Skip to content
Seismograph Stations
Home Resources Overview of UUSS Data Processing and Analysis: 1981 – 2012

Overview of UUSS Data Processing and Analysis: 1981 – 2012

Since 1981, data from University of Utah analog telemetry stations have been digitally recorded on a computer system operating in an event detection mode. From January 1, 1981, to September 17, 1992, these data were recorded on a Digital Equipment Corporation PDP-11/34 computer. The PDP-11/34 system was capable of recording 128 data channels onto 9 track tape using code written by A. Bittenbinder at the University of Washington. The PDP-11/34 was replaced in 1992 by a Concurrent 7200 (Masscomp) computer. The Masscomp computer system is able to record 192 data channels onto disk but is currently configured to record 176 channels. The earthquake detection and recording software of both the PDP 11/34 and the Masscomp was closely modeled after the CEDAR system designed by C. Johnson (1979).

The Masscomp is equipped with a multiplexed A/D converter that samples incoming analog telemetry signals at approximately 100 samples/sec. Data acquisition software (HAWK), obtained from the University of Washington, continuously monitors all seismic signals. Triggering occurs when an abrupt change in the background signal is detected from a cluster or subnet of geographically close seismograph stations. When a seismic event is detected, the digitized signals for the entire network are recorded for the duration of the earthquake and are written to a file stored on the Masscomp’s internal hard disk. All triggers of the event detection system are automatically written to a 1/4″ cartridge tape backup on the Masscomp system. A time code trace is used to establish absolute timing.

Data from digital telemetry stations operated by the UUSS and other organizations, as well from UUSS analog telemetry stations, are continuously recorded using an Earthworm system (see http://folkworm.ceri.memphis.edu/ew-doc/ for more information). The Earthworm system runs on multiple machines, including SUN workstations and PCs. Data from the digital telemetry stations are not currently used in the event detection and triggering. However, digital telemetry data from broadband and short-period stations, and from strong-motion stations located on rock, are merged with the analog telemetry data after each trigger for use in routine data processing. Continuous waveform data from all UUSS analog and digital telemetry stations have been archived on a daily basis since June 2002 at the Earthscope Consortium data center (formerly IRIS Data Management Center; https://ds.iris.edu/ds/).

The basic philosophy of the routine data processing and analysis has remained fairly constant since 1981. The most significant changes to the procedures have been in the magnitude determination methods, as described in Summary of UUSS Magnitude Determinations: 1981-2006. Since 2002, the Earthworm system has been used to provide automatic near real time earthquake locations and magnitudes and maps of ground shaking. However, the results from this automatic processing are not currently integrated into the data stream for the routine analysis.

The following summary applies to the interactive data processing and analysis process in use at UUSS as of December 2006.

  1. Scanning: All event files are automatically transferred from the Masscomp to a SUN Microsystems server where they undergo further analysis. After each trigger occurs, a waveform plot of the triggered stations (up to a maximum of 24), the time code, and miscellaneous information about the trigger is automatically sent to a laser printer. These plots are examined to identify seismic events of interest.
  2. Timing: For events provisionally identified as local earthquakes within the UUSS network, P- and S-wave arrival times are measured as accurately as possible (with reading errors of ± 0.01sec for the best data) using the computer program “UPING”. This routine uses a SUN Workstation interactive graphics terminal to display seismic trace data from individual stations. Each channel may be scaled in either time or amplitude, band pass filtered to reduce site or transmission noise, and saved or deleted from the archive file at the option of the analyst. Arrival times and total signal durations are measured interactively and stored.
  3. Location procedure: The computer program HYPOINVERSE (Klein, 1978) is used for earthquake location as described in Computation of UUSS Earthquake Locations: 1981 – 2012. The hypocentral solution for each event is analyzed and checked for errors. Arrival times having large residuals are retimed. Additional data are sought for events located with a poor spatial distribution of stations. Small earthquakes that cannot be reliably located are excluded from further analysis.
  4. Coda Magnitude (MC) determination: Coda magnitudes are calculated by HYPOINVERSE, in combination with UUSS software, from vertical-component signal duration measurements made using UPING.
  5. Local Magnitude (ML) determination: UUSS-developed software (ML) is used to create horizontal-component synthetic Wood-Anderson seismograms from broadband and short-period digital telemetry data. From these seismograms, the software automatically measures maximum peak-to-peak amplitudes and uses them, along with the earthquake location, to calculate an ML. The amplitude picks are presented to the user for interactive review and revision as needed.
  6. Data archiving: Seismic waveform data for all seismic events recorded are archived on optical disk and 8-mm tape. For local earthquakes, arrival times, signal durations, first motion directions, and maximum amplitudes for local magnitude calculations are also archived on these media.
  7. Elimination of blasts and check for completeness: Located seismic events identifiable as blasts are deleted from the earthquake catalog. Blast identification is accomplished by contacting individual blasting operations and/or correlation with known blasting areas and the time of day of frequent blasting.
  8. Final processing: A final HYPOINVERSE run is made of all data to create a catalog summary of hypocenters and MCs. The final MLs are calculated using an EXCEL spreadsheet and added to the catalog for the earthquakes for which they are available.

References

Johnson, C. E. (1979). I. CEDAR—An approach to the computer automation of short-period seismic networks, II, Seismotectonics of the Imperial Valley of Southern California, Ph.D. Thesis, California Institute of Technology, Pasadena, California, 332 pp.

Klein, F. W. (1978). Hypocenter location program HYPOINVERSE, U. S. Geol. Surv., Open-File Rept. 78-694, 102 pp, doi:10.3133/ofr78694.