|
The AS-1 Seismograph –
Operation, Filtering, S-P Distance Calculation, and Ideas for Classroom Use 1 L. Braile,
November, 2002; Updated
November, 2004 |
|
Examples of Recorded
Seismograms and Filtering of Seismograms: The AS-1 seismometer (Figure 1) has a natural
period of oscillation of about 1 s. To
make the seismometer more useful for recording local as well as distant
earthquakes, the amplifier unit supplied with the AS-1 uses strong filtering to
enhance the long period (greater than 10 s period) response of the
instrument. The combined response of the
seismograph is more broadband and can record signals ranging from about 30 s
period to 2 Hz (0.5 s period). Local and
regional (within about 1200 km or 10 degrees distance from the seismograph
location) earthquake seismograms are usually dominated by short period (near 1
s period) energy. Teleseismic (distant
earthquakes, 10 to 180 degrees) recordings usually consist of lower frequency
(longer period) energy and often have prominent surface wave arrivals of 12 –
30 s period. Microseisms are a nearly
universal source of ground noise in seismic recording and have periods of about
3 – 8 s. Because this period range is
amplified strongly by the AS-1 seismograph, filtering of the records is often
desirable to enhance the earthquake signals.
A useful filter for local and regional events (short period) is a
bandpass filter with frequency cutoffs of 0.5 Hz to 3 Hz (frequency in cycles
per second, or Hertz, is equivalent to 1/Period in seconds, so these cutoffs
are the same as 2 s period and 0.333 s period; in the AmaSeis control
menu, after the seismogram is selected
or extracted, enter the two frequencies, lowest frequency is cutoff one,
highest frequency is cutoff two). An
example of use of a short period filter is shown in Figures 2 and 3 for an
earthquake that was located 2.63 degrees (~292 km) from the station.
[1] 
Last
modified
The web page for
this document is:
http://web.ics.purdue.edu/~braile/edumod/as1mag/as1mag2.htm.
Partial funding for this development provided by IRIS and the National Science Foundation.
ã Copyright 2002-4. L. Braile. Permission granted for reproduction for non-commercial uses.
Figure 1. The AS-1 seismometer showing the main
components of the instrument. The AS-1
is a vertical component seismometer. Up
and down motions of the ground, and therefore of the base and frame of the
seismometer, cause the coil to move relative to the magnet that is suspended by
the spring and boom assembly. The mass
of the seismometer, consisting primarily of the magnet and the washers, tends
to remain steady because of inertia when the base moves. The motion of the coil relative to the magnet
generates a small current in the coil.
The current is amplified and digitized by an amplifier unit (not shown)
and connected to the computer for recording and display. The damping (using oil in the container and a
washer mounted to a bolt extending downward from the boom into the oil) reduces
the tendency for the mass and spring system to oscillate for long duration from
a single source of ground motion (arrival of seismic waves at the location).
Figure 2. AS-1 seismogram recorded at West Lafayette,
Indiana from an earthquake located near Evansville, Indiana, December 7,
2000. The earthquake epicenter was about
292 km away from the seismograph and had a magnitude of about 3.9 (mbLg). Microseisms of about 3-6 second period are
visible before the first arrival (the compressional or P wave) that is located
at about 1.1 minutes (relative time).
The S (Shear) wave and surface waves are the largest arrivals following
the P wave.
Figure 3. Seismogram for the
For
teleseismic or distant earthquakes, lower frequency filtering often enhances
the seismogram, especially the surface wave arrivals that are prominent for
shallow focal depth (0 – 70 km depth) earthquakes. Suggested filter cutoffs for lower frequency
(longer period) signal enhancement are 0.01 Hz (cutoff one) to 0.2 Hz (cutoff
two), equivalent to a 100 s to 5 s period range. Because the seismograph has very low response
at very low frequencies, the lowest frequency cutoff (0.01 Hz) is intended
primarily to avoid amplitude distortion and offset of the zero level of the
trace that is sometimes caused by filtering the extracted trace. An example of the advantages of filtering
seismograms to enhance the long period energy is shown in Figures 4, 5 and
6. Figure 4 is the AmaSeis screen
display for a 24 hour record (each horizontal trace is one hour long; the
display is designed to look like the familiar drum recording on paper; hours
scale is along the left side; minutes scale is at the bottom) including an
earthquake that occurred at 11:41:47.9 (Greenwich Mean Time, GMT), August 9,
2000, in Michoacan, Mexico. The event
had a magnitude of 6.5 (surface wave magnitude, MS) and was located about 25.96
degrees (2887 km) from the seismograph at
Figure 4. AS-1 seismograph recording for
Figure 5. Extracted seismogram for the Michoacan
earthquake. Amplitude scale is in count
or digital units.
The
filtered seismogram (Figure 6) enhances the main arrivals and makes it much
easier to determine arrival times.
Filtering using the control menu in AmaSeis can be performed more than
once to further enhance the frequency range of interest.
Figure 6. Extracted seismogram for the Michoacan
earthquake after bandpass filtering (0.0001 Hz and 0.1 Hz cutoff frequencies)
to enhance the long period (10,000 – 10 s) energy. The P wave arrival (the first arriving energy
is at about 3 minutes, relative time, followed about 40 s later by the PP
arrival. The PP phase is a P wave that
reflects from the Earth’s surface once near the halfway distance between the
epicenter and the seismograph station.
The S (shear) wave arrives at about 7.5 minutes and is visible partly
because of the frequency change (lower frequencies often characterize the shear
waves). Prominent surface waves
(Rayleigh waves) are visible at about 14 – 15 minutes relative time.
S-P Distance Calculation: The AmaSeis software includes a useful tool for determining the epicenter-to-station distance from the S-P arrival time difference on seismograms. For AS-1 seismograms extracted from a 24 hour record or saved as SAC files, or for SAC format seismograms downloaded from the internet, the AmaSeis arrival time picking tool can be used to mark the interpreted arrival times of the P and S waves (Figure 7). Then, by selecting the travel time curve tool, the seismogram is displayed on standard travel time curves. By moving the seismogram on the screen, the interpreted P and S times can be aligned with the arrivals on the travel time curves (Figure 8). The S-P times can also be analyzed using standard travel time curves (Bolt, 1993, p 134; or from the internet at http://lasker.princeton.edu/index.shtml). Once aligned, the epicenter to station distance is determined and displayed.
Figure
7. Vertical component seismogram from
GSN station KIP for the
Figure
8. KIP seismogram from Figure 7
displayed with the travel time curve tool in AmaSeis. In this tool, the seismogram can be moved
around on the screen until the selected P and S arrival times are aligned with
the travel time curves. The epicenter-to-station
distance corresponding to the observed S-P time is then determined by the
position of the seismogram and the specific distance, in this case 58.72
degrees, is displayed adjacent to the distance axis.
Using
seismograms from three or more stations, the location of the epicenter can be
estimated by triangulation. For example,
for the September 30, 1999 Oaxaca earthquake, seismograms for stations KIP,
BINY, COLA and AMNO were displayed in the AmaSeis program, the P and S arrival
time picked, and the epicenter to station distances inferred. Using an inflatable globe, circular arcs are
drawn on the globe corresponding to the epicenter-to-station distance for each
seismograph station (Figure 9). When all
arcs have been drawn, the approximate epicenter location is determined by the
intersections of the arcs (Figure 10).
Although this triangulation method is not the most reliable or accurate
technique for locating earthquakes, it is relatively easy to understand and
illustrates concepts of the travel times of seismic waves at increasing
distances and earthquake location methods.
Figure
9. Using a piece of string marked at the
appropriate distance in degrees (measure along the equator) and a felt pen
(water soluble ink), a circular arc is drawn on an inflatable globe using the
seismograph station (in this case, station AMNO) as the center of the
circle. Arcs are also drawn for other
stations using the appropriate distances inferred from the S-P times.
Figure
10. After all arcs have been drawn, the
epicenter is inferred by the approximate intersection of the arcs. Because of inaccuracies in the estimation of
the S-P times and drawing the arcs on a globe, the intersections of the arcs
may not be exactly a single point.
However, a reasonably good location is determined by this method.
Because
the S wave arrival is not always prominent on a vertical component seismogram,
the epicenter-to-station distance cannot always be inferred from AS-1
seismograms. Furthermore, for
seismograms from earthquakes that are located greater than 105 degrees from the
station, no direct S wave arrivals are present due to the Earth’s core. For earthquakes recorded at the West
Lafayette, Indiana AS-1 station since April, 2000 (Table 1), distances inferred
from the AS-1 seismograms are compared with the calculated distances (from the
USGS reported epicenter to the station location) in Figure 11. Although the distances generally correspond,
the AS-1 distance determination appears to be accurate to only about +/- 4
degrees distance for distances greater than about 20 degrees.
More
information on using the AmaSeis software is available at:
http://web.ics.purdue.edu/~braile/edumod/as1lessons/UsingAmaSeis/UsingAmaSeis.htm. An S minus P earthquake location exercise is
available at:
http://web.ics.purdue.edu/~braile/edumod/as1lessons/EQlocation/EQlocation.htm.
Figure 11. Comparison of calculated distances (using
station location and USGS epicenter location) with distance estimated from the
S-P times on seismograms recorded by the AS-1 seismograph. Distance estimates from the S-P times were
determined using the travel time curve tool in AmaSeis. Comparing actual and AS-1/AmaSeis S-P
calculated distances. N = 75; Standard
Deviation = 2.51 degrees (November,
2004).
Ideas for Classroom
Use: Operating an AS-1 seismograph in a classroom or school building
encourages awareness of earthquake activity around the world and provides an
opportunity for teachers and students to work with real scientific data. There is particular interest and excitement
when a local or regional event occurs and is recorded by the seismograph or
when a significant distant event happens.
Students will be interested in “checking the seismograph” each day and
seeing how their records and magnitude estimates compare with the seismograms
recorded at other stations and official magnitudes. The occurrence of significant earthquakes
around the world (there are about 20 events per year of magnitude of 7 or
greater: many of these events cause significant damage) can be used to
stimulate discussion, learning and research on geography, the causes of
earthquakes, propagation of seismic waves, earthquake hazards, and earthquake
safety. Some specific suggestions for
classroom use are:
1. Maintain a catalog of seismograph recording.
Students can check the seismograph every day (or at regular intervals),
perform routine maintenance (check that it is operating properly and perform a
time check), identify possible recorded events and enter appropriate
information into the catalog. A hand
written catalog is sufficient for recording such information as time corrections
and approximate times of recorded events.
A more
complete catalog can be
created using a spreadsheet (see Table 1, below) with information on earthquake
to station distance and magnitudes.
Distance from the earthquake epicenter can be inferred from the P and S
travel times as described above (AmaSeis provides a simple tool for performing
this estimation; these times will not be able to be determined for all recorded
events). Magnitudes can be estimated
using the procedures described in The AS-1
Seismograph – Magnitude Determination or using the online magnitude
calculator at: http://web.ics.purdue.edu/~braile/edumod/MagCalc/MagCalc.htm.
A comparison between the AS-1 distance
and magnitude information can be made by checking the USGS/NEIC, IRIS event
search or IRIS seismic monitor sites on the Internet. Instructions for accessing earthquake data on
the Internet are included in
http://web.ics.purdue.edu/~braile/edumod/eqdata/eqdata.htm. All data should be written in the catalog to
provide a record of activity of your seismograph station and document the
comparison of distance and magnitude determinations.
Exercises that include
magnitude calculations using AS-1 seismograms are available at:
http://web.ics.purdue.edu/~braile/edumod/as1lessons/EQlocation/EQlocation.htm
and
http://web.ics.purdue.edu/~braile/edumod/as1lessons/magnitude/CalcMagnElect.htm.