Magellan radar image data of Sapas Mons, a 600 km diameter volcano located on the flanks of the Atla Rise, permit the distinction of widespread volcanic units on the basis of radar properties, morphology, and spatial and inferred temporal relations, each representing a stage or phase in the evolution of the volcano. Six flow units were identified and are arranged asymmetrically about the volcano. Although there is some evidence for overlapping of units, the stratigraphy clearly indicates a younging upwards sequence. The estimated volume of this 2.4 km high volcano is 3.1 x 10(4) km3, which is comparable to the largest Hawaiian shield (Mauna Loa, 4.25 x 10(4) km3), but it is significantly less than an estimated volume for the entire Hawaiian-Emperor chain (1.08 x 10(6) km3) and less than the lower diameter (100 x 150 km) island of Hawaii (11.3 x 10(4) km3). Although it is difficult to clearly identify a single lava flow, estimates of apparent single flow volumes range from 4 km3 (for an average unit 5 flow of 3.4 km width, 10 m thickness, and 121 km length) to almost 59 km3 (for a 17.8 km wide, 10 m thick, 330 km long unit 1 flow). Estimates of total volumes for the units show that four of the six flow units have volumes that are within a factor of 1.2 of each other, one unit is approximately three times more voluminous, and the latest unit has a very small volume. Flows within a given unit are very distinct relative to flows in other units with respect to average lengths, aspect ratio, radar brightness, and planimetric outline. There is a weak distinction in rms slope between units and emissivity is correlated with altitude, not unit boundaries. A pair of 25 km diameter scalloped-margin domes occur at the summit and are the source of the last stage of eruptions on Sapas; steep fronts and high aspect ratios suggest that associated flows may have had a high viscosity. Graben form a circumferential structure 75-100 km in diameter surrounding the summit domes and are interpreted to be indicative of subsidence over a central magma reservoir. Radial fractures with associated small edifices cut the lower flanks of the edifice but are not observed within the summit ring of graben; these are interpreted to be the expression of near-surface dykes and may have been emplaced during a period of enhanced activity but correlates with the most voluminous flow unit. Unlike at Hawaii, however, these dykes and small edifices do not seem to be the source of significant flank eruptions. Although some effusive activity may have accompanied their emplacement, the majority of lava flows at Sapas appear to be radial to a single, near-summit point located between the two summit domes. Calculated effusion rates range from 1.5 x 10(3) m3/s to 3.1 x 10(5) m3/s; these values suggest that rates were high compared with the Earth and decreased with time. These rates, and the volumes calculated, give eruption durations for the various units that range from 18 days to over 20 years. If eruption is caused by the influx of magma from depth and rupture of an overpressurized chamber, this suggests a variable flux over the history of the volcano. The late-stage eruptions which formed the summit domes are interpreted to be the result of fractional crystallization and/or volatile build-up in the chamber, following a period of decreased supply from depth. Local topography and gravity, as well as regional geology support the presence of a mantle plume at Sapas. The similar properties of large volumes of magma over the total history of the volcano, as well as the prolonged period of magma supply and gradual waning, are consistent with a plume origin. These inferences and the observations allow us to characterise the history of the volcano as follows: arrival of the mantle plume caused uplift of topography and surrounding plains formation: continued supply of smaller volumes of material permitted construction of the edifice; development of a magma reservoir (predicted by theory to form at shallow depths) modified eruption characteristics by permitting storage and homogenization of magma; unbuffered conditions prevailed for the majority of eruptions, producing flows of similar volumes but decreasing flow lengths; a period early on of enhanced supply led to buffered chamber conditions, resulting in the eruption of the voluminous flow unit and the emplacement of many lateral dykes; evacuations from the chamber and cooling towards the last stages caused distributed summit collapse and formation of the ring graben; and finally the gradual waning of supply allowed evolution of the magma which produced the late-stage, possibly viscous flows and dome construction. Preliminary observation of Sapas and two other volcanoes at different elevations suggests that altitude-dependent chamber development and growth may influence the complexity of lava flows and determine the existence of collapse calderas. Many features at Sapas are representative of large volcanoes on Venus and thus Sapas Mons is a good example of a typical plume-associated edifice. Sapas differs in many ways from Kilauea, a terrestrial type shield volcano, but these differences can be understood in the context of the Venus environment.
