ZTEM탐사보고서 Northwestern Ontario지역

ZTEM Airborne Geophysics
Thunder Bay North Project

Thunder Bay Mining Division
Northwestern Ontario
2011-2012
Panoramic PGMs (Canada) Limited

Formerly Magma Metals (Canada) Limited
P.O. Box 10628, Thunder Bay, Ontario P7B 6V1
1004 Alloy Dr., Thunder Bay Ontario P7B 6A5
Overview
In 2011 a 629 line km airborne ZTEM geophysical survey was flown. Included within this report are the logistic
reports and interpretations produced by the contracted company as well as an additional interpretation report
done by an independent geophysical consulting company. This cover page is meant as a brief summary of the
results, for detailed information refer to the individual attached reports.
During 2011 Geotech Ltd. was contracted to fly a ZTEM survey over the Thunder Bay North Project area. To help
interpret the results Condor Consulting Inc. was contracted to verify the quality of the data and to independently
interpret the results of the airborne geophysical survey. It was concluded that while numerous anomalies were
found they required drill testing to determine whether they represented a subtle change in lithology, structure or
potential mineralization.
Conclusion
Several correlations were made between known intrusives and mineralization and responses on the ZTEM survey.
In addition several similar responses were identified in the ZTEM survey that could represent potential
mineralization. Further investigation of these areas by drill testing the highest ranked targets is recommended.
Attached
Report on a Helicopter-Borne Z-Axis Tipper Electromagnetic (ZTEM) and Aeromagnetic Geophysical Survey –
Geotech Ltd.
Report on Processing & Analysis of ZTEM EM and Magnetics Data Thunder Bay North Property – Condor Consulting
REPORT ON A HELICOPTER-BORNE
Z-AXIS TIPPER ELECTROMAGNETIC (ZTEM)
AND AEROMAGNETIC GEOPHYSICAL SURVEY
Thunder Bay North Block

Thunder Bay, Ontario
For:
Magma Metals Limited
By:
Geotech Ltd.
245 Industrial Parkway North
Aurora, Ont., CANADA, L4G 4C4
Tel: 1.905.841.5004
Fax: 1.905.841.0611
www.geotech.ca
Email: info@geotech.ca
Survey flown May – June 2011

Project 11182
August, 2011
8/12/2011
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited 1
TABLE OF CONTENTS
EXECUTIVE SUMMARY ............................................................................................................................. III
1. INTRODUCTION.................................................................................................................................... 1
1.1 General Considerations..............................................................................................................................1
1.2 Survey Location...........................................................................................................................................2
1.3 Topographic Relief and Cultural Features ...............................................................................................3
2. DATA ACQUISITION............................................................................................................................. 4
2.1 Survey Area .................................................................................................................................................4
2.2 Survey Operations ......................................................................................................................................4
2.3 Flight Specifications ....................................................................................................................................5
2.4 Aircraft and Equipment ...............................................................................................................................5
2.4.1 Survey Aircraft .....................................................................................................................................5
2.4.2 Airborne Receiver................................................................................................................................5
2.4.3 Base Station Receiver ........................................................................................................................6
2.4.4 Airborne magnetometer......................................................................................................................7
2.4.5 Radar Altimeter....................................................................................................................................7
2.4.6 GPS Navigation System.....................................................................................................................7
2.4.7 Digital Acquisition System..................................................................................................................8
2.4.8 Mag Base Station ................................................................................................................................8
3. PERSONNEL......................................................................................................................................... 9
4. DATA PROCESSING AND PRESENTATION .................................................................................... 10
4.1 Flight Path ..................................................................................................................................................10
4.2 In-field Processing and Quality Control..................................................................................................10
4.3 GPS Processing ........................................................................................................................................10
4.4 ZTEM Electromagnetic Data....................................................................................................................11
4.4.1 Preliminary Processing.....................................................................................................................11
4.4.2 Geosoft Processing...........................................................................................................................11
4.4.3 Final Processing ................................................................................................................................12
4.4.4 ZTEM Profile Sign Convention ........................................................................................................12
4.4.5 ZTEM Quadrature Sign Dependence .............................................................................................13
4.4.6 Total Divergence and Phase Rotation Processing .......................................................................14
4.4.7 2D EM Inversion ................................................................................................................................14
4.5 Magnetic Data............................................................................................................................................15
5. DELIVERABLES ................................................................................................................................. 16
5.1 Survey Report............................................................................................................................................16
5.2 Maps ...........................................................................................................................................................16
5.3 Digital Data.................................................................................................................................................16
6. CONCLUSIONS AND RECOMMENDATIONS ................................................................................... 20
6.1 Conclusions................................................................................................................................................20
6.2 Recommendations ....................................................................................................................................20
7. REFERENCES AND SELECTED BIBLIOGRAPHY........................................................................... 21
LIST OF FIGURES
Figure 1 -Property Location........................................................................................................................... 1
Figure 2 – The Block, with ZTEM and Magnetic Base Station Locations...................................................... 2
Figure 3 - Google Earth image of the block................................................................................................... 3
Figure 4 - ZTEM System Configuration......................................................................................................... 6
Figure 5 - ZTEM base station receiver coils.................................................................................................. 7
Figure 6 - ZTEM Crossover Polarity Convention for Tzx and Tzy for the block .......................................... 12
Figure 7 - Illustration of ZTEM In-Phase & Quadrature Tipper transfer function polarity convention.......... 13
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited . i
LIST OF TABLES
Table 1 - Survey Specifications..................................................................................................................... 4
Table 2 - Survey schedule............................................................................................................................. 4
Table 3 - Acquisition and Processing Sampling Rates.................................................................................. 8
Table 4 - Geosoft GDB Data Format........................................................................................................... 17
APPENDICES
A. Survey location maps..................................................................................................................................
B. Survey Block Coordinates...........................................................................................................................
C. Geophysical Maps ......................................................................................................................................
D. ZTEM Theoretical Considerations ..............................................................................................................
E. ZTEM Tests over Unconformity Uranium Deposits.....................................................................................
F. 2D Inversions ..............................................................................................................................................
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited . ii
REPORT ON A HELICOPTER-BORNE Z-AXIS,
TIPPER ELECTROMAGNETIC (ZTEM) AND AREOMAGNETIC
GEOPHYSICAL SURVEY
Thunder Bay North Block

Thunder Bay, Ontario
Executive Summary
During May 24th to June 5th 2011 Geotech Ltd. carried out a helicopter-borne geophysical
survey for the Magma Metals Limited over the Thunder Bay North block situated 40
kilometres northeast of Thunder Bay, Ontario, Canada.
Principal geophysical sensors included a Z-Axis Tipper electromagnetic (ZTEM) system, and
a caesium magnetometer. Ancillary equipment included a GPS navigation system and a radar
altimeter. A total of 629 line-kilometres of geophysical data were acquired during the survey.
The survey operations were based out of the town of Thunder Bay, Ontario. In-field data
quality assurance and preliminary processing were carried out on a daily basis during the
acquisition phase. Preliminary and final data processing, including generation of final digital
data and map products were undertaken from the office of Geotech Ltd. in Aurora, Ontario.
The processed survey results are presented as the following maps:
• Total Magnetic Intensity

• 3D View of In-Phase Total Divergence versus Skin Depth
• In-Phase Total Divergence (30Hz, 90Hz and 360Hz)
• Tzx In-line In-Phase & Quadrature Profiles over 90Hz Phase Rotated Grid
• Tzy Cross-line In-Phase & Quadrature Profiles over 90Hz Phase Rotated Grid
The survey report describes the procedures for data acquisition, processing, final image
presentation and the specifications for the digital data set. 2D inversions over two selected
lines were performed in support of the ZTEM survey results.
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited . iii
1. INTRODUCTION
1.1 General Considerations
These services are the result of the Agreement made between Geotech Ltd. and Magma
Metals Limited to perform a helicopter-borne geophysical survey over the Thunder Bay
North Block located 40 kilometres northeast of Thunder Bay, Ontario (Figure 1).
Allan MacTavish represented Magma Metals Limited during the data acquisition and data
processing phases of this project.
The geophysical surveys consisted of helicopter borne AFMAG Z-axis Tipper
electromagnetic (ZTEM) system and aero magnetics using a caesium magnetometer. A total
of 629 line kilometres of geophysical data were acquired during the survey. The survey area
is shown in Figure 2.
In a ZTEM survey, a single vertical-dipole air-core receiver coil is flown over the survey
area in a grid pattern, similar to regional airborne EM surveys. Two orthogonal, air-core
horizontal axis coils are placed close to the survey site to measure the horizontal EM
reference fields. Data from the three coils are used to obtain the Tzx and Tzy Tipper (Vozoff,
1972) components at six frequencies in the 30 to 720 Hz band. The ZTEM is useful in
mapping geology using resistivity contrasts and magnetometer data provides additional
information on geology using magnetic susceptibility contrasts.
Figure 1 -Property Location
8/12/2011
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited 1
The crew was based out of Thunder Bay, for the acquisition phase of the survey. Survey
flying was started on May 24th 2011 and finished on June 5th 2011.
Data quality control and quality assurance, and preliminary data processing were carried out
on a daily basis during the acquisition phase of the project. Final reporting, data presentation
and archiving were completed from the Aurora office of Geotech Ltd. in August, 2011.
1.2 Survey Location
The block is located approximately 40 kilometres northeast of Thunder Bay, as shown in

Figure 2.
Figure 2 – The Block, with ZTEM and Magnetic Base Station Locations
The survey was flown in a SW to NE (N 60° E azimuth) direction, with a flight line spacing
of 200 metres, as depicted in Figure 3. Tie lines were flown perpendicular to the traverse
lines at a spacing of 2000 metres, in a SE to NW (N 150° E azimuth) direction. For more
detailed information on the flight spacing and direction see Table 1.
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited . 2
1.3 Topographic Relief and Cultural Features
Topographically, the block exhibits a shallow relief with an elevation ranging from 420 to
530 metres above mean sea level over an area of 115 square kilometres (Figure 3). The
survey area has various rivers and streams running throughout which connects various lakes
and wetlands. There are visible signs of culture throughout the survey, such as, roads also,
there is a mining area located along flight lines 1390 and 1380 at the southwest corner.
Figure 3 - Google Earth image of the survey

The survey area is covered by numerous mining claims which are plotted on all maps. The
survey area is covered by NTS (National Topographic Survey) of Canada sheets 052A10,
052A11, 052A14 and 052A15.
2. DATA ACQUISITION
2.1 Survey Area
The survey block (see Location map in Appendix A and Figure 2) and general flight
specifications are as follows:
Table 1 - Survey Specifications
Traverse Line Area Planned Actual 1

Survey block Flight direction Line numbers
spacing (m) (Km2) Line-km Line-km
Traverse: 200 549 565 N 60° E / N 240° E L1000 – L1410

Block 115
Tie: 2000 80 81 N 150° E / N 330° E T1800 – T1870
TOTAL 115 629 646

Survey block boundaries co-ordinates are provided in Appendix B.
2.2 Survey Operations
Survey operations were based out of Thunder Bay, Ontario on May 24th, 2011 until June 5th,
2011. The following table shows the timing of the flying.
Table 2 - Survey schedule
Date Flight # Block Crew location Comments
24-May-2011 Thunder Bay, ON System assembly

25-May-2011 Thunder Bay, ON System assembly
26-May-2011 Thunder Bay, ON System assembly completed
27-May-2011 Thunder Bay, ON Helicopter delayed
30-May-2011 Thunder Bay, ON Helicopter install completed
31-May-2011 Thunder Bay, ON Testing completed
1-Jun-2011 Thunder Bay, ON No production due to weather
2-Jun-2011 1,2 Thunder Bay, ON 149km flown
3-Jun-2011 3 Thunder Bay, ON Production aborted due to weather
4-Jun-2011 4,5 Thunder Bay, ON 141km flown
5-Jun-2011 6,7 Thunder Bay, ON 329km flown – flying completed
1
Actual line-km represents the total line-km contained in the final databases. These line-km normally exceed the Planned
line-km’s, as indicated in the survey NAV files.
2.3 Flight Specifications
During the survey the helicopter was maintained at a mean height of 163 metres above the
ground with a nominal survey speed of 80 km/hour for the survey block. This allowed for a
nominal EM sensor terrain clearance of 93 metres and a magnetic sensor clearance of 108
metres.
The on board operator was responsible for monitoring the system integrity. He also
maintained a detailed flight log during the survey, tracking the times of the flight as well as
any unusual geophysical or topographic feature.
On return of the aircrew to the base camp the survey data was transferred from a compact
flash card (PCMCIA) to the data processing computer. The data were then uploaded via ftp
to the Geotech office in Aurora for daily quality assurance and quality control by trained
personnel.
2.4 Aircraft and Equipment
2.4.1 Survey Aircraft
The survey was flown using a Eurocopter Aerospatiale (Astar) 350 B3 helicopters,
registration number C-FVTM. The helicopter was operated by Geotech Aviation. Installation
of the geophysical and ancillary equipment was carried out by a Geotech Ltd crew.
2.4.2 Airborne Receiver
The airborne ZTEM receiver coil measures the vertical component (Z) of the EM field. The
receiver coil is a Geotech Z-Axis Tipper (ZTEM) loop sensor which is isolated from most
vibrations by a patented suspension system and is encased in a fibreglass shell. It is towed
from the helicopter using an 85 metre long cable as shown in Figure 4. The cable is also used
to transmit the measured EM signals back to the data acquisition system.
The coil has a 7.4 metre diameter with an orientation to the Vertical Dipole. The digitizing
rate of the receiver is 2000 Hz. Attitudinal positioning of the receiver coil is enabled using 3
GPS antennas mounted on the coil. The output sampling rate is 0.4 seconds (see Section
2.4.7)
GPS Antenna
Radar Altimeter
Antenna
55 m
70 m
85 m
Magnetic Sensor
6.5
EM Receiver Coil
15 m (Including 3 GPS Antenna)
46 m
Figure 4 - ZTEM System Configuration
2.4.3 Base Station Receiver
The two Geotech ZTEM base station receiver coils measure the orthogonal, horizontal X and
Y components of the EM reference field. They are set up perpendicular to each other and
roughly oriented according to the flight line direction. The orientation of both units is not
critical as the horizontal field can be further decomposed into the two orientations of the
survey flight. The orientation of the base stations were measured using a compass.
The base station coils each have a diameter of 3.5 meters, with the coil orientations to the
horizontal dipole, as shown in Figure 5.
The base station receiver coils were installed in an open area just outside the block (48°42.42
N, 88°59.91W). The azimuth of the reference coil was N233oE (named as A) and for the
orthogonal component it was N323oE (named as B). Angles A and B are taken into account
together with the survey lines azimuth to calculate the in-line (Tzx) and cross-line (Tzy) field
utilizing a proprietary software.
GPS Antenna
Figure 5 - ZTEM base station receiver coils.
2.4.4 Airborne magnetometer

The magnetic sensor utilized for the survey was a Geometrics split-beam optically pumped
caesium vapour magnetic field sensor, mounted in a separate bird, and towed on a cable at a
mean distance of 55 metres below the helicopter (Figure 4). The sensitivity of the magnetic
sensor is 0.02 nanoTesla (nT) at a sampling interval of 0.1 seconds. The magnetometer will
perform continuously in areas of high magnetic gradient with the ambient range of the sensor
approximately 20k-100k nT. The Aerodynamic magnetometer noise is specified to be less
than 0.5 nT. The magnetometer sends the measured magnetic field strength as nanoTesla to
the data acquisition system via the RS-232 port.
2.4.5 Radar Altimeter

A Terra TRA 3000/TRI 40 radar altimeter was used to record terrain clearance. The antenna
was mounted beneath the bubble of the helicopter cockpit.
2.4.6 GPS Navigation System

The navigation system used was a Geotech PC104 based navigation system utilizing a
NovAtel CDGPS (Canada-Wide Differential Global Positioning System Correction Service)
enabled Propak V3-RT20 GPS receiver. Geotech’s Navigate software, using a full screen
display with controls in front of the pilot, allows him to direct the flight.
5 NovAtel GPS antennas are utilized during the survey; one is mounted on the helicopter tail
(Figure 4), one installed with the Receiver Base Station (Figure 5) and three are mounted on
the airborne receiver (Figure 4). As many as 14 GPS and two CDGPS satellites may be
monitored at any one time. The horizontal positional accuracy or circular error probability
(CEP) is 1.8 m, with CDGPS active, it is 0.6 m. The co-ordinates of the block were set-up
prior to the survey and the information was fed into the airborne navigation system.
2.4.7 Digital Acquisition System
The power supply and the data acquisition system are mounted on an equipment rack which
is installed into the helicopter. Signal and power wires are run through the helicopter to
connect on to the tow cable outside. The tow cable supports the ZTEM and magnetometer
birds during flight via a safety shear pin connected to the helicopter hook. The major power
and data cables have a quick disconnect safety feature as well. The installation was
undertaken by the Geotech Ltd. crew and was certified before surveying.
A Geotech data acquisition system recorded the digital survey data on an internal compact
flash card. Data is displayed on an LCD screen as traces to allow the operator to monitor the
integrity of the system. The data type and sampling interval as provided in Table 3.
Table 3 - Acquisition and Processing Sampling Rates

DATA TYPE ACQUISITION SAMPLING PROCESSING SAMPLING
ZTEM Receiver 0.0005 sec 0.4 sec
Magnetometer 0.1 sec 0.4 sec
GPS Position 0.2 sec 0.4 sec
Radar Altimeter 0.2 sec 0.4 sec
ZTEM Base station 0.0005 sec __
2.4.8 Mag Base Station
A combined magnetometer/GPS base station was utilized on this project. A Geometrics

Caesium split-beam vapour magnetometer was used as a magnetic sensor with a sensitivity
of 0.001 nT. The base station was recording the magnetic field together with the GPS time at
1 Hz on a base station computer.
The base station magnetometer sensors for the block were installed in an open lot close the
hotel (48°22.816 N, 89°21.1795 W) away from electric transmission lines and moving
ferrous objects such as motor vehicles. The base station data were backed-up to the data
processing computer at the end of each survey day.
3. PERSONNEL
The following Geotech Ltd. personnel were involved in the project.
Field:
Project Manager: Adrian Sarmasag (Office)
Data QC: Emilio Schein (Office)
Crew chief: Colin Lennox
Operator: Les Moschuk
The survey pilot and the mechanical engineer were employed directly by the
helicopter operator – Geotech Aviation.
Pilot: Andres Par
Mechanical Engineer: Chris Ward
Office:
Preliminary Data Processing: Emilio Schein
Final Data Processing: Lily Manoukian
Final Data QC: Ali Latrous
2D Inversions: Ali Latrous
Reporting/Mapping: Wendy Acorn
Data acquisition phase was carried out under the supervision of Andrei Bagrianski, P. Geo,
Chief Operating Officer. Processing and 2D Inversions phases were carried out under the
supervision of Jean Legault, P. Geo, P. Eng, Chief Geophysicist (Interpretation). The overall
contract management and customer relations were by Paolo Berardelli.
4. DATA PROCESSING AND PRESENTATION
Data compilation and processing were carried out by the application of Geosoft OASIS
Montaj and programs proprietary to Geotech Ltd.
4.1 Flight Path
The flight path, recorded by the acquisition program as WGS 84 latitude/longitude, was
converted into the WGS 84, UTM Zone 16 North coordinate system in Oasis Montaj.
The flight path was drawn using linear interpolation between x, y positions from the
navigation system. Positions are updated every second and expressed as UTM easting’s (x)
and UTM northing’s (y).
4.2 In-field Processing and Quality Control
In-Field data processing and quality control are done on a flight by flight basis by a qualified
data processor (see Section 3.0). Processing steps and check up procedures are designed to
assure the best possible final quality of ZTEM survey data. A general overview of those steps
is presented in the following paragraphs.
The In-Field quality control can be separated into several phases:

a. GPS Processing Phase: GPS Data are first examined and evaluated during the GrafMov
processing.
b. Raw data, ZTEM viewer phase:
Data can be viewed, examined for consistency, individual channel spectra examined
and overall noise estimated in the viewer provided by the ZTEM proprietary software,
on the raw flight data and raw base station data separately, on the merged data, and
finally on the data that have undergone ZTEM processing.
c. Field Geosoft phase:
Magnetic data, Radar altimeter data, GPS positioning data are re-examined and
processed in this phase. Prior to splitting the lines EM data are examined flight by
flight and the effectiveness of applying the attitude correction evaluated. After
splitting the lines, a set of grids are generate for each parameter and their consistency
evaluated. Data profiles are also re-evaluated on a line to line basis. A power line
monitor channel is available in order to identify power line noise.
4.3 GPS Processing
Three GPS sensor (mounted on the airborne receiving loop) measurements were
differentially corrected using the Waypoint GrafMovTM software in order to yield attitude
corrections to recorded EM data.
4.4 ZTEM Electromagnetic Data
The ZTEM data were processed using proprietary software. Processing steps consist of the
following preliminary and final processing steps:
4.4.1 Preliminary Processing

a. Airborne EM, Mag, radar altimeter and GPS data are first merged with EM base station data
into one file.
b. Merged data are viewed and examined for consistency in an incorporated viewer
c. In the next, processing phase, the following entities are taken into account:
• the Base station coils orientation with respect to the Magnetic North,
• the Local declination of the magnetic field,
• Suggested direction of the X coordinate (North or line direction),
• Sensitivity coefficient that compensates for the difference in geometry between the base
station and airborne coils.
• Rejection filters for the 60 Hz and helicopter generated frequencies.
d. Six frequencies (30, 45, 90, 180, 360, and 720 Hz) are extracted from the airborne EM time-
series coil response using windows of 0.4 seconds and the base station coils using windows
of 1.0 seconds.
e. The real (In-Phase) and imaginary (Quadrature) parts of the tipper transfer functions are
derived from the In-line (X or Tzx) and Cross-line (Y or Tzy) components.
f. Such processed EM data are then merged with the GPS data, magnetic base station data and
exported into a Geosoft xyz file.
4.4.2 Geosoft Processing
Next stage of the preliminary data processing is done in a Geosoft TM environment, using the
following steps:
a. Import the output xyz file from the AFMAG processing, as well as the base Mag data into
one database.
b. Split lines according to the recorded line channel,
c. GPS processing, flight path recovery (correcting, filtering, calculating Bird GPS coordinates,
line splitting)
d. Radar altimeter processing, yielding the altitude values in metres.
e. Magnetic spike removal, filtering (applied to both airborne and base station data).
Calculation of a base station corrected mag.
f. Apply preliminary attitude corrections to EM data (In phase and Quadrature), filter and make
preliminary grids and profiles of all channels.
4.4.3 Final Processing
Final data processing and quality control were undertaken by Geotech Ltd headquarters in
Aurora, Ontario by qualified senior data processing personnel.
A quality control step consisted of re-examining all data in order to validate the preliminary
data processing and to allow for final adjustments to the data.
Attitude corrections were re-evaluated, and re-applied, on component by component, flight by
flight, and frequency by frequency bases. Any remaining line to line system noise was
removed by applying a mild additional levelling correction.
4.4.4 ZTEM Profile Sign Convention

Tzx and Tzy tipper components do not exhibit maxima or minima above conductors, resistors
or at contacts; in fact they produce cross-over type anomalies (Ward, 1959; Vozoff, 1972;
Labson, 1985). The crossover polarity sign convention for ZTEM is according to the right
hand Cartesian rule (Z positive –up) that is commonly used for multi-component transient
electromagnetic methods.
For the southwest to northeast lines of the Block the sign convention for the Tzx in-line
component crossover is positive-negative pointing southwest to northeast for tabular
conductors perpendicular to the profile (Figure 6). The corresponding Tzy component in-
phase cross-over polarity is positive-negative pointing southeast to northwest (90 degrees
counter clockwise to Tzx) according to the right hand Cartesian rule.
For the northwest to southeast tie-lines of the Block the sign convention for the Tzx in-line
component crossover is positive-negative pointing southeast to northwest for tabular
conductors perpendicular to the profile (Figure 6). The corresponding Tzy component in-
phase cross-over polarity is positive-negative pointing northeast to southwest (90 degrees
counter clockwise to Tzx) according to the right hand Cartesian rule.
Conversely, tabular resistive bodies produce In-Phase cross-over’s that are opposite in sign to
conductors. A brief discussion of ZTEM and AFMAG, along with selected forward model
responses is presented in Appendix D.
Figure 6 - ZTEM Crossover Polarity Convention for Tzx and Tzy for the block. Tie-lines on the left and Lines on
the right
4.4.5 ZTEM Quadrature Sign Dependence
One important note regarding the sign of the ZTEM Quadrature, relative to the In-Phase component,
particularly with regards to computer modeling and inversion.
The sign of the magnetotelluric Quadrature relative to the In-Phase tipper transfer function component
pertains to the Fourier transformation of the time series to give frequency domain spectra. There are
two widely used conventions for time dependence in the transformations, exp(+iωt) and exp(-iωt).
That which is implemented largely is a matter of personal preference and precedent. The importance of
the In-Phase and Quadrature sign convention is not critical, provided that it is known and documented.
In ZTEM, the data processing code used for the Fourier transformation the time-series data to frequency
domain spectra adopts a exp(-iωt) time dependence (J. Dodds, Geo Equipment Manufacturing, pers.
comm., Nov-2009). Whereas in the forward modeling and inversion program Zvert2d, the sign of the
Quadrature relative to the In-Phase transfer function assumes an exp(+iωt) dependence 2 .
As a result, for users interested in computer modeling and inversion of ZTEM data, the sign of the
Quadrature will need to be reversed, relative to the In-Phase component, in order to provide a proper
result (Figure 7). Indeed this reverse Quadrature polarity convention is assumed in all forward modeling
and inversion of ZTEM data, as described in Figures 5-7 in Appendix D.
ZTEM In-Phase Z/X (In-line) Tipper
ZTEM Quadrature Z/X (In-line) Tipper
Quadrature Z/X (In-line) Tipper (MT Convention)
Figure 7 - Illustration of ZTEM In-Phase & Quadrature Tipper transfer function polarity convention (e-iωt) relative to
equivalent MT Tipper Quadrature polarity convention (e+iωt) for a graphitic conductor in Athabasca Basin, SK.
2 Phillip E. Wannamaker (2009): Two-dimensional Inversion of ZTEM data: Synthetic Model Study and Test Profile Images,
Internal Geotech technical report by Emblem Exploration Services Inc., January 22, 2009, 32 pp.
4.4.6 Total Divergence and Phase Rotation Processing
In a final processing step DT (Total Divergence) and PR (Phase Rotation) processing are
applied to the multi-frequency In-phase and Quadrature ZTEM data. This is due to the
crossover nature of the Tipper Responses; these additional processing steps are applied to
convert them into local maxima for easier interpretation.
To present the data from both tipper components into one image, the Total Divergence
parameter, termed the DT is calculated from the horizontal derivatives of the Tzx and Tzy
tippers (Lo and Zang, 2008). It is analogous to the “Peaker” parameter in VLF (Pedersen,
1998).
Total Divergence DT: DT = DIV (Tzx, Tzy)

= d(Tzx )/dx+d(Tzy)/dy
This DT parameter was introduced by Petr Kuzmin (Milicevic, 2007, p. 13) and is derived for
each of the In Phase and Quadrature components at individual frequencies. These in turn
allow for minima over conductors and maxima over resistive zones. DT grids for each of the
extracted frequencies were generated accordingly, using a reverse colour scheme with warm
colours over conductors and cool colours over resistors.
The DT gives a clearer image of conductor’s location and shape but, as a derivative, it does
not preserve some of the long wavelength information and is also sensitive to noise.
As an alternative, a 90 degree Phase Rotation (PR) technique is also applied to the grids of
each individual component (Tzx and Tzy). It transforms bipolar (cross over) anomalies into
single pole anomalies with a maximum over conductors, while preserving long wavelength
information (Lo et al., 2009). The two orthogonal grids are then usually added to obtain a
Total Phase Rotated (TPR) grid for the In-Phase and Quadrature.
Total Phase-Rotation TPR: = PR (Tzx) + PR (Tzy)
A presentation of the ZTEM test survey results over unconformity uranium deposits that
illustrates DT and TPR examples, as documented by Lo et al. (2009) is provided in Appendix
E.
4.4.7 2D EM Inversion
2d inversions of the ZTEM results were performed over selected lines using the Geotech
Av2dtopo software developed by Phil Wannamaker, U. of Utah, for Geotech Ltd. The
inversion algorithm is based on the 2D inversion code with Jacobians of de Lugao and
Wannamaker (1996), the 2D forward code of Wannamaker et al (1987), and the Gauss-
Newton parameter step equations of Tarantola (1987). Av2dtopo has been
developed/modified for use with our ZTEM platform by taking into account the ground
topography and the air-layer between the receiver bird and the ground surface.
The 2D code only considers the In-Line (Tzx) data and assumes that the strike lengths of
bodies are infinite and orthogonal to the profile. The code is designed to account for the
ZTEM vertical coil receiver and fixed base station reference measurements. The inversion
uses a model-mesh consisting of 440 cells laterally and 112 cells vertically. Typically the
ZTEM data are de-sampled to 192 pts, in order to allow the inversion to run in 20 minutes or
less. Typically, between 1-2% errors are added to the In-line in-phase (XIP) and Quadrature
(XQD) data obtained at 30,45,90,180,360 & 720Hz. Errors are adjusted until numerical
convergence (<1.0 rms) is attained in 5 iterations or less. All inversions are based on an
apriori homogeneous starting half-space model, usually between 100 – 1000ohm metres, as
determined by the interpreter.
4.5 Magnetic Data
The processing of the total magnetic field intensity (TMI) data involved the correction for
diurnal variations by using the digitally recorded ground base station magnetic values. The
base station magnetometer data was edited and merged into the Geosoft GDB database on a
daily basis. The aeromagnetic data was corrected for diurnal variations by subtracting the
observed magnetic base station deviations.
The corrected magnetic data was interpolated between survey lines using a random point
gridding method to yield x-y grid values for a standard grid cell size of 50 metres. The
Minimum Curvature algorithm was used to interpolate values onto a rectangular regular
spaced grid.
5. DELIVERABLES
5.1 Survey Report
The survey report describes the data acquisition, processing, and final presentation of the
survey results. The survey report is provided in two paper copies and digitally in PDF
format.
5.2 Maps
Final maps were produced at scale of 1:20,000. The coordinate/projection system used was
WGS84, UTM Zone 16 North. All maps show the flight path trace and topographic data;
latitude and longitude are also noted on maps.
The preliminary and final results of the survey are presented as profile plans for the EM data
that were generated for individual real (In-Phase) and imaginary parts (Quadrature) of the
Tzx and Tzy components. Colour contour maps of the corresponding DT (Total Divergence)
or TPR (Total Phase Rotated) grids for three of the six frequencies, (30, 45, 90, 180, 360 and
720Hz), as well as for corresponding Phase Rotated Grids for individual components.
3D views have been constructed by plotting the either DT or TPR grids at their respective
penetration depths using a 1000 ohm-m half space, using the Bostick skin depth rule
(Bostick, 1977) see Appendix D.
Final maps were chosen, in consultation with the client, to represent all collected data, are
listed in Section 5.3.
Sample maps of the related 3D view, Magnetic and Total Phase Rotated are included in this
report and presented in Appendix C.
5.3 Digital Data
• Two copies of the data and maps on a DVD were prepared to accompany the report.
Each DVD contains a digital file of the line data in GDB Geosoft Montaj.
• DVD structure.
There are two (2) main directories;
Data contains databases and grids, as described below.
Report contains a copy of the report and appendices in PDF format.
Databases in Geosoft GDB format, containing the channels listed in Table 4.
Table 4 - Geosoft GDB Data Format
Column Description
X UTM Easting WGS84 Zone 16N, (Centre of the ZTEM loop) (meters)
Y UTM Northing WGS84 Zone 16N , (Centre of the ZTEM loop) (meters)
Longitude Longitude – WGS84 (Centre of the ZTEM loop) (Decimal degree)
Latitude Latitude – WGS84 (Centre of the ZTEM loop) (Decimal degree)
Z Elevation- WGS84 (Centre of the ZTEM loop) (metres)
Radar Helicopter terrain clearance from radar altimeter (metres - AGL)
Radar_B: Calculated ZTEM Bird terrain clearance (metres)
DEM Digital Elevation Model (above mean sea level, meters)
Gtime UTC Time (seconds of the day)
basemag Magnetic base station data, nT
Mag1 Measured total magnetic field, nT
Mag2 Diurnally-corrected total magnetic field, nT
Mag3 Levelled total magnetic field, nT
xIp_030Hz Tzx In-Phase 30 Hz final corrected
xQd_030Hz Tzx Quadrature 30 Hz final corrected
yIp_030Hz Tzy In-Phase 30 Hz final corrected
yQd_030Hz Tzy Quadrature 30 Hz final corrected
PLM Power Line Monitor (60Hz)
• Grids in Geosoft GRD format, as follows:
MAG: Total Magnetic Intensity

DEM: Digital Elevation Model
PLM: Power Line Monitor
XIP_30Hz_PR: Tzx In-Phase Component Phase Rotated grid at 30 Hz
XQd_30Hz_PR: Tzx Quadrature component Phase Rotated grid at 30 Hz
YIP_30Hz_PR: Tzy In-Phase Component Phase Rotated grid at 30 Hz
YQd_30Hz_PR: Tzy Quadrature component Phase Rotated grid at 30 Hz
IP_30Hz_DT: Total Divergence grid from In-phase components at 30 Hz
QD_30Hz_DT: Total Divergence grid from Quadrature components at 30 Hz
QD_720Hz_DT: Total Divergence grid from Quadrature components at 720 H
IP_30Hz_TPR: Total Phase Rotated grid from In-phase components at 30 Hz
QD_30Hz_TPR: Total Phase Rotated grid from Quadrature components at 30 Hz
A Geosoft .GRD file has a .GI metadata file associated with it, containing grid
projection information. A grid cell size of 50 metres was used.
• Maps at 1:20,000 scale in Geosoft MAP format, as follows:
11182_20K_3D_IP_TPR: 3D view of In-Phase Total Phase Rotated versus

Skin Depth (30-720Hz)
11182_20K_3D_IP_DT: 3D view of In-Phase Total Divergence versus Skin
Depth (30-720Hz)
11182_20K_TMI: Total Magnetic Intensity (TMI)
11182_20K_30Hz_IP_DT: 30Hz In-Phase Total Divergence Grid
11182_20K_XIP_profiles_XIP_PR 3 : Tzx (In-line) In-Phase Profiles over 90Hz
Phase Rotated In-Phase Grid
11182_20K_XQD_profiles_XQD_PR3: Tzx (In-line) Quadrature Profiles over a
90Hz Phase Rotated Quadrature Grid.
11182_20K_YIP_profiles_YIP_PR3: Tzy(Cross-line) In-Phase Profiles over 90Hz
Phase Rotated In-Phase Grid
11182_20K_YQD_profiles_YQD_PR3: Tzy (Cross-line) Quadrature Profiles over
90Hz Phase Rotated Quadrature Grid.
• 2D Resistivity Inversion maps:

11182_L1160_res: resistivity inversion of line 1160
11182_L1270_res: resistivity inversion of line 1270
• Maps are also presented in PDF format.
• 1:50,000 topographic vectors were taken from the NRCAN Geogratis database at;
http://geogratis.gc.ca/geogratis/en/index.html.
• A Google Earth file “11182_Magma.kml” is included, showing the flight path of
each block. Free versions of Google Earth software from:
http://earth.google.com/download-earth.html
3
Please note that due to the partial distortion caused by power-lines a clipped database was used to produce to the
EM stacked profiles for these maps in order for the majority of data to be properly displayed. However, the full data
set (not clipped) is provided in the final database.
6. CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusions
A helicopter-borne ZTEM and aeromagnetic geophysical survey has been completed over
the Thunder Bay North Block located near Thunder Bay, Ontario.
The total area coverage is 115 km2. Total survey line coverage is 629 line kilometres. The
principal sensors included a Z-Axis Tipper electromagnetic (ZTEM) system and a caesium
magnetometer. Results have been presented as stacked profiles and contour colour images at
a scale of 1:20,000.
There is no summary interpretation included in this report, however 2D inversions over

selected lines have been provided in Appendix F.
6.2 Recommendations
Based on the geophysical results obtained, a number of interesting conductive structures

were identified across the property. The magnetic results also contain worthwhile
information in support of exploration targets of interest. We therefore recommend a more
detailed interpretation of the available geophysical data, including additional 2D or 3D
inversion in conjunction with the geology, prior to ground follow up and drill testing.
Respectfully submitted6,
____________________________ _________________________
Emilio Schein Jean Legault, P. Geo, P. Eng
Geotech Ltd. Geotech Ltd.
____________________________ ____________________________
Lily Manoukian Ali Latrous
Geotech Ltd. Geotech Ltd.
August 2011
6
Final data processing of the EM and magnetic data were carried out by Emilio Schein and Lily Manoukian.
2Dinversions were carried out by Ali Latrous from the office of Geotech Ltd. in Aurora, Ontario, under the
supervision of Jean Legault, P. Geo, P. Eng, Chief Geophysicist (Interpretation)
7. REFERENCES AND SELECTED BIBLIOGRAPHY
Anav, A., Cantarano, S., Cerruli-Irelli, P., and Pallotino, G.V.(1976). A correlation method for
measurement of variable magnetic fields: Inst. Elect. and Electron. Eng. Trans., Geosc. Elect.
GE14, 106-114.
Bostick, F.X. (1977). A Simple almost exact method of MT analysis, Proceedings of the University of
Utah Workshop on Electrical Methods in Geothermal Exploration, 175-188.
De Lugao, P.P.,and Wannamaker, P.E.(1996). Calculating the two-dimensional magnetotelluric Jacodian
in finite elements using reciprocity: Geophys. J. Int., 127, 806-810
Karous, M.R., and S. E. Hjelt (1983). Linear filtering of VLF dip-angle measurements: Geophysical
Prospecting, 31, 782-794.
Kuzmin, P., Lo, B. and Morrison, E. (2005). Final Report on Modeling, interpretation methods and field
trials of an existing prototype AFMAG system, Miscellaneous Data Release 167, Ontario
Geological Survey, 2005.
Labson, V. F., Becker A., Morrison, H. F., and Conti, U. (1985). Geophysical exploration with
audiofrequency natural magnetic fields. Geophysics, 50, 656-664.
Legault, J.M., Kumar, H., Milicevic, B., and Hulbert, L. (2009), ZTEM airborne tipper AFMAG test
survey over a magmatic copper-nickel target at Axis Lake in northern Saskatchewan, SEG
Expanded Abstracts, 28, 1272-1276
Legault, J.M., Kumar, H., Milicevic, B., and Wannamaker, P.,(2009), ZTEM tipper AFMAG and 2D
inversion results over an unconformity uranium target in northern Saskatchewan, SEG Expanded
Abstracts, 28, 1277-1281.
Lo, B., Legault, J.M., Kuzmin, P. and Combrick, M. (2009). ZTEM (Airborne AFMAG) tests over
unconformity uranium deposits, Extended abstract submitted to 20th ASEG International
Conference and Exhibition, Adelaide, AU, 4pp.
Lo, B., and Zang, M., (2008), Numerical modeling of Z-TEM (airborne AFMAG) responses to guide
exploration strategies, SEG Expanded Abstracts, 27, 1098-1101.
Milicevic, B. (2007). Report on a helicopter borne Z-axis, Tipper electromagnetic (ZTEM) and magnetic
survey at Safford, Giant Hills, Baldy Mountains and Sierrita South Areas, Arizona, USA.,
Geotech internal survey report (job A226), 33pp.
Pedersen, L.B., Qian, W., Dynesius, L. and Zhang, P. (1994). An airborne sensor VLF system. From
concept to realization. Geophysical Prospecting, 42, i.8, 863-883
Peterson, L.B. (1998). Tensor VLF measurements: first experiences, Exploration Geophysics, 29, 52-57.
Strangway, D. W., Swift Jr., C. M., and Holmer, R. C. (1973). The Application of Audio-Frequency
Magnetotellurics (AMT) to Mineral Exploration. Geophysics, 38, 1159-1175.
Tarantola, A.,(1987) Inverse problem theory: Elsevier, New York, 613 pp.
Vozoff, K.(1972). The magnetotelluric method in the exploration of sedimentary basins. Geophysics, 37,
98-141.
Vozoff, K. (1991). The magnetotelluric method. In: Electromagnetic Methods in Applied Geophysics -
Volume 2 Applications, edited by Nabighian, M.N., Society of Exploration Geophysicists, Tulsa.,
641-711.
Ward, S. H. (1959). AFMAG - Airborne and Ground. Geophysics, 24, 761-787.
Ward, S. H, O’Brien, D.P., Parry, J.R. and McKnight, B.K. (1968). AFMAG Interpretation. Geophysics,
33, 621-644.
Wannamaker, P.E., Stodt, J.A., and Rijo, L., (1987). A stable finite element solution for two-dimentional
magnetotelluric modeling: Geophy. J. Roy. Astr. Soc.,88, 277-296.
Zhang, P. and King, A. (1998). Using magnetotellurics for mineral exploration, Extended Abstracts from
1998 Meeting of Society of Exploration Geophysics
APPENDIX A
SURVEY BLOCK LOCATION MAP
Survey Overview Location Map
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited A-1
Mining Claims for block
11079 Report on Airborne Geophysical ZTEM and Magnetic Survey for the Silver Reef Resources Ltd. A-2
APPENDIX B
SURVEY BLOCK COORDINATES

(WGS84 Zone 16N)
X Y
346904.1 5401378.3
358840.9 5408950
360436.9 5406401.2
362485.5 5407663.7
365129.6 5403090.1
355207.8 5396858.5
352902.8 5400742.1
348867.3 5398069.9
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited B-1
APPENDIX C
GEOPHYSICAL MAPS 1
3D View of In-Phase, Total Divergence (DT) grids versus Skin Depth (30 Hz – 720 Hz)
1
Full size geophysical maps are also available in PDF format on the final DVD
8/12/2011
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited C-1
Tzx (In-line) In-Phase Profiles over 90Hz Rotated Tzx In-Phase Grid
11079 Report on Airborne Geophysical ZTEM and Magnetic Survey for the Silver Reef Resources Ltd.. C - 2
Tzy (Cross-line) In-Phase Profiles over 90Hz Rotated Tzy In-Phase Grid
Tzx (In-line) Quadrature Profiles over 90Hz Rotated Tzx Quadrature Grid
Tzy (Cross-line) Quadrature Profiles over 90Hz Rotated Tzy Quadrature Grid
30Hz Total Divergence In-Phase (DT) Grid
Total Magnetic Intensity (TMI)
APPENDIX D
ZTEM THEORETICAL CONSIDERATIONS
A brief section on the theory behind the AFMAG technique is provided for completeness
and a more comprehensive development of the theory can be found in standard texts. The
natural EM field is normally horizontally polarized. Subsurface lateral variations of
conductivity generate a vertical component, which is linearly related to the horizontal field.
Although the fields look like random signals, they may be treated as the sum of sinusoids. At
each frequency the field can be expressed as a complex number with magnitude and
argument equal to the amplitude and phase of the sinusoid. The relation between the field
components can then be expressed by a linear complex equation with two complex
coefficients at any one frequency. These coefficients are dependent upon the subsurface and
not upon the horizontal field present at any particular time and are appropriate parameters to
measure (Vozoff, 1972).
Hz(f) = Tx(f) Hx(f) + Ty(f) Hy(f), (1)
Where
Hx(f), Hy(f) and Hz(f) are x, y and z components of the field,
Tx(f) and Ty(f) are the “tipper” coefficients.
In the case of a horizontally homogeneous environment, Tx and Ty are equal to zero because
Hz =0. They show certain anomalies only by the presence of changes in subsurface
conductivity in the horizontal direction. The real parts of the coefficients correspond to
tangents of tilt angles measured with a controlled source. The complex tensor [Tx, Ty]
known as the “tipper” defines the vertical response to horizontal fields in the x and y
directions respectively.
Tx and Ty are two unknown coefficients in one equation, and we therefore must combine
two or more sets of measurements to solve them. To reduce effects of noise, multiple sets of
measurements can be made, and the coefficients, which minimize the squared error in
predicting the measured Z from X and Y, can be found. This leads to next formulas for
estimating the coefficients.
Tx = ([HzHx*] [HyHy*] – [HzHy*] [HyHx*]) / ([HxHx*] [HyHy*] – [HxHy*] [HyHx*]), (2)
and
Ty = ([HzHy*] [HxHx*] – [HzHx*] [HxHy*]) / ([HxHx*] [HyHy*] – [HxHy*] [HyHx*] . (3)
Where
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited D- 1
[HxHy*] (For example) denotes a sum of the product of Hx with the complex conjugate of Hy.
In practical processing algorithms, all numbers Hx, Hy and Hz can be obtained by applying
the same digital band-pass filters to three incoming parallel data signals. FFT algorithms are
also applicable. All sums like [HxHy*] can be calculated on the basis of a discrete time
interval in the range from 0.1 to 1 sec or on a sliding time base.
Using platform attitude data in the EM data processing can be done at different stages of the
signal processing. The most obvious idea is to transform parallel data from local coordinates
of the platform into absolute geographical coordinates before the main signal processing
procedure. Unfortunately, the proper algorithms of attitude data obtained, often require some
post-processing algorithms such as using post-calculated accelerations based on GPS data
etc. That is why it is preferable to treat x-y-z coordinates in formulas above in the local
coordinate system of the platform and to recalculate resulting local tilt angles into a
geographical or global coordinate system later, during the data post processing.
In weak field conditions where the level of the signal is comparable with input noise levels
in preamplifiers, the bias in the estimated values of Tx and Ty caused by noise in the
horizontal signals become substantial and can not be reduced by any averaging. This bias can
be removed by the use of separate reference signals containing noise uncorrelated with noise
in signals Hx and Hy. (Anav et al., 1976).
Tx = ([HzRx*] [HyRy*] – [HzRy*] [HyRx*]) / ([HxRx*] [HyRy*] – [HxRy*] [HyRx*]), (4)
and
Ty = ([HzRy*] [HxRx*] – [HzRx*] [HxRy*]) / ([HxRx*] [HyRy*] – [HxRy*] [HyRx*]). (5)
Where:
Rx is the reference field x component,
Ry is the reference field y component.
An additional two electromagnetic sensors, providing these reference signals can be placed
at some distance away from the main x, y and z sensors. Currently, though, no additional
remote-reference processing are applied to ZTEM data.
Numerical Modelling
In order to understand the airborne AFMAG responses to conductors for a variety of

geological environments, EMIGMATM modelling code from PetRos EiKon (Toronto, ON)
was obtained to conduct the formulated model studies.
Below are some of the modelling results from their study.
Modelling assumption:
The assumptions for the modelling are that:
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited D-2
3 components of the magnetic field are measured and they are processed according to:
Hz(f) = Tx (f) Hx (f) + Ty (f) Hy (f)
The vector (Tx,Ty) is usually referred to as the ‘tipper’ vector and is determined in the
frequency domain through processing. This is normally done by determining transfer
functions from an extended time series.
For the modelling exercise, the 3 components of the magnetic vector (Hx,Hy,and Hz) are
modelled twice for 2 orthogonal polarizations of a plane wave source field and then the
tipper is calculated from a matrix calculation using the results of the 2 source polarizations’
models. For the 2D forward modelling results, the tipper vectors are shown as a function of
frequency
Basic Model Response
For the initial models, we assume a thin plate-like model. The model is perpendicular to the
flight direction. Initially, we will assume very long strike directions. From this quasi-2D
model, there are 2 basic responses. The so-called TE response and the so-called TM
response.
For the initial models, we will assume the strike is in the y (North) directions and the flight is
in the x (East) direction Sensor heights are 30m above ground.
TE Mode: For the TE response, the electric field excitation flows along strike (current
channelling) and the horizontal H field (Hx) flows perpendicular to strike thus causing
induction through Faraday’s law. The Hz response is generated both from channelling and
induction.
TM Mode: For this response, the electric field excitation flows perpendicular to strike
generating quasi-static charges on faces and the horizontal H field (Hx) flows parallel to
strike. Since, the XZ face is very small for this model, little current is induced. The charges
on the faces have a small dipole moment due to the thinness of the model.
For the rest of the models unless otherwise noted, the parameters used are:
Strike Length: 1km

Depth Extent: 1km
Conductance: 100S
Depth to Top: 10m
Background: Thin-overburden (10m), Resistive Basement (1000 Ohm-m)
Figure D-1 – Calculated Tipper components at 10 Hz for above model parameters.
Figure D1 shows the Tipper (Tx,Ty) Amplitudes at 10Hz using a10Ωm overburden. Note
small Ty (ie quasi-TM response)
Amplitude Response
Figure D-2 – Calculated Tx component of the Tipper at various frequencies
The (Tx) response amplitude at 1,10,100,1000,10000 Hx. Peak amplitude at 100Hz
Inphase and Quadrature Response
Figure D-3 – Calculated In-phase and Quadrature of the Tx component at various frequencies
Figure D-3 shows the In-phase and Quadrature response at 10 and 100Hz. Note the
crossovers in the In-phase and Quadrature, and the phase reversal in the Quadrature
responses from low to high frequencies.
Bo Lo, P.Eng, B.Sc. (Geophysics), Consultant

Geotech Ltd.
September, 2007
AFMAG Source Fields and ZTEM method 1
AFMAG uses naturally occurring audio frequency magnetic fields as the source of the
primary field signal, and therefore requires no transmitter (Ward, 1959). The primary fields
resemble those from VLF except that they are lower frequency (tens & hundreds of Hz
versus tens of kHz) and are usually not as strongly directionally polarized (Labson et al.,
1985). These EM fields used in AFMAG are derived from world wide
atmospheric thunderstorm activity, have the unique characteristic of being uniform, planar
and horizontal, and also propagate vertically into the earth – to great depth, up to several km,
as determined by the magnetotelluric (MT) skin depth (Vozoff, 1972), which is directly
proportional to the ratio of the bedrock resistivity to the frequency (Figure D4).
AFMAG Depth Penetration

Simplest Case: 1D Skin Depth Rule
Earth Surface
360 Hz
30 Hz 80m
290m 270 m
ZTEM AFMAG Minimum

1000 m 840 m
920m Penetration Depth
DEPTH (METRES)
2000 m
δs ~ 503√(ρ/f) [metres]* 360 Hz

2700 m
3000 m 2900 m
ZTEM AFMAG
Maximum Penetration 9200 m
Depth 30 Hz
4000 m
10 Ohm*m 100 Ohm*m 1000 Ohm*m 10,000 Ohm*m
EARTH RESISTIVITY (OHM-METRES)
MT PLANE WAVE SKIN DEPTHS in 1D HALF-SPACE

*Vozoff (1972)
Figure D4: MT Skin Depth Penetrations for ZTEM in 30-360Hz and 10-1000 ohm resistivity
At the frequencies used for ZTEM, the penetration depths likely range between approx.
600m to 2km in this region (approx. 1k ohm-m avg. resistivity assumed), according to the
following equation for the Bostick skin depth δB = 356 * √(ρ / ƒ) metres (Bostick, 1977),
which is considered appropriate as a rule of thumb equivalent depth estimate.
The other unique aspect of AFMAG fields is that they react to relative contrasts in the
resistivity, and therefore do not depend on the absolute conductance, as measured using
inductive EM systems, such as VTEM. Hence poorly, conductive targets, such as alteration
zones and fault zones can be mapped, as well as higher conductance features, like graphitic
units. Conversely, resistive targets can also be detected using AFMAG– provided they are of
a sufficient size and contrast to produce a vertical field anomaly. Indeed resistors produce
reversed anomalies relative to conductive features. Hence AFMAG can be effective as an
1
From: Legault, J.M., Kumar, H., and Milicevic, B. (2009): ZTEM tipper AFMAG and 2D inversion results over an unconformity
uranium target in northern Saskatchewan, Expanded Abstract submitted to Society of Exploration Geophysics SEG conference,
Houston, Tx, Nov-2009, 5 pp.
all-round resistivity mapping tool, making it unique among airborne EM methods. A series
of 2D synthetic models that illustrate these aspects have been created using the 2D forward
MT modelling code of Wannamaker et al. (1987) and are presented in figures D5-D7.
The tipper from a single site contains information on the dimensionality of the subsurface
(Pedersen, 1998), for example, in a horizontally stratified or 1D earth, T=0 and as such HZ is
absent. For a 2D earth with the y-axis along strike, TY=0 and HZ = TX*HX. In 3D earths, both
TX and TY will be non-zero. HZ is therefore only present, as a secondary field, due to a lateral
resistivity contrast, whereas the horizontal HX and HY fields are a mixture of secondary and
primary fields (Stodt et al., 1981). But, as an approximation, as in the telluric-
magnetotelluric method (T-MT; Hermance and Thayer, 1975) used by distributed MT
acquisition systems, the horizontal fields are assumed to be practically uniform, which is
particularly useful for rapid reconnaissance mapping purposes. By measuring the vertical
magnetic field HX, using a mobile receiver and the orthogonal horizontal HX and HY fields at
a fixed base station reference site, ZTEM is a direct adaptation of this technique for airborne
AFMAG surveying.
_________________
Jean M. Legault, M.Sc.A., P.Eng., P.Geo.
Geotech Ltd.
References
Bostickm, F.X., 1977, A simple almost exact method of MT analysis. Proceedings of the
University of Utah Workshop on Electrical methods in Geothermal Exploration,
175-188.
Hermance, J.F., and Thayer, R.E., 1975, The telluric-magnetotelluric method,
Geophysics, 37, 349-364.
Labson, V. F., A. Becker, H. F. Morrison, and U. Conti, 1985, Geophysical exploration
with audio-frequency natural magnetic fields: Geophysics, 50, 656–664.
Murakami, Y., 1985, Short Note: Two representations of the magnetotelluric sounding
survey, Geophysics, 50, 161-164.
Pedersen, L.B., 1998, Tensor VLF measurements: Our first experiences, Exploration
Geophysics, 29, 52-57.
Stodt. J.A., Hohmann, G.W., and Ting, S.C., 1981, The telluric-magnetotelluric method
in two- and three-dimensional environments, Geophysics, 46, 1137-1147.
Vozoff, K., 1972, The magnetotelluric method in the exploration of sedimentary basins,
Geophysics, 37, 98–141.
Ward, S. H., 1959, AFMAG—Airborne and ground: Geophysics, 24, 761–787.
Wannamaker, P.E., Stodt, J.A., and Rijo, L., 1987, A stable finite element solution for
two-dimensional magnetotelluric modelling, Geophy. J. Roy. Astr. Soc., 88, 227-
296.
2D Resistivity Model – Conductive Buried Prism
2-D Model conductor_ew
001002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
| 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | | |
0
Ohm-M
500m
1
1.8
1 ohm-m 400m 50m Air Layer
Depth (METERS)
3.2
5.6
1000
10
200m 18
32
50
100
180
2000
3200 ohm-m 320
560
1000
1800
3200
5600
3000
W E
METERS
0 500 1000 1500 2000 2500 3000
E-W
Transfer Function Profile conductor_ew - 32, 48, 100, 180, 380 Hz

002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
001 | 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
Hz/Hx In-Phase (%)

| | | | | | | | | | | | | | | | | | | | | | | | |
+100%1
0.75 2D Forward Model of ZTEM Tippers
0.5
1
1 1 2 1 1 1
1 1 1 2 2 2 2 2 1
2 2
1 1 2 2 2 1
0.25 1 1 2 1 1 1 2 2 3 3 3 3 3 3 2
1
1 21 2 2 2 3 3 3
2 2
3 3 3 4 4 4 4 4 4 3
4 1
2
3
3 3 3 3
4 4 4 4 4 2
5 5 5 5 5 5 5 1
3
3 3 3 3 4 4 4
TZY
4 5 5 5 5
0%0 145 45 45 54 45 45 45 5 5 5 5 5 3
2
1
4
5
5
4
3 5 5 5
5 5 5
5 5 5 5 5 4 5 4
5 4
4 4 4 3
5
2 5
1 5
4 5 5 5 5 5 5 5
3 4
2
5 5
4 4 4 4 4 4 4 4
3 3 3 3 3
3 3
1 4 4 3 3 3 2
3
1 4
2 3 4 4 4 4 4
4 3 3 3 2 2 2 2 2 2 1
-0.25 2 3 3 3 3 3 3 3 3 2 2
2 2 1 1 1 1 1
1 2 2 1 1 1
1 2 2 2 2 2 1 1
1=32Hz 1 1 2 2
1 1
2
1
2
1 1 1 1 1
Z
-0.5
2=48Hz
-0.75
3=100Hz Note: Positive to Negative
4=180Hz
-100%-1
W
5=380Hz
In-Phase Cross-Overs
X E
METERS
0 500 1000 1500 2000 2500
Tyr
Hz/Hx Quadrature (%)

002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
001 | 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | | |
+100%1
1=32Hz
0.75 2=48Hz Note: Negative to Positive
3=100Hz Quadrature Cross-Overs
0.5
4=180Hz
0.25 5=380Hz 2 2 2
1 1 1
2 2
1 1 1 1
2 2 1 1
3 3 3 2
1 1 3
3 3 2 2 2 1 1 1 1
5
4 5
4
3 4
3
5
2 1
3
4
5
2 1 5 2
4
2
2 1 1
4
5 4
5 4
5 4 4 3
4 3
5 5 5 4 3
4 3
4 3
4 3 3 2
4 4 3 2 2
4 3 3
2 1
3 2
3
5
4 3
2 1 5 5 5 5 5 5 5 4 4 4
1 2 1
3
5 1 5 5 5 4
TZY
5 4
3
2
2 1
4
3 5
0%0 5 5 5 5 5 5 5 1
2 2
3 2
1
4 3
1 4
2 1
5
4
3
5 1
3
2
1
5 4
4 4 4 4 4 4 4 5 5 5 5 5 5 5 2 5 5
5 3
1 4
3
2 2 4 4 4 4 4 4 4 4
3 3 3 3 3 3 3 5 1
5
4
3
2 1
5
2
4
3 2
1
4
3 5 3
5 1
2
4
3 4
2 2
1 1 1 2 2
1 1 2 2
1 1 3 2
2
1 1 3 3 2
1 1
2 3 1 1
3 2
1 1
2 3 3
2
-0.25
-0.5
(Note: Quadrature Polarity Reversed Relative to ZTEM due to
-0.75
Difference in Sign Conventions with 2D Forward Modeling Code)
-100%-1
W E
METERS Tyi
0 500 1000 1500 2000 2500
Figure D5: 2D synthetic forward model Tipper responses (Tzy) for conductive brick model.
2D Resistivity Model – Poorly Conductive Buried Prism

001002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
| 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | | |
0
Ohm-M
500m
1
1.8
Depth (METERS)
3.2
5.6
1000
10
18
200m 32
50
100
180
2000
3200 ohm-m 320
560
1000
1800
3200
5600
3000
W E
METERS
0 500 1000 1500 2000 2500 3000
E-W

002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
001 | 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
Hz/Hx In-Phase (%)
| | | | | | | | | | | | | | | | | | | | | | | | |
+25%0.25
2D Forward Model of ZTEM Tippers
4 4 4 4 4 4 4
3 3
4 4
3 3 3 3
5 5
3 3
5 5 5 4
3 3 4 3
3
4 2
3 3
4 4 5 5
2
2 1 2 2 2 2 2 2 5
3 4
5
4 3
3
3 2 4 2
4 2 2 5 5 5
2 2 1 1 1 1 1 2 3
1 1
3
3 4
2 3 4
4
1 2 1 1 2
1 2 1 1 1 5 1 1 1
1 1 2
1 4
5
3
1412 5 5 5
5 5 2
1 4
5 5 5 5
TZY
5
3
2
1
0
0% 2
1
5
4
3
1
2 5 5 5
5 5 5 5 5
5
3
5 1
4 5 5 1 1 1 1 1 1
4 2 2 2 1
4 1
4 4 4
2 1
2
3
2 1 1 1 5 5
1 1 1 1
2 2 4 2
4 2
2 4
3 3 3 3
5 2 2 2
3
4 1 1
1 1 1 1 2 1
2 2 2 2 2 2
5 5 4 3
4 3
4 3
3 3
3
5 3 2 2 2 5 5 3 3
4 4
3 3
4 5 3
5 5 5 5
3 3 3 3 3 3
4 4
4 4 4
1=32Hz 4 4 4 4
2=48Hz
3=100Hz Note: Positive to Negative Z
4=180Hz In-Phase Cross-Overs
-25%-0.25
W
5=380Hz X E
METERS
0 500 1000 1500 2000 2500
Tyr
Transfer Function Profile conductor_ew - 32, 48, 100, 180 Hz

002003
001 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
|| | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | |
+25%0.25
1=32Hz
2=48Hz Note: Positive to Negative
4=180Hz
1 5=380Hz
2 2
1 1 2 2
1 1 1 2
1 2 1 2
1 2 1 2 2
1 1
1 2
1 2 2 2 3 3 3 3 3 1 1 1 4 4 4 4 4 4 4
4 4 4 4 4 4 4 4 4 4 4
TZY
1 2
1 2 1 2
1 2 3 3
1 1 2
2 2 1 2 3 3 3 3 3 3 2
3 1
2
3 4 4 4 4 4 4 4 3 3 3 3 3 3 2
3 3
0% 0 3 3 3 3 3 3 3 3 3
4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
4 4 4 4 4
4 4 4 4 1
4 3
3
2
2 1 2
2 1 1 2 2
3 3 3 3 3 2
3 3 3 3 3 3 3 3 3 3 3 2
2 1
1
1 1 2
2 2
1 1 2 2
1 1 1 1 2 2
2 1 1 2 2
1 1 1 1 2
2 1 1
2 2
1 2
1 2
1 1

Difference in Sign Conventions with 2D Forward Modeling Code)
-25%-0.25
W E
METERS Tyi
0 500 1000 1500 2000 2500
Figure D6: 2D synthetic forward model Tipper response (Tzx) for poorly conductive brick model.
2D Resistivity Model – Resistive Buried Prism
001002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
| 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | | |
0
Ohm-M
500m
1
1.8
Depth (METERS)
3.2
5.6
1000
10
18
200m 32
50
100
180
2000
100 ohm-m 320
560
1000
1800
3200
5600
3000
W E
METERS
0 500 1000 1500 2000 2500 3000
E-W

002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
001 | 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
Hz/Hx In-Phase (%)

| | | | | | | | | | | | | | | | | | | | | | | | |
+5%0.05 2D Forward Model of ZTEM Tippers

1
0.025
TZY
1 1 1
2 1 1
2 2 1 1
2 2 2 1
2 1
2 1 1 2 1
0% 0 3
2 3
2
5 1
4
1 3
2
5 1
4 3
2
5 1
4 3
2
5 1
4 4 4
3
2
5 1 4
3
5 1 4
3
5 1
2 3
4
5 1
2 2 3
4
5 1
2 3
4
5 1 3
4
5 1
2 2 4
3 4
3
5 2
5 1
2 5 4
4
3
5 2 4
5 3
3 4
5 3 4
5 3
5 3 5 3
4 5 3
4 5 3
4 5 2
4 4
3
2 1 5 2
5 1
2
3
4 3
5 3
4 3 3
5 4
4 5 3 3
5 4
4 5 3 3
5
5 4
4 5 3
4 5
4
2
3
4 5 1
5 2
4
3 4
3 5 1
5 2
4
3 5 1
2
4
3 5 1
2
4
3 5 1
4
2
3 5 1
4
2
3 5 1
4
2
3 5 1
4
2
3 5 1
4
3
2 5 1
4
3
2 5 1
4
3
2 5 1
4
3
2 5 1
4
3
2 5
4
3
2
1 1 2 1
1 2 2
1 2
1 2 2 1
2 1
1 1
-0.025 1=32Hz
2=48Hz Note: Negative to Positive
-5%-0.05
3=100Hz In-Phase Cross-Overs Z
4=180Hz
W
5=380Hz X E
METERS
0 500 1000 1500 2000 2500
Tyr

002 004 006 008 010 012 014 016 018 020 022 024 026 028 030 032 034 036 038 040 042 044 046 048 050
001 | 003 | 005 | 007 | 009 | 011 | 013 | 015 | 017 | 019 | 021 | 023 | 025 | 027 | 029 | 031 | 033 | 035 | 037 | 039 | 041 | 043 | 045 | 047 | 049 |
| | | | | | | | | | | | | | | | | | | | | | | | |
1
1=32Hz
+5%0.05 2=48Hz
3=100Hz
Note: Positive to Negative
0.025
5=380Hz
TZY
2 1
1 2 1
2 1 2 1
2 2
3 1 1 3
3 2
3 4 3
2
1 2 3
4 3
2 2 2
3
0% 0 1
4
3 5 1
5 1
2 2
4
3 5 1
2
4
3 5 1
2
4
3 2
4
3 5 1
5 1
2
3
4 5 1
2
3
4 5 1
2
3
4 2
5 1
3
4 2
3
5 3
4 4 4 5 3
5 3
5 3 4 4 5 4
5 4 5 4
5 5
1
4
5 1
5 4
1
5 4
5 3
4
1 1
2
5 1
4
3
2
5 1
5 1
4
3
2 5 1
4
2
3 5 1
2
4
3
5 4
5 1
4
2
3 4 1
2
3
5 4
5 4
3 1
2
5 4
3 1
2
5 4
3 1
2
5 3
5 3
4
3 1
3 1
2 2
5 3
4 5 2
4
2 1
2 1
4
5 1
4
3
2 1
4
5 1
3
2 5 1
3
2 5 1
4
3
2 5 1
4
3
2 5 1
4
3
2 5 1
4
3
2 2 4
3
5 1
4
3 2 2 5 1
4
3
5 1
4
3
5 1 2 5 1
4
3
2 5
4
3
2
-0.025

-5%-0.05 Difference in Sign Conventions with 2D Forward Modeling Code)
W E
METERS Tyi
0 500 1000 1500 2000 2500
Figure D7: 2D synthetic forward model Tipper response (Tzx) for resistive brick model.
APPENDIX E
ZTEM (AIRBORNE AFMAG) TESTS OVER UNCONFORMITY URANIUM DEPOSITS 6

1 2 31
Bob Lo *, Jean Legault ,Petr Kuzmin Formerly Geo Equipment Manufacturing Ltd., now Exploration Syndicate, Inc.,
2 3
bob.lo@expsyn.com, Geotech Ltd., jean@geotech.ca, Geo Equipment Manufacturing Ltd., petr@geotech.ca
Key Words: ZTEM, AFMAG, electromagnetic, airborne, uranium, Athabasca.
INTRODUCTION
A series of demonstration tests were conducted using the ZTEM, airborne AFMAG system over deep targets in the
Athabasca Basin of Saskatchewan, Canada. These tests were conducted in mid-2008 and were flown to test ZTEM’s
ability to detect large conductive targets at depth; deeper than conventional airborne EM methods. Data are presented
over areas where the conductors are located 450-600 metres beneath the surface. As well, a case of ZTEM following the
plunge of a conductor to over 800 metres depth is shown.
BACKGROUND
The ZTEM system is the latest implementation of an airborne AFMAG system first commercialized in late 2006. ZTEM
uses a large, 8 metre diameter airborne air core coil, slung from a helicopter, to measure the vertical component of the
AFMAG signal. Two 4 metre square coils are deployed on the ground to measure the horizontal field. The ZTEM system
has flown successful demonstration surveys over porphyry copper deposits in the southwest USA (Zang et al., 2008).
ZTEM was tested in the Athabasca Basin in Canada in May of 2008 to determine its depth of investigation and to
determine its suitability for mapping deep conductors in the crystalline basement. Over 30% of the world’s U3O8 is
mined in the Athabasca Basin from unconformity uranium deposits. Unconformity uranium deposits of the Athabasca
Basin are often associated with conductors located in the crystalline basement. The search for economic uranium deposits
is moving to areas of the basin which are deeper and beyond the detection limits of modern airborne instrumentation.
This creates the requirement for a system which can detect conductivity past the detection limits of modern traditional
EM systems. This was the motivation behind the field trials of the ZTEM system in the Athabasca Basin. Several areas
where known deep conductors (450-600m+) were located were flown. Also, a test survey block in the northern part of the
basin was able to trace a deep and plunging conductor to depths that no other airborne EM system has been able to
achieve.
ATHABASCA BASIN GEOLOGY
The high-grade uranium deposits within the Athabasca Basin are associated with the unconformity between the
essentially flat-lying Proterozoic Athabasca Group sandstones and the underlying Archean-Paleoproterozoic
metamorphic and igneous basement rocks. The deposits occupy a range of positions from wholly basement-hosted to
wholly sediment-hosted, at structurally favourable sites in the interface between the deeply weathered basement and
overlying sediments of the Athabasca Basin (Ruzicka, 1997). The locations of These deposits are lithologically and
structurally controlled by the sub-Athabasca unconformity and basement faults and fracture zones, which are localized in
graphitic pelitic gneisses that may flank structurally competent Archean granitoid domes (Quirt, 1989).
In general, most of the known important deposits tend to occur within a few tens to a few hundred metres of the
unconformity and within 500 m of the current ground surface. This may be more of a limitation of exploration
techniques. There is no reason to believe that the distribution of the deposits is dependent on the modern day depth of
6 TH
Extended abstract submitted to 20 ASEG International Geophysical Conference & Exhibition, Adelaide, AU, 22-26 Feb, 2009.
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited E- 1
burial.
Empirically, the geophysical exploration for unconformity type uranium targets have been to search for large basement
structures which post date the sandstone deposition of the basement (Matthews et. al, 1997). All the deposits located so
far are associated with fault structures associated with a graphitic conductive basement. An alteration zone of clay
silicification and enrichment around the deposits probably leads to magnetite destruction causing the magnetic low
observed around the deposits. The clay alteration should give rise to a resistivity low signature about the deposits. The
low conductivity of the clay alteration makes it a difficult target for airborne EM if it is buried at significant depth.
ZTEM INSTRUMENTATION AND PRESENTATION
ZTEM is an airborne AFMAG system introduced by Geotech Ltd. of Canada in early 2007 (Lo et al., 2008). In a ZTEM
survey, a single vertical dipole air-core coil is flown over the survey area in a grid pattern similar to other airborne
electromagnetic surveys. Two orthogonal, air-core, horizontal axis coils placed close to the survey site measures the
horizontal EM fields for reference. A GPS array on the airborne coil monitors its attitude for post-flight corrections.
Figure 1 – Stacked profiles of the x-component Tipper over the gridded values of the phase rotated x-component
data. Note that the cross-overs in the profiles are now peaks on the image.
As the source field is assumed to be far away, the excitation of the ground is more or less uniform. For large structures,
the signal fall-off will be much slower than from a dipole source, such as those energized by traditional airborne systems.
With the ZTEM system being less susceptible to terrain clearance, the planned ground clearance height is higher and the
terrain drape is looser as compared to standard helicopter EM surveys.
The two Tippers obtained from the relationship between the vertical airborne coil and the two ground coils have a cross-
over over a steeply dipping, plate-like body. The cross-overs can be made into local maxima via a 90 degree phase
rotation which allows for easier interpretation of the gridded values. Figure 1 is an example of this transformation.
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited E-2
To present the data of both Tippers as one image, we calculate a parameter termed the DT which is the horizontal
divergence of the two Tippers, much in the same manner as the “peaker” parameter in VLF (Pedersen, 1998). The DT is
typically plotted with an inverted colour bar as it is negative over a steeply dipping thin body.
ZTEM RESULTS – NORTHERN ATHABASCA BASIN
Figure 2 shows gridded values from a number of ZTEM lines over an area where the sedimentary cover is approximately
450-600 metres thick. A number of traditional EM systems have also been flown over this block. While they were able to
detect conductors, the resolution of the conductive features is not nearly as detailed as the information provided by
ZTEM.
Figure 2 – ZTEM results over an area of 450-600 metre thick sedimentary cover.
Figure 3, from another area, shows the data from one of the larger blocks that was flown. It is a 3D composite image
of the DT at various frequencies plotted at the equivalent skin depth assuming a 1,000 ohm-m average resistivity.
Figure 3 - Perspective view of DT’s of different frequencies plotted at the skin depth (using a 1,000 ohm-m Earth.
The data in Figure 3 come from a survey over the north rim of the Athabasca Basin. The sandstone cover is about 500m
on the left hand side of the image, and progressively getting deeper to the right. It is about 700m in the middle part of the
image and over 800 metres thick on the right middle portion where exploration drilling is concentrated. Starting in the
middle left and trending to the right of the image, there is a known graphitic shear.
In the uppermost (600m) “depth slice”, Figure 3 shows a linear conductive feature that progressively weakens as one
moves to the right until it is no longer seen. This is interpreted to be due to the graphitic shear conductor plunging deeper
past the depth of investigation of the 360 Hz data. The lower frequencies penetrate more into the sedimentary cover that
is deeper towards the right. DT’s of decreasing frequency show the linear conductive feature extending more and more to
the right. The feature also strengthens/sharpens into a synformal shape with lower frequencies. This fits with what the
known geology of a plunging conductor at depth is doing.
At the nose of the fold, in the right third of the images, we also see another, broader anomalous zone that trends towards
the back of the image. At this location, two radioactive springs are situated. These spring waters which are anomalously
high in uranium and radon may reflect the upward migration of deep waters along faults, suggesting structural targets in
areas where basinal waters may have tapped a radioactive source. This broad DT trend might be the plunge of the fold
axis that is aligned away from the front of the image. An anomaly along this trend, at the highest frequency, that steadily
grows with each decreasing frequency can be seen. This might represent an alteration zone in the sandstone that is
detected at the shallowest depth. By about the 90Hz DT depth slice or so, we are possibly in the deeper basement and
into a basement graphitic unit.
CONCLUSIONS
A number of successful ZTEM tests were conducted over the Athabasca Basin. The tests demonstrated that ZTEM can
easily detect conductivity to 800 metres beneath relatively resistive sedimentary cover. Assuming a 1,000 ohm-metre
resistivity, the skin depth of the 30 Hz data is approximately 2,000 metres. The 30 Hz data presented have good signal to
noise ratios indicating a deep depth of exploration. The observation that ZTEM may be detecting the clay alteration
above the crystalline basement is a significant advantage for exploration of unconformity uranium deposits.
More demonstration surveys are planned in the Athabasca Basin later this year. And more target types for testing are also
planned.
ACKNOWLEDGEMENTS
The authors thank Geotech Ltd. for allowing them to publish this work and for providing the support required to write
this abstract and to present this paper.
REFERENCES
Labson, V. F., Becker A., Morrison, H. F., and Conti, U., 1985, Geophysical exploration with audiofrequency
natural magnetic fields, Geophysics, Vol. 50, p. 656-664.
Lo, B., Zang, M., Kuzmin, P., 2008, Geotech’s Z-TEM (Airborne AFMAG) Instrumentation, a paper presented at
KEGS PDAC 2008 Symposium, Toronto.
Matthews, R., Koch, R. and Leppin, M., 1997, Advances in Integrated Exploration for Unconformity Uranium
Deposits in Western Canada; in Proceeding of Exploration 97, edited by Arnis Gubins, Prospectors and
Developers Association of Canada, Toronto.
McMullan, S.R., Matthews, R.B, and Robertshaw, P., 1990, Exploration geophysics for Athabasca Uranium
Deposits, in: Proceedings of Exploration 87, Ontario Geological Survey.
Pedersen, L.B, Qian, W., Dynesius, L., Zhang, P., 1994, An airborne tensor VLF system. From concept to
realization, Geophysical Prospecting, Vol. 42.
Ruzicka, V.R., 1997, Metallogenic features of the uranium-polymetallic mineralization of the Athabasca Basin,
Alberta, and a comparison with other parts of the basin; in R.W. Macqueen, ed., Geological Survey of
Canada, Bulletin 500, 31-79.
Wheatley, K., Murphy, J., Leppin, M., and Climie, J.A., 1996, Advances in the Genetic Model and Exploration
Techniques for Unconformity-type Uranium Deposits in the Athabasca Basin; in Ashton, K.E., Harper,
C.T., eds., MinExpo ’96 Symposium – Advances in Saskatchewan Geology and Mineral Exploration:
Saskatchewan Geological Society, Special Publication No 14, p. 126-136.
Quirt, D., 1989, Host rock alteration at Eagle Point South: Sask. Research Council, Publication no. R-855-1-E- 89,
95p.
Ward, S. H., 1959, AFMAG - Airborne and Ground: Geophysics, Vol. 24, p. 761-787.
Zang, M., Lo, B., 2008, The Application of Airborne Natural Field Electromagnetics (ZTEM):
Some Examples from the Southwestern United States, a paper presented at the 2008
PDAC, Toronto
APPENDIX F
2D ZTEM Inversion Results

for
Magma Metals Ltd.
Thunder Bay
Ontario, Canada
Job: 11182
By
Geotech Ltd.
245 Industrial Parkway North

Aurora, Ont., CANADA, L4G 4C4
Tel: 1.905.841.5004
Fax: 1.905.841.0611
www.geotech.ca
Email: info@geotech.ca
August, 2011
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited F- 1
Summary
• Geotech Av2Dtopo program was used to perform the
inversion of ZTEM data with topography and receiver
altitude effects considered. While we collected both in‐
line (Tzx) and cross‐line (Tzy) tippers, inversion makes
use only of Tzx (In‐phase and Quadrature).
• Total of 2 lines of data were inverted with starting model
resistivity 3000 Ω.m and 6 frequency data were used for
all lines
• Slide 3 show the location of the survey lines over IP 90Hz
DT (left) and TMI (right) grids.
• Following slides show the inversion results, the inversion
parameters and the location of each line over IP 90Hz DT
grid.
11182 Report on Airborne Geophysical ZTEM and Magnetic Survey for Magma Metals Limited F-2
Thunder Bay
Line Locations over IP 90 Hz Line Locations over TMI Image

DT Image
* the south-west end of the line L1270 has been cut off to avoid power Line effect during inversion.
REPORT ON PROCESSING & ANALYSIS OF
ZTEM EM AND MAGNETICS DATA
THUNDER BAY NORTH PROPERTY
THUNDER BAY ONTARIO
FOR
MAGMA METALS LIMITED
JANUARY 2012
Condor Consulting
Lakewood Colorado
USA
REPORT ON PROCESSING & ANALYSIS OF
ZTEM EM AND MAGNETICS DATA
THUNDER BAY NORTH PROPERTY
THUNDER BAY ONTARIO
FOR
MAGMA METALS LIMITED
JANUARY 2012
Condor Consulting
Lakewood Colorado
USA
Assessment of TBN (Ontario) ZTEM Survey Magma Metals Limited
CONTENTS
1.SUMMARY ................................................................................................................................. 1
2.INTRODUCTION ....................................................................................................................... 2
ZTEM SURVEY ................................................................................................................. 2
DATA ASSESSMENT ....................................................................................................... 3
3.GEOLOGY AND BACKGROUND DATA ................................................................................... 4

GEOLOGICAL BACKGROUND ........................................................................................ 4
Regional Geology .................................................................................................. 4
Local Geology ....................................................................................................... 5
TARGET MODEL .............................................................................................................. 6
Nickel-PGE Targets ............................................................................................... 6
4.PROCESSING, ANALYSIS AND PRODUCTS .......................................................................... 8

PROCESSING .................................................................................................................. 8
Magnetic Data Processing .................................................................................... 8
ZTEM Data ............................................................................................................ 8
ZTEM 1D and Grid-Based Results ........................................................................ 9
2D ZTEM Inversion.............................................................................................. 10
ANALYSIS ....................................................................................................................... 10
2D Results ........................................................................................................... 11
3D Results ........................................................................................................... 12
PRODUCTS .................................................................................................................... 13
Table 4-1 Survey Products .................................................................................. 13
5.INTERPRETATION .................................................................................................................. 15
DATA QUALITY .............................................................................................................. 15
ZTEM Survey....................................................................................................... 15
PROPERTY SCALE ZTEM GRID BASED OUTCOMES ................................................ 16
Grid Based Outcomes: EM Results ..................................................................... 16
Grid Based Outcomes: EM and Magnetic Results .............................................. 20
Grid Based Outcomes: EM and Gravity Results ................................................. 21
Discussion of Grid-Based Outcomes .................................................................. 22
3D BASED ASSESSMENT ............................................................................................. 24
Comparison 2D-3DH-3DF Inversion of ZTEM....................................................... 27
Comparison of 3D Outcomes for Magnetic and Gravity Results ......................... 28
Comparison of 3DH EM with Magnetic and Gravity 3D Inversions ...................... 29
Deposit and 2D-3D Geophysical Models............................................................. 33
Comparison with IP-R Survey ............................................................................. 35
SUMMARY OF TARGET ZONES ................................................................................... 36
DISCUSSION .................................................................................................................. 36
FOLLOW-UP ................................................................................................................... 38
Table 5-1: Summary of Target Zones.................................................................. 44
Condor Consulting, Inc. i January 2012

6.CONCLUSIONS AND RECOMMENDATIONS ........................................................................ 45
7.REFERENCES ........................................................................................................................ 46
8.APPENDICES .......................................................................................................................... 47
APPENDIX A - INFORMATION ON PROCESSING EM AND MAGNETIC DATA .......... 48
Section 1: Magnetic Processing: ZS and Mag3D ................................................ 49
Section 2: ZTEM Processing ............................................................................... 50
Section 3: ZTEM Processing-Y component MultiPlots ........................................ 51
APPENDIX B – Comparison of Condor and Geotech 2D Processing............................. 54
APPENDIX C – ARCHIVE DVD ...................................................................................... 55
Condor Consulting, Inc. ii January 2012

1. SUMMARY
This report describes the assessment of a ZTEM airborne EM and magnetic survey carried out for
Magma Metals Limited (Magma) over their Thunder Bay North (TBN) property located approximately
50 km NE of Thunder Bay Ontario. Condor Consulting, Inc. (Condor) was commissioned to carry out
an assessment of the survey results and examine these outcomes in light of other available
geoscience data provided by Magma.
This assessment showed that the ZTEM survey has provided new information about both shallow and
deeper-seated features that relate to both specific targets for follow-up, as well as possible information
about aspects of the property geology not previously understood.
The ZTEM outcomes were however, not without inconsistencies in terms of the defined features within
the various grid-based and 2D/3D products. This is attributed to both the fundamental differences of
AFMAG as a geophysical technique when compared with more “precise” techniques such as
magnetics and active source EM, as well as the infancy in the modes of processing available for the
ZTEM data and end user experience with such outcomes. This said, the outcomes are deemed to be
of a quality and internal consistency such that unique and believable contributions have been made to
the understanding of the economic geology on the property.
Condor Consulting, Inc. 1 January 2012

2. INTRODUCTION
ZTEM SURVEY
The ZTEM EM and magnetic survey over TBN property was commissioned by Magma and carried
out by Geotech Ltd. (Geotech) between May 24th and June 5th, 2011. A total of 629 lkm of EM and
magnetic data were recorded. Full details on the survey can be found in Schein (2011). The area
covered is shown in Figure 1.
.
Figure 1: Location of ZTEM EM and magnetic survey TBN project, Thunder Bay Ontario.

DATA ASSESSMENT
Condor was commissioned to assess the outcomes of the ZTEM survey in light of the known
economic geology. This assessment involved the processing of the EM and magnetic data and a full
examination of the outcomes in 2D and 3D.
In addition to the ZTEM results, Condor was given access to additional geoscience data including a
high resolution magnetic survey flown by UTS in 2006 (no reference), a Falcon gravity survey (Diorio
and Witherly 2011) and a limited amount of IP-resistivity (IP-R) results acquired mid-2011 around
Current Lake-Bridge Zone (no reference).
As the earlier magnetic surveys were of higher resolution than the magnetic data acquired during the
ZTEM survey, these results were used for both the grid displays provided in this assessment. The 3D
magnetic and gravity models used in this report were also derived from the Diorio and Witherly (2011)
assessment.
The client as well provided Condor with a number of geoscience layers which were used in a variety of
the map and 3D products. These are defined where used.

3. GEOLOGY AND BACKGROUND DATA
GEOLOGICAL BACKGROUND
Regional Geology
The following is taken from the Magma web site: www.magmametals.com.au.
The TBN project is in the northern part of the Proterozoic Midcontinent Rift region, an important emerging Ni-
Cu-PGM province. The geology of this region is analogous to that of the giant Norilsk - Talnakh Ni-Cu-PGM
camp in Russia. Mineralization has been known in the Midcontinent Rift for many years; however, up until
recently it appeared to be mostly of relatively low grade. The Eagle (3.6Mt at 3.5% Ni, 3.0% Cu & 1.6g/t
Pt+Pd+Au) and Tamarack discoveries in the USA, as well as Magma’s TBN discovery, all made this decade,
have demonstrated that higher grade mineralization is also present and that the province has significant
potential to become a major North American Ni-Cu-PGM camp.
The regional geological situation is outlined in Figure 2.
Figure 2: Regional geology of mid-continent rift system.

Local Geology
The local geology is dominated by a series of four intrusive complexes that are comprised of peri-
dotite and melagabbo. These are outlined in Figure 3. Magma’s area of activity is focused on the
eastern side of the known intrusive complex and to-date, shows the best PGE mineralization en-
countered in the area. A significant resource has been developed associated with what is called
the Current Lake Intrusive Complex (Current Lake IC).
Figure 3: Local geology around Magma’s TBN property.
The Current Lake system is detailed in Figure 4. North of the Quetico Fault Zone is dominantly
Archean granite and to the south are Archean sediments.
Figure 4: Local geology of Current Lake Intrusive Complex.

TARGET MODEL
Nickel-PGE Targets
The local and regional (Figure 2) environment has been shown to host a number of deposits con-
taining significant Ni-PGE mineralization. In most cases however, (Lac des Iles, Hele and Duluth)
the mineralization is disseminated and no strong conductive features are recorded in the environ-
ment. The Eagle deposit does show a conductivity response, but it is in part masked by strongly
conductive serpentinized mafic rocks. Tamarack shows a clear magnetic response, but local cul-
tural sources appear to obscure any airborne EM response. It is reported to produce locally strong
EM responses from ground and borehole surveys.
Figures 5 and 6 show the type geological sections and aeromagnetic responses for the Current
Lake mineralized zone.
Figure 5: Geological sections through Current Lake mineralized zones (source Magma Metals Limited).
Figure 6: Magnetic profiles over three zones of mineralization hosted in Current Lake IC.

Conceptually, the Current Lake system is envisioned to look like the long section shown in Figure
7. Note, a number of massive sulfide pods are suggested along the length of the IC. The lack of
conductive responses from the VTEM survey (W Hughes personal communication) would limit the
possibility of massive sulfides to the southern part of the system where possible feeder dikes are
located at depth.
Figure 7: Long section through the Current Lake IC (source-Magma Metals Limited).
A Noril’sk style deposit is mentioned as a possible target in this environment. This style/size of
deposit is expected to produce a strong conductivity response, assuming that an EM system can
get within the detection limit of mineralization. A reference (West et al 2002) on some geophysical
aspects of the Noril’sk deposit is attached in Appendix C
In this environment with the above noted negative outcome of the VTEM survey, if holes were
drilled deeper than 350-400 m then performing borehole EM logging would be warranted to ‘scan’
for possible targets of high conductance beyond the reach of the VTEM survey.
With the present survey, the primary outcome was expected to be contacts between rock types of
differing resistivities and possibly structures with well-developed fault gouge. The detection of dis-
crete massive sulfide targets at depth are a possibility, but such features, unless very large, would
be outside the range of detection.

4. PROCESSING, ANALYSIS AND PRODUCTS
PROCESSING
Magnetic Data Processing

In addition to the normal filters available in the Geosoft application, additional processing was done
using software and algorithms; these products are termed the ZS suite. Information
pertaining to the ZS processing methodology is provided in Appendix A-Section 1. In this report, all
magnetic products (including the ZS products) are reduced to pole.
ZTEM Data
The primary data Geotech provides are the ratios of the Tzx and Tzy (sometimes shortened to X
and Y) components for the recovered frequencies; in this case 30, 45, 90, 180, 360 and 720 Hz.
The X and Y components are termed the along-line and cross-line components respectively. They
in turn provide information about the geology in the cross-line and along-line directions respective-
ly. For each X and Y component, there is both an in-phase (IP) and quadrature (QD) attribute.
These form the core data provided in the survey.
The primary form of the ZTEM data is a cross-over response such as shown in Figure 8.
Figure 8: Modeling response of ZTEM over a contact (from Lo 2007).

ZTEM 1D and Grid-Based Results

From the primary ZTEM data, a number of 1D and grid-based products are provided in the report;
these various products are listed in Table 4-1 and provided in Appendix C (report archive). The
primary data (1D results) are typically displayed as profiles. Geotech generates most of the grid-
based products which are derived from the profile data. The Geotech report provides definitions of
how these are created. The primary purpose of the grid products is to simplify the primary out-
comes by merging X and Y components and converting the cross-over style response to a positive
peak response (over a contact) similar to what the Fraser performs for VLF-EM data.
The grids are very useful as they provide what are the highest resolution outcomes of the survey in
terms of capturing the spatial changes within the AFMAG fields in the survey area. The processing
undertaken is relatively simple as well (defined in the Geotech report), so the results are fairly easy
to examine. These results do however, contain effects of topography. The primary effect of topog-
raphy is to induce conductive features on peaks and resistive features in troughs or valleys. Figure
9 shows this effect.
Figure 9: Theoretical modeling of topography and ZTEM responses.
In the present study however, topographic effects are not severe and so the grid products contain
little topographic ‘over print’.

2D ZTEM Inversion
The recorded ZTEM data were inverted using a modified 2D MT algorithm to produce a conductivi-
ty depth section (CDS). The algorithm models the along-line tipper data Tzx, taking into account
that the vertical component is airborne and the horizontal component measured in a fixed location
on the ground. The responses of all frequencies (30-720 Hz) were included in the inversion. The
version of the code used for the present study also removed any topographic effects from the data.
Further information on the processing of the ZTEM data is provided in Appendix A-Section 2.
ANALYSIS
The ZTEM methodology is a variation of the AFMAG technique (Ward et al 1966) that is fundamen-
1
tally different than all other types of airborne EM systems in that naturally occurring EM fields flow-
ing through the earth are being mapped rather than sending a primary signal from an artificial
transmitter carried by the aircraft and recording the return or secondary signal. The frequencies
used are in what is termed the AFMAG range, roughly 10 Hz-1 000 Hz. A primary difference with
ZTEM compared with active source EM techniques is that AFMAG fields produce a response via
current channeling rather than through inductive decay. This means ZTEM is typically more sensi-
tive to large scale features such as contacts, faults and changes in lithology that have a corre-
sponding change in resistivity.
As the ZTEM technique is relatively new, the processing and interpretation methodologies are still
being developed and trialed for a variety of deposit styles. As with other forms of airborne EM data,
Condor works with a combination of 1D, 2D and 3D products to represent the outcomes of the
survey data.
The understanding of the results in terms of geology is still evolving. At one level the higher fre-
quencies (which are expected to be derived from shallower depths due to the skin depth phenom-
ena of EM fields) should be expected to provide more detail at shallow depths, but at the same
time the primary AFMAG signal strengths are known to drop off with increasing frequency. Empiri-
cally, while somewhat counterintuitive based on how we assess inductive EM data, it seems the
lower frequencies (30-90 Hz) often contain more detail and appear to be more robust than the
higher frequencies (360-720 Hz).
1
AFMAG has some similarities to the VLF technique but operates at much lower frequencies.

At this stage, it is difficult to give a concise definition of how best to use different components, other
than to state that good practice would be that the various products should be examined whenever
possible and best efforts made to see if the various outcomes are contributing to understanding the
local geology. In reality however, due to the plethora of grids provided (in the present study there
are 48 derived grid products from Geotech along with the 24 primary components). A sub-set of
the provided grids are used in the survey assessment. Condor as well provides a grid product
termed Apparent Conductivity or AppCon. The underlying theory of the AppCon calculation pro-
vides for what we believe to be a more rigorous derivation of conductivity from the AFMAG results,
whereas the Geotech processing produces only relative conductivity outcomes. The AFMAG tech-
nique however, does not produce a robust measure of conductivity in absolute terms, but rather is
seen as being sensitive to differences or gradients in the conductivity.
2D Results
The primary 2D product is the conductivity depth section (CDS) that is derived from rectangular
cells of a certain X and Z dimension (see Appendix A-Section 2 for the grid used on this project).
Geotech has a program similar to Condor’s and we have seen that the two can produce very com-
parable results if the initial inversion model is defined in a similar way and also if the inversion con-
vergence or fitting is done with close to the same degree of precision. The two lines inverted by
Geotech are provided along with the Condor processing in Appendix B. With the 2D inversion
model, we have the best means to define the changes of conductivity with depth. As stated above,
we believe that the relative changes in conductivity are being accurately captured, but the absolute
values of conductivity are not well resolved. Condor has experimented with ‘seeding’ the inversion
with various starting models to help guide the inversion to a better conductivity solution but this
strategy has not produced a significant improvement.
In early 2011, Condor introduced a means to correct for topographic effects in the 2D inversions in
a timely and cost effective manner. This can be very important as topographic effects can not only
distort the shallow conductivity responses, but as well contaminate the conductivity responses at
depth.
While the 2D inversion is normally done on the X component data, the same processing can be
done on the Y component data, either along the normal lines or across lines, which in effect pro-
duces an X component outcome for an orthogonal set of survey lines. This type of processing can
be useful to help define structures which are oriented sub-parallel to the standard survey direction.

A set of the Y component derived sections are provided in addition to the X component sections
(see Appendix A Section 3).
3D Results
The 3D EM results presented for most of this assessment were created by using the 3D gridding
tool inside on the processed 2D Occam line results. As this is not a pure 3D outcome,
we have designated this as a 3D Hybrid result and labeled 3DH. Prior to completion of this assess-
ment, Condor acquired the UBC-GIF produced MTZTEM (Holtham and Oldenburg 2008) code
which allows for the full 3D modeling of ZTEM data. These results are considered as full 3D out-
comes and are identified with the label 3DF.
Condor has produced a background document on the ZTEM technique entitled “ZTEM Primer”; this
is provided with Appendix C (Archive DVD).

PRODUCTS
Table 4-1 lists the products that are provided. All maps are created using the following parameters:
Datum: WGS84
Ellipsoid: WGS84
Projection: UTM Zone 16N
Central Meridian: 87ºW
False Northing: 0
False Easting: 500 000
Scale Factor: 0.9996
Table 4-1 Survey Products

TargetMaps: @ 1: 20 000 (hardcopy + PDFs + Geosoft maps)
2
• TMI-RTP
• TMI-Tilt
• Interpretation: 45 Hz IP DT + Condor Interpretation (Figure 14)
• Interpretation: 180 IP TPR + Condor Interpretation (Figure 15)
• Interpretation: 45 Hz AppCon + Condor Interpretation (Figure 16)
• Interpretation: TMI + 180 Hz IP TPR contours + Condor Interpretation (Figure 17)
MultiPlots™ @ 1:20 000 (PDF)

Set 1: Mini-Plates™: TMI-RTP, TMI-Tilt, 45 Hz AppCon,1800 Hz IP TPR, DEM
On each standard survey line the following content is shown using Tzx In-Line data:
• Profiles-EM Z/X 30-720 Hz IP & QD observed with error fit
• Profiles-30-720 Hz AppCon
• Profiles-TMI, TMI-Tilt, 1st VD, power line monitor
• Section-Conductivity Depth Section-derived from X component + flying height
• Section-Density Depth Section from Gd (Falcon AGG 2010 survey)
• Section-Susceptibility Depth Section (Falcon MAG 2010 survey)
• TrackMap – 45 Hz Inphase DT + flight path
2
Fugro Falcon TMI outcome was used primarily but the UTS result for selective products; from Diorio and
Witherly (2011)

3
Set 2 : Mini-Plates™: TMI-RTP, TMI-Tilt, 45 Hz AppCon, 180 Hz IP TPR, DEM
On each standard survey line the following content is shown using Tzy Cross-line synthetic data:
• Profiles-EM Z/Y 30-720 Hz IP & QD observed with error fit
• Profiles-30-720 Hz AppCon
• Profiles-TMI, TMI-Tilt, 1st VD, power line monitor
• Section-Conductivity Depth Section-derived from Y component + drill holes + flying height
• Section-Density Depth Section + flying height
• Section-Susceptibility Depth Section + drill holes + flying height
• TrackMap-45 Hz Inphase DT + flight path
ZPosters: @ 1:50 000 (hardcopy + PDFs)

* The derivation of these grids is provided in the Geotech report (Schein 2011)
IP = In-Phase; QD = quadrature
Component Frequencies
IP DT* 30-720 Hz
IP TPR* 30-720 Hz
QD DT* 30-720 Hz
QD TPR* 30-720 Hz
X IP PR* 30-720 Hz
X QD PR* 30-720 Hz
Y IP PR* 30-720 Hz
Y QD PR* 30-720 Hz
AppCon 30-720 Hz
Processing and analysis report (1 hardcopy + PDF)

On the archive DVD (Appendix C) the following files are provided:
-Digital archive in Geosoft format
-TargetMaps, ZPosters and MultiPlots™ (PDF)
- session files to produce ZPosters and MultiPlots™ (includes Viewer)
-Processing & Analysis Report (PDF)
3
A set of Y derived MultiPlots has been produced along with the standard X component series. The creation
and use of the Y component series is discussed in Appendix A.

5. INTERPRETATION
DATA QUALITY
ZTEM Survey
The ZTEM data (EM and magnetics) was deemed to be of reasonable quality. A power line was
observed crossing the far western edge of the survey block. This is highlighted in Figure 10. The
power line contaminated a portion of the data in the surrounding survey; this affected area is noted
on the various figures where relevant.
Figure 10: Image of survey grid, claims and geology for TBN property; power line crosses the survey block
on the far western edge.

PROPERTY SCALE ZTEM GRID BASED OUTCOMES
Grid Based Outcomes: EM Results

While a magnetic data set was obtained concurrently with the ZTEM EM results, the client had
previously conducted two higher resolution magnetic surveys over the study area; one a dedicated
magnetic survey and the second as part of the airborne gravity survey (Diorio and Witherly 2011).
Where shown, the magnetic results from these earlier (superior) surveys are used.
Condor generally takes an empirical approach to ZTEM assessments and builds up a subsurface
model based on the various grid, 2D and 3D products, the similarly modeled magnetic results and
the available geology (includes conceptual geological knowledge as well as direct observation).
Figures 11-12-13 show a series of grids of the ZTEM IP results for 30, 45, 180 and 720 Hz. Figure
11 are the Geotech-produced DT grids, Figure 12 the Geotech-produced TPR grids and Figure 13
the Condor-produced AppCon grids.
Figure 11: 30, 45, 180 and 720 Hz IP DT outcomes.

Figure 12: 30, 45, 180 and 720 Hz IP TPR outcomes.
This set of images shows there are marked changes both with frequency and between the different
products. Condor’s experience with ZTEM has been that there are sometimes significant differ-
ences between the various products and frequencies, but not always. The reasons for these differ-
ences (or similarities) are not well understood, as many of these changes occur where there is little
established geological control. There is a definable depth of investigation based on frequency, but
this does not always appear to provide a consistent outcome.
So as to highlight some of the similarities and differences in the grid products, Figures 14, 15 and
16 show the same three components as used in the previous figures; DT, TPR and AppCon. The
frequency selected varies so as to try and capture the maximum information content for that com-
ponent. Comments on these figures are provided below.
Figure 14-45 Hz IP DT: The major intrusive complexes; Current Lake, Steepledge and RT appear
as linears. A third feature just west of Steepledge also is apparent and is designated Ray Lake.

Figure 13: 30, 45, 180 and 720 Hz AppCon outcomes.

.
Another area of anomalous conductivity is designated Ros 1_2. The geology looks to be favorable
but the area has seen limited drilling.
There is a major E-W trending zone of elevated conductivity east of Current Lake. This area is not
currently recognized as hosting favourable geology. A strong N-S trending conductor is mapped on
the far eastern side of the property; this is designated as a ‘Non-Interesting Target” (NIT) by the
client. A complex zone of conductivity designated the Lil South is noted on the far west side of the
survey area. The Lone Island Lake IC lies at the north end of this feature (which is also near the
power line contaminated area). This feature shows a wide variety of responses with both frequency
and component.
Figure 15-180 Hz IP TPR: This image captures some of the same features as apparent in Figure
14, but as well shows responses that are more subdued or different. The three linears, two of

Figure 14: 45 Hz IP DT for with geological notes.
which are associated with the two main intrusive complexes Current Lake and Steepledge, are
mapped but are subtler responses. However, the Current Lake trend extends further to the north
than is apparent in Figure 14, up to what has been designated as the Ros 1_2 area. Lil South is
now a monolithic high rather than being a series of more discrete zones grouped together. There is
also a new zone of elevated conductivity along the south-central part of the survey block. The zone
east of the Current Lake IC is still present, but much more reduced in overall size and amplitude.
The NIT is still apparent.
Figure 16: 45 Hz AppCon. In this image, the three linears; Current Lake IC, Steepledge/RT IC and
possibly Ray Lake are apparent. However, the northern extents of the Current Lake and Stee-
pledge trends are more subdued in response, similar to their character in Figure 15. There seems
to be a Ray Lake trend, but it lies south of where it appears in Figures 14 and 15 so might be
something different. Lil South is smaller in size and amplitude, but still sits in the same location.

Figure 15: 180 Hz IP TPR with geological notes.
There is a zone of elevated conductivity in the south-central part of the survey. This has some simi-
larity to Figure 15, but the location is displaced to the SW. The NIT feature as a stand-alone con-
ductor is no longer apparent in this image, but there is an elevated zone of conductivity along the
eastern side of the survey block. The area of Lone Island Lake IC has been removed due to the
effects of the power line on the calculation of the AppCon response.
Grid Based Outcomes: EM and Magnetic Results

Figure 17 shows a color image of the TMI with the 180 Hz IP TPR response presented as both
contours and a pseudo-relief on the magnetic results. This image shows that portions of the ZTEM
responses correlate with the magnetic character and portions do not. The central part of the Cur-
rent Lake mineralized system appears as a coincident EM and magnetic feature. However, the
segment that extends northwards to the area termed Ros 1_2 only appears as an EM response.

Figure 16: 45 Hz AppCon.
The Steepledge IC also appears to show coincident EM and magnetic responses but the zone
appears more complex that the Current Lake system. The Ray Lake feature only appears as an
EM responder. The RT IC shows as a discrete magnetic high and EM response. The NIT feature
also appears as a discrete magnetic source. The area of elevated conductivity noted in Figure 14
east of the Current Lake IC appears to loosely correlate with a portion of the major E-W trending
magnetic high that crosses the survey block and lies in close proximity to the Quetico Fault Zone.
Grid Based Outcomes: EM and Gravity Results

The ZTEM and AGG outcomes are shown together in Figure 18. The upper image shows a grid of
the Falcon AGG (Gd) results with the Quetico Fault and metasedimentary unit outlined. The lower
image shows the 180 Hz IP TPR grid with the AGG results shown as contours. The metasedimen-
tary unit is the dominant feature in the gravity response, both in terms of amplitude and spatial
extent, spanning the full E-W limits of the survey block. The area south of the gravity high is not

Figure 17: TMI (color) with 180 Hz IP TPR as relief on magnetics and as contours.
explicitly defined geologically but is assumed to be Archean sediments, but of a lower density to
those which the gravity highlights. The Quetico Fault lies along the northern boundary of the
metasedimentary unit and the granite to the north. Most of the TBN intrusive system lies within the
granite and metasedimentary units. Within the granite, there is a variable strike from N30E to
N50E. Within the metasedimentary unit, there is a strong elongation parallel to strike suggesting
that the TBN intrusive system has selectively intruded into the sedimentary unit along possible
zones of weakness parallel to the unit’s strike direction.
There is basically no expression of the metasedimentary unit in the ZTEM results. This implies that
the metasedimentary unit has a resistivity similar to host rocks. Also, neither of the two faults
mapped, Quetico and Escape Lake, show anything but a subtle expression in the ZTEM results.
Discussion of Grid-Based Outcomes

The ZTEM survey shows little influence from the Archean lithologies or structures (the age of the
structures is unclear, but likely have been active over a period of time including during the em-
placement of TBN system). A major part of the E-W trending TBN system (mafic intrusions) ap-
pears to have been introduced between the Escape Lake Fault and the southern margin of the

Figure 18: Falcon AGG outcome with Quetico and Escape Lake Faults and metasedimentary unit (top); 180
Hz IP TPR image with Falcon AGG response as contours (bottom).
metasedimentary unit. This suggests that there was some open space/weakness present prior to
the intrusive event. This suggests that besides being denser, the metasedimentary rocks were also
more brittle than the surrounding rocks. While the 180 Hz IP TPR results do not appear to show
the E-W system, the 720 Hz IP DT and TPR results (Figures 11 and 12) do show character that
correlates with the mafic dike system. As highlighted in Figures 14-17, the Current Lake, Stee-
pledge/RT ICs are reasonably well imaged in the ZTEM results.

Additionally, at Ray Lake an en echelon response is noted west of the above two zones. There is
also a ‘bright spot’ noted at the northern end of the Current Lake Zone around Ros 1_2 and an
elevated response in the western part of the survey area termed Lil South. At Lil South however,
the response does not look the same as noted above at Current Lake and Steepledge/RT, but is a
large zone of higher conductivity with a series of lineations through the zone (refer to Figure 16).
This feature appears to extend south of the survey area.
The commentary above did not focus on one frequency, but ranged basically over the full suite
from 45 Hz to 180 Hz to 720 Hz. While such a range of frequencies implies a range of depths of
investigation, the geological features involved however, are all believed to be shallow and none of
them are thought to be very conductive. Nor is there any appreciable amount of conductive near-
surface material such as lake sediments. The 2D Occam inversion results discussed next contain
what is believed the most reliable presentation of the results with respect to depth.
3D BASED ASSESSMENT
As discussed in Section 4, 2D Occam inversions were run on all the survey lines. A 3DH voxel
model was then created. Figure 19 shows two perspective views of this 3DH model along with cli-
ent-provided layers. The major feature in the voxel model is a large conductive wedge that extends
from the eastern side of the property and dips westward. The model also shows two styles of re-
sponse associated with the Current Lake and Steepledge Lake/RT ICs; at surface there is a con-
ductive ‘rind’ that extends around the ICs. At depth under the ICs, there are two ‘tubes’ or confined
zones of conductivity that extend over a number of survey lines that shows basically the same lin-
4
ear extent as the known extent of the ICs.
Figure 20 provides a cross-sectional view of the voxel model where the major westward dipping
4
Condor has encountered ‘tube’ responses in several other ZTEM projects, but to date not enough geological control has
been available to establish with a high degree of certainty that the tubes represent real geological features and are not
some artefact in the acquisition or data processing. Nevertheless, various components of the data have been processed
and evaluated to try and establish whether there is a legitimate non-geological origin to the tubes, or they represent a real
geological feature. Our opinion is that the tubes do represent some legitimate sub-surface response, but the depth and
geometry of this may not be exactly as shown by the tubes. Condor will continue to assess these data at no cost to the
client and pass along any significant outcomes as they become available.

conductivity zone and two tubes are highlighted. The tubes are fairly deep, both approximately 1
km in depth.
Figures 21 and 22 provide a series of ‘walk-around’ shots of the EM voxel model with various cli-
ent-provided layers. The major westward dipping conductivity zone does not represent any current-
ly understood geological relationship. The grid-based assessment showed that neither the granite
nor the metasediments showed a diagnostic conductivity response. In the Lake Nipigon area to the
north of the TBN project, the Sibley sediments show a variable and sometimes strong conductivity
due to contained saline water. Condor is not aware that Sibley sediments are present in the study
area and if so, would not be expected to be underlying Archean age rocks. The other possibilities
are that this zone represents a previously unrecognized conductive Archean-age rock unit (likely
graphitic) or the ZTEM is mapping a dipping fault zone or thrust that contains conductive gouge.
Figure 19: Perspective views of Occam 3D voxel model; view to 021.
Figure 20: Cross section through voxel model showing major conductive zone and tubes; view to 337.

Figure 21: Four views of the ZTEM 3D inversion model; views various as noted in upper corner of image.
Figure 22: Four views of the ZTEM 3D inversion model with geology layers; views various as
noted in upper corner of image.

Comparison 2D-3DH-3DF Inversion of ZTEM

Although not defined as part of the processing and analysis contract agreed to by Magma, Condor
used the UBC-GIF code MTZTEM (Holtham and Oldenburg 2008) to model the central part of the
survey area. Figure 23 shows a comparison slice through the 3DF model that corresponds to
L1180 and the 2D Occam inversion for that line.
Figure 23: Comparison of 3D and 2D inversion for L1180.
The overall major structure of the line appears quite similar in the two models; the tubes are mapped
with both approaches as is the deeper conductive unit coming in from the east. However, the depths of
the tubes in the 3DF outcome are considerably shallower than in the 2D outcome. The deeper
conductive zone however, appears at approximately the same depth in both outcomes, albeit the 3DF
model conductivity is lower than the 2D Occam shows. As noted earlier, the processing to date while
encouraging still has ambiguity as to the depth and the geological significance of what are appearing
as tubes. While in this setting, the known (shallow) mineralization at Current Lake shows a tube-like
character (Figure 5), the other locations where Condor has observed tubes in ZTEM outcomes lacked
a clear geological model to support this form in the data. One hypothesis has been that the tubes could
represent the top on a buried contact and the ZTEM is highlighting only the upper-most of the feature,
resulting in a tube-like shape to the inversion. The conclusion at this stage is that below the Current
Lake and Steepledge Lake ICs, there lies some geological feature which is deemed part of the system
observed at surface. While the extent of this feature appears to agree with what is mapped at surface,
the depth is uncertain at this time.

Comparison of 3D Outcomes for Magnetic and Gravity Results

Figures 24 and 25 show the 3D models for the magnetic and gravity models (Diorio and Witherly
2011) along with the geology as provided to Condor. In Figure 24, two images of the magnetic
results are shown; the upper image shows the magnetic ‘skeleton’; a DXF with a 0.003 SI isosur-
face. These images show that the magnetic rocks are a mixture of Archean and Proterozoic ages.
The Archean rocks all form fairly continuous linear features that trend ~ N80E. The main line of
TBN intrusive bodies’ form a trend that is offset to the Archean grain and trends almost EW.
Figure 24: Mag3D model with geology layers (above) and with TMI grid (below); view to 018.

Some of the TBN dikes as well trend NE, which suggests there was a pre-existing structural grain
through the study area which the TBI system then ‘occupied’ during the intrusive event. In the 3D
modeling, the variable magnetization that characterizes the TBN system results in apparent holes
in the 3D surface, which is how the inversion code interprets the TMI negatives (absence of the
magnetic material).
The major feature of interest in Figure 25 is how the TBN IC systems have stoped out the
metasedimentary unit. This shows that the rheology of the metasedimentary unit was critical to
localizing the TBC IC during the time of emplacement. The expectation is a high density zone north
of the metasedimentary trend (presumably in granitic rocks) where two TBN dikes appear to coa-
lesce. The Grav3D model and Gd grid both show a localized gravity high in this area.
Comparison of 3DH EM with Magnetic and Gravity 3D Inversions

The major features in the magnetic and gravity results have now been discussed. In this last sec-
tion these outcomes are compared with the ZTEM results. Figures 26-29 show various views of the
combined models as seen from slightly above the model. Figure 29 views the model from below to
better show the Lil South feature. The very shallow conductive patches discussed earlier are left off
these models as they obscure the detail below.
In the suite of images, it is apparent how the oblate Proterozoic magnetic intrusions have intruded
into the metasediments. As noted earlier, E-W trending intrusive dikes are apparent along the trace
of the Escape Lake Fault.
One aspect of the modeling that is not clear at this point is whether the apparent depths to the
bottom of the magnetic and gravity models are geologically ‘correct’ as suggested by the inversion,
or have been imposed by the constraints applied during the inversion modeling. The reason this is
of interest is that the series of images clearly shows that the bottoms of the susceptibility and den-
sity models lie above the major part of the ZTEM conductivity horizon discussed earlier. If this is
correct, then it gives support to the interpretation of the ZTEM surface as ‘cutting off’ the overlying
geology, most likely by a fault/thrust that dips westward. This is only a hypothesis at the moment
and further work will be required to assess how the potential models were created.
The three oblate magnetic intrusive centers that have intruded into the metasediments appear
(based on the magnetic results) to be getting larger moving to the east. The Current Lake deposit

Figure 25: Grav3D model with geology layers (above) and with Gd grid (below); view to 018.
is also closest to the largest of these bodies. All three of these bodies appear finite in depth extent
but presumably there is a plumbing system either below or tied in laterally with these bodies where
the juvenile intrusive material was pumped in. The client’s target model (Figure 9) suggests these
entry points could be prospective for deposition of higher grade mineralization (i.e. massive sul-
fides). As noted earlier, the ZTEM is not expected to be able to image deposits of the expected
size (using Eagle as a type model) at depths greater than 250-300 m depth.

Figure 26: Combined EM, magnetic and gravity models; view to 331.


Deposit and 2D-3D Geophysical Models

In this section, we examine the 3D deposit model and the various 2D-3D models discussed above.
The 3D model of the Current Lake deposit was provided by the client as a DXF file with the desig-
nation WF_BZBL_Pt25_20111027.dxf.
The first comparison is shown in Figure 30; this shows the result of the 2D Occam inversion for
L1150 that cuts through the middle of the deposit. The relationship between the deposit and the
Current Lake tube is apparent. There appears to be a suggestion of a dipping trend to the east off
the end of the deposit; this is highlighted with the double arrow.
Figure 30: Tzx 2D Occam inversion for L1150 over the Current Lake deposit.
Figure 31 shows the results of the 2D Occam processing performed on the Tzy component. Again,
there is a sense of a dipping horizon to the south that is aligned with the dip of the deposit.

Figure 31: Tzy 2D Occam inversion for L1540 over the Current Lake deposit.
Finally in Figure 32, the results of the 2D Occam and MTZTEM 3D inversion are shown. As a result
of the MTZTEM modeling being done largely to the north of the deposit (where the tubes are more
prominent), the line cuts the deposit just at its northern extremity. In this case, the MTZTEM results
show some suggestion of a dip to the east whereas the 2D Occam are less clear.

Figure 32: 3D MZTEM and 2D Occam results for L1180 over the Current Lake deposit.
Comparison with IP-R Survey

Condor has undertaken processing of an IP-R survey conducted for the client over the main min-
eralized area. These data have not been assessed systematically, but an image of one of the de-
rived conductivity images is presented in Figure 30 along with one of the ZTEM TPR grids and the
UTS TMI. The conductivity results are compared with the 90 Hz IP TPR result. The magnetic result
is shown since at first examination the conductivity and magnetic results appear to show a close
correlation. At the north end of the Current Lake Zone (Figure 4), both the conductivity and mag-
netic result shows a sharp westward bend to the horizon. The ZTEM response however, appears
to continue on to the NW towards the Ros 1_2 area.

Figure 33: 90 Hz IP TPR, Conductivity at 75 m below surface and UTS TMI.
This suggests that the ZTEM is either responding to a different aspect of the geology at the same
depth as the IP-R and magnetics is derived, or it is sensing the geological system in the same spa-
tial location but at a greater depth with different petrophysical characteristics. This will be examined
further when the IP-R results are reviewed.
SUMMARY OF TARGET ZONES
An outline of the ZTEM Target Zones is shown in Figure 29. Figures 35-43 are snap-shots of the
main geophysical responses associated with the Target Zones. Summary comments about each
Zone are provided in Table 5-1. The defined ranking is intended to be suggestive and when new
knowledge is acquired in any of the Zones, the assessment of the other Zones could be affected.
DISCUSSION
An assessment of the ZTEM results has been carried out using grid-based, 2D and 3D (3D grid-
ding of 2D and full 3D) products. These results have been assessed in light of the available geolo-
gy and outcomes from earlier processing done to the magnetic and gravity data sets.
The grid-based results show clear responses corresponding to the NW-trending zones of minerali-
zation at the Current Lake and Steepledge Lake/RT zones. A third zone appears to have been
defined that correlates with the Ray Lake Zone. Conductive patches are defined at Ros 1_2 and
east of the Current Lake zone (designated West Greenwich Lake).

Figure 34: ZTEM flight path, TBN IC system and Target Zones.
Based on the geology that defines the known mineralized system at the Current Lake deposit, the
cross-section of the mineralization would appear to be inadequate to cause the observed ZTEM
5
response. A brief examination of the conductivity derived from the IP-R survey over the Current
Lake Zone shows good correlation with the high resolution magnetic survey, but there is not as
consistent correlation with the ZTEM, suggesting that different aspects of the geology (either in
composition or depth or both) are being highlighted in the three surveys.
The 2D and 3D assessments both show shallow conductive zones associated with the Current
Lake and Steepledge Lake/RT zones. As well, under both zones, extensive tube-like zones of con-
ductivity are apparent. While the 2D and 3D processing provides different depths to the tubes, the
two different forms of processing both support the tubes being present. The geological nature of
these features is unclear at this stage.
5
Simply the inversion of the resistivity parameter derived from the IP-R survey but allows for a more direct
correlation with the ZTEM results which are expressed as conductivity.

A discrete deep conductive feature appears associated with the Lil South zone.
A major zone of westward dipping conductivity appears as a floor to the survey area. This feature
appears to form a cut-off to the magnetic and gravity models at depth and if this is the case, could
mean the ZTEM feature represents a thrust plate.
The examination of the magnetic and gravity results did not show any strong correlations with the
ZTEM results. At best, the ZTEM showed a subtle response to either the E-W trending mafic dike
system or the southern edge of the dense metasedimentary unit bounded by the Escape Lake
Fault.
The ZTEM survey presents outcomes which show correlation with the known economic geology
and provides suggestions as to other zones of potential mineralization. The ZTEM results are how-
ever, not totally internally consistent and establishing the depth of features of interest is difficult.
This is considered as a result of two factors; the likely subtle nature of the changes in geology that
can produce AFMAG responses and the limitations of the acquired data and available processing
techniques. More than with most other geophysical techniques, quantifying the data and modeling
uncertainty of ZTEM is still a developing skill.
FOLLOW-UP
While the ZTEM is a resistivity mapping technique, follow-up of ZTEM features has to take into
account aspects of the interpreted geological form of the responses as well as the inferred petro-
physics. As noted earlier, absolute value of resistivity is less important than the resistivity gradients
between different units that are giving rise to the ZTEM response. For shallow features such as
Ray Lake, Ros 1_2_3 or West Greenwich Lake, an IP-R survey similar to that conducted over the
Current Lake-Bridge Zone should suffice.
For the Lil South feature or the Steepledge Lake/RT and Current Lake tubes, further study of the
geophysical results and available geology should be undertaken prior to any more field work. Se-
lective forward modeling of the ZTEM and potential field results may help to better define whether
some of the outcomes suggested in this assessment met a reasonable set of exploration criteria to
justify what would be expensive testing.

Figure 35: Summary geophysical images for Lil South.
Figure 36: Summary geophysical images for Ray Lake.

Figure 37: Summary geophysical images for SL East.
Figure 38: Summary geophysical images for RT.

Figure 39: Summary geophysical images for Ros 1_2.
Figure 40: Summary geophysical images for Ros 3.

Figure 41: Summary geophysical images for Current Lake deposit.
Figure 42: Summary geophysical images for EL 1_2_3.

Figure 43: Summary geophysical images for West Greenwich Lake.

Table 5-1: Summary of Target Zones

Target ID ZTEM response Other data Ranking Figure
L= low
M = moderate
H = high
Lil South Target is located on north side Mag suggests NE L-M 35
of large EM high; some of the trending break going
EM images are complex. 2D through center of the
modeling shows deep feature EM high
along southern edge of survey
block
Ray Lake 45 Hz IP DT shows feature DT response correlates M 36
trending NNW from point loca- closely with outline of
tion of Zone. Trend of geology is Ray Lake; however,
more NE similar response noted
associated with Current
Lake suggesting there
is a structural control to
the lakes
RT Both ZTEM images show NW Zone sits directly over M 37
trending response over Zone gravity high
SL East 45 Hz IP DT response is similar Displaced from mag L 38
to RT Zone and gravity features
Ros 1_2 45 IP DT shows strong response Local gravity high not- M-H 39
ed under Ros_2. Mag
is broken up
Ros 3 Just off 45 Hz IP DT zone but Mag low and along M 40
along 180 Hz IP TPR trend gravity gradient
CLD Major NW trends in 45 Hz IP DT Well defined mag fea- H 41
and 180 Hz IP TPR ture; on north side of
gravity high
El 1_2_3 EM responses are complex and Mag is not coherent. L 42
broken up (45 Hz IP DT) On southern flank of
gravity high
West Major conductive path in 45 Hz Mag appears to show M 43
Greenwich IP DT; no clear 180 Hz response closed form bound by
Lake reversed responses

6. CONCLUSIONS AND RECOMMENDATIONS
An assessment has been made of a ZTEM AFMAG and magnetic survey carried out over the TBN
property of Magma Metals that hosts the Current Lake PGE deposit. The survey produced responses
which showed reasonable correlation with several of the known areas of mineralization and highlighted
several other areas as being of potential interest.
Modeling the data with 2D and 3D codes produced both interesting and somewhat contradictory
results that could not be totally reconciled within the scope of this assessment. While further work can
be done, the 2D and 3D results appear to suggest that there are deeper conductive structures on the
property that are spatially associated with the known shallow mineralization. It has been difficult to
prove conclusively the depth of these features and define a most probable geological nature to these
geophysical features.
The ZTEM results provide information about the possibility of deep-seated structure underling the
study area. When compared with the magnetic and gravity modeling, this suggests a more complex
picture of the 3D geology down to a depth of several kms. The implications of this interpretation in
terms of the overall economic geology are not totally clear to the author, but hopefully will provide the
client with a better overall framework with which to assess the property and its potential.
Several of the defined Target Zones are considered to be shallow and deemed suitable for basic IP-R
surveying similar to what was carried out over the Current Lake-Bridge Zone. The deeper Zones,
including the tubes underlying the Steepledge Lake/RT and Current Lake ICs, need to be studied
further using the existing geology with some further geophysical modeling, specifically a full 3D
modeling of the ZTEM results.
The previously produced magnetic and gravity modeling were examined in light of known economic
geology. Some correlations with the mineralized zones were noted which support earlier assessments,
as well as potentially highlight relationships not previously recognized.
Respectfully submitted,
Ken Witherly
January 2, 2012

7. REFERENCES
Diorio, P., Witherly, K., 2011 Report on Processing & Analysis Of Falcon AGG and Magnetic Data
Thunder Bay North Project, Ontario for Magma Metals (Canada) Limited; report by Condor Con-
sulting, Inc. January 2011.
Holtham, E., and D. Oldenburg, 2008, Three-dimensional forward modelling and inversion of Z-
TEM data: 78th Annual International Meeting, SEG, Expanded Abstracts, 564-568.
Hughes, W. 2011 Results of a VTEM survey over the TBN property; personnel communication.
Lo, B. 2007 Report on Airborne Geophysical ZTEM and Magnetic Survey for JNR Resources Inc.
Geotech Ltd. report; project #8244.
Schein, E. 2011 Report on a Helicopter-Borne Z-Axis Tipper Electromagnetic (ZTEM) and Aero-
magnetic Geophysical Survey Thunder Bay North Block, Thunder Bay, Ontario; by Geotech Ltd.,
Project # 11182, August 2011.
Ward, S., O’Donnell, J., Rivera, R., Ware H., and Fraser D. 1966 AFMAG-Applications and Limita-
tions; Geophysics Vol. 31 No 3 1966 p 576-605.
West, R., Diakov, S., Schissel D., Kritsov, A., Kochnev-Pervoukhov and Migachev, I. 2002 Review
of Geophysical Exploration at Noril’sk Cu-Ni-PGM Deposit; SEG 2002 Ni-PGM workshop Salt Lake
City.

8. APPENDICES

APPENDIX A - INFORMATION ON PROCESSING EM AND MAGNETIC DATA

Section 1: Magnetic Processing: ZS and Mag3D

New enhancement filters for geological mapping
Zhiqun Shi* Graham Butt
Encom Technology Australia Encom Technology Australia
zhiqun.shi@encom.com.au graham.butt@encom.com.au
boundaries, particularly in colour image presentations. They

usually emphasise short wavelength anomalies at the expense
SUMMARY of signal from deeper magnetic sources and the range of
Two types of filters have been developed for the purpose amplitudes remaining in the filtered output may dominate the
of enhancing weak magnetic anomalies from near-surface source boundary information being sought. In addition, some
sources while simultaneously enhancing low-amplitude, traditional filters emphasise noise with resultant impact on the
long-wavelength magnetic anomalies from deep-seated or interpretation of source boundaries.
regional sources. The Edge filter group highlights edges
surrounding both shallow and deeper magnetic sources. This paper identifies new processes which have been
The results are used to infer the location of the developed to address these disadvantages and provide output
boundaries of magnetised lithologies. The Block filter which can improve map-based interpretations.
group has the effect of transforming the data into “zones” Unless otherwise stated, all filters have been operated on TMI
which, similar to image classification systems, segregate data reduced to the pole (RTP).
anomalous zones into apparent lithological categories.
Both filter groups change the textural character of a METHOD AND RESULTS
dataset and thereby facilitate interpretation of geological
Theoretical Model Testing
structures.
A theoretical 2D grid of total magnetic intensity (TMI)
The effect of each filter is demonstrated using theoretical computed at the surface was created by forward 3D modelling
model studies. The models include both shallow and deep of the TMI response from a set of theoretical magnetic sources
sources with a range of magnetisations. Comparative having variable width, strike extent, depth, depth extent (DE),
studies are made with traditional filters using the same dip, magnetic susceptibility and strike azimuth. A list of these
theoretical models. In order to simulate real conditions, parameters is presented in Table 1. In two of the sources,
Gaussian noise has been added to the model response. remanence was simulated using negative magnetic
Techniques for noise reduction and geological signature susceptibility. The TMI of the theoretical models was
enhancement are discussed in the paper. computed at a geomagnetic inclination of -60 degrees using a
The new approaches are applied to actual magnetic notional east–west line spacing of 200 m and a grid cell size
survey data covering part of the Goulburn 1:100 000 of 40 m. The TMI grid was then reduced to the pole (RTP)
scale map sheet area, New South Wales. Some new (Figure 1).
geological inferences revealed by this process are
discussed
Key words: Enhancement filters, magnetic sources,
geological mapping.
INTRODUCTION
High-resolution aeromagnetic survey data represent a rich
source of detailed information for mapping surface geology as
well as for mapping deep tectonic structure. Traditional
enhancement techniques, such as first vertical and horizontal
derivatives (1VD, 1HD), analytic signal (AS), and high-pass
in-line or grid filters are used in enhancing magnetic
anomalies from near-surface geology.
In recent years the potential field tilt filter has been introduced
(Miller and Singh, 1994) and it has achieved recognition for
its value in the analysis of potential field data for structural
mapping and enhancement of both weak and strong magnetic
anomalies (Verduzco et al, 2004). The total horizontal
derivative of the TMI reduced to the pole is also widely used
for detecting edges or boundaries of magnetic sources
(Cordell and Grauch, 1985; Blakely and Simpson, 1986; Figure 1. RTP image derived from multiple theoretical 3D
Phillips, 1998). magnetic sources, shown as wire frame outlines
Several disadvantages pertain to the use of these traditional A set of traditional filters was operated on the theoretical RTP
filters. They often only diffusely identify source location and grid. They include AS, 1VD, modulus of horizontal
derivatives (MS) and Tilt and the results are presented in
th
ASEG 17 Geophysical Conference and Exhibition, Sydney 2004. Extended Abstracts
New enhancement filters for geological mapping Shi and Butt
Figure 2. The output grids variously show discontinuous dominate the edge interpretation. This has the ancillary effect
trending (crossed sources in upper right of AS image), diffuse, that the method can be modified to provide automated edge
weak edges (deep source in centre right of the MS image) and conversion to vectors for use in GIS systems.
lack of precise source edge definition (1VD and Tilt).
Although this filter significantly improves the precision of
Model Dip
Magnetic Strike
Azimuth edge determination, it is subject to normal potential field
Depth (m) Width (m) DE (m) Susceptibility Length
Label (deg)
(SI) (m)
(deg) limitations which determine that source edges cannot be
1 4000 15000 15000 120 0.010 15000 -050 resolved where the source is narrow relative to its depth. The
2 6000 15000 10000 120 0.010 15000 -050
3 10000 15000 10000 120 0.010 15000 -050 filter also can produce a “halo” type artefact due to
4 1000 3000 4000 70 0.010 12000 -055
5 500 5000 2000 60 0.010 7000 -050 superposition of the response of a limited depth extent shallow
6
7
1000
600
800
500
2000
2000
150
120
0.005
0.001
8000
20000
-030
-020
source (Figure 1, Model 6) on that of deeper sources. A
8 200 500 2000 120 0.001 20000 -020 similar “halo” effect can be seen around the edges of
9 500 500 2000 120 0.003 10000 020
10 1000 500 2000 120 0.003 10000 -060 remanently magnetised Model 15, also in Figure 1.
11 1000 500 2000 120 0.003 12000 040
12 200 400 2000 120 0.001 20000 -050 The ZS-Edge filter (Figure 4) has also been developed to map
13 500 400 1000 40 0.002 32000 050
14 500 400 1000 140 0.001 32000 050 source edges. This filter differs from the ZS-Edgezone filter in
15 600 3000 4000 90 -0.002 8000 055
16 400 600 2000 120 -0.010 8000 -010
that a greater contribution of the TMI anomaly amplitude over
the source is retained, thereby improving anomaly
characterisation at the expense of edge sharpness.
Table 1. List of parameters of theoretical magnetic sources
Both these filters produce edges which migrate down-dip
towards the deepest edge of the source. This effect produces
anomaly asymmetry that can assist interpretation of dip,
although this effect is more pronounced for the ZS-Edge filter
than for the ZS-Edgezone filter. Down-dip source extensions
are depicted in cyan in Figure 1.
Figure 3. Anomaly edge and block enhancements using

the ZS-Edgezone (left) and ZS-Block filters (right). Model
positions are shown using wire frames.
Block Filters
In attempting to improve edge detection filters, an obvious
progression is to highlight the magnetic regions whose edges
have been mapped. To do this, a set of filters called “block”
Figure 2. Comparison of enhancement filters of RTP: AS, filters has been developed.
1VD, MS and Tilt filter. The models used are those The Block filter group has the effect of transforming the
depicted in Figure 1. potential field data into “zones” which, similar to image
classification systems, segregate anomalous zones into
Edge Filters apparent lithological categories. These filters can be imported
The first avenue of development was to increase the sharpness for use in image classification systems or displayed in RGB
of the anomalies used to map the edge of the magnetic space with other grids for empirical classification purposes.
sources. The MS grid yields anomaly peaks over the source The block filters, like the edge filters, are linear, derivative-
edge locations, whereas these edges coincide with gradients in based filters which use a combination of derivative and
the 1VD, Tilt and AS filtered outputs. None of these filters amplitude compression techniques to render the magnetic data
produces easily interpreted edges in image form when the into regions whose edges are sharply defined and whose
sources are weakly magnetised or are deep. amplitudes have a reduced range in comparison to the original
A new linear, derivative-based filter termed the ZS-Edgezone TMI.
filter has been developed to improve edge detection in these The ZS-Block filter (Figure 3) and the ZS-Plateau filter
situations. Its effect is shown in Figure 3 using the same (Figure 4) depict the magnetic data as a 2D plan of apparent
theoretical models discussed earlier. The advantages of the magnetic source distribution. Artefacts may occur as
filter are greatly increased anomaly sharpness over source discussed for the edge filters.
edges and compression of the amplitude range so that
differences in the original TMI amplitudes do not persist to
th
The choice of ZS-Block, ZS-Plateau or ZS-Area filters will

depend on the data characteristics of each magnetic survey
and on the end-use requirement. The ZS-Plateau filter, for
example, yields less variation in amplitude “texture” over a
magnetic unit that either the ZS-Block or ZS-Area filters.
Figure 5. Comparison of 1VD of original model RTP data

(top left) with noise-affected RTP data (top right) and
noise-reduced RTP data (lower images)
Figure 4. Comparison of ZS-Edge, ZS-Edgezone, ZS-

Block and ZS-Plateau filtered outputs of RTP data
Effects of Noise
The influence of noise on the operation of these enhanced
grids was tested by adding a large component of noise to the
theoretical TMI profile data. This noise had a Gaussian
distribution with a standard deviation equal to ten percent of
the TMI standard deviation. The noise-modified TMI profile
data were then de-spiked using a non-linear technique. Both
the noise-affected and the de-spiked TMI data were then
gridded and converted to RTP. The RTP data were then
processed both with the traditional and newly developed
filters.
Figure 5 shows the effect of the noise on the computations.
The image of the noise-affected 1VD RTP data (top right) Figure 6. ZS-Block filter using noise-reduced RTP data
shows that weak and deep sources have been severely masked (top left) and examples of filter combinations in RGB
by the noise. Significant improvement can be achieved by space using noise-reduced RTP data
using de-spiked data (lower left) or by low-pass grid filtering
— for example, using an upward continuation filter (lower Application to Field Data, Goulburn 1:100 000 Scale Map
right). Sheet Area, New South Wales
Figure 6 shows that if real data with significant noise is Both the traditional and new enhancement filters were applied
encountered, a standard de-spiking or low-pass smoothing to test their suitability for geological definition to airborne
procedure may be used to achieve successful application of magnetic survey data over the Goulburn 1:100 000 scale map
both the traditional and newly developed filters. sheet area (Johnson et al, 2003). These data were acquired as
part of a joint program between the NSW Department of
Figure 6 also depicts the use of enhanced outputs in RGB
Mineral Resources and Geoscience Australia, with 250 m–
space to provide examples of how the combination of
spaced east–west flightlines. The magnetometer sensor
amplitude information (red colour) with edge information
occupied a nominal terrain clearance of 80 m. This dataset
(green and blue colours) can be used to highlight source
was selected since new detailed geological mapping had been
boundaries and remanence in a single image.
recently completed. All the enhancements have been
computed using TMI data reduced to the pole.
Figure 7 shows a comparison of part of the Goulburn
1:100 000 map sheet area surface geology with the ZS-Area
th
filter output. In the area surrounding location C, the ZS-Area

filter transforms the magnetic data into separate magnetic
units, which comprise the Devonian Bindook Volcanic
Complex. The magnetic regions correlate closely with mapped
andesites (Dkqa–cream coloured unit in Figure 7) whilst the
intervening less-magnetic units correlate with rhyolitic
ignimbrites (Dkqy–red unit in Figure 7)
Figure 9. Comparison of RTP and Tilt filters over the

Bindook Volcanic Complex
CONCLUSIONS
Traditional filters used to enhance magnetic data, including
the more recently developed potential field tilt filter, are
currently used to assist in determination of the location and
Figure 7. Comparison of geology and ZS-Area enhance- extent of magnetic units.
ment over the Bindook Volcanic Complex Newly developed derivative-based filters may be used to
Figure 8 displays some of the advantages of the edge detection improve the precision of source edge detection and, by
filters. At location A, ambiguity concerning the continuity of extension, the determination of the spatial extent of magnetic
Quialigo Formation units (cream and red units in Figure 7) is units. These filters are demonstrated to perform successfully
resolved by the ZS-Edgezone filter. At location B, a subtle on both strongly magnetised features as well as on weakly
lineament is confirmed, whilst at location D, the extent of the magnetised or deep magnetic features. Artefacts may result
Bullamalita Conglomerate (green unit in Figure 7) is clearly particularly where anomaly superposition occurs.
mapped by the ZS-Edge filter. Structural breaks are often The impact of noise in real data may be accommodated by
more easily interpreted using these transforms, for example, these new methods provided noise-reduction techniques are
immediately southwest of location D. employed.
The new filter outputs may be used as part of regional or
detailed geological mapping projects, including in
classification systems or in RGB space, to improve
lithological discrimination and mapping.
The speed of magnetic unit mapping can be considerably
increased through reliance on edge detection filters. Further
improvements in mapping speed can be envisaged through
automated conversion of edge anomalies to vector files.
ACKNOWLEDGMENTS
The authors would like to acknowledge the New South Wales
Department of Mineral Resources for permission to use
aeromagnetic and geological data from the Goulburn
1:100 000 map sheet area and helpful comments by David
Robson during the project.
Figure 8. Comparison of ZS-Edge and ZS-Edgezone The authors wish to acknowledge Encom Technology for
enhancements over the Bindook Volcanic Complex permission to publish the results of research into the
Figure 9 shows standard RTP and Tilt transforms over the proprietary filters used in this paper. The 3D modelling was
carried out using Encom ModelVision Pro software, whilst
same area for reference.
processing and data visualisation were accomplished using
Geosoft OASIS montaj and Encom Geoscape.
REFERENCES
Blakely, R. J. and Simpson, R. W., 1986, Locating edges of
source bodies from magnetic or gravity anomalies,
Geophysics, 51, 1494-1498.
th
Buckingham, A.J, Dentith, M.C., and List, R.D, 2003, Miller, H.G., and Singh V., 1994, Potential field tilt — a new
Towards a system for content-based magnetic image retrieval: concept for location of potential field sources: Journal of
Exploration Geophysics, 34, 195-206. Applied Geophysics, 32, 213-217.
Cordell, Lindrith, and Grauch, V.J.S., 1985, Mapping Phillips, J.D., 1998, Processing and interpretation of
basement magnetization zones from aeromagnetic data in the aeromagnetic data for the Santa Cruz Basin–Patagonia
San Juan Basin, New Mexico pp.181-197. In Hinze, W.J., ed,. Mountains Area, South-Central Arizona: United States
The utility of regional gravity and magnetic maps: Society of Geological Survey Open-File Report 02-98.
Exploration Geophysicists,
Verduzco, B., Fairhead, J. D., Green, C. M., and MacKenzie,
Johnson A.J. et al., 2003, Goulburn 1:100 000 Sheet 8828 C., 2004, New insights into magnetic derivatives for structural
Geology Map, New South Wales Department of Mineral mapping: The Leading Edge, 23 (2), 116-119.
Resources.
th
GEOPHYSICS, VOL. 61, NO. 2 (MARCH-APRIL 1996); P. 394-408, 18 FIGS.
3-D inversion of magnetic data
Yaoguo Li* and Douglas W. Oldenburg*
inherent nonuniqueness. By Gauss’ theorem, if the field distri-

ABSTRACT bution is known only on a bounding surface, there are infinitely
many equivalent source distributions inside the boundary that
We present a method for inverting surface magnetic
can produce the known field. Any magnetic field measured on
data to recover 3-D susceptibility models. To allow the
the surface of the earth can be reproduced by an infinitesimally
maximum flexibility for the model to represent geologi-
thin zone of magnetic dipoles beneath the surface. From a
cally realistic structures, we discretize the 3-D model
mathematical perspective, this means there is no depth reso-
region into a set of rectangular cells, each having a
lution inherent in magnetic field data. A second source for
constant susceptibility. The number of cells is generally
far greater than the number of the data available, and nonuniqueness is the fact that magnetic observations are finite
thus we solve an underdetermined problem. Solutions in number and are inaccurate. If there exists one model that
are obtained by minimizing a global objective function reproduces the data, there are other models that will repro-
composed of the model objective function and data duce the data to the same degree of accuracy. The severity of
misfit. The algorithm can incorporate a priori informa- the nonuniqueness problem for magnetic data is illustrated in
tion into the model objective function by using one or Figures l-3. (The gray scale in all figures indicates suscepti-
more appropriate weighting functions. The model for bility in SI units for model sections and magnetic data in nT for
inversion can be either susceptibility or its logarithm. If data plots.) A 3-D dipping prism of uniform susceptibility in
susceptibility is chosen, a positivity constraint is imposed Figure 1 produces the surface magnetic field shown in Figure 2,
to reduce the nonuniqueness and to maintain physical which consists of 441 data. Slices of a 3-D susceptibility model
realizability. Our algorithm assumes that there is no that adequately reproduces the 441 data are shown in Figure 3.
remanent magnetization and that the magnetic data are That result, however, bears little resemblance to the true
produced by induced magnetization only. All minimiza- model. Susceptibility is concentrated near the surface and
tions are carried out with a subspace approach where displays zones of negative values. This mathematical model
only a small number of search vectors is used at each solution provides little information about the true structure
iteration. This obviates the need to solve a large system that is useful.
of equations directly, and hence earth models with many Faced with this extreme nonuniqueness, previous authors
cells can be solved on a deskside workstation. The have mainly taken two approaches in the inversion of magnetic
algorithm is tested on synthetic examples and on a field data. The first is parametric inversion, where the parameters of
data set. a few geometrically simple bodies are sought in a nonlinear
inversion and values are found by solving an overdetermined
problem. This methodology is suited for anomalies known to
INTRODUCTION be generated by simple causative bodies, but it requires a great
deal of a priori knowledge about the source expressed in the
Magnetic surveying has been used widely over the years, form of an initial parameterization, an initial guess for param-
resulting in a great amount of data with enormous area1 eter values, and limits on the susceptibility allowed (e.g.,
coverage. Magnetic data have been used for mapping geolog- Bhattacharyya, 1980; Zeyen and Pous, 1991). Nonuniqueness
ical structures, especially in the reconnaissance stage of explo- is not generally an issue because only a small subset of possible
ration, but when used in detailed prospecting, robust and models is considered due to the restrictive nature of the
efficient inversion algorithms must be used. However, a prin- inversion algorithm. A related, but unique, approach in Wang
cipal difficulty with the inversion of the potential data is the and Hansen (1990) assumes polyhedronal causative bodies and
Presented at the 63rd Annual International Meeting, Society of Exploration Geophysicists. Manuscript received by the Editor May 2, 1994; revised
manuscript received June 29, 1995.
*UBC-Geophysical Inversion Facility, Dept. of Geophysics and Astronomy, University of British Columbia, 129-2219 Main Mall, Vancouver, BC
V6T 1Z4, Canada.
© 1996 Society of Exploration Geophysicists. All rights reserved.
394
3-D Inversion of Magnetic Data 395
inverts for the position of the vertices of these bodies using the able information. Last and Kubik (1983) choose to minimize
spectrum of the magnetic data. The method is general in the total volume of the causative body so that the final model
principle but has difficulties both in constructing the causative is compact and structurally simple. Guillen and Menichetti
bodies from the recovered vertices and in obtaining the (1984) minimize the moment of inertia of the causative body
susceptibility distribution. with respect to the center of gravity or an axis passing through
In the second approach to inverting magnetic data, the earth it. Their inversion result is guided by the estimate of the central
is divided into a large number of cells of fixed size but of depth and dip of the causative body. These approaches have
unknown susceptibility. Nonuniqueness of solution is recog- merit but they are not flexible enough to handle problems we
nized and the algorithm produces a single model by minimizing are concerned with. This is especially true of methods that
an objective function of the model subject to fitting the data. attempt to collapse the anomalous susceptibility into a single
Green (1975) minimizes a weighted model norm with respect body; such a solution is rarely an adequate representation of
to a reference model, and this allows the interpreter to guide geologic structure.
the inversion by varying the weighting according to the avail- In our inversion approach, we first make a decision about
the variable in which the interpretation is to be made, that is,
whether susceptibility, log susceptibility, or some function of
susceptibility is sought. Next, we form a multicomponent
objective function that has the flexibility to generate different
types of models. The form of this objective function is such that
it can correct for the undesirable aspects of the mathematically
acceptable model in Figure 3, namely-the concentration of
susceptibility near the surface, the excessive structure, and the
existence of negative susceptibilities. Our objective function
incorporates an optional reference model so that the con-
structed model is close to that. It penalizes roughness in three
spatial directions, and it has a depth weighting designed to
distribute the susceptibility with depth. Additional 3-D weight-
ing functions in the objective function can be used to incorpo-
rate further information about the model. Such information
might be available from other geophysical surveys, geological
data, or the interpreter’s qualitative or quantitative under-
standing of the geologic structure and its relation to the
magnetic susceptibility. These 3-D weighting functions can also
be used to answer questions about the existence of suscepti-
bility features found from previous inversions. Negative sus-
ceptibilities are prevented by making a transformation of
FIG. 2. The total field anomaly produced by the slab model in

FIG. 1. Slices through a 3-D magnetic susceptibility model Figure 1. The inducing field has direction I = 75° and D = 25°
composed of a dipping slab in a nonsusceptible half-space. The and a strength of 50 000 nT. Uncorrelated Gaussian noise, with
slab is buried at a depth of 50 m and extends to 400-m depth a standard deviation of 2% of the datum magnitude plus 1 nT,
at a dip angle of 45° . The gray scale indicates the value of is added to the data. The gray scale indicates the magnetic
magnetic susceptibility in SI units. anomaly in nT.
396 Li and Oldenburg
variables and solving a nonlinear inverse problem. The numer- by inverting a field data set over a copper-gold porphyry
ical solution for the inversion is accomplished by dividing the deposit and a subsequent discussion.
earth into a large number of cells so that relatively complex
geologic bodies can be constructed. The computational diffi- INVERSION METHODOLOGY
culties often encountered in solving large matrix systems are
avoided by working explicitly with a generalized subspace
Each magnetic anomaly datum observed above the surface
algorithm.
can be evaluated by calculating the projection of the anoma-
The paper begins by outlining our inversion methodology
lous magnetic field onto a given direction. Let the source
and empirically estimating parameters for the depth weighting
region be divided into a set of rectangular cells by an orthog-
based upon synthetic inversion of single 3-D prisms. Data from
onal 3-D mesh and assume a constant magnetic susceptibility
two synthetic models are then inverted. The paper concludes
value within each cell. Further we assume that there is no
remanent magnetization and that the demagnetization effect is
negligible. Thus only the induced magnetization is considered.
This magnetization is uniform within each cell and is given by
the product of the susceptibility and the inducing geomagnetic
field The magnetic anomaly at a location on, or above, the
surface is related to the subsurface susceptibility by a linear
relationship
(1)
where d = is the data vector and
is the susceptibility in the cells. The matrix has as
elements which quantify the contribution of a unit suscep-
tibility in thejth cell to the ith datum. Closed form solutions for
were first presented in Bhattacharyya (1964) and later
simplified in Rao and Babu (1991) into a form more suitable
for fast computer implementation. The function is the
projection onto a given direction of the magnetic field that is
produced by a rectangular cell, so equation (1) is valid for
computing different magnetic anomalies. For example, a pro-
jection onto the vertical direction gives the vertical magnetic
anomaly while a projection onto the ambient geomagnetic field
direction yields the total magnetic anomaly. Thus, the method
presented here can be used to invert different types of mag-
netic data and in the following, we simply refer to them as the
magnetic data with the understanding that it is direction
specific.
Our inverse problem is formulated as an optimization
problem where an objective function of the model is minimized
subject to the constraints in equation (1). For magnetic
inversion, the first question that arises concerns definition of
the “model.” Two possible choices are and but any
function can, in principle, be used. In general, we prefer
to invert for since the field anomaly is directly proportional
to the susceptibility that varies on a linear scale. But depending
upon the expected dynamic range of susceptibility and the
physical interpretation attached to its value or variation, it may
be that is more desirable. To accommodate this, we
introduce the generic symbol m for the model with the
understanding that it might be or any monotonic
function Having defined a model, we next construct an
objective function, which when minimized, produces a model
that is geophysically interpretable. The details of the objective
function are problem dependent, but generally we need the
FIG. 3. The susceptibility model constructed by minimizing flexibility to be close to a reference model m 0 and also require
subject to fitting the data in Figure 2. As a mathematical that the model be relatively smooth in three spatial directions.
solution, this model provides little, if any, information about Here we adopt a right-handed Cartesian coordinate system
the subsurface susceptibility distribution. It effectively illus-
trates the nonuniqueness inherent to the inversion of static with x positive north and z positive down. Let the model
magnetic field data. objective function be
extra information is incorporated, the inversion derives a

model that not only fits the data, but more importantly, also
has a likelihood of representing the earth. From the viewpoint
of magnetic inversion, such an approach allows one to con-
struct a most-likely earth model that uses all available infor-
mation, and it can also be used to explore the nonuniqueness.
These two aspects form the foundation of a responsible
interpretation.
The kernels (values of for the surface magnetic data
decay with depth. It is for this reason that an inversion that
minimizes dv subject to fitting the
data will generate a susceptibility that is concentrated near the
surface. To counteract the geometric decay of the kernels and
(2)
to distribute susceptibility with depth, we introduce a weighting
of the form into and optionally
where functions ws, wx, wy , and wz are spatially dependent include it in The values of and z 0 are investigated in the
weighting functions while and are coefficients following section, but their choice essentially allows equal
that affect the relative importance of different components in chance for cells at different depths to be nonzero.
the objective function. Here, w ( z ) is a depth weighting The next step in setting up the inversion is to define a misfit
function. It is convenient to write equation (2) as measure. Here we use the 2-norm measure
refers to the first term in equation (2)
and refers collectively to the remaining three terms that (3)
involve variation of the model in three spatial directions. and we assume that the contaminating noise on the data is
The objective function in equation (2) has the flexibility of independent and Gaussian with zero mean. Specifying to
constructing many different models. The reference model m 0 be a diagonal matrix whose ith element is where is the
may be a general background model that is estimated from standard deviation of the ith datum, makes a chi-squared
previous investigations, or it could be the zero model. The variable distributed with N degrees of freedom. Accordingly
reference model would generally be included in but can be
= N provides a target misfit for the inversion.
removed if desired from any of the remaining terms. Often we The inverse problem is solved by finding a model m that
are more confident in specifying the value of the model at a minimizes and misfits the data by a predetermined amount.
particular point than in supplying an estimate of the gradient. This is accomplished by minimizing
The relative closeness of the final model to the reference where is our target misfit and is a Lagrangian
model at any location is controlled by the function w,. For multiplier. To perform a numerical solution, we first discretize
example, if the interpreter has high confidence in the reference the objective function in equation (2) using a finite-difference
model at a particular region, he or she can specify ws to have approximation according to the mesh defining the susceptibil-
increased amplitude there compared to other regions of the ity model. This yields
(4)
model. The weighting functions wx, wy , and wz can be designed where m and m0 are M- length vectors. The individual matrices
to enhance or attenuate structures in various regions in the are calculated straightforwardly once the
model domain. If geology suggests a rapid transition zone in model mesh and the weighting functions ws, wx, wy, wz, and
the model, then a decreased penalty for variation can be put w ( z ) are defined (see Appendix). The cumulative matrix
there, and the constructed model will exhibit higher gradients is then formed. For our formulation, the matrix is
provided that this feature does not contradict the data. There- never computed explicitly but we shall use it to derive our final
fore, the reference model and four 3-D weighting functions equations.
allow for the incorporation into the inversion of additional The inverse problem is solved by minimizing with an
information other than the magnetic data. The additional appropriate minimization technique. To reduce computation
information can be from previous knowledge about the sus- and to invoke positivity, we use a subspace methodology. In its
ceptibility, from other geophysical surveys, or from the inter- general form, the subspace technique allows the model param-
preter’s qualitative or quantitative understanding about the eter to be both positive and negative, and thus to ensure
geologic structure and its relation to susceptibility. When this positive susceptibility, we may need to invoke a transformation
of variables. Whether or not the transformation is required of that achieves the smallest misfit is taken. The search is
depends upon the relationship between mi and If usually accomplished by solving equation (10) a number of
so that interpretations are carried out in the logarithmic times using different values. Once the optimum value of is
domain, then no further transformation is necessary since found, the system is solved again to obtain the coefficients
will be positive irrespective of the sign of m i . However, if and the model perturbation. This iterative process is continued
and is a positive function, then a until the final expected data misfit is achieved and the model
transformation is required. All possibilities can be handled by objective function undergoes no significant decrease with
introducing a new parameter p, such that where successive iterations. Subspace vectors v i are generated mainly
f ( p ) is a monotonic function whose inverse and first-order from the gradients of the data and model objective functions.
derivative exist. This mapping is then incorporated directly into The data are grouped to form subobjective functions of misfit,
the subspace minimization process. and a steepest descent vector corresponding to each subobjec-
Let p ( n ) denote the parameter vector at the nth iteration and tive function is used as a subspace vector. Partitioning of the
denote the sought perturbation. Performing a Taylor ex- data can be formed by grouping data that are spatially close, or
pansion of the perturbed model objective function about the by grouping data such that each group has approximately the
point p ( n ) yields same contribution to the total data misfit. Both approaches
have worked well. The model objective function is partitioned
(5) and the gradient vector associated with each of the four
components in the model objective function provides addi-
where is a diagonal matrix with elements tional subspace vectors. In addition, a constant vector is always
included, and the selected subspace vectors are orthonormal-
(6) ized before being used in the search. More details on the
implementation of the subspace method for the linear inverse
problem can be found in Oldenburg and Li (1994).
A similar Taylor expansion applied to the misfit objective
The final item of practical importance is the specification of
functional yields
the mapping needed to ensure positivity of susceptibility. The
(7) positivity is required since we are dealing only with induced
magnetization, and the presence of negative susceptibility is
At each iteration we desire a perturbation that minimizes negligible in practical geophysical applications. Although our
equation (4) subject to generating a data misfit of formalism permits the minimization of m = the two most
where is the target misfit at the nth iteration. In the common situations are m = ln and m = When m =
subspace technique we represent the perturbation as ln we set p = m and hence the matrix in equation (10)
is the identity matrix. If m = we use the two-stage mapping
proposed in Oldenburg and Li (1994). It is composed of an
(8) exponential segment and a straight line. The two segments are
joined together such that the mapping and its first derivative
are both continuous. The mapping is given by
where the M-length vectors v i ( i = 1, q ) are as yet arbitrary.
Writing the objective function to be minimized in terms of the
coefficients yields
(11)
where p = p 1 is the transition point between exponential and

linear segments, and is selected to be small enough such
(9) that susceptibilities smaller than are not significantly differ-
Differentiating with respect to the coefficients a yields the final ent from zero when the final interpretation is carried out. Here,
equations and hence p l are chosen so that the ratio
does not exceed about two orders of magnitude. This prevents
the elements Fii from becoming too disparate. We note that
the ith row of is multiplied by Fii, and if this value is too
small, the ith row of is essentially annihilated and there will
be no possibility of adjusting the value of the ith cell. However,
if the ratio is too small, the flexibility in the mapping will be
restricted and this affects the convergence rate of the
algorithm. In the limit that the nonlinear mapping
(10) degenerates into a linear truncation and the inversion will not
converge. However, between the above two extremes, there is
We note that the matrix is q x q and therefore the system of a wide range of values for the ratio that can yield a good
equations is easily solved if q is small. At each iteration, we mapping. Based upon numerical experiments (Oldenburg and
search for a value of that yields the target misfit for that Li, 1994), we have chosen a value of 50.0 for this ratio for the
iteration. If the target misfit cannot be reached, then the value examples throughout this paper.
DEPTH WEIGHTING
It is well known that static magnetic data have no inherent The decay factor e-az causes the constructed model m,(z) to
depth resolution. For instance, when minimizing have structure concentrating toward the region of small z in the
structures tend to concentrate near the surface classic model construction that minimizes since the
regardless of the true depth of the causative bodies. In terms of model will be a linear combination of the kernels, i.e.,
model construction, this is a direct manifestation of the nature
of the kernels whose amplitudes rapidly diminish with depth.
The tendency to put structure at the surface can be overcome
(13)
by introducing a depth weighting to counteract this natural
decay. Intuitively, a weighting that approximately compensates
for the decay gives cells at different depths equal probability to This is shown in Figure 4a and 4b for two different models.
enter into the solution with a nonzero susceptibility. Before These models are constructed from five data ( i = 0,4) to which
proceeding with the details of the weighting function for noise has been added. It is apparent that the constructed
magnetic inversion, we illustrate the necessity, and effective- model is shifted toward small z where the amplitude of kernels
ness, of such a weighting function using a simple 1-D problem. is relatively large. One way to counteract the bias is to seek a
Consider a set of data d = ( d 1, , . . . , d N ) T generated from the solution in model space that is spanned by the nondecaying
equation portion of the kernels, in this case just the cosine functions.
The desired model would have the form
(12)
(14)
where the kernels are
FIG. 4. A 1-D example showing the use of a weighting function in the inversion procedures to counteract the natural decay in the
kernel function. In all panels the dashed line shows the true model. Panels (a) and (b) show, for the two different true models,
respectively, the model constructed using the original kernel functions with the decaying factor e-az. Notice the shift of the
recovered model towards the small z region. Panels (c) and (d) show the weighted models recovered by applying a weighting function
w ( z ) = e-az/2. They are better representations of the true model.
where are coefficients. Free from the influence from the in both directions, and 2% Gaussian noise is then added. The
decay factor, a model constructed from this set of basic observation is assumed to be 1 m above the surface and the
functions should have a better chance of having significantly inducing field has I = 75°, D = 25°. The region directly beneath
high values at depth. the data grid is taken as the model domain and discretized into
We accomplish this by finding an appropriate weighting 4000 cells (20 cells in each horizontal direction and 10 along
function w ( z ). We first rewrite the data equation as depth) of 50 m on a side.
Given the stated data parameters and model discretization,
the estimated value of z 0 in the depth weighting function is
( 1 5 ) 25 m. Figure 5 shows the comparison of the kernel beneath a
datum point and the function w 2( z ). This weighting function is
where are the weighted kernels and m w ( z ) is the used to invert surface data caused by the susceptible prism, and
weighted model. Then the inverse problem is solved by mini- the results of minimizing are shown in Figure 6. Each
mizing and the solution is given by panel in the figure is the cross-section through the center of the
model obtained by inverting the data set produced by a cube at
a different depth. They are rather good recoveries in terms of
(16) source depth, which is indicated by the superimposed outline
of the true body in each section.
In the above analysis we have established a practical way for
Dividing by the weighting function and substituting in
estimating an appropriate depth weighting function that dis-
yields
tributes the susceptibility more uniformly with depth. The
weighting is valid when the model objective function consists
only of In general, we like to include a penalty against
roughness and thereby produce a model that is smooth. To
(17) incorporate the above weighting scheme in the spatial varia-
tions, we make the following argument. Since minimizing
This equation can be made identical to equation (14) by
tends to provide a reasonable depth distribution, we wish only
choosing Carrying out the weighted inversion
to improve the model’s smoothness while maintaining the
for the above two data sets produces models shown in
depth characteristic. A conceptually consistent approach
Figures 4c and 4d. They are much better representations of
would be to apply the roughness measures to the weighted
true models.
model. We form a generic model objective function
This methodology is then applied to the inversion of surface
magnetic data by finding the appropriate weighting function
that counteracts the depth decay of the data kernels. There is
no distinct separable factor defining the decay in the kernel,
therefore we resort to an empirical estimate. Since the decay
rate depends upon the observation height as well as the size
and aspect ratios of the cells making up the 3-D model, such
estimates are expected to be problem dependent. Numerical
experiments indicate that the function of the form ( z + z 0)-3
closely approximates the kernel’s decay directly under the
observation point, given a correctly chosen value of z 0. This is
consistent with the fact that, to first order, a cubic-shaped cell
acts like a dipole source whose magnetic field decays by inverse
distance cubed. The value of z 0 can be obtained by matching
-3
the function ( z + z 0) with the kernel function beneath the
observation point. Thus, a reasonable candidate for the depth
weighting function is given by
(18)
The susceptibility model constructed by minimizing a model

objective function consisting of only i.e.,
(19)
subject to fitting the data should place the recovered anomaly FIG. 5. Comparison of the kernel function (solid) directly
at approximately the depth of the causative body. This hypoth- beneath the observation point with the estimated curve
(dashed) given by w 2( z ) = ( z + z 0)-3 with z 0 = 25 m. The
esis is tested by inverting surface data produced by a suscep- source cell is a cube of 50 m on a side. Here, z denotes the
tible cubic body at three different, depths. The cube is 200 m on depth to the center of the cell. Both curves are normalized for
a side. Data are calculated over a 21 X 21 grid of 50-m spacing comparison.
it is observed that straightforward inclusion of the depth weight-

(20) ing derived above into the 3-D weighting function in the form of
can yield reasonable results.
where the depth weighting is applied inside the derivatives of PRACTICAL ASPECTS OF DATA PREPARATION
the roughness components and the reference model m 0 can be
removed from any term if desired. This type of depth weighting The data used in the inversion are the residual data obtained
has proven to work satisfactorily on a number of synthetic by subtracting a regional field from the initial observation. The
examples and is the default choice in our algorithm. The inversion algorithm has been developed under the assumptions
examples to be presented in the following sections all use this that the surface magnetic anomaly is produced by the induced
depth weighting function. magnetization only and that there are no remanent magneti-
Before proceeding further, we remark that the above zation or demagnetization effects present. Incorrect removal of
weighting represents only one possibility. One could poten- regional field, or any deviation from the above assumptions, is
tially design a different weighting by incorporating the depth expected to cause a deterioration in the inversion results.
weighting in the usual 3-D weighting functions ws, wx, wy, wz. Furthermore, the susceptibility distribution is mathematically
Such an approach applies the depth weighting outside the deriv- represented by a piece-wise constant function defined on a
ative operators directly. However, the decay rate of the depth user-specified grid of cells. Magnetic sources, however, have a
weighting for each component will be different, and it is difficult to wide range of physical sizes. In some cases, source dimensions
establish a consistent rule for the choice of the different weight- will be significantly smaller than the size of cells in the
ings. In addition, the extra set of parameters required by such a mathematical model. If measurements are taken close to such
weighting scheme introduces more subjectivity into the inversion a source, the resulting anomaly will have a width that is
process. We have not explored this approach in detail; however, significantly smaller than that produced by a single cell in the
mathematical model and this may produce artifacts. We ame-
liorate this problem by inverting data that have been upward
continued to a height approximately equal to the width of the
surface cells in the model. We arrive at this conclusion from a
numerical experiment. We first generate the magnetic field
from a small localized surface source that is assumed to be a
cube of width At each height h above the surface, a
one-parameter inverse problem is carried out to find a uniform
susceptibility of a large surface cube that has a width of L and
shares a common horizontal center with the small cube. If HL
is the field of the large cell that best reproduces then the
misfit functional,
(21)
FIG. 6. Cross-sections through the center of the recovered

model for a cube at a central depth of 150,200, and 250 m. The
cube is 200 m on a side. The inversion uses the weighting FIG. 7. The misfit between magnetic field as a result of a small
function derived from the kernel decay estimated in Figure 5. cubic source and the field as a result of a larger cubic model
The true position of the cube is outlined in each cross-section. cell having a best fitting susceptibility. The numbers indicate
As the true source depth increases and, as a result, the the ratio of the cell width. The misfit is plotted as a function of
high-frequency content in the data decreases, the recovered the observation height normalized by the width of the model
model becomes increasingly smooth and attains a smaller cell. Note that the misfit decreases rapidly until the height is
amplitude. However, the depth of the recovered model is close approximately equal to the width of the model cell, and that it
to the true value. changes slowly thereafter.
can be computed, where is the surface area of the data map. The inversion uses 54 subspace vectors and achieves the
Figure 7 shows the misfit function r ( h ) for trial values of = expected misfit in 13 iterations. The recovered model is shown
0.1, 0.2, 0.4. We note that r ( h ) decreases rapidly until h L, in Figure 9. It is smoother, has a slightly lower amplitude than
and that it changes slowly thereafter. Since the above misfit the model in Figure 8, and it recovers the essential features of
analysis is a worst case scenario because the contaminating the true model such as the depth and dip angle.
body is located at the surface, the suggestion of upward It is observed, in this example and in other synthetic and
continuing the data to a height approximately equal to the field test examples, that minimizing either the first term in the
width of surface cells may be somewhat conservative, and model objective function in equation (20), or using all
inversionists may want to vary this. However, in many field four terms, generates models that are reasonable representa-
surveys, magnetically susceptible small bodies exist close to the tions of the true structure. In the absence of prior information,
surface and hence upward continuing the data prior to inver- both models can provide useful information about the subsur-
sion is prudent. face susceptibility distribution. However, the model minimiz-
ing can be obtained at less computational cost. Further-
SYNTHETIC EXAMPLES
As the first example, we invert the total field anomaly data

given in the Introduction. The model consists of a 3-D dipping
slab buried in a nonsusceptible half-space (slab model).
Figure 1 shows three slices through the slab model. The
susceptibility of the slab is 0.06 (SI unit). Under an inducing
field with a strength of 50 000 nT and a direction at I = 75° and
D = 25°, the slab model produces the surface total magnetic
anomaly shown in Figure 2, which consists of 441 data over a
21 X 21 grid of 50-m spacing. The data have independent
Gaussian noise added whose standard deviation is equal to 2%
of the accurate datum magnitude plus 1 nT. We invert these
441 noise-contaminated data to recover the susceptibility of an
earth model parametrized by 4000 cells of 50 m on a side (20
cells in each horizontal direction and 10 in depth).
The data are partitioned into 49 groups to provide 49 search
vectors for the subspace algorithm. In addition, each compo-
nent in the model objective function provides one basis vector,
and a constant vector is included. For the depth weighting, the
value of z 0 is estimated as 25 m. The additional 3-D weightings
in the objective function are all set to unity. The reference
susceptibility model is set to zero. For the nonlinear mapping,
we choose = 0.0002 and = 0.01.
First, we invert the data by minimizing an objective function
composed only of the and using m = as the model
parameter. A total of 51 subspace vectors are used at each
iteration. The inversion reaches the expected misfit in 13
iterations but a few extra iterations are performed in an
attempt to further reduce the value of the model objective
function while keeping the misfit at the target value. By
iteration 18, the objective function is decreasing by less than
1% per iteration, and the process is terminated. The con-
structed susceptibility model is shown in Figure 8 and can be
compared with the true model in Figure 1. The tabular shape
of the anomaly and its dipping structure are clear, and the
depth extent is reasonably recovered. The amplitude of the
recovered model is slightly higher than the true value, but the
dip angle inferred from the recovered model is close to the true
value. We point out that the model sections should be plotted
using gray shading for each cell to reflect the piece-wise
constant nature of the model. However, when the model has
only a small number of cells in each spatial direction, the
structural trends are more readily shown when contours are
used. For this reason, we have contoured the model sections. FIG . 8. Model obtained from inverting the data shown in
Figure 2 by minimizing only which has the depth weight-
Next, the same data are inverted using a model objective ing applied. This is to be compared with the true model in
function that includes penalty terms on spatial roughness, Figure 1. The major features in the true model, such as dip
The depth weighting is applied to all terms, as in equation (20). angle and depth exte nt, are evident in the recovered model.
more, the depth weighting in this case is rather well supported as an initial model. The available prior information can be
by mathematical analysis whereas it is an argued extension for incorporated into the second inversion by forming a reference
the three roughness components. Therefore, a reasonable model and 3-D weighting functions, ws, wx, wy, wz.
approach to inverting field data might be a two-step process. We now invert the same data by using m = ln as the
The data can be inverted first by minimizing and the model. It is not possible to incorporate a zero susceptibility as
resultant model may be used in the interpretation as a prelim- the reference model, so we minimize an objective function
inary result. If there are interesting features present and if one consisting of with the reference model removed. The same
desires to refine the model by incorporating prior information depth weighting is applied to all terms of Since = em,
to enhance or attenuate the structural complexity in different the positivity of the susceptibility is ensured without invoking
regions, a second inversion can be carried out using an the transformation of variables. The result is shown in
objective function consisting of both and The model Figure 10a. This is a cross-section at x = 500 m and plotted on
obtained by minimizing can then be used in this inversion a logarithmic scale in accordance with the model used in the
inversion. The inverted susceptibility shows the presence of the
dipping anomaly as a broad region of high susceptibility.
However, the interpretation based upon such a model can be
complicated by the variations of susceptibility that are small
and have little effect on the surface data. We have replotted
the cross-section on a linear scale in Figure 10b and the
anomalous region is now delineated more clearly. Its top
portion indicates the tabular body and defines the depth to the
FIG. 10. The model obtained from inverting the data shown in
Figure 2 by using m = ln as the model and minimizing
FIG. 9. The model derived from inverting the slab model data with the reference model removed. The inverted logarithmic
in Figure 2 by minimizing the model objective function having susceptibility in cross-section at x = 500 m is shown in (a) and
both and The same depth weighting is used. This it is replotted on a linear scale in (b). As a comparison, the
model appears to be smoother and has a smaller amplitude result obtained by using m = and the same objective
than that in Figure 8. function is shown in (c).
top and dipping angle. The anomaly terminates at a shallower scale as demonstrated here, and since the magnetic data are
depth than the true model and has a nearly horizontal exten- linearly related to the susceptibility, we generally prefer to work
sion to the left. As an exact comparison, Figure 10c is the with the susceptibility K as the model in the inversion.
susceptibility model obtained by minimizing but using As the second example we invert the total field anomaly data .
m = as model and invoking the positivity. This is a smoother produced by a slightly more complicated model and with two
model and exhibits more gradual changes in the susceptibility. different inducing field directions. The true model is shown in
It has a slightly deeper extent than the model in Figure 10b. With Figure 11 in the same format as before. It is a dipping slab
the exception of details toward the bottom, however, both models having its top and bottom portions offset to simulate the result
provide almost the same information about the anomalous sus- of a normal faulting. The faulted slab strikes north. The data
ceptibility region. It might be concluded that inversion using from this model, when the inducing field has a direction of I =
either linear or logarithmic susceptibility is viable for practical 45° and D = 45°, are shown in Figure 12. Again Gaussian noise
applications. However, we note that the presentation in has been added to the data. The inversion minimizes an
Figure 10b is inconsistent with the model used in the inversion. objective function consisting of and that have the
Since the inverted susceptibility is easier to interpret on a linear same depth weighting and nonlinear mapping as used to
produce the results in Figure 9. Figure 13 displays the recov-
ered model in three slices. It shows two distinct anomalous
regions of susceptibility that correspond to those in the true
model. The dipping structure is evident from the top block. On
plan view, the strike direction and the strike length of the
anomaly are also well recovered.
When the inducing field direction is I = 0° and D = 45°, the
surface anomaly with added Gaussian noise is that shown in
Figure 14. Carrying out the inversion using an identical model
objective function generates the model shown in Figure 15. It
is similar to the model shown in Figure 13, which is recovered
under an inducing field at 45° inclination. Again, the two
separate blocks, the dipping direction, and the length and
direction of the strike, are all reasonably recovered. This is a
positive result in that, although the surface anomalies have
very different expressions under different inducing field direc-
tions, the inversion algorithm is able to consistently recover the
source structure. Moreover, the algorithm had no difficulty in
inverting data generated from an inducing field having zero
inclination; such data often pose problems in interpretations
that include a reduction to pole.
We emphasize that positivity has played a pivotal role in all
the inversions. Magnetic data generally have regions of nega-
FIG . 11. The second synthetic test example. The top and FIG . 12. The surface total field anomaly produced by the
bottom portions of the anomalous susceptibility are offset to faulted slab in Figure 11, under an inducing field at I = 45° and
simulate a norm fault structure. It also has a large strike length D. = 45°. Uncorrelated Gaussian noise is again added to the
in the north direction. data.
tive values that result from dipping bodies or inclined inducing FIELD EXAMPLE
field, or both. Without positivity, the constructed susceptibility
is often negative and the dipping bodies appear more vertical. As the final example, we invert field data taken over a
Recovery of correct dip and, to some extent, depth to the top copper-gold porphyry deposit at Mt. Milligan in central British
of the anomalous body, are often the result of invoking Columbia. The host rocks for the deposit are early Mesozoic
positivity. Once the positivity is imposed, it is no longer true volcanic and sedimentary rocks and contain intrusive monzo-
that an equivalent stratum that reproduces the data exists at nitic rocks that have accessory magnetite. Porphyry-style alter-
any depth. Therefore, cells of anomalous susceptibility cannot ation and copper-gold mineralization are contemporaneous
be placed arbitrarily close to the surface, and no equivalent with the intrusive events. The copper and gold are known to be
source can be constructed with negative susceptibilities. This concentrated in the potassic alteration assemblage, which is
restricts the class of admissible models and, consequently, mainly around the contact of the monzonite intrusions and
reduces the nonuniqueness. may extend outward and into fractured volcanic rocks. Among
other minerals, magnetite is one of the strong indicators of the
potassic alteration. Ground magnetic data are acquired in the
region at 12.5-m spacing along lines in the east direction and
spaced 50 m apart. Our study of the data set has focused on a
1.2 km x 1 km area, which covers a large monzonite body
known as the MBX stock and contains a reasonably isolated set
of magnetic anomalies. Fairly detailed information about the
geology is available through a major drilling program, but no
susceptibility logs were available.
Magnetic data from a larger area were first upward contin-
ued to 20 m. A regional field was then defined and removed
from the upward continued data. The continuation operation
suppresses the noise in the data and also facilitates the
discretization of the topographic surface for the model so that
all observation points remain above the discretized surface.
Although the original data were collected at 12.5-m spacing,
we use the data at 25-m spacing. This yields 1029 data points at
varying elevations. Figure 16 shows the data contoured accord-
ing to their horizontal locations. The direction of the inducing
field is I = 75° and D = 25.73°. Several major magnetic highs
are observed in the map. However, the influence of anomalies
adjacent to the map is also visible along the edges. We choose
a model domain that is horizontally larger than the data area,
coincides at the top with the highest point on the topographic
surface, and extends to 450-m depth. The model is discretized
horizontally at a 25-m interval beneath the area of data. In the
FIG. 13. The susceptibility model recovered from the data

shown in Figure 12. It is seen that both the top and bottom FIG. 14. The surface total field anomaly produced by the
block of the true model are recovered and the strike direction faulted slab in Figure 11 under an inducing field at I = 0° and
and length are well defined. D = 45°. Uncorrelated Gaussian noise is added to the data.
vertical direction, the first 100 m is divided at a 12.5-m interval hundred subspace vectors generated by dividing the data map
so that the surface can be adequately discretized onto the into small subareas are used in the inversion. We use a
model mesh. Below the depth of 100 m, an interval of 25 m is nonlinear mapping with = 0.0002 and = 0.02. The
used. This results in a mesh with 52 x 44 x 22 cells. Once the recovered model is shown in Figure 17 as one plan-section and
mesh is defined, the topography is discretized onto it. The three cross-sections. From the plan-section, two concentrated
43 428 cells below this surface define the susceptibility model, susceptibility highs are observed in the central region. Sur-
and the inverse problem is therefore formalized by inverting rounding them are three linear anomalies trending northeast.
1029 data to recover the susceptibilities in those cells. The In the cross-sections, the major anomalies are seen at moder-
depth weighting is referenced to the top of the model domain. ate depths but there is considerable variation in the depth to
Each datum is assumed to have an error whose standard the top. There are also smaller anomalies extending to the
deviation is equal to 5% of its magnitude plus 10 nT. The error surface. In general, there are more detailed structures near the
estimate includes not only the repeatability of the instrument surface and the model becomes increasingly smooth at greater
reading but also the geological noise and errors introduced by depths. As required by the objective function, there is no
the inaccurate recording position and by separating the anom- excessive structure associated with each unit of high suscepti-
alous field from the initial total field measurements. One bility region. Comparison with drill logs indicates that the
recovered magnetic susceptibility highs are mostly associated
with the monzonite intrusions and with faults or fracture zones.
Figure 18 compares the recovered susceptibility model with the
geology (Cam DeLong, personal communication) in the cross-
section at x = 600 m. The large susceptibility high is spatially
well-correlated with the MBX stock and reflects the initial
magnetite content in the intrusion. Two smaller susceptibility
highs are present east of the stock. The high at y = 650 m
coincides with the boundary of stock and porous trachytic units
while the high at y = 900 m coincides with the upper portion
of the Rainbow dyke. These are locations of the most intensive
potassic alterations and the susceptibility highs are indicative.
of the magnetite produced by the alteration process. Over all,
this is a rather encouraging result.
CONCLUSION
We have developed an algorithm to invert surface magnetic

data for general 3-D susceptibility distributions. Although we
have illustrated the algorithm using examples on the scale
pertinent to mining applications, the method is general and
applicable to problems on different scales ranging from envi-
ronmental to regional investigations. To overcome the inher-
ent nonuniqueness, we obtain the solution by minimizing a
FIG. 16. The extracted total field anomaly from ground mag-
FIG. 15. The susceptibility model recovered from the data netic data at Mt. Milligan Copper-gold porphyry deposit. The
shown in Figure 14. This model is similar to that shown in data are contoured according to their horizontal locations in
Figure 13. this map, although thev are at different elevations.
specific objective function of the model. Our model objective
function has the ability to incorporate prior information into the
inversion via a reference model and 3-D weighting functions. A
crucial feature of the objective function is a depth weighting
function that counteracts the natural decay of the kernel func-
tions. The parameters of the depth weighting depend upon the
discretization of the model but are easily calculated. The minimi-
zation is carried out using a subspace technique that reduces the
computational effort and allows the positivity constraint of sus-
ceptibility to be incorporated. Both susceptibility and logarithmic
susceptibility can potentially be used as the model in the inver-
sion. Since the data are linearly related to susceptibility, and since
usually absolute values of susceptibility are required for interpre-
tation rather than relative values, especially in regions of very low
FIG. 18. Comparison of the recovered susceptibility model in a
susceptibility, we have generally chosen to work with susceptibil- cross-section ( x = 600) with the geology for the Mt. Milligan
ity. To suppress the noise from small magnetic bodies near the deposit. The susceptibility high within the MBX stock reflects
surface, we recommend in general that the data be upward the initial magnetite in the intrusive while the susceptibility
continued to a height comparable with the width of the surface highs near the Rainbow dyke are related to the magnetite
produced by potassic alteration.
cell before inversion.
Applications of our inversion to synthetic data sets have

produced models representative of the true structures and
demonstrated the ability of the algorithm to construct consis-
tent models at different magnetic latitudes. Inversion of field
data has produced a susceptibility model that is consistent with
the known geology and mineralization information. These
results represent an encouraging conclusion: although the
inversion of magnetic data seems impossibly nonunique when
one has a large number of cells, inversions using a properly
designed model objective function can produce susceptibility
distributions that yield meaningful geologic information.
ACKNOWLEDGMENTS
This work was supported by an NSERC IOR grant and an

industry consortium “Joint and Cooperative Inversion of Geo-
physical and Geological Data.” Participating companies are
Placer Dome, BHP Minerals, Noranda Exploration, Cominco
Exploration, Falconbridge, INCO Exploration & Technical
Services, H u d s o n B a y E x p l o r a t i o n a n d D e v e l o p m e n t ,
Kennecott Exploration Company, Newmont Gold Company,
Western Mining Corporation, and CRA Exploration Pty.
REFERENCES
Bhattacharyya, B. K., 1964, Magnetic anomalies due to prism-shaped

bodies with arbitrary magnetization: Geophysics, 29, 5 17-53 1.
1980, A generalized multibody model for inversion of magnetic
anomalies: Geophysics, 45, 255-270.
Green, W. R., 1975, Inversion of gravity profiles by use of a Backus-
Gilbert approach: Geophysics, 40, 763-772.
Guillen, A., and Menichetti, V., 1984, Gravity and magnetic inversion
with minimization of a specific functional: Geophysics, 49, 1354-
1360.
Last, B. J., and Kubik, K., 1983, Compact gravity inversion: Geophys-
ics, 48, 713-721.
Oldenburg, D. W., and Li, Y., 1994, Subspace linear inversion meth-
ods, Inverse Problems, 10, 915-935.
Rao, D. B., and Babu, N. R., 1991, A rapid methods for three-
dimensional modeling of magnetic anomalies: Geophysics, 56,
1729-1737.
FIG. 17. The recovered susceptibility model shown in one Wang, X., and Hansen, R. O., 1990, Inversion for magnetic anomalies
plan-section and three cross-sections. The plan-section is at the of arbitrary three-dimensional bodies: Geophysics, 55, 1321-1326.
depth of 150 m and the three cross-sections are at x = 600,500, Zeyen, H., and Pous, J., 1991, A new 3-D inversion algorithm for
and 400 m, respectively. magnetic total field anomalies: Geophys. J. Int., 104, 583-591.
APPENDIX
MODEL OBJECTIVE FUNCTION
Our inversion method uses a model objective function of the Each component matrix can be written as the product of three
form individual matrices and one oefficient. That is,
(A-3)
where are diagonal matrices representing the spatially

dependent 3-D weighting functions, are the finite-difference
operators for each component, and is a diagonal matrix
representing the discretized form of depth weighting function
The elements of are given by They are defined

(A-1) over each cell for and over each interface between
adjacent cells in the respective directions for and
The numerical evaluation of this functional is carried out by has elements on its diagonal, where Ax, Ay,
introducing the model mesh and evaluating all terms using a and are the cell width. The matrix has two elements
finite-difference approximation. The discretized model objec- in each row, where is the distance between
tive function has the form the centers of cells adjacent in the x - direction. Similarly,
and have elements and respec-
tively, where and are the distances between centers of
adjacent cells in the y- and z - directions. Once the mesh is
defined and all weighting functions, ws, wx, wy , wz, and w ( z )
are chosen, equation (A-3) is evaluated straightforwardly
(A-2) and is formed.
Section 2: ZTEM Processing

Layered-earth 1-1,000 Ohm-m on 10,000 Ohm-m halfspace
Regolith 10-1,000 Ohm-m on 1,000 Ohm-m halfspace
Regolith wedge 10 Ohm-m on 1,000 Ohm-m halfspace
Saltlake 5 S/m (0.2 Ohm-m) in 10 Ohm-m regolith on 1,000 Ohm-m halfspace
Quad-spaces 10,000 / 10 Ohm-m
Quad-spaces 10,000 / 10 Ohm-m w/ 1000 S dyke
Tabular 1 S/m conductor (200m x infinite x 200m) in 10,000 Ohm-m half-space
200 S shear zone in 1,000 Ohm-m half-space
200 S shear zone under 10 Ohm-m (10 S) overburden in 1,000 Ohm-m half-space
Quad-spaces 10,000 / 10 Ohm-m w/ 1000 S dyke
Tabular 1 S/m conductor (200m x infinite x 200m) in 10,000 Ohm-m half-space
200 S shear zone in 1,000 Ohm-m half-space
200 S shear zone under 10 Ohm-m (10 S) overburden in 1,000 Ohm-m half-space
Section 3: ZTEM Processing-Y component MultiPlots

As discussed in Section 4, the ZTEM system records a Tzx and Tzy component. The standard
product for 2D Occam processing is done on the along-line component Tzx which has the maxi-
mum sensitivity to geological features normal to the flight direction. By sampling across the normal
survey lines, the Tzy component can be then reconstructed into the equivalent of what the Tzx
would be for lines normal to the standard survey lines (i.e. in the direction of the Tie Lines). This
sampling is not as high a density as the normal along line sampling for ZTEM, but due to the gen-
erally long wavelengths ZTEM produces the recovered model is very similar.
An example of this processing is shown below in Figure S3-1 from the Forrestania property, WA.
Here there were two orthogonal surveys and the conductor of interest was mapped on the N-S
lines (L1075). The conductor was not visible on the E-W line (Tzx processed) but when the Tzy
data was used, the conductor was again imaged.
Figure S3-1

The two blocks from the Magma survey are shown in Figure S3-2; the upper image being that of
the ‘as-delivered’ from Geotech and the lower image shows the Tzy processing synthetic lines.
Figure S3-2

Processing of Thunder Bay North ZTEM data for Magma Metals
Limited
2D Inversion
Inversion Algorithm
The recorded ZTEM data were inverted using a 2D MT algorithm, developed by Constable and Wannamaker (deGroot-
Hedlin and Constable, 1990; Wannamaker at al., 1987; deLugao and Wannamaker, 1996). The transverse electric (TE)
response computed by the finite-element algorithm is used to model the along-line tipper data Tzx, taking into account that the
vertical component is measured in the air and the horizontal component is recorded sitting on uniform ground. The responses of all
frequencies (30 – 720 Hz) were included in the inversion.
Model specifications
The model mesh for one line is shown in the following figure, with a close-up shown below. The topography is indicated by a solid
black line.
The model cells are 20 m wide. The number of vertical cells is a function of the flight line topography and bird elevation, hence, is
different for each line. Since the algorithm takes into account the topography along each line, air cells are an integral part of the
finite-element model. 25-m-thick cells are modelled between the topographic low and the highest bird elevation. Beneath the
topographic low, cell thicknesses increase with increasing depth, as listed in the following table:
Layer depth to thickness Layer depth to thickness Layer depth to thickness

bottom (m) (m) cont. bottom (m) (m) cont. bottom (m) (m)
1 25 25 15 776 71 29 2143 127
2 50 25 16 850 74 30 2276 133
3 83 33 17 927 77 31 2426 150
4 123 40 18 1007 80 32 2606 180
5 172 49 19 1090 83 33 2826 220
6 223 51 20 1177 87 34 3086 260
7 276 53 21 1267 90 35 3386 300
8 331 55 22 1361 94 36 3726 340
9 388 57 23 1459 98 37 4146 420
10 447 59 24 1561 102 38 4566 420
11 508 61 25 1668 107 39 5106 540
12 571 63 26 1779 111 40 5646 540
13 637 66 27 1895 116
14 705 68 28 2016 121
Padding was added in symmetric fashion to the left and right of each section. The padding cells to the right have the following
widths (m): 60 60 80 200 200 200 400 400 400 800 800 800 2700 8100 24300
A 1000 Ohm-m halfspace was used as a start model, and a reference resistivity of 10,000 Ohm was used for depths beneath 4566 m.
The inversion results from all lines were merged into a survey GDB, containing modelled conductivities as a function of elevation,
as listed in the following table.
Layer Layer top Layer bottom Layer Layer top Layer bottom Layer Layer top Layer bottom
Z1 (m ASL) Z2 (m ASL) cont. Z1 (m ASL) Z2 (m ASL) cont. Z1 (m ASL) Z2 (m ASL)
1 525 500 15 -58 -121 29 -1218 -1329
2 500 475 16 -121 -187 30 -1329 -1445
3 475 450 17 -187 -255 31 -1445 -1566
4 450 425 18 -255 -326 32 -1566 -1693
5 425 400 19 -326 -400 33 -1693 -1826
6 400 367 20 -400 -477 34 -1826 -1976
7 367 327 21 -477 -557 35 -1976 -2156
8 327 278 22 -557 -640 36 -2156 -2376
9 278 227 23 -640 -727 37 -2376 -2636
10 227 174 24 -727 -817 38 -2636 -2936
11 174 119 25 -817 -911 39 -2936 -3276
12 119 62 26 -911 -1009 40 -3276 -4116
13 62 3 27 -1009 -1111 41 -4116 -5196
14 3 -58 28 -1111 -1218
Noise and target fit
The noise level, specified for the inversion, was the maximum of 1.5% of the ZTEM response (relative noise level) and 0.15%
(absolute noise level). The noise level for the high-frequency responses at 360 and 720 Hz were set twice and five times those
values, respectively. These noise estimates resulted in a stringent target data fit that was never met by the inversion, due to the
actual noise level exceeding the noise estimates and the occasional inappropriateness of the 2D algorithm in a 3D conductivity
scenario. To avoid overfitting the data, while optimizing data fit in a variable noise environment, the final result was chosen from
inversion results a few iterations below the final iteration.
Apparent Conductivity and phase
The derivation of apparent conductivity and phase from VLF data is discussed by Becken and Pedersen (2003). The method has
been applied to the ZTEM data, making joint use of the Tzx and Tzy tipper data.
References
Becken, M., and Pedersen, L.B., 2003, Transformation of VLF anomaly maps into apparent resistivity and phase: Geophysics 68,
497-505.
DeGroot-Hedlin, C. and Constable, S., 1990, Occam’s inversion to generate smooth two-dimensional models from magnetotelluric
data: Geophysics 55, 1613-1624.
De Lugao, P.P., and Wannamaker, P., 1996, Calculating the two-dimensional magnetotelluric Jacobian in finite elements using
reciprocity, Geophys. J. Int., 127, 806-810.
Wannamaker, P.E., Stodt, J.A., and Rijo, L., 1987, A stable finite-element solution for two-dimensional magnetotelluric modeling:
Geophysical Journal of the Royal Astronomical Society, 88, 277-296.
APPENDIX B – Comparison of Condor and Geotech 2D Processing

APPENDIX C – ARCHIVE DVD

ZTEM탐사보고서 Northwestern Ontario지역

Uploaded by

Copyright:

Available Formats

You might also like

ZTEM탐사보고서 Northwestern Ontario지역

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ZTEM탐사보고서 Northwestern Ontario지역

Uploaded by

Copyright:

Available Formats

ZTEM Airborne Geophysics

Thunder Bay North Project

Panoramic PGMs (Canada) Limited

Thunder Bay North Block

Survey flown May – June 2011

Thunder Bay North Block

The processed survey results are presented as the following maps:

• Total Magnetic Intensity

Figure 1 -Property Location

1.2 Survey Location

The block is located approximately 40 kilometres northeast of Thunder Bay, as shown in

Figure 3 - Google Earth image of the survey

2.1 Survey Area

Table 1 - Survey Specifications

Traverse Line Area Planned Actual 1

Traverse: 200 549 565 N 60° E / N 240° E L1000 – L1410

TOTAL 115 629 646

2.2 Survey Operations

Date Flight # Block Crew location Comments

24-May-2011 Thunder Bay, ON System assembly

2.4 Aircraft and Equipment

2.4.1 Survey Aircraft

2.4.2 Airborne Receiver

Figure 4 - ZTEM System Configuration

2.4.3 Base Station Receiver

Figure 5 - ZTEM base station receiver coils.

2.4.4 Airborne magnetometer

2.4.5 Radar Altimeter

2.4.6 GPS Navigation System

2.4.7 Digital Acquisition System

Table 3 - Acquisition and Processing Sampling Rates

Magnetometer 0.1 sec 0.4 sec

GPS Position 0.2 sec 0.4 sec

Radar Altimeter 0.2 sec 0.4 sec

ZTEM Base station 0.0005 sec __

2.4.8 Mag Base Station

A combined magnetometer/GPS base station was utilized on this project. A Geometrics

The following Geotech Ltd. personnel were involved in the project.

Project Manager: Adrian Sarmasag (Office)

Data QC: Emilio Schein (Office)

Crew chief: Colin Lennox

Operator: Les Moschuk

Pilot: Andres Par

Mechanical Engineer: Chris Ward

Preliminary Data Processing: Emilio Schein

Final Data Processing: Lily Manoukian

Final Data QC: Ali Latrous

2D Inversions: Ali Latrous

Reporting/Mapping: Wendy Acorn

4.1 Flight Path

4.2 In-field Processing and Quality Control

The In-Field quality control can be separated into several phases:

4.3 GPS Processing

4.4.1 Preliminary Processing

4.4.2 Geosoft Processing

4.4.4 ZTEM Profile Sign Convention

ZTEM In-Phase Z/X (In-line) Tipper

ZTEM Quadrature Z/X (In-line) Tipper

Tx = ([HzHx] [HyHy] – [HzHy] [HyHx]) / ([HxHx] [HyHy] – [HxHy] [HyHx]), (2)

Ty = ([HzHy] [HxHx] – [HzHx] [HxHy]) / ([HxHx] [HyHy] – [HxHy] [HyHx] . (3)

Tx = ([HzRx] [HyRy] – [HzRy] [HyRx]) / ([HxRx] [HyRy] – [HxRy] [HyRx]), (4)