The distribution and age of early volcanic and tectonic features surrounding the Hellas and Isidis impact basins are shown to fit a four‐stage sequence. First, concentric “canyons” form outside the basin boundary scarp at or near the time of basin formation, followed shortly thereafter by radial troughs extending beyond the boundary scarp. After a hiatus, concentric graben develop inside the basin within the massif ring. Finally, soon after graben formation, widespread volcanic plains are emplaced between the massif ring and the boundary scarp to one side of the basin. Three models of basin‐centered deformation are compared to these observed deformation events: elastic flexure due to loading, elastic flexure due to uplift, and impact fracture. A central load of basin fill can account for concentric fractures by elastic flexure around the load. Formation of the distant concentric canyons, however, requires lithospheric thicknesses a factor of 5 greater than those indicated by planetary thermal history models and an increase in feature width by an order of magnitude relative to flexural graben identified elsewhere on Mars and the Moon. Although inclusion of loading by basin ejecta reduces the required lithospheric thickness to acceptable values, flexure still fails to account for the large canyon widths observed. Impact fracture from a rapid inward flow of asthenosphere during transient cavity collapse, however, is consistent with the three key observational constraints for the distant canyons: their age, width, and distance from the basin. As the basin then undergoes isostatic uplift soon after impact, elastic flexure induces a radial pattern of failure consistent with the observed extent and timing of the radial troughs. Later, after a long period of basin infilling, elastic failure due to loading of the central basin region can account for the ring grabens in the massif ring. The localization of volcanism into a single ridged plains unit on the basin rim indicates a thermal source offset from the basin which is not included in these models and perhaps reflects an additional mechanism unique to Mars.

Endogenic modification in lunar floor‐fractured craters can constrain spatial variations in early lunar conditions. The nature of these constraints, however, depends on the assumed mechanism of crater modification. For viscous relaxation, the extent of crater modification depends on the surrounding crustal viscosity and thus provides loose constraints on the history of crustal heating within a region. For igneous intrusion models, the extent of crater modification reflects magmatically driven deformation and can be inverted to estimate both local magma pressure and intrusion depth. Both models indicate clear differences between regional conditions at Mare Smythii and in the highlands west of Oceanus Procellarum. The uniformly shallow crater depths in Mare Smythii probably indicate a long‐lived period of extreme crater relaxation, whereas the wide range of modified crater depths in the western highlands suggest a much shorter period of partial crater relaxation. For comparable relaxation times, the average viscosity derived for Mare Smythii is over 10 times lower than the average viscosity inferred for the western highlands. Alternatively, if modification reflects deformation over crater‐centered, laccolithlike intrusions, the derived magma pressures indicate a broad, uniform magma source beneath Mare Smythii, whereas the spatial variation of estimated magma pressures in the western highlands suggests the presence of several, smaller magma sources. The derived intrusion depths are partly a function of crater size, but range from ∼1 to 10 km in depth for both regions and may be slightly greater on average in the western highlands. While both viscous relaxation and igneous intrusions can explain the modification of individual craters, the regional variations in these derived modification conditions also allow further testing of each modification model. In particular, the correlation of the lowest viscosities or the longest relaxation times with the smallest craters in both Mare Smythii and the western highlands seems inconsistent with crater modification by the relaxation mechanism. The onset of total relaxation at progressively smaller topographic wavelengths over time might produce such a trend in Mare Smythii, but total relaxation cannot be invoked for the western highlands, where many large craters still preserve significant fractions of their initial relief. In addition, the relaxation of the smallest craters in both regions suggests the presence of exceptionally high near‐surface thermal gradients (∼150–200 K/km). Since the observed regional variations in crater modification can be easily attributed to variations in magma pressure as a function of mantle topography, we conclude that crater floor fracturing in both regions is more consistent with the igneous intrusion mechanism than with viscous relaxation during a crustal heating event.

Abstract. Bright-floored and dark-floored craters on Venus show systematic differences in their size, distribution and apparent modification. Bright-floored craters exhibit the following characteristics: an interior radar brightness comparable to the youngest craters on Venus, a tendency toward smaller crater diameters, and a broad range of crater elevations. In contrast, the dark-floored craters are darker than pristine craters on Venus, are typically much larger, and preferentially occur at lower elevations. They also have larger floors than pristine craters of the same size and are similar in many respects to floor-fractured craters on Venus. Of the four proposed origins for dark crater floors, these observations are most consistent with crater-centered volcanism. Surface weathering or eolian deposition may contribute to floor darkening in some cases, but neither of these mechanisms, nor impact melts, can independently explain the full range of observed modifications.

Regional tectonism and volcanism affect crater modification and crater loss on Venus, but a comparison of Venusian craters to lunar floor‐fractured craters suggests that a third style of more localized, crater‐controlled magmatism also may occur on Venus. Based on lunar models for such magmatism, Venusian crustal conditions should generally favor crater‐filling volcanism over crater‐centered floor fracturing. Nevertheless, three craters on Venus strongly resemble extensively modified craters on the Moon where deformation can be attributed to failure over large crater‐centered intrusions. Models for crater modification over igneous intrusions indicate typical magmatic pressures beneath these three craters of ∼200–300 bars and intrusion depths of the order of 1–6 km. All three craters also share common settings and low elevations, whereas craters embayed by regional volcanism preferentially occur at much higher elevations on Venus. We suggest that the style of igneous crater modification on Venus thus may be elevation dependent, with crater‐centered intrusions primarily occurring at low elevations on Venus. This interpretation is consistent with theoretically predicted variations in magmatic neutral buoyancy depth as a function of atmospheric pressure suggested by other authors.

The pattern of radial and concentric offset dikes at Sudbury strongly resembles fracture patterns in certain volcanically modified craters on the Moon. Since the Sudbury dikes apparently formed shortly after the impact event, this resemblance suggests that early endogenic modification at Sudbury was comparable to deformation in lunar floor‐fractured craters. Although regional deformation has obscured many details of the Sudbury Structure, such a comparison of Sudbury with lunar floor‐fractured craters provides two alternative models for the original size and surface structures of the Sudbury basin. First, the Sudbury date pattern can be correlated with fractures in the central peak crater Haldane (36 km in diameter). This comparison indicates an initial Sudbury diameter of between 100 and 140 km but requires loss of a central peak complex for which there is little evidence. Alternatively, comparison of the Sudbury dikes with fractures in the two‐ring basin Schrödinger indicates an initial Sudbury diameter of at least ∼ 180 km, which is in agreement with other recent estimates for the size of the Sudbury Structure. In addition to constraining the size and structure of the original Sudbury crater, these comparisons also suggest that crater modification may reflect different deformation mechanisms at different sizes. Most lunar floor‐fractured craters are attributed to deformation over a shallow, crater‐centered intrusion; however, there is no evidence for such an intrusion at Sudbury. Instead, melts from the evolving impact melt sheet probably entered fractures formed by isostatically‐induced flexure of the crater floor. Since most of the lunar floor‐fractured craters are too small (<100‐km diameter) to induce significant isostatic adjustment, crater modification by isostatic uplift apparently is limited to only the largest of craters, whereas deformation over igneous intrusions dominates the modification of smaller craters.