Jayne C. Aubele

Venus volcanism: Classification of volcanic features and structures, associations, and global distribution from Magellan data

A preliminary analysis of a global survey of Magellan data covering over 90% of the surface and designed to document the characteristics, location, and dimensions of all major volcanic features on Venus has revealed over 1660 landforms and deposits. These include over 550 shield fields (concentrations of small volcanoes <20 km in diameter), 274 intermediate volcanoes between 20 and 100 km diameter with a variety of morphologies, 156 large volcanoes in excess of 100 km diameter, 86 calderalike structures independent of those associated with shield volcanoes and typically 60–80 km in diameter, 175 coronae (annulus of concentric ridges or fractures), 259 arachnoids (inner concentric and outer radial network pattern of fractures and ridges), 50 novae (focused radial fractures forming stellate patterns), and 53 lava flood-type flow fields and 50 sinuous lava channels (all of which are in excess of 102–103 km in length). The vast majority of landforms are consistent with basaltic compositions; possible exceptions include steep-sided domes and festoons, which may represent more evolved compositions, and sinuous rules, which may represent more fluid, possibly ultramafic magma. The range of morphologies indicates that a spectrum of intrusive and extrusive processes have operated on Venus. Little evidence was found for extensive pyroclastic deposits or landforms, consistent with the inhibition of volatile exsolution and consequent disruption by the high surface atmospheric pressure. The large size of many volcanic features is evidence for the presence of very large magma reservoirs. The scale of resurfacing implied by individual features and deposits is typically much less than 125,000 km2. The areal distribution, abundance, and size distribution relationships of shield fields, arachnoids, novae, large volcanoes, and coronae strongly suggest that they are the surface manifestation of mantle plumes or hot spots and that the different morphologies represent variations in plume size and stage and thermal structure of the lithosphere. Maps of the global distribution of volcanic features show that they are broadly distributed globally, in contrast to the plate boundary concentrations typical of Earth. However, they are not randomly distributed on the surface of Venus. An observed deficiency of many volcanic features in several lowland areas of Venus may be due to an altitude-dependent influence of atmospheric pressure on volatile exsolution and the production of neutral buoyancy zones sufficient to form magma reservoirs; this would favor lava floods and sinuous channels at low elevations and edifices and reservoir-related features at higher elevations. A major concentration of volcanic features is observed in the Beta/Atla/Themis region, an area covering about 20% of the planet and centered on the equator. This region is unique in that it is the site of local concentrations of volcanic features with concentrations 2–4 times the global average, an interlocking network of rift and deformation zones, several broad rises several thousand kilometers in diameter with associated positive gravity anomalies and tectonic junctions, and evidence for volcanically embayed impact craters. Although the region as a whole does not appear to be anomalously older or younger than the rest of Venus, there is evidence that the most recent volcanic activity on the planet occurs here, and the presence of this series of concentrations suggests that the mantle in this region is anomalous. Analysis of the impact crater population shows that it cannot be distinguished from a completely spatially random population (Phillips et al., this issue), and several end-member models for this distribution are possible: (1) single production age or “spasmodic or catastrophic volcanism” model: craters have accumulated subsequent to a global volcanic resurfacing event about one-half billion years ago (Schaber et al., this issue); (2) vertical equilibrium or “leaky planet” model: craters are removed by slow accumulation of lava over the whole planet leading to a range of volcanic degradation states for craters; (3) regional resurfacing or “collage” or “cookie-cutter” model: craters are removed largely instantaneously by superposition of features and deposits; the horizontal scale of resurfacing does not exceed the horizontal scale of randomness of the crater population. Our data on the scale and location of resurfacing are consistent with the regional resurfacing model and with the catastrophic resurfacing model. The nature and abundance of impact craters definitely degraded by volcanism also favor these two models, although uncertainty exists as to whether all such craters have been detected. Although a process toward the regional resurfacing end-member model presently seems most plausible, distinction between the three models requires an understanding of the mode and timing of emplacement of the volcanic plains that make up the majority of the surface and which are not clearly related to the edifices and features mapped in this study. In addition, the resurfacing mechanisms involved in the catastrophic resurfacing models are not yet explicitly enough formulated to test with the existing data. An equilibrium resurfacing model implies a volcanic flux of 0.5 km3/yr, a value similar to the present rate of intraplate volcanism on Earth (0.3–0.5 km3/yr). This value is broadly comparable to that implied by the edifices and deposits on Venus mapped in this study. Geologically recent volcanism on Venus is dominated by features interpreted to be related to mantle plumes.

Small volcanic edifices and volcanism in the plains of Venus

The most widespread terrain type on Venus consists of volcanic lowland plains. Several styles of volcanism are represented in the plains. The most extensive volcanic units consist of flood lavas, the largest of which have volumes of the order of thousands of cubic kilometers. As with terrestrial flood lavas, they are inferred to have erupted at high effusion rates. They show a range of radar backscatter characteristics indicating different surface textures and ages. Small edifices on the plains occur mainly in clusters associated with fracture belts. The majority are shield volcanoes that may be up to a few tens of kilometers across but are generally 10 km or less in diameter. Volcanic cones have the same size range. Volcanic domes have diameters up to several tens of kilometers and volumes of the order of 100 km3. These are interpreted as being constructed of lava erupted with a relatively high effective viscosity and thus possibly composed of more silicic lava. For many domes, the flanks were unstable during and afte eruption and suffered gravity sliding that produced steep, scalloped outer margins. Because of the high atmospheric pressures on Venus, explosive activity is less likely to occur than on Earth. However, n a few plains areas there is evidence of pyroclastic deposits surrounding craters, indicating that volatile contents in some of the magmas may be high in comparison to Earth. The clusters of small volcanic edifices are considered to be analogous to plains volcanism, similar to that of the Snake River Plain of Idaho. There may also be analogues with terrestrial volcanic clusters associated with mid-oceanic ridges.

Learning to Recognize Volcanoes on Venus

Dramatic improvements in sensor and image acquisition technology have created a demand for automated tools that can aid in the analysis of large image databases. We describe the development of JARtool, a trainable software system that learns to recognize volcanoes in a large data set of Venusian imagery. A machine learning approach is used because it is much easier for geologists to identify examples of volcanoes in the imagery than it is to specify domain knowledge as a set of pixel-level constraints. This approach can also provide portability to other domains without the need for explicit reprogramming; the user simply supplies the system with a new set of training examples. We show how the development of such a system requires a completely different set of skills than are required for applying machine learning to “toy world” domains. This paper discusses important aspects of the application process not commonly encountered in the “toy world,” including obtaining labeled training data, the difficulties of working with pixel data, and the automatic extraction of higher-level features.

Vesicle zonation and vertical structure of basalt flows

Abstract Observation and measurement of vertical sections of thin ( Numerical simulations, performed for this study, suggest that these characteristic patterns of vesicle zonation are the result of the growth and rise of gas bubbles in cooling lavas rather than the result of dynamic conditions such as flow movement or convection. As a bubble grows, it begins to ascend, and continues to ascend until it is overtaken by solidification progressing inward from either the upper or lower cooling surfaces of the flow. Bubbles that start out high in the flow will ascend ahead of the lower solidification front and cease rising only after encountering the downward-advancing upper solidification front, and bubbles near the base of a flow will be entrapped by the upward-advancing lower solidification front. Bubbles that start and rise just above the lower solidification front form the lower part of the upper vesicular zone. Such bubbles will also have longer times in which to grow than bubbles that are either higher or lower and are therefore among the largest in the flow. A zone free of vesicles will remain between the last bubbles to ascend to the upper solidification front and the last bubbles to be overtaken by the lower solidification front.

Evidence of Regional Structural Controls on Vent Distribution: Springerville Volcanic Field, Arizona

Quantitative analysis of the geographic distribution of vents and comparison with regional structural, petrologic, and vent age data provide insight into the processes governing the emplacement of vents in the Springerville volcanic field, Arizona. A total of 409 vents in the Springerville volcanic field (SVF) have a mean distance to nearest neighbor vents of 955 m, a much closer spacing than is common in some platform-type volcanic fields. Based upon a cluster analysis search radius parameter of 4500 m, these vents comprise seven geographic clusters, with only five outlying vents occurring in the entire field. Cinder cone clusters in the western portion of the field are significantly older than clusters in the eastern portion of the field (p value of <0.001), and there is a tendency for cluster age to decrease to the east. This is particularly evident when mean cluster ages are calculated for tholeiite, alkaline olivine basalt, and evolved alkaline rock types independently. Application of the two-point azimuth and Hough transform methods demonstrates that regional cinder cone alignments transect these clusters. The most prominent of these alignments trend ENE in the eastern portion of the field and WNW in the western portion of the field, creating an overall arcuate pattern that is subparallel to the trend of the Mogollon Rim and the Colorado Plateau/Transition Zone boundary. These observations suggest that vents (and clusters) migrated from west to east in response to plate motion, but the general pattern of vent migration was complicated by regional structures, which enhanced the volume and duration of magmatism in some areas. The fractures or faults implied by vent alignments indicate that Shmin is oriented radial to the Colorado Plateau in the SVF. Preferred vent alignment orientations may be related to extension resulting from plateau uplift, and to a much smaller degree from a minor Basin and Range imprint. While regional in extent, the implied structures appear to differ significantly from some of those in several other plateau-marginal fields in that they cannot be related to major reactivated Precambrian structures. Our vent alignment data differ from those seen by other workers in the Zuni-Bandera and Mount Taylor fields, suggesting the stress field for the SVF is different from other fields in the proposed Jemez lineament. The stress field implied by vent alignment data, combined with structural data, suggests that the southwestern tectonic boundary of the Colorado Plateau of Brumbaugh (1987) should be extended southeastward to include the SVF at the plateaus southern boundary.

Small domes on Venus: Characteristics and origin

Approximately 22,000 small domes have been identified on the 25% of the surface of Venus imaged by Venera 15/16. The word dome is used to imply a broad, lens-shaped, positive topographic feature. The domes: (1) are generally circular in planimetric outline; (2) range in diameter from the effective limit of Venera resolution (⩾2 km) to 20 km; (3) show flank slopes generally ⩽10 ° and possibly ⩽5 °; and (4) occur in association with mottled plains units. Associated features include summit pits, radar bright surfaces, and basal topographic platforms. There are two significant areas of major dome concentrations approximately 180 ° in longitude apart: (1) the largest concentration occurs in the Akkruva Colles area of Niobe Planitia, centered at approximately 45 ° N/120 ° E, just north of the flanks of the Thetis Regio rise; and (2) another concentration occurs in northwestern Guinevere Planitia, centered at approximately 35 ° N/300 ° E, on the north flank of the Beta Regio rise. In addition to these major areas of concentrations, domes occur in smaller concenrations throughout the imaged area of Venus, in association with coronae, arachnoids, intermediate sized hills interpreted to be volcanic constructs, large volcanic centers and calderas. The characteristics and geologic associations of small domes are consistent with an interpretation of their origin as volcanic, and on the basis of their low slopes, individual characteristics, and geologic associations they are interpreted to represent dominantly effusive low shield volcanoes. The large number of small domes implies a large number of multiple centralized eruptions, each one of which represents a discrete, relatively small, volume of material available to build an edifice over a finite time period. Calculated modal volume is 0.73 km3 for individual edifices. Based on the number identified by Venera, the total number of small domes estimated for the entire planet 4.4 × 106 and total edifice volume over the entire planet represents a minimum volume equivalent to a layer approximately 7 m thick over the planet and representing 0.03% of the estimated crustal volume of Venus. In absolute number, size range, and distribution they appear to be similar to terrestrial oceanic seamounts. The global abundance and distribution, size frequency distribution, minimum size, and changes in these characteristics with latitude for the domes will be particularly important in understanding the way in which the domes form and their relationship to global models of tectonism and heat flow on Venus. Increased spatial resolution and coverage from Magellan data will enable a more thorough assessment of these features and associated questions, particularly where radar incidence angles are ⩽15 °.

Relation of Major Volcanic Center Concentration on Venus to Global Tectonic Patterns

Global analysis of NASA Magellan image data indicates that a major concentration of volcanic centers covering ∼40 percent of the surface of Venus occurs between the Beta, Atla, and Themis regiones. Associated with this enhanced concentration are geological characteristics commonly interpreted as rifting and mantle upwelling. Interconnected low plains in an annulus around this concentration are characterized by crustal shortening and infrequent volcanic centers that may represent sites of mantle return flow and net down-welling. Together, these observations suggest the existence of relatively simple, largescale patterns of mantle circulation similar to those associated with concentrations of intraplate volcanism on Earth.

Calderas on Mars: characteristics, structure, and associated flank deformation

Abstract Calderas and flank structures of martian volcanoes yield insight into general questions of volcano structural evolution and the underlying magma chambers in an environment where erosion is minimal. We have documented, through detailed geological mapping, the structures, associated volcanological features, and the stratigraphical relationships between the flank structures and caldera events during the building of each martian edifice. Two fundamentally different types of calderas are identified on Mars (the Olympus type and the Arsia type) that may represent end member variations in the size and depth of magma chambers. Many of the flank structures adjacent to caldera rims are consistent with the predicted effects of magma chamber inflation as well as deflation that exert significant influences in the structural development of many volcanoes. Large-scale terracing and steepening of the upper flanks of the larger martian volcanoes may originate from magma chamber inflation and radial thrusting. Thus the endogenous component of volcano growth resulting from accumulated magma chamber growth may be significant. Many of the deepest calderas are associated with evidence for voluminous eruptions elsewhere on the flanks and along through-going fissures and appear to result largely from evacuation and deflation of magma chambers without extensive precursor inflation. Draining of the magma chamber in these cases may be aided by the lateral propagation of magma in the form of shallow dykes up to several hundred kilometres in length and the associated formation of linear fissures. Nested caldera sequences, related flank pits, large-scale slumping, terracing, and sector structure are frequently arranged in linear patterns and are part of through-going eruptive lines or fissures several hundred kilometres in length that characterize several martian shield volcanoes. Fissures this long are interpreted to be dykes propagated outward from shallow magma chambers that have followed a minimum regional stress orientation. Comparison of the observed shape and orientation of caldera structures with orientation and style of flank deformation, and with the predictions from theory, indicate that regional stresses have probably been an important influence on the caldera and flank structures of martian volcanoes. The minimum regional stress orientation may be controlled largely by regional slopes associated with the Tharsis region and Elysium regions, and, in the case of Tyrrhena Patera, pre-existing radial fractures associated with the Hellas basin.

The 1999 Marsokhod rover mission simulation at Silver Lake, California: Mission overview, data sets, and summary of results

We report on a field experiment held near Silver Lake playa in the Mojave Desert in February 1999 with the Marsokhod rover. The payload (Descent Imager, PanCam, Mini-TES, and Robotic Arm Camera), data volumes, and data transmission/receipt windows simulated those planned for the Mars Surveyor mission selected for 2001. A central mast with a pan and tilt platform at 150 cm height carried a high-resolution color stereo imager to simulate the PanCam and a visible/near-infrared fiberoptic spectrometer (operating range 0.35–2.5 μm). Monochrome stereo navigation cameras were mounted on the mast and the front and rear of the rover near the wheels. A field portable infrared spectroradiometer (operating range 8–14 μm) simulated the Mini-TES. A Robotic Arm Camera, capable of close-up color imaging at 23 μm/pixel resolution, was used in conjunction with the excavation of a trench into the subsurface. The science team was also provided with simulated images from the Mars Descent Imager and orbital panchromatic and multispectral imaging of the site obtained with the French SPOT, airborne Thermal Infrared Mapping Spectrometer, and Landsat Thematic Mapper instruments. Commands sequences were programmed and sent daily to the rover, and data returned were limited to 40 Mbits per communication cycle. During the simulated mission, 12 commands were uplinked to the rover, it traversed ∼90 m, six sites were analyzed, 11 samples were collected for laboratory analysis, and over 5 Gbits of data were collected. Twenty-two scientists, unfamiliar with the location of the field site, participated in the science mission from a variety of locations, accessing data via the World Wide Web. Remote science interpretations were compared with ground truth from the field and laboratory analysis of collected samples. Using this payload and mission approach, the science team synergistically interpreted orbital imaging and infrared spectroscopy, descent imaging, rover-based imaging, infrared spectroscopy, and microscopic imaging to deduce a consistent and largely correct interpretation of the geology, mineralogy, stratigraphy, and exobiology of the site. Use of imaging combined with infrared spectroscopy allowed source outcrops to be identified for local rocks on an alluvial fan. Different lithologies were distinguished both near the rover and at distances of hundreds of meters or more. Subtle differences such as a contact between dolomite and calcite were identified at a distance of 0.5 km. A biomarker for endolithic microbiota, a plausible life form to be found on Mars, was successfully identified. Microscopic imaging of soils extracted from the surface and subsurface allowed the mineralogy and fluvial history of the trench site to be deduced. The scientific productivity of this simulation shows that this payload and mission approach has high science value and would contribute substantially to achieving the goals of Mars exploration.

Geology of central Chryse Planitia and the Viking 1 landing site: Implications for the Mars Pathfinder mission

1:500,000-scale geologic mapping in the central Chryse Planitia region of Mars was correlated with “ground-truth” data gathered by the Viking 1 lander. Materials within the Chryse basin can be subdivided into plains and channel units that are typically separated from one another by gradational contacts. Hesperian Ridged plains materials, unit 1 (Hr1) are the oldest materials mapped. Typically, these materials contain numerous fresh impact craters and have sharply defined, mare-like wrinkle ridges similar to those appearing on the lunar maria. These materials grade into Hesperian Ridged plains materials, unit 2 (Hr2), which are characterized by buried and eroded impact craters and subdued wrinkle ridges. From analyses of crater age dates and their associated geologic contacts, channel materials appear to have formed at the same time as Hr2 materials, and it is likely both units represent fluvial sediments. Measurements of buried craters contained in Hr2 materials suggest that in places this unit may be ∼50 m thick, but crater size-frequency distribution curves suggest that the areal average may be closer to ∼170 m. Based on these observations, our interpretation is that Hr2 materials were deposited into a standing body of water during channel formation. This interpretation implies that many of the rocks visible in the Viking 1 lander images were emplaced by fluvial processes. Possibly, finer-grained sediments remained in suspension and were subsequently transported out of Chryse Planitia and into the northern plains during draining of the ponded water. East-west trending surface undulations, visible in lander views toward the south, may represent aeolian dunes, lava flow fronts, or sediment waves formed at the bottom of the standing body of water. Broad physiographic units seen at the surface are not clearly visible in Viking orbiter images; however, they can be projected at the resolution of the orbiter images. These units show that concentrations of drift materials are oriented in a northwesterly direction, contrary to the strongest prevailing wind direction which is toward the northeast. These materials were probably deposited on Ridged plains materials, unit 2, during a period of time when aeolian processes were more active in the region. Both Earth-based radar and Viking thermal data suggest that the Mars Pathfinder landing site will be similar geologically to the Viking 1 site. If this is true, then the Mars Pathfinder mission provides the opportunity for building directly on results of the Viking program. Some of the outstanding questions that Mars Pathfinder may be able to address include determining the aeolian modification history of the Chryse Planitia region, the degree and possibly the relative rate of sediment induration, the fraction of rocks and boulders emplaced by impact processes, the possibility that some materials are the result of in situ weathering, and whether materials were emplaced by fluvial processes and the associated depositional environment.