View
0
Download
0
Category
Preview:
Citation preview
PhD. student: Cecilia Contreras I Helmholtz Institute Freiberg for Resource Technology I www.hzdr.de/hif
Mapping Epithermal Alteration Mineralogy with High Spatial Resolution Hyperspectral
Imaging of Rock Samples
Lisbon, December 15th, 2017
GRSG 28th International Annual Conference “Applied Geological Remote Sensing”
Cecilia Contreras
Co-authors: Dr. Chris HeckerDr. Freek van der Meer
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 2
ITC Faculty of Geo-Information Science and Earth Observation
Differentiation betweenhypogene and supergene
composition
crystallinity
mineral associations and patterns
hypogene: primary minerals i.e.: hydrothermal alterations
supergene: secondary minerals i.e.: weathering processes
Introduction
Hypogene and supergene alterations (Modified from White & Hedenquist, 1995)
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 3
ITC Faculty of Geo-Information Science and Earth Observation
To quantitatively study hypogene mineral assemblages to automatically classify the alteration zones of an epithermal alteration system
spectral and chemical differences mineral abundances
SWIR colour composite of sample260
hyperspectral images
Objective and Approach
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 4
ITC Faculty of Geo-Information Science and Earth Observation
(A) Location of the Rodalquilar Au deposit, Cabo de Gata volcanic field, south-eastern Spain. (B) Central part of the field (Arribas et al., 1995)
Cross-section of the Rodalquilar caldera showing the distribution of the alteration zones modified from White & Hedenquist (1995)
Study area: Rodalquilar Caldera Complex, Almeria, Spain
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 5
ITC Faculty of Geo-Information Science and Earth Observation
Compilation of the 22 SWIR color composite images from Rodalquilar samples
288 spectral bandswavelength range: 1.0µm – 2.5µm
pixel size: 0.4mm
Specim Hyperspectral SWIR camera
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Image acquisition: 22 Hyperspectral SWIR images
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 6
ITC Faculty of Geo-Information Science and Earth Observation
Sample 260 Wavelength map between 1200 – 2400nm. Depth stretching 0 – 1470
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Visual analysis : Wavelength maps
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 7
ITC Faculty of Geo-Information Science and Earth Observation
supergene alunite hypogene alunite
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Visual analysis: Alunite spectral and chemical differences
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 8
ITC Faculty of Geo-Information Science and Earth Observation
Supergene
Hypogene
Supergene
Hypogene
ASD spectra
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Visual analysis: Alunite spectral and chemical differences
Spectra from images
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 9
ITC Faculty of Geo-Information Science and Earth Observation
Supergene
Hypogene
Spectra from images
Supergene
Hypogene
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
ASD spectra
Visual analysis: Alunite spectral and chemical differences
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 10
ITC Faculty of Geo-Information Science and Earth Observation
Supergene
Hypogene
Supergene
Hypogene
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Spectra from images
Wavelengths maps, range (1460-1495nm) Supergene - Hypogene
ASD spectra
Visual analysis: Alunite spectral and chemical differences
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 11
ITC Faculty of Geo-Information Science and Earth Observation
Supergene HypogeneNa 13 3402.33K 120*103 50*103
Sr 97.33 533.79Y 1.03 3.31Zr <0.26 <0.26
Na/K <0.01 0.07
LowLow High
Inductively Coupled Plasma Optical Emission Spectroscopy
(ICP-OES) Thermogravimetric Analysis (TGA)
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Validation: Geochemical analysis
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 12
ITC Faculty of Geo-Information Science and Earth Observation
Total of 20 endmembers
Endmember extraction: Wavelength maps
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
Offset and hull-removed spectra of alunite and kaolinite endmembers.
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 13
ITC Faculty of Geo-Information Science and Earth Observation
Color composites to evaluate spatial distribution of the minerals
ISMA sample260 Alunite long
Alunite short-BSilicification
ISMA sample213 Alunite short-AJarosite overall
Montmorillonite
ISMA sample208Alunite short-BJarosite 1936Silicification
Unmixing: Iterative Spectral Mixture Analysis (ISMA)
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 14
ITC Faculty of Geo-Information Science and Earth Observation
ISMA sample260 Alunite long
Alunite short-BSilicification
ISMA sample260 Alunite long
Alunite short-AAlunite short-B
Unmixing: Iterative Spectral Mixture Analysis (ISMA)
i.e.: alunite zonation
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 15
ITC Faculty of Geo-Information Science and Earth Observation
ISMA sample260 Alunite long
Alunite short-AAlunite short-B
Alunite longAlunite short-AAlunite short-B
Alunite long Alunite short-A
Alunite short-B
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
i.e.: alunite zonation
Unmixing: Iterative Spectral Mixture Analysis (ISMA)
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 16
ITC Faculty of Geo-Information Science and Earth Observation
Mineral Abundances per Alteration Zones (Arribas et al 1995)
Vuggy silica Interm. argillic NonAdvance argillic SupergeneAlteration zones
Quantification: Iterative Spectral Mixture Analysis (ISMA)
Image acquisition
Visual analysis Validation Endmember
extraction Unmixing Quantification
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 17
ITC Faculty of Geo-Information Science and Earth Observation
• Hyperspectral imaging is an emerging technique for mineral mapping(detection mineral mixtures and spatial patterns)
• Spectra and geochemical analysis offer a good tool to differentiate between supergene andhypogene samples
• Alunite minerals in the Rodalquilar show spectral variations and geochemical analysis provethat they are also different in their chemistry
• Exploring automatic algorithms for the endmember extraction
• Applying the proposed methodology on a larger dataset to precisely develop the automaticclassification
Conclusions
Future work
• For a complete evaluation of alteration mineralogies, LWIR data would be useful
PhD. student: Cecilia Contreras
Helmholtz Institute Freiberg for Resource TechnologyPage 18
Thank you for your attention!
ITC Faculty of Geo-Information Science and Earth Observation
Gracias!
PhD. Student: Isabel Cecilia ContrerasHIF-Exploration division
E-Mail: i.contreras@hzdr.de
Recommended