INFLUENCE OF NUMBER AND SPACING OF SENSORS ON THE EFFECTIVENESS OF SEISMIC ARRAYS*

Ideally, geophones would be placed in a noiseless environment, in which case there would be no reason to resort to arrays of geophones. If the noise is such that an array is required, the objective of the array is to enhance the signal‐to‐noise ratio and thus to maximize the intelligence that can be derived from a given signal. The design of the array will be a function of the signal characteristics, of the direction and velocity of the noise in the bandpass of the signal, and of the site geology.

It has been demonstrated previously that in a practical sense the optimum array processing is represented by precise beam forming, by which we mean simple time‐delay and summation. Increasing the number N of sensors within a given area decreases the inter‐element spacing and may increase the coherency between noise samples at adjacent sensors, thus yielding poorer results compared to √N improvement one expects to get if the noise is uncorrelated. Increasing the number of sensors by proportionately increasing the area is liable to result in signal deterioration, also yielding an unfavorable comparison to √N improvement in signal‐to‐noise. These two effects, together with economical factors, combine to limit the number of sensors that can be used.

Although the data on which our conclusions are reached were drawn from earthquake seismology, the principles involved are equally applicable to exploration seismology and to other geophysical measurements in which arrays of sensors are required.