August Newsletter GPR SLICE Software

Document Sample
August Newsletter GPR SLICE Software Powered By Docstoc
					                                                                     Key Words:
                                                                             GPS XYZ
                                                                             Differential staggering
                                                                             World file

                                 Newsletter – August 2007

The following organizations have joined the growing community of GPR-SLICE users:

      Division of Anthropology and Archaeology, University of West Florida
      Museum of Anthropology, University of Kentucky
      Universidad Politecnica de Valencia, Spain
      Dept of Anthropology, University of Central Florida
      Dept of Archaeology, Ghent University, Belgium
      Joanneum Research, Hydrogeology and Geophysics, Austria
      Murphys Surveys Ltd, Ireland
      Weston Solutions Inc, Geophysics Group, Pennsylvania

Victor D. Thompson is an Assistant Professor at the Division of Anthropology & Archaeology and was
introduced to GPR-SLICE at the SAA in Austin. Victor was also recommended to the software by
other users and we are grateful for all the positive comments that have been flowing. Victor has been
involved in archaeological remote sensing and just recently received grants to acquire GPR and funds
to train undergraduate students. His research activities can be viewed at

George M. Crothers is the Director of William S. Webb Museum of Anthropology and is an Assistant
Professor of Anthropology at the University of Kentucky. George was recommended to the
software and US Forest Service archaeologists at the Land Between the Lakes were helpful in
influencing George's decision in acquiring GPR-SLICE.

Isabel Rodriguez Abad is a Phd student at the Universidat Politecnica de Valencia in Spain. I had a
chance to first meet Isabel at the Archaeological Prospection meeting in Rome in 2005. She has been
following developments with the software and the introduction of GPR-SLICE in Spain. Isabel's
department was interested in evaluating all the softwares including GPRSIM and GPR-SLICE VIEW
before subscribing. Roger Sala at SOT Prospeccio was very helpful in training Isabel on the software
operations which helped to make their decision to sign on.

John J. Schultz, from the Department of Anthropology at the University of Central Florida has been
involved in forensic archaeological research. He has appeared in several high profile programs

                            GPR_Slice V5.0
including the Discovery channels "Mummy Autopsy" program

Lieven Verdonck is a Phd student in archaeological Geophysics at Ghent University. I had an
opportunity to meet Lieven at the University of Siena workshop last year in Italy where he got to see
the software operations in action during processing of a Roman site.

Rainer Morawetz is a scientist at the Institute of Water Resources Management in Austria. Rainer has
been a user of GPRSIM software over the last year and we were able to convince him to take a
subscription to GPR-SLICE. Rainer attended the University of Granada workshop in Spain in January
and he was able to see GPR-SLICE imaging during our talk.

David Graham at Murphys Surveys Ltd with offices in the UK and in Ireland is the Associate Director of the company, which is involved
in land surveying. David was recommended to the software by another user in Ireland.

John Williams who is the technical manager for the Geophysics Group at Weston Solutions randomly
discovered GPR-SLICE through an internet surfing session. John explained to me that they were
looking for a more powerful imaging software solution than the current manufacturer solution they
were using. John and several of his staff including Mark Saunders, invited me to give Webex internet
training on the software. This was a great experience and a first for me. Weston's Geophysics Group
was able to see a demo of the software at their offices in Pennsylvania while I was in Los Angeles. All
my screen shots and real-time activity on my computer display was broadcast to their
computers. Once they had seen the software in action as well as the images that could be generated,
John and his staff followed up with licensing. (This session was quite a valuable learning lesson for
me too and I hope to be able to follow up on acquiring Webex equivalent capabilities in the future to
provide real time training over the internet using web conferencing tools).

In addition to our new organizations signing on, Rick Trnka from the Seminole Tribe of Florida
ordered an additional 2 licenses of GPR-SLICE and GPR-SLICE VIEW to handle their growing
activity in GPR surveying.

There are now 92 organizations using GPR-SLICE Software!


Several new options were added to GPR-SLICE over the last 2 months:

   1. Automatic World File creation added to the Pixel Map menu
   2. Time slice identifier adjusted to give quick multi-grid definitions in the Pixel Map menu
   3. Range gain graphic increase in the Convert Data menus
   4. GPS XYZ creation options engaged for 0-mean line, 0-mean grid or XYZ line match in the
      Slice/Resample menu
   5. Gain factor added to migration operations in the Filter menu

                             GPR_Slice V5.0
1) Automatic World File creation

   As more and more GIS referenced surveys are required, we have made the options to generate
World Files easier for jpeg and bitmaps outputs to GPR-SLICE. Before the users would have to make
several settings such as removing axes, adjusting the corner shifting of the plots and removing the
autoscaling features. Clicking on the World File check box now conveniently placed in the Pixel Map
menu (Figure 1), the user can automatically create world files which are necessary to open up the UTM
referenced images directly into GIS software. The only requirement when the World File is shut off is
that the user will have to go back into the Options menu and turn back on the axis displays.

2) Multi-grid time slice identifier

    The Pixel Map menu used to have 2 columns in which the user could display 2 different grid maps
side-by-side. In the case where one wanted to put 3 or more grid maps side-by-side for
comparison one would need to explicitly type these into the menu items. The identifier for time slices
now accepts multiple names which require the user to separate them by a comma. In Figure 1, an
example showing gridsets a,b,c is shown. The grids are multiplexed into the menu items
automatically. This feature is very valuable for those working with multiple gridsets in the same
folder. It is particularly valuable for appending operations of multiple grids. The transform settings
for each of the individual grids can of course be conveniently adjusted using the Auto-Gain All button
in the Transform menu on display.

3) Range gain graphic

   The graphic for applying gain during data conversion was increased for all the different radar
formats (e.g. Figure 2). During use of the graphic, particularly for Sensors and Software and Mala
conversion where data are recorded as 16 bit ungained, more on screen resolution is available to adjust
the gain curve. The maximum gaining for Mala was also increased to 400 as it was for Sensors and
Software last year. The raw gained pulse will also track the 0 line slightly better for applying gain. In
cases where the DC drift was quite large at depth for deep surveys with a lot of noise, when gain was
applied the entire signal for SS and MALA, the resulting signal could be completely clipped. The 0
line is now better tracked for the raw pulse before gain is applied to it. This is
accomplished by subtracting average of the signal below the halfway part of the scan where the DC
drift is most prominent. Removal of the DC drift is then instituted by the Batch Gain - Wobble button
as before (or using the BandPass filtering options in the Filter menu).

4) GPS XYZ creation

  Creation of XYZ time slice data for automatic mosaic removal now include GPS surveys as well.
Previously, 0-mean line, 0-mean grid, or line matching creation were limited to regularly gridded data.
This portion of the menu was reprogrammed to also include GPS. As a result, the number of menu
buttons in options have been simplified. 0-mean line operations are the best for automatic mosaic
removals, although 0-mean grid can also work well in some situations. 0-mean line operations though
can remove linear features parallel to the profile line so this operation should only be used after
examining raw time slices to see if linear features are present first. If not, then 0-mean line can work
the best to remove mosaic noises automatically without loss of imaged structures.

                               GPR_Slice V5.0
5) Migration gain factor

   Migration requires applying a gain to the processed radargram to boost the effects of this filter.
Before, a setting called Overall Gain was present in the Filter menu and which could be applied to any
of the filtering processes and not just migration. This gain factor is now redirected solely to migration
 The overall gain was placed below the migration menu item and will default to 4 if the gain is set to
1. Gaining of a migrated radargram is necessary in order to account for the changes in the dynamic
range after the migration process is applied. A value of 4 will often work well for most radargrams,
however, this setting can be adjusted depending on the scan density and the width of the migrator
being used.

The update is on the website, password "Gypsy Ln".

Recent Events

The 3rd annual Strawtown GPR/GPR-SLICE workshop from July 9-11 that was hosted by Dr. Bob
McCullough of Indiana-Purdue University had 15 attendees including students and users from the
Indiana Geological Survey and Southern Illinois University. Several equipments were used at the
workshop this year, including a Sensors and Software Pulse Ekko system supporting a 100 MHz
bistatic antenna with GPS that was brought by Amzie Wenning from the Indiana Geological Survey.
This was the first GPR-GPS survey done at the Strawtown workshop. While other equipments were
being used to train students on standard high density regular surveys, the GPS survey was done using
relatively coarse line spacing along with some overlapping in the recorded antenna track (Figure 4).
Even though the profiling on the random GPS survey was approximately 2-3 meters apart, time slices
generated from the survey were able to show a native Indian enclosure on the shallow time slice
maps. The GPS images were made in about 20 minutes after the data was collected and transferred to
the computer. If anyone would like to process this data for their own practice in doing a GPR-GPS
survey in the software, please send an email and I can post the radargrams and the navigation file on
our ftp site.

The National Park Service workshop in Richland Washington from May 13-18 gave me an opportunity
to meet with several users including Shawn Steinmetz of the Umatilla Tribe. Shawn showed me some
data he collected and he wanted to get some advice on improving the image (Figure 5). Ranges greater
than about x=12 m shows staggering noises in the image, however, ranges less than 12 meters appear
to be fine. The causes for the differential staggering noises across the radargram is not completely
known, however, Shawn believes it was a result of some ground disturbance caused by a trench he
surveyed over on that part of the site. Whatever the cause, to correct this image one does not want to
simply apply staggering to all the reverse lines in the file, but only to those files that the staggering
appears in. To implement a differential staggering correction I went into the Reverse menu and
clicked off those radargrams that did not appear to have any problems (e.g. up until x=12 m) and then
had the radargrams checked on as usual for the reverse files. The edits are then saved to the info.rev
file using the Save Edits button - noting that the radargram reversing operations are NOT run
again. After redefining the info.rev we can go back into the Grid menu, set the staggering length and
in this case to 1 meter, and then the grids are recalculated. The stagger corrected image looks
significantly better than the raw uncorrected image. Users have a easy way to apply staggering
differentially to the images if need be by saving customized reverse files.

                              GPR_Slice V5.0
In late May I was invited to be an instructor for 3 day workshop on GPR and other methods for
archaeological geophysics at Ft Lewis Washington. The workshop involved teaching state
archaeologists on what to expect from geophysical surveys as many of the attendees are responsible for
writing RFPs (request for proposals) and they needed to know what contractors should provide them.
We were able of course to raise the bar very high and introduce the attendees to 2D/3D depth slices
and isosurface rendering animations as a form of data display and output from GPR. Many were
surprised at the quality of subsurface imaging that could be achieved with GPR. Some that attended
had a history of poor results from earlier GPR contracting. The reason they stated was that the few
professional surveys performed in Washington State only contained radargram profiles with no
imaging provided by local contractors.

Magnetics, Resistivity as well as Aerial Photography were made of the Ft Lewis site during the
demonstration of geophysical surveying. The site was believed to contain a destroyed historic ranch
and church that was run by native Indians who were eventually forced off their land during the
acquisition of lands to build the American military fort. A comparison map showing the recorded
geophysics is given in Figure 6. The extent of surface rock (burials) could be identified on the shallow
time slices and the possible continuations of these features just below the ground cover may have also
been imaged at parts of the site. One interesting note that came out of the surveys was that the crop
marks were clearly defined on GPR (top time slices) and were not recorded in the magnetic survey.
Some similar anomalies between GPR and Magnetic can be seen in the diagram.

Upcoming Events

The University of California at Santa Barbara extension class on GPR will be held August 6-10
( There are still several openings should
anyone want to attend. I had an opportunity to join in on the pre-site selection in June and was able to
spend 2 days on the island. Santa Cruz is certainly the most exotic place I have ever been too that is
within 2 hours of Los Angeles by car and boat. The part of the island we are working on is only open
to researchers and the general public can not visit. It is a great chance for those that want to get away
and see some unique geography, camp out, and also to do GPR on a remote part of the island.

Have a great summer...Dean

This newsletter is available in a word document on the Subscribers Only page of the www.GPR- website.

                              GPR_Slice V5.0
Figure 1. Screen shot showing the addition of World File generation, and the new option to include
multiple time slice identifiers in the Pixel Map menu.

                            GPR_Slice V5.0
Figure 2. Adjustments to the conversion menus include increased graphic scaling for the gain curve as
well as better 0 tracking of the raw signal in the case when strong DC drifts in the lower parts of the
gain radargrams exist.

                             GPR_Slice V5.0
Figure 3. GPS XYZ time slice creation buttons are now also included in 0 mean line, 0 mean grid, and
XYZ line match operations.

                            GPR_Slice V5.0
Figure 4. Results from a GPR-GPS survey made at the Strawtown GPR workshop at the Indiana-
Purdue University Archaeological field school in Indiana is shown. (Data courtesy of the Indiana
Geological Survey).

                            GPR_Slice V5.0
Figure 5. An example of applying differential staggering to a time slice image is shown. The
operation include custom settings of the info.rev file along with rerunning the gridding use an
appropriate stagger length.

                             GPR_Slice V5.0
Figure 6. Comparison showing GPR time slices, magnetic, and resistivity surveys made at the Ft
Lewis site, Washington State. The magnetic and resistivity data was generated in GPR-SLICE using
the Import Geophysical Data option in the Grid menu.

Dean Goodman
Geophysicist, Phd
Geophysical Archaeometry Laboratory

 20014 Gypsy Ln
 Woodland Hills, CA 91364
 818-434-9932 cell

                            GPR_Slice V5.0

Shared By: