The following is summarized from a paper by Bent (1992) on the 1925 Charlevoix earthquake. Seismograms of this earthquake were analyzed using advanced waveform modeling techniques unavailable at the time of the earthquake. While the overall data set was poorer both in terms of quality and in number of seismograms available than we would expect for a recent earthquake of comparable size, it was sufficient to provide a fairly detailed description of the source properties of this earthquake.

Data:
Seismograms from 20 seismograph stations in North America, Europe and the Pacific were collected. Data from 15 were of sufficient quality to be used for detailed waveform modeling. Amplitude data from 19 stations could be used for magnitude calculations. A detailed description of data quality and instrument responses can be found in Bent (1993).

Magnitude:
The 1925 Charlevoix earthquake occurred 10 years before the development of the Richter magnitude scale. Gutenberg and Richter (1954) later made an attempt to calculate magnitudes for earlier earthquakes which they considered significant. Because the task of collecting large numbers of seismograms for each earthquake of interest was enormous, their measurements were often calculated from a small number of seismograms. It is believed that their published magnitude of 7.0 for the Charlevoix earthquake was based on a measurement from a single station in southern California.

In theory, the magnitude of an earthquake should not vary from one location to another since the Richter scale has a correction term for the distance between the earthquake and the seismograph station. In practice, we often see variations in the magnitude for the same earthquake measured at different stations. These differences occur because earthquakes do not radiate energy equally in all directions, because some soil conditions may amplify or reduce the seismic signal or because an instrument has been improperly calibrated. To reduce these effects, seismologists determine the magnitude at as many stations as possible and then calculate the average. Usually the difference between one station and another is small, but occasionally the differences can be large.

To further complicate matters, there are actually several magnitude scales. All operate on the same principle: a difference of a factor of 10 in amplitude (at the same distance) corresponds to a magnitude difference of 1.0. The scales measure amplitudes on different parts of the seismogram. While the magnitudes are usually similar, they can be noticeably different. M_S measures surface waves with a period of around 20 seconds; m_B measures body waves; m_b also measures body waves but is restricted to periods of about 1 second; M_L (the original Richter magnitude) measures the largest amplitude regardless of phase or period but is restricted to stations close to the epicenter; m_N or m_bLg measures the high frequency (1 second period) Lg phase (the magnitude favoured in the description of smaller eastern Canadian earthquakes); M_W (or M) measures very long periods and is more closely related the physical earthquake rupture process than the other magnitudes. In general the high frequency magnitudes (M_b, M_N) are important for seismic hazard estimations in the epicentral region and the longer period magnitudes (M_S, M_W) give a better overall picture of the size of the earthquake.

Gutenberg and Richter's magnitude of 7.0 is an M_S from one station. Using data from stations in North America, Street and Turcotte (1977) determined an M_S of 6.6, while Ebel et al. (1986) calculated an M_S of 6.2 using European data.

In a more recent study, Bent (1992) combined North American and European data to obtain an M_S of 6.2 (± 0.3). While magnitudes at most stations were close to the mean value, they ranged from 5.2 to 7.2, emphasizing the need for a large number of measurements over a wide azimuthal range. The m_B is 6.5 (± 0.4). M_W is 6.2 (± 0.2). As would be expected for an earthquake of this size, M_s and M_w are in good agreement with each other. Because M_W corrects for source effects, the variation in value from station to station (5.7-6.5) is much less than for M_S.

For comparison, the magnitudes for the 1988 Saguenay, Quebec earthquake are as follows: M_N 6.5, M_S 5.8, M_b 5.9. A comparison of the M_S values indicates that the 1925 Charlevoix earthquake was 0.4 magnitude units larger, which means that the amplitude of the seismic signal at any given distance would have been 2.5 times larger and that 4 times more energy was released by the 1925 earthquake.

See also our FAQ for more questions and answers on magnitude scales.

Focal Mechanism:
Using advanced waveform modeling techniques, it was determined that the 1925 Charlevoix earthquake occurred 10 km beneath the Earth's surface (probably beneath the St. Lawrence River) on a fault roughly parallel to the St. Lawrence River. These conclusions are consistent with surface geology and with studies of recent, smaller earthquakes. However, because of the depth of the earthquake, we cannot associate it with any particular fault. References

Home page Charlevoix-Kamouraska 1925