Académique Documents
Professionnel Documents
Culture Documents
AusSpec
AusSpec Workshop – 2010 International
Main Aims of Workshop
Review of Principles of spectral data analysis
Practical aspects of sample measurement,
some tips and considerations
Spectral data interpretation and handling
Case study exercises and examples
More data handling methods for LARGE data
sets
Selected case studies
Project data sets
AusSpec
AusSpec Workshop – 2010 International
Firstly: Important questions
What are our requirements for exploration and mining?
AusSpec
AusSpec Workshop – 2010 International
What are our requirements?
Build alteration models
AusSpec
AusSpec Workshop – 2010 International
Why map alteration minerals?
The chemistry associations of alteration minerals are a response
to their conditions of formation - minerals and mineral
assemblages provide information on hydrothermal conditions.
AusSpec
AusSpec Workshop – 2010 International
Mineral assemblage
Alteration minerals
provide vital
information on the
fluid history of a
system
Patterns can be
used to vector
toward prospective
parts of the system
AusSpec
AusSpec Workshop – 2010 International
FIRSTLY SOME BACKGROUND....
AusSpec
AusSpec Workshop – 2010 International
Background to vis-IR Spectroscopy
Electromagnetic spectrum
Remote sensing
Remote sensing – Operational Operational
AusSpec
AusSpec Workshop – 2010 International
Background to vis-IR Spectroscopy
Short Wave Infrared
Vibrational features
Water, OH and lattice
bonds (Al, Fe, Mg)
CO3 and NH4
“Harmonics” of
thermal features
Diagnostic signatures
for different mineral
groups
AusSpec
AusSpec Workshop – 2010 International
Background to vis-IR Spectroscopy
Visible-Near Infrared
Electronic transitions
Sub atomic
Fe, Cu, Mn, Cr
Leads to mineral colour
Broad features
Crystal field
Charge transfer
AusSpec
AusSpec Workshop – 2010 International
SWIR 1300-2500nm
SWIR identification: Minerals with OH, water or carbonate
AusSpec
AusSpec Workshop – 2010 International
Vis-NIR 380-1300nm
Vis-NIR Identification: Fe minerals, Cu and Mn minerals
AusSpec
AusSpec Workshop – 2010 International
Ferric features: Hematite vs goethite
Hematite ferric features ~650nm (shoulder), ~860nm
2nd: tsg_core_import_with_Fe_oxides.I.184 Goethite ferric features 650nm, ~930nm
Aux
0.197
0.271
0.345
0.419
near 860nm, with a
1.4
0.493
0.567
shoulder near 650nm.
0.641
1.2
0.714
Norm. Refl. (Aux. colour: Norm. Refl.)
0.862
0.936
1.01
1.084
1.158
often is a feature near
0.8
1.232
1.306
650nm.
1.38
0.6
1.453
1.601
0.4
1.675
AusSpec
AusSpec Workshop – 2010 International
Ferrous features
Chlorite: Ferric and ferrous
epidote ferrous features
2nd: tsg_core_import_with_Fe_oxides.I.373 chlorite ferrous features
Aux
1.4
0.197
0.419
0.493
extending to 1100nm
1.2
0.567
0.641
0.714
Strong effect on the slope
1.1
Norm. Refl. (Aux. colour: Norm. Refl.)
to the SWIR
0.788
0.862
0.936
1
1.01
1.084
0.9
1.158
1.232
1.306
0.8
1.38
1.453
1.527
0.7
1.601
1.675
0.6
NULL
Tourmaline: Ferrous
features.
Can often be seen even if
tourmaline is only a minor
component, AusSpec
AusSpec Workshop – 2010 International
Cu features
Malachite
Azurite
AusSpec
AusSpec Workshop – 2010 International
Non-responsive or problem minerals
SWIR: Minerals with no OH or water or carbonate
Quartz or feldspar
No diagnostic absorptions in SWIR
Opaques:
Magnetite, pyrite, chalcopyrite, magnetite
Strongly absorbing across vis-NIR-SWIR
AusSpec
AusSpec Workshop – 2010 International
1
2
3
4 Pattern recognition
5
6
7
8
9
2385
2306
Exercise 1: Compare these unknowns
10
with the “library” spectra in your
11
workshop manuals.
12
2352
2260
13
AusSpec
AusSpec Workshop – 2010 International
1350
1550
1720
1860
1880
2040
2160
2220
2230
2296
2306
2365
Wavelengths of the Main
SWIR Absorptions
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
Wavelength (nanometres)
2385
2306
Exercise 2: identify to what wavebands
10 the main absorption features in these
11 spectra belong.
12
2352
2260
13
AusSpec
AusSpec Workshop – 2010 International
Hull Corrections – For SWIR data
Hull Quotient Correction
class:biotite/chlorite mineral:Fe2+
Raw Reflectance
Hull
A Line
1 2
3 3 5
5 0 0
0 0 0
AusSpec
AusSpec Workshop – 2010 International
Hull Corrections and Vis-NIR data
Hull quotient
NOT Suitable....
Hull quotient correction on vis-NIR
spectra can distort feature shapes
and cause wavelength shifts
AusSpec
AusSpec Workshop – 2010 International
GMEX Volume 1 – The “SIFM”
In html format for easy browsing
Includes discussions on:
Background information on SWIR analysis
Absorption bands
Hull corrections
Structured approach to mineral interpretation
Interpretation of spectral mixtures
Includes a searchable spectral library of
common minerals
AusSpec
AusSpec Workshop – 2010 International
SWIR Spectral Interpretation
Interpretation of mineral spectra is based on the fact that
many common alteration minerals:
AusSpec
AusSpec Workshop – 2010 International
1350
Structured Approach
1550
1880
2040
2450
2 4 3 1
Then Exercise 3
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
Wavelength (nanometres)
AusSpec
AusSpec Workshop – 2010 International
2180-2228nm Group
AlOH Minerals:
AusSpec
AusSpec Workshop – 2010 International
Kaolinite Page
AusSpec
AusSpec Workshop – 2010 International
Mineral spectra can also be modified
by variations in
Mineral composition – e.g. illite, muscovite, chlorite
Mineral crystallinity – e.g. kaolinite, illite/sericites
AusSpec
AusSpec Workshop – 2010 International
Crystallinity
Affects the shapes and intensities of
absorption features
Sharpness
Kaolinite
Illite/white mica
AusSpec
AusSpec Workshop – 2010 International
Mineral Composition
Affects the wavelengths of
absorption features
Substitution by different cations
Fe and Mg for Al
White micas
Fe and Mg relative proportions
Chlorite and biotite
Na for K
White mica
alunite
AusSpec
AusSpec Workshop – 2010 International
Kaolinite Crystallinity
~2162nm
AusSpec
AusSpec Workshop – 2010 International
Kaolinite Crystallinity
Increasing Crystallinity =
1. Increasing sharpness and depth of 2160
absorption
2. Apparent shift in wavelength of 2160nm feature
i.e. Alteration halos in epithermal systems
Weathering profiles
AusSpec
AusSpec Workshop – 2010 International
Kaolinite vs Halloysite
Both kaolin minerals
Halloysite – weathering or lower T
hydrothermal alteration than Kaolinite
2162nm
Very similar in AlOH band (2206nm)…BUT
2206nm
Subtle difference in the OH band (near
1400nm)
AlOH Doublet
OH Doublet
AusSpec
AusSpec Workshop – 2010 International
Kaolinite vs Halloysite
AusSpec
AusSpec Workshop – 2010 International
Illite
Crystallinity
Increasing Crystallinity =
1. Increasing sharpness of
AlOH absorption
2. Increasing depth of AlOH
absorption relative to water
absorption
3. Weakening water feature
i.e. Alteration halos in
epithermal systems
AusSpec
AusSpec Workshop – 2010 International
Corbett and Leach AusSpec
AusSpec Workshop – 2010 International
White Mica
Composition
Illite and white mica (i.e.
muscovite, paragonite,
phengite)
Applications
Alteration halos
Overprinting phases
Lithological influences
AusSpec
AusSpec Workshop – 2010 International
White Mica
Composition
AlOH wavelengths
Paragonite – Na substitution
2180-2190nm
High Al
Phengite – Mg-Fe
substitution for Al
2216-2228nm
Low Al
Intermediate wavelengths
Mixed phases
AusSpec
AusSpec Workshop – 2010 International
Chlorite
Composition
Mg number influences
Wavelength of FeOH
Wavelength of MgOH
Applications
Multiple phases of chlorite
Regional from alteration
AusSpec
AusSpec Workshop – 2010 International
Chlorite
Composition
AusSpec
AusSpec Workshop – 2010 International
2365
Chlorite
2360
2355
MgOH vs Mg Number 2
R = 0.85
Composition
Wavelen gth (nm)
2350
2345
2340
2335
2330 2262
2260 FeOH vs Mg Number
2325 2
2258 R = 0.8619
2320
Wavelength (nm)
2256
0 0.2 0.4 0.6 0.8 1
Fe-rich Mg-rich 2254
Mg Number 2252
2250
2248
2246
2244
2242
0 0.2 0.4 0.6 0.8 1
Mg Number
AusSpec
AusSpec Workshop – 2010 International
Carbonate
Composition
Identified by wavelength of
diagnostic CO3 feature
Mg Alteration halos in
carbonates around porphyries
AusSpec
AusSpec Workshop – 2010 International
Carbonate
Composition
Fe Slope
More information can be
gained from the raw
reflectance spectrum
Specific ferrous features
can be seen in vis-NIR
AusSpec
AusSpec Workshop – 2010 International
Smectite
Species
AusSpec
AusSpec Workshop – 2010 International
Spectral Mixtures
Considered simply, spectral mixing can be viewed as a linear
process. For example,
AusSpec
AusSpec Workshop – 2010 International
Spectral Mixtures
BUT the spectral response of a mineral mixture is influenced by various
non-linear factors, which include:
Multiple scattering within the volume of the mixed minerals.
Varying absorption coefficients between different minerals.
Carbonate -------> chlorite -> smectite -> kaolinite/white mica ----> pyrophyllite+talc
Weak .
Strong
source detector
strong
moderate
30-100
microns
AusSpec
AusSpec Workshop – 2010 International
Mixed Spectra
Typically are characterised by:
AusSpec
AusSpec Workshop – 2010 International
Mixed mineral spectra
AusSpec
AusSpec Workshop – 2010 International
Mixed mineral spectra
Chlorite
White mica
Wt Mica+Chlorite
AusSpec
AusSpec Workshop – 2010 International
Mixed mineral spectra
Chlorite
White mica
AusSpec
AusSpec Workshop – 2010 International
Unravelling mixtures
AusSpec
AusSpec Workshop – 2010 International
Unravelling mixtures
AusSpec
AusSpec Workshop – 2010 International
Unravelling mixtures
AusSpec
AusSpec Workshop – 2010 International
Unravelling mixtures
AusSpec
AusSpec Workshop – 2010 International
Mixed Hematite-Goethite
Goethite 910-940nm
Mixed He+Go
Hematite <890nm
Gibbsite + hematite
Kaolinite + hematite
Gibbsite + kaolinite
+hematite
Kaolinite + hematite
Gibbsite + kaolinite
+hematite
Problematic minerals/material
Opaques (magnetite, sulphides)
Carbonaceous material
AusSpec
AusSpec Workshop – 2010 International
Spectral mixtures - sulphides
Sulphides
Noisy and dark
spectra
Weak illite
features
AusSpec
AusSpec Workshop – 2010 International
Unaltered “fresh” rocks
No alteration
minerals or clays
Minerals present
are probably
anhydrous
Noisy and dark
spectra
Need geology to
know that this is
not due to
sulphide
AusSpec
AusSpec Workshop – 2010 International
Exercise 4 TSG Pro, samples 1 to 14 (Aux colour: Norm. HullQ.)
000001
Aux
Mixed spectra
000002
0
000003
0.0417
0.0833
000004
000005
0.125
0.167
0.208
0.25
0.292
000006
0.333
000007
0.375
000008
0.417
000009
No rm. Hu llQ. (St ac k ed )
0.458
0.5
0.542
000010 0.583
0.625
000011
0.667
0.708
000012 0.75
0.792
000013
0.833
000014
0.875
0.917
0.958
NULL
AusSpec
AusSpec Workshop – 2010 International
SAMPLE TYPES AND
MEASUREMENT
AusSpec
AusSpec Workshop – 2010 International
Sample Considerations
All sample types can be measured
Core, pulps, soils, chips/cuttings, outcrop etc
As objective as possible
Not just because its different OR it looks interesting
AusSpec
AusSpec Workshop – 2010 International
File Naming
Name all spectra with hole name and incremental
depth, i.e.
Ddh001.001 etc
Unique names
This will allow you to always keep track of all
spectra
Combining multiple holes within a single TSG file
Integrating spectra and ancillary data (logging, geophysical and
geochemical)
AusSpec
AusSpec Workshop – 2010 International
Core Samples
Hand measurement
Selective measurements possible
Veins, selvages, matrix, clasts
AusSpec
AusSpec Workshop – 2010 International
Geochemical Pulp Samples
Direct comparison with geochemistry
AusSpec
AusSpec Workshop – 2010 International
Soil samples
Can allow widespread sampling and alteration mapping
where there is no outcrop
Can be directly associated with geochemical data.
Tips
Avoid plant/biogenic material
Weathering can mute alteration zoning
AusSpec
AusSpec Workshop – 2010 International
Grain size effects
Fine grain sizes have
brighter spectra but weaker
absorption features
Due to increased scattering
AusSpec
AusSpec Workshop – 2010 International
Quality Control: QA/QC
Vital at the measurement stage
Wet samples
Noisy spectra
AusSpec
AusSpec Workshop – 2010 International
QA: Water in
samples
AusSpec
AusSpec Workshop – 2010 International
QA: Spectral quality and noise
1
Aux
What is noise?
0.0588
0.118
0.8
0.176
0.235
Measurement effects
0.294
0.353
0.471
0.588
0.4
0.647
0.765
0.824
Instrumental effects
0.2
0.882
0.941
NULL
Steps in spectra
0
1400 1600 1800 2000 2200 2400
Wav elength in nm
Periodic noise
AusSpec
AusSpec Workshop – 2010 International
QA: Detector problems
1 2 3
AusSpec
AusSpec Workshop – 2010 International
QA: Artefact features
Artefact features are absorption features unrelated to the sample
mineralogy
Contaminated samples – oil/kerogen
Vegetation
Coatings or impregnations – resin or cellotape or glue
AusSpec
AusSpec Workshop – 2010 International
QA: ASD- Spectral quality and noise
Measurement
Must be at least 3 seconds or 30 averages long
The longer the time set - the better the spectra (5 secs a good option)
Preventing noise
Noise usually in the longer wavelength detector (SWIR2) because of low
signal at these wavelengths
Warm up light source for 1 hour before measurement, the longer ASD is on
the better the spectra
Take reference readings regularly, AND keep spectralon clean (best stored in
a box, easy to access and use with minimal handling)
AusSpec
AusSpec Workshop – 2010 International
The Spectral Geologist (TSG)
TSG is Project oriented
AusSpec
AusSpec Workshop – 2010 International
The Spectral Geologist (TSG)
Spectral analysis software
Input spectra from a variety of spectrometers
Spectra
TSG Data File
Sections, Geological
Logging Data
plots, down Context
hole
Geochemistry
X-Y-Z data
AusSpec
AusSpec Workshop – 2010 International
The Spectral Geologist (TSG)
TSG spectral files contain,
sample, wavelength and layer dimensions
additional layers of data can be imported and added
AusSpec
AusSpec Workshop – 2010 International
TSG - Viewing Spectra
Illite + Chlorite + Carbonate (Ref: Illite)
Aux
0.0717
1.2
0.178
0.285
0.391
0.497
0.604
0.71
0.817
0.923
1
Norm. Refl. (Aux. colour: Norm. Refl.)
1.03
1.136
1.243
1.349
1.456
1.562
1.668
0.8
1.775
1.881
1.988
2.094
2.201
2.307
2.414
0.6
2.52
AusSpec
NULL
International
TSG Downhole Logs
Index Norm. HullQ. Depth Lith1 Interpretations Gold Illite Composition Chlorite vs Biotite FeOxide dAlOH/dMgOHCa
000140:CY RCO479&3_20a 139 St Wtmica + chl + cb
000141:CY RCO479&3_21a 140 St Wtmica + chl + cb
000142:CY RCO479&3_22a 141 St Wtmica + chl + cb
000143:CY RCO479&3_23a 142 St Wtmica + chl + cb
000144:CY RCO479&3_24a 143 St Wtmica + chl + cb
000145:CY RCO479&3_25a 144 St Wtmica + chl + cb
000146:CY RCO479&3_26a 145 St Wtmica + chl + cb
000147:CY RCO479&3_27a 146 St Wtmica + chl + cb
000148:CY RCO479&3_28a 147 St Wtmica + chl + cb
000149:CY RCO479&3_29a 148 St Wtmica + chl + cb
000150:CY RCO479&3_30a 149 St Wtmica + chl + cb
000151:CY RCO479&3_31a 150 St Wtmica + chl + cb
000152:CY RCO479&3_32a 151 St Wtmica + chl + cb
000153:CY RCO479&3_33a 152 St Wtmica + chl + cb
000154:CY RCO479&3_34a 153 St Wtmica + chl + cb
000155:CY RCO479&3_35a 154 St Wtmica + chl + cb
000156:CY RCO479&3_36a 155 St Wtmica + chl + cb
000157:CY RCO479&3_37a 156 St Wtmica + chl + cb
000158:CY RCO479&3_38a 157 St Wtmica + chl + cb
000159:CY RCO479&3_39a 158 St Wtmica + chl
000160:CY RCO479&3_40a 159 St Wtmica + chl
000161:CY RCO479&3_41a 160 St Wtmica + chl + cb
000162:CY RCO479&3_42a 161 St Wtmica + chl + cb
000163:CY RCO479&3_43a 162 St Wtmica + chl + cb
000164:CY RCO479&3_44a 163 St Wtmica + chl + cb
000165:CY RCO479&3_45a 164 St Wtmica + chl + cb
000166:CY RCO479&3_46a 165 Ss Wtmica + chl + cb
000167:CY RCO479&3_47a 166 Ss Wtmica + chl + cb
000168:CY RCO479&3_48a 167 Ss Wtmica + chl + cb
000169:CY RCO479&3_49a 168 Ss Wtmica + chl + cb
000170:CY RCO479&3_50a 169 Ss Wtmica + chl + cb
000171:CY RCO479&3_51a 170 Ss Wtmica + chl + cb
000172:CY RCO479&3_52a 171 Ss Wtmica + chl +/- cb
000173:CY RCO479&3_53a 172 Ss Chlor + biot + wtmica
000174:CY RCO479&3_54a 173 Ss Chlor +/- biot
000175:CY RCO479&3_55a 174 Ss Chlor + biot
000176:CY RCO479&3_56a 175 Ss Chlorite
000177:CY RCO479&3_57a 176 Ss Chlor + biot + wtmica
000178:CY RCO479&3_58a 177 St Chlor + biot + mont
000179:CY RCO479&3_59a 178 St Chlor + biot + wtmica
000180:CY RCO479&3_60a 179 St Chlor + biot + wtmica
000181:CY RCO479&4_1a 180 St Chlor + biot + wtmica
000182:CY RCO479&4_2a 181 St Chlor + wtmica
000183:CY RCO479&4_3a 182 St Wtmica + chl + cb
000184:CY RCO479&4_4a 183 St Wtmica + chl + cb
000185:CY RCO479&4_5a 184 St Mont + chlor
000186:CY RCO479&4_6a 185 Ss Chl + biot + wtmica
000187:CY RCO479&4_7a 186 Ss Wtmica + chl + cb
000188:CY RCO479&4_8a 187 Ss Wtmica + chl + cb
000189:CY RCO479&4_9a 188 Ss Chlor + biot
000190:CY RCO479&4_10a 189 Ss Biotite + wtmica
000191:CY RCO479&4_11a 190 Ss Biotite + wtmica
000192:CY RCO479&4_12a 191 Ss Chlor + biot + mont
000193:CY RCO479&4_13a 192 Ss Wtmica + chl + cb
000194:CY RCO479&4_14a 193 Ss Wtmica + chl + cb
000195:CY RCO479&4_15a 194 Ss Wtmica + chl + cb
000196:CY RCO479&4_16a 195 Ss Wtmica + chl + cb
000197:CY RCO479&4_17a 196 Ss Wtmica + chl + cb +/- biot
000198:CY RCO479&4_18a 197 Ss Wtmica + chl + cb
000199:CY RCO479&4_19a 198 Ss Wtmica + chl + cb
000200:CY RCO479&4_20a 199 Ss Wtmica + chl + cb
000201:CY RCO479&4_21a 200 Ss Wtmica + chl + cb
000202:CY RCO479&4_22a 201 Ss Wtmica + chl + cb
000203:CY RCO479&4_23a 202 Ss Wtmica + chl + cb
000204:CY RCO479&4_24a 203 Ss Wtmica + chl + cb
000205:CY RCO479&4_25a 204 Ss Wtmica + chl + cb
000206:CY RCO479&4_26a 205 Ss Wtmica + chl + cb
000207:CY RCO479&4_27a 206 Ss Wtmica + chl + cb
000208:CY RCO479&4_28a 207 Ss Wtmica + chl + cb
000209:CY RCO479&4_29a 208 Ss Wtmica + chl + cb
000210:CY RCO479&4_30a 209 Ss Wtmica + chl + cb
000211:CY RCO479&4_31a 210 Ss Wtmica + chl + cb
000212:CY RCO479&4_32a 211 St Wtmica + chl + cb
000213:CY RCO479&4_33a 212 Ss Wtmica + chl + cb
000214:CY RCO479&4_34a 213 Ss Wtmica + chl + cb
000215:CY RCO479&4_35a 214 Ss Wtmica + chl + cb
AusSpec
000216:CY RCO479&4_36a 215 St Wtmica + chl + cb
000217:CY RCO479&4_37a 216 St Wtmica + chl + cb
000218:CY RCO479&4_38a 217 Ss Wtmica + chl + cb
International
4.119 0 19.881
TSG Section Plots
AusSpec
AusSpec Workshop – 2010 International
TSG Plan views
AusSpec
AusSpec Workshop – 2010 International
TSG Scatterplots
Scope 1:3676; 3546 points, R² =0.0269; Aux: Au(ppm) Scope 1:3676; 3371 points, R² =0.0362; Aux: dAlOH/dWater
1 2.2
0.917 2.083
20
0.833 1.967
0.75 1.85
1
16
0.667 1.733
0.583 1.617
0.5 1.5
Au(ppm)
KaolXT2
12
0.417 1.383
0.333 1.267
0.9
8
0.25 1.15
0.167 1.033
0.0833 0.917
4
0 0.8
0.8
NULL NULL
0
2196 2200 2204 2208 2212 2216 2220 2196 2200 2204 2208 2212 2216 2220
w a vAl OH w a vAl OH
Scope 1:3676; 3133 points, R² =0.0095; Aux: dAlOH/dMgOH-cb Scope 1:3676; 3133 points, R² =0.0786; Aux: wavMgOH/cb
2.2 2350
20
2360
2.083 2347.5
1.967 2345
16
1.85 2342.5
2350
1.733 2340
1.617 2337.5
2340
dAlOH/dMgOH-cb
12
wavMgOH/cb
1.5 2335
1.383 2332.5
2330
1.267 2330
8
1.15 2327.5
2320
1.033 2325
0.917 2322.5
4
2310
0.8 2320
NULL NULL
AusSpec
0
2196 2200 2204 2208 2212 2216 2220 2196 2200 2204 2208 2212 2216 2220
International
TSG file format
Four main files
*.tsg – contains the file names and information
*.bip – contains actual data
*.ini – contains the layout information
*_cras.bip – linescan image data
AusSpec
AusSpec Workshop – 2010 International
TSG Demonstration
Import data from csv file
Look at the data in excel
Copy data (these will go to the clipboard) including all
columns
AusSpec
AusSpec Workshop – 2010 International
Practice Exercise
Open the file: “Example Dataset 2010 testing”
This will open in the scatter plot screen
Spectrum screen
AusSpec
AusSpec Workshop – 2010 International
DATA ANALYSIS AND
INTERPRETATION
AusSpec
AusSpec Workshop – 2010 International
Approaches to Spectral Data Analysis
Sample Mineralogy
Manual Interpretation
Mineral characteristics
Spectral parameters / digital
mineralogy
AusSpec
AusSpec Workshop – 2010 International
Approaches to Spectral Data Analysis
Sample Mineralogy
Manual Interpretation
Mineral characteristics
Spectral parameters / digital mineralogy
AusSpec
AusSpec Workshop – 2010 International
Manual Interpretation
Important to have
knowledge to verify
automated analysis results
AusSpec
AusSpec Workshop – 2010 International
Manual Interpretation
Some data sets are not
suited to automated
analysis and require manual
interpretation
Soil spectra can be noisy
Overprinting with a
complex assemblage
Weathered-altered samples
AusSpec
AusSpec Workshop – 2010 International
Manual Interpretation
Some data sets are not
suited to automated
analysis and require manual
interpretation
Some minerals in low
proportions may need to be
visually identified (such as
topaz)
Soil spectra can be noisy
Overprinting with a
complex assemblage
AusSpec
AusSpec Workshop – 2010 International
Alunite species
Multiple phases of
alunite can be
observed in many
HS systems
Potentially
distinguish between
hydrothermal and
supergene
phases….
AusSpec
AusSpec Workshop – 2010 International
water features - silicified?
0.98
0.91
0.84
0.49
0.42
0.35
0.28
0.21
0.14
Acid
0.07
0
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
W avelengt h in nm
class: alun mineral: alunite
0.98
0.91
Vuggy silica
0.84
0.77
0.7
0.63
0.56
Norm. HullQ.
alunite
0.49
0.42
0.35
0.28
Higher Temp:
0.21
0.14
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
W avelengt h in nm
class: pyrph mineral: pyrophyllite
0.98
0.91
dickite
0.84
0.77
0.7
0.63
0.56
kaolin
Norm. HullQ.
0.49
0.42
0.35
0.28
0.21
0.14
0.07
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
W avelengt h in nm
Illite, chlorite
0.98
0.91
0.84
0.77
neutral
0.7
0.63
0.56
Norm. HullQ.
0.49
0.42
0.35
0.28
0.21
0.14
0.07
AusSpec
0
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
W avelengt h in nm
000018
0
0.0357
0.0714
Mixtures
0.107
000019
0.143
0.179
High sulphidation
0.214
000020
0.25
0.286
000021 0.321
0.357
Additional absorption
000022 0.393
0.429
features
Complex mixtures
000023
Nor m . HullQ. (Stac k ed)
0.464
0.5
000024
0.536
000025
0.643
0.679
significant
000026
0.714
0.75
000027
0.821
0.857
000028
NULL
AusSpec
AusSpec Workshop – 2010 International
Associated Alteration zones
ID in SWIR
Propylitic: chlorite, epidote (montmorillonite, carbonate)
AusSpec
AusSpec Workshop – 2010 International
Spectral Mixtures – Low Sulphidation
Low Sulphidation Mineral Assemblages
Illite-smectite and montmorillonite
Crystallinity can be a measure of temperature
and separate from phyllic to argillic
Composition can vary with >1 phase (often
paragonitic)
Look at “Illite Phases
Kaolinite example”
Can replace adularia (late stage), although
adularia cannot be seen in the SWIR
Propylitic
Chlorite, epidote
AusSpec
AusSpec Workshop – 2010 International
Illite Phases example data set
These are a series of 28 spectra of outcrop samples from a region of Illite
alteration and Au mineralisation. These spectra represent a traverse from
background unaltered samples through a zone of visible alteration and back
out to background.
(note the spectrum names are a combined easting/northing co-ordinate)
The logging data (log data) and Au values are already imported and displayed in
log screen.
1. Look at what mineralogy characterises the background rocks.
2. Assess the mineralogy of the alteration.
3. From the log data, visible alteration starts at around 3600mE and ends around
4300mE. Can the alteration halo be seen to extend further than this in the
spectral data?
4. Compare the Au values with your interpretation of the alteration, can you
identify an alteration phase that appears to correlate with the highest Au
values?
AusSpec
AusSpec Workshop – 2010 International
Mineral assemblage
AusSpec
AusSpec Workshop – 2010 International
Porphyry-epithermal settings
AusSpec
AusSpec Workshop – 2010 International
Approaches to Spectral Data Analysis
Sample Mineralogy
Manual Interpretation
Mineral characteristics
Spectral parameters / digital mineralogy
AusSpec
AusSpec Workshop – 2010 International
Mineral Identification Software
Rely totally upon training library
AusSpec
AusSpec Workshop – 2010 International
The Spectral Assistant
Automated spectral interpretation against a training
library of pure mineral spectra
Collected from worldwide sites
Note that results are only TSA‟s “best guess” and may
be wrong!!
AusSpec
AusSpec Workshop – 2010 International
Reasons for errors in TSA results
Spectra include minerals not represented in TSA‟s
library
Mixtures of >2 minerals (higher TSA errors)
A common occurrence in HS samples!
Dark or noisy spectra
Spectrometer error/miscalibration – absorption
features at the wrong wavelengths
AusSpec
AusSpec Workshop – 2010 International
TSA Performance
Open “HS Epithermal Mineral spectra TSG 7.tsg”
AusSpec
AusSpec Workshop – 2010 International
Caution when using TSA results
TSA is an assistant to spectral interpretation
The output results are its “best guess”
User must carry out visual checks and monitor the
results
Use knowledge of the geology, to recognise if any
unusual/unlikely minerals have been identified
Keep track of the error levels
Overlay reference spectra to check where uncertain
AusSpec
AusSpec Workshop – 2010 International
Tips for using TSA
TSA results
Mineral identification of different members of the same mineral
group
Chlorite - mg chlorite, fe chlorite, int chlorite
AusSpec
AusSpec Workshop – 2010 International
Settings for TSA
Settings option under File menu – TSA tab
AusSpec
AusSpec Workshop – 2010 International
Plotting mineral %ages from spectra
Based on TSA spectral weightings, visual estimates or other
sources
AusSpec
AusSpec Workshop – 2010 International
AUX MATCHING
AusSpec
AusSpec Workshop – 2010 International
Interpretation by Aux Matching
Search and match by spectral shapes
Project TSG file - unknowns
Library TSG file – fully interpreted (manual)
Custom built
Classification
Output result can be reported from
Library spectrum name OR
Library spectrum comment field, either organised or
not
Class: and/or Mineral:
Name can be a classification AusSpec
AusSpec Workshop – 2010 International
Aux Match Library
Spectra in library can include:
Mixed mineral spectra
Spectra from project area/specific geological setting
Output of results as
Based on observation that
Representative mineral mineral assemblages
there are similarities in
assemblages of the or alteration facies
mineral assemblage
deposit style classifications
between different project
areas from the same i.e. epithermal Cost effective way
deposit style of getting a project
Mixed mineral spectra specific library
AusSpec
AusSpec Workshop – 2010 International
Matching against Custom Libraries
AusSpec
AusSpec Workshop – 2010 International
Using a Custom Library
Open the file “HS Epithermal Mineral Spectra TSG 7.tsg”
Attach “Highsulph Epith Library TSG 7.tsg”
Overlay spectra on spectrum screen (by selecting the aux
toggle)
Open the floater screen and change to “aux match” mode
Look at matches, test out if water should be used or not,
and which spectral layer would work best.
Update aux matches to output to “CustSample”
Look at results in the log screen and compare with the TSA
results (remember to plot the CustScore as well)
AusSpec
AusSpec Workshop – 2010 International
Settings for Aux Match
Matching Algorithm – Subsetting
(input from Peter) exclude/include water
Pearson’s Correlation Other wavelength ranges
AusSpec
AusSpec Workshop – 2010 International
VALIDATION OF MINERALOGY
AusSpec
AusSpec Workshop – 2010 International
Verification of results
AusSpec
AusSpec Workshop – 2010 International
Reasons for Errors in Results
Spectra include minerals or mixtures not represented
in library (TSA or Custom/Aux)
Dark or noisy spectra
Spectrally ambiguous minerals in a sample (i.e.
chlorite/biotite, illite/muscovite)
Non-mineral artefact features
Spectrometer related artefacts/glitches
Spectrometer error/miscalibration – absorption
features at the wrong wavelengths
AusSpec
AusSpec Workshop – 2010 International
Caution when using automated results
The automated algorithms are designed to assist your
spectral interpretation
The output results are the “best guess”
Highly dependent on the input libraries/training set
Highly dependent on the quality of the input project spectra
User must carry out visual checks and monitor the results
Use knowledge of the geology, to recognise if any unusual/
unlikely minerals have been identified
Keep track of the error levels/ score values
AusSpec
AusSpec Workshop – 2010 International
Mineralogy is ONLY the first step
Although accurate mineralogy is a vital first step, there is
more information than just mineralogy in the SWIR data
AusSpec
AusSpec Workshop – 2010 International
TSG SCALARS
AusSpec
AusSpec Workshop – 2010 International
What are they?
Any numerical or text data associated with your
spectral data set
AusSpec
AusSpec Workshop – 2010 International
When to Import Scalars?
Import data early or later?
Early -
Logging data – to compare with mineralogy and
separate lithological controls from possible alteration.
Assay data – highlight mineralised intervals, start to
focus on relationships, …obvious ones or not so…
Later -
External data can have errors (old logging) which can be
misleading
AusSpec
AusSpec Workshop – 2010 International
Going Further…..
There is more information than just mineralogy in the spectral data.
1. Mineral identification
2. Verification of results – we should have accurate mineralogy….
3. Analysis and representation of down hole mineral characteristics
BUT
How do we access specific mineral information?
Composition, crystallinity, relative mineral proportions?
AusSpec
AusSpec Workshop – 2010 International
Approaches to Spectral Data Analysis
Sample Mineralogy
Manual Interpretation
Mineral characteristics
Spectral parameters / digital mineralogy
AusSpec
AusSpec Workshop – 2010 International
The Key to Success
Spectral Plots
Digital Mineralogy
Sericite Composition
2209.76 2.04 2245.59 2.8
2212.77 4.87 2243.38 4.83 Chlorite Composition
2213.86 5.14 2241.98 5.18
2217.41 5.48 2246.49 6.53 Biotite Composition
2215.18 0.8 2256.4 1.72
2207.57 0.93 2266.15 1.43 Chlorite Proportion
2203.71 1.08 2252.32 1.11 Illite Crystallinity
2203.15 0.96 2268.05 0.93
2204.11 1.08 2257.85 1.36 Fe Carbonate Prop’
0 0 2257.73 1.04
0 0 2253.78 1.42
2204.83 0.94 2269.12 1.9 Kaolinite Crystallinity
2208.25 0.93 2250.04 1.46
2209.21 0.99 2246.58 0.88 Smectite Composition
2203.33 0.91 2249.16 1.17
Section 35400N
AusSpec International
E dh12 As (ppm) W
dh13 ddh05
ddh07
AusSpec
International
Drill Hole
Cross Section
Mine site
orientation
survey
~800 Grade
Control
Samples
Mineralogically
simple
AusSpec
AusSpec Workshop – 2010 International
What are Spectral Parameters?
They are measurements of
Wavelength
Depth
Width
Areas of absorption features
Or combinations of these values i.e. commonly as ratios of
depths of features
Influenced by mineralogical factors such as composition and
crystallinity
AusSpec
AusSpec Workshop – 2010 International
Common Spectral Parameters
Wavelengths of absorption features
AlOH – common to kaolinite, white mica, montmorillonite
FeOH – chlorite, biotite
AusSpec
AusSpec Workshop – 2010 International
Ratios
Measure of
Crystallinity
Illite - Depth AlOH/depth water
AusSpec
AusSpec Workshop – 2010 International
Ratios
Measure of
Relative proportions of two
minerals, use relative depths
of their diagnostic
absorptions:
Depth AlOH/depth feoh – illite
vs chlorite
Depth AlOH/depth MgOH –
illite vs Mg mineral
Depth AlOH/depth Carbonate
– illite vs calcite
0.149
0.311
0.472
0.633
Fe Slope
1.2
0.955
Norm. Refl. (Aux. colour: Norm. Refl.)
1.116
1.277
1.599
1.76
1.921
2.082
~1700nm
Fe carbonate
2.243
0.6
2.404
0.149
NULL
0.472
1.2
0.633
0.794
0.955
Fe minerals, such as
Norm. Refl. (Aux. colour: Norm. Refl.)
1.116
chlorite
1
1.277
1.438
1.599
1.76
0.8
1.921
2.082
2.243
2.404
AusSpec
2.565
0.6
NULL
International
600 900 1200 1500 1800 2100 2400
Wavelength in nm
AusSpec
AusSpec Workshop – 2010 International
IMPORTANT.…remember that
Spectral parameters can be influenced by
several factors
AusSpec
AusSpec Workshop – 2010 International
Calculating Spectral Parameters
TSG calculates spectral parameters using any
of these methods:
The spectral profile (i.e. directly from the
spectra)
Arithmetic expressions (i.e. ratios)
Feature parameters (i.e. based on
deconvolution of the spectra)
Class Extraction Scalars (i.e. data filtered by
Class, eg. TSA Mineral 1, lithology etc.)
AusSpec
AusSpec Workshop – 2010 International
Using Spectral Parameters
Example…..
In the data file “SpecParamExample.tsg” you will find a
series of 50 spectra. Examine these spectra and then
calculate and display at least 3 different spectral
parameters that describe the significant mineralogical
changes in these samples
Wavelength
Slope
Relative depth
AusSpec
AusSpec Workshop – 2010 International
Strategies for selecting parameters
What is the best way of selecting suitable parameters for
a data set?
Mineral assemblages
Observed variations – wavelength shifts
mostly look at crystallinity, composition, important relative
AusSpec
AusSpec Workshop – 2010 International
Examples of selecting parameters
Example assemblage:
Illite, chlorite, carbonate, kaolinite, dickite, dravite
AusSpec
AusSpec Workshop – 2010 International
What if there is no absorption minimum?
AusSpec
AusSpec Workshop – 2010 International
Class Extraction – “3 Drill hole” eg
Class extraction – filtering of data according to a class
value 3 Drill Hole example - Would further drilling intersect
mineralisation at depth?
AusSpec
AusSpec Workshop – 2010 International
Making the output data work for us….
Whatever sample measurement or analysis system
used, analysis of the spectra has provided us with
output including:
Mineral interpretations
Spectral parameters
Digital mineralogy - Mineral specific spectral parameters
AusSpec
AusSpec Workshop – 2010 International
Applications
Data analysis, variations in mineral characteristics
Data integration, with ancillary data
Making the data work for us to represent the mineral
variations that are relevant to our project area.
Plots, sections and maps
Scatterplots of relationships
Classification
Histograms to assess multiphase alteration
AusSpec
AusSpec Workshop – 2010 International
Multiple phases and overprinting
Scope 1:1230; 1230 points; Aux: Interpretation
Aux
Distribution of AlOH
57
Sericite +/- wk
Ser + chl
wavelengths dependent on
54
Ser + chl + cb
Ser + cb
51
Ser + cb + chl
Ser + kaol
Kaolinite
Smect + kaol +
45
Ser + nontr + c
Montmorillonite
Chl/bt + ser
39
Ser + ep + cb +
Illite+carbonate (red/orange)
36
Sericite
Ka (disord) + m
33
24
21
18
the illite+carbonate
9
6
assemblage
3
0
2196 2198 2200 2202 2204 2206 2208 2210 2212 2214
kaolinitic
wa v A l O H
AusSpec
AusSpec Workshop – 2010 International
Multiple phases - Victoria
Phengite + chlorite
Fosterville Ballarat
Costerfield Bendigo
AusSpec
AusSpec Workshop – 2010 International
Example: Data Integration
muscovite
Chlorite
phengite
Sericite
O
/K
2
t Geochemical data
rge
Chlorite
gO
O
3
gO
O
Depth
Al
2
Ta
Mg Fe
M
M
K
2
10m Elemental distribution, and
mineralisation
50m
Geophysical data
55m Changing physical
characteristics
60m
Lithology
Spectral Data
100m Associated mineralogical
variation
150m
200m
AusSpec
AusSpec Workshop – 2010 International
Analysis of results:
Scatterplots
AusSpec
AusSpec Workshop – 2010 International
Overprinting
Weathered
kaolinite
increasing Kaolinite response
Regional phase
phengite
muscovite
paragonite
Targetalteration
Target Alteration
AusSpec
AusSpec Workshop – 2010 International
Setting up scatterplots
Using two different parameters
One common to the minerals being
analysed
The other diagnostic of a particular mineral
group to be separated
AusSpec
AusSpec Workshop – 2010 International
003936:EP1601960 Pima0 le was used so that the numbers herein a; Aux: GeneralClasses
1
Aux
Illite +cb+...
0.8
Illite + ch...
Biotite +/-...
0.6
N or m. H ullQ
Chlorite/b...
Dickite
0.4
Dickite+...
Kaolinite
<clip
0.2
001754:03EPE043P016A Pima0>cliple was used so that the numbers herein a; Aux: GeneralClasses
1
NULL Aux
Illite + ch...
Illite + ka...
Biotite +/-...
0.6
Nor m. HullQ
Chlorite/b...
Dickite
Dickite+...
Kaolinite
<clip
0.2
004099:EPE20A02275 Pima0 le was used so that the numbers herein a; Aux: GeneralClasses >clip
1
Aux NULL
Illite + ch...
Illite + ka...
Chlorite/b...
Dickite
Dickite+...
Kaolinite
002972:03h943023B Pima0 le was used so that the numbers herein a; Aux: GeneralClasses
1
<clip
0.2
Illite + ka...
Chlorite/b...
Dickite
environments
0.4
Dickite+...
Kaolinite
<clip
0.2
>clip
Illite +cb+...
0.8
Illite + ch...
Illite + ka...
Biotite +/-...
Chlorite/b...
Dickite
0.4
sample.
Dickite+...
Kaolinite
>clip Aux
0.8
Illite + ka...
Biotite +/-...
Can be used to good advantage as a measure of
0.6
N or m. H ullQ
Chlorite/b...
Dickite
relative proportions of these minerals in mixtures
0.4
Dickite+...
Kaolinite
<clip
0.2
>clip
NULL
0
AusSpec
AusSpec Workshop – 2010 International
Wavelength of AlOH
Can be influenced by a number of factors depending on the
mineralogy of the samples
Kaolinite vs illite
Illite/white mica compositions
Multiphase illite/white mica
Illite vs dickite
Many applications
AusSpec
AusSpec Workshop – 2010 International
Put it all together… large data set
Using the kaolinite XT and (general) AlOH wavelength can allow separation of mineral
mixtures and assessment of relative mineral proportions.
Kaolinite XT – influenced here by kaolinite content relative to illite
AlOH – kaolinite has a constrained AlOH wavelength
Illite has broadly varying AlOH wavelengths indicating different phases of illite with Au
associated with one of them.
AusSpec
AusSpec Workshop – 2010 International
Relative depths of features: Ratios
Can provide a measure of the relative proportions of two or
more minerals.
AusSpec
AusSpec Workshop – 2010 International
EP Drill hole section Example.tsg
Practice some TSG tools
1. Class extraction (Class XT)
White mica composition – class XT the AlOH
Alunite composition – class XT the 1480-90nm alunite
feature
2. Ratio of alunite to pyrophyllite to map out relative
proportions of these two minerals in this data set.
3. Plot these in the scatterplot screen
Change colours of the Assemblage plot….
4. Discuss difference between “fullAlOH” and “AlOH”
scalars.
AusSpec
AusSpec Workshop – 2010 International
Steps to a successful spectral survey
1. Measure samples
Using appropriate spectrometer for application
Controls on quality, noise, sample conditions
Save data with appropriate naming convention
AusSpec
AusSpec Workshop – 2010 International
Conclusions
Spectral data can provide data on
Mineralogy
Mineral characteristics (composition, crystallinity)
Relative mineral proportion
AusSpec
AusSpec Workshop – 2010 International