Yesterday I converted a petroleum industry passive recording in SEG-D format to the SAC format. This week-length seismogram was recorded at a 2 millisecond sample rate and contains about 300 million samples. We can expect to see significant growth is such industry activity and an opportunity to enhance IRIS research with such densely sampled (both temporal and areal sampling) data. However, the binary SAC header isn't fully up to handling these data in (at least) the following two ways:
1) 4 byte single precision floating point numbers, having only 23 bits of mantissa precision, cannot in general provide accurate values for B and E in the headers.
2) 4 byte single precision floating point numbers, having only 23 bits of mantissa precision, provide low accuracy LAT/LONG locations, good only to something like +/-50 meters. Industry passive areal surveys will typically have receiver station spacings of around 250 meters or less and record to better than 10 Hz and so need location accuracy at least an order of magnitude better for proper processing.
For any upcoming v7 SAC header, I recommend these fields, and, realistically, others such as DELTA, be extended to 64-bit double precision values.
- Stewart A. Levin
Consulting Faculty
Dept. of Geophysics
Stanford University
also
Society of Exploration Geophysicists Technical Standards co-chair
BTW, what is the representation of a Logical in SAC binary? All 1-bits for TRUE and all 0-bits for FALSE?
Dr. Stewart A. Levin
SAGE Facility has been
certified as a trustworthy data repository
by the CoreTrustSeal Standards and Certification Board
