Salamba 2019 IOP Conf. Ser. Earth Environ. Sci. 254 012004

IOP Conference Series: Earth and Environmental Science
PAPER • OPEN ACCESS You may also like

- A comparison of radar and optical remote
Identification of alteration zones using a Landsat 8 sensing to detect cyclone-induced canopy
disturbance in two subtropical forest
image of densely vegetated areas of the Wayang landscapes
Jonathan Peereman, Soyeon Bae and
Teng-Chiu Lin
Windu Geothermal field, West Java, Indonesia
- Fault-related fractures characteristic of
Kijang fault at Wayang Windu Geothermal
To cite this article: Kristian Edwin Salamba et al 2019 IOP Conf. Ser.: Earth Environ. Sci. 254 012004 field
Wahyuddin Diningrat, Shanti R A Sugiono
and Yunus Daud
- A New Strategy for Comparison of Land

Surface Temperature Retrieval Methods
View the article online for updates and enhancements. with Landsat Remote Sensing Images
Considering Regional Consistency
Xu Qingyu, Xu Xiong, Xie Huan et al.
This content was downloaded from IP address 103.111.141.191 on 15/01/2023 at 21:28

7th ITB International Geothermal Workshop (IIGW2018) IOP Publishing
IOP Conf. Series: Earth and Environmental Science 254 (2019) 012004 doi:10.1088/1755-1315/254/1/012004
Identification of alteration zones using a Landsat 8 image

of densely vegetated areas of the Wayang Windu
Geothermal field, West Java, Indonesia
Kristian Edwin Salamba1, Arie Naftali Hawu Hede2, Mohamad Nur Heriawan2
1
Mining Engineering Study Program, Faculty of Mining and Petroleum Engineering,
Institut Teknologi Bandung
2
Earth Resources Exploration Research Group, Faculty of Mining and Petroleum
Engineering, Institut Teknologi Bandung
Jl. Ganesha No.10, Bandung, West Java, 40132, Indonesia.
Email: [email protected]
Abstract. Remote sensing is a method usually used for conducting geothermal reconnaissance
surveys by mapping surface alteration anomalies, which are captured by a satellite sensor. This
study analyzes the characteristics of the surface alteration zones in the Wayang Windu
geothermal field (WWGF). It employs a remote sensing method using multispectral Landsat 8
imagery and is validated by ground truth data from field surveys. The WWGF is located in
Pangalengan, West Java with an elevation of 1500–2600 m.a.s.l., and lies in a quaternary
volcanic arc. The rock types in Wayang Windu consist of andesite, basalt, tuff, breccia, and
pumice. Fractures and faults are identified as lineaments in this area and based on their structures,
were directed to be oriented northwest–southeast and northeast–southwest. This research
combined a field survey and remote sensing methods to enhance the spatial data. Field surveys
yield 18 spots for obtaining soil samples and laboratory analyses were performed. Spectral
reflectance analysis was performed to determine the reflectance and mineral composition of the
samples, X-ray diffraction was performed to determine the mineral composition, and X-ray
fluorescence was performed to determine the abundance of elements. A scene from Landsat 8
image acquired on September 10, 2013 was evaluated using a principal component analysis-
based method. The surface alteration zones in the WWGF correlated with joints and faults. Based
on the mineral composition, the surface alteration zones in WWGF were identified as advanced
argillic zone with the occurrence of secondary minerals such as cristobalite and halloysite, and
a propylitic zone with the occurrence of secondary minerals such as epidote and chlorite.
1. Introduction
Remote sensing is a method used in reconnaissance surveys. This method can map surface alteration
anomalies by capturing the spectral reflectance of surface minerals. This method can periodically
observe large areas, but it has limitations in tropical conditions, which have high precipitation and dense
vegetation [1][2][3][4]. To overcome this problem, this research used principal component analysis
(PCA) to enhance mineral spectral reflectance and targets by reducing the impact of atmospheric and
vegetation conditions [1][5].
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The Wayang Windu geothermal field (WWGF) is located in Pangalengan, West Java, Indonesia
(Figure 1) and has a tropical climate with high precipitation and dense vegetation. It is a geothermal
system at the transition between a liquid-dominated and a vapor-dominated system. The reservoir
consists of the WWGF associated with Mt. Wayang–Windu in the south and Mt. Gambung–Puncak
Besar in the north. Geothermal manifestations in this field include fumaroles, steaming and altered
ground, and acid-sulfate springs [6].
Figure 1. Morphology of the WWGF based on a Landsat 8 image acquired on Septembter 10,
2013 and Shuttle Radar Topographic Mission 1 arc-second.
2. Data and Methodology
2.1. Data
This research used a Landsat 8 image with an acquisition date of September 10, 2013. The image
captured spectral reflectance from path 122 and row 65, sun azimuth 66.27º, sun elevation 60.70º, and
scene center time of 03:02:29 GMT. The spatial resolution is 30 m x 30 m. In addition, this research
used ground truth from surface soils gathered from 18 points at 9 km x 8 km area to validate the PCA.
2.2. Methodology
Directed PCA (DPCA) is a method that enhances imagery based on the differences between two band
ratios [1][5]. The input band ratio images are selected on the basis that one band ratio contains
information related to the component of interest (hydrothermal alteration in this case). The second band
ratio should contain information about a spectrally interfering component, which is dense vegetation in
this case. The DPCA is based on the examination of principal component (PC) eigenvector loadings to
decide which of the PC images will concentrate information related directly to the theoretical spectral
signatures of specific target materials. The technique is able to predict whether the target material is
represented by bright or dark pixels in the PC images according to the magnitude and sign of the
eigenvectors. The DPCA was applied to Landsat 8 imagery in the WWGF to identify limonitic and clay
materials and also to map the alteration zones by integrating with geographic information system-based
image classification.
2
3. Result and Discussion
3.1. Surface Alteration Zone based on PCA

Landsat 8 imagery has 11 bands acquired by both the Operational Land Imager (OLI) and Thermal
Infrared Sensor (TIRS). Bands in Landsat 8 OLI capture 9 bands from visible near-infrared (VNIR) and
shortwave infrared (SWIR). The spectral bands used for this study are bands 2−7 to identify surface
material (Figure 2).
Figure 2. Spectral reflectance from minerals and vegetation. Each

range of the Landsat 8 band is shown in the column.
The Landsat 8 image was preprocessed using ENVI 5.3 software and the FLAASH algorithm.
Preprocessing included the identification of regions of interest, radiometric calibration, and atmospheric
correction. The purpose of the preprocessing procedure was to reduce noise in the imagery. After
preprocessing, imagery was processed using the DPCA to determine alteration zones in the surface of
the WWGF.
The DPCA was used to map clay minerals and limonitic minerals in dense vegetation zones by using
two-band ratios [5]. The limonitic zone is represented by band ratio 4/2, and the dense vegetation zone
is represented by band ratio 5/4 (Table 1). The clay minerals zone is represented by band ratio 6/7, and
the dense vegetation zone is represented by band ratio 5/4 (Table 2).
Table 1. Eigenvector analysis of the two-band ratio of Landsat 8 image for limonitic minerals. The
principal component (in bold) represents limonitic minerals.
Eigenvectors
Principal Band ratio 4/2 Band ratio 5/4 Eigenvalues
Components (band ratio of (band ratio of (%)
limonitic minerals) vegetation)
PC 1 0.01 -0.99 18.77
PC 2 -0.99 -0.01 0.08
3
Table 2. Eigenvectors analysis of the two-band ratio of Landsat 8 image for clay minerals. The
principal component (in bold) represents clay minerals.
Eigenvectors
Principal Band ratio 5/4 Band ratio 6/7 Eigenvalues
Components (band ratio of (band ratio of (%)
vegetation) kaolinite)
PC 1 0.99 0.07 18.87
PC 2 -0.07 0.99 0.02
In the WWGF, clay minerals occur as plain ramps (Figure 3). Their occurrence can be related to land
use and is mostly found in plantation or paddy fields, which consist of clay materials. Clayey soils can
easily hold water and become saturated. Limonitic layered in the WWGF occurs in steep terrain,
hillsides, and bare soil.
Figure 3. Clay and limonitic layered distributions in the

WWGF.
DPCA can be combined with GIS-based image classification to analyze the distribution of
hydrothermal alteration zones and altered hydrothermal minerals and hydrothermal alteration zones in
dense vegetation area [1]. The combined technique uses a two-band ratio between the altered mineral
band and the vegetation band. The ratio band used for DPCA of each altered mineral shown in Table 3.
4
Table 3. Band ratio for DPCA in Landsat 8 image.

Band ratio for DPCA in Landsat 8
image
Altered Mineral
Vegetation band Mineral band
ratio ratio
Quartz band 3 / band 4 band 7 / band 2
Alunite band 5 / band 2 band 6 / band 7
Kaolinite band 5 / band 4 band 6 / band 7
Chlorite band 5 / band 3 band 6 / band 2
Epidote band 3 / band 4 band 6 / band 2
In this research, the selection of altered minerals is identified based on the mineral assembly of the
hydrothermal alteration zone. The ratio band that is used in this research is based on the USGS Spectral
Library and contrasts spectral reflectance between mineral and vegetation. Geological setting based on
geological of the Garut and Pameungpeuk quadrangle, Java [7]. The advanced argillic assembly is
represented by secondary minerals: quartz, alunite, and kaolinite; the propylitic assembly is represented
by secondary minerals: chlorite and epidote. Advanced argillic and propylitic assemblies were chosen
based on previous research [8].
Epidote and chlorite were mostly distributed in Malabar-Tilu volcanic (Qmt) formations (Figure 4a)
which consist of tuff and laharic breccia containing minor amounts of pumice and lavas. Other rock
formations such as young volcanic (Qyw), undifferentiated efflata deposits of old volcanic (Qopu),
Waringin-Bedil andesite, and old Malabar (Qwb) formations, feature lower distributions of epidote and
chlorite.
Kaolinite and alunite are mostly distributed in formations of Qopu (Figure 4b). The Qopu formations
consist of fine to coarse dacitic crystalline tuff and tuffaceous breccia containing pumices and old
andesitic-basaltic laharic deposits. Other rock formations, such as Qyw, Qmt, Waringin-Bedil andesite,
and Qwb formations, feature lower distributions of kaolinite and alunite. Quartz is mostly distributed in
Qmt formations (Figure 4c). The distribution of quartz is predominantly in hillsides or steep terrain near
Mt. Gambung, Mt. Wayang, and Mt. Windu. Quartz can occur as a secondary mineral due to
hydrothermal alteration and as a primary product of volcanic activity.
A fuzzy logic algorithm is the last process used in DPCA to combine several raster images into a
single raster based on dominant similarities between them. In this research, the fuzzy logic algorithm
was used to determine the advanced argillic alteration zone and the propylitic alteration zone. The
advanced argillic alteration zone consists of alunite, quartz, and kaolinite; the propylitic alteration zone
consists of chlorite and epidote.
Advanced argillic and propylitic zones occur on the surface of the WWGF (Figure 4d). Some areas
cannot be mapped because there is no similarity between the spectral reflectance on the surface of both
alteration zones and because of very dense vegetation on the surface preventing remote sensing
identification.
3.2. Lineaments Identification from Landsat 8 Image

The lineament pattern identified from Landsat 8 image interpretation is the same as the regional geology
structure (Figure 5a). The lineament pattern in this area is related to regional faults with directions
southwest–northeast and northwest–southeast.
The lineament distribution in the WWGF was translated into lineament density (Figure 5b). The
purpose of lineament density is to identify areas that have high-density lineament. High density
lineament potentially become evidence of permeable zones and media for hydrothermal fluid flowing
from a reservoir to the surface [9].
5
The most areas of the WWGF have moderate to high lineament density (4–7 km/km2), but near the
mountain, this increases to 8 km/km2 (shown in the red area), meaning that lineament density increases
near the mountain (see Figure 5b). This lineament density is related to a hydrothermal alteration
assembly—mostly secondary minerals on the surface that show up near areas with high lineament
density.
Figure 4. Distribution of minerals in the WWGF: (a) chlorite and epidote distribution;
(b) kaolinite and alunite distribution; and (c) quartz distribution. (d) Advanced argillic
and propylitic zones in the WWGF.
6
Figure 5. Lineament in the WWGF: (a) overlay with alteration zone; (b) in lineament density form
and overlay with WWGF’s morphology.
3.3. Identification of soil samples from the surface of the WWGF

The identification of soil samples from the surface were used to validate the data from Landsat 8 imagery
processing. In this research, 18 soil sample spots were used to validate the data (Figure 6).
Manifestations of geothermal activity in the WWGF were also used to validate the Landsat 8 imagery
processing. Identification of soil samples was done using X-ray diffraction (XRD) analysis, spectral
reflectance analysis, and X-ray fluorescence (XRF) analysis to determine the mineral composition of
the sample.
The identification of soil samples was mostly in agreement with the Landsat 8 imagery processing.
However, some were misinterpreted; from Landsat 8 imagery processing, K-9 and K-13 were identified
as propylitic alteration zone, but from XRD analysis and spectral reflectance analysis, K-9 and K-13
have secondary minerals of halloysite, cristobalite, and quartz. This means that according to the XRD
analysis and spectral reflectance analysis, K-9 and K-13 should be advanced argillic alteration zones.
This problem can be explained as follows;
1. The surface features very diverse. Soil samples were collected from a 0–1meter depth from
ground level. The soil sample was blended for XRD, XRF, and spectral reflectance analysis, but
the Landsat 8 image only captured the surface.
2. The spatial resolution of Landsat 8 is 30 x 30 meters, and diverse mineralization in one pixel of
the Landsat 8 image is difficult to see.
The secondary minerals from each spot of the soil samples was an alunite-kaolinite assembly and a
chlorite assembly. This means that the alteration zone in the WWGF is an advanced argillic and
propylitic zone, as identified in previous research [8]. This alteration assembly indicated that the pH of
the formation was acidic to neutral.
7
Figure 6. Location of soil samples and

hydrothermal manifestation in the WWGF.
Based on XRD analysis, minerals on the surface of the WWGF consist of clay and non-clay minerals.
The non-clay minerals associated with clay minerals on the surface are as follows:
1. Iron oxide (hematite [Fe2O3]) and iron hydroxide (goethite [Fe2O3.H2O])
2. Aluminum hydroxide (gibbsite [Al(OH)3])
3. Silicon oxide (quartz [SiO2] and cristobalite [SiO2])
4. Zeolite
5. Sulfate (alunite [KAl3(SO4)2(OH)6)])
6. Feldspar
The results of XRD and spectral analyses were validated using XRF analysis which provided data
about the abundance of elements in the sample, and this data can validate the occurrence of minerals in
the soil sample. The elements that were used to validate the X-ray analysis and spectral reflectance
analysis were as follows:
1. Silicon (Si)
2. Iron (Fe)
3. Aluminum (Al)
4. Sulfur (S)
8
Figure 7. Distribution of (a) Al2O3, (b) Fe2O3, (c) SiO2, and (d) SO3 abundance in the WWGF.
Aluminum is a key element of clay minerals, which occur across the surface of the WWGF (Figure
7a). The mass percentage of Al2O3 in the WWGF ranged from 30–45%. This is because of the hydrolysis
process of aluminum silicate minerals such as feldspar, pyroxene, amphibole, and mica. In this process,
they repeatedly react with water, releasing aluminum ions that move and settle in the surface of the soil.
This is related to the occurrence of clay minerals in the WWGF.
Iron occurrence is related to the distribution of limonitic minerals in the WWGF (Figure 7b). The
area with the highest mass percentage of Fe is likely to be the most weathered area in the northern region
of the WWGF. This correlates with the limonitic mineral distribution in the northern region of the
WWGF observed during Landsat 8 image processing.
The area with the highest percentage of Si is likely to be near an intrusion or fault zone with high
permeability; thus, many silicate minerals, including quartz, cristobalite, and opal, will occur there
(Figure 7c). This is related to the occurrence of Si, which mostly occurs near Mt. Wayang and Mt.
Windu, both of which feature the intrusion of andesitic rocks.
Sulfur is a constituent element of alunite, which requires a high temperature (around 200ºC) and
acidic pH to form (Figure 7d). Alunite occurs with acidic gases such as CO2 and H2S because of the
boiling and evaporation of deep fluids and condensation in rocks near the surface. The distribution of
high sulfur content is related to K-18, K-8, K-30 and K-19, and these sample points which were
identified to be advanced argillic alteration zones. It is correlated with Landsat 8 imagery processing
and surface manifestations of geothermal activity; there is a fumarole in the Wayang crater near K-18,
and the Kertamanah hot spring is near K-8. These findings agree with previous research [10].
9
3.4. Control Factor of Alteration in the WWGF

Triggers for alteration processes in the surface of the WWGF are probably faults, which occur as
lineaments in the surface of the WWGF. Fumaroles and hot springs are evidence of the occurrence of
faults in the subsurface of the WWGF. The correlation between lineament density and the occurrence
of a hydrothermal alteration zone in the surface of the WWGF is also evidence of faults as the main
factor in the hydrothermal alteration zone.
4. Conclusions
The hydrothermal alterations at the surface of the WWGF are advanced argillic and propylitic zones.
The former contains the secondary minerals alunite, cristobalite, and quartz, and the latter contains the
secondary minerals chlorite and epidote. The controlling factor of hydrothermal alteration in the surface
is probably the geological structure of WWGF, which allows hydrothermal fluid to alter the rocks near
the surface. Remote sensing with DPCA method integrated with GIS-based analysis successfully
enhanced the hydrothermal alteration zones. The alteration zones extracted by this combination were
confirmed to be plausible and consistent with the ground truth data. However, there are some
considerations related to unsuccessful of DPCA to enhance the alteration zones that may be caused by
the limitation of spatial resolution of Landsat 8 image and vegetation cover conditions.
Acknowledgments
The authors wish to sincerely thank the The SATREPS (Science and Technology Research Partnership
for Sustainable Development) program through Beneficial and Advanced Geothermal Use System
(BAGUS) project for providing field equipment, and survey accommodation support. This research was
funded by the Institut Teknologi Bandung through the Research, Community Services, and Innovation
Program 2017.
References
[1] Carranza E J M and Hale M 2002 Mineral Imaging with Landsat Thematic Mapper data for
Hydrothermal Alteration Mapping in Heavily Vegetated Terrane International Journal of
Remote Sensing vol 23 4827–4852.
[2] Hede A N H Kashiwaya K Koike K and Sakurai S 2015 A New Vegetation Index for Detecting
Vegetation Anomalies due to Mineral Deposits with Application to a Tropical Forest Area.
Remote Sensing of Environment vol 171 83–97.
[3] Hede A N H Koike K Kashiwaya K Sakurai S Yamada R and Singer D A 2017 How Can Satellite
Imagery be Used for Mineral Exploration in Thick Vegetation Areas? Geochemistry,
Geophysics Geosystems vol 18(2) 584–596.
[4] Sabrian P G Saepuloh A Syafrizal & Hede A N H 2017 Identification of Altered Minerals Based
on Synthetic Aperture Radar (SAR) For Mineral Exploration in a Tropical Area IOP
Conference Series: Earth and Environmental Science. vol 71 012021.
[5] Fraser S J and Green A A 1986 A Software Defoliant for Geological Analysis of Band Ratios
International Journal of Remote Sensing Vol 8 525–532.
[6] Bogie I and Mackenzie K M 1998 The Application of A Volcanic Facies Model to An Andesitic
Stratovolcano Hosted Geothermal System At Wayang Windu, Java, Indonesia. (pp.265–270).
in: Proceeding 20th NZ Geothermal Workshop 1998.
[7] Alzwar M Akbar N and Bachri S 1992 Geological Map of the Garut and Pameungpeuk
Quadrangle, Jawa. Pusat Penelitian dan Pengembangan Geologi. Bandung.
[8] Susanto A 2011 Geology and Surface Hydrothermal Alteration of Malabar Area, Northern Part
of the Wayang Windu Geothermal Field GRC Transaction vol 35
[9] Soengkono S 1999 Analysis of Digital Topograhic Data for Exploration and Assessment of
Geothermal System 21st New Zealand Geothermal Workshop.
[10] Hendrasto F 2005 Penentuan Daerah Resapan Sistem Panas Bumi Gunung Wayang Windu, Jawa
Barat. Institut Teknologi Bandung.
10

Salamba 2019 IOP Conf. Ser. Earth Environ. Sci. 254 012004

Uploaded by

Copyright:

Available Formats

Salamba 2019 IOP Conf. Ser. Earth Environ. Sci. 254 012004

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Salamba 2019 IOP Conf. Ser. Earth Environ. Sci. 254 012004

Uploaded by

Copyright:

Available Formats

IOP Conference Series: Earth and Environmental Science

PAPER • OPEN ACCESS You may also like

- A New Strategy for Comparison of Land

This content was downloaded from IP address 103.111.141.191 on 15/01/2023 at 21:28

Identification of alteration zones using a Landsat 8 image

2. Data and Methodology

3. Result and Discussion

3.1. Surface Alteration Zone based on PCA

Figure 2. Spectral reflectance from minerals and vegetation. Each

Figure 3. Clay and limonitic layered distributions in the

Table 3. Band ratio for DPCA in Landsat 8 image.

3.2. Lineaments Identification from Landsat 8 Image

3.3. Identification of soil samples from the surface of the WWGF

Figure 6. Location of soil samples and

3.4. Control Factor of Alteration in the WWGF

You might also like