Data Services Newsletter

Volume 23 : No 1 : Spring 2021

The Upgraded IRIS DMC Noise Toolkit

The IRIS DMC Noise Toolkit (NTK) is a collection of programs to assist users with 1) computing Power Spectral Density (PSD) of station waveform data using customized parameters, 2) computing microseism energy from PSDs at different frequency bands and 3) performing frequency-dependent polarization analysis of waveform data.

First released in 2014 to facilitate customized ambient seismic noise processing, the NTK has been updated to support:

  • Python 3, the current major version of the language (Python 2 has been dropped in this release).

The repository for the NTK source codes and release versions has also been migrated to an IRIS GitHub repository with downloadable releases.

NTK Product overview:
https://ds.iris.edu/ds/products/noise-toolkit/

NTK release downloads:
https://github.com/iris-edu/noise-toolkit/releases

NTK GitHub repository:
https://github.com/iris-edu/noise-toolkit

NTK data product DOI:
https://doi.org/10.17611/dp/ntk.1

Examples:

The examples below are created using the upgraded NTK tools without any prior data download (see the Data and Resource section below for the commands used to generate these plots).

  • Figure 1. The BHZ PSD for the GEOFON station Vohitsoka, Madagascar (GE.VOI) created using the PDF/PSD package on one hour of BHZ data starting at 2020-11-04T12:00:00 UTC.
  • Figure 2a. A plot of the temporal variation of the median microseism energy (ME) values at the N4 (Central and Eastern US Network) station 545B (Wilberts Farm station, Edgard, LA, N4.545B) as the tropical storm Cristobal moved towards and passed over the station between June 4 and June 10, 2020 (Figure 2b). The ME plot was generated using NTK’s PSD and ME packages on the BHZ channel waveforms. The displayed ME corresponds to the secondary microseism band (SM, periods of 5 to 10 seconds).
  • Figure 3. A plot of the polarization attributes (Koper and Hawley, 2010) for the station GE.VOI using the Polarization attributes package on one hour of station data starting at 2020-11-04T12:00:00 UTC.

GE.VOI.BHZ PSD 2020-11-04T12:00:00
Figure 1. Computed BHZ PSD of the GE.VO station for one hour starting at 2020-11-04T12:00:00 UTC. The computed PSD is shown in blue and the smoothed PSD, using a quarter of an octave smoothing window, is shown in red. The dotted line represents Peterson’s (1993) new low‐noise model (NLNM) and new high‐noise model (NHNM). Click on image for a larger view.

ME representing the median microseismic power
Figure 2a. Microseism energy (ME) representing the median microseismic power values of the PSDs (BHZ channel) at period band of 5s to 10s (secondary microseism band, SM) for station N4.545B during the passage of the tropical storm Cristobal (Figure 2b). The resulting microseismic power values are smoothed using a moving 6 hour median window. Click on image for a larger view.

Cristobal path
Figure 2b. Map of the tropical storm Cristobal path and intensity in blue between June 4 to June 10, 2020 (data from NOAA Center for Satellite Applications and Research, STAR). Location of the the N4.545B station is shown in red. Click on image for a larger view.

GE.VOI Polarization 2020-11-04T12:00:00
Figure 3. Plot of power and polarization attributes obtained by processing one hour of BHZ, BHN, BHE data of the GEOFON Station Vohitsoka, Madagascar (GE.VOI) between 2020-11-04 12:00:00 UTC and 2020-11-04T13:00:00 UTC. Click on image for a larger view.

Acknowledgments

Special thanks go to Robert Anthony (USGS, Albuquerque Seismological Laboratory) for updating the code to use matplotlib’s CSD for PSD computations and to Timothy C. Bartholomaus of University of Idaho for converting the Polarization package to Python 3.

References

  • Koper K.D. and V.L. Hawley, “Frequency dependent polarization analysis of ambient seismic noise recorded at a broadband seismometer in the Central United States”, Earthquake Science, 23, 439-447, 2010.
  • Peterson, J. (1993). Observations and modeling of seismic background noise, U.S. Geological Survey open-file report (Vol. 93-322, p. 94). Albuquerque: U.S. Geological Survey.

Data and Resources

  • The NTK’s Wiki pages provide additional information on how to use NTK.
  • Figure 1: to generate this plot, a single call was made to the NTK’s PSD package:

ntk_computePSD.py net=GE sta=VOI loc=DASH chan=BHZ start=2020-11-04T12:00:00 end=2020-11-04T13:00:00 xtype=frequency plot=1

  • Figure 2: to generate the ME plot, you need a sequence of calls to the NTK’s PSD and ME packages:

1. Generate PSDs for every data day of this time period:
ntk_computePSD.py param=computePSD net=N4 sta=545B loc=00 chan=LHZ start=2020-06-04T00:00:00 end=2020-06-05T00:00:00 xtype=period plot=0 verbose=0
. . .
ntk_computePSD.py param=computePSD net=N4 sta=545B loc=00 chan=LHZ start=2020-06-09T00:00:00 end=2020-06-10T00:00:00 xtype=period plot=0 verbose=0

2. Combine PSDs:
ntk_extractPsdHour.py net=N4 sta=545B loc=00 chan=LHZ start=2020-06-04T00:00:00 end=2020-06-10T00:00:00 xtype=period verbose=0

3. Compute power using the combined PSDs:
ntk_computePower.py net=N4 sta=545B loc=00 chan=LHZ start=2020-06-04T00:00:00 end=2020-06-10T00:00:00 xtype=period file=N4.545B.00.LHZ.2020-06-04.2020-06-10.period.txt verbose=0

4. Compute 12-hour median power:
ntk_medianPower.py param=medianPower net=N4 sta=545B loc=00 chan=LHZ start=2020-06-04T00:00:00 end=2020-06-10T00:00:00 xtype=period win=12 file=N4.545B.00.LHZ.2020-06-04.2020-06-10.txt verbose=0

5. Plot the Secondary Microseism, SM:
ntk_plotPower.py param=plotPower param=plotPower net=N4 sta=545B loc=00 chan=LHZ start=2020-06-04T00:00:00 end=2020-06-10T00:00:00 win=12 bin=SM file=N4.545B.00.LHZ.2020-06-04.2020-06-10.12h.txt ymax=2.0 bin=SM

  • Figure 3: to generate this plot, a single call was made to the NTK’s Polarization package:
    ntk_computePolarization.py param=computePolarization net=GE sta=VOI loc=DASH start=2020-11-04T12:00:00 end=2020-11-04T13:00:00 type=frequency plot=1

by Data Products (IRIS DMC)

16:51:03 v.b4412d20