At the scale range of 1 to 10 km, faults are not continuous surfaces but are tabular bodies composed of a mesh of subparallel unconnected strands, or subfaults. The cumulative length distribution of these subfaults is observed to be a power law with an approximate exponent of −2 and an upper fractal limit at W*, the seismogenic width. If it is assumed that small earthquakes, with lengths LW*, represent the rupture of these subfaults, this offers a physical explanation for small earthquakes having their observed power-law distribution with exponent −2/3. This further implies that any earthquake will be composed of a population of subevents involving the rupture of this population of subfaults: precisely what was proposed by Frankel (1991) to explain the ω−2 high-frequency falloff of earthquake displacement spectra.

First Page Preview

First page PDF preview
You do not currently have access to this article.