LP! Contribution No. 789 119 more nearly symmetric topographic profiles, such asDcvanaChasrna, 238, 1380• [17] McKenzie D. et al. (1992)JGR, inpress. [18] Ford and on the basis of small-scale morphology are clearly extensional P. G. and Pettengill G. H.(1992) JGR, inpress. [19] Sandwell D. T. rifts. and Schubert G.(1992) JGR, in press. [20] Solomon S• C. and Head Much of the tectonic behavior on Venus appears to be more L W. (1982) JGR, 87, 9236. [21] Phillips R. L and Matin M. C. reminiscent of aedvely deforming continental regions than of (1984) Amut. Rev. Earth Planet. Sci., 12, 411. [22] Phillips R• L et oceanic regions on Earth. In particular, as in tectonically active al. (1991) Science, 252, 651. [23] Bindschadler D. L. eta]. (1992) continental areas, deformation is typically distributed across broad JGR, in press. [24] Head J.W. et tl. (1991) Science, 252, 27_. zones one to afew hundred Idlometers wide separated by compara- 'N93-14383 :! 7 =', " tively stronger and less deformed blocks having dimensions of htmdreds of kilometers. On Earth, the continental lithosphere in THE SPATIAL DISTRIBUTION OF C()RONAE OI_VENUS. tectonically active areas is weaker than typical oceaniclithosphere S. W. Squyres !,G. Schubert 2,D. L. Bindschadlc#, D. M. Janes t. because of the greater thickness of more easily deformable cnzst. As L E. Moersch t,W. Moore 2,P. Olson 3,J. T. Ratclift n. E.R. Stofan4, noted above, because of the higher surface temperature on Venus, D• L. Turcotte t, tComell University, 2University of California, Los the likely comparable lithospheric thermal gradients on Venus and Angeles, 3Johns Hopkins University, 4Jet Propulsion Laboratory, Earth [20,21], and the strong temperature dependence of ductile USA. behavior, the lithosphere on Venus should behave inaweak manner forcrustal thicknesses less than are typical of continental regions on Coronae onVenus arelarge,generallycircularsurfacefeatures Earth. thathave distinctivetectonic,volcanic,and topographicexpres- Status of Models: A major challenge in unraveling the tec- sions.They rangeindiameterfrom lessthan 200 km toatleast tonic evolution of Venus is to tmderstand the interaction between 1000 kin.Data from theMagellan spacecrafthave now allowed mande convection and the lithosphere. The hotspot [22] and coldspot complete globalmapping ofthespatialdistributioonfcoronaeon [23] models for the formation and evolution of major highlands on the planet.Unlike impact craters,which show a random (i.e., Venus are distinguishable on the basis of the predicted sequence of Poisson)spatialdistributio[nI],thedisWibutionofcoronacappears events and the time-dependent relationship between topography and tobenonrandom. We investigatethedistributiohnereindetaila,nd gravity. Both models fa.cc difficulties in their present forms, how- explore its implications in terms of mantle convection and surface ever. atleast partly because both the theology of the upper crust of modification processes. Venus and the observed patterns ofmagmatism and deformation arc Figure 1shows the distribution of coronae and _a-related more complex than current models for the deformation and mag- features on Venus, inasimple cylindrical (Plate Carr6e) projection. marie response of the crust to mantle flow• Any global tectonic The map gives the locations of 311 features identified in all of the model, of course, must also consider the formadon and character- Magellan data taken to date• These features include coronae; they istics of the lowlands, including the large apparent depths of also include radially fractured domes, which are believed to be isostade compensation [22] and relatively recent lowland volcan- eoronae in an early or arrested stage of development [2]. The map ism [24]. All dynamical models to date require special pleading to gives the qualitative sense that the distribution is nonrandom, with explain Ishtar Terra. Both the rifted, volcanically active highlands distinct areas of clustering and sparseness. However, this assertion and at least the larger coronae are generally regarded as sites of requires testing, especially because there are some significant gaps upwelling mantle flow and magma generation by pressure-release in the Magellan data that artificially introduce areas of sparseness partial melting of mantle material. These two classes of features, in the corona distribution. however, have very different tectonic and morphological manifes- In order to test for nonrandomness, we apply asimple nearest- tations at the Venus surface• If both are products of mantle up- neighbor test. In this test, we measure the distance, r,of each feature welling, then multiple scales of mantle convection are indicated and from its nearest neighbor on the planet. The percentage of points the different morphologies of the two classes of features must be with a nearest-neighbor distance greater than r is then plotted as a related todifferences in the geometry, buoyancy flux, or duration of function of r in Fig. 2. The next step is to compare this result to the flow in the two types of upwelling regions. The assessment of result expected for a spatially random distribution. There exist existing dynamical models for the tectonic evolution of Venus and theoretical tleatments for a nearest-neighbor curve of the sort in the development of the next generation of models will require an understanding of geological relationships at all scales, from the , , _ , , T_ • _ . T _ ,--7 highest resolution available to global patterns. High-resolution P measurements of the global gravity field later in the Magellan / - o••• mission will provide key data for testing and reining models. 4f_J_.• " . References: {1] Masursky H. ct at. (1980) JGR, 85, 8232. t • • '• i ••• ." • , • [2] Campbell D. B. et al. (1983) Science, 22I, 644. [3] Barsukov •. •- . 3" _t _0• "•-" • • t V. L.etat. (1986)Proc.LPSCI6th, inJGR, 91, D378. [4] Pettcngi]l •: ",t-7.. "_ nl :::• " " G. H. etat. (1991) Science, 252,260. [5] Solomon S.C. etal. (1991) ••.%,• • ! . Science, 252,297. [6] Solomon S. C. et al. (1992) JGR. in press. '° t -:'• •" • "" . • •" [7] Phillips R. J. et at. (1991) Science, 252,288. [8] Schaber G. G. •• • :• "¢ • •,kot• et al. (1992) JGR, in press. [9] Weertman J•(1979) PEPI, 19, 197. dc_ I • [10] Solomon S. C. and Head J. W. (1984) JGR, 89, 6885• l. [I1] Smrekar S. E. and Solomon S. C. (1992) JGR, in press. [12] Baker V. R. etal. (1992)JGR, inpress• [13] McGill G. E.et al. 9o ' [ _J _ [ , J (1981)GRL, 8, 737. [14] Grimm R. E. and Solomon S. C. (1988) o 45 9o 135 180 2;'5 _70 31S 360 JGR, 93, 11911• [151 Zuber M. T. and Parmentier E. M. (1990) Fig. 1. Map showing the distribution of coronae and related features on Icarus, 85, 290• [16] Head J. W. and Crumpler L.S. (1987)Science, Venus. 118 International Colloquium on Venus dependentviscositywiththehelpofarecentlydevelopedapproach Tectonic activity on Venus has continued until geologically [8]and thethermalevolutionofVenus withmeltingand differen- recent time, and most likely the planet is tectordcnlly active at tiationis calculated. present. Several arguments support this inference. The great relief The modelpredicts thatonlyasmallamount ofmeltisextracted andsteep slopes of the mountains and plateau scarps of Ishtar Terra inthefirspteriodofevohdon; however,thiswouldresultinasn-ong and of the equatorial chasm systems are difficult to reconcile with depictioninradioactiveelernentsand devolatilizadonofapartof long-term passive SUplXn't by crustal strength. Because of the high theuppermantle.Indcpendendy ofthethicknessofthisdifferenti- surface temperature on Venus, temperatures atwhich crustal rocks 'atedlayer,only abouta 300-kin layercontributesto thecrust failby ductile flow should be reached atmuch shallowet" depths than formation(adepthbelow whichthemelt issupposedtobe denser on Earth [9,10]. Numerical models suggest that areas of high relief thanthesurroundingrocks).A smallbuoyancy ofthisdepletedlayer and steep slope in the Ishtar region should spread under self-gravity issufficienttostabiliztehislayerwithrespecttotheundifferentiated byductile flow of the weak lower crust on time scales less than about mantle.Thispreventssupplyof theundifferentiatemdaterialtothe 10m.y. [11]. Thus the prtw.esses that build relief and steepen slopes meltingregion.Convectionoccurseventuallyinbothlayers.The must have been active within the last 10m.y. Further, anumber of lower,undifferentiateldayerisheated from within.The upper° features produced by geological processes that have operated more differentiateldayerismainlyheatedfrom below.The temperature or less steadily during the past 500 m.y. show evidence of subse- increasesand reachestheanhydrous solidusonly afterseveral quent tectonic activity. About one-third of all preserved impact billionyearsdepending on the rhcologicalmodel. The young craters on Venus have thoroughgoing faults and fractures, and 1in basalticcrustobserved on Venus isproduced by mehing ofthe 12 are extensively deformed [81. The longest lava channel on the anhydrouslayer.Thismeltingandthecrustgrowthareweak mainly plains of Venus does not progress monotonically tolower elevations because ofthelowerheatfluxconsistingoftheradiogenichcat downstream, indicating that differential vertical motions have production in the undifferentiated lower layer and almost not occurred since the channel was formed [12]. contributed from the secular-cooling heat flux from the core (it is Compared with the Earth, horizontal displacements on Venus even negative) and the remaining radioactivity inthe differentiated over the last 500 m.y. have been limited. Most of the tectonic layer. An additional consequence of the model isthat the magnetic features require modest strains and horizontal displacements of no field was never generated in venusian history. more than afew tens to perhaps afew hundreds of kilometers. Plains References: [1] Zuber M. T. and Parmemier E. M. (1987) thousands of kilometers across record horizontal strains of order Proc. LPSC 17th, inJGR, 92, E541. [2] Zuber M. T. and Parmentier 10-_or less. The great rift systems of Beta and Aria Regiones need E. M. (1990) Icarus, 85, 290. [31 Crrimm R. E. and Solomon S. C. have extended no more than a few tens of kilometers, on the basis (1988) JGR, 93, 11911. [4] Zharkov V. N. and Solomatov V. S. of topographic profiles, extended feamares such as Somerville (1992) In Venus Geology, Geochemistry, and Geophysics, 281, Crater in Devana C"hasma, and analogy with continental rifts on Univ. of Arizona, Tucson. [5] Solomon S. C. and Head J. W. (1991) Earth [13]. For compressional features, the amount of crustal Science, 252, 252. [6] Williams D. R. and Pan V. (1990) GRL, 17, thickening can be estimated from topographic relief and isostatic 1397. [71 Kaula W. M. (1990)Science, 247, 1191. [8]Solomatov V. considerations, but this approach provides only a lower bound on S. (1992) In Flow and Dynamic Modefing of the Earth and Planets, horizontal displacements ifany crustal material is recycled into the == ''N93-14382 : mantle atzones of underthrusting. For ridge belts 100 km in width Univ. ofAlaska, in with up to 1km of relief, horizontal displacements of no more than THE TECTONICS OF V_NU-S: AIN:OVERVII_W. Scan C. 100 km are required for crustal thicknesses of 10--20 km beneath the Solomon, Department ofEarth,Atmospheric, and PlanetarySci- adjacent plains [14,15]. Mountain belts are exceptional in that ences,Massachusetts Instituteof Technology. Cambridge MA greater horizontal displacements are required. For a two- to four- 02139, USA. fold thickening of the crust beneath the 500-kin width of Maxwell Momes, the implied minimum horizontal displacement is 1000- lntrod uctlon: While the Pioneer Venus altimeter, Earth-based 2000 km. radar observatories, and the Venera 15-16 orbital imaging radars Unlike the Earth, Venus does not show evidence for a global provided views of large-scale tectonic features on Venus at ever- system of nearly rigid plates with horizontal dimensions of 103- increasing resolution [I-3], the radar images from Magellan consti- l0 tkm separated by narrow plate boundary zones a few kilometers tute an improvement in resolution of at least an order of magnitude to tens of kilometers across. Predictions prior to Magellan that over the best previously available [4]. Asummary of early Magellan Aphrodite Terra would show features analogous to terrestrial spread- observations of tectonic features on Venus has been published [5], ing centers and oceanic fracture zones [16] now seem to be incor- but data available at that time were restricted to the first month of rect. Evidence for shear is present in the ridge and fracture belts and mapping and represented only about 15% of the surface of the in the mountain belts, but the shearing tends to be broadly distrib- planet. Magellan images and altimetry are now available for more uted and to accompany horizontal stretching or shortening. Few than 95% of the Venus surface. Thus amore global perspective may clear examples have yet been documented of long, large-offset be taken on the styles and distribution of lithospheric deformation strike-slip faults such as those typical of oceanic and many conti- on Venus and their implications for the tectonic history of the planet nental areas on Earth; two such features have been identified in the [61. interior of Artemis Corona [17]. Anumber of the chasm systems of Generalizations: Tectonic features on Venus are widespread Venus have arcuate planforms, asymmetric topographic profiles, and diverse, and comparatively few regions are undefonned. Areas and high relief [18] and have been likened to deep-sea trenches on lacking tectonic features at Magellan resolution are restricted to Earth [17]. These include Dali and Diana Chasmata [17] and the relatively young volcanic deposits and the younger impact craters moat structure of Artemis Corona; such trenches may be the and associated ejecta. Most of the surface, during at least some products of limited underthrusting or subduction of lithosphere period over approximately the last 500 m.y. [7,8] has experienced surrounding large coronae [19]. Elsewhere, however, chasm sys- horizontal strain sufficient to fault or fold near-surface material. tems of somewhat lesser relief display more linear segments and