Seismic Network Operations

As part of the annual calibration process, the USGS runs a sequence that includes a random, a step, and several sine wave calibrations.  The USGS analyzes the random binary calibration signal in order to estimate the instrument response.  The figures below show the results from the analysis of the most recent processed calibration at the station.

We use an iterative three-step method to estimate instrument response parameters (poles, zeros, sensitivity and gain) and their associated errors using random calibration signals. First, we solve a coarse non-linear inverse problem using a least squares grid search to yield a first approximation to the solution. This approach reduces the likelihood of poorly estimated parameters (a local-minimum solution) caused by noise in the calibration records and enhances algorithm convergence. Second, we iteratively solve a non-linear parameter estimation problem to obtain the least squares best-fit Laplace pole/zero/gain model. Third, by applying the central limit theorem we estimate the errors in this pole/zero model by solving the inverse problem at each frequency in a 2/3rds-octave band centered at each best-fit pole/zero frequency. This procedure yields error estimates of the 99% confidence interval.

1. Current Issues
   The KS54000 continues to have high noise one the vertical channel despite work done to address this in March 2012. The newly installed episensor appears to have high noise on the LN2 channel.
2. 2012-03-10
   The Episensor which had excessively noisy channels was replaced. The STS-2 was replaced with a STS-2.5.
3. 2007-11-15
   Upgraded to Q330 digitizer.

• IRIS/USGS GSN Program
• NOAA Pacific Tsunami Warning Center
• USGS GSN Program
• National Science Foundation
• Incorporated Research Institutions in Seismology

• Wake Atoll
• IRIS MetaData Aggregator