Tide Tool Manual

V10.1
March 2014

Tide Tool Version 10.11, Client Version 2.6

Abstract
Tide Tool is an operational tool developed and supported by the Pacific
Tsunami Warning Center for the decoding, display, and manipulation of
sea level data transmitted over the WMO Global Telecommunications
System (GTS). Tide Tool continuously decodes GTS sea level messages
in real-time and displays the time series on a computer screen; users are
able to interactively manipulate the time series to measure wave
amplitude and period, and to remove the tidal signal. Non-operational
versions also exist for data transmissions through the internet (ftp) and
for archived data.
The ITIC distribution of the GEOWARE TTT program has now been
integrated with Tide Tool. This integration enhances the use of TTT and
Tide Tool for the purposes of tsunami monitoring

1. Introduction
Tide Tool is created by the Tide.tcl script, which is based on the Tcl/Tk extension called
BLT. Tcl/Tk is an open source, platform-independent software package. Tcl/Tk at its
heart is a robust shell programming language (tclsh) with a powerful, graphical toolkit
(wish) (Appendix I). Tide Tool v9.x and later also requires the Tk extension BWidget.
Tide.tcl provides end users with the ability to decode, display, and manipulate sea level
data broadcast over the Global Telecommunications System of the World Meteorological
Organization (WMO). Data are continuously downloaded directly from the GTS to a
data file, which is then decoded and displayed. The displays can be converted into
postscript and printed. A version is also available using a non-GTS ftp data
transmission protocol in Linux and Windows XP/Vista for tsunami centers without GTS
links (Appendices II and III). The ftp method should be considered non-operational if
commercial internet service is utilized since connection reliability and timeliness,
especially during tsunami emergencies, cannot be guaranteed.
The software application was originally developed by the US NOAA NWS Pacific
Tsunami Warning Center to provide an operational tool for the real-time continuous
tsunami monitoring in the Indian Ocean. Its primary users would be National
1

Meteorological and Hydrological Service (NMHS), or other agencies with a down link
from the GTS or to a data file containing those data formatted in a similar manner. It has
been tested under UNIX, LINUX, Windows 2000, XP, and Vista environments in
Indonesia, Malaysia, the IODE (Belgium), ITIC, PMEL and at the PTWC. It is currently
being ported to MAC OS.
The software is easily modified for changes in sea-level message formatting and the
addition of sea-level stations. The primary use of this software is as an operational
program run by tsunami warning centers, or other agencies, that need to continuously
monitor sea levels. Mouse-clickable functions include the expansion of the time series,
and measurement of the arrival time, wave height and wave period from the incoming
signal.
Tide Tool version 7.x and later, use dynamic map clients which allow users to select
individual stations for display. The clients have a zoom feature that allows users to
select an area on the map client and enlarge it. Socket handlers were added to the Tide.tcl
script, such that Tide.tcl becomes a server listening for requests on a designated socket.
The clients connect to Tide.tcl over this socket.
Tide Tool version 8.x and later also has the ability to decode old logfiles of GTS sealevel messages. Of course, decoding old logfiles necessitates having the correct metadata
as of the date of the logfile.
Two fundamental changes have been made in Tide Tool V9.x, the introduction of a
stripchart widget and changes in coding that require the addition of another column to the
COMP_META database file. This column contains the name of the Tcl function that will
decode messages from that station. Previously, the Tide.tcl script needed to have all of
the WMO headers used in GTS sea-level messages hardwired into the Tide.tcl script in
an effort to assign a message to a decoder. Hence, products with new headers required a
new release of Tide.tcl which then required all users who wanted to display the new sealevel message to install. I hope from now on, new products wont require a new release of
Tide.tcl unless they require a new decoding function.
This latest release of Tide Tool utilizes the maps and estimated time of arrival (ETA) file
generated by the ITIC distribution of GEOWARE TTT. Tsunami travel times can now
be overlain on the maps used by the Tide Tool clients and displayed on Tide Tools sealevel widgets. The use of TTT with Tide Tool is examined in Section 4.

Requirements:

In order to decode GTS sea-level data, run the dynamic map clients and display the data,
and utilize tsunami travel time information the following are required:
2



Computer running Tcl/Tk software with BLT extension

Sea level data that are continuously archived into a data file
Tide.tcl software
ITIC distribution of GEOWARE TTT

Tide Tool can be downloaded from http://ilikai.soest.hawaii.edu/ptwc/. Look for the

latest version (versions numbered sequentially) and download the accompanying
TideTool_MetaData file and the Tide_Coeffs directory. A complete metadata inventory
of PTWCs sea-level station list can also be found at this link in the COMP_META file.
This will allow you to customize the stations your Tide Tool will decode.
The GEOWARE TTT software and Tide Tool are available on the ITIC Decision
Support DVD. This manual will not discuss TTT in great detail and users are urged to
examine the TTT documentation on the Decision Support DVD.
The IOC maintains a website :
http://www.ioc-sealevelmonitoring.org/
where you can also monitor sea-level and obtain information about sea-level stations.
Computer and Software Requirements:
The software requires the installation of the Tcl/Tk software package and the BLT (or
Wize), BWidget extensions, all of which are freely available for download and easy to
install. The software has been installed and tested under Unix, Linux, and Windows
2000, XP, Vista and Win7 operating systems (OS). Use on other platforms is possible as
it only depends on the Tcl/Tk and BLT (or Wize) software being available. Wize is
better optimized for Win 7. The software does not require substantial computing power,
and can thus run easily on a Pentium III or higher PC system.
Some users download the sea-level messages on a PC running a flavor of Windows and
then transfer the data to a UNIX/LINUX system or vice versa. You must make sure that
the scripts, metadata and sea-level message files are consistent with the platform Tide
Tool is running on. Tide Tool checks what OS it is running under and expects the files to
have the end-of-line and end-of-file terminators appropriate for that OS. Use the
dos2unix or unix2dos utilities to ensure consistency with the OS.
It is not our intention to give a Tcl/Tk tutorial in this manual, as there are several good
books and websites on the subject. Please refer to a short list of them in Appendix I.
There is a large Tcl/Tk user community, and there is a Google group (comp.lang.tcl)
devoted to Tcl/Tk issues.

2. Configuration
For Decoder

To execute Tide.tcl, you must run the bltwish.exe program or wize.exe program
depending on your installation. ITIC now distributes Tide Tool with Wize, a Tk package
that combines Tcl/Tk 8.5 and BLT. Wize is an unofficial BLT 2.5. Note that
bltwish24.exe is an alias to bltwish.exe, so either can be used. UNIX users can turn the
Tide.tcl script into an executable by making the first line in the script:
#!/full_path/bltwish (or
(chmod u+x Tide.tcl).

#!/full_path/wize ) and then making the script executable

When the bltwish or wize console appears, just type source Tide.tcl and the program
should start. Alternatively, you may type bltwish Tide.tcl or wize Tide.tcl on a command
line to start the program.
For Windows PC users, another easy way to start Tide.tcl and the Map clients without
having to open a console to type the command is to simply drop the script onto a
bltwish24.exe or wize.exe icon. To do this, create a short cut/alias of bltwish24.exe (or
wize.exe) and place it as an icon on the desktop. To run Tide.tcl and the map clients,
highlight the tcl script from the directory where they reside and move/drop it onto the
bltwish24.exe (or wize.exe) icon to start the script.
There are 20 customizable parameters. You can assign their values at the top of the
Tide.tcl script. The assignment statements are already placed near the top of the script.
They have the form:
set parameter
If nothing is placed between the , default values will be used or some features will be
disabled. You must however specify the filename parameter.
The 20 parameters are:
1. filename: Tide.tcl assumes the log files which contain the GTS (WMO) data are in the
form rootJJJYY.log where JJJ is a three digit Julian day and YY is the two digit year. The
filename includes the path and filename suffix or root, and should be specified as
path/root. Note that the \ is a special character in the Tcl world and must be escaped.
Therefore if you use the backslash in the path specification, it must be written as \\
otherwise the path will be garbled. The filename parameter must be set otherwise Tide
Tool will not work.
4

An example from PTWCs implementation of Tide Tool is:

set filename /data/logs/nmc/sr
where /data/logs/nmc/ is the path and sr is the root. Hence the filename of a log file for
day 116, 2006 is /data/logs/nmc/sr11606.log.
2. dirname: Tide.tcl also has the ability to write the decoded tide data to files. The
decoded data files have the syntax: station.gauge.JJJ where JJJ is Julian day. The data is
arranged in a two-column ASCII format: time in fractional Julian day and sea-level
height in meters. To enable this feature you must define the dirname parameter in the
script. This parameter specifies where the decoded tide data is written. If this parameter is
not specified decoded data will not be written to your system.
An example from Stus implementation is:
set dirname /home/stuart/Tide/2009/
Note: This directory must be created before running Tide.tcl; it will not be created for
you.
3. hardirname: Tide.tcl will now de-tide the decoded sea-level messages. The files that
contain the harmonic coefficients have the naming convention station_code.tide.coeffs.
They are to be placed in the directory specified by the hardirname variable.
An example from Stus implementation is:
set hardirname /home/stuart/Tide/Tide_Coeffs/
4. metadata_filename: The full pathname for the metadata file which determines which
stations to decode and display. If not specified, ./TideTool_MetaData is assumed.
An example from Stus implementation is:
set metadata_filename ./COMP_META
5. mapdirname: Tide Tool can now display maps showing the geographical context in
which the sea-level station lies. These maps contain a lot of information and they will be
described later in this document.
An example from Stus implementation:
set mapdirname "/data/ttt_dbase/geoware/maps/"
6. tilelistfile: This specifies the name of a file which contains station-sensor code pairs.
After Tide Tool is finished with its initial read of the logfile after startup it
5

will automatically tile the screen with displays corresponding to those stations.
An example from Stus implementation is:
set tilelistfile /home/stuart/Tide/Tile_List
7. record_filename: You can now record measurements of sea-level heights and those
measurements will be written to the file specified by record_filename.
An example from Stus implementation is:
set record_filename ../observations
8. error_log: Messages with bad time stamps, improper formats and unrecognized
headers will be written to the file specified by error_log.
An example from Stus implementation is:
set error_log Error_log
The actual logfiles will be error_log.jd where jd is Julian Day.
9. printer_command: Tide Tool stores the latest 100 messages from each tide station.
These messages are viewable and can be made printable if the variable
printer_command is set to the local systems printer command.
An example from Stus implementation is:
set printer_command enscript fCourier-Bold8 2> /dev/null
10. TWC_DART_FORMAT: If this is set to 1, Tide Tool will write the decoded
DART data files using the same format as does PTWC. An example of that format is
given below:
File: 55015_nha_2008.235
The filename is composed of the DART buoy number, a three-letter code and the year
and Julian day. The three letter code can be one of nha, tra, tha, nhb, trb, thb, which
indicate normal transmission mode (nha, nhb), trigger transmission mode (tra, trb) and
triggered hourly messages (tha, thb). The contents of file 55015_nha_2008.235 look like:
# DART_Tasman_Sea_55015 dtas nha 55015 NDBC NORMAL_MODE -46.9222 160.5622 2008234
# DART m UTC 15 min Water_Depth PTWC Unfiltered
# DCP_PLATFORM_ID 55015P
#
Data
Format:
SampleTIme(seconds
from
midnight
1/1/1970)
WaterLevel
SampleTime(yyyymmddhhmmss)
1219276800 04944.306 20080821000000
1219277700 04944.378 20080821001500
1219278600 04944.442 20080821003000

1219279500
1219280400
1219281300
1219282200
1219283100
1219284000
1219284900
1219285800
1219286700
1219287600
1219288500
1219289400

04944.504
04944.556
04944.602
04944.643
04944.670
04944.694
04944.709
04944.708
04944.700
04944.684
04944.659
04944.625

20080821004500
20080821010000
20080821011500
20080821013000
20080821014500
20080821020000
20080821021500
20080821023000
20080821024500
20080821030000
20080821031500
20080821033000

The first column is the epochal time (seconds since midnight 1/1/1970), the second is the
water column thickness in meters and the right column is the human readable time at the
time of measurement.
11. num_buttons_row: This parameter sets the number of buttons per row on the main
decoder GUI.
The next several parameters control the appearance of the sea-level graphs.
12. plot_width and 13. plot_height: These parameters set the dimensions of the basic
Tide Tool sea level displays like that shown below for Benoa. An example value would
be 3.0i where i stands for inches. Likewise, the dimensions can be specified as 3.0c
where c is for cm. The defaults are 7.0i and 3.0i respectively.
14. plot_bg, 15. plot_obs, 16. plot_cor, 17. plot_pre: These parameters set the
background color for the displays and the colors for the observed, corrected, and
predicted sea-level curves.
18. plot_dir, 19. plot_pri_com specify the directory where postscript files of the
displays are placed and the system printer command used to print them. The printer must
understand postscript.
19. strip_plot_width, 20. strip_plot_height specify the width and height of the plots in
the stripchart widget. The defaults are 8.0i and 1.5i respectively.
Starting with Tide Tool V8.0, these parameters can also be configured using environment
variables. This way you will not have to edit the Tide.tcl script to customize it for your
implementation. For example in C-shell you can have:
setenv FILENAME /data/logs/nmc/sr
setenv DIRNAME /data/logs/nmc/wl
setenv METADATA_FILENAME /dbase/water/COMP_META
7

setenv
setenv
setenv
setenv
setenv
setenv

HARDIRNAME /dbase/tide/
MAPDIRNAME /data/ttt_dbase/geoware/maps/
TILELISTFILE ""
PRINTER_COMMAND "enscript -f Courier-Bold8 >& /dev/null/"
RECORD_FILENAME /home/ptwc/io/params/water/obs/tsu_measures
ERROR_LOG

setenv
setenv
setenv
setenv
setenv
setenv
setenv
setenv
setenv
setenv
setenv
setenv

NUM_BUTTONS_ROW 12
TWC_DART_FORMAT 1
PLOT_WIDTH 6.0i
PLOT_HEIGHT 3.0i
STRIP_PLOT_WIDTH 8.0i
STRIP_PLOT_HEIGHT 1.5i
PLOT_BG white
PLOT_OBS red
PLOT_COR black
PLOT_PRE blue
PLOT_DIR /home/ptwc/io/params/water/station_plot/
PLOT_PRI_COM "lp -d hp3700"

If the environment variable is set for a parameter, its value will take precedence over the
setting in the Tide.tcl script.

For Clients
The clients have 15 required parameters however, you should be able to use the clients
with little modification. The parameter descriptions are as follows:
1. left_lon and 2. rht_lon : These parameters represent the longitude limits of the map.
3. top_lat and 4. bot_lat : These parameters represent the latitude limits of the map.
5. Image_Width_Pixels and 6. Image_Height_Pixels: The width and height of the
map in pixels.
7. Map_File_Name : The name of the map file. As an example, for the Pacific client we
specify the map file using:
set Map_File_Name "./pacific_ocean_map.gif"
8. metadata_filename : This is the name of the metadata file (full pathname).
8

9. client_title : The name of the client that appears at the top of the client.
10. scale : The amount by which the selected area will expand when creating a client
zoom widget. A scale value of two seems to give good results. A scale value greater than
three produces somewhat fuzzy maps. You can use non-integers, but then the generation
of the client zoom widget may take several tens of seconds depending on your computer.
The integration of the Tide Tool Clients with TTT requires 5 additional variables:
11. Image_Width_Pixels_TTT and 12. Image_Height_Pixels_TTT: The width and
height of the TTT generated high resolution maps in pixels.
13. map_filename_ttt and 14. map_filename_ttt_hr: The names of the TTT maps fro
regular and high resolution
15. ttt_grd_filename: The name of the travel time grid file created by TTT.
For the clients there are four total environment variables:
setenv METADATA_FILENAME COMP_META
setenv PAC_MAP_FILENAME ./pacific_ocean_map.gif

IND_MAP_FILENAME ./indian_ocean_map.gif

CAR_MAP_FILENAME ./carribbean_map.gif
The other parameters are not installation dependent.
You can certainly generate your own map files (in the GIF format) and if you change the
above parameters in the client Tcl/Tk scripts you can create your own clients. For now,
the clients can only do the linear/cylindrical projection. However, I expect to add the
math needed to do some other common projections like the Mercator in the future.

The COMP_META File

The COMP_META file is a flat file database containing a list of all stations decoded by
Tide Tool. The COMP_META file has 18 columns and the metadata it contains includes
the geographic coordinates for each station, and a number of parameters that facilitates
the decoding of the sea-level messages. From left to right, the 18 columns are: station
code, geographic name, platform ID number, WMO header, sensor type, sample rate for
that sensor, number of samples for that sensor per message, observation units (F for .01ft
and M for mm), a number that indicates the number of non-data fields in the data
payload, time ordering (1 for chronological and -1 for reverse chronological order), a
9

scale factor, transmission time interval, hour:minute of first transmission of the day,
latitude, longitude, availability of tide coefficients for tide prediction (1 for available, 0
for unavailable), owner of the sea-level station and GLOSS ID (added at the request of
GLOSS). An extra column was added with Tide Tool v9.0 that is used by the Tide.tcl
script to determine which decoding function to use. An example is given below for the
Benoa sea-level station pressure sensor.
beno Benoa_ID 22856700
049 PARSE_GLOSS

SXXX32 prs

30

-1

1.00

015

0007

-8.7500

115.2166

UHSLC

In this case the entry in the last column is PARSE_GLOSS which is the Tcl function that
will be called to decode messages from Benoa.
Any line in the metadata file that begins with # will be ignored. When Tide Tool starts
it performs a quality check on the metadata file and will stop if the metadata file is not
configured properly.
One can construct a customized metadata file using just the stations of interest. An
example of this is the TideTool_Metadata file that is also available at the ftp site. This
metadata file is customized for the Indian Ocean. The metadata can be organized into
regions of your choice by placing a region name like SUMATRA_JAVA on a line by
itself. The region name is your choice, however, white space is not allowed, hence
underscores _ must be used instead of blank space. The presence of region names will
cause the main GUI to be segmented into different regions making it easier to locate
stations on the GUI. An example is given in section 3.
To be consistent, the same Metadata file (parameter metadata_filename) should be used
in both Tide.tcl and the map client. This will ensure that the Tide.tcl station GUI
(showing decoded stations) and the map client (showing dots for each station with colorcoded station health) are correctly shown and updated.

3. Operational Details
Please note that Tide Tool enforces GMT (Greenwich Mean Time) when decoding.
Therefore, you must make sure that your system gives the correct GMT time. This
requires that you set the time zone on your computer to 0 offset from GMT. Do not
simply change the clock on your system; change the time zone. Changing the clock will
cause Tide Tool to assume that the change of day occurs either too late or too early. The
use of GMT is necessary to remove any ambiguities with respect to time, as GTS data
also uses GMT. If your system does not have the correct GMT time, Tide Tools
behavior will be erratic.
1

Tide Tool version 10.0 consists of two basic parts, the decoder (created by the Tide.tcl
script) with the familiar main GUI containing buttons for each sea-level station. The other
part is the dynamic map client. There is one basic map client which takes one argument
that specifies which ocean to support. Tide.tcl listens for requests on port 15500.
Communication between the decoder and clients is two-way. The clients and decoder can
send instructions back and forth. The decoder can support any number of clients, so if
you choose, you can run multiple instances of the same client, or all four clients.
When Tide.tcl is started it will read data from the current day data log. Tide.tcl will keep
up to ~48 hours of data. As more data arrives beyond what Tide.tcl can hold, it will
discard the older data to make room for the new. We have left some active debug code in
Tide.tcl, so if you see a lot of output, do not worry. If Tide.tcl fails, please send PTWCs
Stuart Weinstein ([email protected]) as much of the debug output as you can, up to
the last statement before Tide.tcl failed. If possible, send the log file Tide Tool was
reading at the time. Also send the error_log if you have one.

The Decoder

Tide.tcl starts by reading the previous days log file (if it exists) and continues reading
through the current days log file. A simple GUI will be displayed that has buttons for
each tide station in the metadata file. After the initial read through the current day log file
Tide.tcl goes into the background. Tide Tool sleeps and wakes up every 20 seconds to see
if any new data has arrived. If so, it is decoded. Once started, Tide Tool runs
continuously.
The station name is listed as a four-letter code. The stations codes are obtained from the
metadata file. The gauge code is only three letters; prs stands for pressure sensor, bub
indicates bubbler, enc indicates encoder, rad indicated radar, etc. More information on the
gauge types can be found in Appendix IV, which is extracted from the PTWC
Operations, Systems, and Procedures Manual (Version 1.0, Dec 2008).
When Tide Tool is started it will read the current and previous day's logfiles (if
available). During this time the GUI is deactivated and the lettering on the buttons is
grey. After the initial read is done, the GUI is activated and the lettering turns dark violet.
On the bottom of the GUI is a message widget showing how many bytes have been read
from the current log file.

When Tide.tcl is first started,

the lettering on all buttons is
grey (inactive) except for the
EXIT button. The station
button section is placed on a
scrollable, expandable canvas.
Depending on the size and
resolution of your screen, you
may or may not need to scroll.
Region
names
like
SUMATRA_JAVA are yours
to create and are to be placed in
the metadata file as described
previously.
Clicking on a button in the GUI will display a menu showing the gauges available:
In this snap shot, the Benoa,
Indonesia station was selected
and the menu showing three
gauges (a pressure, radar and
encoder gauge ) plus two
other items Info and
Mess are displayed. Those
will be discussed later. Four
buttons, EXIT, RETILE,
REPORT
and
UPDATE
STATUS appear on the
bottom. If you make changes
to the file referred to by the
parameter tilelistfile, clicking
RETILE will retile the screen.
Clicking
on
UPDATE
STATUS will send station
status information to the clients. Clicking REPORT will generate a table showing the
number of messages received for each station. The last two features will be described in
greater detail later.

With V10.0, additional buttons have been added to the bottom:

The new buttons are the GET ETAs

and PLOT EPI button. Clicking the
GET ETAs ingests the Tsunami ETAs
calculated by TTT. Clicking the PLOT
EPI will plot the epicenter used to
compute the ETAs on the Clients.

The Monitor and Zoom Widgets

Selecting one of the sensors (for example, prs, rad, enc, bub) will create a monitor
widget. This shows the sea-level curve as a function of time. Clicking on the prs
(pressure sensor) for the Benoa (beno) station produces a monitor widget for the Benoa
station pressure sensor:

Two curves are in the display. The red curve is the sea-level curve resulting from
decoding the sea-level messages. The black curve is the de-tided sea-level curve using a
tide prediction model. The tide coefficients are computed by Dr. Dailin Wang at PTWC.

In some cases the available time series for the newer stations are too short to allow
rigorous de-tiding. However, the bulk of the tide signal is usually eliminated. In the
above figure the TIDE menu button has been clicked and it shows 4 options. The
default is to plot both the observed and corrected tides. Other options are to show the
observed or corrected separately or both with the predicted tide. On the bottom of the
display, the time of the next expected transmission (Next XMT) for this station is given.
Clicking the MENU button produces a
menu showing other sensors for this
station:

This way, you wont have to go to the main GUI to display curves for the other sensors.
Another feature is the DESPIKE button. Clicking the DESPIKE button will run the
sea-level curve through a three point median filter. Basically, at each point, the value is
replaced with the median value amongst that point and its nearest neighbors. This gets rid
of solitary bad points. For spikes/glitches lasting more than one sample interval, it may
not eliminate the spike/glitch. In our experience, however, this simple technique has
proven to be effective in the majority of cases. Consider the mareogram obtained at Lime
Tree, Virgin Islands. Clearly this station is having some sort of software or telemetry
problem:

Clicking the DESPIKE button results in:

As you can see, this is quite an improvement. Also note that the DESPIKE button now
reads UNDESPIKE. Clicking this button now will undo the despiking.
The next to last button on the monitor widget is the PRINT button. Clicking the
PRINT button produces a menu with two options; Plot File Only and Make Plot &
Print. The first option will create a postscript file representation of the graph in the
directory specified by the parameter plot_dir with the form ccccsss.ps where cccc is the
station code and sss is the station sensor. The resulting plots are not exactly publication
quality, but more than good enough for record keeping purposes. I expect to improve
these plots in the future.
The last group of numbered buttons sets the time span in hours. Clicking the 12 button
will change the maximum time span to 12 hours.

As new data is decoded by Tide.tcl, the monitor widget will be automatically updated to
reflect the new data. The widget will only display data for no longer a span of time than
what is selected. The maximum is 48 hours and that is the default. If new data is decoded
for a station while a display widget for that station exists, the background of that display
widget will change from black to blue for 0.6 seconds.
What operations can you do? By using the right button or number three button on your
mouse, you can draw an expanding rectangle to zoom in on a section of the sea-level
curve. Hold the right button down and move the mouse diagonally, and you will draw a
thin red rectangle. When the section of curve you wish to examine is in the rectangle,
release the right button, and a zoom widget will pop-up with the selected portion of time
serious expanded. This was done on a mareogram of the 2010 Chile tsunami recorded at
Acapulco (acap). The resulting zoom widget is shown in the figure below:

You can also further zoom on the zoom widget itself. When the right button is released,
the appropriate section of curve will be displayed in the zoom widget. The TIDE button
at the top left has the same function as with the monitor widget.
On the zoom widget you can conduct simple measurements of amplitude and period.
Using the left or number one mouse button, simply click on the curve in the zoom widget,
and a summary inset will appear in the top left corner showing the coordinates of the
point you just clicked on. A cross will mark the spot you picked. After a second point is
clicked on, the coordinates of the second point will be displayed. In addition the
difference in height and time between the two points will be displayed on the summary
inset. Time difference is in minutes:seconds, and height differences are in meters.

A summary inset can be seen on the Acapulco zoom widget shown above. Note the two
crosshairs. If you wish to clear the summary display, click the middle or number two
mouse button any where on the zoom widget canvas. The PRINT button on the zoom
widget has the same functionality as on the monitor widget. The only difference is that
the postscript file will have the form ccccsssZ.ps (note the Z).
The record button REC will allow you
to record sea-level observations in the
file referenced by the record_filename
parameter.
By clicking the point where tsunami waves first arrive and then clicking the First
Arrival menu item, you can record the arrival of the tsunami at a sea-level station. By
using two picks you can record either the Zero to Peak or Peak to Peak variation on
the sea-level curve. Entries in the record_filename file are in this form:
kota prs First Arrival 111/07 08:43 2007111 11:48
kota prs Peak to Peak 111/07 08:43 H 00.208 111/07 09:08 H 00.155 Per 00:24 Amp 00.052 2007111 11:48

The observation records start with the station and gauge followed by the type of
measurement and then the measurement(s). In the case of First Arrival, we have the
day time stamp of the arrival in the format JJJ/YY HH:MM followed by the time this
observation was recorded in format YYYYJJJ HH:MM. In the Peak to Peak or Zero
to Peak cases, both picks are recorded with a time stamp JJJ/YY HH:MM followed by
H (sea-level height flag) followed by the sea-level in meters. After the picks we have
the time difference (after the Per flag) and sea-level difference (after the Amp flag)
between the two picks followed by the time the observation was recorded.
The Info and Mess items allow you
to display basic station data and recent
messages for a given station. This is an
example of an information widget. It
shows basic information about the sealevel station.

Clicking on the Mess item will create a widget that shows the last 100 messages for the
station that were recognizable to Tide Tool. An example of the messages widget is shown
below on the right:

Note that this widget is scrollable and has a print button. However, as different systems
utilize different commands for printing text, the print button requires some configuration
by the user. In particular the PRINTER_COMMAND variable must be set to a string
that is the printer command for your system. If this variable is not set, the PRINT
button will not appear. It may require more work in the future to make this option a bit
more user-friendly. If you cannot get it to work, contact Stuart Weinstein and we will try
to develop a customized solution.
Clicking the REPORT button on the main GUI will give an account of the number of
messages received and the number of messages expected for each station in the metadata
file over a specified time period. You can select 3, 6, 12 or up to 48 hours. You can also
print the report. The report is approximate in terms of the number of messages expected,
however it is accurate enough so that you can keep track of stations that are having
telemetry problems or other issues. Note that receipt of a message as indicated on the
report does not mean the message was successfully decoded; only that it was received.
An example of the report is shown above, to the left.

Another feature included is the MAP

feature. Clicking on the MAP button will
display a map (GIF file imported into a
tk widget) showing the geographic
context of the stations. There are three
views showing the station at Benoa in a
global, regional and local context. The
location of the station is centered on the
map. Contour lines showing tsunami
travel times at one-minute intervals are
drawn using the station as the starting
point. These maps are useful when it
comes to assessing how the station may
be impacted by a tsunami. For example if
the station is relatively sheltered or is
facing away from the likely direction of
tsunami propagation, it may not be ideal
or
useful
for
evaluating
the
destructiveness of the tsunami.

Tide Tool Clients

Tide Tool has one basic client that takes one argument which is one of PACIFIC,
INDIAN, CARIBBEAN or ATLANTIC. The client opens an interactive map interface
for the specified region as shown below:
Indian Ocean Client

Caribbean Client

Pacific Ocean Client

In each interface display window, the small dots show the location of coastal sea-level
stations and the larger dots show the position of the DART buoys. The colors of the dots
indicate status and the meaning of the colors is given by the legend just below the map.
Green indicates that at least one message was decoded for that station within the last 7
hours. Yellow indicates that at least one message was decoded within the last 24 hours,
and Red indicates no messages received in over 24 hours. The reason these criteria were
chosen is that the DART buoys in their normal mode (as opposed to their trigger mode)
only transmit data 4 times a day.
The station status is updated automatically every 30 minutes and can be updated at any
time by clicking the STATUS button on the decoder (Tide.tcl) GUI. These clients do
not require any special configuration. Place the client scripts in the same folder as the
decoder and start them the same way as the decoder. When the clients connect, the
message CONNECTED will appear on the bottom left of the client GUI as shown
above. If the decoder is not running or the clients have not connected yet, the message
will be NOT CONNECTED. If the decoder is shut down after the clients connected the
message will be DISCONNECTED. Note that after the decoder (Tide.tcl) is started, it
will not accept clients until the initial read of the current log file is completed.
With V10.0, the Clients have two additional buttons; PLOT EPI (same function as on
the decoder) and PLOT TTs. Clicking PLOT TTs will overlay the tsunami travel
time contours computed by TTT. More about that later.
A single left button click on a dot will display the station code for that station. A double
left button click
will create a menu button,
just like the one
on the decoder GUI, and
you can choose
to display a sensor.

The Exit selection will delete the menu. Only one menu at a time can appear, so if you
click on another dot, any pre-existing menu will be deleted. You will notice that some
stations have a yellow halo; this is to inform you that the station has been moved,
graphically, from its actual location so that it can be located and clicked on the map.
When you use the zoom feature, most of these halos will disappear as there is more
resolution and less overlap among the station dots.
The DISP/FIND STATIONS menu button on the bottom can be used to find stations on
the map. Clicking this button produces the menu:

The first two items if checked will cause the station codes and/or DART NDBC numbers
to be displayed on the map. The FIND items are cascading menus that allow you to
locate a station by its station code or DART NDBC number:

To zoom in a section of the Client map, use the right mouse button. Click and traverse
from one corner of the region you wish to enlarge, to the opposite corner and release the
button. This will create a zoom client
The zoom client has the same functionality as above. It also has EXIT and SHOW
buttons on the bottom. Clicking EXIT will destroy the zoom widget.

Clicking SHOW will bring up a menu with two options for displaying monitor widgets
for stations within the zoom widget. These are TILE which was available in previous
releases and STRIP which will arrange the monitor widgets in a stripchart. Clicking
TILE will tile the screen with all stations that have data, present in the zoom region
(see figure above). The TILE option will then switch to UNTILE. Clicking
UNTILE will remove the tiled widgets. Client version 2.x supports multiple zoom
widgets. In previous versions of the Client, creating zoom widget would cause any preexisting zoom widget to disappear. The zoom widgets keep track of what stations exist in
the zoom region independently. If you have multiple workspaces, you can tile each
workspace independently using a different zoom widget. Alternatively, one can create a
strip widget in each workspace, or both tiling and stripcharts. This is a fundamental
change in behavior from Client version 1.x and makes version 2.x more versatile.

An example of the stripchart

widget is shown on the left. The
individual widgets in the
stripchart are monitor widgets.
The stripchart is scrollable
(scrollbar on the right) and
currently there are no limitations
as to how many stations can be
placed in a single stripchart
widget and there are no
limitations, aside from system
limitations as to how many
stripcharts can be made.
Individual stations can be placed
in more than one stripchart. The
time span for all plots in the
stripchart can be set by clicking
and of the 4 buttons at the top. A
span of 48 hours is the default.
The time span can also be set for
each station individually. But
any individual settings will be
reset if the strip widget time
span is selected.
Clicking the Station List button produces a menu showing all stations that are present
in the stripchart widget. Clicking any station on the menu results in the stripchart
reorganizing the stations such that the clicked station appears first and the others ordered
by proximity to the clicked station.

The buttons on the monitor widget are

not shown by default but can be invoked
by left clicking anywhere on the monitor
widget graph. In this case the graph
widget for the Prickly Bay station (pric)
was clicked and now the usual monitor
widget buttons have appeared. Clicking
on the graph again will make the buttons
vanish.

The station:sensor indicators are themselves buttons and clicking on them will arrange
the order of the stations in the stripchart widget in the same manner as clicking on the
Station List menu.; the station clicked will go to the top and the other stations rearrange
in terms of proximity to that station

4. Using TTT
The GEOWARE TTT distribution in the ITIC Decision Support DVD comes with the
program ttt_tidetool.bat. This simple script can be executed in the MSDOS prompt. The
scripts ask for basic information, coordinates of the epicenter, origin time of the
earthquake, ocean basin and resolution to use when computing the travel times. The
ttt_tidetool.bat script creates all of the necessary files used by Tide Tool.
In the example below we computed the travel times for an earthquake off of the coast of
Chile (30.3S, 71.9W) using the ttt_tidetool script and then clicked the PLOT_TTs and
PLOT_EPI buttons:

Zooming on a region about Chile results in the following zoom widget:

When the travel times are overlaid the zoom is performed on
a higher resolution map and the resulting zoom widget is
bigger by a factor of 4 (in terms of area). Also the travel time
contour interval is reduced to 15 minutes from the 30
minutes used on the Client map. Now we can click the GET
ETAs button on the main GUI and Tide Tool acquires the
ETA for each station. Clicking the STRIP option under the
SHOW button creates the following stripchart widget:

In this case, the stations are

arranged
in
terms
of
traveltimes. The magenta
lines indicate the tsunami
ETA on the marigram and the
ETA for each station is
written on the bottom of the
plot. In this case, the origin
time was set to about one
hour before the ttt_tidetool
script was run. You can see
that the magenta line moves
further right as one moves
down the strip. These lines
will move leftward as time
continues. An extra hour is
appended to the end, so you will see an indication of the arrival time up to one hour prior
to the ETA for that station. The strip widgets have an internal clock and every 60s they
update the time axis limits. Each marigram in the strip chart has exactly the same time
limits. The maximum time is the current time (to within 6os) + one hour and the
minimum time is the current time the time span selected at the top of the strip chart, 12
hours, in the chart shown above.

5. Future Directions
Among the new features under consideration for future releases of Tide Tool are:
1. Real Time tide modeling. In some cases, not enough data is available to create a good
set of tide harmonic coefficients. In other cases the tide coefficients are no longer
adequate or meteorological effects are skewing the de-tiding. In these cases, it is useful to
compute tide coefficients on the fly using the 48 hours of data held by Tide Tool.
2. Dynamic reconfiguration. Changing station metadata or adding new stations will not
require a restart of Tide Tool.

6. Historical LOGFILES
Tide Tool has the ability to examine old logfiles. To run Tide Tool in historical mode,
add the argument H to the command you use to invoke the Tide.tcl script. For example:
wize Tide.tcl H
There is one important caveat to be aware of in running Tide Tool in historical mode.
You must to be sure that the metadata used by Tide Tool is consistent with the date of the
logfile you wish to examine. Sensors and data units do change through time. Upon
starting Tide Tool in historical mode, a simple entry widget will appear:

This widget requires three entries, Julian day, Year and SPAN which is the number of
days of data you wish to read starting at the Julian day and year entered. After clicking
the DONE button, this widget disappears and the Tide Tool GUI then appears. After
this, everything is basically the same as Tide Tool normally operates. Currently the
maximum span is 2, or two days worth of data decoded.
Author:
Dr. Stuart Weinstein
Asst. Director, Richard H. Hagemeyer Pacific Tsunami Warning Center
Please send all questions, bugs, requests etc. to:
[email protected]

For other related questions or assistance, please contact Dr. Laura Kong, Director, IOC
International Tsunami Information Centre, [email protected]
Acknowledgments
S. Weinstein is indebted to several colleagues for their help and ideas in creating and
improving Tide Tool. The genesis of Tide Tool was a request from Dr. Laura Kong in
late 2005 to develop a quick and dirty tool that can be used to decipher GTS sea-level
data, display the results and allow basic measurements of amplitude and period to be
made from the sea-level record. Tide Tool 1.0 was released in Nov. 2005. The earliest
users (and perhaps victims) of Tide Tool were the Met. Office in Malaysia, and BMG
in Indonesia. Tide Tool has proven so capable that it is now part of the operational
software suite at PTWC.
In Malaysia, Namran Hussein provided valuable assistance in debugging Tide Tool. Dr.
Yap Kok Seng, Low Kong Chiew, and Saw Bun Liong also provided suggestions and
tested Tide Tool on Windows platforms.
Dr. Fauzi (BMG) provided the first detailed set of instructions for obtaining everything
needed to get Tide Tool running on LINUX. Indra Gunawan was also very helpful in
trouble shooting problems with Tide Tool.
Dr. Yohko Igarashi (JMA/ITIC) provided valuable assistance in improving the port of
Tide Tool to both LINUX and Windows and developing the scripts/processes used to
obtain GTS data from the Web.
Dr. Dailin Wang (PTWC) created and maintains the suite of tide coefficients used in detiding and is currently helping me to develop an on the fly de-tiding algorithm in Tcl.
Dr. Nathan Becker (PTWC) designed and created the station maps and the maps for the
Tide Tool clients. Nathans assistance was invaluable in getting Tide Tool integrated with
GEOWARE TTT.
The UHSLC (University of Hawaii Sea-Level Center) has also been a great help to our
efforts, not just for installing sea-level stations world-wide, but for their assistance in
decoding matters and their encouragement.
Peter Pissierssens (IODE) offered several valuable suggestions during my visit to IODE
in 2006. Mark Van Crombrugge (IODE) helped get Tide Tool running on Windows. We
continue to work closely with our colleagues at the IODE (Bart Vanhoorne and Tjess
Hernandez).

Alina Nieves of the US NWS in Puerto Rico helped with testing Tide Tool V9.
Last but not least, I thank Chip McCreery and the PTWC staff for putting up with my
mood swings while working on Tide Tool.

Appendix I.
Reading Materials:
Practical Programming in Tcl and TK by Brent Welch. Prentice Hall.
A super book and introduction to Tcl/Tk. You can find it in many bookstores or obtain it
from an online source.
Effective Tcl/Tk Programming by Harrison and McLennan
This covers fundamentals of event driven programming. For those of you who do mostly
serial programming in Fortran or C, event driven programming is a bit a of a trip.
Tcl and the Tk Toolkit by John Ousterhout
Ousterhout invented Tcl/Tk. So I feel obliged to mention his book.

Websites:
http://www.activestate.com/Products/ActiveTcl/

You can get binaries here. This site has something for everyone.
http://tcl.sourceforge.net/
http://sourceforge.net/projects/blt/

You can get binaries here and keep tabs on the latest FAQs, versions, bugs, and advances
in the Tcl world.
http://phaseit.net/claird/comp.lang.tcl/HowToC.html

Howto, on how to integrate the power of C programming with Tcl/Tk. I use combinations
like this at PTWC. Cameron is one of the great Tcl/Tk experts and frequently answers
Tcl/Tk questions on the comp.lang.tcl Google news group. You may find me there on
occasion as well.
http://wiki.tcl.tk/199

A starting point for BLT

http://pdqi.com/w/pw/pdqi/Wize/Blt
A link to WIZE blt.

Appendix II.
Tide Tool using Internet Data Download
Windows XP and Vista
For more information, contact
Stuart Weinstein ([email protected])
Laura Kong ([email protected])
Version June 2010, July 2012, April 2014
Tide Tool is an operational tool developed and supported by the Pacific Tsunami Warning Center for the
decode, display, and manipulation of sea level data transmitted over the WMO Global Telecommunications
System (GTS). The tool continuously decodes sea level in real-time and displays the time series on a
computer screen; users are able to interactively manipulate the time series to measure wave amplitude and
period, and to remove the tidal signal. A separate process archives GTS-downloaded data continuously
through a dedicated line (recommended), or alternatively by commercial internet ftp (not recommended for
operations requiring 100% reliability). An offline tool (getTide) is available for archived data. Archived
data can also be viewed (see Section 5 of User Manual).
The following are instructions for the installation and running of Tide Tool on a PC under a Windows XP or
Vista operating system and with data collection through internet through wget
Instructions below refer to the generic names of scripts; in general, each script will also include its version
number as part of its file name (e.g., Tide_v10.11.tcl, Client_v2.5.tcl, etc)

Instruction for Tide Tool (Windows version)

1. If you already have an old version of Tide Tool, delete the files. You may use
C:\Tcl\uninst_TideTool.exe to delete all the files
Alternatively, if you already have the wize.exe (TclTk V8.5.9) and BLT (2.5), chikan.exe, etc
loaded, you may instead just replace the old files with the new files, or add the new files to the
correct directory; most often, the updates are to Tide.tcl, map client scripts, COMP_META, or to the
TideTool_MetaData files, or are the addition of new Tide_Coeffs or MAPS files. In this case, you
should stop get_data.tcl and Tide.tcl, and then restart each again (skip to Step 6 in this instruction).
2. Download the latest TideTool_version_install.zip. When unzipped, the directory has the files:
SLdecode_display_summary7.pdf (general summary)
Tide Tool 10.1 Manual Vxx.doc (User manual these install instructions are in Appendix II)
PC_install_dirinfo.txt (contains directory structure specific for this installation)
TideTool_Quickinfo.doc (quick info / cheat sheet for running software)
Setup.exe (installs software)
3. Run Setup.exe to place files in the correct directories and create desktop shortcuts.
Setup.exe will:

copy all files in C folder to C:\

create Shortcut for WIZE (which runs Bltwish) on desktop and names it: WIZE



create Shortcut for map clients on desktop and names them: ATLANTIC, CARIBBEAN,
INDIAN, PACIFIC
create Shortcut on desktop to go to C:Tcl\bin and names it TideTool_bin
create Shortcut on desktop to go to C:Tcl\TideTool and names it TideTool_data
create C:Tcl\uninst_TideTool.exe (Uninstall program) in case there is a need to uninstall this
package. Unist.exe will delete all installed files except the collected data files in the folder
SR_LOG (srJJJYY.log), Decoded_Data (decoded station data by Julian Day), and any other files
you have created since the last installation

Setup.exe in TideTool_v9.20_install.zip places the following files / folders under C:\

FILE/Folder NAME
---------------------------------------------Tcl

<Folder> -->

bin

<Folder> -->

doc
<Folder>
include
<Folder>
lib
<Folder>
licenses
<Folder>
INSTALL.LOG
license-at8.4-thread.terms
MANIFEST_at8.4.txt
README-8.4-thread.txt
uninst.exe
TideTool
<Folder> -->

(station maps)

Tide_vXX.tcl(V10.0 Beta)
get_data.tcl
wget.exe
chikan.exe
Client_vXX.tcl
pacific/indian/caribbean/atlantic_map.gif
Tile_List (stations to display at start up)
wize.exe (TclTK & BLT files)
MetaData
<Folder>--->
COMP_META and
TideTool_MetaData
region files (may be
outdated)

BOX

<Folder>--->

Decoded_Data

<Folder>--->

Error
MAPS
PLOTS
SR_LOG

<Folder>--->
<Folder>--->
<Folder>--->
<Folder>--->

data downloaded from

ftp site
decoded data by
stations
Error_LOG file
maps for stations
time series plots
srJJJYY.log files of
GTS transmissions

Tide_Coeffs
<Folder>--->
(station_name).tide.coeff
CaribbeanAtlantic_SL_Stations_nov2013.pdf
Indian_SL_Stations_2012.pdf
Pacific_SL_Stations_nov2013.pdf
PC_install_dirinfo_mar14.txt
PTWC Operations Manual V1.3.5.doc
README.txt
SLdecode_display_summary10_nov13.pdf
Tide Tool 10.1 Manual V1.0.pdf
TideTool_QuickInfo_mar14.pdf

- WIZE (wise.exe) includes TclTk (V8.5.9) and BLT (v2.5)

- wget.exe is used to download data files (e.g., files from NWS telecommunications gateway)
- chikan.exe is used to convert from Unix (\n) to MS-DOS (\r\n).

4. Set your PC to use UTC (GMT, Greenwich Mean Time).

5. Customize Tide.tcl and map client scripts to specify directories where files to be found or created,
and which files are to be used. For this Windows package, the customization has already been done

using the information in PC_install_dirinfo.txt.

Some variables that you may want to additionally change are:
a. metadata_filename variable, if you want only a subset of the data to be decoded and displayed.
Pre-selected files available are:

COMP_META (default) will decode all stations and list in alphabetical order

COMP_META_region will decode all stations and list by region

TideTool_MetaData_IO is for Indian Ocean

TideTool_MetaData_NEPO is for Northeast Pacific Ocean

TideTool_MetaData_NWPO is for Northwest Pacific Ocean

TideTool_MetaData_SEPO is for Southeast Pacific Ocean

TideTool_MetaData_SWPO is for Southwest Pacific Ocean

TideTool_MetaData_NWAO_CAR is for Northwest Atlantic Ocean, Caribbean, Gulf of
Mexico

TideTool_MetaData_PO_IO_NWAO is for Pacific Ocean, Indian Ocean, and Northwest
Atlantic Ocean, Caribbean, Gulf of Mexico

TideTool_MetaData is to specify user-selected stations to decode/display or getTide.tcl.
b.

tilelistfile, which specifies which sea level station monitors to bring up at Tide.tcl startup. It is
suggested to leave this file blank unless you would like a pre-set set of stations displayed at
startup. An example Tile_List is provided. You can set it not to bring up any monitors at startup
by leaving this file unspecified. For example, if you need to monitor the 3 stations yapi,
mani and davo for prs sensor, then Tile_List would look like
yapi prs
mani prs
davo prs

After Tide.tcl is running, you can edit the file Tile_List to designate other stations, and then hit
the button RETILE to show these station monitors.
c. printer_command or plot_pri_com, which specifies the command for printing. Presently, the
default is set for a linux/unix command or defaulted to no printing.
6. Start get_data.tcl (drop icon onto WIZE icon)
This short cut has already been created by the Package to provide an easy way to start get_data.tcl
without having to open a console to type the command. This is to simply drop the script onto the
WIZE icon, e.g., to run get_data.tcl, highlight the tcl script and move/drop it onto the WIZE icon to
start the script.
get_data.tcl runs wget.exe twice, first to obtain the filename of the most current data file and second
to download that data file. The downloaded data file is then concatenated to the existing srxxx.log
file of transmissions, which is used by Tide.tcl.
7. Start Tide.tcl (drop icon onto WIZE icon)
This short cut has already been created by the Package to provide an easy way to start Tide.tcl and
the Map clients without having to open a console to type the command. For Tide.tcl simply drop the

script onto a WIZE icon, e.g., to run Tide.tcl, highlight the tcl script and move/drop it onto the WIZE
icon to start the script.
Tide.tcl reads and decodes the existing data in the current days SRLOG file (srxxxxx.log), and then
reads Tile_List and automatically displays small same-sized windows for each station data. Note
that at the beginning, when srJJJYY.log is small and/or little station data are obtained, only a small
number of small windows will be displayed. If you rerun the Tide.tcl later on the first day of your
usage, it will display more station data since srJJJYY.log is larger.
8. Start map clients (double click each map client shortcut icon onto the desktop)
Each shortcut opens a map of a different region showing the stations (PACIFIC, INDIAN,
CARIBBEAN, ATLANTIC). More than 1 map client can be run at the same time. Note that Tide.tcl
must always be running since the maps are only clients of Tide.tcl.
You may also start Tide.tcl and the map clients manually by:
a. Double click on the C:\Tcl\bin\wize.exe to display console for each script.
b. In the console window, type % wize Tide.tcl &
c. In the console window, type (example is for Pacific Ocean map)
% wize Client.tcl PACIFIC &
Important Notes:
get_data.tcl must always be running in order to collect the sea level data. This means your computer
must always be on and logged in. If not, data gaps will display for the times when data are not
collected!
Tide.tcl time series plots are output as postscript files. To view and/or print, you may need additional
software, such as Adobe Acrobat, Distiller (to create a .pdf file), ghostscript (to convert .ps to other
formats such as .jpg, .gif, or .png files), or other conversion software
Examining Historical Data
Tide Tool has the ability to examine old logfiles. In historical mode, you must to be sure that the
metadata used is consistent with the date of the logfile you wish to examine (e.g., you may have to edit
in the correct metatdata filename in Tide.tcl before running). Sensors and data units do change through
time. To run, the argument H must be added to the command Tide.tcl.
To run, double click on C:\Tcl\bin\wize.exe, then type
% wize Tide.tcl H
The widget requires three entries: Julian day, Year and SPAN (number of days of data you wish to read
starting at the Julian day and year entered); the maximum SPAN is 2, or two days worth of data
decoded. For more information, see Section 5 of Tide Tool Manual.

Appendix III.
Tide Tool using Internet Data Download
Linux / Unix
Developed and maintained by the Pacific Tsunami Warning Center
Distributed by International Tsunami Information Center

For more information, contact

Stuart Weinstein ([email protected]) or Laura Kong ([email protected])
Version March 2009
(NOTE this description has not yet been updated nor checked for consistency with V9.x)
Tide Tool is an operational tool developed and supported by the Pacific Tsunami Warning Center for the
decode, display, and manipulation of sea level data transmitted over the WMO Global Telecommunications
System (GTS). The tool continuously decodes sea level in real-time and displays the time series on a
computer screen; users are able to interactively manipulate the time series to measure wave amplitude and
period, and to remove the tidal signal. Non-operational versions also exist for data transmissions through
the internet (wget) and for archived data.
Instructions for the installation and running of Tide Tool on a PC under a Linux operating system and with
data collection through the internet ftp are provided below.

Instruction for Tide Tool (Linux version)

1. Put Tide_Tool.tar.gz on the Desktop and extract the files;
> gunzip Tide_Tool.tar.gz
> tar xvf Tide_Tool.tar
Check if
- directory named Tide_Tool is created on the Desktop
- directories named Decoded_Data, MAPS, past_metadata, SR_LOG, Tide_Coeffs and
V8.0 are under Tide_Tool
- files named Tide Tool 8.0 Manual.doc and SLmap-AO/IO/PO.jpg are under Tide_Tool
Create a directory named BOX on the Desktop.
2. Check if the following files are in V8.0;
-- COMP_META
-- TideTool_MetaData, TideTool_MetaData_IO/PO/NEPO/NWPO/SEPO/SWPO/NWAO_CAR
-- URL_FILE
-- Tile_List (sample)
-- Get_data.sh . script to get the logfile srJJJYY.log, which is required to run Tide.tcl
sr23608.log and some past log files are in SR_LOG for reference
-- Tide.tcl script to monitor the sea level data (near real-time)
-- get_data.tcl .. script to continuously download sea level from NWS gateway
-- getTide.tcl script to show the past sea level data
-- Pacific_Client.tcl, Indian_Client.tcl, Caribbean_Client.tcl scripts to display a map
-- pacific_ocean_map.gif, indian_ocean_map.gif, caribbean_map.gif..maps for each ocean

3. Install TclTk & BLT and set the path.

Download software of TclTk (8.4.14 or upper, lower than 8.5) and BLT and install on your machine.
http://www.tcl.tk/software/tcltk/ for TclTk
http://sourceforge.net/project/showfiles.php?group_id=18616 for BLT
How to install is mentioned in the instruction of each source, so please follow the instruction.
How to set the path (commands, filenames, file locations, etc.) depends on your machine.
Here is an example;
> whereis tclsh8.4 .Find the location of tclsh8.4
> whereis wish8.4 .Find the location of wish8.4
> whereis bltwish .Find the location of BLT
> vi .bash_profieOpen the setting file to edit.
File name might be different if yours is not using bash but tcsh and so on.
With the command ls al, you can search which file is for the setting.
Add the path for tclsh8.4, wish8.4 and BLT in the setting file.
For example, when all of the three are in /usr/local/bin, add the following lines;
PATH=$PATH:/usr/local/bin
Export PATH
There should be a "lib" directory where the Tcl/Tk libraries (libtcl8.4.s0, libtk8.4.s0 e.g.) live. This is
where the BWidget package needs to be dropped. The BWidget package can be obtained here:
ftp://ilikai.soest.hawaii.edu/ptwc/Tide_Toolv9/bwidget-1.9.4.tar
Untar this file and drop in the lib directory described above.
4. Edit Tide.tcl to specify different filenames or directory paths than currently specific.
Some variables that you may want to change are:
a. metadata_filename variable, if you may want only a subset of the data to be decoded and displayed,
Some files available are:
COMP_META (default) will decode all stations
TideTool_MetaData_IO is for Indian Ocean
TideTool_MetaData_NEPO is for Northeast Pacific Ocean
TideTool_MetaData_NWPO is for Northwest Pacific Ocean
TideTool_MetaData_SEPO is for Southeast Pacific Ocean
TideTool_MetaData_SWPO is for Southwest Pacific Ocean
TideTool_MetaData_NWAO_CAR is for Northwest Atlantic Ocean,
Caribbean, Gulf of Mexico
TideTool_MetaData to select user-selected stations
b. tilelistfile, which specifies which sea level station monitors to bring up on Tide.tcl startup. An
example Tile_List is provided. Presently, the default is not to bring up any monitors at startup. For
example, if you need to monitor the 3 stations yapi, mani and davo for prs sensor, then Tile_List
would look like
yapi prs
mani prs
davo prs

top

de_ oo

V8.0

de.tc
Pacific/Indian/Caribbean_Client.tcl
Pacific/Indian/Caribbean_map.gif
get_data.tcl
wget.exe
getTide.tcl
URL_FILE
c. printer_command, which specifies
the command for printing. Presently, the default is set for a
TideTool_MetaData
linux/unix
command.
BOX
TideTool_MetaData_IO/PO/NEPP/NWPO/SEPO/SWPO/NWAO_CAR
COMP_META
i r e c t o r y5. Set your machine to use UTC
Tile_List
so set your(tomachine
to use UTC, or your machine will not create a
r u c t u r e Tide.tcl works following UTCError_LOG
be created)
new srJJJYY.log when the UTC
date
changed
and
shows
Error_LOG_past (to be created)errors.

Tide_Coeffs -- coefficient files for the sea level stations

7. Run the scripts
-- decoded data are
archived
in thisstarts.
directory
Type as followsDecoded_Data
under Desktop/Tide_Tool/V8.0,
and
get_data.tcl
youve
set the path,
justdirectory
type ./get_data.tcl &)
> /usr/local/bin/tclsh
./get_data.tcl
& (if
SR_LOG
-- srJJJYY.log
files
are archived
in this
file is successfully
downloaded
and
then
start
the
decoding
/ displaying
Wait until the 1stMAPS
-- maps for each station
Type as follows under Desktop/Tide_Tool/V8.0, and Tide.tcl starts.
past_metadata
-- past
TideTool_MetaData
> /usr/local/bin/bltwish
./Tide.tcl
& COMP_META
(if youve set the&path,
just type ./Tide.tcl &)
When you need
Tide
a map
Toolof8.0
Pacific
Manual.doc
Ocean, type
> /usr/local/bin/bltwish
./Pacific_Client.tcl &
SLmap_AO/IO/PO.jpg
You can add more
maps if you (to
need,
by adding lines in the same way.
Tide_Tool_LOG
be just
created)
be.tcl
created)
Tide.tcl reads Tile_Tool_LOG_past
srJJJYY.log, which get(to
_data
creates.
Tide.tcl reads Tile_List and automatically displays small data windows for each station
Note that at the beginning, srJJJYY.log is small and not all the station data are obtained,
so only a limited number of small windows will be displayed.
If you rerun the Tide.tcl later on the first day of your usage, it will display more stations
since srJJJYY.log is larger. Then, you can keep those windows forever and need not rerun again.

When you run getTide.tcl, type

> /usr/local/bin/bltwish ./getTide.tcl
(if you have set the path, just type ./getTide.tcl)
Then, put the day, year and span in a small window. # As mentioned above, use UTC to count days.

APPENDIX IV
SEA-LEVEL GAUGE INSTRUMENTATION
Extracted from
PTWC OPERATIONS, SYSTEMS AND PROCEDURES MANUAL, Edition 1.0, December 2008
CHAPTER 6. SEA-LEVEL DATA PROCESSING
C. SEA-LEVEL GAUGE INSTRUMENTATION
a. Coastal Gauges
The sea-level networks of the worlds oceans utilize a combination of NOAA/GOES, JMA/GMS or
EUROSAT satellite transmission communications. These data arrive at PTWC via the NESDIS/NWS
communications routing through the NWSTG onto line 58. The sea-level data is collected and
transmitted from the sea-level station using a Data Collection Platform, (DCP) that are procured from
various manufacturers. The DCPs provide a suite of programmable parameters for data (water-level)
monitoring and data transmission via satellite. Other meteorological sensors can also be utilized in
conjunction with these systems. The water-level data is sampled at a 2 second rate averaged over a 2
minute period and transmitted via GOES or GMS on an hourly basis to PTWC. However, NOS (US
NOAA National Ocean Service) is upgrading sea-level stations to transmit 6 one-minute samples every
6 minutes. Already a number of NOS stations in Hawaii and Alaska region have been upgraded to the
higher transmission rate. Supported by the IOC with funds from NOAA, the UHSLC is upgrading the
GLOSS stations in the Indian Ocean. The upgraded gauges in the Indian Ocean transmit every 12 or 15
minutes, with 12 to 30 one-minute samples for the sensors with the highest sampling rate.
Each sea level site may use one or more of 6 types of sensors:

SPS (Submersible Pressure Sensor) measures water-level sea pressure from 0 45 feet. Backup
pressure sensor code is SPB.

PR1,PR2 Pressure sensors.

BUB - (Bubbler Pressure Sensor) measures water-level sea pressure, from 0 45 feet (utilizing a
nitrogen gas purging system) in a cavity submerged in the water. As sea-level changes, the pressure
in the nitrogen gas cavity changes. Backup bubbler code is BBB.

ENC (Incremental Encoder Sensor) measures water changes developed in a stilling well system,
incrementally changing a pulley using a float and weight. As sea-level changes a float moves up and
down. The operation is similar to the float in your toilet. Backup encoder code is ENB.

PAQ (Primary Aquatrak) measures sea-level with acoustics. The transmitter and receiver are above
the water. Sea-level is determined by the round trip time for the acoustic signal. Backup aquatrak
code is BAQ. PAQ designation is used for NTF gauges.

PWL This is the designation for NOS aquatraks. BWL is the backup NOS sea-level gauge
designation, usually a Paros pressure sensor.

RAD (Radar) measures sea-level using radar. The RAD sensor is placed above the sea-level and
bounces radar signals off the sea-surface. Timing the round trip time yields the sea-level.

The sea-level stations systems are generally located in ports or bay areas in secure areas with facilities.
However, they can be located in remote locations on most any shoreline where there is a suitable area

for anchoring the sensor and the Electronic Systems are safe from the elements. The photos below
(courtesy Richard Nygard) show the basic elements of the Installation at Hiva Oa, Marquesas. The Hiva
Oa station primary sensor is a pressure sensor (SPS) while the secondary gauge is an encoder (ENB).
Hiva Oa Installation

Hiva Oa Stilling Well

Vaisala DCP, wiring, 12v battery, and solar regulator.

b. DART Buoys
The DART buoy concept is illustrated here:
The DART sensor is a pressure sensor that sits on the sea
floor in very deep water and measures sea-level fluctuations
averaged over 15 sec to a precision of 1 mm. The data is
relayed by acoustic telemetry to the buoy and is then
transmitted to satellite. The data is transported to PTWC via
line 58.
Why DART buoys? The Achilles heel of the warning
system is the unnecessary evacuations. An evacuation of the
State of Hawaii will easily cost the State tens of millions of

dollars in lost business revenue and the costs associated with executing emergency procedures. Frequent
unnecessary evacuations will undermine public confidence in the warning system. If people stop paying
attention to the warning system, then we have failed the public.
Why do we have unnecessary evacuations? Because forecasting how large a tsunami generated on the
Pacific Rim will be when it arrives on distant shores is still very much in its nascent stages. Coastal tide
gauges provide an imperfect guide as to how destructive a tsunami will be as it propagates away from
the source. This is because the near shore bathymetry and geometry of the harbor where the tide gauge is
situated have as much effect on the wave-height and inundation as the fundamental characteristics of the
tsunami itself. Along the same coast, it is not unusual to find harbors that resonate strongly with
tsunami and those which show much less tsunami activity. Therefore the picture given by coastal tide
gauges is somewhat murky; and forecasts with respect to damage on distant coasts fraught with peril.
What was needed was a way to distinguish between tsunamis that will cross oceans with destructive
power and tsunamis that may produce measurable, but non-damaging waves on distant coasts.
By measuring the properties of a tsunami far from the influences of near-shore bathymetry and
geometry, the DART buoys give us a look at the tsunami in deep water unfettered by resonant effects.
Furthermore, while modeling tsunami propagation in the near shore environment is difficult, modeling
tsunamis as they make their way into deep water is a much simpler problem and can be simulated with
more robustness.
Therefore both Paul Whitmore of the ATWC (EARTHVU) and Vasily Titov of PMEL (SIFT) have
produced a tsunami wave-height forecast database of pre-computed tsunami scenarios. These models
will make predictions based on basic seismic data as to what size tsunami should be recorded by the
DART and coastal buoys. By comparing the predicted results to the observed, the watchstander should
be able to gauge whether or not the tsunami will be destructive when it arrives on distant shores.