HEM System: Tutorial and User Guide
HEM System: Tutorial and User Guide
HEM System: Tutorial and User Guide
www.geosoft.com
The software described in this manual is furnished under license and may
only be used or copied in accordance with the terms of the license.
Written by, Nancy Whitehead with sincere thanks to Tanya Elliott and Nick
Valleau. Please send comments or questions to [email protected]
Geosoft Incorporated
8th Floor
85 Richmond St. W.
Toronto, Ontario
M5H 2C9
Canada
Tel: (416) 369-0111
Fax: (416) 369-9599
Creating a project 5
Spreadsheet View 10
Filtering Data 19
Picking Anomalies 36
Re-Labeling Anomalies 38
Classify Anomalies 42
Single Profiles 48
Multi-channel Profiles 50
DEFINITIONS:
In this Agreement:
"Licensed Program(s)" means the actual copy of all or any portion of Geosoft’s proprietary software technology, computer
software code, components, dynamic link libraries (DLLs) licensed through the Geosoft license server, including any
modifications, improvements or updates provided by GEOSOFT.
“Effective Date” is the date the Geosoft license is installed. This date is recorded by the Geosoft License server when the
Licensed Program(s) is installed.
"Services" means the Services described on Section 4.
"Termination" means the occurrences contemplated by Section 6 and 7.
LICENSE:
GEOSOFT grants to me a non-transferable and non-exclusive license to use the Licensed Program(s) for my own purposes
whereby the Licensed Program(s) are being used only by myself, on one computer, at any one time.
Title and all intellectual property rights in and to the License Program(s), including, without limitation, copyright, trade secrets
and trade marks, shall remain with GEOSOFT. I agree to refrain from raising any objection or challenge to such intellectual
property rights, or from assisting or causing or permitting other(s) to do so, during the term of the Agreement and thereafter
I may not assign this Agreement or any part thereof or sub-license the rights granted herein, or lend, rent, time-share, sell or
lease the software without the prior written consent of GEOSOFT.
I may not attempt to reverse engineer, de-compile or disassemble the software.
I may not make any attempt to circumvent the License Manager that controls the access to the software use.
TERM:
The Term of this Agreement shall commence on the Effective Date and shall continue until termination, as described in Section
6.
SERVICES:
(i) According to the terms of my initial purchase, GEOSOFT shall make available to me, without additional fees such
corrections and improvements to the Licensed Program(s) as may be generally incorporated into the Licensed Program(s) by
GEOSOFT. (Normally this will be for a period of twelve (12) months).
(ii) GEOSOFT has a strong commitment to customer service and product support. GEOSOFT offers me, subject to applicable
Service Charge(s), continuing support in the form of email or telephone advice and other assistance in problem diagnosis and
the correction of errors or faults in the Licensed Program(s) during the life of this License. When a problem occurs which
appears to be related to errors or faults in the Licensed Program(s), I may contact GEOSOFT and GEOSOFT will make an
honest effort to solve the problem. However, GEOSOFT cannot guarantee service results or represent or warrant that all errors
or program defects will be corrected. Also it is to be noted that each Licensed Program is designed to operate on a Windows
NT (sp 6 or later), Windows 2000 or Windows XP platform.
(iii) Further, if I request service relating to the modification of the Licensed Program(s) to meet a particular need or to conform
with a particular operating environment, GEOSOFT may, at its discretion, modify the Licensed Program(s) to meet these
particular needs, subject to applicable Services Charge(s). However, all intellectual property or other rights which may arise
from such modifications shall reside with GEOSOFT.
WARRANTY:
GEOSOFT does not warrant that the functions contained in the Licensed Program will meet my requirements or will operate in
the combinations which may be selected for use by me, or that the operation of the Licensed Program will be uninterrupted or
error free or that all program defects will be corrected.
Each Licensed Program shall be furnished to me in accordance with the terms of this Agreement. No warranties, either
express or implied, are made to me regarding the Licensed Program.
THE FOREGOING WARRANTIES ARE IN LIEU OF ALL OTHER WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING,
BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OR MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE.
LIMITATION OF REMEDIES
I agree to accept responsibility for the use of the programs to achieve my intended results, and for the results obtained from
use of said Program(s). I therefore accept complete responsibility for any decision made based on my use of the
aforementioned Licensed Program(s).
In no event shall GEOSOFT be liable for any damages arising from performance or non-performance of the Licensed
Program(s), or for any lost profits, lost savings or other consequential damages, even if GEOSOFT has been advised of the
possibility of such damages, or for any claim against me by any other party.
GENERAL:
I agree that this Agreement is a complete and exclusive statement of the agreement with GEOSOFT.
This Agreement supersedes all previous Agreements with respect to the Licensed Programs, with the exception of a current
signed Technical Service Agreements.
GEOSOFT is not responsible for failure to fulfill its obligations under the Agreement due to causes beyond its control.
Should any part of This Agreement for any reason be declared invalid, such declaration shall not affect the remaining portion
which shall remain in full force and effect as if this Agreement had been executed without the invalid portion thereof.
The relationship between the parties is that of independent contractors. Nothing contained in this Agreement shall be deemed
to constitute or create a partnership, association, joint venture or agency.
The provision of this Agreement shall be binding upon me and GEOSOFT and my respective successors and permitted
assigns.
This Agreement will be governed by the laws of the Province of Ontario and applicable laws of Canada.
YEAR 2000:
The Licensed Programs have been tested to conform to DISC PD2000 1:1998 Year 2000 Conformity Requirements
(www.bsi.org.uk/disc/year2000/2000.html), with the exception of clause 3.3.2, paragraph b. Section 3.3.2 paragraph b) requires
that inferences for two-digit year dates greater than or equal to 50 imply 19xx, and those with a value equal to or less than 50
imply 20xx. The Licensed Programs will recognize all two digit years as 19xx. This is to prevent errors importing historical data
that pre-dates 1950. All dates that follow 1999 must use four digit dates in the Licensed Programs.
3
The best way to find information in this system is to use the Search tab to perform a
full-text search of all help topics.
The list below provides contact information for Geosoft Technical Support around the
world.
In this manual, we will guide you through the key steps necessary to install and begin
working with the HEM system. We will also walk you through the necessary
procedures you will need to know to process, analysis and present your HEM data.
For more information on working with HEM data we have included HEM Data
Processing – A Practical Overview by Nick Valleau in the Appendix, page 54. This
paper is available on the Geosoft web site at:
www.geosoft.com/resources/papers/index.asp
Tutorial 1: Working with the HEM System 5
This tutorial uses sample data provided on the Oasis montaj CD. This data
(hemdemo.xyz) is located in the “Oasis montaj|Data|hem” directory. Before you
begin the tutorial, copy the data file to a working directory such as D:\Tutorial.
Creating a project
In order to access the QC QA menus in Oasis montaj, you must have an open
Project. An Oasis montaj "Project" encompasses every item in your working project;
from the data files in your project (databases, maps, and grids), to the tools used
(including auxiliary tools such as histograms, scatter plots etc.), to the project setup
including the menus you have displayed and whether you are working on a map or
profile and the state in which you left it the last time you used it.
The project also controls your working directory. Projects are saved as (*.gpf) files. If
you open an existing project from a directory, the system assumes that all your
project files are located in the same directory. To streamline your work, as well as
keep it organized, you may wish to make sure that your project file is in the same
directory as the other files you want to use. We recommend that each project you
work on have its own project (*.gpf) file. If you use a number of applications or add-
on tools in Oasis montaj that have different menus, you can use the project to display
only the menus you require.
The Project Explorer tool enables you to browse as well as open any project item.
The Project Explorer has two tab windows, the Data window that includes all data
files included in the project and the Tools window that organizes and maintains the
project tools. To access the Tools window click the Tools bar on the bottom of the
Project Explorer and click the Data bar on the top to return to the Data window.
6 Tutorial 1: Working with the HEM System
Important Note: Workspace files (*.gws) used in Oasis montaj prior to version 6.0 can
be easily converted to Project files (*.gpf) simply by opening them in
Oasis montaj 6.0. On the Open Project dialog (File|Project|Open)
select File of Type as "Workspaces (*.gws)" and when asked if you
want to convert the old workspace into a new Oasis montaj project
file, select "Yes". The workspace file will be converted to a project file
and all associated workspace information will be transferred to the
new project file. In addition, the workspace file will remain untouched
so that it can be opened in previous versions.
T O C REATE A P ROJECT :
1. On the GX menu, select Load Menu or click the load Menu icon ( ) on the
main toolbar. The Load Menu dialog will be displayed.
2. Select the (HEM.omn) from the list of files and click the [Open] button. The
system adds the HEM menu on your menu bar.
3. After completing these steps, you are now ready to start using the HEM system. If
you require more details about Oasis montaj capabilities, please refer to the
Oasis montaj Quick Start Tutorials, which can be found on the Help|Manuals
and Tutorials menu.
T O C REATE A N EW D ATABASE
1. On the Data menu, select New database. The Create New Database dialog is
displayed.
1. On the Data menu, select Import and then click Geosoft XYZ. The Import XYZ
data dialog is displayed.
Tutorial 1: Working with the HEM System 9
2. Using the [Browse] button locate the XYZ data file (hemdemo.xyz) from your
working directory.
3. In the Import template box, specify (hemdemo) and then click the [Template]
button. The Import .\hemdemo.XYZ dialog will be displayed.
4. To edit the settings for each channel, select (highlight) the channel of interest
(TIME) and make your changes in the Source Data and Output Channel boxes to
the right.
5. In the Source Data box, select a Format for your time as (HH.MM.SS.SS) and
then in the Output Channel box, specify a Name for the channel as (TIME). For
more information on the parameters, click the [Help] button.
10 Tutorial 1: Working with the HEM System
6. When you have completed modifying your data, click the [OK] button to return
to the Import XYZ data dialog box. Click the [OK] button to import the data and
display it in your spreadsheet window.
Note: An asterisk * character (or blank for text strings) is used to indicate that the
point has no valid data and is a dummy place holder. A double asterisk **
indicates that the data are too wide for the spreadsheet column. To change the
width of a column, place the cursor on the dividing line between the column
headers. The mouse becomes a double arrow (see image above). Click on the
left mouse button and drag the line to the right to increase the column width.
Release the mouse button when done.
Spreadsheet View
When you create or open a database, you see a spreadsheet view. The Spreadsheet
view is your “window” to the Oasis montaj database and it also provides you with
flexibility in setting up your working environment. All data is stored securely in the
underlying database — you simply decide which data you want to display in the
spreadsheet and keep all other data in the background, hidden from view.
The spreadsheet is organized in rows, channels (columns) and lines. Rows and
columns work similar to standard spreadsheets in that you can edit and delete them as
needed. For more information on Spreadsheet views, see the online help (Help|Help
Topics).
The data that you have just imported has already undergone some pre-processing.
This includes lag corrections, merging of location and altimeter data and removal of
the calibration portion of flights.
Tutorial 2: Set Up Your HEM Project 11
The (hemdemo.xyz) data file contains data (6.2 m coil separation) for the following:
• 985 Hz, vertical coaxial inphase and quadrature (cx985i, cx985q)
• 7001 Hz, vertical coaxial inphase and quadrature (cx7000i, cx7000q)
• 385 Hz, horizontal coplanar inphase and quadrature (cp385i, cp385q)
• 6606 Hz, horizontal coplanar inphase and quadrature (cp6600i, cp6600q)
• 34135 Hz, horizontal coplanar inphase and quadrature (cp34ki, cp34kq)
T O S ET P ARAMETERS
1. On the HEM menu, select Project setup and then select Set parameters. The
Number of Frequencies dialog is displayed.
2. Specify the No. of frequencies as (5), as the hemdemo dataset has five
frequencies. Click the [OK] button. The Frequency #1 dialog is displayed.
12 Tutorial 2: Set Up Your HEM Project
3. Using the dropdown lists, select the In phase channel as (cx985i) and the
Quadrature channel as (cx985q). Specify the Frequency (Hz) as (985), the Coil
separation (meters) as (6.2), and Coil orientation as (Coaxial).
4. Click the [OK] button. The system accepts the parameters for Frequency #1 and
displays the dialog for the next frequency (Frequency #2).
5. Specify the Frequency parameters for all five frequencies, as shown in the table
below:
In Phase channel Quadrature channel Hz Coil Coil Orientation
Separation
6. Continue this process, clicking the [OK] button until the parameters are set for all
frequencies.
For example, most datasets from the same HEM bird will have the same frequencies
and coil separation. You can use an (*.INI) file for a given HEM system and load this
into the system when required.
T O S AVE P ARAMETERS
1. On the HEM menu, select Project setup and then select Save Parameters. The
Save parameters to an INI file dialog is displayed.
Tutorial 2: Set Up Your HEM Project 13
1. On the HEM menu, select Project setup and then select Load Parameters. The
Load parameters from an INI file dialog is displayed.
2. Using the [Browse] button, select a Parameter file name that you want to load.
Click the [Load] button. The system loads the parameters.
3. Check that the new parameters are in the system.
*Asinh scaling is used so that the values both above and below zero can be stored.
Asinh(x) = log(x + sqrt(1+x*x)), which behaves like –log(-x) for x<<0, log(x) for
x>>0, and linearly around x=0.
Note: The returned depths values are the apparent distances from the bird to the top
of the conductive half space. The apparent thickness of the overlying resistive
layer is obtained from the returned depths by subtracting the true bird height.
Remember too, that the bird height must first itself be calculated, usually from
radar altitude.
The generated nomogram grid files use the following name convention.
14 Tutorial 2: Set Up Your HEM Project
1. On the HEM menu, select Project setup and then Resistivity nomograms. The
system calculates the nomogram grids and displays the Creation of nomogram
grids dialog.
2. This dialog lists the names of the new Resistivity and Depth nomogram grids that
have been created and stored in your project directory. Click the [OK] button to
continue.
3. To view a nomogram, select the Grid|Display Grid|Single grid menu item. The
Place a grid on a map dialog is displayed.
4. Select the Grid name by clicking the [Browse] button and locating the grid file
(R385h.map) in your project directory.
5. Click the [New Map] button and the grid will be displayed on a new map in your
project.
6. You can repeat steps 2 to 4 and select the map (D385h.map) to display the two
maps as shown below.
Tutorial 2: Set Up Your HEM Project 15
For each frequency, this process creates a database with four channels of data: X, Y,
Conductance and Depth. The X channel represents in-phase and Y channel
quadrature. The data is then gridded to create conductance and depth nomograms.
Note that conductance is stored as Log10 values, and that the in-phase and quadrature
values have been transformed using the formula:
This is an arc-sinh type transform, which handles negative inputs. To get the true in-
phase and quadrature values in ppm, use the forward transform:
You can view these nomograms by loading the corresponding grids into maps. Note
that conductance is plotted in Log 10 units and depth is plotted linearly.
Note that since the field calculations are performed with the receiver coil directly
above the plate, both the horizontal coplanar and vertical coplanar systems are null-
coupled, and give no response (in fact, the vertical co-planar system is always null-
coupled to a semi-infinite plate when flown perpendicular to the plane of the plate).
16 Tutorial 2: Set Up Your HEM Project
A warning message is issued if one or more of the preset coil orientations are not
coplanar, to the effect that only coaxial system nomograms are produced.
The generated nomogram grid files use the following name convention.
• Conductance nomogram grids use the following naming convention: “C” +
“Frequency” + “1st letter of coil orientation”. For example, C500C.GRD is a
conductivity nomogram grid for 500 Hz, coaxial coil orientation.
• Depth nomogram grids use the same convention except that the 1st letter of the
file name is 'H' instead of 'C'.
T O G ENERATE C ONDUCTANCE AND D EPTH N OMOGRAMS
1. On the HEM menu, select Project setup and then select Conductance nomograms.
The Creation of nomogram grids dialog is displayed.
2. This dialog lists the names of the new conductance and depth nomogram grids
that have been created and stored in your project directory. This dialog also
informs you that: Conductance nomograms are created only for coaxial coils;
other systems are null-coupled over plate.
3. Click the [OK] button to continue.
4. To view a nomogram, click the Grid|Display Grid|Single grid menu item. The
Place a grid on a map dialog is displayed.
5. Select the Grid name by clicking the [Browse] button and locating the grid file in
your project directory.
6. Click the [New Map] button and the grid will be displayed on a new map in your
project.
Tutorial 2: Set Up Your HEM Project 17
The HEM System automates a large number of steps in HEM processing, analysis
and presentation. From experience, we have identified two common sources of
potential errors in the system. If the following situations are not applicable, please
contact Geosoft technical support for further assistance.
• Keeping track of channel names
• Keeping track of single and multiple lines
It is important to keep track of the channel names you are using as you progress
through the processing cycle. This is especially important during:
• Filtering
• Drift correction – both finding zero-levels and applying corrections.
Results obtained at these stages are used later in anomaly picking. We recommend
using a book-keeping approach at first, until you gain familiarity with the system.
Most problems typically occur when applying corrections since this process requires
reading a table based on previously entered channel values. If you have problems at
this point, you may want to check that channels are not duplicated or missed in this
table.
When you run the filter GX (HEMFILT) you can apply filters to one channel or a
group of channels. The result (filtered data) will be saved into new channels, which as
a default has the original input channel names + “_F”.
When running the drift correction GX (HEMLEVEL) for each input channel to be
levelled this GX creates a temporary channel to contain the level data that is
18 Tutorial 2: Set Up Your HEM Project
interpolated from the data of the related field in the lookup table. It subtracts the data
in the temporary channel from the data in the input channel to be levelled. The result
is then saved in a new channel, which as a default has the input channel name + “_L”.
The temporary channel is automatically deleted.
When picking anomalies the new channels will have a default of the input channel
name + “_A”.
There are several points in the system in which you have the option of selecting the
lines you want to process. These occur at the filtering and single profile parameter
stage. This functionality enables you to select only a portion of your data for initial
processing and parameter selection.
Please note that when you select specific lines for processing, the system changes the
selection status in the database. For example, if you select particular lines for
filtering, say L10 in the HEMDEMO database, you should notice that the system de-
selects all other lines.
If you want to process all data, you must remember to re-select all lines.
Tutorial 3: Pre-Processing Data 19
A key point to note about profile windows is that they are linked dynamically to their corresponding
database. When you select a value or range of values in either the database or profile window
respectively, they are also highlighted in the other window. This capability keeps you in touch with
your data and gives you an interactive means of accomplishing quality control or analysis tasks.
You can display both your In Phase (cp385i) and Quadrature (cp385q) channels in the same profile
panel. Then with your cursor over the Profile window, right click and from the popup menu select
Panel options. The Panel Options dialog will be displayed, check ( ) the Same scale for all profiles
in panel option, and both profiles will be set to the same scale.
Filtering Data
Once the data has been examined the next steps are usually to define best spike removal, and noise
filtering parameters, and then to pre-process the data using a combination of a non-linear filter (to
remove spikes) and low pass filter (to clean background noise). This section assumes that you have
already examined the data to select filter parameters and therefore, only describes the pre-processing
sequence.
20 Tutorial 3: Pre-Processing Data
T O F ILTER D ATA:
1. On the HEM menu, select Filter. The HEM Filters dialog is displayed.
2. Complete the parameters as shown in the dialog above. The following table provides an extended
description of selected parameters and the rational for entering certain values:
Non linear filter. YES to apply non-linear filter and NO not to, use default (YES).
Filter width (data points) Maximum width of the noise in data points. Features that are
wider than this width will not be changed. Use default (3).
Filter tolerance Only noise of greater amplitude than this tolerance will be
changed. Use default (10).
Lowpass filter YES to apply low pass filter and NO not to. Use default (YES).
Cut-off wavelength Short Wavelength cutoff in fiducials. Default is 0.0. Specify (40)
fiducials.
Channels to filter Use default (list of eight comma separated channel names
specified in the project setup). This configuration processes all
the channels in one step.
Note that you can also apply the above filter parameters to one
channel initially and evaluate the results before applying to all
channels.
Lines to filter For the HEMDEMO project, leave this field blank. The system
will process all currently selected lines.
When processing your own data, please note that you can use this
field to select any subset of your data. Simply type a valid line
selection string, or a group of comma separated line selection
strings. This is useful for testing filters on 1 or 2 lines first.
3. Click the [OK] button to filter the data. The system will filter the data and ten new data channels
containing the filtered results will be created. Note that, HEM “filtered” channels use the
following naming convention: “original channel name” + “_F”.
Tutorial 3: Pre-Processing Data 21
You can display your filtered channels with the original data channels, and examine whether the
filtered results are satisfactory.
1. On the Data menu, select Channels|Display all. All of channels in your database will be
displayed in the spreadsheet window, including the newly created “*_F” channels.
2. Show a profile of cx7000i and cx7000i_F with same scale for flight 55. Set the range of the X-
Axis, 40600 to 41200 and the Y-Axis range as, –130 to –60. The output should look like this:
22 Tutorial 4: Applying Drift Corrections
1. Select (highlight) the HEM.gdb. Deselect all the flight lines except for L5601.
2. Then, display the profiles of several (or all) of the inphase and quadrature channels for flight
L56. Examine the profiles and data to determine the background component. This is where there
is zero signal or no ground response. Note that it is often useful to display altimeter data in a
separate profile window to help in selecting the area of highest altitude or least signal.
Tutorial 4: Applying Drift Corrections 23
3. Highlight a range of data points in the profile window. (It does not matter which channel is
highlighted.) In this example above, data points from fiducial 0 to 2000 are highlighted.
4. On the HEM menu, select Drift correction and then select, Pick zero levels. The Pick zero levels
dialog is displayed.
5. In the Table box, specify a name for the new zero table file as (56hemzero). Then using the
dropdown list, select the Save mode as (Insert). Note that, if there is no file called
(56hemzero.tbl) in your current working directory, both modes will do the same and create a
table file with the first row of data. However, if a file with the name (56hemzero.tbl) exists in
the current directory, the mode “New” will over-write the file and the mode “Insert” will insert
the new data into the existing file and sort the datasets in the file using the TIME field.
24 Tutorial 4: Applying Drift Corrections
6. Using the Time reference channel dropdown list, select (TIME). We will leave the Zero level
channel(s) to the default – a list of the10 unfiltered inphase and quadrature channels.
7. In the Level to set box, specify the default (0.0). Note that, if you leave the value as 0.0, the
system averages the values over the range of selected data. If you use a non-zero value, the
system averages the values over the selected range and then adds your specified value (i.e.
applies a level shift to averaged data). This could be used later to add non-zero level to refine
drift corrections. For more detailed information, click the [Help] button.
8. Click the [OK] button and the HEMZERO dialog is displayed. This dialog indicates that the
(56hemzero.TBL) table file was successfully created. Click the [OK] button to continue.
9. The table file is saved in your project directory. You can open this table file in any text editor
(i.e. Notepad) and view the results. At this point the table file will contain one row of data, as
shown below.
Note: To view the 56hemzero.tbl table file, on the Edit menu, select Edit ASCII file. Using the
[Browse] button, locate the file in your project directory and then, using your default text
editor, open and view this file, as shown below:
10. Now highlight data points from fiducials 21700 to 22200 and repeat Steps 4 to 8. Keeping all of
the settings the same in the Pick zero levels dialog.
11. The table file (56hemzero.tbl) will be updated to now include two rows of data.
12. Highlight data points from fiducials 32000 to 32600 and repeat Steps 4 to 8 and then highlight
data points from fiducials 56100 to 56500 and repeat Steps 4 to 8.
13. The table file (56hemzero.tbl) will be updated and will now have four rows of data (see below).
Tutorial 4: Applying Drift Corrections 25
14. This process (Steps 1 to 12) should be repeated for each flight line. You should create a new
table for each line and ensure that only that line is selected when picking the levels. Also note,
you should include the flight information in the zero table names (e.g. 56hemzero.tbl) to ensure
the correct table is used on the correct line.
Note: Four zero table files (flt55_zero.tbl, flt56_zero.tbl, flt57_zero.tbl and flt58_zero.tbl) are
provided with this tutorial. You can use these table files, or create new zero table files
following the procedure above.
1. Open the database (HEM.gdb). Deselect all the flight lines except for L55.
26 Tutorial 4: Applying Drift Corrections
2. On the HEM menu, select Apply Drift correction. The system displays the Apply HEM drift
correction dialog.
3. Using the Table browse button ( ) locate and select the zero table file (flt55_zero.tbl).
4. From the Channel(s) to level dropdown list, note that the default is a list of comma separated
channel names specified in the project setup stage. However, in this case, we want to level the
“filtered” data rather than the original raw data therefore, you MUST edit this list. To edit, type
“_F” at the end of each channel name displayed in the list.
5. For the Field name(s) in the table dropdown list, this is the same list of comma separated
channel names specified in the project setup stage, we will accept the default names as this list
corresponds with the field names in the table file. Then, from the Time reference channel
dropdown list, select (TIME).
6. Click the [OK] button. Ten new channels are added to the database. All of these new channels
will have “_F_L” in their names. Note that, HEM levelled channels use the following naming
convention: “name of channel to be levelled” + “_L”. For example, cp385q_F is the (filtered)
channel to be levelled, and cp385q_F_L is the corresponding levelled channel. These channels
will be used to pick anomalies, calculate resistivity and plot at a later stage of this tutorial.
7. Repeat Steps 1 to 6 for each of the flight lines. Ensure that the correct line is selected with the
correct table file.
You can display your levelled channels with the original data channels, and the filtered data
channels.
1. On the Data menu, select Channels|Display all. All of channels in your database will be
displayed in the spreadsheet window, including the newly created “*_L” channels. You can view
the profiles of the original, filtered and levelled data at the same scale to observe the changes.
Tutorial 4: Applying Drift Corrections 27
2. On the GX menu, select Run GX. The Run a GX dialog is displayed. Using the [Browse] button,
locate the GX file (dbsubset.gx). Click the [OK] button and the Create new subset database
dialog is displayed.
28 Tutorial 4: Applying Drift Corrections
3. Complete the fields as shown above. The following table provides an extended description of
selected parameters and the rational for entering certain values.
Channels It is a good idea to export only the levelled, filtered EM channels as well as
X, Y and auxillary information, since further processing will not use the raw
EM data. In hemdemo.gdb remove the raw data channels from display as
well as the levelled (_L) channels and then choose to export only the
Displayed channels.
Remove mask Continued processing is only relavent for EM data that has a survey location
dummys specified. The database can be windowed to exclude points without a
location by specifying the Mask channel below as X or Y and setting this
parameter to All dummies
Mask channel X
Compression None
Type
4. Click the [OK] button and a new database (hemlines.gdb) will be created and displayed in your
current project.
Now that we have a new subset database (hemlines.gdb) that only includes the filtered and levelled
data, we can now split the data into lines for further processing.
1. Make sure the database (hemlines.gdb) is selected; this is the database you will work with from
here onwards.
2. On the Utility menu, select Split line and then Split on line channel. The Break up a line based
on a line dialog is displayed.
Tutorial 4: Applying Drift Corrections 29
3. Using the Line to break dropdown list, select line (L5501). Then, using the Line reference
channel dropdown list, select (line).
4. Click the [OK] button. The line L5501 will be broken into component lines based on line
numbers in the reference channel line.
5. Repeat Steps 1 to 4 for the remaining flight lines (L5601, L5701, and L5801) and save the
database changes.
T O C REATE A N EW M AP
4. Click the [Next] button. The Create a New Map dialog will be displayed.
30 Tutorial 4: Applying Drift Corrections
5. Enter a Map name (linepath) and click the [Scale] button. The system automatically determines
a scale that will best fit the data on the default map template. Once the “best-fit” scale has been
determined you can then round the scale to a more useful value (70000).
Note: If no scale is specified, a scale will be chosen which will fit all of the data within the default
map template.
6. To change the default map template, click the [Templates] button. The Map Template Manager
dialog is displayed. From this dialog you can select from a list of predefined map templates, or
you can create a [New] template or [Modify] a predefined one. For more detailed information,
click the [Help] button on the Map Template Manager dialog.
6. Click the [Finish] button and a new blank map (linepath.map) will be displayed.
1. On the Mapping menu, select Line Path. The Line path plot dialog will be displayed.
Tutorial 4: Applying Drift Corrections 31
2. You can leave the defaults as they are. Click the [OK] button and the system displays the survey
lines on your map.
3. The line path of your survey area is displayed on the map. If you zoom into the map and examine
the tie-lines, you will note that they are the 49xxx lines, and can be deselected for gridding
purposes.
32 Tutorial 5: Calculating Apparent Resistivity
2. Complete the fields as shown above. The following table provides an extended description of
selected parameters and the rational for entering certain values.
Look-up type Select the type of value to look up from the dropdow list, resistivity or
depth. In this case (Resistivity).
Inphase channel Select from the dropdow list the inphase channels to use (cp6600i_F_L).
Quadrature channel Select from the dropdow list the quadrature channel to use
(cp6600q_F_L).
Output channel Specify the output channel (this is a new channel to be created). Type
(Res6600).
Nomogram File This should be the resistivity nomogram file for the inphase quadrature
channels created at the project setup stage. The nomogramfile is called
R6600H.GRD where R stands for resistivity, 6600 is the frequency and H
stands for horizontal coplaner coil orientation.
Unit scaling Normal or Log 10. We are using the default setting of (Normal).
Threshold This is a minimum cutoff inphase and quadrature value below which data
Tutorial 5: Calculating Apparent Resistivity 33
1. On the HEM menu, select Resistivity and then select Inversion. The Invert Resistivity Value
dialog is displayed.
34 Tutorial 5: Calculating Apparent Resistivity
2. Complete the fields as shown above. The following table provides an extended description of
selected parameters and the rational for entering certain values.
Height channel (m) This is the height of the HEM bird above ground. Select (radar_alt).
Inphase reference
Select from the dropdow list the inphase channels to use
channel (cp6600i_F_L).
Quadrature
Select from the dropdow list the quadrature channel to use
reference channel (cp6600q_F_L).
Output resistivity
Specify the output channel (this is a new channel to be created). Type
channel (Res6600Inv).
Frequency (Hz)
Frequency that corresponds to the reference channels picked above.
Type (6600)
Coil orientation
Coil orientation that corresponds to the reference channels picked
above. Select (Horizontal coplanar)
Fractional error in
When the fractional error in the output resistivity value reaches this
inversion value the inversion stops. Use the default of 1% (0.01)
Threshold for
The forward model solution has errors on the order of 1 ppm. Below
inversion (ppm) this level the output resistivity will not calcuated. Keep this at (1.0)
Invert
You may choose to invert the In-Phase and Quadrature components
alone, or together. Choose (In-phase and Quadrature)
Error Calculation
If one component is close to an order of magnitude larger than the
other, it will dominate the inversion. If this is the case you can choose
Log Scaling for the errors. For this data use the default (Unscaled).
3. Click the [OK] button. The Res6600Inv channel is created and added to the database
(hemline.gdb).
T O G RID D ATA
1. Open and select the database you wish to process (hemlines.gdb). Using the Line Selection Tool
(right-click on the Fiducial header cell and from the popup menu, select Selections|Selection
Tool) deselect all the tie lines (L49011 to L49091).
Tutorial 5: Calculating Apparent Resistivity 35
2. On the Grid menu, select Gridding|Minimum curvature |Dialog controls. The Minimum
Curvature Gridding dialog is displayed.
3. Using the Channel to grid dropdown list, select the (Res6600Inv). In the Name of new grid file
box, specify (Res6600Inv).
4. Then, in the Grid cell size box, specify a grid cell size. For this tutorial we will leave specify a
cell size of (20). This should normally be ½ to ¼ the nominal sample interval. If not specified,
the data points are assumed to be evenly distributed and the default cell size will be. ¼* (sqrt
(grid area / #data points)).
5. Click the [Advanced>] button to specify additional parameters and click the [<Back] button to
return to the Minimum Curvature Gridding dialog. Click the [OK] button grid the data.
6. The Minimum Curvature grid (es6600Inv.grd) will be created and displayed in your current
project.
36 Tutorial 6: Picking Anomalies
Picking Anomalies
Anomalies are picked automatically and placed in new channels in the database. The naming
convention is similar to other HEM processes, where the new channels have the same name as the
starting channels with the addition of “_A” added to the end of the channel name.
T O P ICK A NOMALIES :
1. On the HEM menu, select Anomalies and then select Picking Anomalies. The Anomaly picking
dialog is displayed.
2. The default for the Channels to pick anomaly is a list of channel names given at the Project
Setup stage. However, for this example, we will pick anomalies in channel (cx7000i_F_L and
cx7000q_F_L). To do this, type these two channel names separated by a comma.
3. We will leave the Lines to pick anomaly parameter to the default, which is blank. You can type
any valid line selection strings (comma separated) however, if you leave the entry empty all the
currently selected lines are used. Then, for the Format of results, select (Values) from the
dropdown list.
4. For the Base level specify (0.0), as the minimum value above which highs may be accepted as
anomalies. Then, for the Minimum amplitude, specify (20.0).
Note: For a point in the channel to be picked as an anomaly, its value must be a peak value that is
greater than the ‘Base level’ and the difference between this peak value and the low value of
the adjacent readings at the left and right troughs of the channel profile must be greater than
the ‘Minimum amplitude’.
5. Click the [OK] button and the HEMPICK2 dialog is displayed. This dialog informs you on the
number, in this case 422, anomalies found.
Tutorial 6: Picking Anomalies 37
6. Click the [OK] button. The system locates multiple anomalies and stores them in two new
anomaly channels (cx7000i_F_L_A and cx7000q_F_L_A). (Note that, the “_A” has been added
to the file names to signify that these new channels contain Anomaly locations for the cx7000i
and cx7000q channels that have in turn been filtered “_F” and levelled “_L”.)
7. Display symbols of the 7000q_F_L_A channel with profiles of the related inphase and
quadrature channels to view the anomalies found. The following example shows the levelled
inphase and quadrature channels in profile format and shows the anomalies (plotted as symbols)
for line L40441.
7. After analyzing the results, you may want to add anomalies to or delete anomalies from the
current list. To do this, edit the anomaly channel (7000q_F_L_A). Experiment by adding an
anomaly to Line L40441, fiducial 18409. Then, delete the last anomaly (fiducial 20381). Note
the symbol is automatically displayed or removed from the profile view.
38 Tutorial 6: Picking Anomalies
Re-Labeling Anomalies
After automatically picking and manually editing anomalies, we need to sort anomalies in each
channel and label, (or re-label) them alphabetically. This label channel can then be used in anomaly
classification and plotting.
T O R E -L ABEL A NOMALIES :
1. On the HEM menu, select Anomalies and then select Re-labelling Anomalies. The Re-labelling
anomalies dialog is displayed.
2. Using the Channel to relabel dropdown list, select (cx7000q_F_L_A) and in the Output channel
name box, specify (cx7000q_Anom).
Note: The Output channel name default (entry left empty) is to use the name given in the Channel
to relabel plus “_C” to form a new name for the anomaly label channel.
3. Click the [OK] button and a dialog is displayed with the number (285) of valid labels found.
4. Click the [OK] button and the system creates an anomaly label channel CXQ1_Anom
containing multiple anomalies. Your results should be similar to the ones shown below.
• Line L40441 –7 anomalies (A,B,C,D,E,F,G)
• Line L40571 — 8 anomalies (A,B,C,D,E,F,G, H)
Anomalies in the channel are numbered alphabetically starting from A in each line.
Note that, since the field calculations are performed with the receiver coil directly above the plate,
both the horizontal coplanar and vertical coplanar systems are null-coupled, and give no response (in
fact, the vertical co-planar system is always null-coupled to a semi-infinite plate when flown
Tutorial 6: Picking Anomalies 39
perpendicular to the plane of the plate). A warning message is issued if one or more of the preset
coil orientations are not coplanar, to the effect that only coaxial system nomograms are produced.
Note: The conductance nomograms were generated in Tutorial 2: Setting up your HEM project
page 15.
T O C ALCULATE A PPARENT C ONDUCTANCE :
1. On the HEM menu, select Anomalies and then select Conductance. The Lookup conductance
from a nomogram dialog is displayed.
2. Complete the fields as described in the following table that provides an extended description of
the parameters and the rational for entering certain values.
Inphase reference Select from the dropdow list the inphase reference channel
channel (cx7000i_F_L).
Quadrature reference Select from the dropdow list the quadrature reference channel
channel (cx7000q_F_L).
Anomaly reference Select from the dropdow list the anomaly reference channel
channel (cx7000q_Anom)
Output conductance Specify the output conductance channel (Cond7000).
channel
Nomogram file Browse to find the nomogram file (C7001C.GRD).
Conductance units Select (Normal) from the dropdow list. Note that, we used coaxial inphase
and quadrature channels and a coaxial nomogram to estimate conductance.
This is due to the fact that under the horizontal coplanar coil configuration,
there is no response when the coils are on top of the thin vertical plate,
therefore nomograms were generated.
3. Click the [OK] button. The system creates a conductance channel (Cond7000) that you can
display in the profile window using a symbol format.
4. The above procedure can be repeated to find apparent depth to top of the body. Choose a
different Output channel (Depth7000), the correct nomogram file (H7001C) and use Log10 for
the conductance units.
40 Tutorial 6: Picking Anomalies
T O E XPORT A NOMALIES :
1. On the HEM menu, select Anomalies and then select Export anomalies to ASCII file. The Export
XYZ data dialog is displayed.
2. Specify the name of the new XYZ data file (HEManom) and click the [Template] button. The
Export XYZ template dialog is displayed.
6. Click the [OK] button. You are returned to the Export XYZ data dialog. Click the [OK] button,
the system creates the anomaly file and saves it, in XYZ format, in your working directory.
42 Tutorial 7: Plotting Results
Classify Anomalies
The HEM system enables you to classify anomalies found into eleven groups (levels)
based on the values in a channel and range of values for each level.
T O C LASSIFY A NOMALIES :
1. On the HEM menu, select Plot and then select Classify anomalies. The HEM
classification of anomalies dialog is displayed.
2. From the Anomaly channel dropdown list, select (CX7000q_Anom) and from the
Reference channel dropdown list, select (CX7000q_F_L).
3. In the Classification channel box, specify a new channel name as (CLASS).
Tutorial 7: Plotting Results 43
4. In the Minimum value box, specify the minimum (base level) used in picking
anomalies as (0.0).
5. In the Level 1 maximum value box, specify (50), in the Level 2 box, specify (100),
in the Level 3 box, specify (150), in the Level 4 box, specify (200) and in the
Level 5 box, specify (250).
6. Keep Level 6 to 10 maximum values entries empty. The system will then group
anomalies with Cx7000q_F_L values above 250 into “Level 6”. Therefore, we
will not have any anomalies in Levels 7 to 11.
7. Click the [OK] button. The system will create a CLASS channel.
T O C REATE A B ASE M AP :
6. Accept the default parameters, and click the [Next>] button. The Figure style
base map dialog is displayed.
7. Accept the default parameters, and click the [Next>] button. The Figure titles
dialog is displayed. Specify a title for your map (Anomaly Symbol Plot), and
click the [Finish] button to display a base map on the open map sheet.
T O P LOT A NOMALY S YMBOLS :
4. Complete the fields as shown above. The following table provides an extended
description of selected parameters and the rational for entering certain values.
Classification Specify the classification channel name (CLASS) created at the
channel previous stage.
Symbol font name HEM system comes with a anomaly font file called
(HEMANOM.GFN). This file is the default setting in this entry
which we use.
Tutorial 7: Plotting Results 45
6. Specify the four channels to post, how to draw the posting text (such as size,
offset to symbols, etc.) and titles for legend as shown above. The following table
provides an extended description of selected parameters and the rational for
entering certain values.
Channel to post at top Default is an empty entry which means no right corner of
right corner of symbol symbol posting at this position. Select (CX7000q_Anom)
because we want to post anomaly number at this corner. Note
that the term “top right corner” refers to the position before the
symbols and postings are rotated.
Channel to post at top Select in (Depth7000).
left corner of symbol
Channel to post at Select in (cx7000q_F_L).
bottom right corner of
symbol
46 Tutorial 7: Plotting Results
Note: This example only shows data for a single line since only one line was
selected in the database.
When you plot a single profile, the procedure is to enter plotting specifications (line,
plotting scale, channel, etc.) via the Single profile dialog box. The system then stores
your specifications in a script (*.gs) file with default filename “_protmp_.gs”. When
the Single profile menu item is run, it first reads the script file for plotting instructions
and then determines how to plot the profile.
When you plot multiple profiles, the procedure requires an additional step, namely,
you must manually create the Geosoft Script file that identifies all profiles to plot,
profile offsets, labels and other parameters. When you are finished, you then select
the Multi-channel profile plotting menu item. This then reads your script file and uses
the Single profile dialog to generate all stacked profiles according to your
specifications.
You have the option of plotting all lines or a subset of lines. When you plot lines, the
system creates a series of map (*.map) files that start with “SP_” plus the line name.
For example, if the line number is L100, the profile plot name will be SP_L100.map.
Profile plotting normally takes some effort to determine the best settings for final
presentation. We suggest that you experiment with the HEM|Plot|Single profile…
menu item to determine the basic parameters that work best for your data.
The first time you work with data, you may want to run the option with the defaults
and no scale factors. The system will then calculate starting scales, which you can
view by accessing the menu option again and looking at the values in the dialog box.
This is a quick way of starting to work.
In addition, you can take advantage of the (_protmp_.gs) file created in the system.
You can use this as a template for creating your own script file for plotting multi-
channel profiles. The basic template contains 10 parameter lines with multiple
parameters on each line. If you want to plot multiple profiles, copy the parameter
48 Tutorial 7: Plotting Results
lines from the previous profile and change the vertical offset, scales and other
parameters as required.
An efficient way to learn about parameters is to examine the script file generated the
first time you run the Single profile dialog box. You may want to copy this file and
then experiment with the copy to determine the effect of varying different parameters.
For additional information, click the [Help] button on the Single profiles parameters
dialog. Parameter information, application notes, and links to other topics, including a
link to an example of a GS Script File are available in the help.
Single Profiles
When plotting single profiles, you must provide certain basic parameters (horizontal
axis reference and channel to plot). Note that, you may want to use your Y channel as
the horizontal axis reference channel for testing before plotting all lines, for each
north-south survey line. When starting to plot profiles for a project, you may also
want to select a specific line.
1. On the HEM menu, select Plot and then select Single profile. The Single profile
parameters dialog is displayed.
Tutorial 7: Plotting Results 49
2. Set parameters as shown above. Note that, initially we will leave the Horizontal
and Vertical map scales blank.
3. Click the [OK] button. The system creates the script file (_protmp_.gs) in your
project directory and creates and displays the new map (sp_l40441.map) with the
profile as shown below. Note that, if you select multiple lines, the system creates
multiple maps but only displays the first map in the sequence. This
implementation is designed to save both memory and plotting time especially in
the case where you have many lines to plot. To see the other maps, you must open
them individually (Map|Open map).
4. The script file (_protmp_.gs) created when you ran the Single profile menu
option, should look as below.
50 Tutorial 7: Plotting Results
Multi-channel Profiles
For the purposes of this tutorial, we have provided a starting script file called
(HEMDEMOnew.gs). This script file plots the inphase and quadrature profiles for
all five frequencies, the resistivity for frequency 6600 and the altimeter data for this
dataset.
1. On the HEM menu, select Plot, then select Multi-channel profiles. The Multi-
channel profile parameters dialog is displayed.
2. Complete the fields as shown, remembering to change the Prefix of map names
field, at the bottom of the dialog, to “Demo”. This prevents the system from
overwriting any previously generated profile maps.
3. Click the [OK] button. The stacked profile map (Demo40441.MAP), as shown
below, will be displayed in your project,.
Tutorial 7: Plotting Results 51
52 Tutorial 8: Archiving Data
T O A RCHIVE D ATA:
1. On the HEM menu, select Archive. The Archive XYZ data dialog is displayed.
8. In the Archive data file box, specify the name of the new archive file
(HEMD_ARC), in XYZ format. The Reference channel (X) provides the system
with a means of determining how to locate the start of channels and determine
their fiducial spacing.
9. Click the [Template] button. The Export XYZ template dialog is displayed.
Tutorial 8: Archiving Data 53
10. Specify the template name (HEMD_ARC.o0) and then select the parameters as
shown above (Note that, when archiving in the future you can include any
additional channels (anomaly classifications, conductance, etc.) that you may
require).
11. Click the [OK] button. You are returned to the Archive XYZ data dialog. Click the
[OK] button. The system creates an ASCII file and stores it in your project
directory. This file contains the header information and column data. All date,
flight, line and time information will be in columns so that the data resembles a
flat ASCII file. In addition, dummies (placeholders) will be removed so that the
file is as compact as possible, as shown below.
54 Appendix
Appendix
Paper presented at the 14th ASEG Conference & Exhibition, Perth, March 2000.
ABSTRACT INTRODUCTION
The theory of frequency-domain helicopter electromagnetic There are a growing number of geophysicists working with
(HEM) surveys is relatively straightforward, but in practice there frequency domain helicopter electromagnetic (HEM) data, in both
are many issues to deal with, both in hardware development and in survey contracting and exploration organisations. More
the software to handle the data processing and analysis (DPA). commercial HEM systems are also available worldwide than in the
past. A practical overview of the data processing and analysis
This paper gives an overview of standard HEM data processing (DPA) procedures for HEM surveys is not available in the
and analysis techniques. This is not new technology, and is literature, to the author's knowledge.
intended only to fill a perceived gap in the literature by
summarising existing basic DPA procedures for typical HEM This paper provides a brief, informal review of basic data
datasets. processing techniques specific to HEM data. This is not a
comprehensive review; theoretical information, detailed
Assuming that the inphase (IP) and quadrature (Q) data for each algorithms, and advanced procedures such as modeling and
frequency has been calibrated to parts per million (ppm) during interpretation are not included. However, it does provide enough
data acquisition, the main procedures are as follows. (when combined with adequate software tools) to enable a
geophysicist to carry out the routine HEM data processing of any
A. DATA PROCESSING dataset.
• Apply standard geophysical pre-processing techniques such The procedures described assume an understanding of
as parallax (lag) corrections. frequency-domain EM theory and knowledge of airborne
• Remove non-geological noise from the raw data using geophysical data processing. The most common HEM coil
appropriate filters. configurations in regular use are horizontal coplanar and (vertical)
coaxial. Several frequencies are recorded, each as inphase (IP) and
• Apply drift corrections (zero level corrections). Remove quadrature (Q) components of the secondary electromagnetic field
system and instrumentation drift from all EM channels. at the receiver coil, in parts per million (ppm) of the primary field
• Archive data and analysis results along with complete at the receiver.
documentation of all archived parameters and necessary
metadata. Practical Examples
B. DATA ANALYSIS For each data processing task described below, an example is
given. Four flights (45 lines) of raw data from a small HEM
• Calculate apparent resistivities and depths. survey in Australia's Northern Territory are used for these
• Pick anomalies for conductive target body location and examples. The survey was flown for Ashton Mining (WA) Pty.
analysis. Requires an EM interpreter, but automatic picking Ltd. by Geo Instruments Pty. Ltd., using a Geotech-built
provides a first pass. Hummingbird system. The system has five frequencies of
operation with a transmitter-receiver coil separation of 6.2 m and
• Analyse anomaly targets. Calculate apparent conductances nominal bird altitude of approximately 35 m. (Bettina Townrow,
and depths to conductors. 1999, personal communication):
• Presentation of results; various combinations of:
Horizontal Coplanar Configuration:385 Hz, 6606Hz, 34135 Hz
❏ Standard geophysical maps - survey lines etc.; Vertical Coaxial Configuration: 985 Hz, 7001Hz
❏ Apparent resistivity and depth maps;
Software used for processing the data consists of standard and
❏ Profile maps of IP, Q, resistivity, anomalies;
specialised tools for HEM data, built in Geosoft Inc.'s Oasis
❏ Classified anomaly symbol maps with annotations montaj data processing environment.
(e.g. Identification letters, apparent depths, etc.);
❏ Detailed multichannel profile plots for each survey line. OVERVIEW
❏ Conductivity depth sections. The primary steps in the DPA sequence for HEM data are
❏ Anomaly report. discussed in the following sections. It is best to begin processing
HEM data on a flight-by-flight basis, splitting flights into
individual survey lines only after noise and drift corrections have
1
Geosoft Australia Pty. Ltd. been applied to the data. Prior to carrying out the HEM-specific
32 Richardson Street,
West Perth, WA 6005, Australia
data processing described in this paper, some calibration and
Phone: 61 (8) 9322 8122 routine pre-processing work are required.
Fax: 61(8) 9322 8133
Email: [email protected]
• Correct GPS locations and apply parallax (lag) corrections; Practical Example
• Merge auxiliary data with the HEM database (if required),
such as locations, radar altimeter, magnetics; For this survey, profile data was examined to determine the
typical size (width and amplitude) of spikes and the wavelength of
• Calculate HEM bird altitude from aircraft radar altimeter. background noise. A quick assessment found background noise in
all frequencies, fairly irregular but with typical wavelengths of 10-
As well, an experienced eye should routinely examine data for 20 fiducials (1-2 seconds). Typical noise amplitudes are in the
problems at all stages of processing. The raw data can quickly be range 0 to 5 ppm, depending on EM frequency. Figure 2 shows
assessed by visual examination in profile form, or gridded. For some excessive noise in this dataset. Occasional large-amplitude
HEM, two important aspects of data quality are noise and signal spikes are one to three points wide.
drift.
These wavelengths of noise can be removed without affecting
Practical Example the geological information in the data. A nonlinear filter (Naudy
and Dreyer, 1968) was applied to all EM channels, to remove
Figure 1 shows preprocessed HEM data for one complete spikes with a maximum width of three points and minimum
flight. For simplicity, only the HEM data is shown, but it is also amplitude of 10 ppm. This was followed by a low-pass filter with
useful to display bird altitude and other information at this stage. a wavelength of 40 fiducials (4 seconds). Examination of the
There is non-survey data on these flights that must be cleaned up filtered data against the original was deemed satisfactory (Figure 2).
Fig. 1. Profile display of all 10 raw HEM data channels against fiducials for Flight 55 of the sample dataset (inphase and quadrature channels from
each of the 5 frequencies). Profiles are scaled to view the useful EM data only. On the computer, colours are useful to distinguish the individual
channels more readily.
It is necessary to remove drift from all EM channels. The The applied drift corrections are normally checked by
procedure during the survey is to fly to high altitude (usually 400 m calculating apparent resistivities from the corrected data at each
or higher) at the start and end of each flight, and often in the EM frequency, and displaying the gridded resistivity as an image.
middle of the flight between survey lines. At high altitude there is Remaining drift or other data integrity problems show up as line-
no ground response, so the secondary field signal should be zero. to-line "leveling" busts. In this case, the best approach is for an
The reasonable assumption is made that the drift at high altitude is EM expert to work with the calculated resistivity results in
the same as that at flight altitude. To carry out "leveling" or "drift problem areas and estimate the error in IP and Q values for each
corrections" in post-processing, the processor looks at the high- channel, at locations between high-altitude zero levels. These
altitude flight data segments to determine the signal levels that error values can be added to the table of "zero levels" for each EM
must be subtracted from the secondary field data to yield zero channel, and the drift corrections reapplied to the whole flight.
values. These "zero levels" are tabulated and then interpolated by Several iterations of this process may be required in difficult areas,
time across the entire dataset to provide an adjustment for every which can be very time-consuming.
data point, for every EM channel.
Fig. 2. Detail of raw HEM data in ppm (vertical coaxial 7000 Hz inphase), showing spike and background noise, with filtered version superimposed.
Nonlinear and lowpass filters were applied, as described in the text. X axis is fiducial counter in tenths of seconds.
Fig. 3. Profiles of horizontal coplanar 6600 Hz quadrature channel for Flight 55, raw and after drift (zero level) corrections. The high altitude flight
segments used for selecting zero levels are indicated, as well as locations of the interactively selected zero levels. X axis is fiducial counter in tenths
of seconds.
New hardware technology has reduced drift problems and end of the flight, we would have left significant drift errors (up
somewhat in recent years, and new software tools allow easier to 30 ppm) in the data.
selection of zero levels, and application of the corrections. Good
survey practices and frequent high-altitude flight segments during This process was repeated for the other three flights in the
a survey can also reduce problems considerably. dataset. Few line-to-line leveling problems are visible in apparent
resistivity images calculated from the leveled data (see next
Practical Example section), so the drift corrections for this 6600 Hz horizontal
coplanar data look quite reasonable.
To demonstrate drift corrections, we will work with a single
EM channel. In practice, all EM channels must be drift corrected After completing the drift corrections, the leveled survey data
for each flight, and the software is designed to work on all and auxiliary information from each flight was split into separate
channels at once. I have chosen an example of excessive drift survey lines in a new database for subsequent processing. The
found in this dataset (Figure 3, top profile). The four segments leveled flight data was archived in case it is needed again in the
where the EM signal is low and inactive are the high-altitude future.
portions of this flight. Several survey lines were flown between
each high-altitude segment. Also, in each high-altitude segment a APPARENT RESISTIVITIES AND DEPTHS
small spike of signal is present. These internal calibration signals
will be ignored for this process. While the immediate goal of HEM surveys is often to locate
prospective anomalies, apparent resistivity maps created from the
To select zero levels for this flight, I examined each high- data are also extremely valuable for geological mapping and
altitude segment closely. I interactively selected a range of data interpretation. Apparent resistivities are also used for checking
unaffected by noise, internal calibrations or ground effects. By data integrity, particularly drift problems as discussed in the
selecting a range of data, rather than a single data point, the data is previous section.
averaged over a period of time to reduce the effects of local noise.
Altimeter data is often helpful in picking optimal zero levels Apparent resistivities can be calculated from each frequency
because you can use the maximum height reached as a guide to the used in the survey. However, the horizontal coplanar coil
best data to pick. configurations are most often used for resistivities, because they
have maximum coupling to horizontal layering. Most of the
A table was created with an entry for each EM channel (the standard resistivity data and maps provided by HEM survey
average ppm value over the selected range), and for the average contractors use one of two calculation methods, or variations of
time (fiducial) value for that range. This was repeated for each them. Both methods assume a uniform conductive halfspace and
high-altitude segment of the line, yielding a table with fiducials in work independently at each data point, for each frequency. The
one column, and corresponding signal levels (zero levels) in two methods are:
separate columns for each EM channel. Table 1 shows the table of
zero levels for Flight 55. On this flight there were four high- 1. Pseudolayer method. Inphase and quadrature data (or total
altitude sections, so there are four selected zero levels. The amplitude and phase) are the input, while apparent
locations of the zero level picks are shown in Figure 3. resistivity and apparent distance to the conductive halfspace
are output, as used by Dighem systems historically (Fraser,
To apply the drift corrections, the observed zero levels from 1978). Principal advantages include:
Table 1 were then subtracted from each EM channel. At each point
during the flight the correct value to subtract was interpolated • the apparent resistivity is usually more accurate since
between the known corrections, based on the time at that point. resistive surface layers are ignored in the calculation;
The bottom profile of Figure 3 shows the results of the drift • the apparent depth to the conductive halfspace, a
correction. As expected, the high-altitude survey data now all byproduct of the calculation, is itself a very useful
contain values close to zero, and all other values have been interpretive tool (after subtraction of bird altitude). It
adjusted accordingly. provides some mapping of the resistive surface layer
thickness, if present.
It is very useful to have the two high-altitude segments in the
middle of this flight because it can be seen from the top profile in
Figure 3 that the drift is nonlinear. At the start and end of the flight 2. Altitude-amplitude method. EM signal amplitude and bird
the signal levels are close to zero, whereas in the middle two altitude are the input, and apparent resistivity is output
sections they have significant positive values. If we had applied (Cheesman, 1998). Assumes that the conductive halfspace
drift corrections based on the high-altitude data only at the start is at the earth's surface. This method is normally more
tolerant of drift problems and poor data quality, requiring
less rigorous data corrections to provide useful results.
/ GEOSOFT TABLE - Flight 55 Zero Levels However, it can be misleading, particularly in areas of
/ resistive cover.
/ time cp6600i cp6600q
1043.6 26.42422535 11.68676056 All calculated resistivities and depths using these methods are
2971.4 53.34983871 41.25552419 "apparent" values, since actual ground conditions are not a
4907.7 73.71405941 16.24762376 uniform conductive halfspace. If the earth was indeed a uniform
6116.5 90.13040936 2.942748538 halfspace and the EM data was perfect, both methods would give
/ End of Data. the same results. Since this is never the case, comparisons of
results from the two techniques are very useful for interpretation.
I recommend using both methods on all datasets where possible.
Table 1. Zero level table for use in applying drift corrections to Flight
55. Normally all EM channels would be included; only the 6600 Hz
channels are shown here. The time column is measured in fiducials Calculated apparent resistivities from the two techniques can be
(1 fiducial = 1 second). gridded and displayed as images. The first resistivity maps created
for an area are examined for leveling busts and coherency. resistivity with depth. Generally, higher frequencies provide
Microleveling of resistivity data for final grids can sometimes be information on shallower features. Thus, areas with significant
successfully used for aesthetic purposes, but requires careful resistive cover according to the depth map, correlate with resistive
consideration because of the nonlinear relationship between the IP areas in the 34 kHz resistivity map. The similarity between the 34
and Q channels and the calculated resistivities. It is preferable to kHz pseudolayer resistivity and the 6600 Hz amplitude-altitude
correct the EM channel drift as described earlier, rather than the resistivity reflects the fact that the pseudolayer method at the
calculated resistivities. higher frequency is responding to very shallow features, while at
the lower frequency the pseudolayer method sees further through
There are many refinements and control parameters for these the resistive upper layers than the amplitude-altitude approach.
methods, particularly with regard to handling or avoiding areas of
low signal strength. One approach for the amplitude-altitude ANOMALY PICKING AND TARGET ANALYSIS
algorithm, for example, is to use only the quadrature signal rather
than total amplitude, in areas where the inphase response is very For mineral exploration, the prime goal of most HEM surveys
low. More advanced modeling can be done as well. is anomaly detection. Anomalies are selected and analysed,
usually starting with determination of apparent conductance and
Use of apparent resistivities from different frequencies yields depth to the causative body.
significant information on the variation of resistivity with depth.
Lower frequencies have greater ground penetration than the higher Anomaly picking normally requires an EM interpreter.
frequencies. Automatic picking provides a first pass, which is then interactively
edited by adding or removing anomaly picks on each survey line,
Practical Example while examining the EM profiles. Many simple (or complicated)
algorithms can be used for the automatic picking. By common
Apparent resistivities for the horizontal coplanar coil convention, the final selection of anomalies is alphabetised along
configurations were calculated along all survey lines using the each survey line for identification.
drift-corrected IP and Q data, with both resistivity methods.
Figure 4 shows profiled results for the 6600 Hz data. At each selected anomaly, apparent conductances and depths to
conductors are then calculated. Most basic methods of doing this
While interpretation of the results is beyond the scope of this assume that the causative body is a vertical thin sheet conductor
paper, it is worth the reader examining the relationships between oriented perpendicular to the survey line, in a uniform resistive
the two calculated resistivities, the bird altitude and the calculated halfspace. Conductance is the conductivity-thickness product of
apparent depth. As well, note that the pseudolayer method often the thin sheet, a useful measure since it is difficult to distinguish
calculates the top of the conductive halfspace to be higher than the conductivity from thickness of the sheet using this simple model.
actual ground surface. This reflects either a problem in the EM Depth is calculated from the HEM bird to the body, so bird altitude
data or the fact that the "halfspace" is not uniform in conductivity. must be subtracted to yield depth from ground surface to the top of
the body.
I then gridded the apparent resistivities and depth data for all
survey lines; these images are displayed in Figure 5 for two of the Vertical coaxial coil pairs are normally used for this calculation
frequencies. Again, note the general agreement between the two since they are maximum-coupled with such bodies. In fact the
resistivity calculation methods and the rough inverse correlation horizontal coplanar configuration has no coupling at all with such
between apparent depth and the amplitude-altitude resistivity. a body when the HEM bird is directly above the thin sheet.
In general, all of these maps provide different information, so The EM interpreter will generally classify the anomalies after
all are useful for a complete interpretation. Apparent resistivities this process, commonly using the calculated conductance as a
from different frequencies yield information on the variation of measure. Like the anomaly picking, automatic anomaly
Fig. 4. Profiles of leveled 6600 Hz IP and Q data for line 40691. In the second panel are calculated apparent halfspace resistivities using the
pseudolayer and amplitude-altitude calculation methods. The third panel shows elevations of HEM bird, ground surface and apparent top of the
conductive halfspace from the pseudolayer method. X axis is fiducial counter in tenths of seconds.
Fig. 5. Maps of apparent resistivity and apparent depth (bird to conductive halfspace) of the survey area, calculated from two horizontal coplanar
frequencies, using both calculation methods. The colour schemes used for the resistivities and depths are shown, crosshairs are spaced 1 km apart.
classifications should then be examined by the interpreter and conductance and apparent depth to top of the body (thin sheet), as
possibly reclassified. For example, the interpreter may use special well as classifications and anomaly ID for each anomaly.
classifications such as "culture" or "surficial". Final classifications
are used in presenting the anomalies as various symbols in maps PRESENTATION AND ARCHIVING OF RESULTS
and profile displays. More complicated classification schemes
might combine multiple channels of information in some manner. As with all geophysical data, there are many potential ways to
present HEM survey information. Images of the HEM data itself
As with resistivities, the simple conductance model used is (IP and Q) are rarely displayed in plan maps because EM signal
rarely correct, so the results are always inaccurate. It is strength varies with bird height so that such maps are difficult to
nonetheless a useful approach to quickly analyse and classify interpret. HEM survey parameters, instrumentation details and
observed anomalies in an automated fashion. An EM interpreter other information are normally plotted in the margins (not
should examine the data to ensure that prospective anomalies are included here for space reasons). Survey lines are commonly
not overlooked. Advanced EM modeling methods can also be included on all maps.
applied to this data to yield more accurate information including
body extent, strike, dip and other parameters, especially from the The most common presentations are various combinations of:
vertical coaxial coils.
• Apparent resistivity maps for horizontal coplanar
Practical Example frequencies (Figures 5 and 7). Very useful for geological
mapping and interpretation. Resistivities are usually
Automatic picking was applied to the leveled Coaxial 7000 Hz displayed using an inverted colour scheme with conductive
inphase and quadrature channels. In this simple example, all peaks areas appearing "hot" (red) and resistive areas "cool" (blue).
were picked which had a minimum data value of zero and a • Profile maps of IP and Q data, often superimposed on
minimum anomaly amplitude of 20 ppm above the local resistivity or other (e.g. magnetic) maps. Figure 8 shows an
background. Figure 6 shows the automatically picked targets for example. These maps are useful for the anomaly interpreter.
these two channels on one survey line.
• Classified anomaly symbol maps with annotations, often
In this dataset there were 296 automatic picks using the superimposed on resistivity or other colour maps; symbols
quadrature data. I decided to use the quadrature channel picks as usually represent conductance classifications or interpreted
my starting point, rather than combining the anomalies from all information such as culture. Figure 7 includes an example.
channels. I then manually added two more anomalies into the Annotations generally include anomaly identification,
database anomaly channel for this line (marked as triangles in apparent depth and other parameters such as apparent
Figure 6), and deleted one. This editing is normally done based on conductance or EM data values at that location (Ontario
all EM channels, but for simplicity I have only displayed the 7000 Geological Survey, 1990).
Hz data in this Figure. The final picks were sorted alphabetically • Simple colour symbol maps, proportional symbol maps and
along each line. similar presentations can quickly highlight important
anomalies, based on apparent conductances, apparent
The final selection of anomalies in the database was then fed depths, interpreted classifications or other parameters.
into the Geosoft apparent conductance algorithm, which uses an
implementation of the PLATE forward modeling approach • Multichannel profile plots for each survey line at full
developed at the University of Toronto (Dyck et al., 1980). horizontal map scale (Figure 9). Symbols and anomaly ID
Automatic classifications were then applied. In this example I letters match those in the classified anomaly maps. These
specified conductance breakpoints of 0.5, 0.75, and 1.0, to give plots are also very effective in colour, to help distinguish
four classification levels. Figure 6 shows the resulting apparent channels.
Fig. 6. Profiles of 7000 Hz IP and Q data along line 40441. Automated anomaly picks and manual modifications are shown. The second panel shows
the calculated apparent conductance and depths (from surface to top of a thin sheet model) for each anomaly. Below the anomaly ID, the bottom
panel shows anomaly classifications based on conductance.
Fig. 7. Example HEM survey map presentation for part of survey area. Included are survey lines, colour image of apparent resistivity, and classified
anomaly symbols with annotated ID, apparent conductance and depth, and 7000 Hz Q value. Full-scale maps would normally have more anomaly
symbols, additional survey specifications and other information in the margin.
ACKNOWLEDGEMENTS
I thank Ashton Mining (WA) Pty. Ltd. for allowing the use of
this survey dataset, and the staff at Geo Instruments Pty. Ltd. for
supplying the HEM data. The Geosoft Technical Solutions group
allowed the use of the specialised HEM processing tools
developed by them. I must also acknowledge the numerous
unnamed contributors to my practical HEM education, particularly
during my years at Dighem Surveys and Processing Inc. Thanks
to Chris Bishop for reviewing the manuscript.
REFERENCES
Cheesman, S., 1998, HEMRES2 GX documentation (online Help for the amplitude-
altitude resistivity inversion algorithm in Oasis montaj software): Geosoft Inc.
Dyck, A.V., Bloore, M. and Vallee, M.A., 1980, User manual for programs PLATE
and SPHERE: Research in Applied Geophysics, 14, University of Toronto.
Huang, H. and Fraser, D.C., 1996, The differential parameter method for
multifrequency airborne resistivity mapping: Geophysics, 61, 100-109.
Naudy, H. and Dreyer, H., 1968, Essai de filtrage non-lineaire applique aux profiles
aeromagnetiques: Geophys. Prosp., 16, 171-178.
Fig. 9. Multichannel profiles for survey line 40391. Scales are optimised for different channel groupings. This example includes IP and Q for all
channels, apparent resistivities, classified anomalies, and bird height. Labeling and scales are designed to be efficient for the interpreter rather than
overly detailed.
Telford, W.M., Geldart, L.P., Sheriff, R.E. and Keys, D.A., 1976, Applied Geophysics:
Cambridge University Press.
Practical Geophysics II for the Exploration Geologist (Chapter 6), 1992, compiled by
Richard van Blaricom: Northwest Mining Association.
Oasis montaj - HEM System User Guide, 1997, by Greg Hollyer and Huanjin Wang:
Geosoft Inc.