_ //*/ PHOTOGEOLOGIC ANALYSIS OF THE MAGELLAN STEREO VIEW OF SIX CORONAE A.T.Basilevsky t'2and J.W.Head=. tVernadsky Inst., Russian Academy of Sciences, Moscow 117975, Russia, [email protected]; 2Dept. Geol. Sci., Brown Univ., RI 02912 USA. INTRODUCTION. The present study is one of ef- concentric fractures in the west. it embays asmall massif forts of many researches to progress in the stratigraphic of tessera terrain; Pwr plains deformed by concentric context of local, regional, and global geology of coro- wrinkle ridges and concentric and radial fractures in the nac (see forexample, [1--4,6-9]). This study is based on north; zigzaging wrinkle ridges in the east andconcen- photogeologic analysis of several coronae selected be- tric wrinkle ridges and inward-looking scarp in the cause they were imaged by Magellan in stereo, that is south. The central part of Belet-ili corona is dominated very helpful in determining age sequences among the by fields of numerous small shields correlative to the Psh geologic units. We found in this study that we can de- plains. These fields areseparated and emhayed by thePwr scribe the geology of coronae and their surroundings in plains. terms of units we distinguished in our previous studies GAYA (3.5oN, 21.5oE, 350x500 km). Gaya corona [3,4]. is apear shaped E-Welongated structure, sitting in the VERDANDI (5.5oN, 65.2oE, D ffi180 km). Verdandi Pwr plains with subordinate Psh patches and outlined by sits among the plains with wrinkle ridges (Pwr) embay- sets of wrinkle ridges. The southern part of Gaya is out- ing tessera (Tt) in the central part of Ovda regio. It is fined also by the well seen in stereo inward looking ar- well seen in stereo that corona Verdandi is outlined by custe scarp. The the NW part of the Gaya annnins is si- almost circular annulus standing over the surrounding multaneously the sonthern part of the Belet-ili annulus. Pwrplains. The corona floor is made of the same Pwr This connection looks as mutually concordant without plains which is noticeably lower than the surrounding any evideace of superposition. _ SW part of the Gaya plains.Three _ of the corona annulus are made of annulus is also encircled by aset of fine fracturesconcen- densely fractured terrain (COd0 with almost perfect con- Iric to general corona sUucture. _ cenlral part of co- cenlric trend.Despite very close distance (5-20 km) be- rona Gaya are made of the Pwr plains with patches of tween the corona annulus and tessera terrain no influence densely (radially) fractured terrain (O0df) and fields of of the tessera structural pattern is seea in any corona small shields (Psh)bothembayed by Pwr. element. O3df annulus is emhayed from inside and out- ARAMAITI (26.3oS, 82.0oE, D= 350 km) and neig- side by the Pwrplains. The wrinkle ridges that deform boring corona Ohogetsu (see below) areamong the pre- the Pwr plains trend mostly N-S and NE showing no dominantly Pwr plains of Aino Planitia. In the vicinity alignment with the corona structure. In the SE part of the and inside the corona there aretwo varieties of regional annulus there is a topographic gap fiUed by Pwr plains Pwrplains: the older having intermediate radarbright- criss-cmssed by the dense NE-trcoding fractures which ness (Pwr-i) and the younger which is noticeably dsrke_ areone of the branches of the lx-ClwJ chasma rift zone. (Pwr-d). The network of wrinkle ridges,is mpesposed on Eastof thecorona several arcnate fractures concentric to both of them. Immediately east of Aramaiti there is abelt corona and cutting the Pwrplains arcseen. They arecut of fractured and ridged plains (Pfr) embayed by both va- by another set of the rift-associated fractures. rieties of the Pwr plains. It is well seea in stereo that at THOURUS (6.5oS, 12.9oE, D ffi290 kin). Corona the north Aramaiti has two annulse divided by the trough Thoums sits among plains with wrinkle ridges (Pwr). The and only one annulus at the south. The inner annulus of corona annulus,which completely encircles the corona, the corona north and the annulus of its south form to- is mostly made of densely fractured terrain (COdO with gether a spiral-like feature made of densely (con- almost perfect concentric strucUn'al pattern. It is well centricaUy) fractured temdn (COdOwhich can be subdi- seen in stereo that the annulus stands ovm"the sunound- vided into two mlxmtts: l)COdf-a, which occupies the ing plains and the corona interior is topographically flngrest of the first thirdof the mentioned COdfspiral, lower than the surrounding plains.The COdfring of the and 2) CX3df-bwhich occupies the rest. COdf-a is more annulus isembayed by the materiais of three sorts: 1) the deformed, embayod by the Pwr-i plains and probably material correlative to rite fractured and ridged plains correlative to _ unit of [3,4]. COdf-b has avisi- (Pfr)of [3,4], it participates in composing the annulus; ble plain-like backSmnnd with fine concemric linea- 2) the material co_lative to the shield plains (Psh) [4]; ments and is the deformed Pwr-t plains. The SW sector of and the material of plai_ with wrinkle ridges (Pwr) tbe inner annulus looks as mmutte plate of COdf4 up- which cmbays the corona annulus from outside and forms thrusted into SB direction (probably when the mentioned the dominant part of the corona interior. Pwrplains are ridgebelt was formed). Outerannulus of the noalte_ part deformed by wrinkle ridges with a predominantly B.W of Aramaiti is made of the Pwr-i plains cut by roughly trend which slightly deflects to radial to the corona east concentric set of partly sinuous faults. _ trough be- and west of it. In the northern part of the corona interior tweeo the outer and inner annulae is filled with Pwr-d there is 10xS0 km areaof relatively bright plains (not which embays both the faulted material of Pwr-i andthe deformed by wrinkle ridges) correlative to the lobate heavily deformed material of COdf-& The corona core is a plains (PI) unit of [3,4] dome surrounded by the circular trough filled with Pwr-i BELET-ILi(6.0oN, 20.0oE, D = 300 kin). This co- and Pwr-d, as well as with younger radardarkflows (Pl-d). rona and neighboring corona Gaya (see below) are domical core consists of Pwr-i, Psh and COdf-a ma- among the Pwrplains of the sonthem part of Berighinya terieh. In the arenate depression within the eastern seg- Planitia. Belet -ill has an anulus consisting of different meat of the annulns andoutside the atmulns areobserved materials and t_acUJa'alfeatures which form tngether the young dark plains of the type which was called by [5] as well visible quasi-circular structure: Pfrplains crossed by amochoids (Pda)which areprobably correlative to PI/Ps 13 MAGELLAN STEREO VIEW OF SIX CORONAE: A.T.Basilevsky and J.W.Head units [3,4]. logic events. Only Verdandi is an example of corona OHOGETSU (27.0oS, 85.7oE, D = 175 km). It is the outlined by structures of one age deforming only one corona neighboring Aramaiti so regional geologic situa- material unit. We found no evidence of coronae predating tion for Ohogetsu is mainly same as for Aramaiti. The tessera and deformed by tessera terrain structure. The first most important difference is that corona Aramaiti sits evidence of acorona structure is aformation of an annu- aside of the belt with broad compressional ridges on it, lus due to warping of early regional and giobai units (Pdf; while Ohngetsu is in between two segments of the belt. Pfr) of apparent volcanic origin. Following this initial Corona Ohogetsu, like corona Aramaiti, has an annulus stage, most corona armulae were flooded to different ex- consisting of two varieties of the densely fractured tents by regional volcanic plains (Pwr with patches of retain: 1) COdf-a, 2) _-b. In SW sector of Ohogetsu Psh). Then majority of the studied comnae experienced there is a peculiar arcuate feature. On stereo it looks as a concentric and/or radial fracturing and with a half of them horse-shoe plate upthrusted upon the middle part of the was associated the emplacement of flows otside and/or SW section of the annulus. From consideration of local inside the corona. These later volcanics are unmodified geology it was concluded that the plate material is by wrinkle ridges that relates them to rather recent period probably the Pfr plains whose upthrust was clue to the of geologic history of Venus. These observations are in event of regional compression which formed the belt agreement with our earlier results [2-4] and with results with broad ridges. Inner part of corona Ohogetsu is occu- of other workers[I,6-9] so the general conclusion is that pied by dark Pwr-d plains deformed by N-S trending set of coronae are characterized by both local and regional wrinkle ridges extending southward into the area of Pwr-i stru_zres and stratigraphic units and that their geologic plains. Among the Pwr-d plains of the corona interior historyrepresentstheinterplay of localcoronae events there is a shield of about 10km in diameter (Psh?). West- supe_x_ on regional and global geologic evolution. ern part of corona Ohogetsu anmflus is flooded by the Totallifetimeofcoronae insome cases(e.g.Vernandi) dark plains of the amoeboid type (Pda) which embays all may span sinceearlypost-tesseraperiodtillpre-Pwr or the materials it is in contact including the Pwr-d plains even pre-Pfrtime,that isprobably lessthan lO0 my, and the wrinkle ridges themselves. To SE and SW of the while in other cases (e.g. Ohngetsu) it is significantly corona annulus there are several localities of the lobate longer, from early post-tessera time till post-Pwr, Ps/PI plains correlative to the Pl unit [3,4]. time, that may be as long as 200 to 400 my. DISCUSSION AND CONCLUSION: The described observations on the age relations of different compo- REFERENCES: 1)BaerG. et aL JGR, 99, FA, 8355- nents of the studied coronae can be presented in aform of 8369° 1994; 2)Basilevsky A.T. Annales Geophysicae, the following table: Suppl. to vol. 12, C652, 1994; 3) Basilevsky AT. & Head J.W. PSS, 43, N 12, 1523-1553, 1995; 4) Basilevsky et ai. In: Venus H, The Univ. Ariz. Press, in Corona "It COdf/Pdf Pfr/RB Pwr H/Ps press, 1997; 5) Head J.W. et ai., JGR, 97, Eg, 13153- V' Gm-O...'.i." ......".7.."......."...".........-........... 13197, 1992; 6) Jackson E. et ai., LPSC-26, 665-666, - IIIIIIIIIIIIIIIIII/11111/111II1I1IIIIIIIII/1I1IIIIIIIIIIIIII 1995; 7) Magce Roberts K. &HesdJ. GRL, 20, 1111- Belet-ilnl ? II/IIII/IIII/I/I/I/IIIIIIIInIII 1114, 1993; 8)Magee K. & Head J. JGR, 100, 1529- Gaya I/////II///I/// ? 1/1/1/1/m//// n Ammmfi - lllllllllll//ll ? IIIIIII/11111II1I1IIIIIIIIIIII 1252, 1995; 9)Pmnin A.A. Astron. Vestnik, 31, NI, Ohoge_ - IIIIIII/11111I1II1II/I/111111I1IIIIIIIIIIIIIIIIIIIIIIIIIIIII 1997. Five of considered coronae consist of several material STEREO VIEW OF SIX CORONAP_ A.T.Basil_sky and J.W. Head. units and structures representing a time sequence of geo- 14 THE ANCIENT AGE OF MAXWELL MONTES, VENUS: PRF_ERVATION OF HIGH TOPOGRAPHY UNDER HIGH-SURFACE-TEMPERATURE CONDITIONS. J.D. Bun, J.W. Head, and E.M. Parmentier, Brown University (Dept. of Geological Sciences, Brown University, Providence, RI 02912 jburt_cos.hitc.com). ABSTRACT: Maxwell Montes has the highest terrain increase and the rate of eclogite formation. and surface slopes of Venus, but displays little Smrekar and Solomon [1] have investigated the gravitational relaxation except on its nortbem and evidence for gravitational relaxation in Maxwell. While southern flanks [1]. Evidence thatMaxwell is as old as the extensional features can be found on the northern and surrounding plains [2] leads to the challenge of explaining southern flanks of the belt, the western slopes, where the the preservation of high topography underconditions of steepest slopes occur, bear no recognizable evidence of _'face temperature that would quickly cause the collapse relaxation. They model viscous relaxation of topography, of terrestrial mountain belts [3]. The drydiabase flow laws invoking the flow law of Caristan [10] for Maryland of Mackwell et ai. [4] may provide the key to diabase. They predict that gravitational spreading is understanding this problem. Model results indicate that likely to occur on a time scale shorter than the mean gravitational relaxation should not be important in surface age indicated by the crater statistics. They conclude changing the topography of Maxwell, if the mantle and that dynamic tectonic support may be necessary to crustof Venus arevery dry. preserve the high topography and slopes of the mountain belts. BACKGROUND: Vorder Bruegge and Head [5] examine Freed and Melosh [11] modelled the gravitational the implications of Airy support for Maxwell under collapse of lshtarTerra. Building on the work of Smrekar conditions of low thermal gradients created by and Solomon [I], they find that the dry flow law of compressional deformation. At low strain rates thermal Mackwell [4] permits the crust to be sufficiently viscous equilibration would permit volcanism to occur at the same to support Ishtar's topography for periods on the orderof timeasmountainbeltgrowth. Low viscosityofa deep 500 million years. a_stalrootwouldalsopromotegravitationaclollapseA.t We have formulated two simple models to explore the higher strain rates the thermal gradient would be role of the dry mantle viscosities in the support of depressed, preserving the strength of crustal root and Maxwell. Assuming Maxwelltopographyis supportedby delaying igneous activity. Therefore, they favor rapid thickness variations in a basaltic crust v_ calculate the construction of Maxwell because the mountain belt lacks stress distribution for periodic topography having a evidence of simultaneous deformation and volcanism. wavelength similar to that of Maxwell A steady-state Additionally, at low temperatures the gabbro-eclogite thermal gradient should apply if Maxwell is old. This phase change occurs slowly, but the transformation of allows for estimation of the crustal viscosities beneath the deep crustal material would eventually reduce topography. belt and estimation of the rate of collapse of the Namiki and Solomon [6] investigate further the role topography. of the gabbro-eclogite phase change in determining the mountain belt topography. Tne possible low water THEORETICAL MODEL: A simple assessment of the almndatwe on Venus may mean the gabbro-eelogite phase age of Maxwell canbe obtained using amodel for viscous change occm_ there by solid state (volume) diffusion flow in response to a surface load [12]. Assuming a ratherthan grain boundary diffusion, inhibiting the rate of periodic surface load of theform the transition and preserving metastably basaltic crustal 2gx material within the stability zone of eclogite in mountain c0= C0opgcos"_ room. They concinde that rapid mountain belt formation, resulting in low thermal gradients, favors metastable where mo is the amplitude of the surface load or gabtwo in the cnumd root and preserves high topography, topography, and _.is the wavelength, the characteristic implying that Maxwell must be young relative to the relaxation time is given by mumunding plains. "I'nelower elevations of volcanically active Denu may illuatrate the effects of the temperature- .gffi4ttg enhanced phase change limiting elevations. Also, a young pgZ Maxwell would be consistent with the endefotmed Ifthe wavelength of Maxwell is _ to be Z = 1000 appearance of Ckopatra erater. kin, p : 3 gm/cm2and Itto be 102t Pa s, the relaxation In contrast, Basilevsky [2] finds stratigraphic time is !.2 x 104years. For Maxwell to have arelaxation evidence that Maxwell is at least m old as neighboring time on theorder of 109 years the viscosity would be 8 x ridged plains. He interprets this as evidence that Maxwell 1025Pas. is as old asthe estimated surface age of O.3to 0.5 Cut[7, 8] Msckweil et eL [4] provide flow laws appropriate to and thatlow water abundance allows high topography and materials subject to the dry conditions of Venus. Their steep slopes to last for geologically long periods. flow law forMaryland Diabase is Meanwhile, Herzog et aL[9] point out that the high surface atmospheric pressure would prevent exsolution of I-I20for effi4.2os'texp -(50_5) , abundances below about one weight percent. The dry atmosphere could thus be consistent with hydrous where e isthe strain rate, o the differential stress in MPa, minerals existing in the venusian mantle. This represents R the gas cousUmt, and T the _ This can be a volatile mttrce capable of expediting the gabbro- rewritten, for this model, and using the second invarlant of eclogite phase change. If so, high topography may be the stresstensor, as dynamic, limited by the balance between rate of elevation 21 THE ANCIENT AGE OF MAXWELL MOI'_I'ES: J.D. Burr,J.W.Head, and E.M. Parmentier v,- (-,) E 2, I ' e=4.2 pg COo y _" exp- T '_ At t = 0 this simplifies to 2n 2_x 2gy (505 v=--- g mo y _-" cos"_- exp- "-_ exp-_. _ The model calculates the amplitude of the w From this relation I_can bederived component of the surface velocity to be 1.2 x 10-I; Using viscosities for a thermal gradient governed by This produces a characteristic relaxation time of basaltic crustal radioactivity, a surface temperature of 750 10tl years. K,and auniform mantle temperature of 1321 Kbelow 200 CONCLUSIONS: While these models are km depth, the relaxation rate for Maxwell can be complicated, they do provide an estimation of the 1 evaluated. A mid-crustal viscosity canprovide an estimate collapse of Maxwell Montes. These results indicat( of the relaxation time. Assuming Maxwell to be in gravitational relaxation should not be importaJ isostatic equilibrium a maximum crustal thickness is about changing the topography of Maxwell, if the mantl 120 km. The viscosity for the 60 km depth is tt = 1025"6pa crustof Venus arevery dry. If significant water do_ exist within the interior of Venus, then the mantl s, _ielding acharacteristic relaxation time of about 5 x 10_years. This is close to the value that is required to crustarestrong and the rate of the gabbro-oclogite maintain the topography of Maxwell for 109 years. change is very low. Both of these factors would con! If gravitational collapse may happen slowly enough to the preservation of high topography despite the to permit Maxwell to retain its high elevations, the temperatures of the vennsian mrface. If hydrous mi) transformation of gabbro to eclogite may become more exist [9] within the mantle, however, dy_ important. A characteristic reaction time for the mechanisms may be required to explain the ex/sten conversion from one phase to the other can take the form old high topography. 82 REFERENCES x=_" [I] Smrehar,S.E. and S.C. Solomon, (1992) JGI; where xis the characteristic reaction time, 8 is the grain 16121; [2] Basilevsky, A.T., (1995) LPSC 26: size, and Dis the diffusion coefficient. Taking DAI, Opx Weertman, 1979; [4] Mackwell, S.J. et al., (1994) as given in Namiki and Solomon [6] as alower bound, 25: 817; [5] Vorder Bruegge, R.W., and J.W. Hea_ (1991) Geology 19: 885; [6] Namiki, N., and DAI, Opx = 1.1x 10"5exp C 40R0T kJ_ Solomon, (1993) JGR 98, E8: 15025; [7] Schaber, G. ai., 0992) JGR 97, Eg: 13257; [8] Strum, R.G. • an estimate can be made of the reaction time for a given temperatmx-.. At the base of the crust the temperature is (1994) JGR, 99, E5: 10899; [9] Herzog, S.G. e, about 925 K. At this temperature, the reaction time is on (1995) LPSC 26: 591; [10] Caristan, Y., (1982) JG_ the ordm" of 1014 years. Clearly this indicates that the 6781 ; [I1] Freed, AM., and HJ. Melosh, (1995) ] phase change happens too slowly to significantly affect 26: 421; [12] Turcotte, D.L. and G. Schubert, (1 the isostasy over periods as short as abillion years. Geodynamics Applications of Continuum Physic_ This simple modeling is based on a formulation Geological Problems. John Wiley & Sons, New • assuming uniform, not variable, viscosity. To more 1982; acctwately determine rates of relaxation for a model in which the viscosity varies with temperature, numerical models are necessary. NUMERICAL MODEL: In a model region measuring 1000kmwide and 1000 km deep, and having an imposed surface load of the form 2xx mop gcos'-_,-- , we solve for fluid flow using aLagrsngian finite element method. A free surface and free-slip vertical and basal boundaries comprise the boundary conditions. The imposed depth-dependent viscosity is derived from the flow lawof Mackwell et al. [4]. Flow calculations result in estimates of the surface vertical velocity component. To evaluate the rateof collapse we again appeal to the simple theory. Since the rate of change of the vertical displacement (the vertical velocity) is given by 8. exp-( ) 8t - 4=tt mo _4=_J ' then 22 Tim_cale of Regional Plains Emplacement on Venus Geoffrey C. Collins l, James W. Head 1, Mikhall A. Ivanov 2, and Alexander T. Basilevsky 2. 1Brown University, Dept. of Geological Sciences, Providence, R102912, 2Vemadsky Institute, Russian Academy of Sciences, Moscow, Russia; [email protected]. An outstanding question in the Venus depth contour was obtained by assuming that res,urfacingdebate isthelengthof time over the topography of the older material underlying which a theoretical"catastrophic"resurfacing the plains sloped away from every surface eventmay have occuned. The emplacement of contact with the wrinkle ridged plains at a 1% thewrinkle ridged plainsover -70% of the grade (0.57 ° angle). This will probably give planetoccurredsynchronouslyoverlargeareas an overestimate of the depth of the wrinkle [1,2]and was an importantpartof theglobal ridged plains, since only 14% of the Venusian event. The length of time over which the surface has regional slopes above 0.24 ° [5], wrinkle ridged plains were emplaced can be and the only features on Venus with slopes coustrained by the number of creaters which consistently above this 0.57 ° value are the theyembay. Most of theembayed craterson mountains around Ishtar Terra [6]. Using this Venus are embayed by lobate plains from local slope assumption, 63% of plains with wrinkle volcanic sources, and only 5-8 craters are ridges in this mapping area are thinner than embayed by the vast regions of wrinkle ridged 500 meters, and 37% are thicker than 500 plains [3]. The model of Strom et al. [4] meters. The depth of 500 meters was used as estimates the length of tlie re.surfacing episode a cutoff value because the rims of median-size based on how many craters it embayed, but (30 km diameter) or larger craters are about this model was based on resurfacing by small, 500 meters high, as deduced from topographic randomly distributed volcanic flows. This profiles of craters generated by Sharpton [7]. does not agree with observations of the nature Plains thinner than 500 meters should be thin of the wrinkle ridged plains, which the plains enough for the rims of median-size craters or appear to be simultaneously emplaeed over larger to show through, so at least half of the extensive areas [2]. Here we develop a simple preexisting population of craters below the stati_'cal model based on the plains flooding plains will show through as embayed craters in of extensive areas in order to constrain the these shallow areas. We will assume that length of time over which they were emplaeecL plains thicker than 500 meters will have buried One end-member model would be to all underlying craters. All of these _,ssume that these 5-8 craters embayed by assumptions: the steep underlying slopes, the wrinkle ridged plains represent the entire burial of all craters in thick plains, and the populationof craterson thesurfacebelow the burial of all small cratersin thinplains, will plains. This implies that all of the wrinkle tend to overestimatethe number of craters ridged plains are thin enough that they did not destroyedby the emplacement of wrinkle completely bury the smallest craters. On the riaged plains, thus making the resulang other extreme, the plains could have buried all timespan a maximmn estimate. The eight preexisting craters, and the embayed craters craters possibly embayed by wrinkle ridged we see were formed on the plains during their plains occur in areas mapped in this scheme as emplacement. The relative roles of these two thin plains. Five of the eight craters are larger end-member models can be examined by than median di_aneter. estimating the depth of wrinkle ridge plains If 70% of thesurfaceof Venus iscovered deposits by a means independent of craters. by wrinkle ridged plains and 63% of these We have mapped an area extending from 23°- plains are thinner than 500 meters, 44% of the 35° N latitude, covering over 8% of the planet. surface is covered by wrinkle ridged plains This area is composed of 37.7% material older thinner than 500 meters. The probability is than plains with wrinkle ridges, 10.3% .44 that a crater emplaced randomly on the material younger than the plains, and the other surface falls within this area. Since only 52% is plains with wrinkle ridges. The plains craters larger than median size are guaranteed with wrinkle ridges were separated into areas to be unburied within this area, the probability thinner and thicker than 500 meters. This is .22 that a crater emplaced on the surface 29 Timescale of Venus Plains Emplacement: Collins et al. during the time.span in question will be occurred prior to the emplacement of these embayed but not buried by wrinkle ridged plains to erase the ancient population. plains. Since we observe five craters of Resurfacing such as tectonic resurfacing, median diameter or larger embayed by thin which may have erased the ancient crater areas of wrinkle ridged plains, we can population from the tessera [8] or volcanic calculate the probability that, given a number resurfacing by stratigraphically lower plains of craters emplaced on the surface, exactly five units, must be the primary mechanisms will be larger than median diameter, in the area responsible for the young surface age observed on the surface of Venus. The of thin plains. This gives an expected value of 22 craters, with a 98% confidance interval of wrinkle ridged plains which cover the majority 10-54 craters. In terms of a percentage of the of thesurfaceof Venus areonly a relatively mean age of the surface of Venus (300-500 thinveneerwhich formed quicklyand did little Ma), the expected value is 2.4% (7-12 Ma), torejuvenatethesurface. with a lower limit of 1% (3-5 Ma) and an upper limit of 5.8% (17-29 Ma). This References: [1] Basilevsky, A. T., and J. timespan represents the age of the surface W. Head (1995), Planet. Space ScL 43, 1523- upon which the wrinkle ridged plains were 1553; [2] Basilevsky, A. T., and J. W. Head emplaced plus the length of the emplacement (1996), Geophys. Res. Lett. 23, 1497- of wrinkle ridged plains. 1500;.[3] Collins, G. C., et al. (1996), LPSC Most of the wrinkle ridged plains are thin XXVII; [4] Strom, R. G., et al. (1995), J'. enough that a large number of underlying Geophys. Res., 100, 23,361-23,365; [5] craters would not have been completely Sharpton, V. L., and J. W. Head (1985), ./. buried, but only embayed by their Geophys. Res. 90, 3733-3740; [6] Sharpton, emplacement. Since so few craters are V. L., and J. W. Head (1986), J. Geophys. observed to be embayed by these thin plains, Res. 91, 7545-7554; [7] Sharpton, V. L. they must have formed over a short time on a (1994), in Dressier, B. O., et al., eds., Large young surface, covering a 5-30 Ma timespan. Meteorite Impacts and Planetary Evolution, The emplacement of these plains was not the GSA Special Paper 293; [8] Solomon, S. C. prime mechanism for removing the ancient (1993), LPSC XXIV 1331-1332. population of craters. An event must have 3O REMOTE AND LOCAL STRESSES AND CALDERAS ON MARS; L. S. Crumpler 1, J, C, Aubele 1, and J. W. Head2; (l)New Mexico Museum of Natural History and Science, 1801 Mountain Rd NW, Albuquerque, NM 87104; (2)Dept. of GeoL Sciences, Brown University, Providence, RI 02912; crumple r@nmmnh. -abq.mus.nm.us INTRODUCTION. The detailed structure of calderas is a DISCUSSION. Calderas are limited in absolute number, sensitive indicator of the stress environment existing at so the results of acomparison between observed and pre- the time of caldera formation. Unlike regional patterns of dicted patterns arcof limited quantitative value. However, strain, such as wrinkle ridges, graben, and fractures. logicalgeologicinferenceosf regionalinterestcan be calderas, however, have short time scales for formation madebasedon observedstructuraendlocal/regionaalsso- and record local stress at essentially point sources. Strain ciationsT.he salienptointsarcreviewedinTableIwhich associated with caldera formation and evolution may also summarizestheprincipailnferences. be tied strafigraphically to a relatively well-defined geo- logic unit and well-constrained times. Inthe following, we Table 1.Three caldera strain/remote stress associations: have compiled the structural characteristics of caideras on Volcanoes Timing Origin of Stress Mars [1] and compared the deduced orientation of remote [AM, PM, AM, mid-age vol- regional dynamic stress with predicted patterns of global strain, patterns at- Uranins Patera, canoes lithospheric flex- tributable to regional slopes, and patterns attributable to Hecat_ rElysium] ural stress reties or pre-existing substrate structure. OM, 'l]uu_ mid end young locai crnstal grav- OBSERVATIONS. Several types of t_mcturealignment Tholus, Ceraunins age ity stress (region- can be distinguished in the summit caldems of the larger Tholus al slope andedi- martian volcanoes: (1) overlapping calderas, (2) concen- riceeffect) trations of pits and channeh on flank sectors, and (3) lin- Tyrrbena, Nili, old to mid age regional inheri- ear, through-trending fiuure patterns. Overlapping and Meroe, Hadrlaca, tance from pre-ex- elongated calderas characterize the summits of Olympus Pcoens, Amphi- isting structure Mona and the Thamis Montes (Figure 1). These patterns _tes] aremost prominent in larger edifices or the larger calderas that areindicative of large magma reservoirs. Caideras as- [Biblis, Alber, old No (or little) ap- sociated with smaller volcanoes, such as Biblis Patera, !Jovus, Apolio- parent local orre- Ulysses Pater& Ceraunins Tholns, and Albor Tnolus am ei- naris,_bu(7)] gional ther circular or consist of randomly overlapping caldera segments. Patterns of strain associated with the larger calderas ANALYSIS. Caldera structures pteservo regional stress follow stresses predicted from global isostatic flexure pat- patterns because dikes develop along directions of maxi- terns [6,7,8] particularly within the large volcanic rises of mmn principal stress. Systematic alignments therefore Tharsis. Notable exceptions in Tnanis include Olympus tend to develop in successive magma reservoirs within an Mons, Tnands Tholns, Alba Pate_ and Cemunins Tholus. evolving magmatic system [2]. Theprinciplesof dikeem- Fracam= obeying the predicted pattern of strain occur placementwere reviewed previonsly [3] and areherebriefly around Allm Pater& but do not appear to have operated at re-iterated. A dike is propagated from a magma body, re- the time of magma emp_t. At Olympus Mons, large suiting in magma being either erupted at the surface or in- gravity stresses assodated with _ionai slip of the edifice on the slopes of 'l]utnis may have dominated the local ,:ted laterally, when the wall failure criteriaPl + Pm> Is31 stress field [9]. 'l_is also appears tobe the case for Tharsis +Tis mJsfied[4] and magmafic pressure¢xceedztbesum Tholus. of remote stress (minimum compressive stress, a3) and Parterre of defommflon in older caklems may relate tensile strength (T) of the country rock. Regional patterns to _-_tional andlocalslope effects. "I'neoldest, and lowest, of stress that may influence the orientation of s3 can arise cakle_ appear to have been influeneed by pre-existing muctm_ falx/cs a.odated with basins. from strains resulting from either tectonicprocesses, local REFERENCES C/TED. [11 Cmmplor, e_id, 1995, Lunar and regional relief, or pre-existing, directional variations P/aries _'L, XXVI, 305-306; Crnmpler et aL, 1994, Lunar in the value of T(i.e., pre-existing stmctm_ fabrics)[S]. PinneL _ XXV, 305-306; Crumpler, et al, 1995, Geol. All of these arepredleted to have asign/flcant infinen_ on _7c. London 5_ec. Pub. 110. 307-347; Crnmpler, et aL, regional stress arrangvments on Mars [6,7,8]. The influ- 1990, MEVIV Workshop on the Evolution of Magma ence of regional topography and its associated stress pat- Bod/_ on Mart, Richland, 14-15; [2] Pollard, and Muller, terns on dike orientation has been treedprevionsly in as- 1976. Your. Geophys. R_. 81,975-984; Nakamura,1982, king the orientations of possible dike-related graben Bull. Vok.So¢. Japan, 25, 255-267; [3] Crumpler, Head, on Venus [9]. On Mm, buinsandbasinm_mms arc and Aubele, 1996, LPSCXXVil, 277-278; [4] alsolikelytobe important. Gudmundu_n, 1988. J. Vole. Geoth.Res, 35, 179-194; [5] Figure I is acomparimn of lm_di_d patterns of Nakamura, 1977, J. Vok. Oeotherm. Res., 2, 1-16; [6] ,minimum stress from global topographic relief [8] with orientations of the minimum stress determined from the Banerdt et ai., 1982, Your. Geophys. Re$., 87, 9723-97- 33; [7] Phillips and IAtmbeck, 1980, Rev. Oeophys. Sp. relevant s3 indicators (overlapping calderas, con- Phys. 18. 27-76; [8] Sleep and Phillips, 1985, Your. centrations of pits and channels on flanksectors, and lin- Geophys. Res.90,4469-4489; [9] McGovern and ear, throegh-trending fissure patterns). Solomon, 1993, Your. Geophys. Rex. 98, 23553-23579. 39 CALDERA STRAIN/REMOTE STRESS: Crumpler and others Figure 1.Top: Minimum compressive stress orientations determined from mapped structure of large calderas on Mars [1 Center:. Stress orientations superimposed on global stress field predicted from topography [8]; bottom: Enlargement ( main volcanic regions. Observed orientations are consistent with remote stresses predicted from isostatic flexure in man cases, but some ca]dents appear to have been strongly influenced by su_sses within the crust dueto slope and azimuth: variations in crustal strength. 4O LUNAR LINEAR RILLES, MODELS OF DIKE EMPLACEMENT AND ASSOCIATED MAGNETIZATION FEATURES. J.W. HeadI. L. Wilson1,2 K.A. Anderson and R. P. Lin3; IDepartment of Geological Sciences, Brown University, Providence, RI 02912 USA; 2Environmental Science Division, Institute of Environmental and Biological Sci- ences, Lancaster University, Lancaster LAI 4YQ, UK; 3Space Sciences Laboratory, University of California - Berkeley, Berkeley, CA 94720. Introduction and Background: Lunarlinear and to the surface will cause new fractures to form and allow arcuate rilies form from tectonic deformation associated significant movements along parallel faults to occur. As with near-surface stress fields [1, 2] which have been at- the dike tip furtherapproaches the surface, the main effect tributed a variety of origins, including lithospheric flexure will be for the grabea to become progressively deeper as in response to mare basalt loads [3], and to the emplace- more strainis accommodated. Very shallow intrusion may mentof dikes to near-surface environments [4]. We have lead to further fractures developing on the floor of the gra- been assessing the nature of dike intrusion as asour_ of ben and will encourage the formation of small secondary near-surface stress fields sufficient to producelinear rilles intrusions and possible eruptions. [4, 5] and conversely, developing criteria to distinguish These theoretical predictions support the idea that lin- rilles plausibly cause by near surface dike intrusion from ear rilles may be the sites of subsurface dike emplacement. those formed by other mechanisms [6]. We have devel- Specifically, from a morphological point of view, many opod model predictions [5], photogeologic criteria [6], and dikes may intrude to sufficiently shallow depths that they have most recently been investigating the possibility of will create graben, butthere will be no surface evidence of using uw.astm_ments of magnetic fields by the electron eruptions. "l['neseare clearly the most difficult to distin- reflection method to aid in the identification of candidate gnish from grabea formed from stress fields not related to linear rilles formed by dike intrusions [4]. Inthis abstract dikes. Several examples are known in which other evi- we report on progress in the further assessment of data dence suggests that dikes exist below the observed graben. from these sources and on the identification of areas in Rima Sirsalis, a 380 km long graben in the highlands, is which magnetization features appear to be associated with characterized by alinear magneticanomaly interpreted to linear rilies. be due to an underlying magnetized dike [I0, II, 4]. Thus, Theory, Predictions, and Observations: A magnetometer and electron reflection experiments may dike propagating from depth toward the surface generates a provide additional information on the location of buried stress field in the surrounding rocks. As an upward- dikes and the origin of specific graben. A reasonable in- propagating dike nears the free surface, the accompanying terpretation of this relationship is that the impact event stress field is progressively modified; stresses andassoci- excavated mare material from the top of adike underlying ated strains at the surface become concentrated in two re- the graben In some cases, dikes may propagate suffi- gions parallel with the strike of the dike plane andlocated ciently near to the ma'faceto create agraben, but still not on either side of the line of the potential outcrop of the canse significant eraption of invas [12]; in these situa- dike onthe surface. 'I'noseparation on the surface between tions, &gassing of the upperpartof the dike may cause the the zones of maximum elastic stress is essentially inde- formation of gas/magam mixtur_ which might buoyantly pendent of the mean dike thickness and depends only on rise to the surfsce orbeforced to the surface through over- the dike tip depth [7]. However, the rtmenltmt_, of the preuurizafion of the upperpart of tbe dike. In this particu- maximum stress is directly pmpoctional to the ¢hTgethick- isr configuration, the distribution of stresses anticipated ness, as well as being dependent on tbe depth of tbe dike in the vicinity of the dike tip can cause migration of tip. As adike tip rises towards the surface the stresses magma and exmlved gas from the upperpart of the dike to become sufficiently large that failure of the counUy rocks locationseutwatdofthemain grabenboundingfaults, a takes place in shear or tension at points well outside the phenomenon likely to explain the distribution of cones at proce_ zone on both sides of tbe dike. Wbether these ini- Rlma PartyV [4]. Dikes reaching _ to the surface, but tial failures are actually on the surface or at some finite stilnlot having associated extensivelava flows, should depth below it is a function of relationships between the produce narrower graben, and any pyroclastic cones should excess pressure within the dike, the dike width and the be more timely asmeiated with the graben. mechanical properties of the country rocks [8]. The failure Observations and Analysis: On the basis of the likelihood that nmnemns linear rilles could represent the planes soon intersect the surface, bowever, end a linear graben structure begins to develop. If the dike stalls at a surface manifestation of dikes emplaced to the vicinity of sufficiently great depth tbere will be mine undetectable the lunar tau'face,andthe fact that in at least one place on small amount of surface extension and uplift. Shallower the Moon where there is agood correlation betweea alin- penetration will lead to avolume of melt being exposed to earmagnetization anomaly and alinear rifle [I0, II], we the relatively low pressure environment near the mrface have been mmalyzlng datafrom Apollo magnetometer and and will encourage the gene_tiou of agreater ttum of OO electron reflection experiments to assess whether they can since the chemical reaction producing it is pressure- provide additional information on the location of buried dependent [9]. Sufficiently close approach of the dike tip dikes and the origin of specific graben. The method of 61 LUNARLINF_,RARRLF_M._O,DF_J._ OFDIKE I_API..A_: J.W. Head e_sl. measuring lunar magnetic fields by the electron reflection with formation by dikes which propagated from depth to method has been described elsewhere [10, 13, 14]. near the lunar surface. The lower reflection coefficient Rima Sirsalis is the site of a strong local magnetic values than at Sirsalis may be related to the smaller dike field [10, 11] (Figure 1). It is a NEtrending linear rille widths. about 380 km in length located in the highlands south of Preliminary Conclusions: On the basis of this Grimaldi (Figure 1);it averages less than 4 km wide and is analysis, we find that one of the prime candidates for linear 150-250 m deep. The depth tothe dike top is estimated to rilles associated with dike emplacement on the basis of be-1700 mand dike width is estimated to be in the range morphology and geology (e.g. Rima Parry V and associ- 600-700 m [4]. These values compare with predicted dike ated rilles) is also characterized by a magnetic field anom- widths of 600-800 m for dikes propagating from parent aly plausibly attributed to magnetization of the dike. We magma bodies at depths up to 300 km [15]. The magnetic arcpresently examining the areain more detail to correlate datafor Rima Sirsalis can also beused to estimate the aver- individual peaks with local features, and extending the age width of the subsurface dike given suitable assump- analysis to other areas of the lunar surface covered by the tions about its magnetic properties. Srnka et aL [11] used Apollo 15and 16 data. various configurations of single and multiple dikes extend- ing to various vertical depths from the surface and having References: 1] Golombek, M.P., JGR, 84, 4657, 1979; various degrees of magnetization. Using a single dike 2] MeGili, G. E., Icarus, 14, 53, 1971; 3] Solomon, S.C. model exemplified by their Figure 7 and varying the mean and Head, J.W., RGSP, 18, 107, 1980; 4] Head, J. and Wil- dike width and a range of remanent magnetization values it son, L., PSS, 41, 719, 1993; 5] Wilson, L. and Head, J. W., was found [4] that amean dike width of --430 mis consis- LPSC 27, 1445, 1995; 6] Head, J.and Wilson, L., LPSC 27, tent with the measurements if the magnetization. Thus, on 519, 1996; 7] Mastin"L. and Pollard, D., JGR, 93, 13221, the basis of these considerations, we conclude that adike 1988; 8] Melosh, 14.and Williams, D. JGR, 94, 13961, emplacement model for the Sirsalis graben and magnetic 1989; 9] Sato, M., Eos, 58, 425, 1977; 10] Anderson" K. et anomaly is plausible. al., EP$1, $4, 141, 1977; 11] L. Srnka et al., PEPIo20, An additional anomaly is observed over the Fra 281, 1979; 12] J. Mustard and J. Head, LP$C 26, 1023, Mauro/Bonpland region (Figure 1) where the reflection 1995; 13] Howe, H.et al., GRI_ I01, 1974; 14] Anderson, coefficient is distinctly above background but less than a IC et aL, Space Sci. Instr., 1, 439, 1976; 15] Head, J.W. factor of two of that seen atRima Sirsalis. A large concen- and Wilson, L., G&CA, 56, 2155, 1992; 16] Wilhelms, D. tration of linear rilies is seen in this area. Rima pony V, E.. The geologic history of the Moon, U.S.G.S. Prof. Paper about 50 km in length, is a linear riile that is part of a se- 1348, 302 p., 1987. ries of graben-like features that cut the floors andrims of the craters FraMango, Parry, and Boupiand. It begins in _°I' ' II ...... :,,.._ the south on the floor of Bonpland in the vicinity of Rima f II "oo_I " TM.._O_ ParryVI, andextends NHE across thenorthern tim of Bow O.i I -IGUll pland, descending to the floor of FraMauro, and continu- ing across the floor until it merges tangentially with the ':IJ NNW trendingRima ParryI. Midway between the ends of the rille, the rille wails areobscured by deposits associated with aset of volcanic cones parallel to the western rille margin. Rilles in the vicinity of Rima Parry V cut the early Imbrian Fra Manro Formation and areembayed by later lmbrian-aged maria [16]. On the basis of observa- tions and empirical relationships, the average dike width isestimated to be about 150 m [4]. The estimated depth to 410W ]lOW. • ]lo[ the dike top is -750 m. "Inns,on the basis of these con- siderations and the presence of associated pyroclastic de- Figure 1. posits, we conclude that the characteristics of Rima Parry V, and by association the adjacent rilles, areconsistent 62