35 - Strength and energetics of active fault zones
The strength of active fault zones, i.e., the shear stress level required to cause fault slip, is fundamental to understanding the physics of earthquakes and to assessing earthquake hazard. Although many researchers have concluded that fault zones are weak (shear stresses 10 MPa or less averaged between 0 and ˜20 km depth), others maintain that faults are strong (˜100 MPa average of an approximate linear increase with depth). Thus, despite 30 y of dedicated research, relevant data remain inconclusive and fault strength remains uncertain by an order of magnitude. In part, this is because the main source of energy release in earthquakes is at depths greater than 5 km, inaccessible to direct instrumental observation. Very large earthquakes rupture to the Earth's surface where direct observation of the shallow rupture process is possible. However, the rupture characteristics at shallow depth may differ from those at seismogenic depths. To date no great ( M ≥ 7.75) earthquakes have occurred within a network of modern strong-motion instruments, but the large, well-recorded 1999 earthquakes in Turkey and Taiwan, both M = 7.6, show that this data gap is rapidly being filled. Furthermore, shear stress changes at the earthquake source (and the resulting seismic waves) are nearly linear perturbations of the absolute stress field. Thus, an unknown absolute background stress does not greatly affect the basic characteristics of the observed low frequency seismic waves and the observed geodetic deformation. Other, less-direct data must then be used to infer the physical state and ambient stress levels on active faults. This chapter discusses available evidence and current ideas about fault zone strength and energetics. In our review we begin by outlining the general physical conditions prevailing in the Earth around active faults and summarize the generally agreed upon observational features of faulting and earthquake occurrence ( Section 2 ). In doing so we make as few assumptions as possible, because making uncertain assumptions may lead to logical inconsistencies and apparent paradoxes. For example, the absence of a measurable, frictionally generated heat flow anomaly near active faults, the so-called stress-heat flow paradox, rests upon assumptions that must somehow be incorrect. We wish to avoid such inconsistencies and to begin we introduce the observations and briefly state what they imply about fault zones. Subsequently, in Sections 3–8 , we place these observations in an interpretative context and show how they have been used to infer various measures of fault shear stress. Section 9 summarizes our assessment of the average shear stress state near faults, argues for the general importance of stress heterogeneity in faulting processes, and discusses its implications.