Earthquakes, elevations and continental plateaux: An investigation into the absence of large thrust earthquakes at high elevations within fold-and-thrust belts.

Large thrust earthquakes (M ≥ 5) are rare at high elevations in continental fold-and-thrust belts. For example, Nissen et al’s (2011) work infers a cut-off in M ≥ 5 earthquakes at ~ 1250 m in the Zagros. Data from the Zagros, Himalayas, Qilian Shan and Longmen Shan fold-and-thrust belts are analysed to test three hypotheses for this cut-off phenomenon. (1) Fold-and-thrust belts deform via the critical Coulomb wedge model, which favours thrusting at lower elevations. (2) Fold-and-thrust belts are controlled by minimum work processes, which suppress thrusts at high elevations. (3) A thickened ductile zone within the sedimentary layer of fold-and-thrust belts inhibits earthquake propagation. Earthquake magnitude vs. elevation analysis revealed an abrupt cut-off in M ≥ 5 earthquakes at threshold elevations which were dependant on the locality. These elevations ranged from 1250 m in the Zagros to 4000 m in the Qilian Shan. Cut-offs were then compared to their corresponding slope geometries determined by swath profile analysis. Two significant changes in gradient were identified; one at the edge of the plateau and one correlating with the elevation of the earthquake cut-offs, where the slope gradient decreased but remained positive. Brittle-ductile transitions, first in the basal decollement and then in the lower wedge, are considered to account for such major variations in the slope geometry (Hypothesis 1). The abrupt cut-off in M ≤ 5 earthquakes at the second gradient change supports the theory that the fold-and-thrust belts have a limiting elevation, defined by Gravitational Potential Energy (Hypothesis 2). For this reason, a hybrid of Hypotheses 1 and 2 is proposed to explain the reduction in M ≥ 5 earthquakes with elevation, in which ductile deformation in the critical Coulomb wedge limits M ≥ 5 earthquakes, but convergence continues aseismically until the formation of a plateau with its height determined by minimum work processes. To explore the relevance of Hypothesis 3, dual lithology triaxial loading experiments were conducted as an analogue to study the influence of brittle-ductile behaviour on fault propagation. Brittle faults were found to propagate aseismically into a ductile medium offering an explanation for the high elevation regions of aseismic convergence.