Transcript Mineral Abundance Mapping Using Hyperion Dataset in Part

Mineral Abundance Mapping Using
Hyperion Dataset in Part of Udaipur Dist.
Rajasthan, India
Presented By
Salaj.S.S
PSNA College of Engineering and Technology
Dindigul, TamilNadu
Introduction
• Hyperspectral remote sensing , measuring hundreds
of spectral bands from aircraft and satellite platforms,
provides unique spatial/spectral datasets for analysis
of surface mineralogy (Goetz et al., 1985; Kruse et
al.,2003). Their use for geologic applications is well
established (Goetz et al., 1985; Kruse et al.,
1999,2003; Rowan and Mars, 2003).
• For the present study Hyperion, a hyperspectral
imager on-board of EO-1 satellite is used contains
220 spectral bands ranging from 400-2500nm and the
spatial resolution is 30 meter per pixel, swath width is
7.5 km by 42 or 100 km area.
Comparison between Multispectral and Hyperspectral sensors
Study Area
• Part of Udaipur district – Aravalli
Fold Belt, Rajasthan covering
about 303.43 sq. km
• Extent: 73˚ 33’ 25” E to 73˚42’53” E
24˚ 09’ 34” N to 24˚ 31’ 40” N
• The average height of the area is
598m from the sea level
Hyperion data
(FCC 40 31 13)
Aims and Objectives
The main aim of the study is to test the abilities of Hyperion
data for mineral abundance mapping in the study area. The broad
objectives are as follows
• Atmospheric Correction of Hyperion Data.
• Extraction of Endmembers from the image using advanced
techniques like MNF, PPI etc.
• Mineralogical mapping using the various endmembers
extracted from the image.
Data Used
• Hyperion image (Path : 148, Row : 43)
L1R Image - Radiometrically corrected
L1Gst – Radiometrically as well as geometrically and
Topographically corrected
•
Spectral Library(USGS)
•
Geological Map of the study area
Software Used
• ENVI 4.7
• Arc GIS 9.3
Methodology
Hyperion L1Gst Data
Hyperion L1R Data
Geological Map
Pre processing
Atmospheric correction using
FLAASH
Geometric Correction
Spectral
Library(USGS)
MNF Transformation
Pixel Purity Index (PPI)
n-D visualizer
Resampling
Spectral Analyst
(Endmember Identification)
Mapping Methods
SAM
MTMF
Mineralogical Mapping
Interpretation of
Geological Units
Preprocessing
Bad Column Removal
Each bad column was replaced by average of previous and next column
Before correction
After correction
Atmospheric Correction using FLAASH Module
Atmospheric Correction using FLAASH Module
Sensor type
Hyperion
Pixel size
30
Ground elevation
.6 km
Scene Centre Lat/Long
24.6◦N,73.7◦E
Visibility
40 km
Sensor altitude
705km
Flight date & Flight time
19/01/2004,5:22:17
Atmospheric model
Tropical
Aerosol model
Rural
Water vapour retrieval
1135nm
Wavelength Calibration
Yes
Advanced parameters
Output reflectance Scale factor
10000
MODTRAN resolution
15 cms-1
MODTRAN multi scattering model
Scaled DISORT 8 Stream
CO2
390 ppm
Atmospheric Correction
Geometric Correction
Hyperion L1Gst image as base image and L1R image as warp
image(Auto-image-to-image registration)
Dimensionality Reduction
• Minimum Noise fraction was used to reduce spectral
dimensionality and redundancy
• Noise estimation from image
• Separating noise from image
• New uncorrelated 8 bands from 144original bands
Comparison of Geological map with MNF color composites
Shelf Facies – Delwara
Group and Debari Group
Deep Sea Facies - Jarol
Group and Lunavada
Group with Ultramafics
MNF(R:2 G:3 B:6)
Pixel Purity Index(PPI)
N-Dimensional Visualizer
• selecting the endmembers in n-D space.
• Pure pixels can be viewed in any angle by rotating the n-D scatterplot
• The clusters can be saved as ascii files
• Finalization of clusters after using spectral analyst
Endmember Identification Using Spectral Analyst
•
Endmember identification using Spectral Angle Mapper and
Spectral Feature Fitting with equal weightage
•
Mineral with highest score was identified as material for a
particular cluster
Mapping Methods
• SAM (Spectral Angle Mapper)
• Mixture Tuned Matched
Filtering(MTMF)
Post Classification
• Rule Classifier
End Member Collection Spectra
Mineral Abundance Maps
1) Grossularite (Ca3Al2Si3O12) Grossularite is especially characteristic of both thermally and regionally metamorphosed impure
calcareous rocks. It also occurs in the rocks which have undergone calcium metamorphism. It may
result from replacement of Wollastonite. It also occurs in association with serpentinite and has
been described from highly metamorphosed layered complex possibly resulting from the
alteration of anorthite.
2) Calcite (CaCO3) –
Calcite is one of the most ubiquitous minerals, and in addition to being an important rockforming
mineral in sedimentary environments, it also occurs in metamorphic and igneous rocks and is a
common mineral of hydrothermal and secondary mineralization. In sedimentary rocks Calcite is
the principal constituent of most limestones. It occurs both as a primary precipitate and in the
form of fossil shells. Calcite is the stable form of CaCO3 and although approximately equal
numbers of organisms make their shells of Calcite and aragonite (or, as for some of the mollusca,
of both), the aragonite eventually undergoes recrystallization to calcite.
3) Pyrite (FeS2) Pyrite is similar in appearance to chalcopyrite, pyrrhotite and marcasite, It can be
distinguished from chalcopyrite since the latter mineral has a deeper yellow colour in
reflected light, and is softer, being scratched by a knife, Pyrrhotite is bronze rather than
brasscoloured, is also scratched by a knife and usually magnetic.
4) Andradite (Ca3(Fe+3, Ti)2Si3O12) Typically occurs in contact or thermally metamorphosed impure calcareous sediments
and particularly in the metamorphism and particularly metasomatic skarn deposits. This
involves the addition of Fe2O2 and SiO2. If FeO is also introduced, hedenbergite may
form in addition to andradite, and if insufficient silica is available magnetite may result
giving the typical andradite-hedenbergite-magnetite skarn assemblage. Andradite also
occurs as the result of metasomatism connected with the thermal metamorphism of
calcic igneous rocks such as andesite
Mineral Map
Conclusion
• Mineralogical mapping can be achieved using
Hyperion data.
• Mapping is affected by topography and
vegetation and ground in-situ conditions.
• False positives may be generated
• Limitations due to Standard spectral library.
• Need of Field spectral library for minerals is
essential for accurate mapping.
References
• Kruse, F.A., J.W. Boardman And J.F. Huntington (2003): Evaluation and
validation of hyperion for mineral mapping, IEEE Trans. Geosci. Remote
Sensing, 41 (6), 1388-1400.
• Kruse, F.A., J.W. Boardman And J.F. Huntington (1999), Fifteen years of
hyperspectral data: Northern Grapevine Mountains, Nevada, in Proceedings
of the 8th JPL Airborne Earth Science Workshop, Jet Propulsion
Laboratory Publ. 99-17, 247-258.
• Rowan, L.C. And J.C. Mars (2003): Lithologic mapping in the Mountain
Pass, California area using Advanced Spaceborne Thermal Emission and
Reflection Spectrometer (ASTER) data, Remote Sensing Environ., 84,350366.
• S Sinha-Roy, G Malhotra, M Mohanty (1998), Geology of Rajasthan,
Geological Society of India, 1st edition, 278 p.
• Deer W. A., Howie R. A. and Zussman J, 1966, Mineralogy, Longman
Group Limited, 7th edition, 528p.
Thank You