What's new at the International Deployment of Accelerometer's Data Collection Center
IDA staff are in the process of transforming the Data Collection Center’s (DCC) hardware architecture from a collection of Sun Microsystems servers running Solaris to a Linux cluster utilizing Linux Virtual Machine (VM) under Red Hat. In making this change, we are following the practices of the IRIS DMC and the USArray’s Array Network Facility (ANF), who have also adopted Linux VM as their central architecture. Currently all data received from II stations are fed into a buffer on the VM and immediately forwarded to the IRIS DMC for distribution to the community. Initial processing of the raw data into standard form and many quality assessment functions are still performed on the Solaris systems. Thus, we are temporarily running a hybrid system. Our goal is to transfer these remaining Solaris functions to the VM in the near future. When that change is complete, we will be left with a simpler system that we expect will be much easier to administer.
IDA staff are also updating the IDA website to have a new look and feel. The last major overhaul of the site was in 2002. The new website, which went public on January 24th, has been reorganized and modernized around a Drupal content management framework that should display better under a broader range of platforms. There is much work yet to complete our vision of what we wish to include in the site. In particular, we would like to make available some of the MUSTANG quality assessment information that IDA is integrating into the DCC’s quality control procedures.
An example of one IDA QC application that takes advantage of the DMC’s web service developments is shown in Figure 1.
Here are plotted seismograms from the recent Mw=8.2 Tarapaca, Chile earthquake as recorded on a three-component set of STS1s and an Episensor accelerometer at II station NNA (Ñaña, Peru). For stations so close to such large events (NNA is at an epicentral distance of 9.5 degrees), we wish to check for non-linear behavior on our weak-motion sensors, the STS1s in this example. All waveforms were retrieved from the DMC’s Seattle archive and read directly into Matlab © using
irisFetch. The data streams from the Episensor were then transformed to mimic the response of the STS1, and finally both sets of data were filtered to identical frequency passbands. That the pairs of traces overlay suggests to us that the STS1s were not overdriven by this event and also that the orientations of the STS1s relative to the Episensor are well determined. Lastly the scaling value for each trace (shown as an inset) indicates that the relative response of the pairs of components is close to the 1% level of accuracy desired by the GSN.
To refine knowledge of instrument response at each station, IDA is conducting a field campaign as part of the GSN’s data quality initiative. One part of that initiative, the use of portable instruments to calibrate the sensitivity of permanently deployed GSN sensors, is described on the above page as the SenSorLoc Kit. Figure 2 shows IDA technician Jim Conley as he uses the IDA SenSorLoc to calibrate instruments at II PALK (Palakele, Sri Lanka).
The APS differential GPS unit on the tripod in the foreground is used to orient the portable Compact Trillium 120 sensor placed on the floor of the recording building to the rear. A mirror assembly (see inset) is affixed to the Trillium to permit the sensor to be accurately oriented by the APS. A table of recent test dates is maintained on our web page.
by Pete Davis (IRIS/IDA)