principles and applications of microearthquake networks

18 2. Itistrirmeritation Systems
In practice, however, we are faced with physical constraints such as
accessibility and sources of cultural and natural noise. For example,
Raleigh et nl. (1976) used a 13-station network to study microearthquakes
from a specific source area within a producing oil field near Rangely,
Colorado. Seismicity of primary interest was confined to an area about 2
km wide by 6 km long. Focal depths were in the range 1-5 km. The
purpose of the network was to determine if fluid injection and withdrawal
would be effective in controlling seismic activity. Seismometers were
concentrated around the area of interest and distributed more sparsely at
greater distances. This arrangement provided good hypocentral control in
the experimental area and still allowed seismicity in the surrounding area
to be monitored.
By contrast, a 16-station microearthquake network in southeast Missouri
has a relatively uniform distribution of stations (Stauder et al., 1976).
The primary purpose of the network is to delineate the features of the
New Madrid seismic zone-the site of the major earthquakes of 1811-
1812. In this case the entire area is under investigation, and it is important
to establish the regional seismicity unbiased by station distribution.
Certain physical constraints will also affect station distribution. For
example, Quaternary alluvium should be avoided if possible, as it usually
has higher background noise than more competent rock. In some places in
the USGS Central California Microearthquake Network, station distribution
is relatively sparse (Fig. 3). Sometimes this is due to inaccessible
terrain or to large bodies of water, where the cost of station operation
outweighs the expected benefits. Other places are not instrumented because
of their proximity to large cities that generate too much cultural
noise.
One way to reduce cultural noise is to place seismometers at some
depth. For example, Steeples (1979) placed 10-Hz geophones in steelcased
boreholes 50-58 m deep for a 12-station telemetered network in
eastern Kansas. This was effective in reducing cultural noise (such as that
from nearby vehicles and livestock) by as much as 10 dB. However, it was
less effective in reducing train and other noise in the range of 2-3 Hz.
Takahashi and Hamada (1975) described the results of placing seismometers
in deep boreholes in the vicinity of Tokyo. Velocity seismometers with
a natural frequency of 1 Hz were placed in a hole 3500 m deep, along with
sets of accelerometers, tiltmeters, and thermometers. Within the frequency
range of 1-15 Hz, the spectral amplitude of the background noise
was about 1/300 to 1/1000 of that observed at the surface. This noise level
was small even when compared with that from other Japanese stations
located in quieter areas.