MAKING ISHTAR TERRA, VENUS: MOBILE LID TECTONICS, CONTINENTAL CRUST, and IMPLICATIONS for LIQUID WATER and PLANETARY EVOLUTION Walter S
Total Page:16
File Type:pdf, Size:1020Kb
44th Lunar and Planetary Science Conference (2013) 2541.pdf MAKING ISHTAR TERRA, VENUS: MOBILE LID TECTONICS, CONTINENTAL CRUST, AND IMPLICATIONS FOR LIQUID WATER AND PLANETARY EVOLUTION Walter S. Kiefer, Lunar and Planetary Institute, 3600 Bay Area Blvd., Houston TX 77058, [email protected] Introduction with the mountain belt tectonics. We can quantify the Present-day Venus is very different from Earth. amount of convergence required by a simple Airy Mantle convection on Venus is currently in the stag- isostasy calculation. Based on limited geochemical nant lid regime, with very little surface motion [1]. measurements, the lowlands plains crust on Venus is Due to the 470 °C temperature, liquid water is not cur- basaltic [15]. Assuming crust and mantle densities of rently stable on the surface of Venus. But is it possible 2900 and 3300 kg m-3, the average 3.5 km elevation of that Venus was once more Earth-like? In this work, I Lakshmi Planum requires 29 km of crustal thickening present evidence that the Ishtar Terra highlands on in order to be supported isostatically. The average Venus require thousands of kilometers of crustal con- crustal thickness is estimated to be ~30 km in the vergence and thus imply that mantle convection on plains [16], so the Lakshmi crust is roughly double the Venus once operated in the mobile lid regime. A mo- global mean crustal thickness. Given the size of Ishtar, bile surface requires faults that fully penetrate the brit- doubling the thickness of crust by means of crustal tle lithosphere. In the absence of such faults, the tem- convergence implies ~2000 to 3000 km of crustal con- perature dependence of mantle viscosity results in a vergence. For flow velocities similar to present-day globally continuous, thick, high viscosity layer that plate velocities on Earth (2-10 cm/yr), the necessary enforces stagnant lid convection [2]. Liquid water fa- amount of crustal convergence could occur in ~108 cilitates fault development by lowering pore pressures years. If the relevant density contrast for the isostasy and also lowers the coefficient of friction of materials calculation is between the crust and a mantle melt re- in the fault zone. For these reasons, water has been siduum [14], will be smaller than calculated here, commonly invoked as mechanism for creating weak requiring a greater degree of crustal thickening and plate boundary faults [e.g., 1, 3-6], although other crustal convergence to produce Lakshmi. mechanisms may also be possible [7, 8]. Low density, The mountain belts such as Maxwell Montes are silica-rich crust, similar to continental crust on Earth, significantly taller and their formation by this mecha- may also form portions of Ishtar Terra and would also nism would therefore require greater degrees of crustal imply the presence of water within the Venus mantle convergence. However, the phase transition from ba- [9]. Measurement of seismic receiver functions from a salt to dense eclogite limits the total thickness of basal- small (minimum 2 station) seismic network, supple- tic crust that can form on Venus [17, 18]. An alterna- mented by existing gravity observations, can distin- tive model is that the mountain belts are formed at guish between basaltic and felsic crust in Ishtar. least in part from more felsic material such as granite Crustal Convergence Model [18]. The lower density of this material reduces the Ishtar Terra consists of a flat, central plateau, required amount of crustal thickening and crustal con- Lakshsmi Planum, surrounded on most sides by vergence. Because the mobile lid convection scenario mountain belts. Lakshmi is typically about 3.5 km being considered here likely requires liquid water, fel- above mean planetary radius, whereas the mountain sic igneous material might have formed on Venus over belt peaks are frequently at 6 to 10 km elevation. The the sites of mantle downwelling by a process similar to overall structure is about 2700 km (East-West) by subduction zone volcanism on Earth. 1800 km (North-South); these dimensions do not in- Implications clude Fortuna Tessera to the east or Itzpapalotl Tessera The considerations outlined above suggest that a to the north [10]. Tectonic structures within the moun- crustal convergence model may explain the high to- tain belts indicate an origin by compressional deforma- pography, gravity, and tectonics of Ishtar Terra. This tion, possibly as fold-and-thrust belts [11-13]. Gravity model requires that Venus once existed in a mobile-lid observations indicate that most of Ishtar Terra is com- convective regime, in which the surface experienced pensated at sufficiently shallow depths that the topog- substantial motion, but does not necessarily require the raphy may be supported mostly isostatically by thick- rigid plate rotations that characterizes plate tectonics ened crust [4, 10]. There is a long-wavelength gravity on Earth. Moreover, it is possible that Venus only ex- component, comparable in length to the plateau, that isted in the mobile lid regime on an intermittent basis suggests some deeper support is also present [10, 14]. [19]. Other investigators have previously proposed One way to produce the necessary thickened crust models involving thickening of either the crust or a is in a crustal convergence zone driven by convective near-surface keel of mantle melt residuum over a man- flow in the mantle. Such a model is also consistent 44th Lunar and Planetary Science Conference (2013) 2541.pdf tle convergence zone, but these earlier studies were between the surface and the base of the crust [30], but done prior to the recognition of the importance of stag- the initial seismometers available for high temperature nant lid convection, and thus the earlier studies did not measurements may not be sufficiently sensitive. An emphasize the possible implications of these models alternative approach to resolving the ambiguity is to for liquid water on Venus [14, 20-22]. incorporate ancillary constraints into the model [31]. This model may help to explain observations of In the case of Ishtar, a possible constraint is the differ- 40Ar in the atmosphere of Venus. 40Ar is a radioactive ence in crustal thickness between Ishtar and the low- decay product of 40K and is degassed from the mantle land plains, as constrained by gravity. Such an ap- to the atmosphere over time. Measured on the basis of proach would also require a seismic station in the low- per unit mass of planet, Venus has about half as much lands. One can use existing databases to relate seismic 40Ar in its atmosphere as Earth does [23]. This sug- velocity, density, and rock composition [32] to formu- gests that averaged over the age of the solar system, late the joint gravity-seismology analysis. volcanic degassing has been about half as effective on Venus as on Earth. Assessments of the present-day References [1] Nimmo F. and McKenzie D., Ann. volcanic resurfacing rate on Venus vary considerably Rev. Earth Planet. Sci. 26, 23-51, 1998. [2] Solomatov [e.g., 24, 25], but if Venus spent a portion of its history V. and Moresi L., JGR 101, 4737-4753, 1996. [3] Mo- in the mobile lid convection regime, volcanic out- resi L. and Solomatov V., Geophys. J. Int. 133, 669- gassing during that period of time would make an im- 682, 1998. [4] Simons M. et al., Geophys. J. Int. 131, portant contribution to the atmospheric argon record. 24-44, 1997. [5] Smrekar S.E. et al., in Exploring Ve- Several highland structures on Venus, such as Beta nus as a Terrestrial Planet, AGU Geophys. Monograph Regio and Atla Regio, appear to be best explained as 176, 45-71, 2007. [6] Gerya T.V. et al., Geology 36, hot, rising mantle plumes [26, 27]. On Earth, similar 43-46, 2008. [7] Bercovici D., EPSL 205, 107-121, plumes such as Hawaii produce prominent chains of 2003. [8] Bercovici D. and Ricard Y., PEPI 202-203, volcanos and seamounts that can be used as a record of 27-55, 2012. [9] Campbell I.H. and Taylor S.R., GRL past plate motions. If Venus convected in the mobile 10, 1061-1064, 1983. [10] Kaula W.M. et al., in Venus lid regime for a significant period of time, one might II, Univ. Arizona Press, pp. 879-900, 1997. [11] Keep expect that lithospheric motion over upwelling plumes M. and Hansen V.L., JGR 99, 26,015-26,028, 1994. would create similar chains of volcanos on Venus. [12] Williams C.A. et al., JGR 99, 19,947-19,974, Such volcano chains are not observed, which is a po- 1994. [13] Zuber M.T. and Parmentier E.M., Nature tential problem for the mobile lid hypothesis. Can vol- 377, 704-707, 1995. [14] Hansen V.L. and Phillips canic resurfacing during the transition from mobile lid R.J., Geology 23, 292-296, 1995. [15] Treiman A.H., to stagnant lid convection obscure the record of prior in Exploring Venus as a Terrestrial Planet, AGU Geo- volcanic chains? phys. Monograph 176, 7-22, 2007. [16] Grimm R.E. A Possible Seismic Test and Hess P.C., in Venus II, Univ. Arizona Press, pp. A seismic sensor that functions at Venus’s ambient 1205-1244, 1997. [17] Namiki N. and Solomon S.C., atmospheric temperature is currently in development JGR 98, 15,025-15,031, [18] Jull M.G. and Arkani- [28]. A seismic experiment with a small number of Hamed J., PEPI 89, 163-175, 1995. [19] Loddoch A. et stations could measure both the thickness and average al., EPSL 251, 79-89, 2006. [20] Bindschadler D.L. et composition of the Ishtar Terra crust and thus test the al., GRL 17, 1345-1348, 1990. [21] Lenardic A.