A geological approach to seismicity b-values: Implications for hazard assessment

Assessing the hazard associated with naturally-occurring or induced earthquakes requires a prediction of the likelihood and size of the largest earthquakes that might occur in a region. The size-frequency distribution of seismic events, known as the b-value, can be used to estimate how often earthquakes of a particular magnitude will occur. The global average b-value is 1.0, but it can be larger or smaller if earthquake populations are segregated and analyzed by time, depth, location, or focal mechanism. The fact that this variation is observed has led workers to seek a causal relationship to explain different b-values, suggesting that bvalues are constrained by conditions related to earth stress, hypocenter depth, or fluid pumping rate (in the case of induced earthquakes), for example. Rock heterogeneity is also considered to influence the size-frequency distribution of earthquake failure planes. Rock heterogeneity is often defined by the existing fault geometry, and the aim of this study was to test the hypothesis that the fault size-frequency distribution of an area directly correlates with the seismicity b-value of that area. Two study areas were chosen; natural earthquakes in southern California, and comparison of natural and induced earthquakes in Oklahoma. The results indicate in the study areas there is a direct relationship between fault size-frequency and earthquake size-frequency, and that mapping faults on the surface and in the subsurface could be used to estimate the b-value for an area, thereby providing a method to predict damaging earthquake size and likelihood even in areas which were not previously seismically active.