EARTHQUAKES AND ELASTIC REBOUND THEORY   

A.  INTRODUCTION

        Earthquakes occur along breaks within the earth's crust known as
        faults.  Most earthquakes occur along pre-existing faults, but a new
        fault can be created during an earthquake. 

        Two terms are used to describe the point of origin for an earthquake:

        1.  FOCUS

             This is the actual location where fault movement begins.  Almost all
             of the time, the focus of an earthquake is below the surface.
  

        2.  EPICENTER

             This is the point on the land surface directly above the focus and is
             the location normally reported in the news or shown on maps.  Note
             that an earthquake epicenter need not be located on a fault line.

B.  OBSERVATIONS ABOUT EARTHQUAKES

        1.  WHERE EARTHQUAKES OCCUR 

              Recent U.S. seismicity

              Large historic earthquakes (U.S.) 

             

              a.  Plate Tectonics (boundaries)

                   Most earthquakes (esp. large ones) occur at plate boundaries,
                   but there are many exceptions (e.g.,  Missouri and Hawaii).   

              b.  Volcanoes 

              c.  Post-glacial Rebound

              d.  Meteorite Impact Craters

              e.  Large Water Reservoirs

                   Example:  the Hoover Dam and Lake Mead

              f.  Nuclear Weapons Test Sites

                   Example:  Nevada Test Site:  legacy

        2.  MAGNITUDE AND FREQUENCY

             In general, the larger the magnitude of an earthquake, the larger its 
             recurrence interval  and the greater the amount of fault movement.  
             Very large earthquakes can cause sizeable offset (e.g., the 1906 San
             Francisco earthquake, which had a magnitude of more than 8.0 and
             has a recurrence interval of 300+ years, caused more than 20 feet of 
             of displacement along the San Andreas Fault). 

             At the other end of the strain release spectrum is fault creep , a slow
             but continuous movement that can be detected only by seismometers
             or by the damage it causes (see example from Hayward Fault).

             Note that large faults (e.g., the San Andreas) do not move all at once.
             Instead, only certain segments move during an earthquake.  Segments
             of active faults that haven't moved recently are called seismic gaps.  

        3.  PRECURSORY EVENTS

             a.  Fault Zone "Bulges"

                  Rises in land surface elevation observed along some active faults
                  has been attributed to the development of very small cracks in
                  rocks under stress that cause an increase in volume (dilatancy). 

             b.  Earthquake Lights

                  Although not common, there have been cases where "luminosities
                  in the sky" were reported prior to or during an earthquake.   These
                  occur when certain rock types, under stress, discharge electricity,
                  ionized gases, or natural gas.  

  
             c.  Foreshocks

                  Foreshocks often precede large earthquakes, which suggests that
                  stress is starting to overcome the resistance to fault movement.  

             d.  Animal Behavior

                  Anecdotal evidence suggests that animals can sense very small
                  foreshocks that precede large earthquakes.  The Chinese have 
                  used animal behavior to help predict earthquakes.

C.  ELASTIC REBOUND THEORY

        1.  HYPOTHESIS

             Based on the observations listed above, scientists hypothesize that
             rocks under stress deform elastically (analogous to a rubber band).
             Strain builds up until either: (1) the rocks break (creating a new fault),
             or (2) movement occurs on an existing fault.  

             Earthquakes occur as stored strain is released, and rocks"rebound" 
             to their undeformed shape.  The more strain released, the larger the 
             magnitude of the earthquake.  Residual strain that is not released in
             the initial earthquake may be released in smaller aftershocks.

        2.  TESTING THE MODEL

             Does this hypothesis allow geologists to predict earthquakes?    To   
             test this model, we must have certain information:

             a.  When did the last earthquake occur along the given fault segment?

                  This is either known from the historic record, or must be estimated
                  from the geologic record (which yields only a rough estimate). 

             b.  How often do earthquakes occur along the given fault segment?

                  In some cases (e.g., Parkfield, CA), there are sufficient historical 
                  data to calculate a recurrence interval; in other cases the rate of
                  strain buildup is used to estimate recurrence intervals.
         

             Such information exists only a few faults (e.g., the San Andreas Fault).
             By contrast, little is known about recurrence intervals or strain rates
             for areas where large intraplate earthquakes have occurred.

        3.  DIFFICULTIES

             Testing the elastic rebound model is complicated by the fact that:
           

             a.  Rates of strain, fault creep (or slippage) can change through time

             b.  Strain is sometimes distributed along several parallel faults

             Thus, the "state of the art" is to forecast earthquake probabilities,
             with earthquake prediction having achieved only limited success.

        Note that because earthquakes are not completely random events, the
        method used to calculate their probabilities of occurrence differs from
        that used for random events (such as rolling dice).

 Return to Geology 100 Home Page