Mini-treatise on Earthquake magnitudes

Regarding the comments on magnitude and energy - at the risk ofboring those who have seen
this before. When Richter developed his original magnitude scale in 1935, he wasl ooking for an
empirical way to classify the relative size of smallto moderate earthquakes in California. The
scale was based on thepeak amplitude, or wave height, recorded on a standard seismograph
commonly deployed for locating local earthquakes in California atthat time. Richter's scale was
not initially related to energy, nor could it be used for large earthquakes far away. The
instruments which Richter used are no longer in common use. Modern broadband seismographs
can simulate the response of these early instruments and seismologists can still calculate a
magnitude essentially identical to Richter's original scale (now called local magnitude or ML).
Later work by Richter and other seismologists developed other magnitude scales so that
themagnitude concept could be applied to earthquakes occurring anywhere in the world. Body
wave magnitudes (mb) are based on the amplitude of the initial p wave signal and surface wave
magnitudes (MS) are determined using the peak amplitude of longer period (20 sec.) surface
waves. Relationships were also developed to estimate the amount of energy released by an
earthquake from the magnitude, but none of these early formulations were entirely satisfactory
because they involved extrapolating the energy from the amplitude of a particular wave or phase
and did not consider the entire wave train. These scales also had a problem in that they tended to
saturate at around magnitude 8 and couldn't effectively differentiate between very large
earthquakes (such as the 1906 San Francisco earthquake) andtruly humongous ones like the 1960
Chilean and 1964 Alaskan earthquakes. Seismologists now use seismic moment and moment
magnitude (Mw) to estimate the energy released by earthquakes. Seismic moment depends on
the fault rupture parameters - thearea of the fault, the amount the two sides moved during
theearthquakes, and how tightly the two sides are locked together. Seismic moment and moment
magnitude can be determined by analyzing the complete wave forms of surface waves as
recorded by broadbandinstruments and are now routinely determined for global earthquakes of
magnitude 5.5 or larger, and well-recorded US earthquakes of magnitude 4 and larger.

The recent earthquake near Seattle illustrates how many of these different magnitudes come into
play. The May 2 earthquake (now called the Duvall earthquake because of its proximity to the
smalltown of Duvall on the Snoqualmie River) was located about 25 miles northeast of Seattle or
6 miles east-northeast of Duvall. A magnitude estimate of 4.8 was initially released about ten
minutes after the earthquake by the University of Washington. This preliminary magnitude was
based on automatic processing by a computer recording system picking the duration of shaking
on sensitive instruments. This type of magnitude is readily determined by seismograph networks
linked to computer systems and are routinely determined for California earthquakes within afew
minutes of the event. These are the magnitudes you usually see when you finger the listings of
California earthquakes. Since they are initially determined without the review of a seismologist,
they may be in error. The U.S. Geological Survey issued a body wave magnitude estimate of 5.4
about an hour after the earthquake. About the same time manual analysis of the local
seismograms at the University of Washington determined a local magnitude of 5.3. A moment
magnitude estimate of 5.2 was determined by Oregon State University a couple of hours later by
analyzing the complete wave forms of the earthquake as recorded by broadband seismographs. A
similar analysis the next day at the University of Washington on a subset of data used by Oregon
State determined a moment magnitude of 5.1. None of these different magnitudes were exactly
the same as the original "Richter" magnitude (although the local magnitude comes the closest).
But all of the magnitude scales roughly agree with Richter's scale so that a magnitude 5
earthquake, regardless of scale, is a moderate EQ capable of doing minor damage in the
immediate epicentral area. Log Mo (seismic moment) = 16.74 + 1.22Log(CD^), where

C = maximum peak-to-peak amplitude in mm
D = duration in sec from time of S-wave onset to the peak-to-peak amplitude that exceeds C/3
^ = epicentral distance in km (from recording station to event)
seismogram recorded on Wood-Anderson instrument 300 km from eventBolt and Herraiz, 1983