Incorporated Research Institutions for Seismology (IRIS) Data Management Center (DMC) Data Products Team horizontal to vertical spectral ratio (HVSR)

2019-07-31 V.2019.212

---

DESCRIPTION:

The Horizontal to Vertical Spectral Ratio (HVSR) is a popular method that easily provides the predominant frequency at a given site. In this paper, we introduce the Incorporated Research Institutions for Seismology (IRIS) Data Management Center’s (DMC’s) HVSR station toolbox that is available to the community. This toolbox offers sundry ways to compute the ratio by providing different averaging routines. They go from the simple average of spectral ratios to the ratio of spectral averages. Computations take advantage of the available power spectral density and probability density function estimates of the ambient noise for the seismic stations and as such, can readily be used to estimate the predominant frequency of the many three-component seismic stations available from IRIS. To facilitate identification of the clear HVSR peaks and estimate of the predominant frequency of station sites, this toolbox further processes the results of HVSR analysis to detect and rank HVSR peaks.

This toolbox contains two tools (scripts) to assist with HVSR analysis: 1) the station baseline and 2) the HVSR tool.

• computeStationChannelBaseline.py uses IRIS DMC’s MUSTANG noise-pdf web service to compute channel-specific noise-baseline for a given station following the technique outlined by McNamara et al., 2009. The objective of this script is to provide an insight into station’s ambient seismic field power characteristics. For each channel, the computed baseline represents the probability density function (PDF) characteristics of the PSDs in the form of median, lower and higher percentiles of the available PSDs.

• computeHVSR.py - uses IRIS DMC’s MUSTANG noise-psd and noise-pdf web services to compute HVSR for a given station based on the MUSTANG hourly PSDs. HVSR estimates are obtained by converting hourly PSDs to median daily spectral powers and then HVSRs are computed using one of the 6 different available methods including the Diffuse Field Assumption method (DFA), as described by Sánchez-Sesma et al. (2011) or the average of spectral ratios, similar to the method used by McNamara et al. (2015). Options available include:

  • Remove PSDs that fall outside the station noise baseline as computed by computeStationChannelBaseline.py above (parameter: removeoutliers=0|1).

  • Compute HVSR using one of the methods below (parameter: method=1|2|3|4|5|6).

  • Output a peak rank report with ranking based on SESAME 2004 (not avaiable for DFA method)

  • The HVSR computational methods supported: (1) DFA, Diffuse Field Assumption method (Sánchez-Sesma et al., 2011) NOTE: The MUSTANG noise-psd web service Power Spectral Density estimate for seismic channels are computed using the algorithm outlined here: (http://service.iris.edu/mustang/noise-psd/docs/1/help/) This algorithm involves averaging and normalization that may result in smoothing of some of the peaks that may otherwise be observed by direct computation of FFT and DFA. With this smoothing, the DFA results tend to be closer to the vector summation method, method (4) below.

    or HVSR computation by combining the two horizontal components using one of the methods referenced by Albarello and Lunedei (2013): (2) arithmetic mean, H ≡ (HN + HE)/2 (3) geometric mean, H ≡ √HN · HE (4) vector summation, H ≡ √H2 N + H2 E (5) quadratic mean, H ≡ √(H2 N + H2 E )/2 (6) maximum horizontal value, H ≡ max {HN, HE}

CHANGES.txt
   - a text file containing the history of changes to this bundle

INSTALL.txt
   - installation notes

README.md
   - this file

bin/
   - scripts directory containing:
        + computeStationChannelBaseline.py (described above)
        + computeHVSR.py (described above)

param/
   - parameters directory containing:
        + getStationChannelBaseline_param.py - the configuration parameter file for the computeStationChannelBaseline.py script above
        + computeHVSR_param.py - the configuration parameter file for the computeHVSR.py script above

lib/
   - bundle library files:
        + fileLib.py - a collection of functions to work with files and directories
        + msgLib.py - a collection of functions to print messages

INSTALLATION:

see the INSTALL.txt file

USAGE:

getStationChannelBaseline.py net=netName sta=staName loc=locCode chan=chanCode start=2007-03-19 end=2008-10-28 plot=[0|1] plotnnm=[0|1]verbose=[0, 1] percentlow=[10] percenthigh=[90] xtype=[period,frequency]

net station network code sta station code loc station location code chan station channel code (separate multiple channel codes by comma); default: BHZ,BH1,BH2 xtype X-axis type; default: period start start date of the interval for which station baseline is computed (format YYYY-MM-DD). The start day begins at 00:00:00 UTC end end date of the interval for which station baseline is computed (format YYYY-MM-DD). The end day ends at 23:59:59 UTC

	NOTE: PSD segments will be limited to those starting between start (inclusive) and 
	end (exclusive) except when start and end are the same (in that case, the range will 
	cover start day only).

verbose Run in verbose mode to provide informative messages [0=no, 1=yes]; default:0 percentlow lowest percentile to compute (float); default 5 percenthigh Highest percentile to compute (float); default 90 plot plot values [0|1] plotnnm plot the New Noise Models [0|1], active if plot=1

computeHVSR.py net=netName sta=staName loc=locCode chan=chanCodes start=2013-01-01 end=2013-01-01 plot=[0, 1] plotbad=[0|1] plotpsd=[0|1] plotpdf=[0|1] plotnnm=[0|1] verbose=[0|1] ymax=[maximum Y value] xtype=[frequency|period] n=[number of segments] removeoutliers=[0|1] method=[1-6] showplot=[0|1]

net station network code sta station code loc station location code chan station channel code (separate multiple channel codes by comma); default: BHZ,BHN,BHE xtype X-axis type; default: frequency start start date of the interval for which HVSR to be computed (format YYYY-MM-DD). The start day begins at 00:00:00 UTC end end date of the interval for which station baseline is computed (format YYYY-MM-DD). The end day ends at 23:59:59 UTC

	NOTE: PSD segments will be limited to those starting between start (inclusive) and 
	end (exclusive) except when start and end are the same (in that case, the range will 
	cover start day only).

verbose Run in verbose mode to provide informative messages [0=no, 1=yes]; default:1 plotbad plot rejected PSDs (float) if "plotpsd" option is selected; default 0 plotnnm plot the New Noise Models [0|1], active if plot=1; default 1 plotpsd plot PSDs; default 0 plotpdf plot PSD\DFs; default 1 ymax maximum Y values; default -50 n break start-end interval into 'n' segments; default 1 removeoutliers remove PSDs that fall outside the station noise baseline; default 1 ymax mcompute HVSR using method (see above); default 4 showplot turn plot display on/off default is 1 (plot file is generated for both options)

EXAMPLES:

getStationChannelBaseline.py net=IU sta=ANMO loc=00 chan=BHZ start=2002-11-20 end=2008-11-20 plot=1 plotnnm=1 verbose=1 percentlow=10 percenthigh=90

getStationChannelBaseline.py net=TA sta=TCOL loc=-- chan=BHZ,BHN,BHE start=2013-01-01 end=2014-01-01 plot=1 plotnnm=1 verbose=1 percentlow=10 percenthigh=90

computeHVSR.py net=TA sta=TCOL loc=-- chan=BHZ,BHN,BHE start=2013-01-01 end=2013-01-01 plot=1 plotbad=0 plotpsd=0 plotpdf=1 verbose=1 ymax=5 xtype=frequency n=1 removeoutliers=0 method=4

computeHVSR.py net=TA sta=TCOL loc=-- chan=BHZ,BHN,BHE start=2013-01-01 end=2013-02-01 plot=1 plotbad=0 plotpsd=0 plotpdf=1 verbose=1 ymax=5 xtype=frequency n=1 removeoutliers=1 method=4

computeHVSR.py net=TA sta=M22K loc= chan=BHZ,BHN,BHE start=2017-01-01 end=2017-02-01 plot=1 plotbad=0 plotpsd=0 plotpdf=1 verbose=1 ymax=6 xtype=frequency n=1 removeoutliers=0 method=4

computeHVSR.py net=TA sta=E25K loc= chan=BHZ,BHN,BHE start=2017-01-01 end=2017-02-01 plot=1 plotbad=0 plotpsd=0 plotpdf=1 verbose=1 ymax=5 xtype=frequency n=1 removeoutliers=0 method=4

computeHVSR.py net=TA sta=E25K loc= chan=BHZ,BHN,BHE start=2017-07-01 end=2017-08-01 plot=1 plotbad=0 plotpsd=0 plotpdf=1 verbose=1 ymax=5 xtype=frequency n=1 removeoutliers=0 method=4

REFERENCES:

Albarello, Dario & Lunedei, Enrico. (2013). Combining horizontal ambient vibration components for H/V spectral ratio estimates. Geophysical Journal International. 194. 936-951. 10.1093/gji/ggt130.

Francisco J Sanchez-Sesma, Francisco & Rodriguez, Miguel & Iturraran-Viveros, Ursula & Luzon, Francisco & Campillo, Michel & Margerin, Ludovic & Garcia-Jerez, Antonio & Suarez, Martha & Santoyo, Miguel &

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.

Rodriguez-Castellanos, A. (2011). A theory for microtremor H/V spectral ratio: Application for a layered medium. Geophysical Journal International. 186. 221-225. 10.1111/j.1365-246X.2011.05064.x.

Guidelines for the Implementation of the H/V Spectral Ratio Technique on Ambient Vibrations, December 2004 Project No. EVG1-CT-2000-00026 SESAME. ftp://ftp.geo.uib.no/pub/seismo/SOFTWARE/SESAME/USER-GUIDELINES/SESAME-HV-User-Guidelines.pdf

HISTORY - 2019-07-31 Release R.1.0 - 2018-07-10: prerelease V.2018.191 - 2017-11-28: computeHVSR V.2017.332 - 2017-11-16: initial version V.2017.320

COMMENTS/QUESTIONS:

Please contact [email protected]