Corona Associations and Their Implications for Venus Mary G. Chapman U.S. Geological Survey, 2255 N. Gemini Drive, Flagstaff; Arizona 86001 E-mail: mchapman@flagmail.wr.usgs.gov and James R. Zimbelman Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, D. C. 20560 Received September 2, 1997; revised December 6, 1997 1992; Head et al., 1992). Coronae occur in linear chains Geologic mapping principles were applied to determine ge- and clusters (Stofan et al., 1992). The morphology of coro- netic relations between coronae and surrounding geomorpho- nae and the proposed sequence of events in the evolution logic features within two study areas in order to better under- are consistent with corona formation bv the ascent of a s6nd venusian eoronae. The study areas contain coronae in a plume of hot mantle material to the base if the lithosphere versus a contrasting chain and are ('1 west (Schubert etal., 1989,1990; Stofan and Head, 1990; Squyres of Phoebe Regio (quadrangle V-40; centered at latitude lSOS, longitude 250") and (2) west of Asteria and Beta Regiones et al., 1992; Stofan et al., 1992; Janes et al., 1992; Sandwell @tween latitude SON, longitude 23p and latitude 3 , l o n g and Schubert, 1992; Watten and Janes, 1995). A tude 275"). Results of this research indicate two groups of plume origin has also been proposed for the much larger coronae on Venus: (1) those that are older and nearly coeval regional rises such as Beta Regio (Kiefer and Hager, 1991). with regional plains, and occur globally; and (2) those that are younger and occur between Beta, Atla, and Themis Regiones or along extensional rifts elsewhere, sometimes showing system- atic age progressions. Mapping relations and Earth analogs suggest that older plains coronae may be related to a near-global resurfacing event perhaps initiated by a mantle superplume or plumes. Younger coronae of this study that show age progres- sion may be related to (1) a tectonic junction of connecting rifts resulting from local mantle upwelling and spread of a quasi-stationary hotspot plume, and (2) localized spread of post-plains volcanism. We postulate that on Venus most of the young, post-resurfacing coronal plumes may be concentrated within an area defined by the bounds of Beta, Atla, and Themis Regiones. 0 1998 Academic Press INTRODUCTION Some of the most intriguing geologic terrains seen in the new Magellan data contain numerous assemblages of coronae. These features are defined by a dominantly circu- lar structure consisting of an annulus of concentric ridges or fractures, an interior that is either topographically posi- tive or negative, a peripheral moat or trough, and, com- monly, numerous volcanic and tectonic landforms in the interior (Barsukov et al., 1986; Stofan and Head, 1990; Pronin and Stofan, 1990; Stofan et al., 1992; Squyres et al., On Earth, volcano-tectonic centers apparently related to mantle convection have a wide range of styles and sizes; the range of length scales in mantle convection and volcanic activity is an area of active research (Rabinowicz et al., 1990,1993). The Magellan radar data for Venus present the scientific community with as many new questions about coronae as they answer. For example, can a genetic connection be established between adjacent coronae and can these fea- tures be dated stratigraphically? Is there a pattern to the age of coronae? Do the venusian "hotspot" features mi- grate or are they stationary? In order to better understand the genesis of venusian coronae and perhaps answer some of these questions, we applied geologic mapping principles to determine strati- graphic relations between coronae and surrounding geo- morphologic features. Two study areas were selected that contain many asymmetric, circular, multiple, and overlap- ping coronae occurring within a contrasting cluster and a chain. Both areas lie within the major volcanic center concentration between Beta, Atla, and Themis Regiones (i.e., Crumpler et al.'s (1993) BAT anomaly). The first area is west of Phoebe Regio (quadrangle V-40; centered at latitude 15"S, longitude 250"); it contains a cluster of many coronae that appear to be randomly distributed (Fig. 1). 0019-1035198 $25.00 Copyright O 1998 by Academic Press All rights of reproduction in any form reserved. CORONA ASSOCIATIONS AND IMPLICATIONS FIG. 1. Merged Magellan color altirnetry and SAR data showing area of interest (USGS, 1997), location of 1 : 5,000,000-scale quadrangles of V-40 and area 2; north toward top; color key on Fig. 12. In contrast, the second area, west of Asteria and Beta Regiones, contains a long chain of coronae that lead from approximately latitude 23"N, longitude 239" to about lati- tude 43"N, longitude 275" (Fig. 1). Earlier studies indicated no systematic variation in age along chains of coronae (Stofan et al., 1992) and most coronal growth was fairly old relative to the global stratigraphy (Basilevsky and Head, 1995). However, results of this study suggest two groups of coronae: (1) those that are older and formed at about the same time, possibly in conjunction with a global resurfacing event; and (2) those that are younger than regional plains, some of which show systematic age progression along rifts or away from volcanic centers. Systematic age progression of coronae may be due to spread of large mantle plumes as they impinge on the lithosphere and perhaps minor migration of small mobile hotspots. Our work combined CHAPMAN AND ZIMBELMAN with that of other researchers implies that most of the young, post-resurfacing coronal plumes may be concen- trated within an area defined by the bounds of Beta, Atla, and Themis Regiones. METHODS Relative ages for most units were established by strati- graphic relations; however, some deposits without deci- pherable stratigraphy could only be mapped as undivided material units. Owing to the side-looking mode of acquisi- tion of the radar images, terrain sloping toward the imaging sensor appears brighter and spatially compressed com- pared to terrain that slopes away from the sensor (Ford and Plaut, 1993). Differences in surface roughness of units result in variation in brightness (Ford and Plaut, 1993); units with nearly identical surface morphologies are diffi- cult to discriminate. Because of these characteristics of the Magellan radar images, many stratigraphic relations are based on very limited observable contacts and many tec- tonic features could only be mapped as lineaments. In addition, the cumulative densities of impact craters are too low on Venus for determining ages at the 1 :5,000,000 mapping scale. However, studies of impact craters on the entire planet of Venus from the Magellan images have led workers to hypothesize that some perva- sive, near-global resurfacing event occurred a few hundred million years ago (Parmentier and Hess, 1992; Phillips et al., 1992; Schaber et al., 1992; Turcotte, 1993; Herrick, 1994). Additionally, Strom et al. (1992) noted that a global resurfacing event yields thermal evolution results consis- tent with numerical calculations of mantle convection in Venus (Arkani-Hamed and Taksiiz, 1984). Therefore, the present surface, which still could be active, and the coronae are no older than that resurfacing event. The age of a venusian geologic unit can only be determined from super- position relations of its distal materials; because these de- posits may not be representative of the entire age span of the unit, their placements on the correlation keys should be considered data points without quantitatively constrained error bars. AREA 1: PARGA CHASMA (V-40 QUADRANGLE) Trending northeast across the northeast quadrant of quadrangle V-40 is the tessera high that is the terminal end of the western arm of Phoebe Regio (Fig. 2). Trending northwest across the south half of the map area is Parga Chasma, a 1,870-km-long fractured depression that roughly connects Themis Regio (latitude 353, longitude 285") and Maat Mons (latitude OOS, longitude 195"). Southwest of the intersection of Parga Chasma with the western arm of Phoebe Regio, tessera material occurs as several isolated hills following the same northeast trend as the Phoebe Regio arm. East of its intersection with the tessera high, the chasma is offset 200 km to the south. Eighteen coronae are scattered throughout quadrangle V-40. A closer inspec- tion of these coronae reveals nine closely spaced coronae (six that bound the linear fracture zone of Parga Chasma and a line of three coronae trending northeast of Parga Chasma, along another fracture zone); four distal coronae that bound Parga Chasma; and a distal uncinate (hook- shaped) chain of five small coronae unrelated to Parga Chasma (Fig. 2). For this paper, the nine closely spaced coronae will be referred to as area 1 and will be contrasted with coronae occurring in area 2, west of Beta Regio (Fig. 1). Relations of coronae on quadrangle V-40 seem to sug- gest minor migration of small mantle plumes and spread of larger plumes at rift junctions. For example, previous mapping of V-40 revealed a systematic age and dimen- sional progression of the five small coronae (archaically termed arachnoids, i.e., coronae with radial fractures, to emphasize their fractures; Head et al., 1992) occurring in the uncinate chain, indicating that the features likely formed from a small mantle plume in a manner similar to terrestrial hotspot features (Fig. 3; Chapman and Kirk, 1996). Other observations led to the interpretation that this mantle plume has migrated beneath a stationary sur- face (Fig. 3; Chapman and Kirk, 1996). Volcanism of these small corona features postdates emplacement of the sur- rounding plains. Mapping of V-40 also revealed details about the area 1 coronae and the evolution of the Parga Chasma fracture zone (Chapman, 1996,1998). In area 1, all of the coronae postdate the surrounding regional plains and six of the nine coronae found along Parga Chasma appear to have formed coeval with and are probably related to the devel- opment of the chasma. The usual, relatively symmetric pattern of concentric and radial structures of coronae is absent around those coronae associated with Parga Chasma. Instead, these structures have been bent around and deflected about the fractures, graben, and depressions of the chasma. Depressions of Parga Chasma even form sections of the outer moats or troughs of these coronae. These aspects and the interlayered stratigraphy of coronal flows adjacent to Parga indicate that these coronae are related to the growth of Parga Chasma. Other aspects indicate that area 1 coronae are related to a tectonic junc- tion of rifts. With the exception of two coronae (mapped as units cd and cf, Fig. 4) at latitude IS'S, longitude 255" and at latitude 19"S, longitude 251?, all area 1 coronae are elevated, plateau-like features with mostly incomplete outer moats. GEOLOGIC RELATIONS OF AREA 1 The oldest unit in area 1 consists of tessera material (unit t; Fig. 4). Tessera or complex ridge terrain (CRT, CORONA ASSOCIATIONS AND IMPLICATIONS Y clarana of area 1 * I FIG. 2. Three-dimensional perspective view of quadrangle V-40 showing geomorphic features of interest (composed of left-looking synthetic aperture radar i n lox ; Kirk et al., 1992). Bindschadler and Head, 1991) was defined as a terrain previously used to relate coronae and extensional belts type characterized by a completely deformed surface of elsewhere on Venus (Baer et al., 1994). ridges and troughs, intersecting at various angles, typically At latitude 9"S, longitude 248.5" flows of corona unit a lying at higher elevations than surrounding materials, and from the central asymmetric corona are overlain by flows likely resulting from extensive horizontal deformation from Javine Corona at latitude SOS, longitude 251?, mapped (Basilevsky et al., 1986; Bindschadler and Head, 1989, as corona unit b (unit cb). Javine Corona is heart-shaped, 1991). The tessera material comprises the highland terrain a morphology that is controlled by regional tectonic activity of the western arm of Phoebe Regio (Fig. 2) and an isolated prior to or during corona formation (Pronin and Stofan, hill along the trend of the arm within real 1 (Fig. 4). 1990). Eastern flows of Javine Corona are overlain by flows Plains material (unit p) embays tessera material but is and cut by graben of corona unit c (unit cc) at latitude superposed by edifice unit 1 (unit el) at latitude 7.5"S, 3"S, longitude 254.5". Flows of edifice unit 2 (unit e2) that longitude 255" (Fig. 4). The outcrop of the edifice unit 1 were emitted from a caldera at latitude 3"S, longitude 257" is embayed by a flow of corona material unit a (unit ca) are interbedded between unit cc flows and those of a corona that can be traced to a 525-km-wide, unnamed, asymmetric to the north mapped as corona material unit d (unit cd; corona at latitude 13"S, longitude 250". Some flows from Fig. 6). this corona have been deformed by and therefore appear Deposits from unnamed corona unit a are overlain to to predate Parga Chasma. For example, south of the corona the southeast by flows of corona unit e (unit ce) that were the chasma is 2.5 to 3 km deep, and the earlier unit ca emitted from asymmetric Atete Corona (latitude 16"S, lon- flows from the corona extend to the opposite side of the gitude 244"; Fig. 4), whose north boundary forms a 1- to chasma (Figs. 4 and 5). Other parts of the corona appear 3-km-high cliff above Parga Chasma (Fig. 5). This portion to be coeval with the chasma, as the southeast section of of Parga is an arcuate segment that curves sharply to define the corona's moat or trough is a part of the chasma (Fig. the corona's boundary (Schubert et al., 1994, Fig. 1). Be- 5). Judging the age of features such as coronae and Parga cause the moat or trough of Atete contains a section of Chasma based on deformation patterns alone is suspect Parga Chasma, the growth of Atete was likely coeval with because of the phenomenon of rejuvenated sympathetic development of Parga Chasma. structural trends; however, it is a valid method in the data- Flows of corona material unit f (unit cf) superpose co- limited field of planetary geology and one that has been rona unit e at latitude 20?S, longitude 248.5". Some graben CHAPMAN AND ZIMBELMAIV FIG. 3. Three-dimensional perspective view of northeast corner of V-40 (composed of left-looking synthetic aperture radar images; vertical exaggeration lox; Kirk et al., 1992); 100-km scale bar accurate for front of view; 2-km-high volcano marked "a" shown in FIG. 10; Nagavonyi Corona in foreground. Outlined coronae chain shown in rectangular view in lower right corner. Arrows in rectangular view show migration of coronae centers with time and may represent direction of possible movement of small migratory plume relative to a stabile surface. and fractures that emerge from Parga Chasma extend south away from the Parga depression to the unit f corona, where the structures form part of the corona's annulus (Fig. 7). Therefore, emplacement of corona unit f was likely coeval with the development of Parga Chasma. Flows of corona unit f are overlain by outcrops of coronae material g and h (units cg and ch). Both outcrops of corona units g (Fig. 8) and h (Fig. 9) have associated annulus rings deflected and deformed to be symmetrical with structures of Parga Chasma, indicating the coronae units g and h were also likely coeval with the formation of Parga. Materials of corona unit h are in turn superposed by outcrops of corona material i (unit ci; Fig. 4). The material of corona unit i was emitted from Nagavonyi, a multiple corona at latitude 18S0S, longitude 259.5" (Fig. 3). The east sections of Nagavonyi's annular rings are deflected from their circular paths to form straighter trends that match the trends of graben extending out of Parga Chasma. Where these Parga Chasma graben cut the flank of Nagavonyi, they emit flows mapped as queried corona unit i material. The unit designa- tion of the flows is based on the interpretation that the deposits are flank fissure flows from Nagavonyi. The distor- tion of Nagavonyi's rings and the possible flank flows sug- gest that unit i was also deposited in conjunction with the development of Parga Chasma. Squyres et al. (1992) suggest that volcanism occurs early in the growth of a corona, followed soon after by early updoming and radial fracture. Mapping of area 1 supports this relation as all the coronal volcanic flows are cut by coronae fractures or Parga structures, with the exception of these late stage Nagavonyi flows. Altimetry data show that the deepest sections of the northwest-trending depression of Parga Chasma lies within area 1 (up to 3 km deep, Figs. 4 and 5). Part of the deep section follows an arcuate path about the north end of Atete Corona (latitude 163, longitude 244", Fig. I), form- ing a sector of the corona's moat or outer trench. Another less dominant fracture zone, associated with three of the area 1 coronae, extends from the convex edge of this path, around the west side of the unnamed corona (at about CORONA ASSOCIATIONS AND IMPLICATIONS FIG. 4. Geologic map of nine coronae shown in Fig. 2 (part of geologic map of V-40; Chapman, 1998); t-tessera material, p-plains material, unit el-older edifice material, units ca through ci-corona material units, and unit e2-younger edifice material; northwest trending 2- to 3-km- deep depression marks location of Parga Chasma. i latitude 12S0S, longitude 250?), and continues to the north- east; this fracture forms a triple junction with Parga (Fig. 10). This northeast triple junction forms a 2-km-deep de- pression and part of the west moat of the unnamed asym- metric corona (Fig. 5); Parga forms the unnamed corona's south moat. This corona is centered at the triple junction and has produced the oldest stratigraphic volcanic flows (unit ca); the flows from the other coronae are systemati- cally younger as the distance from each source coronae to the center of the junction increases (Fig. 4). A 38-mgal positive gravity anomaly and a small positive geoid height are associated with Atete Corona; these data have led Schubert et al. (1994) to postulate formation by either thermally induced thickness variations in a moder- ately thick (about 100 km) lithosphere or a deep positive mass anomaly due to subduction or underthrusting. These authors prefer the interpretation that the gravity anomaly is due to a buried mass anomaly resulting from subduction because (1) the anomaly is associated with the concave side of the arcuate segment of Parga and (2) similar high gravity anomalies over other portions of the same chasma are absent. However, map relations do not appear to support sub- duction at Parga (Chapman, 1996,1998). The prolific num- CHAPMAN AND ZIMBELMAN FIG. 5. Three-dimensional perspective view of northeast corner of Atete and an unnamed carona (co@osed o$&i-looking synthetic aperture radar images; vertical exaggeration 10X; Kirk et al., 1992); Parga Chasma forms the depression between them; viey't,@ward the southeast looking along Parga Chasma. c I ber of graben in the map area and possibly the broken, extended nature of the Phoebe Regio belt argue for exten- sion. Coronae clustering along the entire length of Parga Chasma have caused others to suggest that the chasma is a zone of rifting and extension (Stofan et al., 1992). In addition, there is no evidence of a compressed highland or arc along and coeval with Parga Chasma. Rather, elevated coronae bound either side of the chasma and occur along the fracture zone to the north (Fig. 1). Furthermore, the arcuate bend of Parga seems to form part of a triple junc- tion rift zone. On Earth, in three-branched rifts, generally two of the branches concentrate the tensional stresses, and the less active one functions as a buttress (Carracedo, 1994). Coronae, along both Parga and the less dominant northeast fracture zone, may represent rift-type clusters of aligned volcanic edifices. In terrestrial rift zones, most of the basaltic section is likely to be erupted from an elevated regioh above a rift and flow laterally away from the rift onta an adjacent continent (White and McKenzie, 1989). The placement of coronae on either side of Parga Chasma and their extensive flows away from the chasma are in agreement with this terrestrial eruption mode. The three branches of the Parga triple junction suggest a "least effort" fracture as a result of magma-induced verti- cal upward loading (Luongo et al., 1991), possibly related to local mantle upwelling of a quasi-stationary hotspot plume that could result in thermally induced crustal thick- ening (Fig. 10). Movement of the surface above the plume would tend to be uplift above and extension away from the triple junction as the plume rises and spreads beneath the surface (Olson and Nam, 1986; Koch, 1994). This rise and spread of a solitary large plume is in accordance with the systematic age reduction of the coronae away from the center of the triple junction. The relation can be modeled with plume spread propagating the rifts, coronae growth, and volcanism similar to the movement of 3 zippers away from the triple junction. In addition, the direction of coro- nae asymmetries trends along the rifts. Curiously, Ki Corona (450 km in diameter) exhibits structures that sup- port lateral spreading of another diapir along southeast Parga Chasma (Willis and Hansen, 1996). AREA 2: KAWELU PLANITIA AND BETA REG10 Area 2 consists of a 6,000-km-long chain of 12 circular coronae between Kawelu Planitia and Beta Regio, begin- ning at latitude 22.S0N, longitude 239" (Fig. 1). The coronae in the chain are mostly in low-lying plains, but a fracture CORONA ASSOCIATIONS AND IMPLICATIONS FIG. 6. Left-looking, synthetic-aperture radar (SAR) image centered at latitude L-a, ~ong~tude 255" showing stratigraphic relations of units; annulus ring of corona (unit cd) in upper left corner about 70 km wide; north toward top. Volcanic flows (marked a) of edifice unit 2 (q) flowed toward and changed direction after they encountered preexisting topography of older material of corona unit c (cc). These flows are in turn overlain by material from corona unit d (cd) at arrows marked b. zone that connects them intersects a mixture of plains and a western arm of Beta Regio tessera that extends through tesserae (Figs. 1 and 11). In contrast to the high relief Asteria Regio; the chain shifts direction east to Asteria coronae of area 1, the coronae of area 2 are only slightly Regio, then shifts northeast, impinging on the north edge topographically elevated in relation to their surrounding of Beta Regio, and ends at latitude 43"N, longitude 275". plain (Figs. 1 and 12).From the west, two arms of a bifurcat- The minor fracture zone which connects the 12 coronae ing corona chain trend northeast and coalesce after crossing of area 2 is not associated with a trough or elevated terrain CHAPMAN AND ZIMBELMAN FIG. 7. Left-looking, synthetic-aperture radar (SAR) image centered at latitude 193, longitude 251" showing graben and fractures (arrows) extending from Parga Chasma to the annulus of a corona (unit cf in Fig. 4); north toward top. (Fig. 1). Mapping of area 2 revealed no systematic age or dimensional progression of 7 of the coronae occurring in the 12-member-long chain and a minor north to south age progression of the 5 westernmost coronae (Fig. 11). The coronae of area 2 have fully to partly raised circular rims. With the exception of a corona at latitude 23"N, longitude 240" (unit ce), all of the coronae have depressed interiors (Fig. 12). GEOLOGIC RELATIONS OF AREA 2 As was true of area 1, the oldest unit in area 2 is also tessera material (unit t; Fig. 11). The tesserae occur in a western arm of Beta Regio and some nearby hills. Older plains material (unit PI) embays tessera material and is superposed by younger plains (unit p2) at about latitude 42"N, longitude 270" and by varied corona materi- als at the west end of the area. Seven coronae occur within the older plains material as tectonic features. One of the seven, Rauni Corona, in the northeast corner of the area, is partly flooded by younger plains (unit p2). The coronae of the older plains are only associated with very minor volcanic material or none at all and therefore flow superpo- sition could not be used to date them stratigraphically. Their annulus ring fractures appear to be coeval, with the possible exception of a corona at latitude 32"N, longitude 258" (Zamin Corona), whose annular rings appear to termi- nate those of a corona directly west of it (Fig. 11). These CORONA ASSOCIATIONS AND IMPLICATIONS FIG. 8. Left-looking, synthetic-aperture radar (SAR) image centered at latitude 23"S, longitude 251" showing northwest trend of corona's annulus (unit cg on Fig. 4) mimicking trend of Parga Chasma fractures. Radial fractures from corona extend east into Parga Chasma; north toward top. seven coronae show no systematic age progression along the chain nor do they appear to have obvious random age relations, but instead they seem to have formed synchronously (or perhaps randomly in a short period of time). Flow materials of Mawu Corona (unit ca; feature lies just west of area 2) overlie older plains on the northwest corner of the map area but are cut by the fracture zone. These flows are superposed by corona material b (unit cb) emitted from an edifice to the southeast of Mawu at latitude 29"N, longitude 243". Corona material b is embayed to the south by corona material c emitted from a small corona that is the northern, younger part of a multiple corona at latitude 24"N, longitude 244". Directly southwest of the corona source of material cb is a corona that emitted a large amount of lava. Mapped as corona material d (unit cd), the flows partly flooded the annular rings of the older cb corona and flowed south over older plains material; the material postdates the fracture zone (Figure 13). The youngest unit in area 2 is corona material e (unit ce) emitted from a feature at about latitude 23"N, longi- tude 240". The seven synchronous coronae of area 2 appear to have formed as structural features coeval with their fracture zone. They are associated with only minor volcanic flows and show no age progression; in fact, they appear to have formed at about the same time. They are constructed of tessera-embaying plains that have been deformed by the coronae and the fracture zone. Relations between the frac- ture zone and surrounding units indicate the zone and the coronae are fairly old and might have formed during or shortly after emplacement of the plains. This age assign- ment is compatible with the interpretation by Stofan et al. (1992) that the oldest coronae are in low-lying plains. Rauni Corona was previously described as an old feature based on topography and superposition (Pronin and Stofan, 1990). Like the coronae of Hecate Chasma that show no sys- tematic age relations (Hamilton and Stofan, 1996), a single CHAPMAN AND ZIMBELMAN FIG. 9. Left-looking, synthetic-aperture radar (SAR) image centered at latitude 173, longitude 255" showing northwest trending fractures of Parga Chasma wrapping around corona's annulus (unit ch on Fig. 4); north toward top. stationary thermal anomaly underlying a moving litho- sphere is an unlikely origin for these coronae. The coronae may have been formed by upwelling similar to those ob- served at terrestrial mid-ocean ridges, but the lack of a trough and other evidence of much extension suggest little spreading along the upwelling zone. Instead, they are similar to the chains of seamounts and guyouts in the west- ern Pacific. These seamounts are not associated with rifts, most date from the mid-Cretaceous, they occur randomly or in randomly oriented chains in a slightly older mid- Cretaceous oceanic plain, and they show no evidence of edifice age progression along the chains (Lincoln et al., 1993; Larsen, 1995). One of the hypotheses of origin for the western Pacific is that all of the seamounts and the surrounding ocean floor resulted from a mid-Cretaceous superplume episode (Coffin and Eldholm, 1993; Larsen, 1995). Many workers are now considering previous su- perplume events and the possibility that the Earth has a "heartbeat" (Larson, 1995). (In fact, some believe that thermal effects on Earth act on 500 million year intervals FIG. 10. Three-dimensional perspective view of quadrangle V-40 (composed of left-looking synthetic-aperture radar images; vertical exaggeration lox; Kirk et al., 1992) with illustrations; north is toward top; tessera outlined in black; arrows indicate inferred direction of surface movement; quadrangle is approximately 3,300 km wide. Flows from 2-km-wide volcano marked "a" embay arm of Phoebe Regio tessera. Deepest section of Parga Chasma indicated by white zone marked "b." Cross-hatched zone marked "c" is less deformed fracture zone trending northeast away from Parga. White zone marked "d" is an extension of Parga Chasma offset to south after encountering trend of tessera high. Relatively young extension zone marked "e" intersects volcano. Illustration F (in corner) shows surface movement away from triple zone fracture possibly similar to that formed by zones marked "b" and "c" on quadrangle. CORONA ASSOCIATIONS AND IMPLICATIONS 355 CHAPMAN AND ZIMBELMAN FIG. 11. Geologic map of area 2 (location on Fig. 1); t-tessera material, pl-regional plains material, unit pz-younger plains, and units ca through ce-corona material units. periodically causing the terrestrial continents to assemble into a single landmass; Murphy and Nance, 1992). In contrast, the five western coronae of area 2 occur along a bifurcation of the chain and are associated with numerous volcanic flows that superpose older tesserae em- baying plains; four are younger than the structures of the fracture zone. The age trend of these coronae is perpendic- ular to the end of the western arm of Beta Regio and lies on the south edge of Kawelu Planitia (a high plain) northeast of Ulfrun Regio. Numerous, large volume volca- nic flows occur in Kawelu Planitia (Zimbelman, in review) and are perhaps related to localized upwellings. The age progression of the five coronae away from Kawelu Planitia may be related to an increase of volcanism away from the planitia in an event that continued after the formation of the fracture zone; the increase in volcanism could also be due to the spread of a large plume or perhaps multiple small plumes impinging on the lithosphere in a wider area. DISCUSSION Numerous models of hotspot dynamics exist and there are primarily two competing models for the generation of FIG. 12. Color altimetry merged with SAR data (USGS 1997) showing coronae of area 2 and those of area 1; note that coronae of area 2 are generally topographically lower than and more symmetric, circular features in comparison to those of area 1. Area 2 coronae: A (Zamin Corona and unnamed corona to the east; latitude 32ON, longitude 258"), B (latitude 35"N, longitude 270?), C (latitude 30?N, longitude 246"), D (Rauni Coronae; latitude 41?N, longitude 270.5"), E (two coronae at latitude 2g0N, longitude 242"), F (latitude 23"N, longitude 240?), G (latitude 23"N, longitude 244"), H (latitude 31?N, longitude 250"); and area 1 coronae: I (latitude 13"S, longitude 250?), J (Javine Corona; latitude 5"S, longitude 251?), K (latitude l.SOS, longitude 255"), L (latitude 3 3 , longitude 254.5"), M (Atete Corona; latitude 16% longitude 244"), N (latitude 19"S, longitude 251?), 0 (latitude 16"S, longitude 256"), P (Nagavonyi Corona; latitude 18.5"S, longitude 259.5"), Q (latitude 25"S, longitude 251'). CORONA ASSOCIATIONS AND IMPLICATIONS Planetary Radius (km) CHAPMAN AND ZIMBELMAN FIG. 13. Three-dimensional perspective view of part of area 2 centered at latitude 26"N, longitude 242" (composed of left-looking synthetic- aperture radar images; vertical exaggeration lox; Kirk et al., 1992) showing the voluminous flows (unit cd) of a corona superposed on older corona (mapped as unit cb) and fractures of regional plains. coronae from mantle plumes. The phenomena responsible for the coronae may be directly analogous to the well- known terrestrial hotspots. However, it has been suggested that the process giving rise to terrestrial mantle plumes is inactive or very weak on Venus (Arkani-Hamed and Tok- soz, 1984; Arkani-Hamed et al., 1993; Tackley et al., 1992). Tackley et al. (1992) recognized this difficulty, and they proposed that a novel, Rayleigh-Taylor-like instability might take place within the broad upwelling zones and might account for localized magmatism. In their model, plumes are driven by the buoyancy provided by partial melting (Tackley and Stevenson, 1993). The process is self- accelerating; the degree of melting, and hence the buoy- ancy, increases as the material ascends (Tackley and Ste- venson, 1993). Scaling their numerical simulations to condi- tions on Venus, Tackley et al. (1992) estimated that the plumes would have a typical size of 250 km and that they would ascend on a timescale of 30 m.y. Magma-generation rates on the order of 1,000 km3/Myr could be sustained over this period. This 30-Myr timescale is much shorter CORONA ASSOCIATIONS AND IMPLICATIONS 359 than the inferred few hundred million year average age of Larsen, 1995). However, if many coronae are old and date the surface (Parmentier and Hess, 1992; Phillips et al., 1992; from regional plains, then their widespread global occur- Schaber et al., 1992; Strom et al., 1992, 1994; Turcotte, rence area is much larger in size than that of the terrestrial 1993); therefore, many plumes could have formed and "superplume" in the western Pacific. It is more likely, dissipated over the period preserved in the geologic record. based on the coronae's age and relation to regional plains, The two competing models of plume formation do lead that they are related to the hypothetical, near-global resur- to different predictions for the spatial and temporal distri- facing event that occurred a few hundred million years ago bution of coronae. Terrestrial experience suggests that (Parmentier and Hess, 1992; Phillips et al., 1992; Schaber plumes are localized, relatively stable, and long-lived (Wil- et al., 1992; Turcotte, 1993; Herrick, 1994). Global resurfac- son, 1963, 1965). Given that the terrestrial lithosphere ing could be expected to be responsible for the formation moves relatively rapidly with respect to the mantle, the of a great many corona features planet-wide. Most likely, surface signature of a hotspot is a well-defined chain of this may be the reason preliminary mapping of about 90% volcanic features in which age increases systematically in of the Venus surface by Stofan et al. (1992) detected no one direction. If the lithosphere did not move with respect systematic variation in age along chains of coronae. Simply to the mantle, then the hotspot ~ o u l d build a single large stated, most coronae may be old and date from the time volcanic construct or it might leave a track indicative of of resurfacing. Based on morphology and superposition, the motion of the plume through the mantle such as that Stofan et al. (1992) found that the majority of coronae are by Chapman Kirk (1996, Fig. 3). Alterna- in the middle to late stages of evolution, with about a tively, plumes triggered by melt-driven instability would quarter of the population having very subdued morphology occur throughout the zones of mantle upwelling (Schubert and topography indicating an old age. et al., 1990; ~chubert, 1992; Tackley et al., 1992). Such ~f the hypothesis that most coronae are old is correct, ~ l u m e s would tend to form much like bubbles in boiling how does that account for the obsemations of younger, water, leaving coronae with age relations that were virtu- systematically age-progressed coronae in both this and ally random. other studies (Chapman and Kirk, 1996)? What is the ex- rather than sysf emafic or random age relations planation for the ongoing, nonsystematic formation of co- among coronae, what is observed in this study and another ronae along and coeval with the evolution of Hecate and Kirk, 1996) is stratigraphically Chasms (Hamilton and Stofan, 1996), or what seems to ble systematic and synchronous relations between linearly be ongoing formation of coronae along and coeval with trending coronae. the extension belts of northern Lada Terra (Baer et al., The seven synchronous 'Oronae of area appear 1994)? The answer to these questions could be that all have formed as structural features within older (tessera- of these younger coronae lie in a major volcanic center emba~ing) plains (2) might have formed during concentration between the Beta, Atla, and Themis Regi- or shortly after emplacement of the older plains; (3) are ones (Crumpler et al.'s BAT anomaly, 1993) or are found all circular features; (4) have interior depressions; and (5) elsewhere on extension belts. Relative to the have raised topography only along their annuli. Basilevsky entire plant, a great many coronae are between Beta, Atla, and Head (1995) note that following tessera formation, extensive volcanic flooding resurfaced at least 85% of the and Themis regiones (Stofan et al., 1992). In fact, Squyres planet. The seven synchronous coronae are likely old fea- et al. (1993) show that while the distribution of coronae tures as they occur within tessera embaying plains. Further- over most of the planet is indistinguishable from random more, they have morphologies consistent with an advanced (density equals 0.73 coronae/106 km2), the BAT region state of development (Stofan et al., 1992; Squyres et al., contains a significantly higher coronae concentration (>2 1992). Although it may be too premature to make global coronae/106 km2). Although preliminary mapping by Sto- stratigraphic correlations, many other coronae elsewhere fan et al. (1992) found no systematic variation in age along on Venus may also be old features. McGill (1994) noted corona chains, they did find a number of overlapping coro- that coronae in the region of Eisla Regio began to form nae that show a clear age progression, and 30 out of 35 of very shortly after emplacement of the regional (tessera their multiple coronae are within the BAT anomaly. We embaying) plains. addition, from mapping of their agree that volcanism is concentrated in the BAT anomaly random sample of 36 test areas on Venus, Basilevsky and as noted by Crum~ler et az. (l993)~ but we also suggest Head (1995) suggest that most coronal growth was fairly that much of this volcanism has occurred post-resurfacing old. and may still be active. This suggestion is supported by By analogy (see above), the synchronous coronae and analysis of Magellan gravity data that confirm that Beta the surrounding plains of area 2 are suggested to be the and Atla Regiones are active hotspots (Smrekar, 1994). result of a superplume similar to the short-lived mid- Ironically, Basilevsky and Head (1995) based their sugges- Cretaceous superplume (Coffin and Eldholm, 1993; tion that most coronal growth was fairly old on their ran- CHAPMAN AND ZIMBELMAN dom sample of 36 test areas on Venus, only one of which lies within the BAT anomaly. Area 1 coronae (1) lie along a triple junction composed of a major fractured trough zone (Parga Chasma) and a less dominant fracture zone; (2) are high relief features; (3) contain numerous stratigraphically datable lava flow lobes; (4) are mostly asymmetric or multiple features; and (5) are younger than the surrounding tessera embaying plains. Briefly, area 1 contains an elevated rift with a tec- tonic rift junction and extensive coronal volcanism that postdates the regional plains. Keddie and Head (1995) suggest that there are two classes of volcanic rises on Venus: an areally large type with significant volcanism and rift zone formation and a smaller-scale type with extensive volcanism but moderate rifting. Stofan et al. (1994) point out that the volcanic rises of Atla and Beta Regiones lie at the major tectonic junctions of connecting rift zones (triple junctions). The tectonic rift junction of area 1 lies in the center of the BAT anomaly. Associated with Crum- pler et aL's (1993) anomaly are geologic characteristics interpreted as rifting and mantle upwelling. Could area 1's tectonic rift junction represent the incipient formation of an additional smaller region of the rift-type class within the BAT anomaly? If the above suppositions are correct, what are the impli- cations for the two competing models of plume formation? Older coronae plumes associated with resurfacing appear to be coeval, but could have formed randomly during a short period of time. Perhaps these plumes are indeed triggered by Rayleigh-Taylor-like melt-driven instability within the broad upwelling zones (Tackley et al., 1992). Corona plumes that postdate resurfacing, discussed in this paper and others (Baer et al., 1994; Hamilton and Stofan, 1996), appear to be mostly associated with rifting; the ma- jority seem to occur either within the BAT anomaly or on active extension belts elsewhere. Younger coronae ob- served in this study may be the surface expression of spreading stationary mantle plumes. Minor motion of an- other younger small plume has also been observed (Chap- man and Kirk, 1996). Therefore, post-resurfacing plumes are not occurring randomly like boiling bubbles in a pot, but are confined to certain areas on the planet and are long-lived to the point that local plume spread is evident, as is limited motion of small plumes. CONCLUSION Results of this study indicate that coronae either form in conjunction with older plains and show no systematic age progression or postdate older plains and occur in confined global locations, in some cases showing systematic age progression. The systematic age progression, rather than indicating motion of the surface relative to the mantle, indicates spread and movement of local mantle plumes. Mapping relations and Earth analogs cause us to suggest that older plains coronae may be related to a near-global resurfacing event perhaps initiated by a superplume or plumes. Other coronae of this study that are noticeably younger than the surrounding plains all lie within Crumpler et al.'s (1993) BAT anomaly, a concentration of major volcanic centers between the Beta, Atla, and Themis Regi- ones. We postulate that on Venus there currently exists a large-scale pattern of mantle circulation that concentrates most post-resurfacing coronal plumes within the BAT anomaly or along active extension zones elsewhere. ACKNOWLEDGMENTS We thank Lisa Gaddis and Jeff Kargel of the U.S. Geologic Survey and Icarus reviewers E. M. Parmentier and D. M. Janes for their timely and informative efforts as regards this manuscript. REFERENCES Arkani-Hamed, J., and M. N. Toksoz 1984. Thermal evolution of Venus. Phys. Earth Planet. Interiors 34, 232-250. Arkani-Hamed, J., G. G. Schaber, and R. G. Strom 1993. Constraints on the thermal evolution of Venus inferred from Magellan data. J. Geophys. Res. 98,5309-5316. Baer, G., G. Schubert, D. L. Bindschadler, E. R. Stofan 1994. Spatial and temporal relations between coronae and extensional belts, north- ern Lada Terra, Venus. J. Geophys. Res. 99, 8355-8369. Barsukov, V. L., and 29 colleagues 1986. The geology and geomorphology of the Venus surface as revealed by the radar images obtained by Veneras 15 and 16. J. Geophys. Res. 91,378-398. Basilevsky, A. T., and J. W. Head, I11 1995. Global stratigraphy of Venus: Analysis of a random sample of thirty-six test areas. Earth Moon Planets 66,285-336. Basilevsky, A. T., A. A. Pronin, L. B. Ronca, V. P. Kryuchkov, A. L. Sukhanov, and M. S. Markov 1986. Styles of tectonic deformations on Venus: Analysis of Venera 15 and 16 data. J. Geophys. Res. 91,399-411. Bindschadler, D. L., and J. W. Head, 111 1989. Characterization of Venera 15/16 geologic units from Pioneer Venus reflectivity and roughness data. Icarus 77, 3-20. Bindschadler, D. L., and J. W. Head, I11 1991. Tessera Terrain, Venus: Characterization and models for origin and evolution. J. Geophys. Res. 96,5889-5907. Carracedo, J. C. 1994. The Canary Islands: An example of structural control on the growth of large oceanic-island volcanoes. J. Volcan. Geotherm. Res. 60,225-241. Chapman, M. G. 1996. The evolution of Parga Chasma and coronae growth on Venus. In Abstracts 27th Lunar Planet. Sci. Con$, pp. 205- 206. Lunar and Planetary Institute, Houston. Chapman, M. G. 1998. Geologiclgeomorphic map of the Galindo (V-40) quadrangle. U.S. Geological Survey Miscellaneous Investigations Series, Map 1-2613. [ l : 5,000,000-scale] Chapman, M. G., and R. L. Kirk 1996. A migratory mantle plume on Venus: Implications for Earth? J. Geophys. Res. 101, 15,953-15,967. Coffin, M. F., and Olav Eldholm 1993. Large igneous provinces. Sci. Amer. 269,42-49. CORONA ASSOCIATIONS AND IMPLICATIONS Crumpler, L. S., J. W. Head, and J. C. Aubele 1993. Relation of major volcanic center concentrations on Venus to global tectonic patterns. Science 261,591-595. Ford, J. P., and J. L. Plaut 1993. Magellan image data. In Guide to Magellan Image Interpretation, JPL Publication 93-24, 7-18. Hamilton, V. E., and E. R. Stofan 1996. The geomorphology and evolution of Hecate Chasma, Venus. Icarus 121, 171-194. Head, J. W., L. S. Crumpler, J. C. Aubele, John Guest, and R. S. Saunders, 1992. Venus volcanism: Classification of volcanic features and struc- tures, associations, and global distribution from Magellan data. J. Geo- phys. Res. 97, 13,153-13,198. Herrick, R. R. 1994. Resurfacing history of Venus. Geology 22,703-706. Janes, D. M., S. W. Squyres, D. L. Bindschadler, G. Baer, G. Schubert, V. L. Sharpton, and E. R. Stofan 1992. Geophysical models for the formation and evolution of coronae on Venus. J. Geophys. Res. 97, 16,055-16,067. Keddie, S. T., and J. W. Head 1995. Formation and evolution of volcanic edifices on the Dione Regio rise, Venus. J. Geophys. Res. 100,11,729- 11,754. Kiefer, W. S., and B. H. Hager 1991. A mantle plume model for the equatorial highlands of Venus. J. Geophys. Res. 96, 20,497-20,966. Kirk, R. L., L. A. Soderblom, and E. M. Lee 1992. Enhanced visualization for interpretation of Magellan radar data. J. Geophys. Res. 97 [supplement to Magellan Special Issue], 16,371-16,380. Koch, D. M. 1994. A spreading drop model for plumes on Venus. J. Geophys. Res. 99,2035-2052. Larson, R. L. 1995. The mid-Cretaceous superplume episode. Sci. Amer. 272,8246. Lincoln, J. M., M. S. Pringle, and I. P. Silva 1993. Early and Late Creta- ceous volcanism and reef-building in the Marshall Islands. In The Meso- zoic Pacific: Geology, Tectonics, and Volcanism (M. S. Pringle, W. W. Sager, W. V. Sliter, and S. Stein, Eds.), pp. 279-305. Amer. Geo- phys. Union. Luongo, G., E. Cubellis, F. Obrizzo, and S. M. Petrazzuoli 1991. A physical model for the origin of volcanism of the Tyrrhenian margin: The case of the Neapolitan area. J. Volcan. Geotherm. Res. 48, 173-185. McGill, G. E. 1994. Hotspot evolution and venusian tectonic style, J. Geophys. Res. 99, 23,149-23,161. Murphy, J. B., and R. D. Nance 1992. Mountain belts and the superconti- nent cycle. Sci. Amer. 266, 84-91. Olson, Peter, and I. S. Nam 1986. Formation of seafloor swells by mantle plumes. J. Geophys. Res. 91,7181-7192. Parmentier, E. M., and P. C. Hess 1992. Chemical differentiation of a convecting planetary interior: Consequences for a one-plate planet such as Venus. Geophys. Res. Lett. 19,2015-2018. Phillips, R. J., R. F. Raubertas, R. E. Amidson, I. C. Sarkar, R. R. Herrick, N. Izenberg, and R. E. ~ r i m m 1992. Impact crater distribution and the resurfacing history of Venus. J. Geophys. Res. 97, 15,923-15,948. Pronin, A. A., and E. R. Stofan 1990. Coronae on Venus: Morphology, classification, and distribution. Icarus 87, 452-474. Rabinowicz, M., G. Ceuleneer, M. Monnereau, and C. Rosemberg 1990. Three-dimensional models of mantle flow across a low viscosity zone: Implications for hotspot dynamics. Earth Planet. Sci. Lett. 99,170-184. Rabinowicz, M., S. Rouzo, J. C. Sempere, and C. Rosemberg 1993. Three- dimensional flow beneath mid-ocean ridges. J. Geophys. Res. 98,7851- 7869. Sandwell, D. T., and G. Schubert 1992. Flexural ridges, trenches, and outer rises around coronae on Venus. J. Geophys. Res. 97, 16,069-16,083. Schaber, G. G., R. G. Strom, H. J. Moore, L. A. Soderblom, R. L. Kirk, D. J. Chadwick, D. D. Dawson, L. R. Gaddis, J. M. Boyce, and J. Russell 1992. Geology and distribution of impact craters on Venus: What are they telling us? J. Geophys. Res. 97, 13,257-13,301. Schubert, G. 1992. Numerical models of mantle convection. Annu. Rev. Fluid Mech. 24,359-394. Schubert, G., D. Bercovici, D. J. Thomas, and D. B. Campbell 1989. Venus coronae: Formation by mantle plumes. In Abstracts, 20th Lunar Planetary Science Conference, pp. 968-969. Lunar and Planetary Insti- tute, Houston. Schubert, G., D. Bercovici, and G. A. Glatzmaier 1990. Mantle dynamics in Mars and Venus: Influence of an immobile lithosphere on three- dimensional mantle convection. J. Geophys. Res. 95, 14,105-14,129. Schubert, G., W.-B. Moore, and D. T. Sandwell 1994. Gravity over coronae and chasmata on Venus. Icarus 112,130-146. Smrekar, S. E. 1994. Evidence for active hotspots on Venus from analysis of Magellan gravity data. Icarus 112, 2-26. Squyres, S. W., D. M. Janes, G. Baer, D. L. Bindschadler, G. Schubert, V. L. Sharpton, and E. R. Stofan 1992. The morphology and evolution of coronae on Venus. J. Geophys. Res. 97,13,611-13,634. Squyres, S. W., D. M. Janes, G. Schubert, D. L. Bindschadler, J. E. Moersch, D. L. Turcotte, and E. R. Stofan 1993. The spatial distribution of coronae and related features on Venus. Geophys. Res. Lett. 20,2965- 2968. Stofan, E. R., and J. W. Head 1990. Coronae of Mnemosyne Regio: Morphology and origin. Icarus 83,216-243. Stofan, E. R., V. L. Scharpton, G. Schubert, G. Baer, D. L. Bindschandler, D. M. Janes, and S. W. Squyres 1992. Global distribution and character- istics of coronae and related features on Venus. Implications for origin and relation to mantle processes. J. Geophys. Res. 97,13,347-13,378. Stofan, E. R., D. L. Bindschandler, D. A. Senske, and S. E. Smrekar 1994. Large topographic rises on Venus: Implications for mantle upwelling. J. Geophys. Res. 100, 23,317-23,327. Strom, R. G., G. G. Schaber, J. Arkani-Hamed, and M. N. Toksdz 1992. Global resurfacing of Venus. In Abstracts, 23rd Lunar and Planetary Science Conference, pp. 1379-1380. Lunar Planet. Inst., Houston. Strom, R. G., G. G. Schaber, and D. D. Dawson 1994. The global resurfac- ing of Venus. J. Geophys. Res. 99, 10,899-10,926. Tackley, P. J., and D. J. Stevenson 1993. A mechanism for spontaneous self-perpetuating volcanism on the terrestrial planets. In Flow and Creep in the Solar System (D. B. Stone and S. K. Runcorn, Eds.), pp. 307-321. Kluwer, Norwell, MA. Tacklev. P. J.. D. J. Stevenson. and D. R. Scott 1992. Volcanism bv - . melt-driven Rayleigh-Taylor instabilities and possible consequences of meltine for admittance ratios on Venus. Abstracts, International " Colloquium on Venus, Aug. 10-12, Pasadena, CA, pp. 123-125. Turcotte, D. L. 1993. An episodic hypothesis for venusian tectonics. J. Geophys. Res. 98, 17,061-17,068. USGS 1997. Altimetric radar image map of Venus. U.S. Geological Sur- vey Miscellaneous Investigations Series Map 1-2444, sheet 3 of 4. [I : 50,000,000 scale] Watters, T. R., and D. M. Janes 1995. Coronae on Venus and Mars: Implications for similar structures on Earth. Geology 23,200-204. White, R., and D. McKenzie 1989. Magmatism at rift zones: The genera- tion of volcanic continental margins and flood basalts. J. Geophys. Res. 94,7685-7729. Willis, J. J., and V. L. Hansen 1996. Conjugate shear fractures at "Ki Corona," southeast Parga Chasma. In Abstracts, 27th Lunar Planetary Science Conference, pp. 1443-1444. Lunar Planet. Inst., Houston. Wilson, J. T. 1963. A possible origin of the Hawaiian Islands. Can. J. Phys. 41,863-868. Wilson, J. T. 1965. Evidence from ocean islands suggesting movement in the earth. Roy. Soc. London Philos. Trans. 258,145-165. Zimbelman, J. R., submitted for publication. Geologic map of the Kawelu Planitia (V 16) quadrangle. U.S. Geological Survey Miscellaneous In- vestigations Series Map. [I : 5,000,000 scale]