Abstract Of the features that characterize large shield volcanoes on Mars, flank terraces remain the most enigmatic. Several competing mechanisms have been proposed for these laterally expansive, topographically subtle landforms. Here we test the hypothesis that horizontal contraction of a volcano in response to the down-flexing of its underlying basement leads to flank terracing. We performed a series of analogue models consisting of a conical sand–plaster load emplaced on a basement comprising a layer of brittle sand–plaster atop a reservoir of viscoelastic silicone. Our experiments consistently produced a suite of structures that included a zone of concentric extension distal to the conical load, a flexural trough adjacent to the load base and convexities (terraces) on the cone’s flanks. The effects of variations in the thickness of the brittle basal layer, as well as in the volume, slope and planform eccentricity of the cone, were also investigated. For a given cone geometry, we find that terrace formation is enhanced as the brittle basement thickness decreases, but that a sufficiently thick brittle layer can enhance the basement’s resistance to loading such that terracing of the cone is reduced or even inhibited altogether. For a given brittle basement thickness, terracing is reduced with decreasing cone slope and/or volume. Our experimental results compare well morphologically to observations of terraced edifices on Mars, and so provide a framework with which to understand the developmental history of large shield volcanoes on the Red Planet.

The definition of a volcano is discussed, and a new encompassing version is provided. The discussion focuses on the observations that volcanism is a self-similar process that ranges many orders of magnitude in space and time scales, and that all kinds of geologic processes act on volcanoes. Former definitions of volcano , such as that from the Glossary of Geology (1997, p. 690)—“a vent in the surface of the Earth through which magma and associated gases and ash erupt” or “the form or structure, usually conical, that is produced by the ejected material” are clearly insufficient. All definitions that we encountered tend to consider volcanoes from the point of view of a single discipline, each of them neglecting relevant aspects belonging to other disciplines. For the two cases mentioned above a volcano is seen only from the point of view of eruptive activity or of morphology. We attempt to look at volcano holistically to provide a more comprehensive definition. We define a volcano as a geologic environment that, at any scale, is characterized by three elements: magma, eruption, and edifice. It is sufficient that only one of these elements is proven, as long as the others can be inferred to exist, to have existed, or to have the potential to exist in the future.

Edifice-collapse phenomena have, to date, received relatively little attention in Central America, although ∼40 major collapse events (≥0.1 km 3 ) from about two dozen volcanoes are known or inferred in this volcanic arc. Volcanoes subjected to gravitational failure are concentrated at the arc's western and eastern ends. Failures correlate positively with volcano elevation, substrate elevation, edifice height, volcano volume, and crustal thickness and inversely with slab descent angle. Collapse orientations are strongly influenced by the direction of slope of the underlying basement, and hence are predominately perpendicular to the arc (preferentially to the south) at its extremities and display more variable failure directions in the center of the arc. The frequency of collapse events in Central America is poorly constrained because of the lack of precise dating of deposits, but a collapse interval of ∼1000–2000 yr has been estimated during the Holocene. These high-impact events fortunately occur at low frequency, but the proximity of many Central American volcanoes to highly populated regions, including some of the region's largest cities, requires evaluation of their hazards. The primary risks are from extremely mobile debris avalanches and associated lahars, which in Central America have impacted now-populated areas up to ∼50 km from a source volcano. Lower probability risks associated with volcanic edifice collapse derive from laterally directed explosions and tsunamis. The principal hazards of the latter here result from potential impact of debris avalanches into natural or man-made lakes. Much work remains on identifying and describing debris-avalanche deposits in Central America. The identification of potential collapse sites and assessing and monitoring the stability of intact volcanoes is a major challenge for the next decade.