Mechanical Modeling of the Discovery Rupes Thrust Fault: Implications for the Thickness of the Elastic Lithosphere of Mercury

Mercury: Space Environment, Surface, and Interior (2001) 8070.pdf

MECHANICAL MODELING OF THE DISCOVERY RUPES THRUST FAULT: IMPLICATIONS FOR THE THICKNESS OF THE ELASTIC LITHOSPHERE OF MERCURY. T. R. Watters1, R.A. Schultz2, M. S. Robinson3, and A. C. Cook1, 1Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, D.C. 20560 ([email protected]); 2Geomechanics–Rock Fracture Group, Department of Geological Sciences, Mackay School of Mines, University of Nevada, Reno; 3Department of Geological Sciences, Northwestern University, Evanston, Illinois 60208.

Introduction: One of the most remarkable frictional/constitutive properties of the fault are discoveries of the Mariner 10 mission to Mercury unknown [13, 14]. The magnitude and sense of offset was the existence of hundreds of landforms described are specified along the fault, then the stresses and as lobate scarps [1, 2, 3]. Based on morphology and material displacements are completely determined offsets in crater wall and floor materials, lobate using the stress functions for an elastic halfspace scarps are interpreted to be the surface expression of [15]. An acceptable match between the model and the thrust faulting [1, 2, 3, 4, 5]. The largest lobate scarp topography constrains admissible values. We then on the hemisphere imaged by Mariner 10 is calculate the displacement vectors to predict changes Discovery Rupes. Located in the southern in topography due to the surface-breaking thrust fault hemisphere, Discovery Rupes is over 500 km in beneath Discovery Rupes. length [1, 2]. New topographic data for Mercury, Results and Implications: Iteratively adjusting derived from digital stereoanalysis, using updated the displacement D, fault dip θ, and depth of faulting Mariner 10 camera orientations [6, 7], indicates that T, good fits to the topography are obtained for a Discovery Rupes is up to ~1.5 km high [4, 5]. These relatively narrow range of the fault parameters data are also providing the first quantitative (Figure 2, 3). Depths of faulting T < 30 km and T > measurements about the morphometry of Discovery 40 km (Figure 2) produce unacceptable fits to the Rupes. Using the new topographic data, it is possible topography. The best fits to the topography across to test the validity of kinematic models proposed for Discovery Rupes are for a depth of faulting T = 35 to lobate scarps by mechanically modeling the long and 40 km, fault dip angle θ = 30° to 35°, and D = 2.2 km short wavelength topography across Discovery (Figure 4). A tapered displacement distribution with Rupes. minima at the fault tips is assumed based on Topography and Analysis: Topography of examining the offset where the fault breaks the Discovery Rupes was obtained using an automated surface. Where the Discovery Rupes thrust fault cuts stereo matching process that finds corresponding the floor of Rameau crater (Figure 1), there is no points in stereo images using a correlation patch [cf. significant offset suggesting that the cumulative 8]. A stereo intersection camera model is then used to structural relief developed while the fault was blind find the closest point of intersection of matched and propagating up toward the surface. points which specifies their location and elevation. Our results suggest that the Discovery Rupes The derived digital elevation model (DEM) has a grid thrust fault cuts the mercurian crust to a depth of up spacing of 2 km/pixel (Figure 1) [also see 9, 10]. to 40 km. There are examples of terrestrial thrust The greatest relief on Discovery Rupes occurs faults that cut to comparable depths. The Wind River roughly midway long the length of the scarp (south of thrust in the Rocky Mountain foreland in Wyoming the 60-km-diameter Rameau crater, Figure 1). The extends to a depth of 36 km (with a uniform dip of average relief of the scarp in this area is about 1.3 km 30° to 35°) [16], cutting the entire elastic lithosphere (Figure 2) [4]. The new topographic data reveal a Te. On Earth’s continents, Te typically coincides with shallow trough roughly 100 km west of the base of the thickness of the seismogenic crust Ts (ranging the scarp (Figure 1). This trough is interpreted to be from ~10-40 km), with Te often less than Ts [17, 18]. evidence of a trailing syncline and the distance On Mars, the Amenthes Rupes thrust fault extends to between it and the surface break defines the cross- a depth comparable to estimates of Te for the strike dimension of the upper plate of the Discovery highlands [19] (~20-30 km) [20, 21]. If the Discovery Rupes thrust fault. Rupes thrust fault extends to a depth of ~40 km, it Mechanical Model: The 3-D boundary element may cut the entire mercurian elastic and seismogenic dislocation program Coulomb 2.0 was used to predict lithosphere. An estimate of Te from depth of faulting the surface displacements associated with the provides insight into the thermal structure of the Discovery Rupes thrust fault. The dislocation method mercurian crust at the time the faults formed. The has been successfully applied to terrestrial faults [11, effective elastic thickness is though to be controlled 12] where the magnitude of offset along the fault is by the depth of the 450° to 600°C isotherm below known and when the remote stress state or which the lithosphere is too weak to support long- Mercury: Space Environment, Surface, and Interior (2001) 8070.pdf

LARGE-SCALE LOBATE SCARPS ON MERCURY: T. R. Watters et al.

term stresses [17, 22]. Our results suggest a thermal and Cook, A.C. (2001) this volume. [10] Cook, A.C., gradient of ~8°K km-1 and a heat flux of ~24 mW m-2 Watters,T.R. and Robinson, M.S. (2001) this volume. [11] King, G.C.P., Stein, R.S. and Rundle, J.B. (1988) J. Geophys. at the time Discovery Rupes formed. These estimates Res., 93, 13307-13318. [12] Taboada, A., Bousquet, J.C. and will be testable when MESSENGER [23] and Bepi Philip, H. (1993) Tectonophysics, 220, 223-241. [13] Rudnicki, Colombo returns new gravity and topographic data. J.W., (1980) Annu. Rev. Earth Planet. Sci., 8, 489-525. [14] References: [1] Strom R.G., Trask N.J. and Guest J.E. (1975) Bilham, R. and King, G. (1989) J. Geophys. Res., 94, 10204- J. Geophys. Res., 80, 2,478-2,507. [2] Cordell, B.M. and Strom 10216. [15] Okada, Y., (1992) Bull. Seismol. Soc. Am., 82, R.G. (1977) Phys. Earth Planet. Inter., 15, 146-155. [3] 1018-1040. [16] Brewer, J.A., Smithson, S.B., Oliver, J.E., Melosh H.J. and McKinnon W.B. (1988) in Mercury, 374-400. Kaufman, S., and Brown, L.D. (1980) Tectonophysics, 62, [4] Watters T.R., Robinson M.S. and Cook A.C. (1998) 165-189. [17] McKenzie, D. and Fairhead, D. (1997) J. Geology, 26, 991-994. [5] Watters T.R., Robinson M.S. and Geophys. Res., 102, 27523-27552. [18] Maggi, A., Jackson Cook A.C. (2001) Planet. Space Sci., in press. [6] Robinson, J.A., McKenzie, D. and Priestley, K. (2000) Geology, 28, 495- M.S., Davies, M.E., Colvin, T.R. and Edwards, K.E. (1999) J. 498. [19] Schultz, R.A. and Watters, T.R. (2001) Submitted to Geophys. Res., 104, 30847-30852. [7] Robinson, M.S. and Geophys. Res. Lett. [20] Zuber, M.T., et al., (2000) Science, Lucey, P.G. (1997) Science, 275, 197-200. [8] Day, T., Cook, 287, 1788–1793. [21] Nimmo, F., (2001) Lunar Planet. Sci., A.C. and Muller, J.P. (1992) Intern. Arch. Photo. Rem. Sensing XXXII, abs. 1370. [22] Watts, A.B. (1994) Geophys. J. Int., 29, 801-808. [9] Wilkison, S.L., Robinson, M.S., Watters, T.R. 119, 648-666. [23] Solomon et al., (2001) Planet. Space Sci., in press.

Figure 1. Color-coded DEM generated using Mariner 10 stereo pair 27399 and 166613, overlaid on an image mosaic. The Figure 2. Comparison between predicted structural relief white line indicates the location of the topographic profile and a topographic profile across Discovery Rupes. Depth of across Discovery Rupes (white arrows) shown in Figure 2 and faulting T is varied while the fault-plane dip and 3. The black arrow and dashed line shows the location of a displacement are constant. Profile location is shown in shallow depression. Rameau crater is indicated by an R on the Figure 1. Vertical exaggeration is 30X. image. Elevations relative to 2439.0 km reference sphere.

Figure 3. Comparison between predicted structural relief Figure 4. Difference between topography and predicted and a topographic profile across Discovery Rupes. Fault- structural relief for model runs shown in Figure 2. Plots plane dip θ is varied while the depth of faulting and indicate that the best fit is obtained for depth of faulting T = displacement are constant. Profile location is shown in 35-40 km. Vertical exaggeration is 30X. Figure 1. Vertical exaggeration is 30X.