Skip to main content
NCBI home page
NASA Author Manuscripts logoLink to NASA Author Manuscripts
. Author manuscript; available in PMC: 2019 Sep 18.
Published in final edited form as: J Geophys Res Planets. 2017 May;122(5):889–900. doi: 10.1002/2017JE005274

The anatomy of a wrinkle ridge revealed in the wall of Melas Chasma, Mars

Hank M Cole 1, Jeffrey C Andrews-Hanna 2,3
PMCID: PMC6750226  NIHMSID: NIHMS1049587  PMID: 31534880

Abstract

Wrinkle ridges are among the most common tectonic structures on the terrestrial planets, and provide important records of the history of planetary strain and geodynamics. The observed broad arches and superposed narrow wrinkles are thought to be the surface manifestation of blind thrust faults, which terminate in near-surface volcanic sequences and cause folding and layer-parallel shear. However, the subsurface tectonic architecture associated with the ridges remains a matter of debate. Here we present direct observations of a wrinkle ridge thrust fault where it has been exposed by erosion in the southern wall of Melas Chasma on Mars. The thrust fault has been made resistant to erosion, likely due to volcanic intrusion, such that later erosional widening of the trough exposed the fault plane as a 70 km-long ridge extending into the chasma. A plane fit to this ridge crest reveals a thrust fault with a dip of 13° (+8°, −7°) between 1 and 3.5 km depth below the plateau surface, with no evidence for listric character in this depth range. This dip is significantly lower than the commonly assumed value of 30°, which, if representative of other wrinkle ridges, indicates that global contraction on Mars may have been previously underestimated.

1. Introduction and background

Wrinkle ridges are quasi-linear compressional tectonic features observed on the volcanic surfaces of Mercury, Venus, the Moon, and Mars, being particularly abundant on the Hesperian-aged volcanic plains that cover a large fraction of the martian surface [e.g., Head, 2002; Knapmeyer et al., 2006]. Wrinkle ridges typically have a morphology consisting of a narrow ridge ~2–5 km wide superimposed onto a broad arch ~10–20 km wide, with lengths up to 100’s of km [Watters, 1988]. The observed surface structures are thought to form when a blind thrust fault (i.e., one which does not directly rupture the surface) causes folding and layer parallel shear of the near-surface volcanic plains sequences [Schultz, 2000a]. However, in some cases continued fault propagation may cause surface rupture. Wrinkle ridges can occur either singly or in sets of parallel ridges with common asymmetrical profiles [Golombek et al., 2001].

Studies of wrinkle ridges are able to provide rich commentary on the geodynamic processes at regional and global scales. The compressional stresses responsible for wrinkle ridge faulting may arise either from local to regional-scale loading and flexure, from global scale compressional stresses due to planetary cooling, or from a combination of these two processes. At local to regional scales, wrinkle ridges provide a record of the nature and orientation of the principal stresses within the upper lithosphere. For example, wrinkle ridges over much of the martian surface are generally oriented concentric to Tharsis [Chicarro et al., 1985; Head et al., 2002] and record a combination of Tharsis loading stresses and superimposed uniform compressional stresses from global contraction [Tanaka et al., 1991; Mège and Masson, 1996; Andrews-Hanna et al., 2008]. The tectonic and geodynamic implications of wrinkle ridges hinge on the inferred shortening across individual structures, which in turn depend upon the assumed geometry of faults underlying the ridges. For example, Nahm and Schultz [2011] estimated the magnitude of global contraction due to cooling by integrating the net strain indicated by the surficial expression of wrinkle ridges and other faults. For broad geodynamic studies, it is common to assume a wrinkle ridge fault dip of 25–30° [e.g., Golombek et al., 2001; Nahm and Schultz, 2011], based on the assumption of Andersonian faulting criteria [Anderson, 1905], and calculate the displacement and strain from either the observed topographic step across the ridges or displacement-length relationships.

Studies of the mechanics of wrinkle ridge tectonics have used the observed surface deformation in order to constrain the underlying fault properties. Such studies commonly use elastic dislocation boundary element models [King et al., 1994; Toda et al., 1998] in order to solve for the elastic deformation in a semi-infinite half-space surrounding a fault, and iteratively adjust the fault parameters to match the observed surface topography [e.g., Okubo and Schultz, 2004; Watters, 2004]. Schultz [2000] proposed a model incorporating a blind thrust fault, a shallow back-thrust fault, layer-parallel shear in the near-surface layers, and a possible localized décollement at depth, favoring a typical thrust fault dip of 30°. Watters [2004] used elastic dislocation models to constrain the fault geometry underlying wrinkle ridges in Lunae and Solis Planum, the latter of which includes the target of the analysis in this work. The results of Watters [2004] support a thrust fault with a dip of 30° extending to a depth of 2 km, a dip of 10° from 2–4 km depth, transitioning to a horizontal décollement at 4.5 km depth. Okubo and Schultz [2004] analyzed a wrinkle ridge in Solis Planum neighboring the ridge analyzed by Watters [2004], using a similar approach but with a different starting geometry. Okubo and Schultz [2004] assumed a listric fault geometry with a secondary backthrust, but with no transition to a horizontal décollement at depth. Results of that analysis favored a thrust fault with a dip of 30° extending from the near surface to depths of 2–6 km, and shallower dips of 8–15° then extending down to the maximum depth of faulting at 6.5–11 km depth. The results of those two studies are broadly similar, both favoring listric geometries with 30° dips in the shallow subsurface, transitioning to gentler dips at deeper levels. However, the results of those studies differ in the depth of transition to shallow dips (2 km vs. 2–6 km) and in the existence of a horizontal décollement at depth.

In contrast, other studies have inferred more steeply dipping faults extending to greater depths. Golombek et al. [2001] used the stair-step topography across sets of wrinkle ridges to infer thrust faults with typical dips (25–30°) extending to depths of 10’s of km, and possibly to the base of the lithosphere. Montési and Zuber [2003] used models of localization instabilities to infer that the regular spacing of wrinkle ridges in swarms is controlled by the thickness of the lithosphere, with a model predicting lithosphere-scale thrust faults dipping at 45° (though this model was not intended to reproduce the details of the tectonics or the observed small-scale surface topography of the ridges). An average dip of 28° was measured for small-scale thrust faults exposed in outcrop in the Valles Marineris interior layered deposits [Okubo, 2010], though these faults are not likely representative of wrinkle ridges due to their different scale and geologic setting. On the Moon, wrinkle ridges encircling the mascon gravity anomaly within the Crisium impact basin were modeled as extending between top depths of 0–1.5 km and bottom depths of ~20 km, with surface dips ranging from 15° to 25°, and evidence for a listric structure beneath some segments [Byrne et al., 2015]. However, the tectonic architecture of that wrinkle ridge may not be representative of other ridges, given the unique setting of that wrinkle ridge at the edge of a mascon basin.

As a result of the lack of adequate seismic data suitable for imaging faults and the limited access to clear exposures of wrinkle ridge faults in outcrops on other planets, all previous studies of wrinkle ridges have relied on inferences based on surface deformation and/or assumptions based on terrestrial analogs to constrain the thrust fault dip. However, mechanical models of wrinkle ridge fault geometry are faced with a large parameter space encompassing a range of mechanical properties, fault dips, depths to the tops and bottoms of the faults, fault displacements, and secondary back-thrust fault geometries. While geodynamic studies commonly assume a thrust fault dip of 30°, this dip is an idealized average that is not representative of many terrestrial faults. On Earth, typical thrust fault dips lie in the range of 12°–60°, with broad peaks in the distribution approximately centered at 30° and 60° representing simple thrust faults and reactivated normal faults, respectively [Sibson and Guoyuan Xie, 1998].

In this work, we present the first direct observations of a thrust fault associated with wrinkle ridge formation. We analyze a feature exposed in the wall of Melas Chasma, a part of the Valles Marineris canyon system. Here, erosional widening of the original tectonic basin of Melas Chasma has exhumed the fault plane of a wrinkle ridge that has been made resistant to erosion. This exposure allows direct measurement of the fault dip through various methods.

2. Observations and interpretations

The target feature analyzed in this study is a topographic ridge protruding from the southern wall of Melas Chasma that extends ~70 km and descends ~5 km from the plateau to the floor of Valles Marineris, hereafter referred to as the Melas ridge (Figure 1). This feature is unique because its length is more than twice that of the next longest wall ridge in Melas, and thus it is clearly not a typical feature resulting from trough wall erosion. The Melas ridge is aligned with a prominent wrinkle ridge on the adjacent plateau. The alignment of this anomalously long and straight ridge projecting from the trough wall with a wrinkle ridge on the adjacent plateau surface suggests that the ridge is related to the fault plane responsible for that wrinkle ridge [Sharp, 1973]. As will be shown below, this supposition is further supported by the results of this study, which reveal that the Melas ridge can be accurately fit by a planar surface with a dip angle consistent with that inferred from other studies of similar wrinkle ridges, and a dip direction consistent with that inferred from the topography of the wrinkle ridge.

Figure 1.

Figure 1.

Open in a new tab

(a) MOLA topography map of the central Valles Marineris region showing the broader context of the Melas ridge. (b–c) HRSC topography of the Melas ridge extending into Melas chasma, and the aligned wrinkle ridge in the plateau to the south. In b, the black line indicates the wrinkle ridge profile shown in Figure 2b, and the black and white lines show the general trends of the wrinkle ridge and Melas ridge. (d) THEMIS daytime IR image mosaic of the Melas ridge and wrinkle ridge. Inset shows a 4× enlargement of a 1.6 km diameter crater that has been mantled by a surface deposit. Smaller craters are mantled to the point of being difficult to discern in the images.

The wrinkle ridge south of the Melas ridge exhibits typical wrinkle ridge morphology and topography, with a narrow ridge superimposed on a broader arch, and one or more small-scale wrinkles tracing an irregular path along the ridge. The wrinkle ridge becomes indistinct in both topography and image data within ~18 km of the edge of the trough where it intersects the Melas ridge. The area close to the trough wall in which the wrinkle ridge is no longer visible corresponds with a broad unit paralleling the trough wall that is dark in THEMIS nighttime IR images. Within this unit, a number of other ridges and graben become muted or indistinct in appearance, and a number of mantled craters are found. These observations suggest the presence of some manner of sedimentary cover, possibly related to sedimentary deposits found adjacent to the wall of Melas chasma further to the west [Weitz et al., 2010]. The heavily mantled appearance of a number of craters up to 1 km in diameter (Figure 1d) indicate that the deposit is comparable in thickness to the original crater depths, calculated to be up to 0.2 km for a depth-diameter ratio of 0.2, which is comparable to the relief on the wrinkle ridge. We suggest that this wrinkle ridge, as well as a number of other wrinkle ridges and graben in the region, continues uninterrupted to its intersection with the wall of Melas chasma, but is buried beneath a younger mantling deposit within 18 km of the trough wall.

The prominence of the Melas ridge implies that it is comprised of an erosion-resistant material that has been exhumed as the erosional expansion of Valles Marineris preferentially removed the relatively weaker surrounding plateau material. The southern wall of Melas Chasma has retreated southward by ~150 km relative to the original Valles Marineris border fault [Schultz and Lin, 2001] as a result of the erosional enlargement of the trough [Lucchitta et al., 1994]. The alignment of the Melas ridge with a wrinkle ridge suggests that the source of erosion resistant material is associated with the wrinkle ridge fault plane. Given the prominence of the ridge and the igneous environment associated with Tharsis, the intrusion of a dike into the fault zone is the most likely explanation for its increased resistance to erosion, as first suggested by Sharp [1973]. Previous studies found evidence for spectrally-distinct dikes intruded into the ancient crust beneath Tharsis exposed in the lower walls of Valles Marineris [Flahaut et al., 2011], and larger scale intrusions have been proposed to have played an instrumental role in the formation of the Valles Marineris troughs [Andrews-Hanna, 2012b]. In the case of the Melas ridge, intrusion of a basaltic dike into basaltic host rocks would not impart an easily-discernable unique spectral signature to the ridge, consistent with the similarity between the Melas ridge and other wall ridges in multi-spectral HiRISE images. If a propagating dike intersected a preexisting structural weakness, such as a wrinkle ridge thrust fault, it would tend to follow that plane as the path of least resistance. Although we refer to this structure as a dike here, since a non-horizontal orientation is required to explain the Melas ridge, the intrusive body need not be vertical and our results below support a low-angle intrusion.

Although a basaltic dike would not be stronger than the surrounding basaltic plateau rocks by virtue of its composition, the slow crystallization of the dike may lead it to be more competent and resistant to erosion than the surrounding fractured and rubbly surface-emplaced flows. Furthermore, the strong plane of the dike would cut obliquely across the layered strong and weak zones of the horizontal volcanic flows, resulting in a single erosion-resistant plane extending across a range of depths. However, our analyses here are independent of the specific mechanism for increasing the resistance to erosion of the fault plane. Alternative mechanisms of strengthening the fault zone (e.g., aqueous alteration within the fault plane and damage zone, or frictional heating and melting along the fault plane) would not change our interpretations or results. Channeling of fluids through the fault zone and the associated precipitation of aqueous minerals can lead to an increase in strength of weaker rocks [e.g., Okubo and McEwen, 2007], and this could contribute to a strengthening of the weak interflow zones within the plateau volcanic sequences along the fault, thereby leading to the formation of the Melas ridge. Given the extensive erosional enlargement of the southern wall of Melas Chasma, the wall rocks here may be weaker than typical for Valles Marineris, which may make it relatively easier to create an erosion resistant plane. Igneous intrusion is preferred in this case because the other processes are less likely to produce a substantial volume of material more resistant to erosion than the surrounding material, and the lack of a distinctive signature of the ridge in HiRISE color images is more easily explained by a basaltic intrusion into a basaltic host rock.

A topographic profile along the crest of the Melas ridge reveals a steeply sloping upper section dropping down from the plateau surface, which transitions abruptly into a gently arching segment in the central section of the ridge, before dropping steeply again to the chasma floor (Figure 2a). The steep sections at the beginning and end of the Melas ridge are similar to the short steep ridges making up the spur and gully morphology that is common along the walls of Valles Marineris [e.g., Sharp, 1973]. We interpret the first steep section of the crest to be an exposure of material above the erosion-resistant plane, resulting in slopes more typical of the walls of Melas Chasma. Similarly, the final steep slope of the profile is interpreted to result from more erodible material beyond the end of the erosion resistant fault plane, reflecting the maximum northward extent of either the wrinkle ridge fault or the dike intruded into that fault. Alternatively, the final steep slope could result from the complete erosion of the erosion resistant fault plane.

Figure 2.

Figure 2.

Open in a new tab

(a) HRSC topography profile along the crest of the Melas ridge (gray line) compared to the elevation along the best-fit plane at the horizontal position of the ridge crest (dashed black line). (b) MOLA topography profile orthogonal to the wrinkle ridge, revealing typical geometry of a narrow ridge atop a broad arch and topographic step, indicating a fault that dips down toward the west (profile is taken from west to east; see Figure 1b for profile location).

Prior to the quantitative analysis below, we note four independent observations that provide a qualitative indication of the fault dip direction. Topographic profiles orthogonal to the wrinkle ridge associated with the Melas ridge (Figure 2b) reveal the typical pattern of a broad arch and narrow ridge [Watters, 1988, 2004]. The location of the narrow ridge on the east side of the broad arch indicates a fault that dips downward to the west [Watters, 2004]. The topographic step down toward the east across the wrinkle ridge profile also implies a thrust fault dipping to the west [Golombek et al., 2001]. Profiles taken further to the south show the narrow ridge on the opposite side of the broad arch, which suggests a possible reversal in the vergence further south along the same wrinkle ridge, as is commonly observed [Watters, 1988, 1991, 2004; Okubo and Schultz, 2004]. There is a 6–8 km horizontal offset between the southern end of the Melas ridge and the northern end of the wrinkle ridge where they meet along the trough wall, suggesting a fault plane dipping to the west if the Melas ridge represents an exposure of the fault below the surface. Similarly, the strike of the Melas ridge is rotated slightly counter-clockwise relative to the strike of the wrinkle ridge, as would be the case if it were exposing successively deeper sections of a west-dipping fault. Finally, the arched shape of the central segment of the ridge profile is curved in map view, protruding further to the east in its highest central section, again implying a fault that dips to the west. Thus, all qualitative indications of the fault dip direction support a west-dipping fault. The following sections will provide a quantitative constraint on the fault dip angle from direct observations of the fault plane in the form of the Melas ridge and wrinkle ridge morphologies.

3. Methods

3.1. Vector analysis

A simple approximation of the fault plane can be acquired by taking the cross product of two non-parallel vectors that are inferred to lie along the fault plane – in this case a near-horizontal vector along the wrinkle ridge and an inclined vector following the crest of the Melas ridge. The cross product results in a vector perpendicular to the plane that contains both of these vectors, which we infer to be the fault plane (Figure 3). This method operates under the assumption that both the Melas ridge and the associated wrinkle ridge each either lie on the fault plane or are parallel to a vector that lies on the fault plane (e.g., the surface expression of the wrinkle ridge may lie vertically above the upper tip of the thrust fault and thus may be offset from the fault plane itself). This method also relies on the assumption that the fault surface is adequately approximated by a plane over the depth range in which it is exposed as the Melas Ridge and over the length range of the combined wrinkle ridge and Melas ridge. The vector used to represent the Melas ridge was taken from an approximate fit to the central arched section of the profile that appears to be fault controlled. The vector used to represent the wrinkle ridge was taken from an approximate fit to the wrinkle, which is interpreted as the expression of backthrust faulting.

Figure 3.

Figure 3.

Open in a new tab

Cartoon depicting a perspective view (looking south) of the Melas ridge (represented by vector MR), the wrinkle ridge (vector WR), and the plane calculated from its normal vector n. Vectors and plane shown are for schematic purposes only.

3.2. Least-squares plane fitting

The cross product method depends on both the morphology of the wrinkle ridge and the Melas ridge. We support our assumptions that the two are related by performing a second analysis that is independent of the wrinkle ridge geometry. For this analysis, we used HRSC (High Resolution Stereo Camera; [Neukum et al., 2004; Jaumann et al., 2007]) topography to extract a cloud of points in three dimensions that represent the fault-controlled section of the Melas ridge crest. The height accuracy of the HRSC topography of ~10 m [Jaumann et al., 2007] is much less than the RMS misfit of our best-fit planes to the ridges found in the next section, and therefore has no impact on our results. We then found the best-fit plane to these points using a least squares regression.

3.3. Monte Carlo inversion

In order to quantify the uncertainty in the geometry of the fault plane fit to the Melas ridge crest, we used a Metropolis Hastings Markov chain Monte Carlo (MH-MCMC) inversion to explore the range of allowable strikes and dips [Metropolis et al., 1953; Andrews-Hanna et al., 2013]. This model interrogates the parameter space (strike, dip, and position of the plane) beginning at a random starting location and calculating the RMS misfit between the points along the Melas ridge and a plane of a given orientation. The assumed parameters were perturbed in each step, and solutions were accepted or rejected based on the RMS misfit. Solutions with smaller misfits were always accepted, while those with greater misfits were sometimes accepted based on a likelihood function (see Supplementary Information of Andrews-Hanna et al. [2013] for a detailed description). The model produces a probability distribution for the parameters of interest.

As a result of the large lateral extent of the Melas ridge compared to its vertical relief, a wide range of near-horizontal planes exhibit only modest misfits to the Melas ridge crest. However, we can use the observed wrinkle ridge strike as an a priori constraint on the parameter space explored in the inversion. The strike was limited to be within ±10° of the mean wrinkle ridge strike, based on the measured variability of the wrinkle ridge strike. The dip was allowed to range from 0° to 45°, based on preliminary results indicating that this range was sufficient to capture the full range of acceptable solutions. The horizontal position of the plane was allowed to vary by up to 5 km. The MH-MCMC was run for 100,000 iterations to explore the parameter space.

3.4. Forward modeling of ridge formation and test of the model assumptions

Although it is a simple matter to fit a plane to an observed structure, in this case we do not directly observe the structure responsible for the ridge as being distinct from its surroundings in any way other than its relief. The ridge is clearly more resistant to erosion than the surrounding plateau material, and it is likely associated with the plane of the wrinkle ridge thrust fault, but these inferences alone do not necessarily require that the crest of the ridge follow that plane. We cannot a priori rule out the possibility of a systematic bias, in that the ridge crest may depart further from the fault plane in more deeply eroded sections, introducing errors into the dips derived from the ridge crest.

As a test of our assumption that a plane fit to the ridge crest should be representative of the underlying erosion-resistant plane, we forward modeled the formation of a structurally controlled ridge as an erosion-resistant plane (representing the dike-intruded wrinkle ridge fault plane) is exposed during the erosion of an unrelated and approximately orthogonal fault scarp (representing the original wall of Melas chasma). A tectonic step was formed in the model along an assumed normal fault with a dip of 60° that accumulates a net vertical displacement of 4 km over a period of 106 years. Throughout this period of faulting and beyond, the landscape was subjected to erosion by advective (e.g., stream runoff) and diffusive (e.g., mass wasting and creep) processes in a finite difference model [Kooi and Beaumont, 1994; Petit et al., 2009]. A plane was imposed in the model, with a strike of 0.0° and a dip of 30°, along which the resistance to both forms of erosion was increased by a factor of 5 relative to the surrounding rock. The erosion resistance scale factor was tapered to 1 in the top kilometer of the domain simulating either a narrowing of the dike or it’s weakening by near-surface aqueous alteration. In order to approximate the effects of natural variations in the width or strength of the dike, the erosion resistance of the plane was scaled by a periodic function of position in the N-S direction in the model domain ranging from 1/3 to 1. The erosion resistance in the model was chosen to produce a ridge that is, to first order, comparable to the Melas ridge in scale, and should not be interpreted as having physical meaning.

We emphasize that this model is not intended to recreate the formation of Valles Marineris, the faceted spurs characterizing its walls, or the Melas ridge – it is intended only to test the hypothesis that the topographic expression of a ridge can be used to reconstruct the dip of the associated plane of erosion-resistant material if that material has been subjected to advective and diffusive erosional processes. The topography created by this forward model was then analyzed in the same manner as the topography of the Melas ridge crest, and used to test the fidelity of the resulting best-fit plane to the plane input into the model.

4. Results

The cross product analysis of vectors representing the wrinkle ridge and Melas ridge yields a normal vector to a plane that strikes N17°E and has a dip of 18° to the west. The least-squares analysis of the best-fit plane to the Melas ridge results in a strike of N21°E and a dip of 13° to the west. We assess how well the plane fits the ridge crest by calculating the orthogonal distance to the best-fit plane for each point, which results in a RMS misfit of 298 m relative to the 2.5 km of relief along the profile (Figure 2a).

In order to test the possibility that the fault is listric (i.e., has a dip that decreases gradually with depth) over the depth range represented by the Melas ridge, we repeat the least-squares plane fitting for subsections of the Melas ridge with lengths of 50 points (~4 km) progressing horizontally along the ridge, and thus also progressing in depth below the plateau surface (Figure 4). Although these subsections of the ridge result in substantial scatter in the dip of the planes due to the small-scale variability along the ridge crest, there is no clear trend of dip with location along the ridge.

Figure 4.

Figure 4.

Open in a new tab

Dip of the least-squares fits of planes to 50-point (~4 km) subsections of the Melas ridge crest as a function of distance north along the ridge. A limited number of subsections exhibited large dips that were excluded from the plot.

From the Monte Carlo model, after 100,000 iterations the plane with lowest RMS error (269 m) has a strike of N21.6 °E and dip of 12.2°NW. We note that the Monte Carlo model is not intended to find the best-fit solution, but rather to interrogate the broader parameter space in order to find the range of acceptable solutions. The misfit from the Monte Carlo model cannot be directly compared to the misfit from the least-squares approach, since for efficiency the misfit in the Monte Carlo model is defined as the vertical distance between points along the ridge and the points on the plane. The range in dips from the Monte Carlo model includes values between 6° and 21° at the 1σ level, or between 2° and 29° at the 2σ level, excluding the commonly assumed value of 30° (Figure 5a–b). The RMS misfit as a function of strike and dip for the accepted models reveals a minimum that is extended in strike at low dip angles (Figure 5c), since the broad horizontal range and narrow vertical range of the Melas ridge crest results in modest misfits for a wide range of near-horizontal planes. As discussed above, the a priori information on the allowable range of strikes from the wrinkle ridge was used to limit the allowable range of parameter space considered.

Figure 5.

Figure 5.

Open in a new tab

Normalized probability curves of allowable (a) dips and (b) strikes of the plane from the Monte Carlo model. (c) A scatter plot of the RMS misfit as a function of strike and dip reveals a minimum extended in strike at low dip angles. Note that the Monte Carlo model allows variations in the location of the plane as well as its orientation, resulting in the observed scatter in RMS misfit at a given strike and dip.

The landscape evolution model produced a ridge with a length of ~20 km and vertical relief of ~2 km, protruding from an eroded fault scarp (Figure 6). The ridge and scarp superficially resemble the Melas ridge and the south wall of Melas chasma, respectively. The best-fit plane for this ridge has a strike of 0.5° and dip of 29.9°, in close agreement with the true strike and dip of 0.0° and 30.0°, respectively. The results of this test supports the assumption that the crest of a ridge that formed as a result of diffusive and advective erosion of a resistant fault plane closely follows that plane, and can be used to reconstruct the strike and dip of the associated fault.

Figure 6.

Figure 6.

Open in a new tab

(a) Topographic contour map of the synthetic ridge created by the landscape evolution model of an erosion resistant plane being exposed by a retreating tectonic scarp. (b) Profile along the crest of the synthetic ridge along with the elevation of the best-fit plane at the horizontal location of the ridge crest.

5. Conclusions and discussion

The goal of this work is to use the Melas ridge as a direct observational constraint on the thrust fault geometry related to one of the Melas Dorsa wrinkle ridges. Our interpretation, that the Melas Ridge is the expression of an erosion resistant fault plane, is well supported by agreement between the results of the three methods of investigation, as well as the three qualitative indicators of the fault dip direction. The close agreement between the cross product method (treating the wrinkle ridge and Melas ridge as simple vectors along a plane) and least squares methods (fitting only the points along the Melas ridge for the best-fit plane) supports our interpretations that the two features are related and that the Melas ridge crest preserves the thrust fault geometry. The relationship between the two structures is further supported by the agreement between our derived dip angle and the results of mechanical modeling studies of nearby wrinkle ridges (see discussion below), and is also supported by the agreement of the dip direction to the west determined from qualitative observations of the wrinkle ridge profile and its relationship to the Melas ridge and the dip direction from the quantitative analyses. The agreement of all these separate lines of evidence strongly support the interpretation that the Melas ridge is an exhumed wrinkle ridge fault plane that has been strengthened by an unknown mechanism (possibly dike intrusion) and exhumed during the erosional enlargement of Melas Chasma. The relatively low RMS misfits from the least-squares and Monte Carlo analyses support our assumption that the fault surface is adequately approximated by a plane and implies that there is minimal listric character down to 3.5 km beneath the plateau surface. This is further supported by the best-fit fault dips for subsections of the Melas ridge, which despite larger uncertainties from the shorter segments, show no systematic trend of dip with depth. The strikes derived from each method are in agreement with the measured strike of the wrinkle ridge of N16.9°E.

The dip supported by all three methods of 13° (+8°, −7°) is significantly lower than the 30° value commonly assumed for planetary thrust faults [Schultz, 2000a; Golombek et al., 2001; Nahm and Schultz, 2011], but is within the range typical of terrestrial thrust faults [Sibson and Guoyuan Xie, 1998]. The best-fit dip is lower than the dip of 15–45° determined from seismic data of a wrinkle ridge analog in the Yakima fold belt in Washington state, though the fault in that study was poorly resolved and the dip poorly constrained [Lutter et al., 1994].

The fault-controlled section of the Melas ridge has a minimum depth below the adjacent plateau surface of 1 km, supporting the common interpretation of wrinkle ridges as blind thrust faults and consistent with the depth to thrust fault top predicted by some boundary element models [e.g., Okubo and Schultz, 2004]. This places an upper bound on the depth of the upper fault tip, though we cannot rule out the possibility that the fault propagated to shallower depths but was not preserved by emplacement of erosion resistant material, or that the depth to the top of the thrust fault could vary along the length of the wrinkle ridge. Similarly, we cannot constrain the maximum depth of faulting or the geometry of the fault below 3.5 km depth.

The broader applicability of this result depends on whether this wrinkle ridge is representative of wrinkle ridges in general. Although terrestrial thrust faults exhibit a wide range of dips [Sibson and Guoyuan Xie, 1998], these faults occur in a wide range of tectonic settings and host rock lithologies. In contrast, wrinkle ridges on other terrestrial planets are remarkable in their similar morphology and setting within layered volcanic sequences. The Melas Dorsa wrinkle ridge associated with the Melas ridge is unremarkable in every respect, with the exception of its proximity to Valles Marineris, which ultimately resulted in its erosional exhumation. The wrinkle ridges on Solis Planum are of similar length, width, and relief, and thus one might expect similar fault dips.

These results cannot be applied to a subset of wrinkle ridges that differ in their morphology and tectonic setting. In particular, some wrinkle ridges show evidence of shallow structural control associated with buried topography, such as the circular wrinkle ridges inferred to overlie buried impact craters [Allemand and Thomas, 1995; Mangold et al., 1998; Watters, 2004]. On the Moon, wrinkle ridge thrust faults bordering mascon basins appear to follow the uplifted crust-mantle interface beneath those basins [Byrne et al., 2015], and thus may differ markedly in their formation and structure. However, for more typical wrinkle ridges similar to the Melas Dorsa, the results of this study suggest that a shallow dip of ~13° may be a more appropriate assumption than the commonly assumed value of 30°.

If the Melas ridge fault geometry is representative of other wrinkle ridges, our results are incompatible with inferences of thrust faults with dips of ~30° extending from the near-surface to depths of 10’s of km [Golombek et al., 2001], though our analysis does not place a constraint on the maximum depth of faulting itself. Similarly, the observed dip is incompatible with models of wrinkle ridge formation by re-activation of steeply-dipping (55°) normal faults [Tate et al., 2002]. The expression of the fault plane exhumed from within the thick layered volcanic plateau of Tharsis [McEwen et al., 1999] is incompatible with models in which a thrust fault at deeper levels terminates at the base of the layered volcanic pile, and all strain in the overlying volcanic layers is accommodated by folding and layer-parallel shear [Schultz, 2000b; Mueller and Golombek, 2004]. Strain is accommodated by the Melas ridge fault to depths at least as shallow as 1 km within the thick volcanic sequence of Solis Planum, implying that strain accommodation solely by folding and layer-parallel shear is limited to shallower depths. This behavior can easily be understood as a consequence of the increasing frictional strength of the interflow zones between separate volcanic flows with increasing depth and lithostatic pressure.

In comparison with more recent studies using forward mechanical modeling to match the surface deformation above wrinkle ridges, our measured dip of 13° between 1 and 3.5 km depth is consistent with the low dip angles of 10° or 8–15° at greater depths inferred by both Watters [2004] and Okubo and Schultz [2004], respectively. Mechanical models generally support a steeper 30° dip immediately below the surface transitioning to a lower dip. Although our analysis cannot constrain the dip at depths shallower than 1 km, our results are not compatible with the presence of a fault dip of 30° extending from the surface to depths of 2–6 km before transitioning to the lower dip as inferred by those studies [Okubo and Schultz, 2004; Watters, 2004]. Our analysis can neither confirm nor deny the existence of a more steeply dipping fault segment at depths shallower than 1 km, the presence or geometry of backthrust faults, nor the transition to a horizontal décollement at greater depths. Although mechanical models typically assume discrete changes in fault dip for simplicity [e.g., Okubo and Schultz, 2004; Watters, 2004], the resulting geometry is often described as listric. Our analysis supports a planar rather than listric fault in the depth interval of 1–3.5 km, indicating that any steepening of the fault dip at shallower levels must occur rather abruptly above 1 km.

Although we cannot constrain the fault at greater depths, if a dip of 13° is representative of the entire vertical extent of the thrust fault, then faulting down to the brittle-ductile transition depth of 30–50 km [Montési and Zuber, 2003] would result in the faults extending 130–220 km in the cross-strike horizontal direction, which greatly exceeds the observed ridge spacing of 50 km. This favors either thin-skinned models of wrinkle ridge formation with shallow maximum depths of faulting, or a transition to steeper dips at depth. Thin-skinned deformation, however, leaves unanswered the question of the mechanism of accommodation of shortening at deeper levels in the lithosphere [Mueller and Golombek, 2004]. For more discussion of the debate between thick- and thin-skinned deformation, see Golombek and Phillips [2010].

The general agreement in dip between our results based on direct observation and the results of elastic dislocation models [Okubo and Schultz, 2004; Watters, 2004] supports the applicability of such models for understanding compressional tectonics in settings where direct observations of the fault in the subsurface are not possible. On Earth, elastic dislocation models are generally applied to single earthquakes in which most deformation away from the fault plane is elastic [e.g., Simons et al., 2002], before the topography has been modified by the accumulated long-term viscous response at deeper levels [Pollitz et al., 2001]. Thus, although elastic dislocation models rely on some assumptions that are not strictly correct (e.g., a lack of shallow plastic deformation or deep viscous deformation), our results indicate that such models still have relevance for ancient tectonic structures.

If the cause of fault plane preservation is indeed a dike, the Melas ridge also has implications for the tectonic and magmatic evolution of this portion of Tharsis. Capture of a sub-vertical dike by a steeply dipping fault is supported by both observations and modeling [Gaffney et al., 2007], but requires a fault that is within 45° of the angle of the dike. However, it is only the orientation of faults and dikes relative to the principal stress directions that matters, and thus this same result should apply for a horizontal sill being captured by a low-angle thrust fault. Indeed, volcanism is known to occur in compressional tectonic settings and to be associated with thrust faults on Earth [Tibaldi, 2005; Galland et al., 2007b], and physical analog modeling shows that intrusion into thrust fault planes can occur [Galland et al., 2007a]. Thus, if the Melas ridge is the expression of a dike intruded along the thrust fault, it likely represents the capture of a horizontal sill.

The intrusion must have formed after the wrinkle ridge but before the erosional enlargement of Melas Chasma. The generally accepted sequence of tectonism in this region places Valles Marineris formation in the Late Noachian to Early Hesperian, followed by wrinkle ridge formation in the Late Hesperian and beyond [Anderson et al., 2001]. Differential loading across the buried dichotomy boundary beneath Tharsis may have caused a narrow belt of extensional stresses around the present-day location of Valles Marineris [Andrews-Hanna, 2012a]. This extension allowed for large magma bodies to form parallel to Valles Marineris, resulting in chains of collapse pits [Mège et al., 2003] and possibly playing a key role in Valles Marineris formation [Andrews-Hanna, 2012b]. Sub-vertical dikes exposed in the walls of Valles Marineris [Flahaut et al., 2011] likely formed within this extensional belt. In the surrounding plateau, stresses were dominantly compressional after the transition to membrane-flexural support of Tharsis [Tanaka et al., 1991; Banerdt and Golombek, 2000; Golombek and Phillips, 2010] resulting in long swarms of wrinkle ridges concentric to the center of Tharsis [Head et al., 2002]. Continuing magmatism at this time would have likely been in the form of sills, creating conditions favorable for the intrusion into the low-angle fault plane. Erosion of the interior layered deposits [Andrews-Hanna, 2012b] and erosional widening of the troughs [Lucchitta et al., 1994; Schultz, 1998] would have then exposed the intruded fault plane as the Melas ridge. This ridge is particularly prominent because it is located in the region of maximum erosional widening of the troughs. However, the lack of features similar to the Melas ridge, despite ~10 other wrinkle ridges intersecting the south wall of Melas Chasma, implies that this particular wrinkle ridge may have had some unique interaction with the magma chambers or sills around Valles Marineris.

Lower dips for wrinkle ridge faults have broad implications for the study of Martian tectonics, estimates of lithosphere thickness and regional heat flow, as well as the inferred amount of planetary contraction. For example, the horizontal strain estimated by the vertical relief across a wrinkle ridge is increased by 83% for an assumed dip of 13° in comparison to an assumed dip of 30°. If horizontal strain is instead calculated from the fault displacement deduced from displacement-length ratios [e.g., Watters et al., 2000; Schultz et al., 2006], the horizontal strain for a 13° dip is 13% greater than that for a 30° dip. The lower dip would similarly increase estimates of global contraction [Nahm and Schultz, 2011], implying that Mars experienced a greater degree of global contraction than previously thought. Low fault dips for wrinkle ridges on other planetary bodies would have similar implications for geodynamics on those bodies.

Key Points:

  • Direct observations constraining the fault geometry underlying a wrinkle ridge

  • Observed fault plane dip of 13° is lower than expected

  • Dip angle has broad impact on studies of martian geodynamics

Acknowledgments and Data

This work was supported by grant NNX14AO75G from the NASA Planetary Geology and Geophysics program to JCAH, and by the generous support of the Southwest Research Institute. We thank Chris Okubo and Jim Dewey for their comments on the manuscript, and Ernst Hauber and two anonymous reviewers for their reviews. The authors declare no competing financial interests. Topography data from the HRSC and MOLA instruments are freely available from http://hrscview.fu-berlin.de and http://pds-geosciences.wustl.edu/missions/mgs/mola.html, respectively.

References.

  1. Allemand P, and Thomas PG (1995), Localization of Martian ridges by impact craters: Mechanical and chronological implications, J. Geophys. Res, 100, 3251–3262, doi: 10.1029/94JE03081. [DOI] [Google Scholar]
  2. Anderson EM (1905), The dynamics of faulting, Trans. Edinburgh Geol. Soc, (8), 387–402. [Google Scholar]
  3. Anderson RC, Dohm JM, Golombek MP, Haldemann AFC, Franklin BJ, Tanaka KL, Lias J, and Peer B (2001), Primary centers and secondary concentrations of tectonic activity through time in the western hemisphere of Mars, J. Geophys. Res, 106, 20563–20585, doi: 10.1029/2000JE001278. [DOI] [Google Scholar]
  4. Andrews-Hanna JC (2012a), The formation of Valles Marineris: 2. Stress focusing along the buried dichotomy boundary, J. Geophys. Res, 117, E04009, doi: 10.1029/2011JE003954. [DOI] [Google Scholar]
  5. Andrews-Hanna JC (2012b), The formation of Valles Marineris: 3. Trough formation through super-isostasy, stress, sedimentation, and subsidence, J. Geophys. Res, 117, E06002, doi: 10.1029/2012JE004059. [DOI] [Google Scholar]
  6. Andrews-Hanna JC, Zuber MT, and Hauck SA (2008), Strike-slip faults on Mars: Observations and implications for global tectonics and geodynamics, J. Geophys. Res, 113, E08002, doi: 10.1029/2007JE002980. [DOI] [Google Scholar]
  7. Andrews-Hanna JC, Asmar SW, Head JW, Kiefer WS, Konopliv AS, Lemoine FG, Matsuyama I, Mazarico E, McGovern PJ, Melosh HJ, a Neumann G, Nimmo F, Phillips RJ, Smith DE, Solomon SC, Taylor GJ, Wieczorek MA, Williams JG, and Zuber MT (2013), Ancient igneous intrusions and early expansion of the Moon revealed by GRAIL gravity gradiometry., Science, 339, 675–678, doi: 10.1126/science.1231753. [DOI] [PubMed] [Google Scholar]
  8. Banerdt WB, and Golombek MP (2000), Tectonics of the Tharsis region of Mars: Insights from MGS topography and gravity, Lunar Planet. Sci. Conf, 31, Abstract 2039. [Google Scholar]
  9. Byrne PK, Klimczak C, Mcgovern PJ, Mazarico E, James PB, Neumann GA, Zuber MT, and Solomon SC (2015), Deep-seated thrust faults bound the Mare Crisium lunar mascon, Earth Planet. Sci. Lett, 427, 183–190, doi: 10.1016/j.epsl.2015.06.022. [DOI] [Google Scholar]
  10. Chicarro AF, Schultz PH, and Masson P (1985), Global and regional ridge patterns on Mars, , 174, doi: 10.1016/0019-1035(85)90025-9. [DOI] [Google Scholar]
  11. Flahaut J, Mustard JF, Quantin C, Clenet H, Allemand P, and Thomas P (2011), Dikes of distinct composition intruded into Noachian-aged crust exposed in the walls of Valles Marineris, Geophys. Res. Lett, 38, L15202, doi: 10.1029/2011GL048109, doi:10.1029/2011GL048109. [DOI] [Google Scholar]
  12. Gaffney ES, Damjanac B, and Valentine GA (2007), Localization of volcanic activity: 2. Effects of pre-existing structure, Earth Planet. Sci. Lett, 263, 323–338, doi: 10.1016/j.epsl.2007.09.002. [DOI] [Google Scholar]
  13. Galland O, Cobbold PR, De Bremond J, and Hallot E (2007a), Rise and emplacement of magma during horizontal shortening of the brittle crust: Insights from experimental modeling, , 112, B06402, doi: 10.1029/2006JB004604. [DOI] [Google Scholar]
  14. Galland O, Hallot E, Cobbold PR, and Ruffet G (2007b), Volcanism in a compressional Andean setting : A structural and geochronological study of Tromen volcano (Neuquen province, Argentina), Tectonics, 26, TC4010, doi: 10.1029/2006TC002011. [DOI] [Google Scholar]
  15. Golombek MP, and Phillips RJ (2010), Mars Tectonics, in Planetary Tectonics, edited by Watters TR and Shultz RA, pp. pp. 183–232, Cambridge University Press. [Google Scholar]
  16. Golombek MP, Anderson FS, and Zuber MT (2001), Martian wrinkle ridge topography- Evidence of subsurface faults From MOLA, J. Geophys. Res, 106, 23,811–23,821. [Google Scholar]
  17. Head JW, Kreslavsky MA, and Pratt S (2002), Northern lowlands of Mars: Evidence for widespread volcanic flooding and tectonic deformation in the Hesperian Period, J. Geophys. Res, 107, 5004, doi: 10.1029/2000JE001445. [DOI] [Google Scholar]
  18. Jaumann R, Neukum G, Behnke T, Duxburry TC, Eichentopf K, van Gasselt S, Giese B, Gwinner K, Hauber E, Hoffmann H, Hoffmeister A, Köhler U, Matz K-D, McCord TB, Mertens V, Oberst J, Pischel R, Reiß D, Ress E, Roatsch T, Saiger P, Scholten F, Schwarz G, Stephan K, Wählisch M, and HRSC-Co-Investigator-Team (2007), The High Resolution Stereo Camera (HRSC) Experiment on Mars Express: Instrument Aspects and Experiment Conduct from Interplanetary Cruise through Nominal Mission, Planet. Space Sci, 55, 928–952. [Google Scholar]
  19. King GCP, Stein RS, and Lin J (1994), Static stress changes and the triggering of earthquakes, Bull. Seism. Soc. Am, 84, 935–953. [Google Scholar]
  20. Knapmeyer M, Oberst J, Hauber E, Wählisch M, Deuchler C, and Wagner R (2006), Working models for spatial distribution and level of Mars’ seismicity, J. Geophys. Res, 111, E11006, doi: 10.1029/2006JE002708. [DOI] [Google Scholar]
  21. Kooi H, and Beaumont C (1994), Escarpment evolution on high-elevation rifted margins: Insights derived from a surface processes model that combines diffusion, advection, and reaction, J. Geophys. Res, 99, 12,191–12,209. [Google Scholar]
  22. Lucchitta BK, Isbell NK, and Howington-Kraus A (1994), Topography of Valles Marineris: Implications for erosional and structural history, J. Geophys. Res, 99, 3783–3798, doi: 10.1029/93JE03095. [DOI] [Google Scholar]
  23. Lutter WJ, Catchingst RD, and Jarchow CM (1994), An image of the Coloumbia Plateau from inversion of high-resolution seismic data, Geophysics, 59, 1278–1289. [Google Scholar]
  24. Mangold N, Allemand P, and Thomas PG (1998), Wrinkle ridges of Mars: structural analysis and evidence for shallow deformation controlled by ice-rich décollements, Planet. Space Sci, 46, 345–356, doi: 10.1016/S0032-0633(97)00195-5. [DOI] [Google Scholar]
  25. McEwen AS, Malin MC, Carr MH, and Hartmann WK (1999), Voluminous volcanism on early Mars revealed in Valles Marineris, Nature, 397, 584–586, doi: 10.1038/17539. [DOI] [Google Scholar]
  26. Mège D, and Masson P (1996), Stress models for Tharsis formation, Mars, Planet. Space Sci, 44, 1471–1497, doi: 10.1016/S0032-0633(96)00112-2. [DOI] [Google Scholar]
  27. Mège D, Cook AC, Garel E, Lagabrielle Y, and Cormier M-H (2003), Volcanic rifting at Martian grabens, J. Geophys. Res, 108, 5044, doi 10.1029/2002JE001852. [DOI] [Google Scholar]
  28. Metropolis N, Rosenbluth AW, Rosenbluth MN, Teller AH, and Teller E (1953), Equations of state calculations by fast computing machine, J. Chem. Phys, 21, 1087–1091. [Google Scholar]
  29. Montési LGJ, and Zuber MT (2003), Clues to the lithospheric structure of Mars from wrinkle ridge sets and localization instability, J. Geophys. Res, 108, 5048, doi: 10.1029/2002JE001974. [DOI] [Google Scholar]
  30. Mueller K, and Golombek M (2004), Compressional Structures on Mars, Annu. Rev. Earth Planet. Sci, 32, 435–464, doi: 10.1146/annurev.earth.32.101802.120553. [DOI] [Google Scholar]
  31. Nahm AL, and Schultz RA (2011), Magnitude of global contraction on Mars from analysis of surface faults: Implications for martian thermal history, Icarus, 211, 389–400, doi: 10.1016/j.icarus.2010.11.003. [DOI] [Google Scholar]
  32. Neukum G, Jaumann R, and C.-I. H and Team E (2004), HRSC: The High Resolution Stereo Camera of Mars Express, in Mars Express: The scientific payload, edited by Wilson A, pp. 17–35, ESA, Noordwijk, The Netherlands. [Google Scholar]
  33. Okubo CH (2010), Structural geology of Amazonian-aged layered sedimentary deposits in southwest Candor Chasma, Mars, Icarus, 207, 210–225, doi: 10.1016/j.icarus.2009.11.012. [DOI] [Google Scholar]
  34. Okubo CH, and McEwen AS (2007), Fracture-controlled paleo-fluid flow in Candor Chasma, Mars., Science, 315, 983–985, doi: 10.1126/science.1136855. [DOI] [PubMed] [Google Scholar]
  35. Okubo CH, and Schultz RA (2004), Mechanical stratigraphy in the western equatorial region of Mars based on thrust fault-related fold topography and implications for near-surface volatile reservoirs, Bull. Geol. Soc. Am, 116, 594–605, doi: 10.1130/B25361.1. [DOI] [Google Scholar]
  36. Petit C, Gunnell Y, Gonga-Saholiariliva N, Meyer B, and Séguinot J (2009), Faceted spurs at normal fault scarps: Insights from numerical modeling, J. Geophys. Res, 114, B05403, doi: 10.1029/2008JB005955. [DOI] [Google Scholar]
  37. Pollitz FF, Wicks C, and Thatcher W (2001), Mantle Flow Beneath a Continental Strike-Slip Fault: Postseismic Deformation After the 1999 Hector Mine Earthquake, Science, 293, 1814–1818. [DOI] [PubMed] [Google Scholar]
  38. Schultz RA (1998), Multiple-process origin of Valles Marineris basins and troughs, Mars, Planet. Space Sci, 46, 827–834. [Google Scholar]
  39. Schultz RA (2000a), Localization of bedding plane slip and backthrust faults above blind thrust faults: Keys to wrinkle ridge structure, J. Geophys. Res, 105, 12,035–12,052, doi: 10.1029/1999JE001212. [DOI] [Google Scholar]
  40. Schultz RA (2000b), Localization of bedding plane slip and backthrust faults above blind thrust faults: Keys to wrinkle ridge structure, J. Geophys. Res, 105, 12035, doi: 10.1029/1999JE001212. [DOI] [Google Scholar]
  41. Schultz RA, and Lin J (2001), Three-dimensional normal faulting models of the Valles Marineris, Mars, and geodynamic implications, J. Geophys. Res, 106, 16,549–16,566, doi: 10.1029/2001JB000378. [DOI] [Google Scholar]
  42. Schultz RA, Okubo CH, and Wilkins SJ (2006), Displacement-length scaling relations for faults on the terrestrial planets, J. Struct. Geol, 28, 2182–2193, doi: 10.1016/j.jsg.2006.03.034. [DOI] [Google Scholar]
  43. Sharp RP (1973), Mars: Troughed terrain, J. Geophys. Res, 78, 4063–4072, doi: 10.1029/JB078i020p04063. [DOI] [Google Scholar]
  44. Sibson RH, and Guoyuan Xie (1998), Dip range for intracontinental reverse fault ruptures: Truth not stranger than friction?, Bull. Seismol. Soc. Am, 88, 1014–1022. [Google Scholar]
  45. Simons M, Fialko Y, and Rivera L (2002), Coseismic Deformation from the 1999 M w 7 . 1 Hector Mine , California , Earthquake as Inferred from InSAR and GPS Observations, Bull. Seism. Soc. Am, 92, 1390–1402. [Google Scholar]
  46. Tanaka KL, Golombek MP, and Banerdt WB (1991), Reconciliation of the stress and structural histories of the Tharsis region of Mars, J. Geophys. Res, 96, 15,617–15,633. [Google Scholar]
  47. Tate A, Mueller KJ, and Golombek MP (2002), Kinematics and structural inversion of wrinkle ridges on Lunae and Solis Plana - Implications for the early history of Tharsis, Lunar Planet. Sci. Conf, 33, Abstract 1828. [Google Scholar]
  48. Tibaldi A (2005), Volcanism in compressional tectonic settings : Is it possible?, Geophys. Res. Lett, 32, L06309, doi: 10.1029/2004GL021798. [DOI] [Google Scholar]
  49. Toda S, Stein RS, Reasenberg PA, Dieterich JH, and Yoshida A (1998), Stress transferred by the 1995 Mw = 6.9 Kobe, Japan, shock: effect on aftershocks and future earthquake probabilities, J. Geophys. Res, 103, 24543–24565. [Google Scholar]
  50. Watters TR (1988), Wrinkle ridge assemblages on the terrestrial planets, J. Geophys. Res, 93, 10,236–10,254, doi: 10.1029/JB093iB09p10236. [DOI] [Google Scholar]
  51. Watters TR (1991), Origin of Periodically Spaced Wrinkle Ridges on the Tharsis Plateau of Mars, J. Geophys. Res, 96, 15,599–15,616. [Google Scholar]
  52. Watters TR (2004), Elastic dislocation modeling of wrinkle ridges on Mars, Icarus, 171, 284–294, doi: 10.1016/j.icarus.2004.05.024. [DOI] [Google Scholar]
  53. Watters TR, Schultz RA, and Robinson MS (2000), Displacement-length relations of thrust faults associated with lobate scarps on Mercury and Mars: Comparison terrestrial faults, Geophys. Res. Lett, 27, 3659–3662, doi: 10.1029/2000GL011554. [DOI] [Google Scholar]
  54. Weitz CM, Milliken RE, Grant J. a., McEwen a. S., Williams RME, Bishop JL, and Thomson BJ (2010), Mars Reconnaissance Orbiter observations of light-toned layered deposits and associated fluvial landforms on the plateaus adjacent to Valles Marineris, Icarus, 205, 73–102, doi: 10.1016/j.icarus.2009.04.017. [DOI] [Google Scholar]

RESOURCES