The array instruments consisted of one-component and three-component sensors, and the data recording units were Reftek DAS 72A-06 (16-bit digitizer) and 72A-07 (32-bit), 72A-08 (16-bit), and 72A-02(16-bit). These included Reftek CPU Versions 2.45, 2.53, 2.53H, and 2.53J. Thus, the combination of a one-component sensor with any digitizer recorded data at 50 samples/sec, three-component with 16-bit digitization recorded 50 samples/sec, and three component with 32-bit digitization recorded 25 samples/sec. The sensor types included L4C3D 1Hz (3-component), L4C1D 2Hz (1-component), L4C3D 2Hz (3-component), and L22 2Hz (3-component) and the data were collected by exchanging recorders with internal disks or by copying to tape. Each instrument recorded data in record lengths of 30 minutes and each 30-minute interval was automatically written to the hard disk. The sensors were borrowed from the collections of UCLA, Los Alamos National Lab, the U.S.G.S., PASSCAL at Lamont-Doherty Geophysical Observatory, the Southern California Earthquake Center (SCEC), and USC.

The DAS 72A-06 and 72A-07 recorders had 230-Mb internal hard drives. The DAS 72A-02 and 72A-08 recorders had no internal drives. Either the external hard disks from these were swapped or the data was copied to tapes in the field using field Exabyte and DAT drives. All were powered by batteries which had to be changed each week. The sensors were interspersed so that no segment of the array would have only one type. However, all stations near the San Andreas fault were three-component.

Every station was visited approximately once a week. Operators recorded information related to timing receivers, power, temperature, and data acquisition in the log files for each DAS.


The timing of each instrument was controlled by GPS receivers (version 2.2), Omega receivers, or by the Reftek clock alone. GPS receivers were connected during installation and removal of the stations with no permanent external clocks in order to correct the internal Reftek clock. Those stations that did not receive external time pulses should be considered unreliable in timing. Some Omega stations were pulsed the first week with GPS receivers. Each station with a log file that indicates permanent time locks contains no time drift and, therefore, does not require a timing correction. Note that all of the Omega station data display a time error of exactly one second (due to incorrect leap second input) and some display, in addition, a 10 second error due to incorrect time pulses.

Most of the San Gabriel Mountain section stations were calibrated with GPS time receivers; the Mojave stations were calibrated with either GPS and/or Omega, or had no external timing receivers. Most of the stations with no permanent external receivers were located at the northern end of the array. All of these stations were supposed to be pulsed with GPS receivers at the beginning of the experiment; however, some were not. Corrections made for timing errors will be described in the section on data processing . The timing receivers were borrowed from PASSCAL at Lamont.