This package provides S4 classes and functions for calculating metrics from seismological data available from EarthScope (https:/www.earthscope.org) (formerly IRIS) or data centers offering FDSN web services. This package is part of the MUSTANG project.
The IRISMustangMetrics package depends upon the IRISSeismic package which defines new S4 classes and methods for manipulating seismic data. Please see the "seismic-intro" vignette for introductory examples on using IRISSeismic.
In 2023, IRIS (Incorporated Research Institutions for Seismology) and UNAVCO merged to form EarthScope Consortium. IRIS (now EarthScope) webservices are unchanged but can now be accessed at https://service.earthscope.org as well as https://service.iris.edu.
Jonathan Callahan jonathan@mazamascience.com
version 2.4.7
(updated references from IRIS to EarthScope)
version 2.4.6
(updated email addresses)
(ISPAQUtils, add psd_corrected metric name to allow the ISPAQ tool to return uncorrected psds)
version 2.4.5
(typo fix for email address and updated link in documentation)
(updated error handling code to meet CRAN requirements)
version 2.4.4
(additional error handling for sample_rate_resp)
version 2.4.3
(improved error handling for functions that call IRIS web services)
(additional error handling for sample_rate_resp)
version 2.4.2
minor modification to ISPAQUtils.R for sample_rate_resp, sample_rate_channel
version 2.4.1
fix for max_range where trace length not evenly divided by window length
max_range now rounds window*samplerate and increment*samplerate to integer values
corrects typo error in ISPAQUtils.R
version 2.4.0
adds new metrics sample_rate_resp, sample_rate_channel, max_range
version 2.3.0
modifies PSD metrics for edge case involving metadata and trace start times
adds noCorrection option to PSDMetric, if noCorrection=TRUE it only returns uncorrected PSDs and not corrected PSD or PSD-derived metrics; default noCorrection=FALSE
version 2.2.0
removes dead_channel_exp, metric has been retired
renames pdf_aggregator in ISPAQUtils.R to pdf
version 2.1.3
fixed bug related to getGeneralValueMetrics,getMustangMetrics error handling
added pdf_aggregator to ISPAQUtils.R, for multi-day pdf plotting within ISPAQ
version 2.1.2
version 2.1.1 unintentionally removed the pct_above_nhnm metric. This version restores it.
made sample rate sanity rate check consistent between correlationMetric and crossCorrelationMetric
version 2.1.1
fixed bug in PSDMetrics that affected dead_channel_gsn results
version 2.1.0
transfer_function requires sample rates to be within a factor of 10 to avoid decimation effects on amplitude
transfer_function uses 7 order Chebyshev filter in the decimate function, to correct 1% error occurring with default 8 order Chebyshev
fixed bug in transfer_function trace start and end time comparisons
transfer_function when determining if sample rate < 1, round to 5 digits first
getGeneralValueMetrics added metric_error, ts_channel_continuity, ts_channel_up_time, ts_gap_length, ts_gap_length_total, ts_max_gap, ts_max_gap_total, ts_num_gaps, ts_num_gaps_total, ts_percent_availability, ts_percent_availability metrics
aliased the getGeneralValueMetrics function to getMustangMetrics
dailyDCOffsetMetric now returns error when result is NaN or NA
version 2.0.9
removed dependency on pracma package
removed channel restrictions for pct_above_nhnm,pct_below_nlm
cross correlation sampling rates of < 1 will round to 2 digits
getGeneralValueMetrics better handles case of no targets found
improved error handling in spikesMetric.R
version 2.0.8
minor bug fix to ISPAQUtils.R, spikes=numSpikes
version 2.0.7
fixed bug in getGeneralValueMetrics that didn't return measurements if there was more than one for any day
crossCorrelationMetric filter now defaults to a butterworth 2 pole 0.1Hz (10 second) low pass filter
version 2.0.6
fixed bug related to NA -> NULL replacement in Class-Metric
version 2.0.5
fixed dplyr version dependencies
version 2.0.4
adds additional sanity check to getGeneralValueMetrics()
createBssUrl() adds "&nodata=404" to url
version 2.0.2
updates to ISPAQUtils.R
version 2.0.1
removed dependency on tidyr package
version 2.0.0 -- GeneralValueMetrics
GeneralValueMetric class introduced, SingleValueMetric class deprecated. All metrics that previously returned
SingleValueMetric now return GeneralValueMetric
getGeneralValueMetrics() function added. Retreives metrics measurements from BSS database
crossCorrelationMetric() does not return timing_drift. The metric proved unreliable
users can now supply instrument response information in the form of frequency, amplitude, phase
to the function PSDMetric, in place of the getEvalresp webservice call
version 1.3.1 -- PSDs
getPsdMetrics reworked
version 1.3.0 -- latency
getLatencyValuesXML() removed from package.
documentation improvements.
additional error checking for getSingleValueMetrics().
version 1.2.7 -- PSDs
PSDMetrics() metrics percent_above_nhnm and percent_below_nlnm limited to frequencies less than nyquist/1.5.
version 1.2.6 -- PSDs
Depends on IRISSeismic (>= 1.3.0).
dead_channel_exp and dead_channel_lin metrics will only return values for station channel codes matching "BH|HH".
version 1.2.5 -- ISPAQUtils
ISPAQUtils.R contains functions for use with the ISPAQ standalone metrics system.
version 1.2.4 -- package version dependencies
Depends on IRISSeismic (>= 1.2.3). Imports seismicRoll (>=1.1.2).
version 1.2.2 -- correlationMetric tweak
correlationMetric() allows trace sample lengths to differ by 2 samples without stopping.
version 1.2.1 -- PSDs
Better fix to very low powers issue in PSDMetrics() dead_channel_gsn metric.
PSDMetrics() shifts PDF bin centers by 0.5 dB.
version 1.2.0 -- PSDs
PSDMetric() returns corrected PSD and PDF dataframes in addition to uncorrected PSDs and PSD derived metrics.
Depends on R (>= 3.2.0) and IRISSeismic (>=1.1.7).
Imports tidyr, dplyr.
version 1.1.3 -- bug fix, import version increased
Fixes typo in SNRMetric() function windowSecs argument default value.
Imports seismicRoll (>=1.1.1)
version 1.1.2 -- modifications
Improves error handling messages.
dailyDCOffsetMetric() removes unused selectivity argument and adds argument controlling output type.
Fixes bug in dailyDCOffsetMetrics() related to outlier removal and vector length.
Fixes bug in PSDMetrics() dead_channel_gsn metric related to very low power values.
PSDMetrics() only returns metrics that generate numeric values.
version 1.1.1 -- bug fix
crossCorrelationMetric() exits if either input trace is flatlined (all values equal).
version 1.1.0 -- updates package dependencies
Depends on IRISSeismic (>= 1.1.0).
version 1.0.8 -- new metric and bug fix
Improves error handling messages.
Adds new dead_channel_gsn metric to PSDMetric() function output.
Fixes bug in STALTAMetric() involving required trace length.
version 1.0.7 -- bug fix
Fixes issue with spikesMetric() passing argument values to findOutliers.
version 1.0.6 -- function argument changes
Changes spikesMetric() default argument values thresholdMin=10,selectivity=NA,fixedThreshold=TRUE.
transferFunctionMetric() now requires input of evalresp fap spectra, new arguments evalresp1 and evalresp2.
Additional sanity checks for transferFunctionMetric() and PSDMetric().
Depends on IRISSeismic (>= 1.0.10). Imports seismicRoll (>=1.1.0). Imports stats.
version 1.0.5 -- new PSD metric
Changes URL syntax for MUSTANG web services to use "format=..." instead of "output=...".
Adds new sample_unique metric to PSDMetric() output.
version 1.0.3 -- new functionality and bug fixes
Adds new metricList2DF() function.
Adds new dead_channel_lin metric to PSDMetric() output.
Fixes typo in Class-Metric.R value string format.
version 1.0.0 -- First Public Release
EarthScope web services: https://service.earthscope.org
# Open a connection to EarthScope webservices
iris <- new("IrisClient", debug=TRUE)
# Get the seismic data
starttime <- as.POSIXct("2010-02-27 06:45:00",tz="GMT")
endtime <- as.POSIXct("2010-02-27 07:45:00",tz="GMT")
result <- try(st <- getDataselect(iris,"IU","ANMO","00","BHZ",starttime,endtime))
if (class(result) == "try-error" ) {
print(geterrmessage())
} else {
# Apply a metric and show the results
metricList <- basicStatsMetric(st)
dummy <- lapply(metricList, show)
}
Run the code above in your browser using DataLab