Journal of Spatial Science

ISSN: 1449-8596 (Print) 1836-5655 (Online) Journal homepage: http://www.tandfonline.com/loi/tjss20

Identification and mapping of high-potential iron

ore alteration zone across Joda, Odisha using
ASTER and EO-1 hyperion data

Arnab Sengupta, Manik Das Adhikari, Sabyasachi Maiti, Soumya Kanti Maiti,
Pankajini Mahanta & Siddhartha Bhaumick

To cite this article: Arnab Sengupta, Manik Das Adhikari, Sabyasachi Maiti, Soumya Kanti Maiti,
Pankajini Mahanta & Siddhartha Bhaumick (2018): Identification and mapping of high-potential iron
ore alteration zone across Joda, Odisha using ASTER and EO-1 hyperion data, Journal of Spatial
Science, DOI: 10.1080/14498596.2018.1485120

To link to this article: https://doi.org/10.1080/14498596.2018.1485120

Published online: 25 Jun 2018.

Submit your article to this journal

Article views: 19

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tjss20
JOURNAL OF SPATIAL SCIENCE
https://doi.org/10.1080/14498596.2018.1485120

Identification and mapping of high-potential iron ore

alteration zone across Joda, Odisha using ASTER and EO-1
hyperion data
Arnab Sengupta, Manik Das Adhikari, Sabyasachi Maiti, Soumya Kanti Maiti,
Pankajini Mahanta and Siddhartha Bhaumick
Department of Geology and Geophysics, Indian Institute of Technology Kharagpur, West, Bengal, India

ABSTRACT KEYWORDS
Identification and mapping of high-potential Iron Ore indicative altera- High-potential iron ore
tion zone is important for new prospects. Here we have performed indicative alteration zone;
same through ASTER and Hyperion data by considering the in-situ and multi- and hyper-spectral;
reported spectral signatures and Band Ratio techniques. The Minimum FLAASH-MNF- PPI-SAM-band
ratio; Singhbhum shear zone
Noise Fraction and Pixel Purity Index have been applied for the noise
reduction and pure pixel extraction from the Hyperion data. Spectral
Angle Mapper algorithm has been used to identify the Hematite and
Goethite exposures based on the spectral assembling. On the other
hand, Ferric Iron, Ferrous Iron, Ferrous Silicates, and Ferric Oxide index
have been applied on ASTER VNIR and SWIR data for demarcating
alteration zones indicating subsurface occurrence of iron ore.
Thereafter, the consistency index has been established between the
ground truth/field observations and classified alteration through
kappa statistics demonstrates the accurate identification and remote
mapping of Iron Ore alteration zone in and around Joda region.

1. Introduction
The steel industry is one of the major sectors of the country which plays an essential role
in strengthening the economy. Iron Ore is generally found in the form of magnetite
(Fe3O4), hematite (Fe2O3), goethite (FeO(OH)), limonite (FeO(OH)nH2O) or siderite
(FeCO3) which are used for the manufacture of steel. The demand for Iron Ore has
escalated globally in recent times led by rising in steel production. The high demand
from steel industry would bring Iron Ore supply deficit of about 20–25 million tons
globally in the forthcoming years (Thangavelu et al. 2011). In the recent time, India
produces 65 million tons of steel, though, as per the ‘National Steel Policy’ country is
anticipated to increase this production to about 180 million tons by the year of 2020. All
these issues indicate that the demand of Iron Ores is continuously increasing, therefore,
necessitating the adoption of rapid and accurate approaches for identification of Iron
Ore depositional zones, exploration and grade assessment.
The detection and mapping of surface minerals using remote-sensing technique
provide an opportunity to improve the initial steps for mineral exploration
(Moghtaderi et al. 2007, Zhang et al. 2007, Gabr et al. 2010, Pour and Hashim 2011,

CONTACT Manik Das Adhikari [email protected]

© 2018 Mapping Sciences Institute, Australia and Surveying and Spatial Sciences Institute
2 A. SENGUPTA ET AL.

2014, Raj et al. 2015, Ducart et al. 2016). Several geoscientists have used multispectral,
hyperspectral, and in-situ data to detect the various minerals present in the earth surface
through Image Enhancement, Principal Component Analysis (PCA), Spectral Analysis,
and Band Ratio technique (Gabr et al. 2010, Pour and Hashim 2012). In remote sensing,
satellite and airborne multispectral or hyperspectral sensors are widely used to measure
the properties of land, sea, and atmosphere (Gabr et al. 2010, Pour and Hashim 2012).
Geological image interpretation through Band Ratio techniques are widely performed
on Landsat Multispectral Scanner (MSS) data in which the VNIR bands are used for the
iron oxide mapping whereas SPOT and LandsatTM data are widely used for lithological
mapping (e.g. Van Deer Meer and De Jong 2012, Mia and Fujimitsu 2012, Ahmed and
Iqbal 2014, Pour and Hashim 2014). Similarly, Hyperion data are also used for identifica-
tion of minerals present in the lithosphere because of its two spectrometers in visible &
near-infrared (VNIR) and shortwave infrared (SWIR) region covering 356 to 2577 nm
(Gabr et al. 2010). These attributes have been identified for the analysis of electronic
absorption features in different transitional metals, especially iron, and molecular
absorption features in carbonate, hydrate and hydroxide minerals (Zadeh et al. 2014).
The Hyperion data consist of 242 bands among which the first 70 bands include VNIR
and the second 172 bands in SWIR region. However, an overlap band exists in between
900 and 1000 nm (Pearlman et al. 2003, Pour and Hashim 2012, 2014). The reflectance
spectra of different minerals are subjugated in the VNIR region by the presence or
absence of transition metal ions (e.g. Fe, Cr, Co, Ni) which results in absorption features
due to electronic processes (Van Deer Meer and De Jong 2012). However, iron oxide
minerals, i.e. limonite, jarosite, and hematite have the spectral absorption features in the
visible to middle infrared from 400 to 1100nm of the electromagnetic spectrum (Hunt
and John 1974; Hunt and Ashley 1979). On the other hand, the presence of transition
elements viz. Fe2+, Fe3+ Mn, Cr, and Ni in the crystal structure of the minerals electronic
processes produce absorption features in the VNIR radiation (400 to 1100 nm) (Hunt and
Ashley 1979, Pour and Hashim 2011). Thus, the primary objective of hyperspectral
remote sensing is to quantitatively measure the constituents of the earth system from
calibrated spectra in narrow contiguous spectral bands. The quantification for mineral
mapping has been done by using the spectral absorption features to map the earth’s
surface. The spectral attributes of magnetite quartzite and the associated lithologies of
garnet ferrous reveal the absorption features around the ASTER spectral bands of 1, 3, 5,
and 7. On the basis of the absorption features of ASTER data, Rajendran et al. (2011)
used RGB colour composite image of ASTER band ratios i.e. (b1 + b3)/b2, (b3 + b5)/b4
and (b5 + b7)/b6 for the mineralogical mapping. However, Cudahy et al. (2002) also
proposed numerous band ratio techniques for mineralogical mapping. Mars and Rowan
(2010) has compared the ASTER image spectra with other spectral libraries to identify
the amendment areas which demonstrate spectrum matching and also for the mapping
of Argillic and Propylitic minerals. Therefore, in this study, Iron Ore depositional mapping
has been performed through ASTER, Hyperion, and in-situ data on the basis of band ratio
and spectral reflectance or emittance. The extracted Ore potential map has further been
validated through field spectra as well as ground truth sampling based on the accuracy
statistics.
 JOURNAL OF SPATIAL SCIENCE 3

2. Surface geological settings

Singhbhum Shear Zone (SSZ) is the largest iron ore province in Peninsular India which is
spread over the state of Odisha, West Bengal and Jharkhand. Joda is one of the largest
iron ore deposit regions in Odisha and also the major manufacturer of steel in India
which is located approximately at the 68.5 km north of the district headquarters
Keonjhar as depicted in Figure 1(a-e). The study area is bounded by 21º47′17.189″N to
22º9′35.621″N latitude and 85º13′36.247″E to 85º32′41.287″E longitude with an average
elevation of ~538m.
The SSZ of north Odisha is associated with banded iron ore formation (BIF) and is
surrounded by the iron ore series, which are imposed sporadically by the basic lavas and at
certain places by the ultrabasic rocks (Banerjee 1982, Murthy et al. 2009). The iron ore group
existing in the study area is classified into three BIFs which is associated with the periphery of
Archaean continental nucleus such as, the North Odisha Iron Ore Craton (NOIOC), the Older
Metamorphic Group (OMG) and the Older Metamorphic Tonalitic Gneiss (OMTG) (Satpathy
and Beura 2013, Deb 2014). The major surface geological formations of Joda region are BIF,

Figure 1. Study region considered for the iron ores investigation: (a) Generalized geological map in
and around Singhbhum Shear Zone exhibits the (i) Gondwana sedimentaries (G), (ii) Aeolian Sand
(AS), (iii) Alluvium (A), (iv) Basement Crystallines (BC), (v) Deformed metasedimentaries/metavolcanic
with minor granitoid and other intrusives (DM), (vi) Deltaic/Costal Sediment (CS), (vii) Greater/Trans/
lohit Himalayan mountains of Central Crystallines, Tethyan sedimentaries (G), (viii) Laterite and
Lateritic Soil (LS), (ix) Older sedimentaries (OS), and (x) Younger Sedimentaries (YS); (b) Regional
surface geological features in Joda region (modified after Thangavelu et al. 2011, Satpathy and Beura
2013); (c) False Colour Composition (FCC) of ASTER VNIR combination of band3, band2, and band1
exhibits the aerial extension of study area; (d) Surface exposure of iron ore alteration zone exhibits
on Google Earth Imagery; (e) FCC Image of Hyperion data (band 40, 31 and 13).
4 A. SENGUPTA ET AL.

Kolhars, Simhpal Group, Singhbhum Granite, and Older Meta Granite as depicted in Figure 1
(b). The lithological group of BIF includes the Banded Hematite Quartzite, the Banded
Hematite Jasper, the Banded Hematite Shale, the Banded Shale, the Banded Manganese
Formation, the Ferruginous Shale and the Iron Ore bodies. Banded Hematite Jasper and
Banded Hematite Quartzite are the predominant rock types in the study region characterized
by alternate bands of iron mineral and silica of varying thickness. High-grade hematite ores of
the iron ore group in the region are hosted by a lateral extension of 220 m thick BIF in a folded
greenstone belt succession of Paleoarchean age (Ogg et al. 2008). Single ore bodies which are
up to 3.0 km long along the strike and several hundred meters wide are strata bound
(Thangavelu et al. 2011). Banded shale shows bands of different colours and occurs at the
base of BIF. The ferruginous shales younger to BIF occur discontinuously in the area. The iron
mineral in the region is predominately oxide facies (James 1954). The iron oxide minerals are
hematite, magnetite, martite, specularite and goethite, whereas hematite is abundant and
predominant mineral (Beura and Singh 2005). The mineral assemblages reveal primary
sedimentary signatures and subsequently modified by diagenesis and metamorphism
(Acharya 1976, Beura and Singh 2005). The iron minerals are formed by the conventional
processes like martyrization, goethitization and hematitization. Hematite and Goethite is the
major mineral in the study region (Satpathy and Beura 2013). The entire meta-sedimentary
succession spans over 500 million years of supracrustal sedimentation history and is described
to have suffered from relentless thrusting, multiphase folding, and metamorphism. The rocks
were thrusted towards the Cratonic core, along with the prominent and laterally extensive
semi-circular SSZ which encompasses mainly Dhanjori quartzites and schists but the group
consists of basalt, conglomerate, arkose, quartzite and extensive lava flows. Kumar (1985)
explains the rock succession exposed in the southern Singbhum and Keonjhar which lay
unconformably over the older metamorphics named as iron ore series. Large deposits of iron
ore in Joda region of Keonjhar districts exhibiting 30–40% of iron content in ores of these
areas. The generalized stratigraphic succession of the Precambrian formations of Singbhum-
Orissa is illustrated in Table 1.

3. Methodology
3.1. Data
Most of the surface minerals exhibit diagnostic spectral signature in VNIR and SWIR of
the electromagnetic spectrum which enables their detection on the basis of character-
istics of spectral signature (Vicente and de Souza Filho 2011, Yitagesu et al. 2011,
Hewson et al. 2012). Therefore, in this study, the iron ore deposits and associated
alteration zones have been identified in Joda region using ASTER, Hyperion, and
ancillary data. Open file Hyperion Level 1R and Level 1T data from United States
Geological Survey (USGS), ASTER (VNIR & SWIR) Level 1B data from Land Processes
Distributed Active Archive Center (LPDAAC), and geological map (modified after Roy
and Venkatesh 2009, Thangavelu et al. 2011, Singh et al. 2016) have been used for the
purpose as detailed in Table 2. The Hyperion image consists of 224 spectral bands in the
wavelength ranges 356-2577 nm of the electromagnetic (EM) spectrum with the spectral
and spatial resolution of 10 nm and 30 m respectively. However, in this study, only 198
bands have been processed due to the poor signal to noise ratio. The Hyperion
 JOURNAL OF SPATIAL SCIENCE 5

Table 1. General stratigraphic succession of iron ore super group of Singhbhum-Odisha

(modified after Kumar, 1985; Acharya 2005, Thangavelu et al. 2011, Satpathy and Beura
2013, Beura et al. 2016).
Kolhan Group (1500-1600ma)
Unconformity
Newer Dolerite Arkasani Granophyre (Soda Granite)
Mayurbhanj Granite (2000–2100ma) Chakradharpur-Granite
Ultramafic Intrusives
Dhanjori Group
Dhanjori-Simplipal Lava (2100ma) Dalma Lava
Quartzites, Conglomerates
Unconformity
Singbhum Group
Dalbhum Formation
Chaibasa Formation
Unconformity
Singbhum Granite (2950ma)
Iron Ore Group
Upper Lava Formation
Upper Phyllite Formation
Banded Iron Ore Formation
Lower Phyllite Formation
Lower Lava Formation
Sandstone and Conglomerates
(local) Unconformity
Biotite-tonalite Gneiss (3200ma)
Older Metamorphic Group (3800ma)

hyperspectral sensor provides fine resolution and make possible to detect several
minerals viz. clays, carbonates, oxides, chlorites etc. (Magendran and Sanjeevi 2013a;
Zadeh et al. 2014). Several surface mineral mapping has been carried out with significant
accuracy which demonstrates the utility of Hyperion data (Jafari and Lewis 2012, Kusuma
et al. 2012, Farifteh et al. 2013). Geological map of Joda region was referred as litholo-
gical guide and also used in the geological field investigation for representative iron ore
sample collection.
A reconnaissance geological field investigation has been carried out in order to under-
stand the geological settings of the region as well as the exposed broad rock types, their
disposition, areal extent, soil patterns, etc. A spectrometer ‘AvaSpec-NIR256/512–1.7-EVO’
with artificial light source ‘Avasphere-50-LS-HAL’ has been used to collect the in-situ spectra
of iron ore over the 350 to 2500 nm range of the EM spectrum. In this study, about 100 iron
ore samples have been collected from different BIF units in the Joda region. The spectral

Table 2. Detail description of ASTER and Hyperion data used in this study for demarcation of iron ore
depositional zones in Joda region.
Attributes Hyperion Data ASTER Data
Entity ID EO1H1400452006016110PF_SGS_01 AST_L1B_00302012006045946
Acquisition Date 2006–16-01 2006–02-01
Acquisition Time 04:28:35(Greenwich Mean Time) 04:59:46 (Greenwich Mean Time)
Path 140 140
Row 40 127
Sun Azimuth 144.141461 148.030124
Sun Elevation 38.450226 44.749498
Spatial Resolution 30m VNIR(15m), SWIR(30m), TIR(90m)
6 A. SENGUPTA ET AL.

Figure 2. (a) Field photograph of iron ore samples (i.e. hematite and goethite), (b) Field-based
spectral profile of Hematite, and (c) Field-based spectral profile of Goethite taken during the field
investigation by a spectrometer instrument ‘AvaSpec-NIR256/512–1.7-EVO’over the 500 to 2500 nm
ranges of the EM spectrum.

profiles of collected samples exhibit maximum absorption from 850 nm to 900 nm and
maximum reflectance in the 1100–2500 nm wavelength regions as shown in Figure 2.

3.2. Iron ore mapping using ASTER and hyperion data

The pertinent mineralogical information on earth surface can easily be extracted through
spectral analyses of ASTER and Hyperion data. The region influenced by iron ore minerals
have been identified by ASTER data through band rationing techniques while Hematite
and Goethite mapping have been carried out by Hyperion data through spectral angle
mapper (SAM) algorithm. The spectra of diagnostic alteration iron ore have been extracted
from pre-processing imagery using ENVI’s n-dimensional visualizer and a priori knowledge
of geology (Zadeh et al. 2014). In order to obtain information about the spectral char-
acteristics of alteration iron ores, some representative field samples have been collected
and spectrally measured using a spectrometer instrument ‘AvaSpec-NIR256/512–1.7-EVO’.
The measured spectra are compared with the image-driven spectra, USGS spectral library
(Clark 1999), CRISM (NASA) spectral library (http://crismtypespctra.rsl.wustl.edu/) and spec-
tral signatures published by Singh et al. (2016). Thereafter, SAM has been implemented on
Hyperion data to discriminate the mineral species in the target areas based on the field
spectral profile. The spatial distribution of iron ore identified by the Hyperion and ASTER
 JOURNAL OF SPATIAL SCIENCE 7

data have been validated by the in-situ inspection and geological field observation as well
as examining their correspondence to geological and alteration maps of the study area.
The computational framework for iron ore alteration zone using ASTER and Hyperion data
are depicted in Figure 3.

3.2.1 Pre-processing of ASTER and hyperion data

The Hyperion and ASTER data have been used in this study for mapping of iron ore
alteration minerals. Pre-processing has been implemented on the ASTER and Hyperion
data in order to remove noise and to achieve the accurate surface reflectance (Zadeh
et al. 2014). Mars and Rowan (2010) suggested that if radiance and crosstalk correction
has been performed before the atmospheric correction more accurate ASTER reflectance
data can be obtained (Son et al. 2014). Therefore, the ASTER L1B data have been initially
corrected for crosstalk illumination effects by using the crosstalk correction modules
(Iwasaki and Tonooka 2005, Vicente and de Souza Filho 2011). On the other hand,
Hyperion data have been affected by different noise sources and had to be corrected
for abnormal pixels, striping, and smile, prior to application (Acito et al. 2011). The first

Figure 3. Computational protocol for detecting and mapping of iron ore alteration zones based on
ASTER and Hyperion data.
8 A. SENGUPTA ET AL.

70 VNIR and 172 SWIR bands of Hyperion data are affected by striping and bad line
columns and it is also apparent that an uncorrected striping effect leads to the flawed
elucidation of the result (Scheffler and Karrasch 2014). The vertical stripe or bad line
effect has been eliminated by replacing the DN value of the affected column by average
DN values of the adjacent columns through ‘Destriping modules’ using ENVI software
(Research Systems Inc 2003, Van Ede 2004, Zadeh et al. 2014) as depicted in Figure 4.
The crosstalk corrected ASTER L1B data have been converted to radiance values through
the sensor calibration gains as derived from the L1B metadata files (Caccetta et al. 2013).
On the other hand, by considering Hyperion L1T geocoded data, Hyperion L1R data has
been georectified and simultaneously, the selected ASTER L1B scenes have been recti-
fied and re-projected to geodetic coordinates. Thereafter, the Fast Line-of-sight
Atmospheric Analysis of Spectral Hypercubes (FLAASH) algorithm, available in ENVI
software has been applied to determine the surface reflectance data. This algorithm
was developed by spectral sciences, inc (Adler-Golden et al. 1999) under the sponsorship
of the U.S. Air Force Research Laboratory, is a MODTRAN4-based atmospheric correction
software package (Zadeh et al. 2014). It was designed to eliminate the atmospheric
effects through the derivation of atmospheric properties. Thus, both the ASTER and

Figure 4. Before and after de-stripping of band8 of Hyperion data with the spectral resolution of
426.82 nm.
 JOURNAL OF SPATIAL SCIENCE 9

Hyperion data have been atmospherically corrected through FLAASH module and con-
verted to reflectance by considering the following parameters viz. (i) digital elevation
model of the region, (ii) sensor altitude, (iii) data acquisition time, (iv) atmospheric
modules (v) aerosol modules, (vi) scene centre coordinates, and (vii) water vapour
content in the atmosphere (Vicente and de Souza Filho 2011).
Hyperion data also undergo from smile effect which primarily affects the bands in the
VNIR region. The smile, also known as ‘frown’ curve, is a spectral distortion that is
naturally found in push-broom sensors (Zadeh et al. 2014). It refers to an across track
wavelength shift from centre wavelength, which is due to the change of dispersion
angle with field position (Goodenough et al. 2003, Dadon et al. 2010, Zadeh et al. 2014;
Bhattacharya et al. 2012, San and Süzen 2011). The effect of smile is not observable in
individual bands. However, it becomes detectable when the image is transformed into
MNF space (Green et al. 1988). For Hyperion images with a significant smile, there is a
brightness gradient appearing in the first eigenvalue image. On the other hand, there is
no brightness gradient in MNF for images without a significant smile (Zadeh et al. 2014).
The smile effect may vary from one image to another. MNF algorithm implemented on
the Hyperion data exhibits no brightness gradient in MNF eigenvalue image as depicted
in Figure 5.
To minimize uncorrelated spatial noises, MNF has been used to separate the noises
(Green et al. 1988; Zadeh et al. 2014). The first few MNF bands generally convey the most
useful information, whereas subsequent bands gradually have higher noise. MNF bands
with calculated eigenvalues generally do not carry useful information and mainly contain
noise (Jensen 2007; Zadeh et al. 2014). In order to eliminate the noise and to get satisfying
results, MNF bands with eigenvalues more than 1.9 has been selected. Finally, the efficacy
of atmospheric correction and noise removal has been assessed by comparison of dense
green vegetation spectra extracted randomly from the image with the same spectra
obtained from a spectral library as depicted in Figure 6. The overall shape, including the
characteristic NIR region between 700 and 1300 nm, as well as absorption bands related
to chlorophyll (i.e. 498 and 680nm) and leaf water (i.e. 980, 1190 nm) are clearly evident in
the reduced Hyperion data.
These pre-processing procedures have the critical importance for the calibrating
pixels within the ASTER and Hyperion imagery as compared to USGS, NASA, field spectra
and reported spectral profiles. Finally, by artificially increasing the spatial resolution to
15 m of ASTER data, the resampling of SWIR bands has been accomplished which
simultaneously makes conceivable to work with the all nine VNIR and SWIR bands
(Vicente and de Souza Filho 2011).

3.2.2 Iron ore indicative alteration zone mapping using ASTER multispectral data
The various lithological mapping such as identification of granites (Watts and Harris
2005, Massironi et al. 2008, Son et al. 2014), ophiolite sequences (Van Deer Meer and De
Jong 2012) and basement rocks (Gad and Kusky 2007, Qari et al. 2008) on ASTER data are
widely available on several studies. Band ratios can be utilized to apprehend the
physiochemistry of the investigated material through which the depth and shape of
the mineral-related absorption features in VNIR, SWIR and TIR can be obtained. BR is a
technique where the DN value of one band is divided by the DN value of another band
and is very convenient for emphasizing certain features which cannot be visible in the
10 A. SENGUPTA ET AL.

Figure 5. MNF algorithm implemented on the Hyperion data which exhibits the first eigenvalue
image of MNF transformation with no brightness gradient.

raw bands (Inzana et al. 2003). In the recent years, it is observed that several researchers
for geological mapping have widely accepted ASTER BR technique (Gad and Kusky 2007,
Khan et al. 2007, Aboelkhair et al. 2010, Amer et al. 2010, Pour and Hashim 2011). The
AlOH group of minerals, which comprises albearing hydroxylated sheet silicates, such as
muscovite, illite, phengite, kaolinite, and Al smectite, were identified based on a band
ratio of (b5 + b7)/b6. In contrast, the AlOH group composition was measured by using a
ratio of b5/b7, and was masked to include those pixels where there is an adequate
amount of AlOH group of minerals. A comprehensive gestalt of different mineral indices
on ASTER multispectral data is described in ASTER mineral index processing manual as
compiled by Kalinowski and Oliver (2004) and it also recommends a range of band
combinations and false colour composites for the identification of various mineral
groups that highlight alteration intensity (Van Deer Meer and De Jong 2012). In this
study, emphasis has been given on ASTER data which highlights the presence of iron ore
zones in BIFs that contains (i) Ferric iron oxide minerals (e.g. hematite and goethite) and
(ii) Ferrous iron in silicate and/or carbonate minerals (e.g. actinolite, chlorite, and side-
rite).Therefore, as proposed by Van Deer Meer and De Jong (2012), we have used Ferric
 JOURNAL OF SPATIAL SCIENCE 11

Figure 6. Representative spectrum of dense vegetation as randomly extracted from the reflectance
Hyperion image (a-c) and compared with USGS spectral library exhibits the efficacy of atmospheric
correction and noise removal.

Iron, Ferrous Iron, Ferrous Silicates and Ferric Oxide Index i.e. band 2/band 1, band 5/
band 3, band 5/band 4 and band 4/band 3 respectively for the semi-automatic mapping
of iron ore deposit in the study area as depicted in Figure 7. Figure7(a-d) also depicts
that the high-intensity colour epitomizes the concentration of ferric iron, ferrous iron,
ferrous silicates and ferric oxide deposited areas.
The BRs threshold values have been identified through a correlation of pixel to pixel
information between Google Earth imageries, and subsequently the iron ore contained
pixel have been extracted based on the threshold of Ferric Iron, Ferrous Iron, Ferrous
Silicates, and Ferric Oxide Index value. The spatial distribution of ferric iron, ferrous iron,
ferrous silicates and ferric oxide in the study area is depicted in Figure 8 which provides
the accurate spectral information for vegetation, minerals, and lithological mapping.
Thus, the resulting map yields excellent spatial coherency and strong correlations
among the published geological maps of the study area. However, for the consistency
purposes, we have used ground truth and Google Earth imagery to exhibit the stepped
landscape features which are industrialized due to the extensive mining in Joda region.

3.2.3 Hematite and goethite mapping using hyperion data

Hyperion data has been widely used for geological and mineralogical mapping by
several researchers, such as Cudahy et al. (2001) evaluated the capacity of VNIR and
SWIR subsystems of Hyperion data for mineral mapping at Mountain Fitton, South
Australia, Kruse et al. (2003) used Hyperion SWIR (2000-2400nm) data to produce
12 A. SENGUPTA ET AL.

Figure 7. Demarcation of iron ore deposits through BR techniques using ASTER VNIR and SWIR data:
(a) Ferric Iron Index i.e. band 2/band 1, (b) Ferrous Iron Index i.e. band 5/band 3, (c) Ferrous Silicates
Index i.e. band 5/band 4, and (d) Ferric Oxides Index i.e. band 4/band 3.

convenient mineralogical information. Hubbard et al. (2003) used ALI, Hyperion, and
ASTER data for alteration mineral mapping in the Central Andes between Volcan
Socompa and Salar de Liullaillaco located in the border region between Chile and
Argentina. Hubbard and Crowley (2005) utilized ALI, ASTER, and Hyperion data for
mineral mapping in a volcanic terrain of the Chile-Bolivian Altiplano. Bishop et al.
(2011) used Hyperion and ASTER data for mineral mapping in the Pulang, Yunnan
Province, China. Zadeh et al. (2014) used EO-1 Hyperion data for Sub-pixel mineral
mapping of a porphyry copper belt region. Magendran and Sanjeevi (2013b) performed
Hyperion image analysis and linear spectral unmixing to evaluate the grades of Iron Ores
in parts of Noamundi, Eastern India. Mielke et al. (2014) used Hyperion data for
Spaceborne Mine Waste Mineralogy Monitoring in South Africa.
In this study, the spectral signatures have been extracted from the image and
compared with the field and published spectral library available on USGS, CRISM and
Singh et al. (2016). Thereafter, the spectral library resampling technique has been used
 JOURNAL OF SPATIAL SCIENCE 13

Figure 8. Spatial distribution of iron ore deposits based on the combination of Ferric Iron, Ferrous
Iron, Ferrous Silicates, and Ferric Oxide index; Subplots (a-d) exhibits stepped landscape feature
identified from Google Earth on the earth surface which is developed due to extensive mining in
Joda region.

to match the spectral signature which is extracted from the reflected image file over the
mineral spectral library of USGS, CRISM, Singh et al. (2016), and the one derived from the
field which results in library spectra to respond as the image spectral information. The
spectral analysis algorithm has been used to identify the unknown end-member spectra
of the image by matching with specific spectra of the mineral spectral library through
SAM. Since, the study area is a reserve forest and has the largest area of mines, therefore,
the end-members extracted by SAM has been found to be on the clusters of hematite
whose mineral form is an iron oxide (Fe2O3). As the large ore bodies of hematite are
usually of sedimentary origin, therefore it is widely distributed in the sedimentary rocks.
The red colouring of soils all over the study area is due to the hematite because it is also
an accessory mineral in igneous rocks. It is observed in Figure 9 that the soils where
hematite is the dominant iron oxide contains spectral absorptions nearly 404, 444, 529
and 880nm respectively while those having predominately goethite characteristics con-
tains absorptions nearly 434, 480, 650 and 920nm respectively. However, the similar
observation have also been found by Sherman and Waite (1985) and Rossel et al. (2009).
14 A. SENGUPTA ET AL.

During the recent years, the mapping of minerals and vegetative cover of the multi-
spectral dataset have generally been performed by classification and feature extraction
method. However, due to the high dimensionality, standard classification algorithm
cannot be applied on hyperspectral data. SAM is a physically based spectral classification
technique which uses an n-dimensional angle to match the pixels to reference spectra
and has been widely used as a measure of spectral similarity. The algorithm determines
the spectral resemblance between the reference reflectance spectrum and the test
spectrum by calculating the angle between the spectra and is equal to the number of
bands or wavelengths. However, the angle between two spectra is used as a measure of
discrimination where there is a comparison between the vector of each pixel and the
end-member spectra (Van Deer Meer and De Jong 2001). In this study, the SAM classifier
has been applied with a maximum angle of 0.10°, and the smaller angles represent
closer matches to the reference spectra, but the pixels which are further away from the

Figure 9. Atmospheric correction and false colour composite of Hyperion data where left side panel
represent the Goethite spectra of (a) image derived, (b) NASA CRISM spectral profile, (c) spectral
profile reported in Singh et al. (2016), (d) USGS spectral profile derive from the mineral library and
(e) field spectra; On the other hand, right side panel represent the Hematite spectra of (a′) image
derived, (b′) NASA CRISM spectral profile, (c′) spectral profile reported in Singh et al. (2016),
(d′) USGS spectral profile derive from the mineral library and (e′) in-situ/field spectra.
 JOURNAL OF SPATIAL SCIENCE 15

threshold are not classified. Since, this method is relatively insensitive to the illumination
effects, therefore, the MNF, PPI, and n-Dimensional Visualizer have been applied to
identify the Hematite and Goethite spectral end-member.
The extracted spectra have been used as a reference for subsequent processing. These
spectra determine a number of end-members, some of which could be attributed to iron ore
alteration zones. The extracted image spectra have been characterized using spectral
analysis procedure available at ENVI and visual inspection as well. These spectra were also
compared to the laboratory spectra of field samples and existing reference library spectra
such as USGS, CRISM and Singh et al. (2016) as depicted in Figure 9.Since SAM has been
applied to identify the altered iron ore zone based on the extracted spectral end-members
therefore, the spectral resampled library has been used as a reference to identify the end-
member spectra using ‘Spectral Analyst tool’ for detecting the Hematite and Goethite
deposited region as shown in Figure 10. The mineral map derived from Hyperion data
indicates the spatially coherent mineral distributions that are reliable with the geology map.

4. Result and discussion

This study demonstrates the feasibility of using ASTER and Hyperion data for discrimi-
nating the various grades of iron ore indicative alteration zone in and around Joda

Figure 10. Hematite and Goethite depositional zones identified in the study region based on the
user (image spectra), field spectra, and the reference spectra of USGS, CRISM &Singh et al. (2016)
through SAM algorithm; Subplots (a-d) represent Mines areas demarcated on Google Earth.
16 A. SENGUPTA ET AL.

mining areas. Hyperion data have been used to extract the mineralogical information
through SAM for identifying the hematite and goethite deposits in the western part of
the study area, while ASTER data have been used to determine the Iron Ore deposits
through different BR techniques of the entire study region. The results obtained exhibits
the potential uses of Hyperion and ASTER data for mineral mapping where the minerals
are exposed on the earth surface. Therefore, this study indicates the feasibility of
assessing the occurrences of iron ores with good accuracy by using the spectral para-
meters as derived from the atmospherically corrected Hyperion images and the available
spectral library. Furthermore, Ferric Iron, Ferrous Iron, Ferrous Silicates and Ferric Oxide
Index have been applied on ASTER data to identify the mineral abundance region of this
mining area as depicted in Figure 11. Satpathy and Beura (2013) carried out extensive
geological field investigation for the establishment of lithostratigraphy of some BIFs in
Joda region as depicted in Figure 12. The surface and subsurface strata exhibit a high
concentration of iron ore alteration zone which is accurately detected through the
ASTER and Hyperion data.
The accuracy of discriminated iron ore have been assessed by the spectral compar-
ison of collected samples with published one, checking the veracity of identified iron ore
by comparison to corresponding large-scale alteration maps of Joda region and per-
forming the accuracy statistics in between the classified and ground truth information.
The visual inspection of discriminated alteration iron ore exhibits a good correlation with
the classified alteration maps. Initially, the confirmation of detected iron ore has been
performed by visual interpretation of image-driven spectral signature and overall spec-
tral curve with the help of field spectra, USGS, CRISM, Singh et al. (2016) and Thangavelu
et al. (2011) mineral spectral library. In this study, about 100 iron ore samples have been
collected from different BIF units in the Joda region. The results of spectral reflectance
measurement exhibit maximum absorption from 850nm to 900nm and maximum
reflectance in the 1100-2500nm wavelength indicating the study region is covered
mostly of BIF (Thangavelu et al. 2011, Singh et al. 2016). On the other hand, spectral
analyst tool in ENVI software has been used for the spectral feature fitting and the binary
encoding to measure the similarity between the unknown end-member image spectra
and object spectra of the resampled spectral library (i.e. field spectra, USGS, CRISM,
Thangavelu et al. 2011, Singh et al. 2016) which exhibits 70–80% matching of the image
spectral information. Though, there is some variation identified between the field
spectra and library spectra due to the difference in quality of field samples. On the
other hand, dissimilarity is also been observed between the image spectra and field or
spectral library due to coarser spatial resolution of the image data. Since, in the present
investigation, the non-avialability of any portable spectroradiometer and geochemical
analysis therefore, to determined the accuracy of classified image a virtual verification
method has been adopted. This approach is successfully used earlier by Molan et al.
(2014) and Kumar et al. (2014) to assess the confidence level of classification results of
the mineral map.
A universal way of representing the confidence level for the assessment of remote-
sensing data is in the form of an error matrix which establishes a comparison between the
classified image and the validated areas through standard post classification techniques
(Congalton 1991). It is widely based on the accuracy assessment technique of statistical
correlations between two map data – one is the ‘reference’ and the other is the ‘classified’
 JOURNAL OF SPATIAL SCIENCE 17

Figure 11. Iron ore alteration zones in Joda region extracted from Hyperion and ASTER data.

data (Story and Congalton 1986, Jensen 2007). The correlation indicators used in this study
includes ‘overall accuracy’ i.e. the percentage of matched data between the ‘reference’ and
the ‘classified’ maps, ‘user’s accuracy’ i.e. the percentage of matched data in the ‘classified’
map, ‘producer’s accuracy’ i.e. the percentage of matched data in the ‘reference’ map, and
the kappa value defining a measure of differences between the ‘reference’ and the chance
agreement between both the maps (Congalton and Mead 1983, Jensen 2007). The kappa
statistics >0.80 suggests a ‘strong’ agreement and a value within a range of 0.60–0.80
suggests a ‘good’ agreement but the chance of understanding is remote while kappa is
close to 0 which indicates ‘poor’ agreement (Landis and Koch 1977, Jensen 2007). In this
study, the iron ore alteration zone mapped from Hyperion and ASTER data have been used
as ‘classified data’ while the ‘reference’ data which have been used for the accuracy
assessment have been collected from field investigation and Google Earth at 641 randomly
selected locations. During the field investigation, the extensive mining exploration has
been observed in the region those surface expressions are identified from remotely
18 A. SENGUPTA ET AL.

Figure 12. Subsurface lithological strata of (a) Jajang Langalota eastern part, (b) Baspani block, (c)
Jilling block, (d) Jajang Langalota Central part, and (e) Jajang Langalota western block in Joda region
(modified after Satpathy and Beura 2013).

through Google Earth imagery. Initially, we have collected some random ground control
point (GCP) of mines area and thereafter, superimposed the GCP on the Google Earth
imagery to understand the reliability of Google Earth data as a ground truth for accuracy
statistics. On the other hand, object oriented image segmentation technique has also been
used for the selection of GCP form Google Earth as compared to the standard FCC image.
Figure 13 depicts some representative mining area as captured from Google Earth, FCC and
Field in and around Joda region and the accuracy statistics between the ‘reference’ and the
‘classified’ map as illustrated in Table 3 establish the confidence level of this mineralogi-
cal map.

5. Conclusion
This study establishes the likelihood of ASTER and Hyperion data for discerning the
various grades of iron ores alteration zone spread over the Joda mining area. ASTER data
have been utilized to locate the target areas characterized by iron ore, while Hyperion
data have been used for the comprehensive mineralogical mapping. It is observed from
the field spectra that the range of spectra has been used to categorize the well-defined
NIR (750–1000nm) and SWIR (2150–2250nm) absorption features of iron oxide. The
demarcation of the potential iron ore alteration zone has been performed through
SAM algorithm and BR techniques. It is observed that the classified iron ore alteration
distribution map has a strong agreement with the reference data (i.e. k = 0.826). Thus,
the results indicate that the combination of multispectral and hyperspectral data is very
much beneficial for evaluating the occurrences of iron ores with good precision and
mapping of mineral zones on the earth’s surface.
 JOURNAL OF SPATIAL SCIENCE 19

Figure 13. Some representative mining area randomly captured from (a) multispectral, (b) Google
Earth imagery, and (c) Glimpse of iron ore samples photographs collected from Joda mining region.

Table 3. Error matrix derived for assessment of confidence level of remotely based identified iron ore
alteration zones in Joda region.
Iron Ore alteration based on Google Earth
image and Field data User’s accuracy (%)
Iron Ore alteration mapping Iron Oxide Others class Total
using Remote Sensing data Iron Oxide 399 35 434 91.93
Others class 31 176 207 7.24
Total 414 227
Producer’s accuracy (%) 96.37 15.41
Overall accuracy (%) 92.19
Kappa value 0.826
Kappa variance 0.001

Acknowledgments
We greatly appreciate the tremendous support received from the Editorial Board of JSS during the
review process. The anonymous reviewers critically examined the manuscript and suggesting
several valuable changes that greatly helped in bringing up the manuscript to its present shape
with enhanced quality and scientific exposition. The Editor in Chief Prof. G. L. Wright also critically
examined the manuscript and suggested several valuable changes that helped in scientific value
addition and greater exposition.

Disclosure statement
No potential conflict of interest was reported by the authors.
20 A. SENGUPTA ET AL.

Funding
This work was supported by ISIRD (SRIC, IIT Kharagpur).

References
Aboelkhair, H., et al., 2010. Processing and interpretation of ASTER TIR data for mapping of rare-
metal-enriched albitegranitoids in the Central Eastern Desert of Egypt. Journal of African Earth
Sciences, 58 (1), 141–151. doi:10.1016/j.jafrearsci.2010.01.007
Acharya, S., 1976. Iron formation and iron ores of Orissa, their stratigraphy and correlation. In:
Proceeding of symposium on geological exploring mining mineral processing and metallurgical
ferrous and ferroalloy minerals. Bangalore, India: Bangalore University, 86–100.
Acharya, S., 2005. Genetic modeling of iron and manganese deposits of the Joda-Koira Iron Ore
basin (BIF-3). In: India-its application to exploration. Sem. Proc. Vision Mineral development
2020 organised by SGAT. Bhubaneswar, India: SGAT, 44–67.
Acito, N., Dhani, M., and Corsini, G., 2011. Subspace-based striping noise reduction in hyperspectral
images. IEEE Transaction on Geoscience and Remote Sensing, 49 (4), 1325–1342. doi:10.1109/
TGRS.2010.2081370
Adler-Golden, S.M., et al., 1999. Atmospheric correction for short-wave spectral imagery based on
MODTRAN 4. In Imaging Spectrometry V. International Society for Optics and Photonics, 3753,
61–70.
Ahmed, Z. and Iqbal, J., 2014. Evaluation of Landsat TM5 multispectral data for automated
mapping of surface soil texture and organic matter in GIS. European Journal of Remote
Sensing, 47 (1), 557–573. doi:10.5721/EuJRS20144731
Amer, R., Kusky, T., and Ghulam, A., 2010. Lithological mapping in the Central Eastern desert of
Egypt using ASTER data. Journal of African Earth Sciences, 56 (2), 75–82. doi:10.1016/j.
jafrearsci.2009.06.004
Banerjee, P.K., 1982. Stratigraphy, petrology, and geochemistry of some Precambrian basic volca-
nic and associated rocks of Singhbhum District, Bihar, and Mayurbhanj and Keonjhar districts,
Orissa. Geological Survey of India, 111, 1–54.
Beura, D., et al., 2016. Field relationship among the three iron ore groups of iron ore super group
encircling the North Odisha Iron Ore Craton, India: a comparison study. Journal of Geosciences
and Geomatics, 4 (3), 53–60.
Beura, D. and Singh, P., 2005. Geological setting and mineral deposits of Archaean schist belt-A
case study around Badampahar belt, North Orissa, India. In: Proceeding of international seminar,
6–8 September Thailand. Khon Ken, Thailand: Khon Ken University, 326–329.
Bhattacharya, S., et al., 2012. Utilization of hyperion data over Dongargarh, India, for mapping
altered/weathered and clay minerals along with field spectral measurements. International
Journal of Remote Sensing, 33 (17), 5438–5450. doi:10.1080/01431161.2012.661094
Bishop, C.A., Liu, J.G., and Mason, P.J., 2011. Hyperspectral remote sensing for mineral exploration
in Pulang Yunnan Province China. International Journal of Remote Sensing, 32 (9), 2409–2426.
doi:10.1080/01431161003698336
Caccetta, M., Collings, S., and Cudahy, T., 2013. A calibration methodology for continental scale
mapping using ASTER imagery. Remote Sensing of Environment, 139, 306–317. doi:10.1016/j.
rse.2013.08.011
Clark, R.N., 1999. Spectroscopy of rock and minerals and principles of spectroscopy. Manual of
Remote Sensing, 3, 3–58.
Congalton, R.G., 1991. A review of assessing the accuracy of classifications of remotely sensed
data. Remote Sensing of Environment, 37 (1), 35–46. doi:10.1016/0034-4257(91)90048-B
Congalton, R.G. and Mead, R.A., 1983. A quantitative method to test for consistency and correct-
ness in photointerpretation. Photogrammetric Engineering and Remote Sensing, 49 (1), 69–74.
Cudahy, T.J., et al., 2001. The performance of the satellite-borne Hyperion Hyperspectral VNIR-SWIR
imaging system for mineral mapping at Mount Fitton, South Australia. In: Proceedings of IEEE
 JOURNAL OF SPATIAL SCIENCE 21

2001 International Geoscience and Remote Sensing Symposium (Cat. No.01CH37217). Sydney: IEEE,
314–316.
Cudahy, T.J., et al., 2002. Regional to prospect scale exploration for porphyry-skarn-epithermal
mineralisation at Yerington, Nevada, using ASTER and airborne hyperspectral data. Australia:
CSIRO Exploration and Mining Report.
Dadon, A., Ben-Dor, E., and Karnieli, A., 2010. Use of derivative calculations and minimum noise
fraction transform for detecting and correcting the spectral curvature effect (smile) in Hyperion
images. IEEE Transactions on Geoscience and Remote Sensing, 48 (6), 2603–2612. doi:10.1109/
TGRS.2010.2040391
Deb, M., 2014. Precambrian geodynamics and metallogeny of the Indian shield. Ore Geology
Reviews, 57, 1–28. doi:10.1016/j.oregeorev.2013.08.022
Ducart, D.F., et al., 2016. Mapping iron oxides with Landsat-8/OLI and EO-1/Hyperion imagery from
the Serra Norte iron deposits in the Carajás Mineral Province, Brazil. Brazilian Journal of Geology,
46 (3), 331–349. doi:10.1590/2317-4889201620160023
Farifteh, J., Nieuwenhuis, W., and García-Meléndez, E., 2013. Mapping spatial variations of iron
oxide by-product minerals from EO-1 hyperion. International Journal of Remote Sensing, 34 (2),
682–699. doi:10.1080/01431161.2012.715776
Gabr, S., Ghulam, A., and Kusky, T., 2010. Detecting areas of high-potential gold mineralization
using ASTER data. Ore Geology Reviews, 38 (1), 59–69. doi:10.1016/j.oregeorev.2010.05.007
Gad, S. and Kusky, T., 2007. ASTER spectral ratioing for lithological mapping in the Arabian-Nubian
shield, the Neoproterozoic Wadi Kid area, Sinai, Egypt. Gondwana Research, 11 (3), 326–335.
doi:10.1016/j.gr.2006.02.010
Goodenough, D.G., et al., 2003. Processing hyperion and ALI for forest classification. IEEE Transactions
on Geoscience and Remote Sensing, 41 (6), 1321–1331. doi:10.1109/TGRS.2003.813214
Green, A.A., et al., 1988. A transformation for ordering multispectral data in terms of image quality
with implications for noise removal. IEEE Transactions on Geoscience and Remote Sensing, 26 (1),
65–74. doi:10.1109/36.3001
Hewson, R.D., et al., 2012. Investigations into soil composition and texture using infrared spectro-
scopy (2–14 m). Applied and Environmental Soil Science, 2012, 1–12. doi:10.1155/2012/535646
Hubbard, B.E. and Crowley, J.K., 2005. Mineral mapping on the Chilean–bolivian Altiplano using co-
orbital ALI, ASTER and Hyperion imagery: data dimensionality issues and solutions. Remote
Sensing of Environment, 99 (1), 173–186. doi:10.1016/j.rse.2005.04.027
Hubbard, B.E., Crowley, J.K., and Zimbelman, D.R., 2003. Comparative alteration mineral mapping using
visible to shortwave infrared (0.4-2.4/spl mu/m)) Hyperion ALI, and ASTER imagery. IEEE Transactions
on Geoscience and Remote Sensing, 41 (6), 1401–1410. doi:10.1109/TGRS.2003.812906
Hunt, G.R. and Ashley, R.P., 1979. Spectra of altered rocks in the visible and near infrared. Economic
Geology, 74 (7), 1613–1629. doi:10.2113/gsecongeo.74.7.1613
Hunt, G.R. and John, W.S., 1974. Mid-infrared spectral behavior of igneous rocks (No. AFCRL-TR-74-
0625). Hanscom AFB, MA: Air Force Cambridge Research Labs.
Inzana, J., et al., 2003. Supervised classifications of Landsat TM band ratio images and Landsat TM
band ratio image with radar for geological interpretations of central Madagascar. Journal of
African Earth Sciences, 37 (1), 59–72. doi:10.1016/S0899-5362(03)00071-X
Iwasaki, A. and Tonooka, H., 2005. Validation of a crosstalk correction algorithm for ASTER/SWIR. IEEE
Transactions on Geoscience and Remote Sensing, 43 (12), 2747–2751. doi:10.1109/TGRS.2005.855066
Jafari, R. and Lewis, M.M., 2012. Arid land characterisation with EO-1 Hyperion hyperspectral data.
International Journal of Applied Earth Observation and Geoinformation, 19, 298–307. doi:10.1016/
j.jag.2012.06.001
James, H.L., 1954. Sedimentary facies of iron-formation. Economic Geology, 49 (3), 235–293.
doi:10.2113/gsecongeo.49.3.235
Jensen, J.R., 2007. Remote sensing of the environment: an earth resource perspective 2/e. London:
Pearson Education India.
Kalinowski, A. and Oliver, S., 2004. ASTER mineral index processing manual. Remote Sensing
Applications, Geoscience Australia, 37, 36.
22 A. SENGUPTA ET AL.

Khan, S.D., Mahmood, K., and Casey, J.F., 2007. Mapping of Muslim Baghophiolite complex
(Pakistan) using new remote sensing and field data. Journal of Asian Earth Sciences, 30 (2),
333–343. doi:10.1016/j.jseaes.2006.11.001
Kruse, F.A., Boardman, J.W., and Huntington, J.F., 2003. Comparison of airborne hyperspectral data
and EO-1 Hyperion for mineral mapping. IEEE Transactions on Geoscience and Remote Sensing,
41 (6), 1388–1400. doi:10.1109/TGRS.2003.812908
Kumar, C., et al., 2014. Sub-pixel mineral mapping using EO-1 Hyperion hyperspectral data. The
International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 40
(8), 455. doi:10.5194/isprsarchives-XL-8-455-2014
Kumar, R., 1985. Fundamentals of historical geology and stratigraphy of India. New York, NY: Wiley.
Kusuma, K.N., Ramakrishnan, D., and Pandalai, H.S., 2012. Spectral pathways for effective delinea-
tion of high-grade bauxites: a case study from the Savitri River Basin, Maharashtra, India, using
EO-1 Hyperion data. International Journal of Remote Sensing, 33 (22), 7273–7290. doi:10.1080/
01431161.2012.700131
Landis, J.R. and Koch, G.G., 1977. An application of hierarchical kappa-type statistics in the
assessment of majority agreement among multiple observers. Biometrics, 363–374.
doi:10.2307/2529786
Magendran, T. and Sanjeevi, S., 2013a. A study on the potential of satellite image-derived
hyperspectral signatures to assess the grades of iron ore deposits. Journal of the Geological
Society of India, 82 (3), 227–235. doi:10.1007/s12594-013-0145-0
Magendran, T. and Sanjeevi, S., 2013b. Hyperion image analysis and linear spectral unmixing to
evaluate the grades of iron ores in parts of Noamundi, Eastern India. International Journal of
Applied Earth Observation and Geoinformation, 26, 413–426. doi:10.1016/j.jag.2013.09.004
Mars, J.C. and Rowan, L.C., 2010. Spectral assessment of new ASTER SWIR surface reflectance data
products for spectroscopic mapping of rocks and minerals. Remote Sensing of Environment, 114
(9), 2011–2025. doi:10.1016/j.rse.2010.04.008
Massironi, M., et al., 2008. Interpretation and processing of ASTER data for geological mapping and
granitoids detection in the Saghro massif (eastern Anti-Atlas, Morocco) interpretation of ASTER
data for geological mapping and granitoids detection. Geosphere, 4 (4), 736–759. doi:10.1130/
GES00161.1
Mia, M.B. and Fujimitsu, Y., 2012. Mapping hydrothermal altered mineral deposits using Landsat 7
ETM+ image in and around Kuju volcano, Kyushu, Japan. Journal of Earth System Science, 121 (4),
1049–1057. doi:10.1007/s12040-012-0211-9
Mielke, C., et al., 2014. Spaceborne mine waste mineralogy monitoring in South Africa, applications
for modern push-broom missions: hyperion/OLI and EnMAP/Sentinel-2. Remote Sensing, 6 (8),
6790–6816. doi:10.3390/rs6086790
Moghtaderi, A., Moore, F., and Mohammadzadeh, A., 2007. The application of advanced space-
borne thermal emission and reflection (ASTER) radiometer data in the detection of alteration in
the Chadormalu paleocrater, Bafq region, Central Iran. Journal of Asian Earth Sciences, 30 (2),
238–252. doi:10.1016/j.jseaes.2006.09.004
Molan, Y.E., Refahi, D., and Tarashti, A.H., 2014. Mineral mapping in the Maherabad area, eastern
Iran, using the HyMap remote sensing data. International Journal of Applied Earth Observation
and Geoinformation, 27, 117–127. doi:10.1016/j.jag.2013.09.014
Murthy, B., et al., 2009. Geophysical exploration for manganese-some first hand examples from
Keonjhar district, Orissa. Journal Industrial Geophys Union, 13 (3), 149–161.
Ogg, J.G., Ogg, G., and Gradstein, F.M., 2008. The concise geologic time scale. Cambridge:
Cambridge University Press.
Pearlman, J.S., et al., 2003. Hyperion, a space-based imaging spectrometer. IEEE Transactions on
Geoscience and Remote Sensing, 41 (6), 1160–1173. doi:10.1109/TGRS.2003.815018
Pour, A.B. and Hashim, M., 2011. Identification of hydrothermal alteration minerals for exploring of
porphyry copper deposit using ASTER data, SE Iran. Journal of Asian Earth Sciences, 42 (6), 1309–
1323. doi:10.1016/j.jseaes.2011.07.017
Pour, A.B. and Hashim, M., 2012. The application of ASTER remote sensing data to porphyry copper
and epithermal gold deposits. Ore Geology Reviews, 44, 1–9. doi:10.1016/j.oregeorev.2011.09.009
 JOURNAL OF SPATIAL SCIENCE 23

Pour, A.B. and Hashim, M., 2014. ASTER, ALI and hyperion sensors data for lithological mapping
and ore minerals exploration. Springerplus, 3 (1), 130. doi:10.1186/2193-1801-3-130
Qari, M.H.T., et al., 2008. Utilization of Aster and Landsat data in geologic mapping of basement
rocks of Arafat Area, Saudi Arabia. Arabian Journal Sciences Engineering, 33, 99–116.
Raj, S.K., et al., 2015. Iron oxides mapping from E0-1 Hyperion data. Journal of the Geological
Society of India, 86 (6), 717–725. doi:10.1007/s12594-015-0364-7
Rajendran, S., et al., 2011. Discrimination of iron ore deposits of granulite terrain of Southern
Peninsular India using ASTER data. Journal of Asian Earth Sciences, 41 (1), 99–106. doi:10.1016/j.
jseaes.2011.01.004
Research Systems Inc., 2003. ENVI tutorial, ENVI software package version 4.0. Boulder, CO: Research
Systems, Inc.
Rossel, R.V., et al., 2009. In situ measurements of soil colour mineral composition and clay content
by vis–NIR spectroscopy. Geoderma, 150 (3), 253–266. doi:10.1016/j.geoderma.2009.01.025
Roy, S. and Venkatesh, A.S., 2009. Mineralogy and geochemistry of banded iron formation and iron
ores from eastern India with implications on their genesis. Journal of Earth System Science, 118
(6), 619–641. doi:10.1007/s12040-009-0056-z
San, B.T. and Süzen, M.L., 2011. Evaluation of cross-track illumination in EO-1 Hyperion imagery for
lithological mapping. International Journal of Remote Sensing, 32 (22), 7873–7889. doi:10.1080/
01431161.2010.532175
Satpathy, B. and Beura, D., 2013. Establishment of lithostratigraphy of some banded iron forma-
tions of iron ore super group of Odisha, India. International Journal of Engineering and Innovative
Technology, 2 (7), 146–150.
Scheffler, D. and Karrasch, P., 2014. Destriping of hyperspectral image data: an evaluation of
different algorithms using EO-1 Hyperion data. Journal of Applied Remote Sensing, 8 (1),
083645. doi:10.1117/1.JRS.8.083645
Sherman, D.M. and Waite, T.D., 1985. Electronic spectra of Fe3+ oxides and oxide hydroxides in the
near IR to near UV. American Mineralogist, 70 (11-12), 1262–1269.
Singh, M., et al., 2016. Spectral characteristics of banded iron formations in Singhbhum craton,
eastern India: implications for hematite deposits on Mars. Geoscience Frontiers, 7 (6), 927–936.
doi:10.1016/j.gsf.2015.11.003
Son, Y.S., Kang, M.K., and Yoon, W.J., 2014. Lithological and mineralogical survey of the Oyu Tolgoi
region, Southeastern Gobi, Mongolia using ASTER reflectance and emissivity data. International
Journal of Applied Earth Observation and Geoinformation, 26, 205–216. doi:10.1016/j.
jag.2013.07.004
Story, M. and Congalton, R.G., 1986. Accuracy assessment: a user’s perspective. Photogrammetric
Engineering and Remote Sensing, 52 (3), 397–399.
Thangavelu, M., Shanmugam, S., and Bhattacharya, A.K., 2011. Hyperspectral radiometry to quan-
tify the grades of iron ores of Noamundi and Joda mines, Eastern India. Journal of the Indian
Society of Remote Sensing, 39 (4), 473–483. doi:10.1007/s12524-011-0109-z
Van Deer Meer, F.D., and De Jong, S.M., eds. 2001. Imaging spectrometry: basic principles and
prospective applications. Dordrecht: Springer.
Van Deer Meer, F. and De Jong, S.M., 2012. Multi-and hyperspectral geologic remote sensing: A
review. International Journal of Applied Earth Observation and Geoinformation, 14 (1), 112–128.
doi:10.1016/j.jag.2011.08.002
Van Ede, R., 2004. Destriping and geometric correction of an ASTER level 1a image. Ultrecht,
Netherlands: Department of Physical Geography, Faculty of GeoSciences, Ultrecht University.
Vicente, L.E. and de Souza Filho, C.R., 2011. Identification of mineral components in tropical soils
using reflectance spectroscopy and advanced spaceborne thermal emission and reflection
radiometer (ASTER) data. Remote Sensing of Environment, 115 (8), 1824–1836. doi:10.1016/j.
rse.2011.02.023
Watts, D.R. and Harris, N.B., 2005. Mapping granite and gneiss in domes along the North
Himalayan antiform with ASTER SWIR band ratios. Geological Society of America Bulletin, 117
(7–8), 879–886. doi:10.1130/B25592.1
24 A. SENGUPTA ET AL.

Yitagesu, F.A., et al., 2011. Spectral characteristics of clay minerals in the 2.5–14μm wavelength
region. Applied Clay Science, 53 (4), 581–591. doi:10.1016/j.clay.2011.05.007
Zadeh, M.H., et al., 2014. Sub-pixel mineral mapping of a porphyry copper belt using EO-1
hyperion data. Advances in Space Research, 53 (3), 440–451. doi:10.1016/j.asr.2013.11.029
Zhang, X., Pazner, M., and Duke, N., 2007. Lithologic and mineral information extraction for gold
exploration using ASTER data in the south chocolate Mountains (California). ISPRS Journal of
Photogrammetry and Remote Sensing, 62 (4), 271–282. doi:10.1016/j.isprsjprs.2007.04.004