Triggering Earthquakes

Why do earthquakes happen? At one level, this age-old question was solved by the plate tectonics revolution in the 1960's. It was found that earthquakes accommodate motion as large, nearly rigid plates slide past each other. Yet, this broad explanation for slip leaves many of the key questions unanswered. Why is slip sometimes accommodated by gradual creeping and at other times by rapid failure? Why do some earthquakes stop after only a few meters of rupture while others continue for 1000 km? How do earthquakes interact? Most importantly, what is the trigger for slip?

We are trying to unravel why earthquakes happen by examining long-range triggered seismicity. Long-range triggering is one of the few natural situations where the immediate cause of an earthquake is apparent and easily isolated. Therefore, it provides a fundamental clue to the earthquake initiation process. The large distances involved suggest that seismic waves, rather than static stress changes are the trigger. The fact that the triggered sites are often geothermal implies that fluids may be an important part of the triggering process.

We are currently taking two different routes into this problem. One strategy involves investigating the effects of seismic waves on hydrological systems through direct observations of phenomena like coseismic water steps. These observations have motivated a new physical mechanisms for triggering such as redistribution of pore pressure by flushing fractures (Brodsky et al., 2003; Elkhoury et al., 2006). Work currently in progress along this line includes measuring the response of water-filled fractured rocks to seismic waves in the lab. The second line of attack is to observationally measure thresholds for triggering. We have discovered that frequency is an important characteristic of triggering waves in addition to the waves' amplitudes (Brodsky and Prejean, 2005) . The amplitude and frequency of the triggering strains pose strong constraints on the physical models. We have also found the distribution of aftershocks is consistent with their being generated by seismic waves (Felzer and Brodsky, 2006).