Elysium Region, Mars: Tests of Lithospheric Loading Models for the

JOURNAL
OF GEOPHYSICAL
RESEARCH,
VOL. 91, NO. Bll,
PAGES 11,377-11,392, OCTOBER
10, 1986
Elysium Region, Mars' Tests of Lithospheric Loading Models
for the Formation
of Tectonic
Features
J. LYNN HALL AND SEAN C. SOLOMON
Departmentof Earth, Atmospheric,
and PlanetarySciences,
Massachusetts
Instituteof Technology,
Cambridge
JAMES W. HEAD
Departmentof GeologicalSciences,
BrownUniversity,Providence,
RhodeIsland
The.ce.c.c•ncl
largestvolcanicprovince
on Mar• liesin the E!?ium regionI•ike the largerTharsis
province,
Elysiumis markedby a topographic
riseanda broadfreeair gravityanomalyandalsoexhibits
a complexassortment
of tectonicand volcanicfeatures.We testthe hypothesis
that the tectonicfeatures
in the Elysiumregionare the productof stresses
producedby loading of the Martian lithosphere.We
considerloadingat threedifferentscales:local loadingby individualvolcanoes,regionalloadingof the
lithospherefrom aboveor below,and quasi-globalloadingby Tharsis.A comparisonof flexuralstresses
with lithosphericstrengthand with the inferred maximum depth of faulting confirmsthat concentric
grabenaroundElysiumMons can be explainedas resultingfrom local flexureof an elasticlithosphere
about 50 km thick in responseto the volcanoload. Volcanicloadingon a regionalscale,however,leads
to predictedstresses
inconsistentwith all observedtectonicfeatures,suggesting
that loadingby widespreademplacement
of thick plainsdepositswasnot an importantfactorin the tectonicevolutionof the
Elysiumregion.A numberof linearextensional
featuresorientedgenerallyNW-SE may havebeenthe
result of flexural uplift of the lithosphereon the scale of the Elysium rise. The global stressfield
associatedwith the supportof the Tharsisrise appearsto have influencedthe developmentof many of
thetectonicfeatures
in the Elysiumregion,includingCerberusRupesandthe systems
of ridges,ineastern
and westernElysium.The comparisons
of stressmodelsfor Elysiumwith the preservedtectonicfeatures
supporta succession
of stressfieldsoperatingat differenttimesin the region.While the orderin which
those stressfields operated cannot be determinedfrom presentgeologicalobservations,thermal and
mechanical
arguments
favorthe hypothesis
that any flexuraluplift of the lithosphereby a mantlethermal
anomalypreceded
or occurredcontemporaneously
with emplacement
of thelargestvolcanicloads.
INTRODUCTION
The tectonichistory of the planet Mars has beendominated
by the formation and evolution of several major volcanic
provinces,notably those in the regions of Tharsis and Elysium. The Tharsis province, becauseit is the largest, has received the most attention. Two general categoriesof models
have been proposedto explain the topography, gravity, and
tectonic features of the Tharsis region. According to the first
type of model, uplift of ancientlithosphereby a chemicalor
thermal anomaly in the crust or mantle createda broad topographic dome and led to widespreadfracturing and to the
emplacementof relatively thin volcanicplains units and isolated volcanic shields[Hartmann, 1973; Cart, 1974; Sleepand
Phillips, 1979; Wise et al., 1979a, b; Plescia and Saunders,
1980]. This uplift may have occurred isostatically or by a
flexural uplift of the lithosphere[Banerdt et al., 1982]. In the
secondtype of modelthe Tharsisrisewascreatedprimarily by
volcanicconstruction,
and the tectonicfeaturesof the province are a signatureof the responseof the lithosphereto loading by these volcanicunits [Solomonand Head, 1982; Willemann and Turcotte, 1982; Banerdt et al., 1982]. The distribution of preserved tectonic features in the Tharsis region
appearsto rule out flexural uplift as a significantsourceof
lithosphericstressand insteadfavorsdistinctepisodesof local
isostaticsupport of the Tharsis rise and partial support of the
downward load by the finite strengthof the Martian lithosphere[Banerdt et al., 1982]. The temporal ordering of these
episodesis not presentlyresolvablefrom geologicalobservations ['Banerdtet al., 1982; Sleepand Phillips, 1985].
Like the Tharsis region, the Elysium volcanic province is
marked by both a topographicriseand a broad free air gravity anomaly[Sjogren,1979;danleandRopers,1983]. The Elysium region also exhibits a complex assortmentof tectonic
and volcanic features.In this paper we test whether the competingmodelsfor the originof tectonicfeaturesin Tharsiscan
be distinguishedon the basisof how well they accountfor the
tectonic evolution of Elysium. Specifically,we compare the
characteristicsof tectonic featuresin the Elysium region with
the stressfields predicted both by volcanic loading and by
uplift of the Martian lithosphere.
We begin with a brief descriptionof the physiographicfeatures in the Elysium region which are of probable tectonic
origin.We then test the hypothesisthat thesefeaturesare the
productof loadingof the lithosphere.We considerloadingat
three different scales: local volcanic loading (individual
shields),regionalloading(at the scaleof ElysiumPlanitia) of
the lithospherefrom above(volcanicplains)or below(flexural
uplift),and quasi-globalloadingby Tharsis.By comparingthe
predictedstressfieldsfrom suchmodelswith the distribution
and orientaticn of tectonic features, we evaluate the relative
importanceof' loading at thesedifferentscales,we constrain
the mechanicalpropertiesof the Martian lithosphere,and we
offer a further step toward a general understandingof the
evolutionof major volcanicprovinceson Mars.
GEOLOGIC SETTING
Copyright 1986by the AmericanGeophysicalUnion.
Paper number 5B5794.
0148-0227/86/005B- 5794$05.00
The Elysiumregion(Figure 1) is the secondlargestvolcanic
provinceon Mars [Cart, 1973; Malin, 1977]. It consistsof a
broad topographichigh, 2400 by 1700 km in extent and centered at about 25øN, 212øW, rising about 4 km above the
11,377
11,378
HALL ET AL.' TECTONICSOF THE ELYSIUMREGION,Mars
240øW
I
230 ø
I
I
220 ø
I
210 ø
I
I
I
200 ø
I
I
190ø
I
I
180,•0o
N
!i Hecates
Tholus
Phlegra
Montes
'"•
Hephaestus
'•"k.• Fossae
Cerberus Rupes
Fig. 1. Sketchmap of the Elysiumregion,showingmajor volcanicconstructsand tectonicfeatures.Light linesindicate
graben and narrow linear features;darker areas denote wide depressionsand volcaniccalderas;notchedlines indicate
ridges;and dashedlinesoutline the edgesof the volcanoes.Identificationof eachfeaturewas madeusingU.S. Geological
Surveycontrolledphotomosaics(Elysiumquadrangleand portionsof Amenthesand Cebreniaquadrangles)and selected
Viking orbiter photographs.The approximateboundariesof ElysiumPlanitia, which generallycoincideswith the Elysium
topographic rise,are 0ø-40øN, 180ø-260øW.
6.1-mbar referenceequipotential surface [Batson et al., 1979].
On top of this rise sit the three volcanoesElysium Mons,
cipally thermal mechanismfor the origin of the present broad
topographic rise of Elysium is unlikely [Solomon and Head,
Hecates Tholus, and Albor Tholus. The volcanic and tectonic
19823.
A number of physiographic features in the Elysium region
characteristicsof the region have been discussedby Scott and
Allincham [1976], Malin [1977], and Mou•linis-Mark et al.
[1984].
On the basis of crater density, volcanic activity in the Ely-
sium region ceasedprior to the time of most recent activity in
the Tharsis region. The plains of Elysium are more densely
cratered than their Tharsis counterparts [Neukum and Wise,
1976; Malin, 1977]. Plescia and Saunders[1979] found that
the surface of Elysium Mons has fewer craters greater than
1-1.5 km in diameter per unit area than the surfacesof Hecates Tholus and Albor Tholus. Elysium Mons is also the
source of the youngest major lava flows in the region
[Mou•linis-Mark et al., 1984]. The time of last major volcanic
activity in the Elysium region is poorly constrained;the surfaces of the three volcanoes are estimated to have average ages
of at least 1 b.y. according to diverse models of the Martian
cratering flux [Plescia and Saunders, 1979], but isolated
younger eruptionshave also been postulated[Mouginis-Mark
et al., 1982].
A broad positive free air gravity anomaly is associatedwith
the Elysium region. The anomaly is nearly centeredover Elysium Mons but is much broader than the volcano itself [Sjo•lren, 1979]. Gravity models suggestthat the topography of
Elysium is isostaticallycompensatedat wavelengthsof 10002000 km Idarile and Ropers, 1983]. The fact that the topographic rise and gravity anomaly have persistedfor so long
after the cessation of volcanic activity suggeststhat a prin-
either are of tectonic origin or have likely been influenced by
tectonicstress
duringtheir•ormation(Figure1).An extensive
set of concentric fractures and graben encircle Elysium Mons
at distancesof approximately 150-350 km from the volcano
center. Some of these faults, particularly those to the west of
the volcano (Figure 2), are quite fresh and sharply defined and
apparently postdate the most recent volcanic units that they
cut. Others appear to have been partially buried by subsequent volcanic deposits(Figure 2, lower center). On the basis
of photogeologicmapping, Mou•linis-Mark et al. [1984] conclude that most of the circumferentialfracturing occurredafter
formation of the intermediate stage volcanic unit (the compound plains) but before the end of late stage effusive flank
activity on Elysium Mons. No comparable graben surround
the other two volcanoes, but these constructs have been partially buried by extensivelava flows from the vicinity of Elysium Mons [Mou•linis-Mark et al., 1984], so that any evidence
for early episodes of fracturing around Hecates Tholus and
Albor Tholus could have been obscured.
A number of linear depressionsof probable extensional
origin display a predominantly northwest-southeast trend.
Some of these features are sufficiently narrow so that they
show no discernible floor in Viking orbiter photographs
(Figure 3), while others have flat floors and widths up to 15
km (Figure 4). In the first category are the Elysium Fossae
(Figure 1) and a number of fault systems, including He-
HALL ET AL.' TECTONICSOF THE ELYSIUM REGION, Mars
11,379
Fig. 2. Viking orbiter view of concentric fractures and graben west of Elysium Mons. Graben both cut and are
embayed by flows of the youngestmajor volcanic plains unit [Mouginis-Mark et al., 1984]. Frame V0541A42' width of
image is 170 km.
phaestus Fossae [Schumm, 1974; Hiller, 1979], Cerberus
Rupes [Scott and Ailingham, 1976; Scott and Carr, 1978], and
the polygonal fracture set west-southwest of Elysium Mons.
The associationof the linear depressionsof Elysium Fossae
with sinuous rilles, braided channels, and other erosional features of northwest Elysium indicates that they have at least
been modified by volcanic or fluvial processes,a conclusion
that also applies to the wider, flat-floored depressions.The
consistent orientations of these features, however, are likely
indicative of, and may have been controlled by, the direction
of the regional stressfield at the time of their formation [Carr,
1980, 1981].
A system of ridges (Figure 5), similar to lunar mare ridges
[Maxwell, 1982; Chicarro et al., 1985], is found at the eastern
edge of Elysium Planitia (Figure 1), where the volcanic plains
material merges with the older terrain of the Phlegra Montes
[Gifford, 1981; De Hon, 1982]. The ridges are oriented approximately north-south, and the majority are located between latitudes 15ø-40øN and longitudes 180ø-205øW [G/fford, 1981; Chicarro et al., 1985]. There are also several maretype ridge segmentsextending approximately 600 km across
southwest Elysium Planitia and trending northeast-southwest
(Figure 1). In general, ridges are taken to be evidence of horizontal compression of near-surface material [Howard and
Muehlberger, 1973; Muehlberger, 1974; Lucchitta, 1976, 1977;
Sharpton and Head, 1982].
Also found in the Elysium region, primarily to the northwest of Elysium Mons, are a number of sinuous depressions.
Some of these features strongly resemble lunar sinuous rilles
[Carr, 1981], while others are wide and channellike, presum-
11,380
HALLETAL.' TECTONICS
OFTHEELYSIUMREGION,Mars
Fig. 3. Viking orbiter view of narrow linear depressionssouth of Elysium Mons. Frame V0844A21' width of image is
320 km.
ably erosional or fluvial in origin [Sharp, 1980; Carr, 1981;
Baker, 1982; Mouqinis-Mark et al., 1984]. Becausethesesinuous featuresdo not likely have a significanttectonic component to their origin, they are not consideredfurther here.
STRESS MODELS
Our working hypothesisis that the tectonic featuresof the
Elysium region are the result of stressescausedby loading of
the lithosphere from above or below. The hypothesis that
large-scaletectonic features associatedwith volcanic regions
of the terrestrial planets are products of the flexural response
to loading of the lithosphere has proven to be a fruitful approach for constraining the tectonic history of lunar mare
basins [Melosh, 1978; Solomonand Head, 1979, 1980; Comer
et al., 1979] and the regions surrounding individual Martian
volcanoes [Thurber and Toksoz, 1978; Comer et al., 1985]. As
noted earlier, this hypothesiscan also account for the distribution and origin of many of the tectonic features in the
Tharsis province [Willemann and Turcotte, 1982; Banerdt et
al., 1982; Sleep and Phillips, 1985].
We consider as specific models the load imposed by Elysium Mons, regional loading on the scale of Elysium Planitia
and the other volcanoes, and the quasi-global loading of the
Tharsis rise. For the loads of individual volcanoeswe employ
the theory of Brotchie and Silvester [1969] and Brotchie
[1971] for flexure of a thin elastic shell overlying an inviscid
fluid interior in responseto an axisymmetricload. Becausethe
flexureequation
is linear,thestresses
dueto anydistributed
load can be representedby the superimposedstressesdue to
an equivalent distribution of circular loads. We use this ap-
HALL ET AL..'TECTONICS
OF THE ELYSIUMREGION,Mars
11,381
'•:!/.:?"i
"k•-•:-":•L
• ..........
•.i:•,•:.i..:•...:
.....
i%•.,,.."
--•.... ;-.•.
•-:'•-,
....
** •:•,
-•--':
......
,.'::;
.....
:C.
,
,"'..•
l
i
[
"•' j7ll:•:•j•
•
*'l;k•7l
fw:"
"•'••}
•;<:*'""'
-.f•l:ll ll:•,
....
•ll
l,.lll•l.
•
*;.• ll}
;l'l:•
•
::a
I'[•l
I
.:l.-
'•
ll
,
71•
II•
•
;:l
•ll
ß
ll
ll•
......l;-':'t.
X:..;•*-R
......
';.(:-(}-*%
:*{;.
i;:.
?S•(}{....':--':-::
:-:'*:
.....
.
.
.. ." . ' .".....J.......
'..':....'.'.'..-...'
....
'L'....L.'Z.
"...
'.-j.
2.'......fJ'-'.'. '....... ..'..'......
. .'-. '..-..'....".
.....
'.1'.....'....2
........
Fig. 4. Viking orbiter view of a wide depressionwest-southwestof Elysium Mons. Its associationwith the s•stem of
concentric kactures around the volcano suggeststhat it is at least partl• tectonic in origin. Frame V0844A20; width of
image is 260 km.
proach to determine the stressdistribution predicted by regional loading models, including the models proposed by
danle and Ropers [1983] to fit line of sight gravity data. For
the Tharsis load we use the spherical harmonic solutions of
Banerdt et al. [1982] for displacement and stressin a thick
elasticsphericalshellgiven the topography and gravity.
taken to have a uniform thickness or flexural rigidity. Such
time-dependent effects as viscoelasticityand load growth are
ignored. Finally, loads estimatedfrom presenttopography and
gravity need not be correct representations of loads at the
time of formation
of ancient
tectonic
features.
The thin shell approximation is not a serioussourceof error
The elastic theories used in these stress calculations
are
for local loading problems [Melosh, 1978; Pullan and Lainclearly idealizations of actual mechanical behavior, and it is beck, 1981; Comer, 1983], even for lithosphereseveralhundred
important to recognizethe limitations of these simplified de- kilometers in thicknesssuch as that in the vicinity of Olympus
scriptions of the stress field. The local and regional stress Mons [Thurber and Toksoz, 1978; Comer et al., 1985]. The
models are based on thin shell theory and are valid only for
approximation of horizontally limited loads in the Brotchie
loads of horizontal dimension small compared with the plane- formulation of flexure presentsno problem for individual voltary radius.For all models,perfectelasticityis assumedfor the canoesbut may introduce a small error for loads on the scale
lithosphere.Further, for each type of model the lithosphereis of Elysium Planitia. The lithosphere, of course,is not perfectly
11,382
HALL ET AL.' TECTONICSOF THE ELYSIUMREGION,Mars
Fig. 5. Viking orbiter view of ridgesin the PhlegraMontes regionof easternElysium.Frame V0580A09; width of image
is 200 km.
elastic. For the more plausible rheological models in which
strength varies with depth [e.g., Goetze and Evans, 1979; Brace
and Kohlstedt, 1980], however, the stress differences induced
by flexure are supported by an elastic "core" of the plate even
after near-surfacefailure, and the effective flexural rigidity is
similar to that of the elasticcore. If there are important variations in the thicknessof the elasticlithosphere,then estimates
of lithosphericthicknessderived from a comparisonof model
predictions with observed tectonic features will be intermediate between the extremes in actual values [Pullan and Lainbeck, 1981]. Several time-dependenteffectsin loading problems have been consideredby Comer et al. [1985]; they found
that neglect of load growth probably introduces the most
serioussourceof error for the estimation of lithosphericthickness but that this problem can be minimized by considering
only tectonic featurescomparable in age to the youngestcon-
tribution to the load. The somewhat related problem that
present topography and gravity must be used to constrain
models for stressfields that gave rise to ancient tectonic features, while of seriousconcern for the complex Tharsis province, is lessenedfor Elysium by the simpler volcanic and tectonic history of the region.
Elysium Mons Loading
We consider first the hypothesis that the circumferential
fractures surrounding Elysium Mons are the product of flexure of the lithospherein responseto that load. In a successful
test of that hypothesis,Comer et al. [1985] found that a thicknessof about 50 km for the elastic lithosphereyields a distribution
of deviatoric
stresses that
best matches
the observed
distribution of concentricgraben. For this type of calculation
the magnitude of the load is not critical, but a reliable model
HALL ET AL.' TECTONICS
OF THEELYSIUMREGION,Mars
of the areal distribution
of the load is essential. The load
216øW
11,383
212 ø
208 ø
model of Comeret al. [1985] consistedof a stack of concentric
cylindersapproximatinga conicalvolcano.We haverepeated
the calculation
with an alternative
model with the same excess
massbut with a load distribution conforming more closelyto
the shape of the volcano,as given by the elevation contours of
Batsonet al. [1979], which indicatea pronouncednorth-south
elongation to the topographicrelief. We found that this load
model predictsstressesnot significantlydifferent from thoseof
the axially symmetricmodel for the same thickness(i.e., flexural rigidity) of the elasticlithosphere.The lithospheric thickness estimate of Comer et al. [1985] is therefore robust with
respectto uncertaintiesin the areal distribution of the Elysium
Mons
28øN
load.
24 ø
A further test of the hypothesisthat the graben concentric
to Elysium Mons are the result of flexure may be made by
comparing the predicted stress magnitudes with the lithospheric strength. In the upper lithosphere of the terrestrial
planets,strength is likely to be limited by friction on preexisting faults [Goetze and Evans, 1979; Brace and Kohlstedt,
1980]. Byeflee [1968] demonstratedthat frictional strengthis
well approximated by piecewise linear functions of depth
under both horizontal extension and horizontal compression,
20*
relations that are largely independentof temperature and rock
type. The strength dependson the effective confining pressure
Fig. 6. Sketch map of Elysium Mons and vicinity showing loand thus on the presenceof pore fluids. Fluidized crater ejecta, cations at which graben widths were measured, and the inferred
chaotic terrain, and fluvial channels have been cited as evi-
dencefor extensivevolcano-ground ice interactions within the
Elysium region [Mouginis-Mark et al., 1984]. Because there
may have been extensivemelting of permafrostand ground ice
during the time of volcanic activity in Elysium, the strengthin
both "wet" and "dry" situationsshould be considered.
depths (in kilometers) at which the normal faults bounding each
graben would intersect [Golombek, 1979]. Graben widths and wall
widths were measuredfrom Viking orbiter photographs(V0541A30,
V0844A19, V0846A16).
ter while varying Young'smodulusE and the load q by factors
If a flexuraloriginfor theconcentric
grabenis •orrect,then of 2 and 1.5, respectively,greater than and lessthan the values
normal faulting should occur at radial positions and over
depth intervals for which the flexural extensional stressexceeds the frictional strength [Solomon, 1985]. Golombek
[1979] has proposeda simple rule to estimate the maximum
depth of extensionalfaulting from the widths of graben and of
their bounding walls. This method is based on the assumptions that the graben are the product of simple extension,that
the faults bounding the graben dip at approximately 60ø, and
that faulting does not extend below the projected intersection
at depth of the two faults.We measuredgraben widths at nine
locations on six different graben (Figure 6). The measurement
points are located 150-215 km from the load center; they are
distributed around Elysium Mons from the northeast(N30øE)
to the southwest(S30øW).Graben widths range from 1.0 + 0.5
km to 3.0 +_0.5 km. Wall widths in all cases(both measured
directly and determined by subtracting the floor width from
the total graben width) are approximately 0.50 + 0.25 km.
The inferred maximum depth of faulting ranges from 0.3 to
2.1 km for the various graben considered. These estimates
have large uncertainties,however, particularly associatedwith
uncertainties in the dip angles of the bounding faults. For
assumeduncertaintiesof 10ø in the fault dip and 5ø in the
slope of the graben walls the range of possiblefault intersection depths varies from 0.1-1.6 km for the narrowest graben
measuredto 0.6-5.1 km for the widest graben.
For each graben the maximum depth of faulting may be
compared with the depth at which flexural stressesfall below
the extensional strength. The flexural stresses,including both
bending and membrane stresscontributions, are taken from
the loading model for Elysium Mons of Comer et al. [1985].
As a measure of the uncertaintiesin the predicted stresses,we
repeated the flexure calculationswith a fixed flexural parame-
assumedby Comer et al. [1985]. These values represent our
best estimateof the uncertaintiesin each of thesequantities.
One such comparisonfor one of the widest graben is shown
in Figure 7. While the inferred maximum depth of normal
faulting is somewhatshallower than that predicted from the
intersectionof the flexural stressdistributionand Byerlee'slaw
for extensionalstrength under "dry" conditions,the two predicted depths agree to within the estimated uncertainties in
both the maximum fault depth and the flexural stressmagnitudes.If the differencein thesedepthsis real, it may be due
to partial release of stressby the formation of neighboring
graben, finite extensional strength of relatively unfractured
surficial volcanic material, or a superpositionof stressesfrom
other sources.In all casesthe faulting depth predicted from
the intersection
of the flexural
stress curve
with
the exten-
sional strength distribution under "wet" conditions is greater
than that inferred from graben geometry. This comparison
suggeststhat the depth of faulting inferred for the formation of
concentricgraben is consistentwith a flexural origin as long
as there was no significant reduction of maximum effective
pressure by near-surface pore fluids during the period of
graben formation. However, becauseof the large uncertainties
in the predicted flexural stresses,the presenceof pore fluids
cannot be completelyexcluded.
Regional Scale Loading
The stressesdue to any arbitrary surfaceload q(r) on a thin
shell may be obtained by convolving the load with the stress
tensor ao resulting from a point load of unit magnitude:
•(r)
=ffq(r')•o(rr')dA'
(1)
11,384
HALLETAL.' TECTONICS
OFTHEELYSIUM
REGION,
Mars
o'N--or v , kbar
o
i
I ---
I0
--
/
Maximum
graben
//
--/
/
//
-
'
Elysium Mons
conc,n,r,c
ro,n
-
I
20
30
gth
/
.
•
.
.
I
..
I
.
_
I
.
,
Fig. 7. Comparisonof th• maximumd•pth of faultingduringth• formationof on• of th• wider grabcncircumferential
to Elysium Mons, inferredfrom the grab•n width by the m•thod of Golombek[1979], with that predictedfrom the
intersection
of th• predictedfl•xural stressdistributionand th• frictionalstrength•nvdop•. Th• quantitiesar and •n ar•
th• verticalprincipalstressand th• most extensionalhorizontalprincipalstress,r•sp•ctivdy; both are positiv• in extension.The •xtcnsionalstrengthv•rsusd•pth z is shownunderboth dry and w•t (• = 0.3)conditions,wh•rc • is th• ratio of
flnid pressureto lithostaticstress[BraceandKohlstedt,19B0].Th• fl•xural stresses
ar• from th• ElysiumMons modalof
Comeret al. [1985] at the appropriat•radial distanc•(r = 220 km); the rang• indicatedby arrowsincludesth• •ff•cts of
uncertaintiesin Young's modulusfor th• alasticlithosph•r• and in th• magnitudeof the load. The uncertaintiesin th•
maximumdepthof extensional
faultinginferredfrom grab•n g•om•try ar• du• principallyto th• uncertaintyin fault dip.
Th• shadedr•gion indicatesth• overlapin th• predictedmaximumfault d•pthsd•t•rmin•d by th• two methods.
wherer and r' are positionvectorson the planetarysurface the ith disk. Obviously,the stresstensorsgd must be cast in
and dA' is a unit of surfacearea.As an approximationto (1), terms of a global coordinatesystembeforesummation;stanwe replaceg0 with the stressdistributiongdresultingfrom a
dardspherical
coordinates
(r, 0, •p)withthepolaraxisat the
finite disk load of unit magnitude [Brotchie, 1971; Solomon geographicpole are usedthroughout.The principal horizontal
and Head, 1979], and we replacethe integralin (1) with a sum stressesare found by meansof standard relations.
over the number of disks:
A number of different combinationsof loads for Elysium
Planitia
volcanicunits and constructswere examined(Table
N
1). Each of these individual loads correspondsto a surface
excessmassin the gravity modelsof Janle and Ropers[1983]
i=1
for the Elysium region. The Elysium Mons and Hecates
whereqi and ri are the load and the positionof the centerfor Tholus loads correspondto the disc loads TD5 and TD4 of
TABLE 1. Contributionsto Elysium RegionalLoad Models
Load
Geometry
ElysiumMonsa
cone
HecatesTholusb
truncatedcone
Radius,km
Latitude,
øN
Longitude,
øW
Excess Mass,
102•g
100(base)
25.2
213.5
30 (top),
32.0
209.6
0.54
0.22
26.9
210.4
3.6
26.5
208.0
8.6
25.5
204.8
8.8
16.5
176.5
1.4
33.8
176.5
1.0
92 (base)
Elysium Planitiab
JR1
JR2
beveleddisc
beveleddisc
450 (top),
500 (base)
630 (top),
700 (base)
JR3
beveleddisc
900 (top),
1000 (base)
JR6
beveleddisc
510 (top),
570 (base)
JR7
beveleddisc
430 (top),
480 (base)
aFrom Comeret al. [1985].
bFromJanleandRopers[1983].
H^LL ETAL.:TECTONICS
OFTHEELYSIUM
REGION,
Mars
11,385
\
\
-
40ON
--
30 ø
-
20 ø
-I0
o
-- 0 •
240øW
230 ø
220 ø
210ø
200 ø
190ø
180ø
Fig.8. Principal
horizontal
stress
orientations
andmaximum
stress
differences
at thesurface
ofthelithosphere
fora
loadmodelconsisting
of Elysium
Mons,Hecates
Tholus,
anda regional
loadof Elysium
Planitiabasalts
approximately
900 km in horizontal
extent(loadJR1 in Table 1). Crosses
indicatedirections
of horizontalprincipalstresses,
with
arrowheads
denoting
extension;
where
bothhorizontal
principal
stresses
areofthesamesign,theoneoflargermagnitude
is indicated.
Dashedcontours
of themaximumstress
difference
Aa = a• - •3 (in kilobars)arealsoshown.Themagnitudes
of theloadsarebasedonthegravitymodels
ofJanleandRopers
1-1983].
Thethickness
oftheelastic
lithosphere
istakento
be 100km.Young's
modulus
andPoisson's
ratioareassumed
to be 10•2dyn/cm
2 and0.25,respectively.
Circles
represent
diskloads,
withonlythelargest
diskshown
fortheindividual
volcanoes
andwiththelargest
andsmallest
disksshown
for
the regionalload.
JanleandRopers[1983],exceptthat the excess
massfor ElysiumMons is the slightlylargervalueassumed
by Comeret al.
[1985].A separate
loadmightalsohavebeenconsidered
for
Albor Tholus, but we did not do so becauseof its small size;
its volumeis lessthan that of ElysiumMons by nearlya factor
three-disc
gravitymodelsof JanleandRopers[1983].For this
as well as other load models(Table 1), stressfieldswerecalcu-
lated for a rangeof assumed
valuesfor the thickness
of the
elastic lithosphere.
All models with the thicknessof the elastic lithosphere less
of 20 [Blasiusand Cutts,1981].The ElysiumPlanitialoads thanabout150km predictstressfieldswith largestressdifferare basedon topographicreliefand correspond
to the disc ences(Aa = a•- cr3, wherea• and cr3 are the greatestand
principalstresses,
respectively)
andwith the
loads TD1-3 and TD6-7 of Janle and Ropers [1983]. Cylin- leastcompressive
extensional
andcompressive
principalstresses
orientdrical discs,however,are poor approximationsto the areal greatest
ed horizontallyandapproximately
radialandcircumferential,
load distribution for loads of horizontal dimension at least
to the centerof massof the totalload(a point
comparableto the local flexurallength.We thereforerepre- respectively,
ElysiumMonsandHecates
Tholus).Suchstress
fields
sentedeachload by a stackof cylindersapproximatinga cone between
to strike-slip
faulting,though
for individual volcanoesand a truncatedcone (or beveleddisc) wouldgiverisepredominantly
of an additionalstress
field,grabencircumferfor largervolcanicunits.In all of theserepresentations
the in thepresence
ential to the Elysiumrise or compressive
featuresapproxitotal excessmassis preserved.
The simple
The stressfieldfor onesuchregionalload modelis shownin matelyradialto therisemightalsobe produced.
grabenaround
Figure8. The modelincludes
loadsfor ElysiumMons and modelof Figure8 failsto predictconcentric
ElysiumM ons at the distances
observed
because
the stress
Hecates Tholus and a load (JR1 in Table 1) equivalentto
fieldin the vicinityof the shieldis dominatedby the compresabout a 1.5-km-thick basalt unit over central Elysium Planitia. This modelcorresponds
to the leastcompensated
of the sivestressfieldof theregionalElysiumPlanitiaload.
11,386
H^LL ET AL.: TECTONICS
OFTHEELYSIUMREGION,Mars
For an upward lithosphericload equal in magnitude,distriJanle et al. [1984] reported stress calculations similar to
that shown in Figure 8. They argued that suchstressfields can bution, and location to the volcanic load JR1 in Table 1, the
account for the formation of the tectonic featuresin Figure 1, resulting stressfield (Figure 9) is generally consistentwith the
and on the basisof this argument they made further inferences positions of radial extensional features. We have taken the
as to the thicknessof the lithosphere and its lateral variation. thickness of the elastic lithosphere to be 100 km for this
In their comparison of predicted stresseswith tectonic fea- model; a thicker lithospherewould require greater excessprestures, however, Janle et al. [1984] examined only the second sures in order to produce surface stressesof a given maginvariant of the deviatoric stress tensor, which is a measure
nitude. As shown in Figure 9, the greatestextensionalstresses
only of the magnitudes and not the directions of the principal are approximately circumferentialto the center of the Elysium
deviatoric stresses.Although the regional stressmodelsfor a rise and are significant in magnitude (tr _• 0.2-0.6 kbar) over
the area in which those features are found. Also, because the
thicknessof the elastic lithosphere lessthan about 200 km do
predict stressdifferencessufficient to induce faulting at shal- extensional stressfield produced over the regional uplift adds
low depths [Brace, 1964; Brace and Kohlstedt, 1980], the cal- constructively to the local extension produced by flexure in
responseto the Elysium Mons shield,the regionaluplift model
culated stresstrajectories are not consistentwith the orientaof Figure 9 is also consistent with the formation of graben
tions and types of tectonicfeaturesobserved.
We recognizethe possibilitythat our loading models,based concentric to Elysium Mons. The predicted maximum upward
on the approximation of Brotchie [1971] in which the dimen- deflectionof the lithospherefor the assumedvaluesof upward
sion of the load is small compared with the planetary radius, load and lithospherethicknessis approximately1 km. Suchan
may tend to underestimate the magnitude of membrane uplift could be the result of mantle thermal expansion for a
stressesfor loads of large lateral extent. Membrane stresses temperature contrast of about 300øC extending to a depth of
supporttopographyof long wavelength[Turcotte et al., 1981; 100 km, values not unreasonablein comparison to thermal
Willemann and Turcotte, 1981, 1982], while bending stresses anomaliesassociated
with mid-oceanridgeson the earth [Parare dominant in the support of short-wavelengthtopography. sonsand $clater, 1977].
Because this model is offered as a mechanism of formation
For Mars the transition between bending-dominated and
membrane-dominated stress fields occurs for loads with horiof tectonic features rather than as an explanation of the rezontal dimensionsgreater than about 2000 km [Willemann gional topography and gravity, any causativemantle thermal
and Turcotte, 1981]. While the Elysium rise has approxi- anomaly is not required to persist to the present. An uplift
mately this dimension,the portion of the risethat may exert a model in which the deformation matched the presenttopograload on the lithosphereis substantiallysmallerin extent,per- phy (44 km over the center of the rise) would predict very
haps900 km across[Janle and Ropers,1983]. The analysisof large (several kilobars) extensional stressesover most of the
Willemann and Turcotte [1981] indicatesthat for a load of surfaceof the rise. With suchhigh stressesit would be difficult
this dimension the membrane stressesare not likely to be the to account for the relatively small numbers of radial extensional tectonicfeatures.It is thus likely that much of the presdominant support of the load.
On the basisof a comparisonof the stressmodelsin Table 1 ent regional topography is due to some combination of volcanic constructionand isostaticsupport.
with the distribution of tectonic features in Elysium there apIf significantvolcanicconstructionof Elysium Planitia postpears to be no evidencefor downward loading of the lithosphereby volcanicunits at a regionalscale.This implieseither dated regional uplift and the formation of radial extensional
that the regional volcanic units are largely isostaticallycom- fractures, an important question concernsthe extent to which
pensated[Janle and Ropers,1983] or that the uncompensated the evidence of such fracturing would have been preserved.
portion of the load is supportedby the strengthof a very thick One mechanism for preserving the tectonic signature of an
elasticlithosphere.The locally thinner lithospherein the vicin- early uplift phase would be if the radial extensionalfractures
ity of Elysium Mons at the time of formation of concentric later served as preferred sites for lava tube collapse or for
grabenmay have beenthe resultof heatingassociatedwith the
fluvial channel formation.
formation
servation [Cart, 1981; Mouginis-Mark et al., 1984] that the
wide grabenlike depressionsof central Elysium Fossae grade
downslopeinto sinuouschannelsof volcanicor fluvial origin.
The uplift model of Figure 9 does not reproduce the azimuthal asymmetry in the distribution of extensionalfeatures,
which are preferentiallyoriented NW-SE (Figure 1). The observed distribution of faulting could be due to corresponding
asymmetries in the distribution of uplift (which need not
match the present topography) or in lithospheric thickness,
but there is no means to test these possibilitiesat present.
Since Elysium lies near the boundary between the ancient
southern uplands and the northern lowland plains of Mars,
one potential sourceof heterogeneityin lithosphericthickness
of the volcano. Similar lateral variations
in litho-
sphericthicknesshave been inferred for the Tharsis region
[Solomonand Head, 1982; Comeret al., 1985].
Regional Uplift
As an alternative to volcanic loading models, we next consider the possibility of flexural uplift on a regional scale.The
primary constraints are the positions and orientations of the
extensional tectonic features in Elysium Planitia. These features are found to a radial distance of approximately 500 km
from the center of the rise. This suggeststhat the area of uplift
was also of the order of 500 km in radius because extensional
stressis predicted only over the uplifted area. The pattern and
magnitudeof uplift, or of equivalentupward load on the lithosphere, are otherwise poorly constrained. Presumably, this
upward load representsan excesspressureexerted on the base
of the lithosphere by a mechanismsuch as thermal expansion
of a volume of underlying mantle material. Flexural uplift of
the lithospherecreatesstresseswhich balance the excesspressure; if the uplift is due to differential thermal expansion,the
volume of thermally anomalous material and the percent
volume changeare not separatelyresolvable.
This idea is consistent with the ob-
or of preferredzonesof weaknesswas the processthat gave
rise to the Martian hemisphericalasymmetry. Wilhelms and
Squyres [1984] have recently suggested that the uplandlowland boundary marks the rim of a 7700-km-diameter
impact basin centeredabout 50øN, 190øW. While the formation of such a huge impact basin might be expectedto have
left a strong signatureon the distribution of later formed tectonic features, neither the Elysium Fossae nor the Cerberus
Rupes are concentricto the proposed location of the basin
center.
HALLETAL.;TECTONICS
OFTHEELYSIUM
REGION,
Mars
11,387
40øN
,0,2
\\
,4
30 ø
I aI
20 ø
I0 ø
o
240ow
2;.'50
ø
220 ø
210ø
200 ø
190ø
180ø
Fig.9. Principal
horizontal
stress
orientations
andmaximum
stress
differences
atthesurface
ofthelithosphere
fora
model
consisting
ofloading
byElysium
MonsandHecates
Tholus
superposed
ona regional
flexural
upliftofElysium
Planitia;
theupward
loadisapproximately
900kminextent
andequal
inmagnitude
toloadJR1used
forthemodel
in
Figure8.Thethickness
oftheelastic
lithosphere
istakento be100km.
Tharsis Loading
R23 =-
- n(n + 1)
r •
(/t+ 2y)
Finally,we consider
the effecton stressin the Elysium
regionof the long-wavelength
loadingof the Martianlithosphere
bytheTharsis
rise.Weemploy
thesolutions
ofBanerdt
r • + 2n(n
+ 1)(/t+ 2y)
et al. [1982] for displacement
andstressin a self-gravitating
elasticspherical
shelloverlying
an inviscidfluidinterior.Banerdtet al. [1982] usethe formulationof Altermanet al.
R63
--r sin0 •0•½
--cotan
0
[1959] and Arkani-Hamed
[1973],in whicha spherical
harmonicexpansion
of the gravityand topography,
combined and wherethe two componentsof Y,m are the harmoniccoef-
R3
3_ 2Z[
a2
with a model for material propertiesand density of the
ficientsof radial and tangentialdisplacement,
respectively,S,•
medium,yield a representation
of the internaldeformation is the unnormalized surfacesphericalharmonic of degree n
field,subjectto the assumed
sphericalharmonicboundary and order m, and • and y are the Lam• constants.
conditionsand the isostaticresponse
function.The elementsof
The stressmodels for Tharsis loading of Banerdt et al.
the stresstensor are derived from the deformation via a gener-
[1982] are basedon sphericalharmonicexpansions
of gravity
andtopographyto fourthdegreeandorderfromSjogrenet al.
[1975] and Bills and Ferrari [1978], respectively.
At these
long wavelengths
there doesnot appearto be a significant
contributionto either the gravity or the topographyfrom the
Elysium region.While Banerdtet al. [1982] did not show
predictedstresstrajectoriesin the Elysiumregion,W. B. Banerdt(personalcommunication,
1983)kindlyprovideduswith
the valuesof Y,mfor two of theirmodels:theirisostaticmodel
alized Hooke'slaw operatormatrix [Arkani-Hamed,1973].
For the calculation of stressesat the planetary surface,the
operatorequationreduces
to therelation(W. B. Banerdt,personal communication, 1983)
3
•GO,/n
m
0
R63/n
nm
where the elementsof the operator matrix R are
R2•:
R3• -
2• (3i + 2•)
r (2 + 2•)
and their flexural loading model. With these coefficients,we
evaluated(3) by analyticdifferentiationof the surfacespherical
harmonicsS,m.(This procedurediffersslightlyfrom the one
employedby Banerdtand colleagues.)
11,388
HALLETAL.:TECTONICS
OFTHEELYSIUM
REGION,
Mars
240øW
I
220"
I
I
200 ø
I
I
180"
I
40øN
0.05
0.05
0.10
0.15
0.20
20 ø
o
Fig. 10. Principalhorizontalstressorientationsand maximumstressdifferences
at the surfaceof the lithosphere
in the
Elysiumregionpredictedby the isostaticmodelfor Tharsisof Banerdtet al. [1982]. SeeFigure8 for furtherexplanation.
Surface stressesin the Elysium region for the isostatic
model for compensationof Tharsistopography[Banerdtet al.,
1982] are shown in Figure 10. This isostatic model, taken
from Sleepand Phillips [1979], has a mean crustal thicknessof
150 km and a (thermal or compositional)lithosphere400 km
thick. The stressdistribution predicted for this model in the
Tharsis region was shown by Banerdt et al. [1982] in their
Figure 3b; these workers argued that the predicted stresses
provide a good match to the observeddistribution of tectonic
featuresin the central Tharsisregion. In the Elysium region
this model yieldssurfacestresses
that are compressive
over the
entire area shown in Figure 10 except for the portion west of
about 230øW. The magnitudeof the predictedstressdifference
Elysium region this model predicts large extensional surface
stressesin the eastern portion of Figure 11, with maximum
extensional stressesoriented approximately N-S to NE-SW.
Although these stressesare not consistent with the ridge
systemin eastern Elysium, they are of an orientation and a
magnitude to have contributed to the formation of linear extensionalfeaturesin southeasternElysium, including Cerberus
Rupes (Figure 1), a possibility first suggestedby Cart [1974].
In western Elysium the flexural loading model predicts compressivehorizontal stresses,with the direction of greatestcompressivestressoriented NW-SE. Such a stressfield would not
yield the NW-SE trending extensional features observed in
this region, but it might have contributed to the formation of
Aa decreases from 200-250 bars between 180 ø and 195øW to
the ridgesin southwesternElysium (Figure 1).
less than 100 bars for the region west of about 210øW. In
In principle, the attribution of tectonic features in Elysium
easternElysium the direction of maximum compressivestress to lithospheric stressesgenerated by Tharsis might help to
is nearly E-W; sucha stressfield might be a contributorto the resolve the temporal sequenceof compensationmechanisms
formation of the N-S trendingridgesin the area (Figure 1). In for the Tharsis province. The relative ages of the generally
western Elysium the stressdifferencesare too small to have a isolatedtectonicfeaturesin the Elysium area (Figure 1), howsignificant effect on the formation of tectonic features.
ever, cannot readily be determined.
Surfacestressesin the Elysium region for the Tharsis flexCombined Models
ural loading model of Banerdt et al. [1982] are shown in
Figure 11. This model is based on an assumed 150-km thickIt is, of course, unlikely that only one stressfield (local,
ness for the crust and a 200-km thickness for the elastic lithoregional,or global) would be operatingin the Elysium region
sphere;the predictedstressdistributionin the Tharsisregion during the time interval in which the tectonic features now
was shownby Banerdtet al. [1982] in their Figure 3c. In the visible were formed. Despite the greater surface age of the
HALL ET AL.: TECTONICS
OF THE ELYSIUMREGION,Mars
240oW
I
220 ø
I
I
11,389
180 ø
200 ø
I
I
40øN
7
0.4 0.3
0.7
1.0
1.3
1.6
1.9
2:.2
20 <,
Fig. 11. Principal
horizontal
stress
orientations
andmaximum
stress
differences
at thesurface
ofthelithosphere
in the
Elysium
region
predicted
by theflexural
loading
modelfor Tharsis
of Banerdt
et al. [1982].SeeFigure8 for further
explanation.
featuresof
Elysiumvolcanoes
comparedwith their Tharsiscounterparts, centricto ElysiumMons,andthe linearextensional
it is likely that the periodsof major volcanicactivityin the Elysium Fossae.
two provinces
overlapped
I-Plescia
andSaunders,
1979,1982].
CONCLUDING DISCUSSION
It is thusappropriateto consider
the superposition
of regional
In this paper we have testedthree scalesof lithospheric
and local stressfieldswith the global scalestressfield due to
loadingmodelsfor the formationof tectonicfeaturesof the
loading by the Tharsisrise.
In constructing
a varietyof combinedstressmodelswe have Elysiumregion:local modelsof loadingby ElysiumMons,
also allowed for the addition of a horizontally isotropic stress regionalmodelsof loadingon the scaleof ElysiumPlanitia,
modelsof loadingby the Tharsisrise. A
field, suchas would be introducedby planetarythermal ex- and quasi-global
of flexuralstresses
with lithospheric
strengthand
pansionor contraction
[Solomon,
1978].A superposed
hori- comparison
zontal extensionalstressof globalextent,for instance,would the inferred maximum depth of faulting confirm the conlead to the formation of extensionalfeatures in regions where
clusion of Comer et al. [1985] that the concentric graben
simpleloadingmodels(e.g.,Figure8) wouldpredicttheformation of strike-slipfaults [e.g.,Solomonand Head, 1980; Comer
aroundElysiumMons can be explainedas the resultof local
flexureof an elasticlithosphereabout 50 km thick in response
et al., 1985].
to the volcano load.
In no case does the stressfield predicted by a singlecombined stressmodel achievea simultaneouslysatisfactoryfit to
stressfields inconsistent with all of the observed tectonic fea-
all of the observed tectonic features. There is no reason, of
course,to expectthat all tectonicfeaturesin the regionformed
contemporaneously.
One of the bettermodelsconsists
of the
Banerdtet al. [1982] isostaticmodelfor Tharsistogetherwith
the regionalupliftmodeland the individualvolcanicloadsof
ElysiumMons and HecatesTholus (Figure 12). The stress
field from sucha modelmight have contributedto the formation of the ridge systemin easternElysium,the grabencon-
The volcanicloadingmodelson a regionalscalepredict
tures,suggesting
that loadingby widespread
emplacement
of
thick plainsdeposits
wasnot an importantfactorin the tectonic evolution of the Elysium region. This may have been
eitherbecauseregionalvolcanicunitswereisostaticallycom-
pensated
withoutlithospheric
loadingor because
the elastic
lithospherewas sufficientlythick that it could supportthe
regionalload withoutsurfacestresses
reachinglevelswhich
would
cause fracture.
11,390
HALL ETAL..'TECTONICS
OFTHEELYSIUMREGION,Mars
40øN
30*
20*
10*
o
240øW
230'
220*
210 ø
200*
190'
180'
Fig. •2. Principal horizontal stressorientationsand maximum stressdi•½renc½s
at the surfaceof the lithospherefor a
superpositionof the volcanoloading and regionaluplift model of Figure 9 and the B•nerdt et •L •1982-1isostaticmodel for
Tharsisfrom Figure !0. SeeFigure 8 for furtherexplanation.
The global stressfield associatedwith the support of the
Tharsis rise appears to have influencedthe developmentof a
number of tectonic features in the Elysium region. The stress
field predicted by the isostaticmodel for Tharsis of Banerdt et
al. [1982] is generally consistent with the formation of the
ridge system of eastern Elysium; the flexural loading model
for Tharsis [Banerdt et al., 1982] predicts extensionalstresses
generallyconsistentwith the formation of the CerberusRupes
system of linear extensional fault scarps and compressional
stressesin western Elysium consistent with the formation of
generalbasedon an analogy with the Tharsis region, which is
much larger but displayssimilar typesof tectonicfeatures.It is
now recognizedthat at the long wavelengthsof the Tharsis
rise, radially orientedextensionalfeaturesare the product of a
stress field dominated by membrane stress and inconsistent
with an origin by flexural uplift [Willemann and Turcotte,
1982; Banerdtet al., 1982]. For the smaller Elysium province,
in contrast,radial extensionalfeaturescan resultfrom bending
stressesproduced by regional flexural uplift. It thus appears
that while early argumentsabout the nature of Elysium tectonics were based on a faulty analogy, their ultimate conclusionsconcerningthe origin of many of the tectonicfeatures
the ridges in that area.
The linear extensional features of Elysium Fossae can best
be explained as the result of modest flexural uplift of the litho-
may yet be valid.
sphereon the scaleof the Elysium rise, possiblyin responseto
a mantle thermal anomaly beneath the volcanic province.The
simple,circularly symmetricuplift model doesnot explain the
preferred NW-SE trends of the features; such an asymmetry
could be the result of heterogeneitiesin the lithosphericthickness over the region (with local thinning in the areas where
fracturing occurs)or of unmodeled asymmetriesin the extent
of uplift.
It should be noted that other workers have previously
argued [e.g., Cart, 1973; Scott and Ailingham,1976] that many
of the tectonic featuresof the Elysium province were the result
of regional uplift. These earlier arguments, however, were in
Overall, the comparisonof stressmodelsfor Elysium with
the distributionof preservedtectonicfeaturessuggestsa succession of stress fields operating at different times in the
region. The order in which those stressfields were present,
however, cannot be specifiedwith confidence.Becauseof the
apparentlylimited extentof lithosphericextensionin Elysium
it is reasonableto expect that a mantle thermal anomaly
would have beenpresentat the outsetof any major episodeof
volcanicconstruction.By this argument,flexural uplift of the
lithosphereby suchan anomalywould have precededor occurred contemporaneouslywith volcanic loading. The fact
that someof the linear depressions
in Elysiumappearto have
H^LLET^L.:TECTONICS
OFTHEELYSIUM
REGION,
Mars
beenmodifiedby volcanicor fluvialflow processes
alsoindicatesthat the fracturingthat createdthe originalfeaturesoccurredat leastbeforethe final stageof volcanicactivityin the
11,391
Hartmann,W. K., Martian surfaceand crust:Reviewand synthesis,
Icarus, 19, 550-575, 1973.
Hiller, K. H., Geologicmap of the Amenthesquadrangleof Mars,
Map I-1110,U.S.Geol.Surv.,Reston,Va., 1979.
region.We conjecture
thattheearliesttectonicactivityin the Howard, K. A., and W. R. Muehlberger,Lunar thrustfaultsin the
Tharsisprovincemayalsohavebeenthe resultof lithospheric Taurus-Littrowregion,Apollo 17 PreliminaryScientificReport,
NASA Spec.Publ.,SP-330,31-22-31-25,1973.
uplift,but eitherthe scaleof suchuplift was considerablyJanle,
P., and J. Ropers,Investigationof the isostaticstateof the
smallerthan the presenthorizontalextentof the Tharsisrise
Elysiumdomeon Mars by gravitymodels,Phys.Earth Planet.
or the tectonic evidenceof this uplift has been long since
Inter., 32, 132-145, 1983.
Janle,P., F. Roth,andJ. Voss,Limitsof thelithospheric
thickness
of
erasedby subsequent
volcanismandfaulting.
theElysiumareaon Marsfromcorrelations
of surface
stresses
with
the lineamentsystem(abstract),Lunar Planet.Sci., 15, 401-402,
Acknowledgments.
We thankBruceBanerdtfor providinguswith
details of his stresscalculationsfor Tharsis. We also thank Rob
1984.
manuscript
preparation.
Thisresearch
wassupported
by NASA PlanetaryGeologyandGeophysics
ProgramthroughgrantsNSG-7297to
Sci.Conf.,8th, 2691-2703, 1977.
Comer, Marcia McNutt, Peter Mouginis-Mark,and Roger Phillips Lucchitta,B. K., Mare ridgesand relatedhighlandscarps--Aresult
of verticaltectonism?,Proc.Lunar Sci.Conf.,7th,2761-2782,1976.
for constructivecomments;Linda Meinke and Leigh Hall for assiststructure,
andmareridgesin southern
ancein preparing
thefigures;andJanNattier-Barbaro
for helpwith Lucchitta,B. K., Topography,
the MassachusettsInstitute of Technologyand NGR-40-002-116 to
Brown University.
REFERENCES
Alterman,Z., H. Jarosch,
andC. L. Pekeris,Oscillations
of the Earth,
Proc. R. Soc. London,Ser. A, 252, 80-95, 1959.
Arkani-Hamed,J., Stressdifferences
in the moon as evidencefor a
coldmoon,Moon,6, 135-•63,1973.
Baker, V. R., The Channelsof Mars, 198 pp., Universityof Texas
Press,Austin, 1982.
Banerdt,W. B., R. J. Phillips,N.H. Sleep,and R. S. Saunders,
Thick
shelltectonicson one-plateplanets:Applicationsto Mars, J. Geophys.Res.,87, 9723-9733,1982.
Batson,R. M., P.M. Bridges,and J. L. Inge,Atlasof Mars, NASA
Spec.Publ.,SP-438,146pp., 1979.
Mare Imbrium and northern Oceanus Procellarum, Proc. Lunar
Malin,M. C., Comparison
of volcanic
features
of Elysium(Mars)and
Tibesti(earth),Geol.Soc.Am.Bull.,88, 908-919,1977.
Maxwell,T. A., Orientationand originof ridgesin the LunaePalus-
Coprates
regionof Mars,Proc.LunarPlanet.Sci.Conf.13th,Part
1,J. Geophys.
Res.,87, suppl.,A97-A108,1982.
Melosh,H. J., The tectonics
of masconloading,Proc.LunarPlanet.
Sci.Conf, 9th, 3513-3525,1978.
Mouginis-Mark,
P. J.,L. Wilson,andJ. W. HeadIII, Explosive
volcanism on HecatesTholus, Mars: Investigationof eruption conditions,J. Geophys.
Res.,87, 9890-9904,1982.
Mouginis-Mark,
P. J.,L. Wilson,J. W. Head,S.H. Brown,J. L. Hall,
andK. D. Sullivan,ElysiumPlanitia,Mars: Regionalgeology,
vol-
canology,
andevidence
for volcano/ground
iceinteractions,
Earth
Moon Planets,30, 149-173, 1984.
Muehlberger,
W. R., Structural
historyof southeastern
Mare Serenitatisand adjacenthighlands,
Proc.LunarSci.Conf.,5th, 101-
Bills, B. G., and A. J. Ferrari, Mars topographyand geophysical 110, 1974.
Neukum,G., and D. U. Wise,Mars: A standardcratercurveand
implications,
J. Geophys.
Res.,83, 3497-3508,1978.
Blasius,K. R., and J. A. Cutts,Topography
of Martian centralvolpossible
newtimescale,
Science,
194,1381-1387,
1976.
canoes,Icarus, 45, 87-112, 1981.
Parsons,B., and J. G. Sc!ater,An analysisof the variationof ocean
Brace,W. F., Brittlefractureof rocks,in Stateof Stressin theEarth's
floor bathymetryand heat flow with age,J. Geophys.
Res.,82,
803-827, 1977.
Crust,pp. 110-178,editedby W. R. Judd,ElsevierScientific,
New
York, 1964.
Plescia,J. B., and R. S. Saunders,
The chronologyof the Martian
volcanoes,
Proc.LunarPlanet.Sci.Conf.,loth, 2841-2859,1979.
Brace,W. F., and D. L. Kohlstedt,Limitson lithospheric
stressimposedby laboratory
experiments,
J. Geophys.
Res.,85, 6248-6252, Plescia,J. B., and R. S. Saunders,
Estimationof the thickness
of the
1980.
Tharsislava flowsand implications
for the natureof the topograBrotchie,J. F., Flexureof a liquid-filledsphericalshellin a radial
phy of the Tharsisplateau,Proc.LunarPlanet.Sci.Conf.,11th,
gravityfield,Mod. Geol.,3, 15-25,1971.
2423-2436, 1980.
Brotchie,J. F., and R. Silvester,On crustalflexure,J. Geophys.
Res., Plescia,J. B., and R. S. Saunders,
Tectonichistoryof the Tharsis
74, 5240-5252, 1969.
region,Mars,J. Geophys.
Res.,87,9775-9791,1982.
Byerlee,
J. D., Brittle-ductile
transition
in rocks,J. Geophys.
Res.,73, Pullan,S.,and K. Lambeck,Masconsandloadingof the lunarlitho4741-4750, 1968.
sphere,
Proc.LunarPlanet.Sci.,12B,853-865,1981.
Carr, M. H., Volcanismon Mars, J. Geophys.
Res.,78, 4049-4062, Schumm,S. A., Structuralorigin of large Martian channels,Icarus,
1973.
Carr, M. H., Tectonism
andvolcanism
of theTharsisregionof Mars,
J. Geophys.
Res.,79, 3943-3949,1974.
Carr, M. H., The morphology
of the Martian surface,
SpaceSci.Rev.,
25, 231-284, 1980.
Carr, M. H., The Surfaceof Mars, 232 pp., Yale UniversityPress,
New Haven, Conn., 1981.
22, 371-384, 1974.
Scott,D. H., and J. W. Ailingham,Geologicmap of the Elysium
quadrangle
of Mars, Map 1-935,U.S. Geol. Surv.,Reston,Va.,
1976.
Scott,D. H., and M. H. Carr, Geologicmap of Mars, Map 1-1083,
U.S. Geol. Surv., Reston, Va., 1978.
Sharp,R. P., Geomorphological
processes
on terrestrialplanetary
surfaces,
Annu.Rev.Earth Planet.Sci.,8, 231-261, 1980.
Chicarro,A. F., P. H. Schultz,and P. Masson,Global and regional
Sharpton,
V. L., andJ. W. Head,Stratigraphy
andstructural
evoluridgepatternson Mars,Icarus,63, 153-174,1985.
tion of southern Mare Serenitatis:A reinterpretationbased on
Comer,R. P., Thick plate flexure,Geophys.
J. R. Astron.Soc.,72,
101-113, 1983.
Comer, R. P., S.C. Solomon,and J. W. Head, Elastic lithosphere
thickness on the moon from mare tectonic features: A formal inver-
Apollolunarsounder
experiment
data,J. Geophys.
Res.,87, 10,9831(•,998, 1982.
Sjogren,W. L., Mars gravity:High resolution
resultsfromViking
Orbiter II, Science,203, 1006-1010, 1979.
sion,Proc.LunarPlanet.Sci.Conf, loth, 2441-2463,1979.
W. L., J. Lorell,L. Wong,andW. Downs,Marsgravityfield
Comer,R. P., S.C. Solomon,andJ. W. Head,Mars: Thicknessof the Sjogren,
basedon a short arc technique,J. Geophys.Res.,80, 2899-2908,
lithosphere
fromthe tectonicresponse
to volcanicloads,Rev.Geophys.,23, 61-92, 1985.
1975.
modelfor the Tharsis
De Hon, R. A., Thickness
of Tharsisvolcan•ic
materials,J. Geophys. Sleep,N.H., and R. J. Phillips,An isostatic
Province,Mars, Geophys.
Res.Lett., 6, 803-806, 1979.
Res.,87, 9821-9828, 1982.
N.H., andR. J. Phillips,Gravityandlithospheric
stress
on the
Gifford,A. W., Ridgesystems
of Mars, Advancesin PlanetaryGe- Sleep,
terrestrialplanetswith reference
to the Tharsisregionof Mars,J.
ology,NASA Tech.Memo.,TM 84412,220-363,1981.
Geophys.
Res.,90, 4469-4489,1985.
Goetze,C., and B. Evans,Stressand temperaturein the bending
lithosphere
as constrained
by experimental
rock mechanics,
Geo- Solomon,S.C., On volcanismand thermaltectonicson one-plate
planets,Geophys.
Res.Lett.,5, 461-464,1978.
phys.J. R. Astron.Soc.,59, 463-478,1979.
among
Golombek,M.P., Structuralanalysisof lunar grabensand the shal- Solomon,S.C., The elasticlithosphere:Somerelationships
low crustal structureof the moon, J. Geophys.Res.,84, 4657-4666,
1979.
flexure,
depthof faulting,
lithospheric
thickness,
andthermalgradi-
ent (abstract),
LunarPlanet.Sci.,!6, 799-800,1985.
11,392
HALL ETAL.:TECTONICS
OFTHEELYSIUMREGION,Mars
Solomon, S.C., and J. W. Head, Vertical movement in mare basins:
Relation to mare emplacement,basin tectonics,and lunar thermal
history, J. Geophys.Res., 84, 1667-1682, !979.
Solomon, S.C., and J. W. Head, Lunar mascon basins: Lava filling,
tectonics,and evolution of the lithosphere,_Rev.Geophys.,18, 107141, 1980.
Solomon, S.C., and J. W. Head, Evolution of the Tharsis province,
Mars: The importance of heterogeneouslithosphericthicknessand
volcanic construction,J. Geophys.Res.,87, 9755-9774, 1982.
Thurber, C. H., and M. N. Toksoz, Martian lithospheric thickness
from elasticflexure theory, Geophys.Res. Lett., 5, 977-980, 1978.
Turcotte, D. L., R. J. Willemann, W. F. Haxby, and J. Norberry, Role
of membrane stressesin the support of planetary topography, J.
Geophys.Res., 86, 3951-3959, 1981.
Wilhelms, D. E., and S. W. Squyres,The Martian hemisphericdichotomy may be due to a giant impact, Nature, 309, 138-140, 1984.
Willemann, R. J., and D. L. Turcotte, Support of topographic and
other loads on the moon and on the terrestrial planets,Proc. Lunar
Planet. Sci., 12B, 837-851, 1981.
Willemann, R. J., and D. L. Turcotte, The role of lithospheric stressin
the support of the Tharsis rise, J. Geophys.Res., 87, 9793-9801,
1982.
Wise, D. U., M.P. Golombek, and G. E. McGill, Tharsis province of
Mars: Geologic sequence,geometry, and a deformation mechanism, Icarus, 38, 456-472, 1979a.
Wise, D. U., M.P. Golombek, and G. E. McGill, Tectonic evolution
of Mars, J. Geophys.Res.,84, 7934-7939, 1979b.
J. L. Hall and S.C. Solomon,Department of Earth, Atmospheric,
and PlanetarySciences,
Massachusetts
Institute of Technology,Cambridge, MA 02139.
J. W. Head, Departmentof GeologicalSciences,Brown University,
Providence, RI 02912.
(ReceivedSeptember4, •1985;
revisedApril 18, 1986;
acceptedJuly 7, 1986.)