InfraMonitor 3.

1 Users Guide

Users Guide

InfraMonitor 3.1

Stephen Arrowsmith
([email protected]) and Omar
Marcillo ([email protected])

Last Edited: 05/02/13

Page 1
InfraMonitor 3.1 Users Guide

Table of Contents
Overview.................................................................................................................................................... 2
Algorithms.............................................................................................................................................2
InfraMonitor detection algorithm: The Adaptive F-detector............................................................ 2
InfraMonitor association algorithm.................................................................................................. 3
InfraMonitor location algorithm: The Bayesian Infrasound Source Locator (BISL).......................3
Code.......................................................................................................................................................5
Installation..................................................................................................................................................5
Pipeline processing using InfraMonitor 3.1............................................................................................... 6
Additional Pipeline Processing Utilities..............................................................................................12
Interactive Processing using InfraMonitor 3.1.........................................................................................12
Ray tracing using InfraMonitor 3.1..........................................................................................................18
Input Parameters...................................................................................................................................... 22
File Input/Output......................................................................................................................................22
Under the Hood........................................................................................................................................ 23
References................................................................................................................................................ 27
APPENDIX 1: Change Log..................................................................................................................... 28

Overview
InfraMonitor is a Matlab-based tool for infrasound monitoring that incorporates LANL-developed
detection, association, and location algorthms (see descriptions below) in addition to a Matlab
implementation of the Tau-P raytracing technique (Garces et al., 1998, Drob et al., 2010). The
InfraMonitor package has been developed to run in two different modes: (1) As a pipeline process, and
(2) interactively. The pipeline mode is designed for processing large quantities of data, without user
interaction, in order to detect and locate infrasound events. The input for pipeline mode comprises data
from multiple infrasound arrays, and the output comprises an event catalog (containing calculated
origin times and locations). The interactive mode is designed for a more thorough user-driven analysis
of individual events (or short time intervals), providing a suite of tools that aid in the analysis of
infrasound data. Interactive tools include support for reading different data formats (SAC, CSS),
filtering, spectra, spectrograms, array plots, display of ground-truth, arrival time picking, amplitude
measurements, and ray tracing.

Algorithms

InfraMonitor detection algorithm: The Adaptive F-detector

The Adaptive F-detector (AFD) technique is outlined by Arrowsmith et al. (2008, 2009). In contrast
with other routinely used infrasound detectors, the AFD utilizes a contextual detection hypothesis that
adaptively accounts for temporally variable correlated ambient noise (see Figure 1). This technique
reduces the number of false alarms caused by background noise sources such as ocean noise, wind
farms, and other continuous wave sources.

Page 2
InfraMonitor 3.1 Users Guide

InfraMonitor association algorithm

The association technique uses a grid search approach to relate arrivals at multiple arrays (Arrowsmith
et al., 2008). InfraMonitor searches for groups of arrivals with backazimuths and inter-array delay
times that are consistent with each grid node. When such groups of arrivals are identified, they are
associated and input to the localization procedure described below (see Figure 2).

InfraMonitor location algorithm: The Bayesian Infrasound Source Locator (BISL)

The BISL technique (Modrak et al., 2010) uses a statistical approach for infrasound location that can
appropriately account for both model and measurement uncertainties inherent to infrasound. Phase
identification is replaced by the use of a Bayesian prior on group velocity, with both arrival times and
backazimuths contributing to the estimation of a location polygon for a specified confidence level (see
Figure 3).

Figure 1: Illustration of the AFD method in comparison with a conventional F-detector for
infrasound data containing five transient signals from mine shots but contaminated by wind farm
noise. The conventional detector identifies the full data segment as a detection (represented by
shaded gray areas on the right panel) whereas the AFD only detects the mining explosion signals.

Page 3
InfraMonitor 3.1 Users Guide

Figure 2: Illustration of the association algorithm. Gray areas represent grid nodes associated with
test events (stars) at three arrays.

Figure 3: An illustration of the application of BISL to a rocket motor detonation conducted at the
Utah Test and Training Range and recorded by three arrays. The location polygons are calculated
by taking slices through the posterior PDF (shown as a color plot).

Page 4
InfraMonitor 3.1 Users Guide

Code
InfraMonitor 3.1 is a Matlab GUI interface and toolbox for regional infrasound monitoring. Features
include:
• Array detection (Adaptive F-detector)
• Interactive Signal Analysis
• Signal Association
• Event location
• BISL method
• Polygon location method
• Tau-P infrasound ray-tracing model
• Graphical display and Interactive GUI interface
• Command-line alternatives to GUI functionality

The full InfraMonitor 3.1 package comprises Matlab scripts, Matlab GUI templates, Python scripts,
SAC data, CSS3.0 tables, and a users manual (pdf file). This distribution does not contain binary
(compiled) code.
The following supporting software is needed to run this tool:
• Matlab
◦ Signal Processing Toolbox
◦ Statistics Toolbox
• Python

Installation
Note that InfraMonitor 3.1 has been tested on Mac OS-X, Sun Solaris, and Linux operating systems.
The current version is not supported for Microsoft Windows.

The following files are provided as part of the InfraMonitor 3.1 package: InfraMonitor_3.1.tar.gz,
ExampleData.tar.gz
To get started, uncompress and untar the source folder:
gunzip InfraMonitor_3.1.tar.gz
tar xvf InfraMonitor_3.1

This will generate a folder (the root folder, the whole path to which is referred to hereafter as
<InfraROOT>). Add <InfraROOT> to your Matlab path:
addpath(genpath(<InfraROOT>))
The best idea is to add this line to your Matlab startup.m file (see the Matlab documentation for more
details)

Uncompress and untar the example data folder (this can be in a totally separate location from the
source code):
gunzip ExampleData.tar.gz

Page 5
InfraMonitor 3.1 Users Guide

tar xvf ExampleData.tar

Now, enter Matlab and type InfraMonitor3 (or select InfraMonitor from Matlabʼs
start menu).

Pipeline processing using InfraMonitor 3.1

Running InfraMonitor 3.1 in pipeline mode is based on two separate functions. To test these with the
ExampleData folder, run:
InfraMonitor_Detector('.','utah.wfdisc',[1 5],2,'css3.0');
This processes all the data defined in a wfdisc file for detections, outputting .arrival and .detect files. A
site file with the same name as the wfdisc file must exist in the same location as the wfdisc file. Both
wfdisc and site files must conform to either the CSS3.0 or NNSA formats, provided below.
Input parameters to InfraMonitor_Detector are, in order:
• Path to the wfdisc file (Use '.' if calling function from the directory containing the wfdisc file)
• Name of wfdisc file
• Frequency band [f_min f_max]
• Filter order (InfraMonitor uses the Butterworth filter in the Signal Processing Toolbox)
• Schema type (Use either 'css3.0' or 'nnsa')
The precise NNSA and CSS3.0 formats are defined in the following four pages for the wfdisc and site
files. InfraMonitor uses the refsta parameter to associate individual elements into arrays. It is strongly
recommended that the user populates the refsta parameter (see utah.site in the example dataset).
The output files include a .detect file and .arrival file. The .detect file has the following (free-form
format):
• Station name
• start date of detection
• start time of detection
• end date of detection
• end time of detection
• backazimuth
• trace velocity (km/s)
• Cross-correlation coefficient.

Page 6
InfraMonitor 3.1 Users Guide

The .arrival file conforms to the CSS3.0 arrival format, specified below (NA values are important).
STORAGE CHAR NA
# COLUMN EXT FORMAT DESCRIPTION
TYPE POS VALUE
Not
1 sta varchar2(6) a6 1-6 station identifier
Allowed
Not
2 time float(53) f17.5 8-24 epoch time
Allowed
Not
3 arid number(8) i8 26-33 arrival id
Allowed
4 jdate number(8) i8 35-42 -1 julian date
5 stassid number(8) i8 44-51 -1 stassoc id
6 chanid number(8) i8 53-60 -1 instrument id
7 chan varchar2(8) a8 62-69 - channel code
8 iphase varchar2(8) a8 71-78 - reported phase
9 stype varchar2(1) a1 80-80 - signal type
10 deltim float(24) f6.3 82-87 -1.0 delta time
11 azimuth float(24) f7.2 89-95 -1.0 observed azimuth
12 delaz float(24) f7.2 97-103 -1.0 delta azimuth
13 slow float(24) f7.2 105-111 -1.0 observed slowness (s/deg)
14 delslo float(24) f7.2 113-119 -1.0 delta slowness
15 ema float(24) f7.2 121-127 -1.0 emergence angle
16 rect float(24) f7.3 129-135 -1.0 rectilinearity
amplitude, instrument
17 amp float(24) f10.1 137-146 -1.0
corrected, nm
18 per float(24) f7.2 148-154 -1.0 period
19 logat float(24) f7.2 156-162 -999.0 log (amp/per)
20 clip varchar2(1) a1 164-164 - clipped flag
21 fm varchar2(2) a2 166-167 - first motion
22 snr float(24) f10.2 169-178 -1.0 signal to noise ratio
23 qual varchar2(1) a1 180-180 - signal onset quality
24 auth varchar2(15) a15 182-196 - source/originator
25 commid number(8) i8 198-205 -1 comment id
a17:YY/MM/DD Not
26 lddate date 207-223 load date
HH24:MI:SS Allowed

Page 7
InfraMonitor 3.1 Users Guide

NNSA, wfdisc file:

STORAGE CHAR
# COLUMN EXT FORMAT NA VALUE DESCRIPTION
TYPE POS
1 sta varchar2(6) a6 1-6 Not Allowed station code
2 chan varchar2(8) a8 8-15 Not Allowed channel code
3 time float(53) f17.5 17-33 Not Allowed epoch time
4 wfid number(9) i9 35-43 Not Allowed waveform identifier
5 chanid number(8) i8 45-52 -1 channel identifier
6 jdate number(8) i8 54-61 -1 julian date
7 endtime float(53) f17.5 63-79 9999999999.999 ending time
8 nsamp number(8) i8 81-88 Not Allowed number of samples
9 samprate float(24) f11.7 90-100 Not Allowed sampling rate
10 calib float(24) f16.6 102-117 1 calibration factor
11 calper float(24) f16.6 119-134 -1 calibration period
12 instype varchar2(6) a6 136-141 - instrument type
13 segtype varchar2(1) a1 143-143 - segment type
14 datatype varchar2(2) a2 145-146 - numeric data storage
15 clip varchar2(1) a1 148-148 - clipped data flag
16 dir varchar2(64) a64 150-213 Not Allowed directory
17 dfile varchar2(32) a32 215-246 Not Allowed data file
byte offset of data
18 foff number(10) i10 248-257 Not Allowed
segment within file
19 commid number(9) i9 259-267 -1 comment identifier
a17:YY/MM/DD
20 lddate date 269-285 Not Allowed load date
HH24:MI:SS

Page 8
InfraMonitor 3.1 Users Guide

NNSA, site file:

STORAGE CHAR NA
# COLUMN EXT FORMAT DESCRIPTION
TYPE POS VALUE
Not
1 sta varchar2(6) a6 1-6 station code
Allowed
2 ondate number(8) i8 8-15 -1 turn on date
3 offdate number(8) i8 17-24 2286324 turn off date
4 lat float(53) f11.6 26-36 -999 geographic latitude
5 lon float(53) f11.6 38-48 -999 geographic longitude
6 elev float(24) f9.4 50-58 -999 elevation
7 staname varchar2(50) a50 60-109 - station name/description
station type (single station,
8 statype varchar2(4) a4 111-114 -
array)
reference station for array
9 refsta varchar2(6) a6 116-121 -
members
Not north offset from array
10 dnorth float(24) f9.4 123-131
Allowed reference (km)
Not east offset from array
11 deast float(24) f9.4 133-141
Allowed reference (km)
a17:YY/MM/DD Not
12 lddate date 143-159 load date
HH24:MI:SS Allowed

Page 9
InfraMonitor 3.1 Users Guide

CSS3.0, wfdisc file:

STORAGE CHAR
# COLUMN EXT FORMAT NA VALUE DESCRIPTION
TYPE POS
1 sta varchar2(6) a6 1-6 Not Allowed station identifier
2 chan varchar2(8) a8 8-15 Not Allowed channel code
3 time float(53) f17.5 17-33 Not Allowed epoch time
4 wfid number(8) i8 35-42 Not Allowed waveform id
5 chanid number(8) i8 44-51 -1 instrument id
6 jdate number(8) i8 53-60 -1 julian date
7 endtime float(53) f17.5 62-78 +9999999999.999 end epoch time
8 nsamp number(8) i8 80-87 Not Allowed number of samples
sampling rate in
9 samprate float(24) f11.7 89-99 Not Allowed
samples/second
10 calib float(24) f16.6 101-116 Not Allowed nominal calibration
11 calper float(24) f16.6 118-133 Not Allowed calibration period
12 instype varchar2(6) a6 135-140 - instrument type
13 segtype varchar2(1) a1 142-142 - indexing method
14 datatype varchar2(2) a2 144-145 - numeric storage
15 clip varchar2(1) a1 147-147 - clipped flag
16 dir varchar2(64) a64 149-212 Not Allowed directory
17 dfile varchar2(32) a32 214-245 Not Allowed data file
18 foff number(10) i10 247-256 Not Allowed byte offset
19 commid number(8) i8 258-265 -1 comment id
a17:YY/MM/DD
20 lddate date 267-283 Not Allowed load date
HH24:MI:SS

Page 10
InfraMonitor 3.1 Users Guide

CSS3.0, site file:

STORAGE CHAR NA
# COLUMN EXT FORMAT DESCRIPTION
TYPE POS VALUE
Not
1 sta varchar2(6) a6 1-6 station identifier
Allowed
Not
2 ondate number(8) i8 8-15 Julian start date
Allowed
3 offdate number(8) i8 17-24 -1 Julian off date
4 lat float(24) f9.4 26-34 -999.0 estimated latitude
5 lon float(24) f9.4 36-44 -999.0 estimated longitude
6 elev float(24) f9.4 46-54 -999.0 elevation
7 staname varchar2(50) a50 56-105 - station description
station type: single station,
8 statype varchar2(4) a4 107-110 -
virt. array, etc.
reference station for array
9 refsta varchar2(6) a6 112-117 -
members
Not offset from array reference
10 dnorth float(24) f9.4 119-127
Allowed (km)
Not offset from array reference
11 deast float(24) f9.4 129-137
Allowed (km)
a17:YY/MM/DD Not
12 lddate date 139-155 load date
HH24:MI:SS Allowed

Page 11
InfraMonitor 3.1 Users Guide

In InfraMonitor 3.1 the detector and associator/locator are made up of separate functions. This provides
flexibility for running these functions using your own scripts.

To run the association/location routines for the example event, run:

InfraMonitor_Locator('utah',[37.0 43.0 -115.0 -109.0],[0.1 0.1],4)

Input parameters to InfraMonitor_Locator are, in order:

• Database name (name of the wfdisc and site files, without any extension added)
• Grid bounds [min_lat max_lat min_lon max_lon]
• Grid spacing [min_spac max_spac]
• Minimum number of arrays for association.
The output files include a .origin file, a .IMassoc, file and a .origin file. The .IMassoc file and .IMpoly
files have the following formats:
Table Extension Description How to generate
IMassoc Associates location polygons (IMPoly table) Automatically generated
with events (origin table) by location routine
Format:
%17.5f (Arrival time) %8d (EventID/PolyID)
%8d (ArrayID)
IMpoly Contains information on location polygons Automatically generated
Format: by location routine
%9d (PolyID) %9d (PointID)
%11.6f (lat) %11.6f (lon)

Page 12
InfraMonitor 3.1 Users Guide

The .origin file uses the CSS3.0 origin format, provided below.

STORAGE CHAR NA
# COLUMN EXT FORMAT DESCRIPTION
TYPE POS VALUE
1 lat float(24) f9.4 1-9 -999.0 estimated latitude
2 lon float(24) f9.4 11-19 -999.0 estimated longitude
3 depth float(24) f9.4 21-29 -999.0 estimated depth
Not
4 time float(53) f17.5 31-47 epoch time
Allowed
Not
5 orid number(8) i8 49-56 origin id
Allowed
6 evid number(8) i8 58-65 -1 event id
7 jdate number(8) i8 67-74 -1 julian date
number of associated
8 nass number(4) i4 76-79 -1
phases
number of locating
9 ndef number(4) i4 81-84 -1
phases
10 ndp number(4) i4 86-89 -1 number of depth phases
geographic region
11 grn number(8) i8 91-98 -1
number
12 srn number(8) i8 100-107 -1 seismic region number
13 etype varchar2(7) a7 109-115 - event type
estimated depth from
14 depdp float(24) f9.4 117-125 -999.0
depth phases
Not
15 dtype varchar2(1) a1 127-127 depth method used
Allowed
16 mb float(24) f7.2 129-135 -999.0 body wave magnitude
17 mbid number(8) i8 137-144 -1 mb magid
18 ms float(24) f7.2 146-152 -999.0 surface wave magnitude
19 msid number(8) i8 154-161 -1 ms magid
20 ml float(24) f7.2 163-169 -999.0 local magnitude
21 mlid number(8) i8 171-178 -1 ml magid
22 algorithm varchar2(15) a15 180-194 - location algorithm used
23 auth varchar2(15) a15 196-210 - source/originator
24 commid number(8) i8 212-219 -1 comment id
a17:YY/MM/DD Not
25 lddate date 221-237 load date
HH24:MI:SS Allowed

Page 13
InfraMonitor 3.1 Users Guide

Additional Pipeline Processing Utilities

A set of Python scripts are provided in the PYTHON subdirectory for performing various useful
processing functions in analyzing pipeline results. These Python scripts are provided as examples only
and will likely take some tweaking to get these working on your system. These utilities are summarized
below:
1. Scripts for MakeWfdisc.py Make a wfdisc file from a bunch of SAC files
pre- MakeSite.py Make a site file from a wfdisc file and SAC files
processing RobustSiteReader.py Formats a wfdisc file with nnsa format (which
InfraMonitor can read)
RobustWfdiscReader.py Formats a site file with nnsa format (which
InfraMonitor can read)
2. Scripts for CleanEvents.py Removes events from IMassoc, IMpoly, origin tables
post- with duplicate arrivals
processing CatTables.py Concatenates a bunch of separate IMassoc, IMpoly,
origin tables
RemoveEvents.py Removes events from IMassoc, IMpoly, origin tables
based on the area of uncertainty
MakePolygonKML.py Makes a Google Earth kml file for a given IMpoly table

For additional help on using these scripts, type the script name with the -h option from your UNIX
terminal window.

Interactive Processing using InfraMonitor 3.1

This tutorial guides you through using the InfraMonitor GUI using the example dataset included with
this package. The goal of the tutorial is not to cover every detail, but to provide a broad introduction to
using InfraMonitor 3.1 and to some of the key features.
Reading in and filtering data:
• Change into the ExampleData directory that you created in the installation. Open Matlab and
run InfraMonitor3.
• Select File>Open>CSS3.0 from the menu bar.
• Select utah.wfdisc in the file dialog box and click the Open button.
• In the dialog box that comes up highlight all the traces by clicking and dragging, and select OK
(you should note that there are 6 four-channel arrays – i.e., 24 channels – with 1 hour of data
per waveform).
• Once the data are read in, they are displayed in the main InfraMonitor window as white traces
on a black background. The labels on the left hand side denote the channel names. Individual
arrays are not explicitly highlighted but can be identified because they typically have the same
channel names (with different digits appended).
• Select Tools>Filtering>Apply on the menu bar.

Page 14
InfraMonitor 3.1 Users Guide

• Select No to the Filter only selected traces? dialog and click OK.
• Use the default filter coefficients (Filter band = 1 – 5 Hz, Order = 4) by clicking OK in the
Filtering dialog box. Wait until the screen refreshes with the filtered data.
Processing data using the adaptive F-detector:
• Process all the data using the Adaptive F-detector by selecting Detect>Process Everything from
the menubar. This will cycle through all the arrays that are read into InfraMonitor and add
detections.
• If you just wish to process a single array, you first need to select the array. Imagine you wish to
select only BGU (Channels 1-4). You can do this in two different ways. Either press 's' on your
keyboard and then click and drag over the waveforms associated with channels 1-4, or click on
the Select button on the screen and choose channels 1-4 (some people find this second step
easier on certain operating systems). Once channels 1-4 are highlighted, they will appear red
in the main GUI window (see Figure 4). You could now process this array by selecting
Detect>Process Array from the menu bar.
• While the detector is running, a text-based progress bar will appear in the Matlab Command
Window (it will appear as though nothing is happening in the InfraMonitor window!). When the
detector has finished running, a pop-up box will display. Detections are denoted by gray boxes.

Figure 5. After running the detector on all 6 arrays.

• When the detector has completed processing all arrays your window should look like Figure 5.
Select Detect>View from the menu bar and select WMU1 as the channel to plot in the popup
window that opens. A plot (entitled IMPlot) containing the detection processing results will
appear (detections are denoted by gray boxes).

Page 15
InfraMonitor 3.1 Users Guide

Analyzing the detection results:

• The detection plot window (IMPlot) has a number of different functions enabling you to
analyze the detection results. Try some of these features out now. Start by clicking the Extract
Data button, then click on a point on the time axis that you are interested in (e.g., the highest
adaptive F-statistic point). An FK plot will appear, and the azimuth, correlation, and time will
be displayed in the IMPlot display (Note: the trace velocity is also output in the Matlab
command window).
• To zoom in on the IMPlot window, click the x-zoom button then click and drag over the time
period you wish to zoom over. To return to displaying the full waveform, click the Refresh
button (Figure 6).
• To zoom on the y-axis, click the y-zoom button then click and drag over the y limits you want
to zoom on using whichever panel you want to zoom on. To zoom on a different panel, click the
y-zoom button again and repeat the process for that panel.
• To magnify the waveform, click on the Magnification button and enter a magnification factor
(e.g., 10). This enables you to view small signals that may be saturated by a larger signal
(similar to the y-zoom feature).
• There are three features that enable you to measure amplitudes. These features assume that the
data have been instrument corrected to physical units, however, you may simply input a
nominal calibration factor into the yellow box titled Conv Fact.
◦ To measure an RMS signal amplitude, click on a point in time centered on your signal (the
RMS amplitude in a 10 s window centered on the signal will be displayed).
◦ To measure the RMS noise amplitude, click on the RMS noise button and then click at the
start, then the end of a noise interval.
◦ To measure the peak to peak amplitude and period, zoom in on the signal of interest, click
the Amp/Period button, a window will display. Select a peak, then a trough, then a peak to
obtain measurements of the peak-to-peak amplitude and period.

Page 16
InfraMonitor 3.1 Users Guide

Figure 6. IMPlot window showing adaptive F-detector detections as gray boxes.

Modifying the input parameters:
• Select Tools>Preferences from the menu bar on the main InfraMonitor window.
• The preferences window is broken down into three categories: Detector parameters, Associator
parameters, BISL Locator parameters. These parameters are discussed in further detail in the
Input Parameters section. For now, try changing the detection parameters and re-running the
detector on a single array to get a feel for the effect of the input parameters on the results.
Some other features before we get to associating and locating events:
• With the channels for WMU highlighted (Figure 5) select Tools>Plot>Array to plot a map of
the array configuration. In the figure that appears, select the data cursor tool and click on one of
the points to get the array element name associated with that point.
• Now select Tools>Plot>Spectrogram with the same four channels highlighted, choose an nfft
equal to 512, click on a start point, then an end point for plotting the spectrograms. Four plots

Page 17
InfraMonitor 3.1 Users Guide

will appear – one for each waveform that is selected.

• Click on Tools>GT and enter the parameters as shown in Figure 7.

Figure 7. Ground truth parameters for example data.

• Lines should appear on the main window corresponding to predicted arrival times for seismic
arrivals (red, magenta, and purple lines) and infrasound (green lines denote the predicted time
window for this event).
• Try viewing the detection results again. The ground truth predicted backazimuth is denoted by
the horizontal gray bar in the azimuth window.
Running the associator and locator:
• Select Locate>Grid>New and enter the parameters as shown in Figure 8.

Figure 8. Generating a grid for a rectangular region to use for associating and locating events.
• Select Locate>Run>BISL, and use the default minimum number of arrays of 3.
• Some progress bars will appear in the Matlab command window, once the association and
location processing has finished, a dialog box will appear.
• Select Locate>View>BISL and click on the Select all button followed by OK.
• Select yes in the Backazimuths dialog box and click OK.

Page 18
InfraMonitor 3.1 Users Guide

• Two map plots – one for each event found – will be displayed. The ground-truth location you
entered previously is denoted by the red star. You should get a solution corresponding to the
example in Figure 9.

Figure 9. InfraMonitor localization solution for the ground-truth event in the example dataset.
• This example illustrates an important point because one false alarm is obtained at three stations
near the boundary of the box. One must be careful with associations where the network
resolution is poor (i.e., all arrays are in the same direction) and with only a few stations. As the
number of stations associated with the event increases, and the network resolution improves,
one can have more confidence in the solution (e.g., Figure 9 would indicate a highly confident
solution).
Introducing the output files:
• The InfraMonitor GUI writes out some output files as you go along (Note: to write out an
arrival file you must manually select Detect>Save from the file menu). Other output files that
are written out automatically include an origin file, an IMassoc file, and an IMpoly file. The
arrival and origin files conform to the CSS3.0 schema guidelines, the IMassoc and IMpoly files
conform to a structure unique to InfraMonitor 3.1 (see the section below entitled File
Input/Output). All output files are written into the current Matlab working directory (check in
your directory to make sure they are there).
Saving the configuration:
• Select File>Save and save the configuration as Config.mat (you will use this later). Saving a
configuration allows you to load your InfraMonitor session as it was without repeating all the
same steps as before. To open a configuration file, select File>Open>Configuration.

Page 19
InfraMonitor 3.1 Users Guide

Ray tracing using InfraMonitor 3.1

InfraMonitor 3.1 incorporates the Tau-P prediction model.
• To illustrate the implementation of Tau-P, open InfraMonitor3 in Matlab and load the
configuration you saved earlier (Config.mat) by selecting File>Open>Configuration 1.
• Change the grid file coordinates by selecting Locate>Grid>New and inputting the following
parameters:
◦ Grid limits: 36,46,-120,-108
◦ Grid spacing: 0.1,0.1
◦ Output file: utah.mat
• Select Predict>Bounce Points from the menu bar and select the met file included in the
Example data directory – exampleProfile.met.
• A met file is a simple 1D atmospheric profile with the following format: Elevation (km),
Temperature (K), Zonal wind (+E) (m/s), Meridional wind (+N) (m/s)
• After the run is finished, the plot shown in Figure 10 should appear, providing predicted bounce
points color coded by group velocity (or celerity) for the ground-truth event parameters that you
provided earlier.

Figure 10. Predicted bounce points for the example ground-truth event, color coded by group
velocity.
1 Your configuration must include ground-truth information in addition to a grid file generated by Locate>Grid>New

Page 20
InfraMonitor 3.1 Users Guide

A much more complete set of raytracing functions can be obtained by using the InfraTaup GUI
interface (run in Matlab using the command: infrataup).

For any run using InfraTaup, the following basic parameters should be set: Met-file, Resolution,
Azimuth, Launch angle, Range
• Met-file: This should be set to the name of an appropriate ASCII file in the current (working)
directory.
• Resolution: Ensure that the resolution corresponds to the vertical sampling in your met-file (but
note that it should be entered in meters).
• Azimuth: The azimuth in degrees clockwise from North. For shooting rays at multiple
azimuths, enter the minimum azimuth, increment, and maximum azimuth (e.g., 0:1:359 will
shoot rays from 1° to 359° in increments of 1°).
• Launch angle: The launch angle in degrees from the horizontal. For shooting rays at multiple
launch angles, enter the minimum launch angle, increment, and maximum launch angle (e.g.,
1:1:89 will shoot rays from 1° to 89° in increments of 1°).
• Range: The maximum range in kilometers for which to compute the ray path.
Shooting Direction:
• Choose ‘Forward from source’ or ‘Backward from receiver’. For the latter option, the
meteorological profiles will be reversed such that shooting backward from the receiver at a
particular azimuth will simulate a ray arriving at the receiver from that direction.
Methodology:
• The final solution should always be run using the ‘Formal integration method’ which uses the
approach outlined by Drob et al. (2010). The ‘Quick/Dirty method’ is not recommended for
anything other than a quick survey of a large simultation before running the more
computationally expensive ‘Formal integration method’.
Perspective:
• Set the perspective to x-z or x-y before clicking on the Compute button to define what
perspective the ray solutions are plotted from. Note that the perspective can be modified by
clicking on the Rotate button after the full solution has been obtained. Clicking Rotate will
bring up the Matlab rotate tool, enabling the user to rotate the axes by clicking and dragging on
the ray figure. Clicking Rotate a second time will deactivate the rotate tool.
Turning Rays Pop-up Menu:
• Select ‘Yes’ to compute and plot only rays that turn and are refracted back to the surface. Select
‘No’ to compute and plot all rays, including those that do not refract back to the surface, noting
that this option is slower and may overcomplicate the plot in some cases.
Geographic Coordinates Pop-up Menu:

Page 21
InfraMonitor 3.1 Users Guide

• The default coordinate system is (range, transverse offset). Selecting ‘Yes’ in the Geographic
Coordinates pop-up menu will instruct InfraTaup to convert all coordinates to a geographic
coordinate system (longitude, latitude). InfraTaup will then use the parameters specified in the
‘Geographic Parameters’ box (Lat and Lon correspond to an event location – or receiver
location for shooting rays backwards) This option is recommended when computing multiple
azimuths because otherwise all solutions will be projected onto the same azimuth. Selecting this
option also instructs InfraTaup to generate a map of predicted bounce points, color coded by
group velocity, for the region specified by Lat limits and Lon limits.
Output Data Structure:
• InfraTaup creates a global Matlab structure array called ray. Each element in the structure array
corresponds to a different ray. To view this structure array (only after performing a
computation), type:
global ray
in the Matlab command window. This will make available to the user all the ray parameters
calculated by InfraTaup. These include:
◦ ray.phi: The launch azimuth in degrees for the ray.
◦ ray.theta: The launch angle in degrees for the ray.
◦ ray.x: A vector containing either (i) the range in km at discrete points along the ray, or (ii)
the longitude at discrete points along the ray (depending on whether Geographic
Coordinates is selected as No (i) or Yes (ii)).
◦ ray.y: A vector containing either (i) the transverse offset in km at discrete points along the
ray, or (ii) the latitude at discrete points along the ray (depending on whether Geographic
Coordinates is selected as No (i) or Yes (ii)).
◦ ray.z: A vector containing the elevation in km at discrete points along the ray.
◦ ray.v_g: The group velocity of the ray.
◦ ray.bp: Bounce points for the ray (only gets computed for a geographic coordinate system).

Page 22
InfraMonitor 3.1 Users Guide

Figure 11. InfraTaup GUI interface with ray tracing simulations in (range, transverse offset)
coordinates.
Equivalent Command Line Functions:
• Ray computations can also be run from the Matlab command line. This option enables the user
to call InfraTaup functions from their own Matlab scripts.
To shoot a ray, use the taup function:
taup(metfile,dz,h,phi,theta,max_range,plotflag,I,reverse,fullray,turningRays)
Where the input parameters are:
◦ metfile: An ASCII file containing a vertical profile of winds and temperatures [format:
elevation (km), temperature (K), zonal wind (m/s), meridional wind (m/s)]%
◦ dz: Vertical sampling distance for meteorological profiles (m)
◦ h: Ground elevation (km)
◦ phi: Launch azimuth (degrees)
◦ theta: Launch angle from horizontal (degrees)
◦ max_range: Maximum straight line distance from source (km)

Page 23
InfraMonitor 3.1 Users Guide

◦ plotflag: 1 for plotting meteorological data, 0 otherwise

◦ I: Ray number (set I=1 if processing a single ray)
◦ reverse: Forward propagation from source (=0) or reverse propagation from receiver (=1)
◦ fullray: Flag to indicate whether or not to compute the full ray coordinates (0) or just the
bounce-points (1)
◦ turningRays: Flag to indicate whether to compute only turning rays (=1) or all rays,
including those that do not turn (=2)
To convert a ray to geographic coordinates:
BouncePoints(ray(i))
Where i is the index of the ray that you wish to transform. To transform all rays, simply call
BouncePoints in a for-loop.

Input Parameters
All the free parameters used by InfraMonitor 3.1 can be viewed and edited by selecting:
Tools>Preferences in the InfraMonitor GUI. Note that when you make a change, clicking ʻSaveʼ will
retain the parameters for future runs (either GUI or command-line runs).

Table 1 provides a summary of the free parameters.

File Input/Output
InfraMonitor 3.1 can read waveform data in SAC (typically .sac or .s) or CSS3.0/NNSA format (.w
files). In either case, the waveform data must be linked to the following CSS3.0/NNSA table files:
wfdisc & site. If wfdisc and site files are not available, they can easily be generated for .sac waveform
data using the MakeWfdisc.py and MakeSite.py utilities provided in the PYTHON subdirectory.

The results generated by InfraMonitor 3.1 can be output as a series of CSS3.0 tables (in addition to
specific InfraMonitor tables). A summary of the output tables is given in Table 2.

Detector parameters Time window The processing time window should be set to the
approximate signal durations of interest.
Overlap The overlap between processing time windows. An
overlap of 50% is recommended as an appropriate
choice for most applications.
Adaptive window The adaptive window length must be set long enough
to obtain a sufficient sample distribution, but short
enough to account for temporal variations in ambient
noise. We have found that an adaptive window length
of 1 hour is appropriate for typical infrasound data.

Page 24
InfraMonitor 3.1 Users Guide

p-value The p-value affects the numbers and signal-to-noise

ratios of detections obtained.
Associator parameters Azimuthal deviation The allowed backazimuth deviation should be set
sufficiently large to account for the effects of wind on
measured backazimuths.
Min. group velocity The slowest possible infrasonic group velocity or
celerity to use in the Polygon locator
Max. group velocity The fastest possible infrasonic group velocity or
celerity to use in the Polygon locator
Picking Error An estimate of the picking error (s)
BISL Locator Azimuth std. dev. An estimate of the standard deviation in azimuth
parameters Arrival time std. dev. An estimate of the standard deviation in arrival time
Vector length The number of origin-times and group velocities to
search over
Flat Earth 1=flat earth, 2=spherical earth (much slower)
Table 1. Summary of InfraMonitor 3.1 Input Parameters.

Under the Hood

InfraMonitor stores all pertinent information during a session under a global Matlab variable called
InfraConfig. InfraConfig is a Matlab structure array that contains all the data and derived parameters
from each run. To access InfraConfig, type:
global InfraConfig
in the Matlab command window while you have an InfraMonitor session open. To test this, try loading
the configuration you made – Config.mat – then type global InfraConfig in the Matlab command
window. You should see the following:

Page 25
InfraMonitor 3.1 Users Guide

Table Extension Description How to generate

arrival Standard CSS3.0 table containing information After running detector on
on each arrival all arrays, select Detect>Save and
the arrival file will be created
origin Standard CSS3.0 table containing information Automatically generated
on each event by location routine
IMassoc Associates location polygons (IMPoly table) Automatically generated
with events (origin table) by location routine
Format:
%17.5f (Arrival time) %8d (EventID/PolyID)
%8d (ArrayID)
IMpoly Contains information on location polygons Automatically generated
Format: by location routine
%9d (PolyID) %9d (PointID)
%11.6f (lat) %11.6f (lon)
Table 2. Output files generated by InfraMonitor 3.1

>> global InfraConfig

>> InfraConfig
InfraConfig =
db: [1x1 struct]
trace: [1x1 struct]
array: [1x1 struct]
detect: [1x1 struct]
grid: [1x1 struct]
BISL: [1x1 struct]
gt: [1x1 struct]
This section describes the InfraConfig structure array to help to access data or solutions for your own
codes.
InfraConfig.db: Contains database information acquired from wfdisc and site files.
• dir: Full path to the data directory.
• name: Name of CSS3.0 table.
• wfdisc: Contains wfdisc table data.
• site: Contains site table data.
• select: Contains a list of data read into memory.
• stime: The start-time of the data read into memory (in Matlab datenum format).

Page 26
InfraMonitor 3.1 Users Guide

• etime: The end-time of data read into memory (in Matlab datenum format).
• dir_data: Similar to dir with ./ appended.
• f_band: The frequency band used in the bandpass filter.
InfraConfig.trace: Contains the waveform data.
• name: Sensor names corresponding to each waveform.
• s_f: Sampling frequency associated with each waveform.
• fname: File names corresponding to each waveform.
• data: Waveform data associated with each waveform.
• time: Time vectors associated with waveform data.
• select: The current selected waveforms (waveforms displayed in red in the main window).
• y_ax: The locations on the y-axis at which each waveform is plotted.
• gt1: Earliest infrasound arrival times based on the ground-truth information.
• gt2: Latest infrasound arrival times based on the ground-truth information.
• gt_azi: Predicted backazimuths at each sensor based on the ground-truth information.
• GTTime_Pn: Predicted Pn arrival times based on the ground-truth information.
• GTTime_Pg: Predicted Pg arrival times based on the ground-truth information.
• GTTime_Lg: Predicted Lg arrival times based on the ground-truth information.
InfraConfig.array: Contains array and arrival information.
• loc: Central locations for each array.
• arr: Detections obtained at each array.
• trc: A listing of trace numbers (to link with InfraConfig.trace information) associated with each
array.
InfraConfig.detect: Contains FK processing results:
• time: Times associated with FK results.
• slofk: Trace velocities.
• az: Azimuths.
• fstat: Adaptive F-statistics.
• corr: Correlation coefficients.
• trc: A listing of trace numbers (to link with InfraConfig.trace information) associated with each
set of FK results.
InfraConfig.grid: Contains the grid used for association and location.

Page 27
InfraMonitor 3.1 Users Guide

• glat: Grid latitude boundaries.

• glon: Grid longitude boundaries.
• dmin: Minimum grid spacing.
• dmax: Maximum grid spacing (the default is for dmin = dmax).
• D_STN: Distances from each array to each grid point.
• BA_STN: Backazimuths from each array to each grid point.
• LAT: Latitudes at each grid point.
• LON: Longitudes at each grid point.
InfraConfig.BISL: Contains BISL processing results:
• lat: Locations of latitudes used in the BISL grid.
• lon: Locations of longitudes used in the BISL grid.
• len: Dimensions of BISL search [max(lat)-min(lat), max(lon)-min(lon), max(time)-min(time)
(days), max(v)-min(v) (km/s)].
• PDF4: Combined likelihood value over 4 dimensions (lat, lon, velocity, origin time)
• PDF3: Combined likelihood value after integrating out velocity.
• PDF2: Combined likelihood value after integrating out velocity and origin time.
• Credibilities: Empty array used for storing temporary credibilities.
• CREDIBILITIES: Stores location credibilities for each event found.
• Baz: Stores backazimuths at each array that were used in the location.
• Meanloc: Stores the mean locations for each event found.
InfraConfig.gt: Contains ground-truth information
• loc0: Origin location.
• t0: Origin time.
• v_g: Group velocities used for plotting ground-truth predicted arrival times.

Page 28
InfraMonitor 3.1 Users Guide

References
Arrowsmith, S.J., Whitaker, R.W., Katz, C. & Hayward, C., 2009. The F-detector Revisited: An
Improved Strategy for Signal Detection at Seismic and Infrasound Arrays, Bull. Seism. Soc. Am., 99,
449-453.
Arrowsmith, S.J., Whitaker, R.W., Taylor, S.R., Burlacu, R., Stump, B., Hedlin, M.A.H., Randall, G.,
Hayward, C. & ReVelle, D., 2008. Regional monitoring of infrasound events using multiple arrays:
application to Utah and Washington State, Geophys. J. Int., 175, 291-300.
Drob, D.P., Garces, M., Hedlin, M.A.H. & Brachet, N., 2010. The Temporal Morphology of Infrasound
Propagation, Pure appl. geophys.
Garces, M., Hansen, R.A. & Lindquist, K.G., 1998. Traveltimes for infrasonic waves propagating in a
stratified atmosphere, Geophys. J. Int., 135, 255-263.
Modrak, R.T., Arrowsmith, S.J. & Anderson, D.N., 2010. A Bayesian framework for infrasound
location, Geophys. J. Int., 181, 399-405.

Page 29
InfraMonitor 3.1 Users Guide

APPENDIX 1: Change Log

March 26, 2012 – Version 3.1
• New Pipeline processing code, written in Python, that simplifies pipeline processing by
providing a single script that takes input as a series of arguments. The script can be pointed to a
single wfdisc file that may contain multiple data segments.
• Support for reading e1 (compressed) data provided.
• Updated reading of flat files. Support for switching between css3.0 and nnsa format provided.
• Enables use of refsta information in site file.

Page 30