
Blackwell, D. D., J. L. Steele, and L. S. Carter, Heat-flow patternsof
the North American continent: A discussionof the geothermalmap
of North America, in Neotectonicsof North America, Decade Map,
ContinentalUnited States,editedby L. C. PakiserandW. D. Mooney,
Geol. Soc. Am. Mern., 172, 711-742, 1989.
Foxall, W., A. Michelini, and T. V. McEvilly, Earthquaketravel time
tomography
of the southernSantaCruzMountains:Controlof fault
ruptureby lithologicalheterogeneity
of the SanAndreasfault zone,
vol. 1, editedby D. B.Slemmons,
E. R. Engdahl,M.D. Zoback,and
J. Geophys.
Res.,98, 17,691-17,710,1993.
D. D. Blackwell, pp. 423-436, Geological Society of America, Gephart,J. W., andD. W. Forsyth,An improvedmethodfor determinBoulder, Colo., 1991.
ingtheregionalstresstensorusingearthquake
focalmechanism
data:
Bond, G. C., and M. A. Kominz, Constructionof tectonicsubsidence
Applicationto the San Fernandoearthquakesequence,
J. Geophys.
curves for the early Paleozoicmiogeocline,southernCanadian
Res., 89, 9305-9320, 1984.
Rocky Mountains:Implicationsfor subsidence
mechanisms,
age of
Goetze, C., and B. Evans, Stressand temperaturein the bendinglithobreakup,and crustalthinning, Geol. Soc.Am. Bull., 95, 155-173,
sphereas constrained
by experimentalrock mechanics,Geophys.J.
1984.
Brace,W. F., and D. L. Kohlstedt,Limits on lithosphericstressimposed
by laboratoryexperiments,
J. Geophys.
Res.,85, 6248-6252, 1980.
Braile, L. W., R. B. Smith, G. R. Keller, R. M. Welch, and R. P. Meyer,
Crustal structure across the Wasatch Front from detailed seismic re-
fractionstudies,J. Geophys.Res.,79, 2669-2677, 1974.
Braile, L. W., R. B. Smith, J. Ansorge,M. R. Baker,M. A. Sparlin,C.
Prodehl,M. M. Schilly, J. H. Healy, St. Mueller, and K. H. Olsen,
The Yellowstone-SnakeRiver Plain seismic profiling experiment:
Crustal structureof the easternSnakeRiver Plain, J. Geophys.Res.,
87, 2597-2609, 1982.
Braile, L. W., W. J. Hinze, R. R. B. von Frese,and G. R. Keller, Seismic
R. Astron. $oc., 59, 463-478, 1979.
Gough,D. I., andJ. S. Bell, Stressorientations
from oil well fracturesin
Alberta and Texas, Can. J. Earth Sci., 18, 638-645, 1981.
Hansen,F. D., and N. L. Carter,Creepof selectedcrustalrocksat 1000
MPa, Eos Trans. AGU, 63, 437, 1982.
Hecker,S., Quaternarytectonicsof Utah with emphasison earthquakehazard characterization,Utah Geol. Surv. Bull., 127, 157 pp., 1993.
Hickman, S. H., and M.D. Zoback, The interpretationof hydraulic
fracturingprocess-timedata for in situ stressdetermination,in
HydraulicFracturingStressMeasurements,
editedby M.D. Zoback
andB.C. Haimson,pp. 44-54, NationalAcademyPress,Washington,
D.C., 1983.
propertiesof the crustand uppermostmantleof the conterminous Hill, D. P., and L. C. Pakiser,Crustalstructurebetweenthe NevadaTest
United Statesand adjacentCanada,in GeophysicalFrameworkof
Site and Boise, Idaho, from seismicrefractionmeasurements,
in The
the Continental United States, edited by L. C. Pakiser and W. D.
EarthBeneaththe Continents,
Geophys.
Monogr.Ser.,vol. 10, edited
Mooney,Geol.Soc.Am. Mem.,172, 655-680, 1989.
by J. S. Steinhart
andT. J. Smith,pp.391-419,AGU, Washington,
D.
C., 1966.
Byrd, J. O. D., R. B. Smith,andJ. W. Geissman,
Neotectonics
of the
Tetonfault,Wyoming,J. Geophys.
Res.,99, 20,095-20,122,1994.
Humphreys,E. D., and K. G. Dueker,WesternU.S. uppermantle
Chen,W., andP. Molnar, Focaldepthsof intracontinental
andintraplate
structure,
J. Geophys.
Res.,99, 9615-9634, 1994.
earthquakes
and their implicationsfor the thermaland mechanical Jackson,
S. M., I. G. Wong,G. S. Carpenter,D. M. Anderson,
andS. M.
Martin, Contemporary
seismicityin the easternSnakeRiver Plain,
properties
of thelithosphere,
J. Geophys.
Res.,88, 4183-4214,1983.
Condie, K. C., Archean Greenstone Belts, Elsevier Scientific, New
Idaho, based on microearthquakemonitoring,Bull. Seisrnol.Soc.
York, 1981.
Das, S., Appropriateboundaryconditions
for modelingvery longearthquakes and physical consequences,
Bull. Seismol.Soc. Am., 72,
1911-1926, 1982.
DeMets, C., R. G. Gordon, S. Stein, and D. F. Argus,A revisedestimate
of Pacific-NorthAmericamotionand implicationsfor westernNorth
America plate boundaryzone tectonics,Geophys.Res. Lett., 14,
911-914, 1987.
DePaolo,D. J., Neodymium
isotopes
in the Colorado
FrontRangeand
Am., 83, 680-695, 1993.
Jordan,T. H., Compositionand development
of the continentaltectosphere,Nature, 274, 544-548, 1978.
Jordan,T. H., Continentsas a chemicalboundarylayer,Philos.Trans.R.
Soc. LondonA, 301,359-373, 1981.
Kirby, S. H., Rheologyof the lithosphere,
Rev. Geophys.,
21, 14581487, 1983.
Kusznir,N.J., andR. J. Park,The extensional
strengthof the continental
lithosphere:Its dependence
on geothermal
gradient,andcomposition
LOWRY
AND
SMITH:
STRENGTH
AND
and thickness,in ContinentalExtensionalTectonics,edited by M. P.
Coward, J. F. Dewey and P. L Hancock, Geol. $oc. Spec. Publ.
London, 28, 22-52, 1987.
Leeman, W. P., J. S. Oldow, and W. K. Hart, Lithosphere-scalethrusting in the western U. S. Cordillera as constrainedby Sr and Nd isotopic transitionsin Neogene volcanic rocks, Geology, 20, 63-66,
1992.
Lees, J. M., and P. E. Malin, Tomographicimages of P wave velocity
variation at Parkfield, California, J. Geophys. Res., 95, 21,79321,804, 1990.
Lees, J. M., and C. Nicholson, Three-dimensionaltomography of the
1992 southernCalifornia earthquakesequence: Constraintson dynamic earthquakerupture?Geology,21,387-390, 1993.
Lehman, J. A., R. B. Smith, M. M. Schilly, and L. W. Braile, Upper
crustal structureof the Yellowstone caldera from delay time analysesand gravity correlations,J. Geophys.Res.,87, 2713-2730, 1982.
Lowry, A. R., and R. B. Smith, Flexural rigidity of the Basin-RangeColoradoPlateau-Rocky Mountain transitionfrom coherenceanalysis of gravity and topography,J. Geophys.Res., 99, 20,123-20,140,
1994.
Lyon-Caen, H., and P. Molnar, Constraints on the structure of the
Himalaya from an analysisof gravity anomaliesand a flexural model
of the lithosphere,J. Geophys.Res.,88, 8171-8192, 1983.
Manduca,
C. A., L. T. Silver,andH. P. Taylor,87Sr/86Sr
and180/160
isotopic systematicsand geochemistryof granitoid plutonsacrossa
steeply-dippingboundarybetweencontrastinglithosphericblocks in
western Idaho, Contrib. Mineral. Petrol., 109, 355-372, 1992.
Martinez, L. J., C. M. Meertens, and R. B. Smith, GPS surveys of the
Wasatch fault zone, Utah, Eos Trans. AGU, 75(44), Fall Meeting
suppl., 179-180, 1994.
McCamy, K., and R. P. Meyer, A correlationmethodof apparentvelocity measurement,J. Geophys.Res.,69, 691-699, 1964.
McGarr, A., and N. C. Gay, State of stressin the earth's crust, Annu.
Rev. Earth Planet. $ci., 6, 405-436, 1978.
McNutt, M. K., Lithospheric flexure and thermal anomalies, J.
Geophys.Res., 89, 11,180-11,194, 1984.
McNutt, M. K., and H. W. Menard, Constraintson yield strengthin the
oceaniclithospherederived from observationsof flexure, Geophys.
J., 71,363-394, 1982.
Meertens, C. M., R. B. Smith, and D. W. Vasco, Kinematics of crustal
deformationof the Yellowstonehotspotusing GPS, Eos Trans.AGU,
74(43), Fall Meeting suppl.,63, 1993.
Minster, J. B., and T. H. Jordan, Vector constraints on western U.S. deformation from spacegeodesy,neotectonics,and plate motions,J.
Geophys.Res., 92, 4798-4804, 1987.
Mueller, S., and M. Landisman, An example of the unified method of
interpretationfor crustal seismicdata, Geophys.J. R. Astron. $oc.,
23, 365-371, 1971.
Nakamura, K. K., Volcanoes as possibleindicators of tectonic stress
orientation- principle and proposal,J. Volcanol. Geotherm.Res., 2,
1-16, 1977.
O'Hara, N. W., and P. L. Lyons, New map updates gravity data,
Geotimes, 28, 22-27, 1983.
Parry, W. T., and R. L. Bruhn, Fluid inclusionevidencefor minimum 11
km vertical offset on the Wasatch Fault, Utah, Geology, 15, 67-70,
1987.
Peyton,S. L., Contemporarytectonicsof the Yellowstone-HebgenLake
region from earthquakefocal mechanismsand stressfield inversion,
M.S. thesis,90 pp., Univ. of Utah, SaltLake City, 1991.
Roller, J. C., Crustal structurein the easternColoradoPlateausprovince
from seismic-refraction measurements, Bull. Seismol. Soc. Am., 55,
107-119, 1965.
Savage, J. C., M. Lisowski, and W. H. Prescott,Strain accumulation
acrossthe Wasatch fault near Ogden, Utah, J. Geophys. Res., 97,
2071-2083, 1992.
Severinghaus,J., and T. Atwater, Cenozoicgeometryand thermal state
of the subductingslabsbeneathwestern North America, in BasinRange Extensional Tectonics Near the Latitude of Las Vegas,
RHEOLOGY
OF THE CORDILLERA
17,963
Nevada, editedby B. P. Wernicke, Mem. Geol. Soc.Am., 176, 1-22,
1990.
Sheriff, S. D., and M. C. Stickney, Crustal structureof southwestern
Montana
and east-central
Idaho:
Results of a reversed seismic re-
fractionline, Geophys.Res.Lett., 11,299-302, 1984.
Sibson,R. H., Frictional constraintson thrust,wrench and normal faults,
Nature, 249, 542-544, 1974.
Sibson,R. H., Fault zonemodels,heatflow, andthe depthdistributionof
earthquakes in the continental crust of the United States, Bull.
Seismol. Soc. Am., 72, 151-163, 1982.
Simpson,R. W., R. C. Jachens,R. J. Blakely, and R. W. Saltus, A new
isostaticresidual map of the conterminousUnited Stateswith a discussion on the significance of isostatic residual anomalies, J.
Geophys.Res., 91, 8348-8372, 1986.
Smith, R. B., and W. J. Arabasz,Seismicityof the Intermountainseismic
belt, in Neotectonicsof North America,DecadeMap, vol. 1, editedby
D. B. Slemmons,E. R. Engdahl, M.D. Zoback, and D. D. Blackwell,
pp. 185-228, GeologicalSocietyof America,Boulder,Colo., 1991.
Smith, R. B., and L. W. Braile, The Yellowstone hotspot,J. Volcan.
Geotherm. Res., 61,121-188,
1994.
Smith, R. B., and R. L. Bruhn, Intraplate extensionaltectonicsof the
eastern Basin-Range: Inferences on structural style from seismic
reflection data, regional tectonics, and thermal-mechanicalmodels
of brittle-ductile deformation, J. Geophys. Res., 89, 5733-5762,
1984.
Smith, R. B., and M. L. Sbar, Contemporarytectonicsand seismicityof
the westernUnited Stateswith emphasison the Intermountainseismic
belt, Geol. Soc. Am. Bull., 85, 1205-1218, 1974.
Sparlin, M. A., L. W. Braile, and R. B. Smith, Crustal structureof the
eastern Snake River Plain determined from ray-trace modeling of
seismicrefraction data, J. Geophys.Res., 87, 2619-2633, 1982.
Tse, S. T., and J. R. Rice, Crustalearthquakeinstabilityin relationto the
depth variation of frictional slip properties, J. Geophys. Res., 91,
9452-9472,
1986.
Turcotte, D. L., and G. Schubert, Geodynamics: Applications of
ContinuumPhysicsto Geological Problems,JohnWiley, New York,
1982.
Wernicke, B. P., Cenozoic extensional tectonics of the U.S. Cordillera,
in The Cordilleran Orogen: ConterminousU.S., Geol. of N. Am., vol.
G-3, edited by B.C. Burchfiel, P. W. Lipman, and M. L. Zoback, pp.
553-581, Geological Society of America, Boulder, Colo., 1991.
Wernicke, B. P., R. L. Christiansen,P. C. England, and L. J. Sonder,
Tectonomagmatic evolution of Cenozoic extension in the North
American Cordillera, in ContinentalExtensionalTectonics,edited by
M.P. Coward, J. F. Dewey, and P. L. Hancock, Geol. Soc. Spec.
Publ. London, 28, 203-221, 1987.
White, J. D. L., D. L. White, T. Vallier, G. D. Stanley, and S. R. Ash,
Middle Jurassicstratalink Wallowa, Olds Ferry, and Izee terranesin
the accreted Blue Mountains island arc, northeastern Oregon,
Geology,20, 729-732, 1992.
Wong, I. G., and J. R. Humphrey, Contemporaryseismicity, faulting,
and the state of stress in the Colorado Plateau, Geol. Soc. Am. Bull.,
101, 1127-1146, 1989.
Zoback, M. L., First- and second-orderpatternsof stressin the lithosphere: The world stressmap project,J. Geophys.Res., 97, 11,70311,728, 1992.
Zoback, M. L., and M.D. Zoback, Tectonic stressfield of the continental United States, in Geophysical Framework of the Continental
United States, edited by L. C. Pakiser and W. D. Mooney, Mem.
Geol. Soc. Am., 172, 523-539, 1989.
Zoback, M.D., et al., State of stressin the San Andreas fault system,
Science, 238, 1105-1111, 1987.
A. R. Lowry and R. B. Smith, Department of Geology and
Geophysics,College of Mines and Earth Sciences,Universityof Utah,
Salt Lake City, UT 84112-1183. (e-mail: [email protected])
(Received September2, 1994; revisedFebruary21, 1995;
accepted February28, 1995.)