[data-issues] multiple segments from same station in SEED files

Sheila Peacock sheila at blacknest.gov.uk
Thu Feb 18 05:09:05 PST 2010

We are trying to develop software to write our data in
SEED files when the time span includes either a gap
in the data or a change of metadata (e.g. a change in
seismometer gain or orientation).

As far as I can see, there is no problem in principle
with a SEED volume containing several segments of
data from the same station/channel/location-code,
but I have not been able to find an example yet of a
SEED file containing more than one segment of data
per channel.  I did find one where there are two
(overlapping) segments from the same channel distinguished
only by the "data header/quality indicator" (p. 98 of manual)
being "R" for one and "Q" for the other.

Do list members know whether a SEED volume containing
two or more segments of data from the same station/channel/
location-code/quality-indicator can be decoded by
"rdseed" and other such programs, and would be acceptable to IRIS?

I am assuming that repeats of Blockette 52 (for
orientation changes), 53 (for pole-zero changes), and 58
(for channel sensitivity/gain changes) for each data segment
would be the correct way to denote changes in the
channel metadata, with items 22 and 23 of Blockette 52 holding
the start- and end-dates of the metadata being in force.
If not, then is the only way to track
the metadata change to issue a whole new SEED volume
at the time of the change?

I also assume that it would be incorrect to use items
22 and 23 of Blockette 52 just to denote the edges of gaps
in the data where there are no metadata changes, and that
the start time in the "Fixed section of Data Header" (p. 98)
should be used for that.

Advice would be most welcome, as "digging" blindly for SEED files
in the hope of finding one that shows what I'm looking
for would be time-consuming.


Sheila Peacock

More information about the data-issues mailing list