MESURE DU BRUIT ET SISMOS MULTIPLES (etude experimentale dans les Landes)*

To improve the ratio of signal to noise resulting from the shot, seismologists have increased the number of geophones in each recording channel and have also increased the length of each spread. As a rule, however, the effectiveness of this multiple recording is based on theoretical considerations, and it appeared desirable to carry out an experimental investigation to obtain a direct measurement of the reduction in noise as a function of the geophone pattern etc. The work was done in the Landes where experience had shown that recording was difficult.

The experimental arrangement, which allowed the accurate measurement of the noise without interference from reflected energy, is described.

Shots, repeated under apparently identical conditions, revealed the random nature of the noise and pointed to the necessity of a large number of experiments to yield an accurate result. Two geophone patterns were employed, one of 12 geophones in a straight line and the other of two parallel lines of geophones, the directions of the geophone lines being either along the shot line, perpendicular to the shot line or at some intermediate angle.

The results obtained, in two widely spaced locations were in good agreement and allowed definite conclusions to be drawn. When the 12 geophone line is long enough the noise is reduced by a factor of 0.36 and with two lines, each 55 metres long and containing 12 geophones, the reduction was 0.29, if the lines were sufficiently widely separated. The results agree with the factor K N‐½ for the reduction in noise where N is the number of geophones per trace and K is a factor close to unity. The results also show the influence of the length of the geophone line its orientation and its mean distance from the shot.