The boundaries of large-scale collapse on the flanks of Mount Etna, Sicily

Proposals for flank collapse on Etna find agreement from recent authors but models differ considerably in the scale and nature of the instability. Our work, including a new compilation of subsurface data, identifies the Ragalna system of linked faults and cinder cones as defining the southern margin of a very wide sector of flank instability comprising over 50% of the volcanic edifice. The Mascalucia and Trecastagni faults, previously identified as the southern boundary, occur within this sector and are now regarded as two examples of a number of relatively superficial faults accommodating differential movement. By comparison, faulting on the Ragalna system has produced very fresh-appearing scarps up to 20m in height associated with dominantly dip-slip displacement. This pattern is expected on a boundary fault, the unstable side of which is moving downslope. The problem of the enigmatic triangular horst which abruptly terminates the surface expression of the Ragalna system at its upslope end may now be viewed in the light of new scale-model experiments which predict that such features are the likely result of spreading within a volcanic edifice. This helps explain the lack of a clear surface connection between the obviously active Ragalna faulting and the summit rifting. If this explanation is correct the Ragalna system assumes similar continuity and dimensions as the Pernicana - a system of linked structures generally accepted as defining the northern boundary of instability. The scale of the collapse is well within that predicted by the model experiments. Topographic buttressing on the northwestern arc of the volcano is mirrored by the absence of such support on the unstable southeastern arc where the flanks are bordered either by the sea or by a sedimentary trough (Gela-Catania Foredeep) occupied by recent plastic clays close to sea level. In addition, the rapidly growing volcanic edifice, approaching 3.5 km in some 300 ka, straddles the thrusted nappes of the Appennine-Maghrebian Chain to the northwest and the weak sediments of the Gela-Catania Foredeep to the southeast, with their contact beneath the volcano probably contributing to the location of the unstable sector. Minimum long-term slip rates determined from outcrop information on the Ragalna, and determined independently from subsurface data on the basal detachment underlying the collapsing sector are similar, arguing for a mechanical connection as would be expected if the Ragalna system formed the boundary to the basal detachment. A deep detachment model is consistent with focal depths on the Ragalna and with recent data on the timing of regional deformation in the Chain compared to the later timing of local (collapse-induced) deformation within the Etnean realm. The amplitude and wavelength of the local deformation, particularly on the southern margin of the volcano, is inconsistent in magnitude with that expected mechanically if the detachment was relatively superficial. A deep detachment is also indicated by the space requirements of a plutonic sequence postulated on geological and geophysical grounds below the summit region of the volcano.