Description of a Seismic Installation
    Early in the summer of 1999, I accepted a position at the University of Alaska, Fairbanks campus.  I am now one of two seismic data managers at the Alaska Earthquake Information Center (AEIC) at the Geophysical Institute.  Brent Foshee and I are in charge of bringing seismic data into the lab in an efficient manner to be processed by lab staff and then archived for research purposes.
    Alaska is an amazing place to study seismic activity.  The state has experienced three of the ten largest earthquakes in recorded history.  There are also more earthquakes here than in the rest of North America combined.  This is due to the tectonic environment of Alaska, which is different than the more familiar and common case of California.  California contains a large series of mostly transform faults between two crustal plates.  These plates essentially just slide next to each other.  In Alaska, the situation is much more complicated.  In southeast Alaska, there exists a similar case of mostly transform faults.  But as an observer moves further north and west, these faults change to a subduction zone environment where the Pacific plate is directly colliding with the North American plate.  A subduction zone is created, where the Pacific plate is being forced underneath the North American plate.  This allows for a much larger buildup of stress, therefore larger earthquakes.
    The subduction zone environment is also the basis for over a hundred volcanoes in Alaska, most of which are located in the Aleutian Island chain.  AEIC monitors the seismic activity associated to these volcanoes and shares the data with the Alaska Volcano Observatory (AVO) in order to help monitor these volcanoes.

 A general map of major Alaskan faults and earthquakes of interest.

    AEIC has a network of about 220 seismic stations around the state of Alaska.  They are placed to be near the major faults zones or volcanoes in order to carefully monitor the activity there.  This is where my job comes into play.  I monitor the performance of these stations and attempt to keep them in top condition.  This usually involves a visit to almost every station in the network once a year, a big job for a field season of essentially May to October.  I am also involved in installation of new instruments, and designing their installation environment.

    A typical seismic instrument is quite a simple device.  It is metal cylinder about ten inches high and about 6 inches in diameter and coupled directly to the ground with a strap.  Inside the cylinder a heavy mass is suspended (with as little friction as possible) inside a tight, thick coil of wire.  Many of these instruments have environment sensors in the top of the instrument that report the local temperature and humidity.  These are variables that come into effect when the instrument dampens the spring effect on the mass slightly in order to keep it from bouncing on its suspension or becoming too stiff to move properly.  Some of the instruments have the ability to remotely center their mass when a proper signal is received from our laboratory.  This helps keep the instrument in tune in case the base becomes off level over time without a technician having to visit the installation.


cutaway of seismic instrument
    Cutaway view of a single component geophone.  In the center of the instrument is the mass,
    surrounded by red wire.  Underneath is a seismograph of a 5.6 earthquake in Turkey as recorded
    by an extremely long period instrument, each tic mark corresponds to one minute.  Ground motion
    continued to be received in Alaska for about two hours.

    When there is no ground motion, the mass just sits inside the coil of wire, with no change in current passing between the mass and coil.  (This is not completely true because at all times there are small fluctuations in the materials of the instrument and small vibrations in the ground.)  When movement of the ground occurs, the coil will displace around the mass, creating  a fluctuation in the electrical current.  This change in current is actually a change in distance over a change in time (velocity).   This is different than many motion detectors, such as in the ride tunable suspensions in cars that report their data as changes in velocity over continuous time (acceleration).  A properly functioning instrument is always on, always transmitting.  In order to get a location for an earthquake, a model is used that has established a velocity of waves traveling through the rock in a geographic region by using previous earthquake data.  The change in distance over time is then established by observing the data received at several different stations.  This method is made more accurate by measuring the time between two characteristic changes in the earthquake wave, the first arriving in a compression wave, the next as a shear (up and down) wave.  The magnitude of the earthquake is determined by the amplitude of the shear wave.
    Since timing of the seismic wave is so crucial, each installation is timed using its own gps clock.  This provides a nearly synchronized system of all stations.  However this is not without numerous complications, such as snow covering a gps receiver.  If a stations is not reporting on time, its data will appear to show up late or early in the archived records and cannot be used to locate.

    There are basically three types of seismic stations.  The first and most basic type is an instrument that is tuned to specifically respond to short period movement in the base surrounding it.  Short period movement comes from a source somewhat near the instrument in order to get an accurate signal for a localized event and shows much detail.  Most instrument of this type were originally installed or upgraded from older designs in the early 1980's.  A common name for this instrument is the geophone.
    The next step up is a station with three of the same type instrument as above, except the instruments are arranged geometrically in order to best receive signals from an east-west, north-south, or up-down direction.  This arrangement enables us to most accurately see which direction the earthquake is coming from and how much vertical displacement it is causing at the station.

 See a three component station.

    The third type of instrument is the most modern.  A single metal casing houses instruments measuring the three vectors mentioned above.  These are the ones that enable us to center the masses and change the dampening remotely.  The newest instruments have a higher dampening factor built in.  This enables the instrument to pick up a signal with a longer period, where the shorter period signal is filtered out.  This instrument can see smaller movement from further away from the source of activity, thus the name  broadband seismometer.  This ability comes at some cost of detail, so many of these stations are installed with a short period instrument in order to see more localized movement in detail.

    The most quiet stations are able to see the most detail of an earthquake.  A nice quiet station can detect my footsteps as I tiptoe up to it from about 150 feet away.  Interfering noise can come from human sources such as cars, trains, automobiles, quarry blasts, wind vibrations from telephone poles and towers.  Natural sources can be stormy weather, animals, and wind vibrations on trees.
    To avoid these complications, seismic sites are selected to be in remote locations away from all possible noise sources.  Another choice is to put the seismometer in a borehole dug into bedrock.  Thirty feet or more is an ideal depth.  Bedrock is chosen because it is a solid surface that transmits seismic waves with the most detail instead of a sand or gravel bed which will disperse the waves.

 Remote station at Waxell Ridge, Alaska.

    A remote seismic station is ideally composed of a watertight container partially or completely buried into the surface.  It either covers a borehole or a concrete/bedrock pad where the seismometer rests.  A minimum of thirty feet away (to avoid wind noise) is another container with the batteries, charging system, and telecommunications from the instrument to a microwave tower.  One or two solar panels are mounted on a rack several feet off the ground.  For an alternative remote power source I am currently experimenting with wind generators.  One was installed on an Augustine volcano station, however it quickly seized up due to dusty conditions and poor materials.
    Telecommunications to the phone system occur in one of two ways.  Older systems use a standard radio set at a specific frequency to transmit the seismic signal to the phone system.  Newer broadband systems have a digitizer at the site that converts the analog seismic signal to a digital format that is sent over a multispectrum radio that serves as a modem.  Both require a visible line of site to a receiver at a microwave tower.
    The tradeoff in a remote installation is the lack of power and telemetry to get the signal to the laboratory.  A station is powered with a relatively low voltage of about 1.5 volts.  This is easily supplied with 12 volt batteries and solar power to recharge them.  Radios are used to transmit the signal to the state of Alaska microwave phone system.  But it is much cheaper, easier, and more reliable to install a station that is powered directly from a microwave tower and wired directly to the phone system.
    At the lab, all data is separated out from the phone lines into each own channel and any analog signals are converted to digital format.  Data from each channel is then sent to a preliminary archive database to be processed by an analyst using a computer software program to determine earthquake content.  The signals from a select number of stations of interest are sent to helicorders which record the data as a line trace on a continuously rotating and translating drum.

    I am very excited about this job.  It involves travel throughout this very beautiful state and is related to the studies I am pursuing at UAF.  I expect to stay here for several years and pursue the electronics experience I am gathering.

gps station

    That's me at a gps station on Glacier Island, Prince William Sound.

 Home Page