Limitations of seismic field instrumentation for converted-wave recording

Malcolm B. Bertram

The recording of converted-wave data, in particular multicomponent data, means that several of the traditional field techniques need to be revised. Over the past few years, the CREWES project together with the Field School run by the Department of Geology and Geophysics has attempted to redefine some of these parameters. One is the use of single geophones which give us point recording, rather than the traditional geophone array; another is the necessity of recording data with no low cut filters applied. Recording data this way leads to several problems, which the geophone arrays and filters were developed to overcome. The main problem is ground roll, which has usually been attenuated by geophone arrays and low cut filters; another is other shot-generated noise which tends to be in the same frequency range as the convened wave signal. A discussion of the noise problems on converted wave data was presented at last years CREWES meeting in Banff (Miller et al., 1990). Analysis of some of the records from a multicomponent data set recorded during the 1990 Geophysics Field School revealed that the real limitation on signal recovery from these data was the dynamic range of the recording instruments. It is surprising how few seismic data processors have a good working knowledge of seismic instrumentation and its operation. Too often assumptions are made regarding dynamic range and signal resolution. A brief description of their operation is necessary to indicate their limitations.