-
IRIS DMC: New Data Distribution Advisory
-
posted: 2010.05.19
This bulkmail is an advisory to all scientists and recipients who
access SEED format data from the IRIS DMC at http://www.iris.edu/
The IRIS DMC is committed to providing scientists with seismic data
of high quality. We are announcing an improvement in our customized
request processing which will assist with occasional overlapping
data issues.This enhancement will be put into operation in the next
few weeks. A new version of 'rdseed" software might be needed.
Executive summary:
* BREQ_FAST data requests will default to an 'M' quality merged product
to reduce time tears
* The DMC will begin shipping SEED volumes up to 30GB in size, (an
upper limit imposed by the SEED format), where the new rdseed version
5.0 required.
* Test volumes for you to download and use for testing (if you have
your own SEED software and don't use rdseed) can be found on http://www.iris.edu/pub/pickup
Background:
Beginning in 2002, the IRIS DMC has accepted data from recording
stations in two general forms, real-time and quality-controlled.
These are tagged in SEED records as 'R' and 'Q' quality respectively.
(Older data has retained the original 'D' designation when such
distinctions were not available.)
A complication that arises is that occasionally users receive data
of two separate qualities for the same time period. Depending on
the software used to read the SEED data this could result in unnecessary
time tears (apparent gaps and overlaps). What is needed is a way
to deliver a single, post-archiving quality of data that brought
these disparate waveforms together in a seamless, sample-level join.
The benefit to the user will be more continuous data, enabling easier
data processing and less ambiguity.
What's New:
The DMC has developed procedures to create, on-the-fly, the most
continuous data series possible from the highest quality data available
in the DMC archive. In order to deal with the problems associated
with multiple quality indicators and to differentiate these data
from the SEED records supplied by the network operators, all data
that has gone through this process will be marked as 'M' quality
data (M standing for merged). The new process for creating 'M'
quality data is to join archival 'Q','D', and 'R' data at the sample
level such that a continuous waveform results. The 'M' quality
data will NOT replace the original data in the archive as we consider
it of utmost importance to preserve the original recordings as
delivered. Instead, 'M' quality will be generated dynamically and
will become the default. We will always give users the option to
specifically request one of the original waveform streams (Q, D,
or R) when desired.
Our first rollout of 'M' quality data will be for our BREQ_FAST
email request mechanism. When users specify a .QUALITY of 'B' (best),
they will receive an 'M' quality data stream.
Though there may be some software programs that cannot read this
data quality yet, users can rest assured that the most recent release
of rdseed v4.8 and later can recognize and process 'M' quality data.
Other software at the IRIS DMC has also been adapted to 'M' quality.
Another change occurring in the very near future is that the DMC
will be producing SEED volumes up to 30GB in size instead of splitting
large shipments into 2GB increments. In order to use such large files
rdseed users will need to upgrade to a new 64-bit version 5.0 rdseed
which will be released before large shipments commence.
Thank you-
IRIS DMC Operations
Questions: engine_room AT iris.washington.edu
|
