LECTURE 6 NOTES - SEISMIC INSTRUMENTATION (updated 10/15/97)
Instructor: Professor Barbara Romanowicz
Director of Seismological Laboratory
Office Hours: Thursday 2-4 pm , upon appointment only
475 Mc Cone Hall
SEISMIC INSTRUMENTATION
Because we can infer so much about the structure of the earth's interior
and seismic sources from measuring times of travel of different waves, it
is very iimportant to keep accurate record of BOTH vibrations (their
amplitude, periods, phases) and TIME. Need not only a means of recording
motion, but also an accurate clock attached to it. Because of the
polarization properties that help identify different waves, it is also
important to measure the vibrations in 3 dimensions: three components of
motion necessary -->
typically : one vertical and two horizontal (corresponding to orthogonal
reference frame - typically North and East).
Key applications: location of remote earthquakes
measuring earthquake size
determining the mechanism of earthquake rupture
(what fault was it on and in which direction did the fault rupture?)
The design of such an instrument is not simple, because of the large
range of earthquake amplitudes and frequencies:
20Hz to 54 mn
strong motion, weak motion (10-6 g to 2g)
Earliest instruments: seismoscopes
did not record time
~132 A D CHang Heng
designed to indicate 1) the occurrence of an earthquake. 2) the direction
of approach.
metal bowl 2 m in diameter - 8 dragons attached to the sides, frogs
beneath them. Mouth of dragon held a ball secured in place by a lever
connected to some interior mechanism which we do not know anything about
(pendulum?). At onset of eq. one of the dragons would release a ball into
a frog's mouth: direction faced by dragon would indicate direction from
which the waves came.
Story was that it actually worked: there was at least one instance when
people did not feel vibrations but a ball fell and later it was confirmed
that a distant earthquake had occurred.
Very simplified, we now know: it is more complicated than that to know
where the waves came from because of the different types of particle
motions involved in earthquake waves.
Modern seismograph: much later
18th century, earliest recording devices: pendulums employed to indicate
ground motion: but no record of time of arrival nor permanent record of
the ground motion.
Mid 19th century: Luigi Palmieri:"seismografo electro-magnetico (1856)
recorded time and detected vertical motion of the ground: movement of a
mass on a spiral spring and, horizontal motion: motion of mercury in U tubes.
gave info on the direction, intensity and duration of earthquake shaking,
and 3 components of motion.
-> Japan, 1892 John Milne (visiting professor from England) with James
Ewing. built instrument that recorded ground shaking as a function of
time: in particular, installation at Lick Observatory (UCB) in 1897.
Basic principles were the same as modern ones, although not as sophisticated:
Principle of seismograph
If we could float in the air, unaffected by earthquake and hold a pencil
down ,
allowing it to move across a sheet of paper fastened to the shaking
ground: we could make a seismogram.
In practice, freely suspend a mass from a frame attached to the ground:
mass
"independent" of the frame's motion. When frame is shaken, the inertia of
the mass causes it to lag behind the motion of the frame: relative motion
recorded by pen and ink on paper wrapped on rotating drum, or by light
spot and film, producing a seismogram.
3 components of motion: one instrument records only one if these three
components ---> 3 instruments to reconstruct complete motion.
vertical: spring + mass
horizontal: mass attached to a horizontal pendulum swinging on hinges
like a door.
The motion recorded is not the actual motion of the ground: this must be
deduced/calculated taking into account the physics of the pendulum's motion.
Seismographs must record waves with amplitudes as small as 10-9 m, size
of a molecule of gas. Such small motions need to be magnified: in old
instruments, done by mechanical means (series of jointed levers or
swinging spot of light). Now: electromagnetic sensors: motion produces an
electrical field which is magnified electronically thousands of times
before it is used to drive an electric stylus writing on paper. Now, in
fact , this is gradually abandonned and recording is directly into a
computer or magnetic tape (digital versus analog recording).
Modern seismographs record continuously and contain accurate clocks (now
GPS) that provide time continuously (before, radio time, minute marks).
Instrument response
Pendulum swings or an elastic spring vibrates at a characteristic or
natural rate whih depends on the length and spring's elasticity. The
period of the vibration is constent, independent of the amplitude:
natural frequency (or natural period). When the ground shakes at
frequencies much smaller than the natural frequency of the pendulum, the
displacement of the pendulum's mass relative to the frame closely follows
the acceleration of the ground. If close to natural frequencies:
vibrations greatly amplified; if frequencies well above natural
frequencies, mass does not move much at all: relative displacements can
be amplified to give a true measure of the displacement in seismic waves:
design instruments with varied sensitivity to the frequencies of an
earthquake (20Hz to 54 mn or ~1000 sec). Generally, an earthquake
observatory will contain a range of instruments to record the whole range
of ground motions.
most current transducers: either acceleration or velocity
damping
if you just have a spring and a mass, you will find that pendulum
continues to oscillate even after the hand has been brought to rest.
Tells us nothing about the ground shaking and must therefore be damped by
some mechanical or electrical means.
2 issues :
dynamic range: small to large signals, now with electronics A/D
converters, is very large but still require special instruments to record
"strong motion"/
strong motion
weak motion
strng motion accelerometers: record in triggered mode (rarely need to record)
frequency range
long period
short period instruments
now broadband - micro seismic noise
background noise: installation issues - seismic vaults, get away from
surface disturbances due to atmospheric fronts, also to complication of
signals due to weathered rock (many reflections, conversions). Borehole
instrumentation
bacground noise
signal generated noise
Observatories and networks
Von Reubeur Paschwitz 1889 (Japan earthquake observed in Germany) -->
started global surveillance
by 1957 ~ 600 seismographic stations around the world
standard observatory practice:
pick phases (identify the waves), compute times of arrival, measure
amplitudes of waves, periods
global networks
1960's : WWSSN - standardized instrument response, each observatory set
of 3 component short period and 3 component long period instruments,
about 100 instruments around the world: could directly compare
recordings. Drawback: analog recording (on photographic paper).
1970's: first digital recordings
1980's: broadband digital seismometer systems with sophisticated digital
recording and high dynamic range (map of distribution)
we are now talking about installations on the ocean floor....
regional networks:
traditionally short period and analog strong motion
now: broadband
a typical BDSN station: 3 component broadband seismometers, 3 component
strong motion accelerometers, GPS clock, other geophysical instruments:
GPS receivers, electromagnetic sensors, barometers, thermometers, Radon
measurement.....etc...
continuous telemetry: instantaneous analysis ---> REDI display.