358536

MINISTRY OF RESEARCH, TECHNOLOGY AND HIGHER EDUCATION
UNIVERSITAS GADJAH MADA

POSTGRADUATE PROGRAM OF FACULTY OF ENGINEERING
POSTGRADUATE PROGRAM STUDY OF GEOLOGICAL ENGINEERING
THESIS:
MINERAL POTENTIAL MAPPING USING GEOGRAPHIC INFORMATION

SYSTEM (GIS) FOR GOLD MINERALIZATION IN WEST JAVA,
INDONESIA
Prepared By:
Seang Sirisokha
ID: 13/358536/PTK/09380
YOGYAKARTA
2015
i
ii
DECLARATION
This is to state that this thesis does not contain work that previously been
submitted to obtain any degree at any higher educational establishment center, and
to best of my knowledge, does not contain the previously published work or
opinion of any other person, with the exception of those cited in writing this thesis
and acknowledged in the references.
…., 2015
Yogyakarta, July 09
Prepared by
Seang Sirisokha
SIN: 13/358536/PTK/09380
iii
DEDICATION
This thesis is dedicated to my beloved parents, sister, brothers, cousins

and aunties.
iv
ACKNOWLEDGEMENT
I am deeply indebted to many people who have helped me in various ways

to complete this thesis.
First of all, I would like to express my gratitude to the ‘ASEAN University

Network/Southeast Asia Engineering Education Development Network
(AUN/SEED-NET)’ and ‘Japanese International Cooperation Agency (JICA)’
providing me to study ‘Master Degree’ at Gadjah Mada University in Indonesia.
I would like to express my deepest gratitude to my Supervisor, Dr. Lucas

Donny Setijadji and co-advisor, Dr. Ir. I Wayan Warmada, Lecturers of
Department of Geological Engineering, Faculty of Engineering, Gadjah Mada
University, for their technical guidance, constructive comments, valuable advises,
and encouragement.
I am deeply grateful to Dr. Sugeng Sapto Surjono, Head of Geological

Engineering Department, Gadjah Mada University and Dr. Agung Harijoko,
Program manager, Geological Enigneering Department, Gadjah Mada University
for kindness and help everything during my study in Gadjah Mada University.
I would like to express my gratitude to Dr. Arifudin Idrus and Dr. Nugroho
Imam Setiawan, examiners, for their valuable comments, sugguestions, and
constructive criticism where necessary during my defense.
I also thank to the faculty and staffs of the host institution, Geological
Engineering Department, Gadjah Mada University. Thanks to all my AUN/SEED-
Net colleagues and my friends from here and overseas who made my time so
much pleasurable and for supporting me throughout my study in here.
Finally, I would like to express my sincere gratitude to my parents, sister

and brothers for their unlimited loves, encouragement, help, advices and financial
support while I was conducting my master research.
v
ABSTRACT
Western Java is a part of the Sunda Banda magmatic belt. This belt is well
known to be host for several gold deposits in Indonesia, the distribution of 107 Au
occurrences in this area was examined in terms of spatial association with various
geological phenomena. In this study area, GIS and the method of weight of
evidence are used for gold potential mapping.
The weight of evidence method is one of the most important data-driven

methods for mapping in GIS. This method is a probability based on technique for
mapping mineral potential using the spatial distribution of known mineral
occurrences. There are six evidences maps such as NE-SW lineaments, NW-SE
Lineaments, host rocks, heat source, clay alteration and limonitic alteration, have
been combined using a weights of evidence model to predict gold potential in
West Java.
The best predictive map generated by this method defines 21.62% (9902
km2) of study area as favourable zones for gold mineralization further exploration
work. It predicts correctly 74 (92.5%) of the 80 model deposits and predicts
correctly 26 (96.35%) of the 27 validation deposits, has 6 main prospective
target for further exploration are located in Bayah Dome, southern mountain,
Honjie Igneous Complex and Bogor zone, Purwakarta. Bayah Dome is highest
potential area for gold deposit like Gunung Pongor, Cikidang, Cirotan, Ciawitali,
Cikotok districts and other deposits. The potential area of Au occurrences in
research area is associated with NE-SW and NW-SE structure/ lineaments,
dominated surrounding the Tertiary extrusive volcanic rock and intrusive rock
unit and hosted in Miocene to Pleistocene lithology rock unit.
Keyword: GIS, weight of evidence method, mineral potential mapping, Bayah

Dome, Bogkor, West Java.
vi
SARI
Jawa Barat adalah bagian dari sabuk magmatik Sunda Banda. Sabuk ini
dikenal sebagai host dari beberapa endapan emas di Indonesia, distribusi dari 107
keterjadian Au di area ini diperiksa dalam hal hubungan spasial dengan berbagai
fenomena geologi. Pada daerah penelitian, GIS dan metode weight of evidence
digunakan untuk memetakan potensi emas.
Metode weight of evidence adalah salah satu metode data driven yang
paling penting untuk pemetaan di GIS. Metode ini adalah peluang berdasarkan
teknik untuk pemetaan potensi mineral menggunakan distribusi spasial dari
keterjadian mineral yang diketahui. Terdapat 6 peta evidence seperti kelurusan
timur laut-barat daya dan kelurusan barat laut-tenggara, batuan dasar, sumber
panas, alterasi lempung, dan alterasi limonitic dikombinasi menggunakan model
weight of evidence untuk memperkirakan potensi emas di Jawa Barat.
Peta prediksi terbaik yang dihasilkan dengan metode ini menghasilkan
21.62% (9902km2) daerah penelitian yang merupakan zonasi mineralisasi emas
yang dapat dilakukan eksplorasi lebih lanjut. Diprediksi secara akurat 74 dari 80
model endapan (92.5%) dan diprediksi dengan benar 26 dari 27 endapan secara
valid, dimana terdapat 6 prospek utama yang menjadi target eksplorasi lebih lanjut
yang berlokasi di kubahBayah, pegunungan selatan, kompleks batuan beku Honjie
dan zona Bogor, Purwakarta. Kubah Bayah adalah area dengan potensi endapan
emas tertinggi seperti di Gunung Pongkor, Cikidang, Cirotan, Ciawitali, Cikotok,
dan distrik endapan lainnya. Daerah potensi mineralisasi Au di daerah penelitian
berasosiasi dengan kelurusan NE-SW dan NW-SE, yang didominasi dengan unit
batuan intrusif berumur Tersier, dengan batuan dinding berupa unit litologi
berumur Miosen-Pleistosen.
Kata Kunci: GIS, metode weight of evidence, pemetaan potensi mineral, Kubah
Bayah, Bogkor, Jawa Barat
vii
CONTENTS
APPROVAL SHEET ............................................................................................ ii
DECLARATION ................................................................................................... ii
DEDICATION ...................................................................................................... iii
ACKNOWLEDGEMENT ................................................................................... iii
ABSTRACT ........................................................................................................... v
SARI ...................................................................................................................... vi
CONTENTS ......................................................................................................... vii
LIST OF TABLES .............................................................................................. xii
LIST OF FIGURES ........................................................................................... xiv
LIST OF ABBREVIATIONS ........................................................................... xix
CHAPTER I: INTRODUCTION ........................................................................ 1
1.1. Background ...................................................................................................... 1
1.2. Problem Formulation ....................................................................................... 2
1.3. Objectives of the Study .................................................................................... 2
1.4. Research Outcome ........................................................................................... 3
1.5. Scope of Research ............................................................................................ 3
1.5.1. The Study Area ...................................................................................... 3
1.5.2. Scope of Work ....................................................................................... 3
1.5.3. Research Limitation ............................................................................... 4
1.6. Previous Studies ............................................................................................... 4
1.7. Organization of the Thesis ............................................................................... 8
CHAPTER II: LITERATURE REVIEW......................................................... 10
2.1. Regional Geology of Western Java ................................................................ 10
2.1.1. Regional Geological Setting ................................................................ 10

viii
2.1.2. Physiography of West Java .................................................................. 12
2.1.3. Stratigraphy.......................................................................................... 14
2.1.4. Quaternary Volcanism ......................................................................... 16
2.2. Style of Gold Deposits in West Java .............................................................. 17
2.2.1. Mineralization in Bayah Dome/Mountains.......................................... 18
2.2.2. Mineralization in Southern Mountains Range ..................................... 22
CHAPTER III: THEORETICAL BACKGROUND ....................................... 29
3.1. Gold Deposits ................................................................................................. 29
3.1.1. Hydrothermal Gold Deposit ........................................................................ 29
3.1.2. Porphyry Cu-Au Mineralization ................................................................. 30
3.1.3. Epithermal Gold ................................................................................... 32
3.1.3.1. Terminology...................................................................................... 32
3.1.3.2. Classification of Epithermal Deposit ................................................ 33
3.1. 3.3. Base Metal Veins ............................................................................. 38
3.1.4. Skarn Deposit ....................................................................................... 39
3.2. Remote Sensing (RS) ..................................................................................... 41
3.2.1. Landsat Images .................................................................................... 41
3.2.2. ASTER GDEM .................................................................................... 43
3.3. Software Defoliant Method ............................................................................ 43
3.4. Geographic Information System (GIS) .......................................................... 45
3.4.1. Mineral Potential Mapping using GIS ................................................. 45
3.4.2. GIS Modeling Techniques for Producing Mineral Potential Maps ..... 47
3.5. Weight of Evidence Method .......................................................................... 48
3.5.1. Test Conditional Independence ........................................................... 51
1. Pairwise test of Conditional Independence ................................................ 52

ix
3.6. Application of Weight of Evidence in Mineral Potential Mapping ............... 55
3.7. Extraction of Evidences Maps ....................................................................... 55
3.7.1. Lithology Map ..................................................................................... 55
3.7.2. Geological Structure Map .................................................................... 56
3.7.3. Hydrothermal Alteration Map ............................................................. 56
3.8. Hypothesis ...................................................................................................... 57
CHAPTER IV: RESEARCH METHODOLOGY ........................................... 58
4.1. Variable Data and Materials........................................................................... 58
4.2. Software ....................................................................................................... 58
4.3. Research Methodology................................................................................... 58
4.3.1. Research Preparation ........................................................................... 58
4.3.2. Spatial Data Processing and Analysis .................................................. 59
4.3.3. Predictive Modeling ............................................................................. 60
4.3.4. Model Validation ................................................................................. 61
CHAPTER V: SPATIAL DATASETS AND EVIDENCE MAPS ................. 64
5.1. Mineral Deposit Database .............................................................................. 64
5.2. Lithology ....................................................................................................... 65
5.3. Geological Structure....................................................................................... 68
5.4. Hydrothermal Alteration ................................................................................ 73
5.4.1. Clay Alteration ..................................................................................... 73
5.4.2. Limonitic Alteration ............................................................................ 74
5.4.3. Accuracy Assessment .......................................................................... 76
CHAPTER VI: ANALYSIS AND RESULTS .................................................. 80
6.1. Evidence Maps ............................................................................................... 80
6.2. Analysis ....................................................................................................... 81

x
6.2.1. Spatial Association of Au Deposit with Each Evidence Maps ............ 81
6.2.1.1.Calculating Weights of Evidence of NE-SW Trending

Lineaments ............................................................................................ 82
6.2.1.2.Calculating Weights of Evidence of NW-SE Trending

Lineaments ............................................................................................ 84
6.2.1.3. Calculating Weights of Evidence of Host rocks ............................. 86
6.2.1.4. Calculating Weights of Evidence of Heat source ........................... 88
6.2.1.5. Calculating Weights of evidence of Clay alteration ....................... 90
6.2.1.6.Calculating Weights of evidence of Limonitic

Alteration ............................................................................................... 92
6.2.2. Test of Conditional Independence ....................................................... 94
6.3.Results ....................................................................................................... 96
6.3.1. Probabilistic mapping of Au Minerals Potential ................................. 96
6.3.2. Validating of Predictive Maps and Determining best predictive

map................................................................................................................. 96
6.3.3. Overall Test.......................................................................................... 97
CHAPTER VII: DISCUSSION ...................................................................... 102
7.1. Au Potential Map .................................................................................. 102
1. Prospective Target 1: ............................................................................... 102
4. Prospective Target 4, 5, and 6 ................................................................ 104
7.1.2. Weight of Evidence Approach ........................................................... 112
CHAPTER VIII: CONCLUSIONS AND RECOMENDATION ................. 114
8.1. Conclusions........................................................................................... 114
8.2. Recommendation .................................................................................. 116

xi
REFERENCES .................................................................................................. 117
APPENDICES ................................................................................................... 125
A.GLOSSARY .................................................................................................... 126
B. Useful ArcGIS operations, tools and tips ....................................................... 127
C. Statistics of Landsat TM data ......................................................................... 128
E. Example for calculation of weight and contrast ............................................. 130
F. Example for calculation of Chi-square values ................................................ 134
G.Example for calculation of posterior probability map .................................... 134

xii
LISTS OF TABLES
Table 2.1: History of the Bogor Basin simplified .......................................... 15
Table 2. 2: Table of Summary mineralization in West Java.............................25
Table 3.1: Characteristics of epithermal systems ........................................... 33
Table 3.2: Summary of relationships between sulfidation state of Ore-
Forming Environment, related Igneous Rock Compositions,
and Tectonic Setting .......................................................................35
Table 3.3: Characteristic of epithermal deposit types .....................................36
Table 3. 4: Characteristics of Landsat 4-5 TM and 7 ETM ..............................42
Table 3.5: The models exist for mapping mineral potential ............................48
Table 3. 6: Contingency table for pairwise testing conditional
independence based on pixels that contain a mineral deposit. .......54
Table 4.1: Research Schedule ......................................................................... 63
Table 5. 1: Result of Rose Diagram................................................................. 70
Table 6.1: Variation of weights and Contrasts for cumulative distances
from NE-SW trending lineaments with respect to gold
deposit……………………………………………………....... 83
Table 6.2 : Variation of weights and Contrasts for cumulative
distances from NW-SE trending lineaments with
respect to gold deposit ....................................................................85
Table 6.3: Variation of weights and Contrasts for cumulative
distances from host rocks with respect to gold deposit ..................87
Table 6.4 : Variation of weights and Contrasts for cumulative distances
from heat sources with respect to gold deposit. ............................ 89
Table 6.5: Variation of weights and Contrasts for cumulative
distances from clay alteration with respect to gold
deposit.............................................................................................91
Table 6.6 : Variation of weights and Contrasts for cumulative
distances from limonitic alteration with respect to gold
deposit.............................................................................................93
xiii
Chi-square values    for testing conditional independence

2
Table 6.7:
between host rocks (B1) and heat sources (B2) evidence map
with respect to the Au occurrences. ................................................95
Table 6.8: Calculated    values

2
for testing conditional independence
between all pairs of binary maps with respect to gold
occurrences .................................................................................... 95
Table 6.9 : Summarized of overall test conditional independence in
posterio probability map shown in figure 6.15 ............................. 98
Table 6.10: An appropriated the evidence maps base on conditionally
independent and statistically significant contrasts for predict
the Au occurrences in research area .............................................. 99
Table 6.11: Inventory of Au occurrences in zones of different posterior
based on figure 6.14 and 6.15 ........................................................ 99
xiv
LIST OF FIGURES
Figure 1.1: The locality of research area in West Java, Indonesia .....................3
Figure 2.1: Tectonic setting of Java Island, Indonesia .....................................11
Figure 2.2 : Major geological structure of Java Island ....................................12

Figure 2.3 : Physiographic map of West Java ...................................................13
Figure 2.4: Quaternary volcanoes in Java, with volcano-type terminology
follow Tatsumi and Eggins (1995). Note the presence of
several segments of volcanic configuration . .................................17
Figure 2.5: Location of gold deposit in West Java ..........................................18
Figure 2.6: Geological map showing the location of the Cirotan Au–Sn–
W deposit, as well as other epithermal Au deposits in the
region. Insert shows the location of major epithermal gold
deposits of Indonesia within the plate-tectonic setting of the
region. .............................................................................................19
Figure 2.7: Geology and alteration map of the Cikidang vein, Western
Java, Indonesia (Modified from ANTAM, 1994) ..........................20
Figure 3.1: Schematic diagram of porphyry, epithermal and skarn
deposits (Kirkham and Sinclair, 199) .............................................30
Figure 3.2: Hydrothermal alteration zones of porphyry Cu-Au deposits .........31
Figure 3.3: Schematic cross section through an intrusive centered
hydrothermal system outlining the environments of porphyry,
high-sulfidation and low-sulfidation systems.................................35
Figure 3.4: Stages in the development of skarn deposits ..................................40
Figure 3.5: Typical reflectance spectra of vegetation, iron oxides and
clays ......................................................................................44
Figure 3.6: Venn diagram of schematic spatial overlap relationships
between binary predictor pattern and binary deposit pattern. ........49
Figure 3.7: Venn diagram to illustrate the concept of conditional
independence. .................................................................................53
xv
Figure 4.1: Flow Chat of Methodology in this research ...................................62
Figure 5.1 : Distribution of Au occurrences in study area (Physiography

from Van Bemmelen, 1949 and Mineral deposits are
modified from Setijadji, 2014, Unpublished data). ........................64
Figure 5.2: Lithology map of the study area .....................................................65
Figure 5.3: Tertiary- Pleistocene host rock .......................................................67
Figure 5.4: Tertiary intrusive rock ...................................................................67
Figure 5.5: Four shaded relief images derived from DEM with different
illumination directions ( sun azimuth) of ,0°, 45°, 90°, and
315°, with a solar elevation of 45°. ...............................................68
Figure 5.6: The Combined shaded relief DEM image created by
combining serveral shaded relief images with different
illumination.....................................................................................69
Figure 5.7: Lineaments map, interpreted from shaded relief DEM image .......69
Figure 5.8: Rose diagram of lineament .............................................................70
Figure 5.9: a) Lineaments derived from shaded relief of Arinem
districtDEM, b) Regional geological map of the Arinem
district, West Java (modified after Alzwar et al., 1992) .................72
Figure 5.10: a) Lineaments in Cibaliung derived from shaded relief DEM,
b) Simplified geological map of Cibaliung area.............................72
Figure 5.11: a) Lineaments derived from shaded relief DEM, b)
Geological map of Cupunagara, Subang District, West Java .........73
Figure 5.12 : Clay alteration map using the software defoliant technique ..........75
Figure 5.13: Limonitic alteration map using the software defoliant
technique.........................................................................................76
Figure 5.14: The Normalized Difference Vegetation Index (NDVI) map,
derived from Landsat 5 TM data ....................................................77
Figure 5.15: a) Clay alteration from Landsat 5 TM, b) Alteration zone of
Cupunagara, Subang, West Java , c) Alteration zone of
Cihurip, Garut Regency ..................................................................78
xvi
Figure 6.1: Evidences maps: a) Host rock, b) Heat sources, c) NE-SE

trending lineaments, d) NW-SE trending lineaments, e) Clay
alteration, and f) Limonitic alteration ...............................80
Figure 6. 2: Buffer distance s to the NE-SW lineaments ...................................82
Figure 6.3: Graph showing the studentized contrast, C and percent of Au
occurrences related with buffer at distance for NE-SW
trending lineaments. .......................................................................83
Figure 6.4: Binary predictor pattern of NE-SW trending lineamens with
2000 m buffer zones. ......................................................................84
occurrences related with buffer at distance for NW-SE
trending lineaments. .......................................................................85
Figure 6.6: Binary predictor pattern of NW-SE trending lineamens with
1000 m buffer zones. ......................................................................86
occurrences related with buffer at distance for host rocks. ............87
Figure 6.8: Binary predictor pattern of host rocks. ...........................................88
occurrences related with buffer at distance for heat sources. .........89
Figure 6.10: Binary predictor pattern of Heat sources with 1000 m buffer
zones. ..............................................................................................90
occurrences related with buffer at distance for Clay alteration. .....91
Figure 6.12: Binary predictor pattern of clay alteration with 50 m buffer
zones. ..............................................................................................92
occurrences related with buffer at distance for Limonitic
alteration. ........................................................................................93
Figure 6.14: Binary predictor pattern of limonitic alteration with 150 m
buffer zones ....................................................................................94
xvii
Figure 6.15: Posterior probability map of Au in research area, deveried

from all predictor patterns present in table 6.10. ..........................100
Figure 6.16: Posterior probability map of Au in research area, resulting
from exclusion of clay and limonitic alteration binary
predictor patterns. .........................................................................101
Figure 7.1: Percentage of Potential class ....................................................... 104
Figure 7.2: Predictive map of gold in West Java and generate prospective
target area for Au ..........................................................................105
Figure 7.3: Gold potential target areas 1: A) Mineral potential map
overlain by Au occurrences as training and validation data,
B) Mineral potential map superimposed on shaded of
topography, overlain by lineaments and Au occurrences .............106
Figure 7.4: Gold potential target areas 2 ; (A) Mineral potential map
overlain by Au occurrences as training and validation data ;
(B) Mineral potential map superimposed on shaded of
overlain by Au occurrences as training and validation data:
Figure7.6: Gold potential target areas 4: A) Mineral potential map
overlain by Au occurrences as training and validation data:
Figure7.7: Gold potential target areas 5: A) Mineral potential map
xviii

Figure 7.9: Graph showing the positive weight (W+) of all evidence
maps which used to predict Au-Cu occurrences map. .................113
Figure 8. 1: Predictive map of gold occurrences in West Java ........................115
xix
LIST OF ABBREVIATIONS
Abbreviations Meaning
CI Conditional Independence
C Contrast
C/s(C) Studentized Contrast
NDVI Normalized Difference Vegetation Index
DEM Digital Elevation Model
DN Digital Number
DPC Directed Principal Component
E East
ETM+ Enhanced Thematic Mapper Plus
GIS Geographical Information System
MSS Multi spectral Scanner
N North
NE North East
NW North West
OLI Operational Land Imager
PCS Projection Coordinate System
S South
SE South East
SW South West
SRTM Shuttle Radar Topography Mission
TIRS Thermal Infrared Sensor
TM Thematic Mapper
UTM Universal Transverse Mercator
USGS United States Geological Survey
W West
WGS WGS World Geodetic System
WofE Weights of evidence method
W+ Positive weight
W- Negative weight
1
CHAPTER I
INTRODUCTION
1.1. Background
Java is essentially a volcanic island, and a long history of arc volcanism is

records in Cenozoic stratigraphy. An east-west-trending chain of >30 volcanoes,
forming part of the Sunda Arc, creates the central spine of the Java island (Hall,
2002; Smyth et al., 2008). Mineral occurrences in Java Island, Indonesia, are
located in the physiographic area known as the Southern Mountains. This
physiographic area contains the most extensive exposures of Tertiary magmatic
rocks in the island ranging age from Eocene to Upper Miocene. In turn, this
physiographic area is considered to be the most prospective region for metallic
mineral deposits (Setijadji, 2009). Tectonics during Pliocene to Pleistocene is
believed to be responsible for the occurrences of gold mineralization along the
Southern Mountains in the southern part of West Java Island (Koesmono et al.,
1996).
In general, the oldest rocks exposed on West Java are Late Cretaceous –
Eocene in age. Sedimentary and volcanic rocks interdigitated until the Pliocene
when the whole area was uplifting, and some areas were covering by Quaternary
volcanics. Intrusive activity also occurred during the period of uplift. The Western
Java is one area of richest gold in Indonesia areas like gold mine in Cikotok,
Cirotan, Ciawitali, Cikidang and active gold mine in Pongkor (Milesi et al., 1999).
Several epithermal ore deposits associated with volcanism in West Java including
gold deposits within and on flanks of the Bayah Dome and southern mountain of
West Java (Yuningsih et al., 2012). Many occurrences of mineral deposits are
reported from southern part of West Java. The gold mineralization in Western
Java is typical of epithermal type deposit and hosted by Tertiary volcanic rocks.
Many epithermal-style alterations and mineralization are associated with
volcaniclastic and intrusive rocks, including epithermal gold deposits and few
prospect areas for possible mineralization of porphyry systems (Basuki et al.,
2012)
2
Geographic Information System (GIS) methods are very useful for

processing and combining data within maps in mineral potential mapping. The
development of GIS-based methods for integration and analysis of regional
exploration datasets has an important role in assisting the decision-making
processes for geologists in selection of exploration area. The spatial modeling
techniques has been proposed for mineral potential mapping, such as weights of
evidence model (Saro and Hyun, 2011). Gold potential mapping was evaluated by
GIS technique using the weights of evidence modelling with respect to the
bivariate statistical approach.
Although there are many mineral occurrences in West Java, the used of
geological data such as geological structure, lithology, hydrothermal alteration for
considered as gold potential occurrences areas are have not worked out. So, the
current research is carried out to create gold potential map by using weight of
evidence method in Geographic Information System (GIS), for the benefits of
future exploration of gold in the study area.
1.2. Problem Formulation
According to the study area is located in West Java, Indonesia is considered

as gold occurrences area. The factors of evidence maps in the study consists of
geological data, such as structure, lithology, and hydrothermal alteration to show
mineral deposits favorability of study area, which are needed to solve some of
problems below:
1. What are the geologic criteria associations with gold deposits in research
area?
2. How is the distribution of gold in West Java?
1.3. Objectives of the Study
The research objectives are as follows.

1. To determine and evaluate the relationship between geological criteria and
gold deposits in West Java.
2. To delineate prospect areas of gold in West Java.
3
1.4. Research Outcome
1. To understand the relationship between geological data and gold deposit
2. To indicate the prodictive target aera of gold deposits in West Java.
1.5. Scope of Research
1.5.1. The Study Area
The study area is located in the western part of Java Island in Indonesia,
within longitude 105°-109°E and latitude 6°-8°S and measure approximately
400km by 200km (Figure 1.1)
Figure 1.1: The locality of research area in West Java, Indonesia (modified from van
Bemmelen, 1949)
1.5.2. Scope of Work
The objectives of the research will be achieved through the following steps:
 Compilation of spatial datasets on mineral deposits, geological data
(lithology and structures), alteration-mineralization maps, DEM, satellite
images
4
 Creating database and generating evidence maps by GIS

 GIS spatial modeling for predicting favorable areas for gold, and
generating gold potential map.
 Validations of model for confirmation of the generated prospective target
areas for gold mineralization.
1.5.3. Research Limitation

 Geological data was done from secondary data.
 There are only two alteration maps like clay and limonitic alteration and
other alteration cannot be used in the research because the limit of time
has limitation.
 Even the geophysics data is more accurate for alteration and structural
maps. But, due to the study area are too large and its scope of research as
well so the geophysical data could not be conducted or available in this
researched study.
 Used only secondary and primary data such lithology, mineral deposit,
ASTER GDEM and Landsat 5 TM. There have no field work and
laboratory work.
1.6. Previous Studies
There are some previous researchers who have done research about gold
deposit in West Java are as follows:
1. Tun et al. (2014) conducted research about Fluid Inclusion Studies of the
Cijulang High-sulfidation Epithermal Prospect, West Java, Indonesia. Cijulang
prospect is located in Talegong Sub-District of Garut Regency, West Java,
Indonesia. The prospect is characterized by enargite-gold mineralization and
associated acid sulfate alteration. Mineralization and alteration are hosted by
Tertiary volcanic and volcaniclastic rocks. Host rocks belong to lava andesite, tuff
and breccia. Typical ore minerals include pyrite, enargite, luzonite, tennantite,
chalcopyrite, covellite, galena, emplectite, and tellurides. Principal hypogene
hydrothermal alteration types are silicic (vuggy/massive), advanced argillic,
argillic and propylitic.
5
2. Idrus et al. (2013) reviewed about styles and characteristics of metal

Mineralization at Southern part of West Java, Indonesia. The result can suggested
that Cibaliung gold deposit is a representative well known low sulphidation
epithermal styles occurred within the Honje Igneous Complex. In Bayah
Dome/Mountains, two mineralization styles of low sulphidation epithermal
deposit identified consist of ‘Pongkor-type’ and ‘Cirotan-type’ (Marcoux, 1994).
3. Basuki et al. (2012) carried out of gold mineralization systems in
Southern Mountain Range, West Java. It comprises Tertiary extrusive volcanic
rocks and shallow intrusive rocks, as well as sedimentary rocks (limestone and
siliciclastic). Many epithermal-style alterations and mineralizations are associated
with this volcaniclastic and intrusive rocks, including epithermal gold deposits
and few prospect area for possible mineralisation of porphyry systems. There are
four types of gold mineralization like low sulfidation epithermal, high sulfidation
epithermal, polymetalic gold and base metal mineralization, and porphyry system.
They share similarities in type of host rocks, age of host rocks, structural control
in mineralization, age of mineralization, and to some extent, associated alteration
and vein types as well as metal association. The low sulphidation epithermal
prospect areas even share similarities with some of gold deposits in the Bayah
Dome complex.
4. Lubis et al. (2012) conducted research about Geology and Exploration
for Low Sulfidation Epithermal Gold-Silver Mineralization in Kerta, Banten .The
Kerta’s studied area is occupied by a thick sequence of flat lying pyroclastics,
volcaniclastics, and sediments, including Clastic Sediment,Coarse Grained
Volcaniclastic, Tuffaceous Sediment, Rhyolite Lava-Breccias and intruded by
Dacite (Quartz Feldspar Porphyry and Andesite) considered these rocks sequence
to be the Early Miocene Cimapag Formation, the field evidence is not obvious
since the rocks are flat lying, homogene and lack of deformation. Therefore based
on the similarity on rock composition, it is considered that all of these sequences
could be belong to the Pliocene Cipacar Formation or part of the Malimping Tuff.
Kerta displays many hall-marks of a highly dynamic paleo-geothermal system
responsible for low sulfidation adularia-sericite type Au-Ag mineralisation.
6
5. Yuningsih et al. (2012) conducted research about The Arinem Deposit:

An Epithermal Gold-Silver-Base Metal Mineralization System, West Java
Province, Indonesia. The Arinem gold-silver-base metal deposit of Late Miocene
(8.8–9.4 Ma) age is located in the south western mountain part of Western Java
Island, Indonesia. These veins are hosted by andesitic tuff, breccia, and lava of the
Oligocene–Middle Miocene Jampang Formation (23–11.6 Ma) and overlain
unconformably by Pliocene–Pleistocene volcanic rocks composed of andesitic-
basaltic tuff, tuff breccia and lavas. The Arinem vein is composed predominantly
of quartz, calcite, illite and kaolinite, with variable amount of manganese oxide
and limonite and with large amounts of sulfides. The mineralization is also
grouped at least within three stages. Early stage mineralization (stage I) of vuggy–
massive–banded crystalline quartz-sulfide was followed by second stage (stage II)
of banded–brecciated–massive sulfide-quartz and then by last stage (stage III) of
massive-crystalline barren quartz. Arinem deposit is currently the most significant
in the area with the inferred reserve is approximately 2 million tonnes containing
5.7 g/t Au and 41.5 g/t Ag (at a cut-off of 4 g/t Au) for a total of an approximately
12.5 tonnes t of gold and 91.4 tonnes of silver.
6. Sunarie et al. (2011) conducted the research about Gold Bearing Quartz
Veins of Tanggeung Prospect,Cianjur, West Java. His conclusion are The gold
bearing quartz veins in Tanggeung prospect are trending mostly NW-SE and
hosted by mainly tuff of Pleistocene-Pliocene ages. The mineralized quartz veins
exhibit fine–medium crystalline, vuggy, saccharoidal, colloform banding and
massive textures. The mineralized zone occurrences are classified into vein type
and hydrothermal breccias. The Cicelak type vein consists of six mineralized
quartz veins, they are Cicelak, Cigadobras, Pasir Bedil, Cilangkap, honey hole and
ST-7 veins. While the hydrothermal breccias type (Cicengal type) consists of
Cicengal and Cibogo hydrothermal breccias.
7. Subandrio and Basuki (2010) conducted research about Alteration and
Vein Textures Associated with Gold Mineralization at the Bunikasih Area,
Pangalengan, West Java. The Bunikasih vein system in the Pangalengan district of
West Java is a low-sulfidation, adularia sericite epithermal gold deposit. It is
hosted by Late Miocene andesitic volcanic and volcanoclastic rocks occurring in
7
the south western margin of Malabar Volcano complex. Gold ore and alteration
minerals related to deposition of gold in Bunikasih deposits superimposed on Late
Tertiary-Quaternary andesitic formation that were altered and mineralized by
some hydrothermal events. The veins consist almost entirely of quartz, with small
amounts of adularia, bladed calcite, pyrite, and gold. Gold ore shoots are
vertically restricted and are more continuous horizontally.
8. Ismayanto et al. (2009) conducted research about characteristic of
hydrothermal mineralization at Gunung Subang Area, Cianjur District, West Java.
The mineralization in Gunung Subang area is Epithermal Low Sulfidation and
dominated by base metal horizon. There is indication of precious metal horizon at
Cigadobras vein which has higher elevation and lower salinity. This indication
need to be further study, Mineralization occur in vein as dilatation condition at
NE-SW, affected by fault trending in NW-SE. The alterations consist of silisic,
argilic and prophylitic.
9. Harijoko et al. (2007) carried out of characteristics of the Cibaliung gold
deposit in West Java, Indonesia. Middle Miocene (11.18 – 10.65 Ma) low sulfi
dation-type epithermal gold mineralization situated about 70 km west of the
Bayah dome complex. It is hosted by andesitic to basaltic andesitic lavas of the
Middle Miocene Honje Formation (11.4 Ma) and is covered by Pliocene
Cibaliung tuff (4.9 Ma). The exploration estimates mineral resource of
approximately 1.3 million tonnes at 10.42 g/t gold and 60.7 g/t silver at a 3 g/t Au
cut-off. The ore mineral assemblage of the deposit consists of electrum,
naumannite, Ag-Se-Te sulfide minerals, chalcopyrite, pyrite, sphalerite and
galena. Those ore minerals occur in quartz veins showing colloform – crustiform
texture. They are enveloped by mixed layer clay illite/smectite zone, which grades
into smectite zone outward.
10. Warmada et al. (2003, 2007) conducted the research about Polymetallic
Sulfides and Sulfosalts of the Pongkor Epithermal Gold–Silver Deposit, West
Java, Indonesia. The Pongkor gold silver deposit, of Pliocene age (2.05 ± 0.05
Ma), is the largest low-sulfidation epithermal precious metal deposit in Indonesia.
It consists of nine major sub parallel quartz – “adularia” – carbonate veins with
very low sulfide content. The association of sulfides is the main gold carrier and is
8
dominated by pyrite, chalcopyrite, sphalerite, ± galena (grey sulfide – quartz

stage).
11. Utoyo (2007) conducted research about Alteration and Mineralization
of Ancient Caldera in Cupunagara, Subang, West Java. Mineralization is known
by the present of pyrite, which is formed in the epitermal condition. The altered
on rocks consist of 5 zones, i.e: Argilite Silica, Argilite, Propylite Pyrite, Propylite
and Cloritize Andesite. The area is covered by volcanic rocks, which consist of
pyroxene andesite. K/Ar age results yielded between 58,99 ± 1,94 Ma (Upper
Paleocene) and 36,88 ± 3,96 Ma (Upper Eocene).
12. Marcoux and Milési (1994) conducted research about epithermal gold
deposits in West Java, Indonesia. Most of the gold deposits lie within a general
NE-SW belt . They are hosted by steeply dipping dextral NNW-SSE and sinistral
NNE-SSW strike-slipfault resulting from N-S to N 020°E compressional event
related to subduction beneath the island of Java. Epithermal gold mineralization of
the adularia-sericite type in West Java belongs to two distinct vein types, hosted
by intrusive and volcanic rocks of Miocene to Pliocene age.
1.7. Organization of the Thesis
This thesis consists of seven chapters as described below:
Chapter one is introduction with includes the background, problem

formation, objectives, research outcome, scope of research and previous studies.
Chapter two is description about regional geology and mineral deposits in

West Java.
Chapter three is about theoretical backgrounds association with this

research, includes theoretical of mineral deposits, Remote sensing, Geographic
information system (GIS) and Weight of evidence method.
Chapter four is describing about methodology and how to collect the data
and the rule to validation of the result.
Chapter five is explaining how to extract the data, how to generate the data
to create evidence maps like generating the lineaments from ASTER GDEM
data, clay and limonitic alteration from Landsat 5 TM data.
9
Chapter six is giving detail calculation of weight of each evidences map,

how to get binary maps and result of Au potential by using weight of evidence
method.
Chapter seven is discussion on the prospective target areas of Au potential

in study area, also mention about the relationship of structure and lithology with
Au occurrences and conclusion of thesis.
Chapter eight is conclusions and recommendation. This chapter indicates

about the results and some recommendation for next research.
10
CHAPTER II
LITERATURE REVIEW
2.1. Regional Geology of Western Java
2.1.1. Regional Geological Setting
Indonesia is an archipelagic nation in Southeast Asia. There are a lot of

islands, for example, Sumatra, Java, Kalimantan, Sulawesi, and Papua Island.
Tectonically, Indonesia is situated between two continental plates, the Eurasian
Plate and Australian Plate, and between two oceanic plates, the Indian Plate and
Pacific Plate. On the southern part of Indonesia, there is the Sunda Trench which
is formed by the northward subduction of the Indian-Australian plate. This
subduction introduced island arc during Cenozoic time and has been active since
at least Eocene time (Katili, 1975; Hamilton,1979). This subduction is also
generating magmatic arc along the Sumatera and Java island arc that called Sunda
Banda magmatic arc. The Sunda-Banda magmatic arc is the longest arc among
other magmatic arc in Indonesia and extending from north Sumatra through Java
to east of Damar.
Java Island represents part of Sunda-Banda arc due to the subduction of the
northward-moving Indian-Australian Plate beneath the Eurasian Plate, which
consists of a belt of active calc-alkaline volcanos (Claproth, 1989). These features
are built upon older volcanic and volcaniclastic rocks that are intercalated with
Paleogene and Neogene sediments and intruded by small plutonic masses of
composition similar to the volcanics. Basement rocks consist of a melange of Late
Cretaceous or Paleocene age. The early volcanic event produced the so-called
"Old Andesites" (van Bemmelen, 1949), which are exposed mostly along the
south coast of Java and are predominantly tholeiitic in character, whereas the later
Neogene event gave rise to medium to high K calc-alkaline volcanic products that
form a magmatic arc. There are a lot of volcanic activities along this arc and most
of major gold and copper deposits in Sumatera and Java Island is inconection with
this arc. The Sunda-Banda arc is one of the Earth's longest active island arc
systems. The arc extends from the northern tip of Sumatra island through Java to
east of Damar island with a total length of about 4,000 km. This long arc
11
progressively developed from west to east since the Mesozoic, and can be divided
in three segments: the Western Sunda Arc (Sumatra), the Eastern Sunda Arc (Java
to Sumbawa or Flores islands), and the Banda Arc for the islands east of Flores.
The Indian Ocean flow is now sliding approximately northward beneath Java and
Sumatra at a velocity probably near 6cm/yr (Hamilton, 1979).
The main structural features of Java, there are three main strike slip-faults
found in Java (Figure 2.2). In western Java has a still active NE-SW Cimandiri
fault crosscut the whole of West Java. The second strike slip-fault named the
Citandui fault, occurs in western Java and trends NW-SE. this old and inactive
fault was interpreted from gravity data. The third system occurs in central Java,
namely central Java fault as a NE-SW left-lateral strike-slip fault which crosscut
the whole island (Setijadji et al., 2006).
Figure 2.1: Tectonic setting of Java Island, Indonesia (after Setijadji, 2005)
12
Figure 2.2 : Major geological structure of Java Island (Setijadji et al., 2006).
2.1.2. Physiography of West Java

Bogor Zone is an area of hills and Mountains, about 40km across, to the
south the Plain of Jakarta. Hills are composed of strongly folded strata where the
mountains consist of instruction such as Sanggabuwana Complex, West of
Purwakarta.
The Bandung Zone is a series of longitudinal ridges and intermontane

depressions, situated in the central part of West Java, and 40 km wide. The ridges
comprise old sediments, whereas the depressions are filled with young volcanics
and alluvial deposits derived from the surrounding areas.
Bayah Mountain is located in southwest of Jakarta (Figure 2.3). This

geological unit, exposed over an area about 40×80km, consists of Oligocene to
Quaternary calc-alkaline rhyolitic to andesitic rock and small intrusive stocks with
a few intercalations of Miocene limestone and sandstone (van Bemmelen, 1949
and Milesi et al., 1999). The Citorek area, in the Northern part of dome, is large
(60 km2) depression, possibility a caldera, filled with dacitic ignimbrite and
intruded by andesitic dacitic plugs (Milesi et al., 1999).
Physiographically, the Soutern Mountains of West Java region belongs to

the Sunda-Banda magmatic arc (Figure 2.3). This arc has been formed since Early
Tertiary and still active until now. The area is a part of southern slope regional
13
uplift (van Bemmelen, 1949). The Southern mountains, some 50km wide, extend
from Pelabuhan Ratu Bay to Nusakambangan Island. These represent the southern
flank of the Java synclinal structure, an uplifted crustal block dipping to the south.
Rock of Tertiary to recent age occur in this area and consist of andesitic to basaltic
intrusive and extrusive rocks, such as tuff, breccias, and lava flows (Alzwar et al.,
1992). It is sometimes difficult to distinguish intrusive and extrusive rocks from
different ages since the mineralogical composition of those lithologies are similar.
The volcanic rocks are suffered pervasive hydrothermal alteration (Yuningsih et
al., 2011). Most of the primary minerals in the volcanic rocks of the early-Middle
Miocene Jampang Formation are altered. The Jampang Formation (early Middle
Miocene) consists of lava, andesite breccias and propylitic tuff, being a part of the
intruded by quartz diorite (late Middle Miocene). The older rocks are
unconformably overlain by conglomerate and tuffaceous sandstone with
interbedded claystone of the Bentang Formation of Late Miocene-Early Pliocene
age. The volcanic rock unit is dominantly composed of glassy tuff, tuff breccias
and andesitic dyke is uncomfortably overlain by Pliocene. The youngest rocks are
Plio-Pleistocene consisting of glassy tuff, scoria tuffaceous breccias, breccias and
andesitic lava.
Figure 2.3 : Physiographic map of West Java (after van Bemmelen, 1949 and
Martodjojo, 1984 in Suparka et al., 2007)
14
2.1.3. Stratigraphy
The Bogor Basin includes Bogar, Bandung and Southern Mountains Zones
(van Bemmelen, 1949). The sediments in this area are mostly derived from
igneous and sedimentary rocks (e.g. andesite, basalt, tuff and limestone) and are
estimated to be more than 7000 m thick. These rocks are interpreted as gravity
flow deposits in an intra-arc basin overlying Paleogene rocks of fore-arc basin
(Martodjojo, 1984).
The Banten Area (Rangkasbitung Basin) is located in the western part of
West Java . It has changed from fore-arc (Cretaceous – Early Tertiary) to intra-arc
(Late Eocene – Oligocene) through back-arc basins during Early Miocene. The
Rangkasbitung Basin is probably separated from the Bogor Basin by major fault.
Most Quaternary volcanoes are situated in the Bogor Basin (Bronto,1989).
The general stratigraphy of the Bogor Basin is shown in Table 2.1. The
oldest rocks (Cretaceous – Eocene age) in West Java are amelange complex
comprising metamorphosed basic and ultrabasic rocks (peridotites, gabbros and
pillow lavas) and sedimentary rocks (serpentinite, chloritic schists, phyllites and
quartzites) together with chert, black shale, greywacke and limestone.
The Ciletuh melange appears to represent the old subduction zone, whereas
eruptive centres in the Java Sea and an area between Cikotok and Jatibarang may
represent a magmatic arc. The Ciletuh melange appears to represent the old
subduction zone, whereas eruptive centres in the Java Sea and an area between
Cikotok and Jatibarang may represent a magmatic arc. The Ciletuh melange
appears to represent the old subduction zone, whereas eruptive centres in the Java
Sea and an area between Cikotok and Jatibarang may represent a magmatic arc.
The Ciletuh melange appears to represent the old subduction zone, whereas
eruptive centres in the Java Sea and an area between Cikotok and Jatibarang may
represent a magmatic arc.
Pliocene volcanic and sedimentary rocks both occur in West Java. The
volcanic rocks comprise andesitic breccias, lavas, dikes and tuffs which are
locally distributed in the eastern part of West Java. The volcanic rocks form the
Cijulang Formation in the Tasikmalaya quadrangle and the Kumbang Formation
to the east. The maximum thicknesses of the Cijulang Formation and the
15
Kumbang Formation are approximately 1000 m and 2000 m respectively. Many

small intrusions of andesite and dacite, e.g. Sanggabuwana Complex in southwest
Purwakarta, around Ciremai volcano and in Majenang area are considered to be
associated with the Pliocene volcanic rocks (Bronto, 1989).
Table 2.1 : History of the Bogor Basin simplified (Bronto, 1989)

16
2.1.4. Quaternary Volcanism
Java island is currently populated by approximately fifty (50) Quaternary

volcanoes that occupy the median line of the island. Paleocene(?)-Eocene volcanic
centers are identified in few places in Java. These include Bayah dome (Cikotok
Formation) and Jatibarang Volcanic Formation (JVF) in west Java. The Cikotok
Formation is composed of submarine volcanic breccia which was reported to be
Late Eocene-Late Oligocene in age. Meanwhile, the JVF has been dated as Late
Cretaceous-Oligocene in age. Other locations of possible Paleogene volcanics
include the volcaniclastic tuff and gabbro in Ciletuh, west Java (50.1-50.9 Ma).
In west Java, Oligocene volcanics are represented by the Cikotok and
Jampang Formations, and the Cihara granodiorite. Radiometric data include a tuff
sample (33.9±2.0 Ma or Lower Oligocene), a dacite intrusion (32.3±0.3 Ma,), a
basalt dyke (28.1± 6.2 Ma) and a volcanic breccia (33.6±3.8 Ma). In west Java,
such volcanic units are represented by the Cimapag and Citarete Formations, and
the Cihara granodiorite (22.4±0.4 Ma). Andesitic lavas at Cirotan are dated at
15.3±0.7 Ma (Marcoux and Milesi, 1994). To the east the breccia of the Jampang
Formation are exposed at Ciletuh-Ciemas and Pangandaran area, from where a
quartz andesite was dated 22.4±1.5 Ma . In addition a quartz diorite porphyry has
an age of 17.8 and 16.8 Ma ,with lava flows at 17.6±0.6 and 17.9±0.9 Ma.
In western Java the Honje Formation was dated 11.4±0.8 Ma (Harijoko et
al., 2004). In the Bayah dome area, Marcoux and Milesi (1994) reported several
radiometric dates of Upper Miocene from Ciawitali (andesitic pyroclastics,
5.7±0.4 Ma) and Cirotan (rhyolite ignimbrite, 9.6±0.3 Ma). The age of 13.7±1.8
Ma for basaltic lava south of Bayah dome near the beach. The age of 7.2 Ma for a
quartz diorite intrusion. At Cianjur there are many exposures of andesitic
intrusions, some of which have an Upper Miocene age (e.g., 6.0±0.7 Ma). 12.1
Ma for andesite lava from the base of Wayang volcano. At Cineam in west Java,
from 13.5 to 8 Ma for hydrothermal activities related with volcanism and
epithermal mineralization in this area.
Several older volcanic centers such as the Bayah, Ciletuh-Ciemas, and
Jampang were still exist but the new volcanic centers had moved northward from
the older ones. Double volcanic chains were likely to develop at Bayah, Ciemas-
17
Cianjur (western Java), The most important Pliocene rock units are those located
in the Bayah dome region that host the majority of gold mineralization (Marcoux
and Milesi, 1994). Other Pliocene volcanics in west Java are concentrated around
the Bandung Basin. A lava dome in Cianjur area was dated 2 Ma (Soeria-Atmadja
et al., 1994). Near the Saguling dam and Kromong mountain, several dacitic
intrusions and andesitic lava were dated at 3.07 to 4.08 Ma. In the lower part, the
Jampang Formation is dacitic to andesitic composition, and comprises mostly
fine-grained volcaniclastics up to lapilli size. The upper part is dominated by
basaltic breccias and lavas with occasional clasts of limestone. The thickness of
the formation is at least 2000 m.
Figure 2.4: Quaternary volcanoes in Java, with volcano-type terminology follow

Tatsumi and Eggins (1995). Note the presence of several segments of
volcanic configuration (Setijadji, 2005).
2.2. Style of Gold Deposits in West Java
Mineral occurrences in Java are located in the physiographic area known as

the Southern Mountains. This physiographic area contains the most extensive
exposures of Tertiary magmatic rocks in the island that range in age from Eocene
to Upper Miocene (Setijadji et al., 2006; Smyth et al., 2005; Soeria-Atmadja, et
al., 1994), this physiographic area is considered the most prospective region for
metallic mineral deposits. Western Java hosts number of epithermal deposits of
18
precious metals associated with the calc-alkaline volcanism , still active in the
area . Its major metallogenic event during the Miocene and the Pliocene are
located in the Bayah Dome, a Tertiary-Quaternary volcanic structure at western of
Java. There are two noticeably different styles of epithermal gold deposits within
the Bayah dome, both are gold bearing quartz veins; they are known as Breccia
vein of the “Cirotan type,1.7 Ma” and discordant gold-bearing quartz veins or
crustiform banding veins of “Gunung Pongkor type, 2.05 Ma” They are
characterized by very low sulfide content, occurrence of adularia, calcite and
manganese oxide. Gold ore of West Java deposited in many districts (Figure 2.5).
Figure 2.5: Location of gold deposit in West Java (Setijadji, Unpublished data and from
previous research Nurcaho et al., 2012; Syafrizal et al., 2011; Subandrio et
al., 2010; Suparka et al., 2007; Utoyo, 2007 and Ismayanto et al., 2005)
2.2.1. Mineralization in Bayah Dome/Mountains
The West Java gold deposits partly lie within and on flanks of the Bayah
Dome, southwest of Jakarta (Figure 2.6). This geological unit, exposed over an
area about 40×80 km, consists of Oligocene to Quaternary calc-alkaline rhyolitic
to andesitic rock and small intrusive stocks with a few intercalations of Miocene
limestone and sandstone (van Bemmelen, 1949 and Milesi et al., 1999). The
Cikotok area, in the Northern part of dome, is large (60 km2) depression,
19
possibility a caldera, filled with dacitic ignimbrite and intruded by andesitic

dacitic plugs. Most of the important gold deposits of the Bayah dome occur in a
N-S striking structural corridor, which is defined by system of conjugate NNW-
SSE right-lateral and NNE-SSW left-lateral strike-slip faults. The fault zones are
related to N20oE directed compressional regime, and some of the faults are
interpreted as mega tension cracks (Mielesi et al., 1999).
Figure 2.6: Geological map showing the location of the Cirotan Au–Sn–W deposit, as
well as other epithermal Au deposits in the region. Insert shows the
location of major epithermal gold deposits of Indonesia within the plate-
tectonic setting of the region (after Milesi et al., 1994 and Milesi et al.,
1999).
The Gunung Pongkor deposit is a typical low-sulphidation (‘adularia-

sericite’) epithermal vein deposit containing approximately 6.02 million metric
tonnes ore reserves at average grade of 17.4g/t Au and 154.28 g/t Ag (Basuki et
al, 1994). Gunung Pongkor-type including Pongkor, Ciawitali district, Cikidang,
Cibarengkok and possibly Nirmala is typically characterized by quartz-manganese
20
oxides-gold-bearing veins (Marcoux and Milesi, 1994). It is hosted by Miocene

andesitic tuffs and breccias, and a sub volcanic andesite intrusion.
The Cikidang gold deposit (2.4 Ma) is also a typical low-sulphidation
deposit comprising four major sub-parallel quartz-adularia-sericite-calcite veins.
The veins vary from 0.5 to 2.7 m in width and extend for up to 1,000 m in length
trending roughly north east and dip 60 to 86o towards the west. The deposit
contains ore grades varying from trace to 74.9 g/t Au and 1.2 to 225 g/t Ag. The
ore minerals are represented by electrum, aguilarite and pyrite. The Cikidang
system comprises four groups of mineralized quartz veins, named as the Barat
vein, Tengah vein, Cikidang vein and Timur vein (Figure 2.7). The propylitic
alteration is widely distributed in the Cikidang area. The degree of alteration as
shown by the mineral content is different from place to place. This alteration is
indicated by the presence of smectite /chlorite mixed layer mineral, epidote,
carbonate, sericite, quartz and disseminated pyrite. The argillic alteration is found
at the contact with the ore bodies, and is mostly in the hanging walls and/or
between the ore bodies. The argillic zone includes kaolinite, quartz, sericite,
limonite, and carbonate and manganese oxide (Rosana and Matsueda, 2002).
Figure 2.7: Geology and alteration map of the

Cikidang vein, Western Java, Indonesia
(Modified from ANTAM, 1994)
21
The Cikadu-Cisungsang area is located in the Bayah gold district, Banten

Province of Western Java. The base metal mineralization occurred in the
limestone host rocks of Miocene age. The ore is occurring as banded and quartz
vein within the limestone. The mineralized body is identified as hydrothermal
breccias, banded sulfide-quartz and silicified limestone. The ore minerals consist
of sphalerite, galena, arsenopyrite, pyrite, marcasite, silver minerals, chalcopyrite,
and pyrrhotite. Those are occurring as hydrothermal breccias (Rosana et al.,
2006).
The Cirotan deposit is located in the Bayah dome, a Tertiary–Quaternary

volcanic structure covering an area of about 40 by 80 km, which is separated from
the Bandung basin to the east by a major N 70o E left lateral strike-slip fault zone
(Milesi et al., 1994; Malod et al., 1995). The central part of the Bayah dome
contains Oligocene to Miocene andesitic–dacitic volcanic rocks and Pliocene to
Pliocene–Quaternary diorite to andesite intrusions. The Cirotan epithermal gold
deposit is hosted by a system of mineralized fractures in a right-lateral strike-slip
fault, which subsequently evolved into a normal fault (Genna et al., 1996). The
vein structure, which strikes N 10o W and dips 50–60o to the E, extends for a
distance of 800 m and reaches a maximum thickness of 25–30 m. Ore grade
mineralization has been mined over a length of 560 m and vertically from the top
of Gunung Dahbow mountain (843 m) down to a depth of 350 m .The Cirotan
vein cuts Miocene rhyodacitic ignimbrites (K/Ar: 9.5 ± 0.3 Ma), older dacitic to
andesitic lavas (K/Ar: 14.3 ± 0.7 Ma), and a stock of weakly propylitized Pliocene
quartz microdiorite (K/Ar: 4.5 ± 0.3 Ma), which has intruded the Miocene
volcanics. Hydrothermal adularia from the Cirotan deposit has been dated at 1.7 ±
0.1 Ma (Marcoux and Milesi, 1994). Hydrothermal alteration is dominated by
intense silicification, particularly in the footwall, and is accompanied by weak to
intense sericitization; both types of alteration are superimposed onto regional
propylitic alteration. Based on relative age relationships, five distinct
mineralization/alteration stages can be distinguished: (1) wallrock silicification;
(2) siliceous breccia with minor sulphides; (3) polymetallic cockade breccia in
which cockades are characterized by numerous concentric rims of rhodochrosite,
quartz and base metal sulphides; (4) high-grade precious metal ore breccia; and
22
(5) late drusy quartz (Marcoux et al., 1993; Milesi et al., 1994). These bonanza
veins have average gold grades of 9–12 g/t (locally gold grades can attain 700
g/t), and anomalous concentrations of Sn, W, and Bi. Ore textures demonstrate
that electrum is spatially associated with wolframite, cassiterite, and scheelite
(Milesi et al., 1994).
2.2.2. Mineralization in Southern Mountains Range
Some precious and base metal mineralization are also currently discovered
within the Southern Mountains of West Java region including low sulphidation
epithermal systems such as Gunung Subang prospect, Cianjur, Cigaru prospect,
Subabumi, Tanggeung prospect, Arinem prospect, Papandayan district, Garut,
Cihurip/Ciparay porphyry prospect, Garut and Gunung Gupit HS Epithermal in
Magelang (Idrus et al., 2013).
The Gunung Subang LS epithermal prospect is located in the Cianjur
District, West Java province, approximately 120 km southwestern of Bandung,
capital city of West Java province.The mineralized system in Gunung Subang area
is hosted by Lower Bentang Formation (Late Miocene – Pliocene), comprising
crystal tuff, pumiceous tuff, lithic tuff, tuffaceous sandstone with lenses of
andesitic breccia, conglomerate, and tuff breccia. This indicates alteration-
mineralization also occurred associated with magmatism after Pliocene (note: it is
believed that most mineralization in southern West Java is hosted by Oligo-
Miocene rocks). Lithology of this area is composed of tuff and andesitic rock unit
that look place as host rock of mineralization. The NW-SE fault trending is major
of fault in the area. Base on lineament analysis, there is coincident similarity of
trending between extension patterns and vein in NE-SW direction. The
mineralization associated with quartz vein which has NNE-SSW trending. The
alteration type consists of silisification, argilic and prophylitic. The major element
data show the adularia (K-feldspar) – sericite alteration tendency. The mineral
assemblage consists of galena, sphalerite, pyrite, and chalcopyrite and quartz as
dominant gangue mineral. There are indication of boiling zone at Cicelak vein,
which is indicated by the occurrence of bladed calcite and the occurrence of vapor
rich together with other phase in fluids inclusion sample (Ismayanto et al., 2009).
23
One of other gold-base metal prospects in the area is Cigaru prospect,

Sukabumi. It might tend to be included within the Southern Mountain zone of
West Java region as its location is out of Bayah Dome/Mountain and the host
rocks of mineralization are comparable to Southern Mountain stratigraphy (i.e.
Early Miocene Jampang Formation). Geology of this area consists of andesitic
rocks and dacite aged Miocene and layered by tuff volcanic. Gold mineralization
in Cigaru mineralized in main vein system with some minor veins in N320oE-
N345oE and 75-90odips. Main alteration minerals dominated are montmorilonite,
illite, and clinochlore (Syafrizal et al, 2011). Cihar – Cibaliung – Bunikasih –
Tutugan Areas are closed to each other located in about 5-6 km southwest of Situ
Cileunca. The host rocks are andesite lava / flows and breccias, porphyritic with
some mafic mineral phenocrysts, probably Late Miocene-Pliocene in age (based
on regional map by Alzwar et al., 1992). They show argillic and/or prophilitic
alteration and, locally, silicification (Subandrio and Basuki., 2010).
Pakenjeng area is located at south of Arinem area and is one of known
deposits that are hosted by Pliocene volcanic rocks. The area is occupied by
Pliocene volcanic rocks of the Bentang Fm and Quaternary volcanic rocks (based
on regional map by Alzwar et al., (1992)). Pyroxene and hornblende andesite
intrusions are also found (Setiawan et al, 2010). Mineralization in this area is
associated with quartz veins that crosscut volcanic breccia and andesitic-basaltic
lava, and are present in N10-40oE and N32- 330oE trends. Silicified rocks are
found along wallrocks near veins that gradationally change into prophylitic altered
rocks away from the veins. Overprinting of prophylitic mineral assemblage by
argillic is locally found. Calcite, adularia, clay mineral (illite), sericite, and zeolite
are commonly found in association with quartz veins (Setiawan et al., 2010).
In Cimanggu, gold mineralization is associated with N345E trending quartz
vein and hydrothermal breccia that contain chalcopyrite, bornite, chalcosite,
galena, sphalerite, pyrite, marcasite and arsenopyrite. Veins show pinch and
swell-structure associated with erratic high grade Au, indicating a lower level
epithermal and most likely porphyry/intrusive related. The presence of porphyry
related Cu-Au (contain up to 0.3% Cu) has been reported in Cimanggu, which is
associated with quartz stockworks and disseminated fine grained sulfide in altered
24
andesite, diorite, and dacite. The Cipaku prospect appears along the intersection of
N345°E with N295°E trending structural corridors. The high grade quartz veins
are confined to a narrow, pinch and swell veins (maximum 1.5 m) with erratic
good Au grade. In Cijiwa, gold bearing quartz vein associated with N345E
trending structures and the best quartz vein samples returned 1.34 m average
thickness @ 11.65 g/t Au over 175 m strike length associated with various amount
of base metal (Average 7 g/t Au) (Basuki et al., 2012).
Salopa Area is occupied by Oligo-Miocene Jampang Fm. which is
composed of andesitic volcaniclastic breccias, tuffaceous sandstones and shales,
with dacitic dome or sub-volcanic intrusive rock. Milled matrix fluidised breccias
(diatreme) can be observed cross cutting the southern edge of the dacitic unit. The
breccia bodies are elongated in N-S trends, and they are polymictic, with altered
clasts (andesitic to silica-clay altered dacite, and quartz veins) set in tuffaceous
rock flour matrix.
Another important ore deposit type discovered in this region is
Cihurip/Ciparay porphyry Cu-Au prospect, Papandayan district, Garut regency,
West Java. The Churip deposit is hosted by sedimentary volcaniclastic-pyroclastic
rocks of Jampang Formation (early Middle Miocene) consisting of lava, andesitic
breccias and phyrolitic tuff, being a part of the intruded by quartz diorite (late
Middle Miocene). The porphyry Cu-Au occurrence is characterized by the
presence of quartz vein/veinlet stockwork within altered mineralized host rock
(Suparka et al, 2007).
Cijulang HS epithermal gold prospect is situated in Papandayan district,
Garut, West Java. Alteration and mineralization found in volcanic rock of
Jampang and Bentang formations. Alteration types are identified as advanced
agillic (kaolinite, dickite, pyrophyllite, alunite, and sercite-pyrite), massive silica-
vuggy quartz and propylitic (chlorite, smectite, and pyrite). Ore mineralization
occurs as dissemination and frature filling by pyrite, enargite, gold, tennanite,
sphalerite, galena and chalcopyrite associated with advanced argillic and massive
silica-vuggy quartz. Ore grade within advanced argillic is more than 0.4 g/t
(Verdiansyah et al., 2012).
25
Table 2.2: The mineralization of previous studies in West Java

No Name of author Style of gold Mineralization of gold
and year deposits in West deposit and host
Java rock/structural control.
1. Tun et al. (2013) High-sulfidation Mineralization is hosted by

Epithermal Tertiary calc-alkaline
volcanic and volcaniclastics
Prospect in
rocks.Host rocks belong to
Cijulang, West
lava andesite, tuff and
Java
breccia. Typical ore minerals
include pyrite, enargite,
luzonite, tennantite,
chalcopyrite, covellite,
galena, emplectite, and
tellurides. Principal
hypogene hydrothermal
alteration types are silicic
(vuggy/massive), advanced
argillic, argillic and
propylitic.
2. Yuningsih et al. An Epithermal The Arinem deposit is

(2012) Gold-Silver-Base characterized by a quartz–
Metal illite–calcite–sulfide vein,
Mineralization in which occurs along fault
Arinem, West Java planes in Oligocene–Miocene
volcanic rocks.The Arinem
vein trends N20oE to N10oE.
3. Syafrizal et al. Low Sulfidation Mineralization associated

(2011) Epithermal in with Jampang formation,
Kertajaya Village, consists to undersea volcanic
Cigaru, Sukabumi, rocks aged Early Miocene to
West Java Middle Upper Miocene.
Cigaru veins show N320°E -
N345°E orientation and 75°
90° dips.
4. Subandrio and Low-sulfidation, It is hosted by Late Miocene

Basuki. (2010) adularia sericite andesitic volcanic and
epithermal gold volcanoclastic rocks. The
26
deposit in veins consist almost entirely

Bunikasih Area, of quartz, with small amounts
Pangalengan, West of adularia, bladed calcite,
Java pyrite, and gold. Two faults
are NE-SW and the others are
NW-SE.
5. Sunarie et al. (2009) Epithermal low The gold bearing quartz veins
sulfidation type in in Tanggeung prospect are
Tanggeung, West trending mostly NW-SE and
Java hosted by mainly tuff of
Pleistocene-Pliocene ages.
The ore minerals consist of
chalcopyrite, pyrite,
sphalerite, galena and oxide
minerals such covellite,
chalcocite and hematite. The
gangue minerals are mainly
quartz and some clay
minerals.
6. Ismayanto et al. Epithermal Low Mineralization occur in vein

(2009) Sulfidation in as dilatation condition at NE-
Gunung Subang SW, affected by fault
Area, Cianjur trending in NW-SE. The
District, West Java. alterations consist of silisic,
argilic and prophylitic.
7. Harijoko et al. Low Sulfidation- Deposit is characterized by

(2007) Type- Epithermal occurrence of gold-bearing
Gold Deposit in quartz veins. They are
Cibaliung of West enveloped by mixed layer
Java clay illite/smectite zone.
Deposit is hosted by volcanic
rock of the Middle Miocene.
8. Suparka et al.(2007) Porphyry Cu-Au The Churip deposit is hosted

Deposit in Cihurip by sedimentary
and surrounding volcaniclastic-pyroclastic
area, Garut rocks of Jampang Formation
Regency, West (early Middle Miocene). The
Java porphyry Cu-Au occurrence
27
is characterized by the
presence of quartz
vein/veinlet stockwork within
altered mineralized host rock.
9. Warmada et al. Epithermal Gold– The Pongkor deposit consists

(2003, 2007) Silver Deposit in of at least nine major
Pongkor, West subparallel quartz – adularia
Java – carbonate veins rich in
manganese oxides and
limonite in the oxidation
zone, and very poor in
sulﬁdes.The host rock of this
deposit is hosted by Pliocene
age (2.05 ± 0.05 Ma). The
main ore components are
pyrite, chalcopyrite, galena,
sphalerite, electrum, acanthite
– aguilarite and polybasite –
pearceite, with trace amount
of proustite, tetrahedrite and
stromeyerite/mckinstryite
10. Wagner et al. Low-sulphidation The Pliocene Cirotan low-

( 2005) epithermal gold sulphidation epithermal gold
deposit in Cirotan, deposit is characterized by
West Java complex polymetallic
assemblages and progressive
enrichment in Sn–W and Au–
Ag in the late stages of
mineralization. The vein
structure, which strikes N 10o
W and dips 50–60o to the E.
11. Rosana and Low-sulfidation Cikidang deposit is hosted by

Matsueda (2002) quartz-adularia- Volcanic rocks of early
sericite(-calcite) Miocene and it characterized
vein deposits in by argillic and propylitic
Cikidang, West alteration. The vein system
Java comprises four sub-parallel
quartz adularia-sericite (-
calcite) veins that are rich in
manganese oxide and
28
limonite with very poor

amount of sulfides. The vein
trends roughly N-S and dip
60 to 86° toward west.
12. Marcoux and Epithermal gold The Deposit is hosted by

Milési, (1994) deposits in West intrusive and volcanic rocks
Java, Indonesia. of Miocene to Pliocene age.
They are hosted by steeply
dipping dextral NNW-SSE
and sinistral NNE-SSW.
Mineralization consists of
two styles: "Pongkor" type,
and "Cirotan" type.
13. Basuki et al. (1994) Low Sulfidation The Gunung Pongkor deposit
Au-Ag Epithermal occurs in a sequence of
Vein Deposit in Tertiary igneous rocks,
Gunung Pongkor, consisting of tuff breccia,
West Java lapilli tuff and intrusive
andesite. Hydrothermal
alteration is Propylitic
alteration and Argillic
alteration. The main trend of
the veins is northwest
(N 330 ° W), locally
changing to N 30 ° E; dips
vary from 60 to 85 °.
29
CHAPTER III
THEORETICAL BACKGROUND
3.1. Gold Deposits
Most of the important gold deposits, which are considered so primarily for
their gold content, belong to one of the following seven types (Misra, 2000):
a) Young placer deposits

b) Deposits hosted by quartz-pebble conglomerates (Witwatersrand-type)
c) Volcanic-associated epithermal deposits
d) Deposits hosted by banded iron-formation
e) Intrusion related deposits
f) Lode deposits
Knowledge of the geographic distribution and geologic characteristics of
gold deposits is an important part of an analysis of resources. The presence of
deposits confirms that specific ore-forming processes have occurred within the
geologic environments that are believed to be permissive for the occurrence of
certain deposit types, and adds confidence to the delineation of mineral resource
tracts. The presence or absence of known deposits affects the estimates of
undiscovered deposits in a region.
3.1.1. Hydrothermal Gold Deposit

Gold deposits compose the most representative group among hydrothermal
deposits. The theory of hydrothermal ore formation was largely formulated in the
past few decades by study of the composition and physicochemical property of
ore-bearing solutions, source of ore components and ore-forming fluids, and
conditions and mechanisms of ore deposition. The origin of hydrothermal
solutions that take part in the formation of most commercial base metal deposits is
one of the most important and complicated problems of the theory of ore
formation (Betekhtin,1955). Ore vein created by hydrothermal alteration are
comprised of valuable minerals and ore. Many of these veins are mined for gold
copper, lead, and many other valuables. There are many hydrothermal gold
deposits. But in this study area highly focused on porphyry Cu-Au, low
30
sulphidation epithermal, high sulphidation epithermal, and skarn deposit as show

in figure 3.1.
Figure 3.1: Schematic diagram of porphyry, epithermal and

skarn deposits (Kirkham and Sinclair, 199)
3.1.2. Porphyry Cu-Au Mineralization
Porphyry copper deposit refers to large, low grade stockwork to

disseminated deposit of copper (Evans, 1993). The major products from porphyry
copper deposits are copper and molybdenum or copper and gold.The porphyry
copper-gold mineralization occupies in various phases of tonalite porphyry
intrusion which emplaced along the contact of the andesitic volcaniclastic
succession and quartz diorite. The highest grade copper gold mineralization and
quartz vein are hosted by the old tonalite (>1% Cu and >1 g/t Au), the adjacent
volcanic and, quartz diorite and associated intrusive breccia (Mitchell et
al.,1998). Porphyry Cu-Au deposits develop as a result of focusing of the
mineralizing fluids at depths of 1 – 2 km in the cooler apophyses to magmatic
sources at greater depths, and so extend from intrusion host rocks into the wall
rocks. Some of the better ore systems are characterized by multi-phase intrusion
emplacement into spine-like vertically attenuated intrusion complexes.
Overprinting intrusions provide multiple events of mineralization and locally
recycle ore minerals into settings with higher metal grades, but may also overprint
and obliterate mineralization related to earlier porphyry Cu-Au intrusions,
therefore downgrading the total ore system (Corbett, 2004).
31
Many porphyry Cu-Au deposits are not linked to associated extrusive

volcanic rocks suggesting volatiles and mineralisation may have been retained
within the cupola rather than vented. Initial intrusion emplacement is
characterised by zoned prograde alteration grading outwards as potassic
(magnetite, secondary biotite and K-feldspar), to inner propylitic (actinolite,
epidote), and outer propylitic (chlorite, carbonate) alteration . Porphyry Cu-Au
deposits have principal ore minerals are chalcopyrite, bornite, chalcocite,
tennantite, other Cu minerals, native Au, electrum, and tellurides; associated
minerals include pyrite, arsenopyrite, magnetite, quartz, biotite, K-feldspar,
anhydrite, epidote, chlorite, scapolite, albite, calcite, fluorite, and garnet (Sinclair,
2007).
Porphyry Cu-Au deposits display complex patterns of zonation in alteration

and mineralisation which result from the overprinting of many prograde and
retrograde events as figure 3.2.
Figure 3.2: Hydrothermal alteration zones of porphyry Cu-Au deposits (after Sillitoe,
1995)
32
3.1.3. Epithermal Gold
3.1.3.1. Terminology
Epithermal deposits occur largely in volcano-plutonic arcs (island arcs as

well as continental arcs) associated with subduction zones, with ages similar to
those of volcanism. The deposits form at shallow depth, of <1 km – 2km, and are
hosted mainly by volcanic rocks. Epithermal deposits occur as small vein systems
(less than a million tons in size) and breccia pipes, but with good grades.
Epithermal systems are predominantly “near surface” features that can be
associated with the development of gold, silver and base metal mineralization
mainly subaerial hydrothermal systems. Such hydrothermal systems commonly
develop in association with calc-alkaline to alkaline magmatism, in volcanic arcs
at convergent plate margins, as well as in intra-arc, back-arc, and postcollisional
rift settings (Simmons et al., 2005; Sparkes, 2012).
Epithermal Au-Ag deposits are distinguished as high and low sulphidation

on the basis of ore and gangue mineralogy, derived from distinctly different fluid
typess, and for the LS deposits there is a further distinction between the group of
base metal rich deposits which commonly display a relationship with intrusion
source rocks, and the banded adularia-sericite style quartz veins Figure 3.3
(Corbett, 2002). Many epithermal Au-Ag deposits contain flat dipping ore zones
which might be accounted for by factors such as host rock competency,
permeability, structure, or zones of fluid mixing, with no relation to fluid boiling.
Permeability controls to mineralisation include variable combinations of structure
(dilational feeder structures), lithology (permeable tuffs) and breccias (often
phreatomagmatic breccias in flow dome complexes) (Corbett, 2007b).
Epithermal Au deposits are commonly consider to comprise one of three

subtypes: high sulphidation, intermediate sulphidation, and low sulphidation, each
denoted by characteristic alteration mineral assemblages, occurrences, textures,
and, in some cases, characteristic suites of associated geochemical elements (e.g.
Hg, Sb, As, and Tl). Base metal (Cu, Pb, and Zn) and sulfide minerals may also
occur in addition to pyrite and native Au or electrum (Corbett, 2009).
33
Table 3.1: Characteristics of epithermal systems (after Mitchell and Leach 1991)
Depth of formation Surface to 1.000m.
Temperature of formation 50-300oC
Thin to large veins, stockworks, disseminations,
Form of deposits
replacements.
Open-space filling, crustification, colloform banding,
Ore textures
comb structure,brecciation.
Ore elements Au, Ag, As, Sb, Hg, Te, Tl, U, Pb, Zn, Cu.
Silicification, argillization, sericite, adularia,
Alteratlon
propylitization.
Fine-grained chalcedonic quartz, quartz pseudomorphs
Common features
after calcite, brecciation.
3.1.3.2. Classification of Epithermal Deposit
a. High Sulfidation
High-sulfidation systems are characterized by development of broad

alteration haloes, resulting from high temperature, acidic, oxidized, hydrothermal
fluids, which result in pervasive alteration of the original host rock (Sparkes,
2012).
High sulfidation Au-Ag deposits develop as a magmatic vapour rich fluid
depressurises during the rapid rise to higher crustal levels without interaction with
ground waters, and so forms an extremely acidic fluid. The progressive cooling
and neutralisation of this acid fluid by wall rock reaction produces the
characteristic zoned alteration which is overprinted by sulphides, typically pyrite-
enargite and local covellite, with a gangue including alunite and barite (Corbett,
2009).
High sulfidation Au + Cu ore systems develop from the reaction with host
rocks of hot acidic magmatic fluids to produce characteristic zoned alteration and
later sulphide and Au + Cu + Ag deposition. Ore systems display permeability
controls governed by lithology, structure and breccias and changes in wall rock
alteration and ore mineralogy with depth of formation. One of the challenges is to
distinguish the mineralised systems from a group of generally non-economic
acidic alteration styles including lithocaps or barren shoulders, steam heated,
magmatic solfatara and acid sulphate alteration.
34
b. Intermediate-sulfidation
Intermediate sulfidation gold systems also occur in mainly in volcanic

sequences of andesite to decide composition within calcite-alkaline volcanic arcs.
Some Au-rich IS systems are spatially associated with porphyry systems. Igneous
rocks as silicic as rhyolite are related to a few intermediate sulfidation deposits.
These deposits form from fluids spanning broadly the same salinity range as those
responsible for the high-sulfidation type, although Au-Ag, Ag-Au, and base
metal-rich Ag-(Au) subtypes reveal progressively higher ore-fluid salinities
(Sillitoe and Hedenquist, 2003).
c. Low-sulfidation
Low-sulfidation epithermal deposit is the classic style of epithermal deposit,

forms in a distal position relative to the heat sources. However, the term may
reflect the degree of water-rock interaction, rather than the actual distance from
magma (Hedenquist and Arribas, 2000).
 Low sulfidation epithermal Au-Ag deposits are distinguished from

high sulphidation deposits primarily by the different sulphide mineralogy (pyrite,
sphalerite, galena, chalcopyrite) typically within quartz veins with local carbonate,
and associated near neutral wall rock alteration (illite clays), deposited from dilute
hydrothermal fluids (Corbett and Leach, 1998). Mineralisation grades temporally
and spatially from initial quartz sulphide Au + Cu mineralisation to that termed
carbonate-base metal Au. Low sulphidation epithermal-quartz Au-Ag deposits,
which display links the quartz sulphide Au deposits, are gangue-poor and may
host bonanza Au. The low sulfidation epithermal banded chalcedony-ginguro Au-
Ag deposits (formerly termed adularia-sericite epithermal Au-Ag) host local
bonanza Au-Ag in the ginguro sulphidic layers and may display links to deeper
level polymetallic veins, and are common in rift settings.
 Low-sulfidation silver-gold-base metal deposits: They show many of
the textural and alteration characteristics of gold-silver deposits, but are usually
dominated by silver and lead-zinc mineralization. Gold may be present, but is
typically (though not always) much less significant than silver. Copper may be
significant at deeper levels. Manganese carbonates are common as gangue, and
35
adularia and fluorite may be present (White and Poizat, 1995). These deposits
commonly extend much deeper than low-sulfidation gold-silver deposits, and to
higher paleotemperatures. In some cases there may be an increase in thin minerals
at greater depths.
Table 3.2: Summary of relationships between sulfidation state of Ore-Forming

Environment, related Igneous Rock Compositions, and Tectonic Setting (Sillitoe and
Hedenquist, 2003).
Sulfidation Igneous rock
Tectonic setting
state composition
Magmatic arc in a neutral to mildly extensional
Calc-alkaline, stress state; compressive stress state
High
andesite-dacite uncommon but serves to suppress volcanic
activity
Calc-alkaline, Magmatic arc in a neutral to mildly extensional
Intermediate
andesite-rhyolite stress state; compressive stress state rare
Calc-alkaline,
alkaline, tholeiitic Magmatic arc undergoing extension leading to
Low
bimodal basalt- rifting; postcollisional rifting
rhyolite
Figure 3.3: Schematic cross section through an intrusive centered hydrothermal system
outlining the environments of porphyry, high-sulfidation and low-sulfidation
systems (Hedenquist and Lowenstern, 1994 in Sparkes, 2012).
36
Table 3.3: Characteristic of epithermal deposit types (after Cook and Gemmel, 1996)
Low sulfidation High sulfidation
Host rocks Acid to intermediate Acid to intermediate

subaerial volcanic, and subaerial volcanic, and
underlying basement underlying basement
rocks of any type. rocks of any type.
Localizing controls Any faults or fracture Major regional faults or

zones especially closely subvolcanic intrusions.
related to volcanic
centres.
Depth of formation Mostly 0 to 1000 m. Mostly ?500 to ?2000m
Temperature of 100 to 320oC (mainly 150 100 to 320oC

formation to 250oC)
Character of ore fluids Low salinity. Mostly low (some high)

salinity.
Meteoric waters, Magmatic fluid source
interaction with mixing with meteoric
magmatic fluid possible. waters
pH near neutral, may
become alkaline from pH acid from magmatic
boiling; phase separated HCl, and by
gases may be oxidized to disproportionation of
produce an acid fluid. SO2 , becomes
Reduced. neutralized by wall-rock
Total S content typically reaction, and dilution
low. Base-metal content Oxidized.
low (Pb,Zn).
Total S content typically
high. Base-metal content
may be high (Cu).
37
Table 3.3: Characteristic of epithermal deposit types (Continued)
Low sulfidation High sulfidation
Associated alteration Extensive propylitic Extensive propylitic alteration in

alteration in surrounding surrounding regions with low
regions with low water: water: rock ratios.
rock ratios. Deep deposits have intense
Intensive white mica in pyrophyllite-white mica
regions with high water: alteration.
rock ratios.
Clay alteration becomes Shallow deposits have ore core if
dominant with decreasing massive silica (from acid leaching
temperature. Boiled off and silica mobilization), with
gases may produce narrow margin of alunite and
argillic and advanced kaolinite, out to white mica and
argillic alteration interlayered clays. Near-surface
peripheral to, or deposits may have pervasive clay
overlapping alteration alteration.
from deep fluids
Character of Ore mineralization Ore mineralization typically
mineralization characterized by open disseminated, either in white
space and cavity filling, mica-pyrophyllite, or in massive
typically with sharp silica. Open space and cavity
walled veins. Layered filling not common.
vein fillings typical,
commonly with multi-
stage brecciation.
Near surface may be Mineralization usually associated
stockwork or with advanced argillic alteration,
disseminated, depending and pyrite typically very
on nature of local primary abundant.
and secondary
permeability.
Characteristic Crustification banding, Vuggy silica (fine-grained
textures fine comb texture, quartz). Massive silica (fine-
colloform banding, grained quartz).
banded quartz-
chalcedony, drusy
cavities, vugs, vein
breccia, silica
pseudomorphs after
bladed calcite (lattice
texture).
Characteristics Chalcedony veins Chalcedony mostly absent.
mineralogy common. Adularia in Adularia absent. Alunite may be
veins and disseminated. abundant. Pyrophyllite may be
Alunite minor. abundant. Enargite-luzonite
Pyrophyllite minor. typically present.
Enargite-luzonite absent.
38
3.1. 3.3. Base Metal Veins
Among the deposit formed relatively near the surface by ascending thermal
waters in genetic connection with igneous rocks, ores rich in the base metals and
worked principally for these metals are rather exceptional (Lindgren, 1933). The
ore occur mainly in the interstices of a brecciated rhyolite in a series of fractures
arranged in almost circular form. The pyrite is often crystallized; the gangue is
quartz the ore is poor in gold and silver. When the solutions depositing veins in
volcanic rocks leave the flows and enter into the surrounding limestones and other
sedimentary rocks, deposition by selective precipitation comes into play and ores
rich in sulphides, particularly galena, may be formed (Lindgren, 1933).
The metals component of the vein filling is zoned with respect to the
boiling level: base metals (Pb, Zn, Cu) tend to be deposited below it, while silver
and gold are dominantly deposited above the boiling level. Boiling may occur at
different elevations for different mineralizing episodes (in response to the degree
of pressure buildup before rupture), so a broad transition zone often exists
between the precious metals rich upper part of the vein and the more base metal
rich root zone. In the most extreme cases the boiling level may change abruptly by
hundreds of meters during the life of the hydrothermal system (Simmons, 1991).
This can result in temporally separate stages of precious metal and root
zone mineralization occurring side by side in “composite veins”, or repetitions of
the complete zoning separated by 100 meters of barren in “stacked” veins.
 Characteristics:
Vein deposits include most gold mines, many large silver mines and a few
copper and lead-zinc mines.
Veins commonly consist of quartz (sometimes of several varieties such as

amethyst or chalcedony) usually occurring as interlocking crystals in a variety of
sizes or as finely laminated bands parallel to the walls of the vein. Minor amounts
of sulphide minerals and other gangue minerals such as calcite and various
clay minerals often occur; gold is rarely visible. Gold may be associated with
pyrrhotite, arsenopyrite, pyrite, and chalcopyrite and with minor sulphides – the
classic 'free gold'.
39
3.1.4. Skarn Deposit

Skarns are generally thought of as being the result of contact metamorphism
of impure limestone. However, although the majority contains at least some
limestone skarns can form during regional or contact metamorphism and from a
variety of metasomatic processes involving fluids of magmatic, metamorphic,
meteoric, and/or marine origin (Meinert, 1992).
Skarn deposits are developed by Ca-Fe-Mg-Mn silicates which are formed

by metasomatic replacement of carbonate rocks during either contact or regional
metamorphic processes (Einaudi et al., 1981), as a respon for intrusion effect from
many kind of its composition. A descriptive skarn classification based on the
dominant economic minerals, including copper, iron, zinc-lead, gold, tin,
tungsten, and molybdenum. Each class of skarn deposit has a characteristic,
though not necessarily unique, size, grade, tectonic setting, granitoid association,
and mineralogy (Evans, 1993).
Skarn Au deposit is associated with diorite-granodiorite plutons and

generally containing of sub-economic mineralizations of Cu, Pb, Zn. K-feldspar,
scapolite, vesuvianite, apatite, and Cl-rich amphibole are common.The presence
of arsenopyrite and phyrhottite as main sulfides indicate that these menerals
formed in reduction environment and most of Au occurs as electrum.
Most important geological characteristics of skarn (Meinert, 1998, 1992)

are:
 Deposits are hosted by carbonate rocks (limestones, dolostones, etc)

associated with igneous activity, typically with plutons or batholiths.
 Deposits are epigenetic, formed at high temperature (>250º C), cut across
host rocks, and have sharply discordant contacts with host rocks.
 Depth of formation varies from near-surface to about 10-12 km and controls
size, geometry, and type of alteration.
 Deposits cluster around igneous intrusions that typically include other
intrusion-related deposit types (e.g., porphyry deposits, polymetallic veins).
40
 Deposit types and metals are zoned spatially with respect to intrusions such
that copper and gold are proximal to intrusions; zinc and lead are distal to
intrusions.
 Ore-mineral assemblages typically include pyrite or pyrrhotite, and other
sulfide minerals, and calcsilicate and carbonate gangue minerals.
 Local concentrations of ore can consist of massive concentrations (>50%) of
sulfide minerals and typically ore minerals are coarsely crystalline.
 Deposits form in stages, overprinting one another in time and space.
 Late-stage pyrite and other alteration minerals can be disseminated over a
wide area
Figure 3.4: Stages in the development of skarn deposits (Evans, 1993)

41
3.2. Remote Sensing (RS)

Remote sensing provides information on the properties of the surface of
exploration targets that is potentially of value in mapping alteration zones and
lithology units (Dehnavi et al, 2010). Remote sensing is the science of acquiring,
processing, and interpreting images and related data, acquired from aircraft and
satellites that record the interaction between matter and electromagnetic energy
(Sabins, 1999).
Remote sensing images in mineral exploration are well understood by
enhancing geological map, regional lineaments and structural trends. Local
fracture and faults are delineated for localized ore deposits and recognize
hydrothermally altered rocks by their spectral signatures. Landsat thematic
mapper (TM) satellite images are widely used to interpret both structure and
hydrothermal alteration. Digitally processed TM ratio images can identify
two assemblages of hydrothermal alteration minerals; iron minerals, and clays
plus alunite (Sabins, 1997, 1999).
3.2.1. Landsat Images
NASA has launched two generations of unmanned Landsat satellites that

have acquired valuable remote sensing data for mineral exploration and other
applications. Both generations were placed in sun-synchronous orbits that provide
repetitive images of the entire earth, except for the extreme polar regions
(Sabins,1999).
Landsat Thematic Mapper (TM) images consist of seven spectral bands

with a spatial resolution of 30 meters for Bands 1 to 5 and 7. Spatial resolution for
Band 6 (thermal infrared) is 120 meters, but is resampled to 30 m pixels.
Approximate scene size is 170 km north-south by 183 km east-west (106 mi by
114 mi). Landsat Thematic Mapper (TM) data have been used to detect the
presence of altered rock associated with ore mineral deposition.
Landsat Enhanced Thematic Mapper Plus (ETM+) images consist of
eight spectral bands with a spatial resolution of 30 meters for Bands 1 to 7. The
resolution for Band 8 (panchromatic) is 15 meters. All bands can collect one of
two gain settings (high or low) for increased radiometric sensitivity and dynamic
42
range, while Band 6 collects both high and low gain for all scenes. Approximate
scene size is 170 km north-south by 183 km east-west (106 mi by 114 mi).
Table 3. 4: Characteristics of Landsat 4-5 TM and 7 ETM + (USGS, 2013).
Wavelength Resolution
Band Useful for mapping
(μm) (m)
Bathymetric
Band 1 – blue 0.45-0.52 30 mapping,distinguishing soil
from vegetation and deciduous
from coniferous vegetation
Emphasizes peak vegetation,

Band 2 – green 0.52-0.60 30 which is useful for assessing
plant vigor
Band 3 – red 0.63-0.69 30 Discriminates vegetation slopes
Band 4 – Near Emphasizes biomass content

0.77-0.90 30
Infrared and shorelines
Band 5 – Short- Discriminates moisture content

1.55-1.75 30 of soil and vegetation;
wave Infrared
penetrates thin clouds
Band 6 – Thermal 120*(30) Thermal mapping and estimated

10.40-12.50
Infrared 60*(30) soil moisture
Band 7 – Short- Hydrothermally altered rocks

2.09-2.35 30
wave Infrared associated with mineral deposits
Band 8
Panchromatic 15 meter resolution, sharper
0.52-0.90 15
(Landsat ETM+ image definition
only)
Moreover, TM Band 6 was acquired at 120 m resolution, but products

processed before February 25, 2010 are resampled to 60 m pixels. Products
processed after February 25, 2010 are resampled to 30 m pixels. ETM+ Band 6 is
acquired at 60 m resolution. Products processed after February 25, 2010 are
resampled to 30 m pixels.
43
3.2.2. ASTER GDEM
The Advanced Space borne Thermal Emission and Reflection Radiometer

(ASTER) is an advanced multispectral imager that was launched onboard
NASA’s Terra spacecraft in December, 1999. ASTER has three spectral bands in
the visible near-infrared (VNIR), six bands in the shortwave infrared (SWIR), and
five bands in the thermal infrared (TIR) regions, with 15m, 30m, and 90m ground
resolution, respectively. The VNIR subsystem has one backward-viewing (27.70
off-nadir) instrument for stereoscopic observation in the along-track direction,
making imagery acquired by this satellite suitable for DEM generation. Other
important properties that make ASTER data suitable for DEM generation are the
platform altitude (705 km) and its base-to-high ratio of 0.6 (Abrams, 2000)
Digital elevation models (DEMs) are increasingly used for visual analysis of
topography, landforms, as well as modeling of surface processes. DEM data can
be used for visualization and interpretation of the area in terms of geology and
geomorphology. In geological investigation we can delineate the changes in
structures of the area and in the geomorphological features. We can identify the
geomorphological features like lineaments, faults, fractures but not in details.
DEM data, however cannot be ignored all together as it can be very useful in
giving an overview of the geology and geomorphology of any area especially
where any other data (Sharma and Kujur, 2012).
3.3. Software Defoliant Method
An image enhancement technique based on directed principle components

(DPC) analysis of two band ratio images is called the software defoliant technique
(Fraser and Green, 1987, in Carranza, 2002). The input band ratios are selected on
the basis that one band ratio contains information related to the component of
interest (hydrothermal alteration), the spectral response of which suffers
interference from the spectral response of another component (vegetation). The
second band ratio should thus contain information about this spectrally interfering
component. The input band ratio images, which are computed from unstretched
data, need to be standardised through a histogram equalization stretch so that their
range extends from 0 to 255 and their means approximate 128. Once computed,
44
the DPC that has loadings of similar signs on both input band ratio images
explains the variance due to similarities in the spectral responses of the interfering
component and the component of interest. The other DPC, whose loadings are of
different signs on either of the input band ratio images, highlights contributions
unique to each of the components. The sign of the loadings dictates whether the
component of interest is represented as bright or dark pixels in the DPC image; a
positive loading implies bright pixels, a negative loading implies dark pixels.
Based on the generalized reflectance curves of selected cover types, band ratio 4:3
serves as an excellent vegetation indicator while band ratio of 3:1 and 5:7 are
commonly used to enhance iron oxides and clay, respectively (Fraser and Green,
1987, in Carranza, 2002).
Landsat TM bands 5 and 7 are thus potentially useful in detecting

clay zones. In the spectral region covered by Landsat TM band 3, iron oxides
and clays have high reflectance whilst vegetation has strong absorption. In the
spectral region covered by Landsat TM band 4, vegetation has strong
reflectance whilst iron oxides and clays show absorption features. Landsat
TM bands 3 and 4 can thus be used to differentiate areas of iron oxides and/or
clays from areas of vegetation (Carranza, 2002).
Figure 3.5: Typical reflectance spectra of vegetation, iron oxides and clays
(Fraser and Green, 1987, in Carranza, 2002).
45
3.4. Geographic Information System (GIS)
A Geographic Information System (GIS) is a computer system for

capturing, storing, querying, analyzing, and displaying geospatial data. The word
geographic implies that location of the data items are known, or can be calculated
in terms of geographic coordinates (latitude, longitude). The word information
implies that the data in GIS are organized to yield useful knowledge, often as
coloured map and images, but also as statistical graphics, tables and various
onscreen responses to interactive queries. The word system implies that a GIS is
made up from several interrelated and linked components with different function
(Bonham-Carter, 1994).
The Geographic Information System (GIS) now is used in many fields

among such as agriculture and land use, forestry and wildlife management,
monitoring of desertification, archaeology, city planning, municipal applications
and geology. The applications of GIS in geology are very broad and diverse.GIS
also can be used to generate a mineral potential map, in the absence of
comprehensive systematic mineral exploration programs, it is important to
develop alternative methodologies of mineral potential classification.
Mineral exploration procedures always need to integrate data in order to

consider a vast range of combinations and to underline different hypotheses. The
analysis of spatially located data is one the basic concerns of exploration
geologists, and can be more efficiently executed with assistance of Geographical
Information Systems (GIS). One of the major applications of a GIS is the ability
to integrate and combine multiple layers of lithology, structure, geophysical and
geochemical characteristics to delineate mineral prospectivity map (Washington et
al, 2003).
3.4.1. Mineral Potential Mapping using GIS
Mineral potential assessment or classification is a multi-stage activity with

the ultimate objective of delineating mineralized zones that can be exploited under
prevailing economic conditions. Mineral potential is characteristics attributed to a
particular area that describes the probability for the presence of mineral deposits
or existence mineralization and determined by how well the geological and
46
mineral deposit data fit established mineral deposit models and existing
knowledge about the mineralization of particular area (Carranza, 2002).
Mineral potential mapping (MPM) is a process of spatial prediction that has

the objective of delineating area with different probabilities of hosting certain
types of mineralization. Mineral potential mapping is carried out by integrating
different types of geo-data that provide useful indications of mineral deposits of
interest. A GIS is therefore useful for mineral potential mapping. The main steps
in generating mineral potential maps are (Wright and Bonham-Carter, 1996):
1. Establishing the exploration conceptual model
2. Building a spatial database
3. Spatial data analysis (extraction of evidence maps and assigning of
weights
4. Combination of evidence maps to predict mineral potential
Mineral resource potential mapping is a very complex analytical procedure

which requires simultaneous consideration of a number of spatial evidences –
geological, geomorphological, structural, geochemical, geophysical etc. The
geologically-constrained predictive mineral potential maps are based on two
factors: favorability and validity. Favorability is determined by integration of
geological variables that are considered essential for mineral occurrence. Validity
is determined by how well the predictive models delineate correctly know mineral
deposits that were not used to generated the models (Carranza, 2002).
The methodology known as ‘mineral potential mapping’ has been used for
many years to predict the most likely locations in which various minerals and
metals might be found (Bonham-Carter, 1994). The mineral potential map will
show the locations of mineral deposit and ranking of mineral potential sites,
where is high or low potential in area for interpret detailed of mineral for next step
of exploration and mined (Harris, 2006).
47
3.4.2. GIS Modeling Techniques for Producing Mineral Potential Maps
There are two types of GIS based mineral potential mapping methods,
empirical methods and conceptual methods (Table 3.5). Empirical methods are
data driven and involve the examination and analysis of the locations of known
deposit with respect to the surrounding geology (Bonham-Carter et al., 1989).
Data driven approaches require that ‘a prior’ knowledge (expressed in term of a
prior probability) exists in the form of mineral deposits or occurrence (e.g.,
prospects) for the study area. Spatial relationships between the input data
(evidence maps) and the spatial location of the mineral prospects are used to
establish the importance (weight of evidence maps) and the spatial location of the
mineral prospect are used to establish the important (weight) of each evidence
map. In the other data driven approaches, training area can be established over
each mineral deposit from which diagnostic signatures of the mineralization can
be calculated from the various data (e.g., geochemistry, geophysics etc.) used in
the modeling process. Methods such as logistic regression, weight of evidence
(WofE) (Bonham Carter, 1994), decision tree analysis and neural networks are
examples of data driven approaches.
Knowledge-driven approaches rely on the geologist’s input (expert opinion)

to weight the importance of each data layer (evidence map) as they relate to
particular exploration model being used. This approach is more subjective but has
the advantage of incorporating the knowledge and expertise of the geologist in the
modeling process. Examples of knowledge-driven approaches include Boolean
logic, index overlays, analytical hierarchy process (AHP) and fuzzy logic.
Weight of Evidence Method was used in this thesis , this method is a

quantitative data-driven method used to combine datasets. It uses a log-linear
from of the Bayesian probability model to estimate the relative importance of
evidences by statistical means.
48
Table 3.5: The models exist for mapping mineral potential (Bonham-Carter, 1994;
Harris, 2006)
Type Model Parameters Example
Logistic regression
Calculated from
Data-driven Weight of Evidence
training data
Neural Network
Fuzzy logic
Dempster-Shafer belief theory
Estimated by an
Knowledge-driven Analytical Hierarchy Process
expert
Boolean Operations
Index Overlay
3.5. Weight of Evidence Method
The weight of evidence method uses the Bayesian approach of combining

datasets, which is based on a probability framework. The method developed by
the Geological Survey of Canada has been in use since the late 1980s (Bonham-
Carter et al., 1989).
In the Bayesian approach, prior and posterior probabilities are amongst the
most important concepts. Given an area of study that contains certain number of
mineral deposits, the prior probability that a deposit occur per unit area is
calculated as the total number of deposits over the total area. This initial estimate
can be later increased or diminished in different areas by the use of other
evidences. The prior probability can be multiplied by a factor to obtain a posterior
probability that a deposit occurs given certain evidence. The posterior probability
can in turn be multiplied by a second factor (calculated from another evidence)
being used, in this way, as a prior probability to obtain a second posterior
probability.
Given a study area, T, composed by a number of unit area N {T} (unit cell
in a raster map) containing a number of deposits N {D} (Figure 3.6) and assuming
that each deposit occupies one unit area (one unit cell), the probability that one
randomly selected unit area or cell in the map contain a deposit when on other
information is available, is:
49
N D
P D  (3-1)
N{T}
The favorability of finding a mineral occurrence given the presence of a
predictor pattern is given by
P D  B P B/D
P D/B = =P{D} (3-2)
P{B} P{B}
Where P{D/B} is the conditional or posterior probability of a mineral
occurrence given the presence of the predictor pattern, P{D/B} is the conditional
probability of being in the predictor pattern B, given the presence of a mineral
occurrence D, P{B} is the prior probability of being in the predictor pattern.
Figure 3.6: Venn diagram of schematic spatial

overlap relationships between
binary predictor pattern and binary
deposit pattern.
The favorability of finding a mineral occurrence given the absence of a

predictor pattern is expressed by

P DB  =P{D} P B/D
 
P D/B =
P{B} P{B}
(3-3)
Where P{D/ B } is the conditional probability of a mineral occurrence given

the absence of a predictor pattern, P{ B /D} is the conditional probability of the
absence of a predictor B given the presence of a mineral occurrence. P{ B } is the
probability of being in the absence of a predictor pattern.
50
Probability (P) can be expressed as odds (O) or vice versa using the
equation:
O=P/ (1-P) (3-4)
Therefore, equations 3-2 and 3-3 can be expressed, respectively, in an odds

formulation as:
P B/D
O D/B =O{D} and (3-5)
P{B/D}
O{D/ B}=O{D}
 
P B/D
(3-6)
P{B/ D}
where O{D/B} and O{D/ B } are respectively the posterior odds of a

mineral deposit given the presence and absence of a binary predictor pattern, and
O{D} is the prior odds of a mineral deposit. The terms P{B /D} / P{ B/ D } and
P{ B /D} / P{ B /D } are known as the likelihood sufficiency ratio (LS) and
likelihood necessity ratio (LN), respectively (Bonham-Carter, 1994).
The weights of evidence for the binary map relationships are defined as
P B/D
W  =log e (3-7)
P{B/ D}
W  =log e
 
P B/D
(3-8)
P{B/D}
Where W+ and W− are the weights of evidence when a binary is present

and absent respectively.
The variances of the weights can be calculated by the following expressions
(Bishop et al., 1975; Agterberg et al., 1990):
1 1
S2 (W + )= + and (3-9)
N{B  D) N{B  D}
1 1
S2 (W - )= + (3-10)
N{B  D) N{B  D}
51
The contrast C, defined as
C=W + - W - , (3-11)
Provides a useful measure of the spatial association between a binary

predictor pattern and the mineral occurrence points. For a positive spatial
association, C is positive and for a negative association C is negative. The
standard deviation of C is calculated as
S(c)= s 2 (W + )+s 2 (W - ) (3-12)
The studentised C, defined as
sigC = c/stdc (3-13)
Serves as a guide in selecting the optimum proximity distance to a geologic

feature according to Boham-Carter (1994) and shows the statistical significance of
such a spatial association.
Using the log-odds formulation of Bayes’s rule, two or more binary

predictor patterns can be combined to generate a predictor map using the
expression
n
log e O{D/B1k  B2k  B3k .......Bnk =W1k  loge O{D}, (3-14)
j 1
Where the superscript k is positive (+) or negative (-) if the binary predictor
pattern is present or absent, respectively (Carranza, 2002).
3.5.1. Test Conditional Independence
Conditional Independence (CI) is assumed to exist when combining two or

more maps with the Bayesian models. Where weights of evidence are being
calculated from data, CI can be checked with statistical test to show the magnitude
of the problem, and pinpoint the maps that are causing the most difficulty. These
maps can be reject or modified to reduce the problem.
52
Conditional independence between the input binary predictor patterns was

determined through a pairwise test (between two binary predictor patterns) and/or
through an overall test (among all input binary predictor patterns). The conditional
independence between or among the binary predictor patterns was tested prior to
(by pairwise test) and/or after combining them (by an overall test).
1. Pairwise test of Conditional Independence
The connotation of pairwise conditional independence can be shown

graphically by a Venn diagram as illustrated in Figure 3.7. Conditional
independence implies that the number of pixels representing the spatial
coincidence between the binary predictor patterns and the binary pattern of
mineral deposit points ( N{B1 ∩ B2 ∩ D} ) equals the ratio of the product of the
number of pixels representing spatial coincidence between B1 and
D (N{ B1 ∩ D}) and the number of pixels representing spatial coincidence
between B2 and D ( N{ B2 ∩ D} ) to the number of pixels or area of D ( N{D}).
Conditional independence is determined only if N {B1 ∩ B2 ∩ D} is not

equal to zero, (if there is spatial coincidence between B1 and B2 and if their
overlap coincides with D). If there is no spatial coincidence between B1 and B2,
then N {B1∩ B2 ∩ D} is equal to zero and there is no conditional independence
between B1 and B2.
When there is spatial coincidence between two binary predictor patterns, B1

and B2, their conditional independence of each other with respect to a binary
pattern of mineral deposit points is determined not only from the spatial
relationships depicted in the enlarged circle in Figure 3.7, but also in other parts of
the area where both B1 and B2 are absent or where either B1 or B2 are absent.
Thus, two binary predictor patterns are conditionally independent of each

other with respect to a binary pattern of mineral deposit points if the following
spatial relationships are also satisfied.
53
Figure 3.7: Venn diagram to illustrate the

concept of conditional
independence.
N B1  D N B2  D
N{B1  B2  D}= (3-15)
N D
N{B1  B2  D}=

N B1  D  N B2  D
(3-16)
N D
N{B1  B2  D}=

N B1  D N B2  D  (3-17)
N D
N{B1  B2  D}=

N B1  D N B2  D  (3-18)
N D
B1 and B2 Binary predictor patterns present
B1 and B2 Binary predictor patterns absent
The spatial relationships between the binary predictor patterns and the
binary pattern of a mineral deposit lead to a contingency table calculation for
testing the conditional independence of two binary predictor patterns (Table 3.6).
54
The four cells in the table correspond to the four overlap conditions between B1
and B2 where mineral deposits are present.
Table 3. 6: Contingency table for pairwise testing conditional independence based

on pixels that contain a mineral deposit.
B1Present B1Absent Total
B2 Present N {B1  B2  D} N{B1  B2  D} N {B2  D}
B2 Absent N{B1  B2  D} N{B1  B2  D} N {B2  D}
The left hand side of the equation is the observed number of occurrences in
the overlap zone of B1 and B2. The right-hand side is the predicted number of
deposits in this overlap zone. A contingency calculation table is used to test the
conditional independence of the two binary maps. The chi-square test is then
calculated to test the hypothesis by the expression
4
(observedi -predictedi ) 2
χ 2 = (3-19)
i=1 predicted i
An overall test of conditional independence can also be applied after

combining the binary predictor patterns. The predicted number of occurrences
N{D}pred can be calculated as the sum of the products of the number of pixels,
N{A}, and their posterior probabilities, P, for all pixels on the map, thus
m
N{D}pred =  Pk *N{A}k , (3-20)
k=1
Where N{A} is the number of pixels with posterior probabilities P and k =

1, 2, …, m pixels in the map. In practice, the number of predicted mineral
deposits is usually larger than the observed number of mineral deposits. If the
predicted number of occurrences is larger by 15% than the number of observed
mineral deposits, then the assumption of conditional independence is seriously
violated. Problematic maps should then be removed from the analysis or
combined prior to calculation of the weights.
55
3.6. Application of Weight of Evidence in Mineral Potential Mapping
The steps in applying weight of evidence (Bonham-Carter, 1994) are follow:
1. Choose a series of maps that are likely to be evidence for predicting mineral
deposit. A particular deposit type and associated conceptual deposit model
should be used to guide the selection process (section 3.7)
2. For each map, assumed to be multi-class (categorical or higher measurement

levels), determine the optimum reclassification scheme to binary form,
maximizing the spatial association between the map and the deposit point.
Weight of evidence calculation can be used for this optimization process.
3. Check for pairwise Conditional Independence, CI between the binary maps.

Delete problem maps, or combine binary maps to reduce effect of CI.
4. Combine the binary maps with weights determined in step 2 with Equation
3.14. This can be carried out either with a modeling language internal to
GIS, or with external modeling program. Make new maps showing posterior
probability. Optionally, calculate the effects of uncertainty in the weight,
and uncertainty due to missing information, and produce an uncertainty
map.
3.7. Extraction of Evidences Maps
3.7.1. Lithology Map
Geological maps are usful to efforts into prospective ground during

exploration. One of the most important factors to consider during the analysis of
lithological evidences is the timing of mineralization in relation with the rock age.
Since different lithology units have different potential area, they are very
important in providing data for mineral potential map. The lithology map will
show the host rock and heat source that tend to be ore deposit. Selection the host
rock and heat source, it depends on rock types associated geological of
mineralization in study area. The host rock and heat source are extracted by using
selection tool in GIS (select by attribute table).
56
3.7.2. Geological Structure Map
A qualitative knowledge of spatial association s between known mineral

deposits and curvi-linear geological features in a particular area is more important.
Mapping of geological structure, or lineament features in image that represent the
surface expression of faults, joints, folds, lithologic contacts, or other geologic
discontinuities.
According to these information is an essential component for the mineral
exploration. The lineaments are linear or curvilinear discontinuities in direct
connection with the faults and composite fractures; and are associated with
geomorphological features and various tectonic structures such as faults, fractures,
folds axes and lithological contacts that may lead to the location of individual ore
bodies, which is one of important condition to controls on ore deposits (Abdullah
et al., 2010). Lineaments can be generated from hillshade and hillshades are
created in GIS using 3D analysis tool (surface >> hillshade) and then digitized the
lineament by manual technique. After that, Rose diagram software are use to
know the orientation of lineament. To validity the lineaments, they compare the
lineament with lithological map and previous research.
3.7.3. Hydrothermal Alteration Map
Many ore bodies are deposited by hot aqueous fluids, call hydrothermal
solution, that invade the host rock or country rock. During formation of the ore
mineral, these solutions also interact chemically with the country rock to alter the
mineral composition for considerable distances beyond the site of ore deposition.
The hydrothermal altered country rocks contain distinctive assemblages of
secondary, or alteration, minerals that replace the original rock constituents.
Alteration minerals commonly occur in distinct sequences, or zone of hydro the
mal alteration, relative to the ore body (Sabins, 1987).
Ore deposits are accompanied by hydrothermal alteration of the adjacent

country rocks. Not all alteration is associated with ore bodies, and not all ore
bodies are accompanied by alteration, but the presence of altered rocks is a
valuable indicator of possible deposits. Prospectors have long been aware of the
association between hydrothermally altered rocks and ore deposits. Many mines
57
were discovered by recognizing outcrops of altered rocks, followed by assays of

rock samples. Clay and limonitic alterations are very important for gold deposit
like epithermal, porphyry, skarn deposit. Clay alteration is characterized by illite,
smectite, alunite, kaolinite and dickite. Limonite usually forms from the hydration
of hematite and magnetite, from the oxidation and hydration of iron rich sulfide
minerals, and chemical weathering of other iron rich minerals such
as olivine, pyroxene, amphibole, and biotite.
The spectral bands of Landsat TM are well-suited for recognizing

assemblages of alteration minerals. Landsat 5 TM is use to gentrate the clay and
limonitic ateration. They use software defoliant method to extract the clay and
limoniti alteration (section 3.3) by using devided tool (spatial analysis tools >>
math >> devided) and principal components (spatial analysis tools >> multivariate
>> principal components) tool in GIS.
3.8. Hypothesis
1. The mineral deposit occurrences are strongly associated with NE-SW and
NW-SE trending fault/fracture, veins types and these fracture system may
have acted as hydrothermal fluid pathways, hosted by Tertiary extrusive
volcanic rock and shallow intrusive rocks as heat source.
2. The distributions of gold deposits in West Java are mostly located in
Honje Ignous Complex, Bayah Dome and Southern Mountain zone.
58
CHAPTER IV
RESEARCH METHODOLOGY
4.1. Variable Data and Materials
 Geological maps at scale 1:100,000

 Mineral occurrence data
 Landsat 5 Thematic Mapper (TM) data.
 Advanced Spaceborne Thermal Emission and Reflection Radiometer Global
Digital Elevation Model (ASTER GDEM) data.
 Scanner.
 Printer with printable A3 and A4 size.
4.2. Software
The software used in this research is as follows.

 GIS software (ArcView9.3 ) from the Environment Systems Research
Institute Inc (ESRI) was used for: 1) Digitising and editing of geological
map 2) Extraction and analysis of spatial features 3) Modelling (data
integration)
 RockWorks 16 software: to evaluate the orientation of lineamentsis in
rose diagrams.
4.3. Research Methodology
4.3.1. Research Preparation
The first steps, red carefully about the basic theory and previous study
related to the gold deposit in West Java and also collect datas to make gold
potential map like geological map, mineral deposit, ASTER GDEM and Landsat
Thematic Mapper (TM).
For secondary data like geological maps and mineral deposit database are
received from Dr. Lucas Donny Setijadji, Geological Engineering Department,
Faculty of Engineering, Gadjah Mada University, Digital geological map is
derived from digital version of many 1: 100,000 scale geological map by
Geological Agency of Indonesia. Landsat 5 TM data consist of Path 123, Row 64
59
(1997), Path 122, Row 64 (1997), Path 121, Row 64 (2000), Path 123, Row 65,
Path 122, Row 65, Path 121, Row 65 (1994), all were downloaded from United
States Geological Survey (USGS) websites (http://earthexplorer.usgs.gov) and
ASTER GDEM version 2 data were download from ASTER GDEM websites,
which was released October 17, 2011 (http://gdem.ersdac.jspacesystems.or.jp).
4.3.2. Spatial Data Processing and Analysis
A powerful program for mapping, ArcGIS plays a major role in creating

data for using in analysis procedure. All collected data like lithology, mineral
database, ASTER GDEM, and Landsat 5 TM map are input in to the program and
prepare to become news database for analyzing. The preliminary spatial data
manipulation and analysis techniques to create the evidence maps for predicted
gold mineral potential map in this research are map reclassification/generalization
map overlay/combination, data management tools, and spatial analyst tools
 Geological Map
Mineral potential map is greatly controlled by the lithology units. Since

different lithology units have different potential area, they are very important in
providing data for mineral potential map. The lithology map will show the host
rock and heat source that tend to be ore deposit. Based on the characteristics of
mineral deposits in West Java, all of Tertiary to Pleistocene is selected for host
rocks and Tertiary intrusive rocks are considered as heat sources (See detail in
chapter 5). In GIS, using selected tool to extract host rock and heat sources map
and after that find the favorable distances to heat source and host rock using
buffering tool in GIS. Then calculate the evidence map of host rock and heat
sources with known mineral deposit by weight of evidence method.
 Geological Structure Map
Geological structures (lineaments) are created from hillshade image, which

were derived from ASTER GDEM. In order to identify lineaments features, two
hillshade images were generated, which are differences sun angle (azimuth) 45°
and 315° with Altitude of 45° for good contrast of result. Then digitize the
lineament (structure) from hillshade map. The orientation of the lineament is
60
analyzed by rose diagram. Verify the structure by previous research and

geological structure map of West Java (See detail in chapter 5). Next, find
favorable distances to structure with known mineral deposit using buffering tool
in GIS. After that calculated evidence map of structure with known mineral
deposit by weight of evidence method (formula in Chapter 3).
 Hydrothermal Alteration Map
Two hydrothermal alteration maps as clay and limonitic alteration are

derived from Landsat 5 TM data. The first step in this methodology is to use the
software defoliant technique to enhance the spectral response of each alteration
mineral (clay and iron oxide alteration zone). Clay alteration is generated using
band ratio 4:3 and 5:7 and limonitic alteration by using band ratio 3:1 and 4:3.
Band ratio 4/3 refers to vegetation index, band ratio 5/7 refers to clay mineral
index and band ratio 3/1 refers to limonitic index. Next, Direct Principle
Component (DCP) was analyzed by using multivariate tool. The second step is
used reclassify tool for separate zones of alteration and non-alteration, used
mosaic technic to combination. The third step, find the Normalized Difference
Vegetation Index (NDVI) map to compare with the result of alteration zone. A
fourth step was to incorporate DEM image or the previous research for improving
the results of the classification (see detail in Chapter 5). After get the alteration
map, using the buffering tool to find the favorable hydrothermal alteration
assemblage and then calculated evidence map of alteration with known mineral
deposit by weight of evidence method (formula in Chapter 3).
4.3.3. Predictive Modeling
Conditional Independence (CI) is assumed to exist when combining two or

more maps with the Bayesian models. Check for pairwise Conditional
Independence, CI between the binary maps. Deleting problem maps, or combining
binary maps to reduce effect of CI. Conditional independence between the input
binary predictor patterns was determined through a pairwise test (between two
binary predictor patterns) and/or through an overall test (among all input binary
predictor patterns). The conditional independence between or among the binary
predictor patterns was tested prior to (by pairwise test) and/or after combining
61
them (by an overall test) with Equation (3.20). If the total predicted number of
deposit is more by 15% than number of observed mineral deposits, then the
assumption of conditional independence is seriously violated. After combining
evidence maps to generated predictive mineral potential map. Finally, a small set
of deposit and mineral occurrence was used to validate the predictive map.
4.3.4. Model Validation
Mineral deposits are split into 2 subsets. First subset of 80 (75% called as
traning data) of the total 107 known mineral deposits occurrences is used to
generate the probabilistic models. The other subset of 27 (25%) of 107 known
mineral deposits occurrences is used to validate the probabilistic models (called as
validation data). The mineral deposits that comprise the validation subset were
chosen randomly. A pixel size of 30 x 30 m is used in the probabilistic mapping
of Au mineralization potential.
For the probabilistic map to be useful, they must be validated for sureness of
the prediction (Carranza, 2003). So, for validated of result in this research were
used four ways like described below:
1. Perform an overall test of conditional independence, according to equation
3.20.
2. Determine the percentage area covered the favourable target zones (those
areas with posterior probabilities greater than prior probability), acceptable
if percentage areas of favourable zone may set to 25% and by determine the
number and percentage of known occurrences predicted correctly in model.
3. Determine the number and percentage of known occurrences predicted
correctly in model that is number of occurrences fall within the favourable
target zones, is at least 60 % of known mineral occurrences are accepted.
4. Determine the number and percentage of validation model that is number of
validation model fall within the favourable target zones, is at least 70 % of
validation model are accepted.
62
Study Background
Research Preparation
Literature Review: Secondary Data: Primary Data
 Basic study  Digital data of geological
 Previous study  Mineral deposits database Landsat 5 ASTER
TM data GDEM

Hypothesis
GIS ANALYSIS
Spatial Data
Input
Mineral Deposit Lithology Map Alteration Lineament

(Mineral deposit (Digitised lithologic map) Map Map
database)
Reclassification and quantify spatial association with known mineral

Spatial Data
deposits
Extract
Extract host rock Clay and NE-SW

and heat source Limonitic NW-SE
Spatial Processing
Create buffers
and Analysis
Host rock and Heat Clay and NE-SW

source binary map Limonitic NW-SE
binary map binary map
Predictive Modeling
Combine Spatial Binary Maps:

WofE Method and Test of CI
Posterior Probability Map:

Overall Test
No
Validation
Validation
Model
Yes
Gold Potential
Map
Finish
REPORTING
Figure 4.1: Flow Chat of Methodology in this research
63
Table 4.1: Research Schedule
2014 2015
Research
No Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul
Phase
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
I. Research
Preparation
1. Literature
review
2. Secondary
data collection
Software
III.
analysis
1. Building
Spatial
Database
2. Spatial
processing and
Analysis
3.Predictive
Modeling
IV. Field Work
1. Varification
Research
V.
Writing
1. Proposal
2. Thesis draft
3. Final thesis
64
CHAPTER V:
SPATIAL DATASETS AND EVIDENCE MAPS
5.1. Mineral Deposit Database
As mentioned in chapter I and II, styles of gold mineralization in West Java

are dominated by epithermal Au-Ag, porphyry Cu-Au, hydrothermal and
Polymetallic (Bushki et al., 2012; Yuningish et al, 2012; Lubis et al 2012;
Setijadji and Maryono, 2012; Syafizal et al, 2011; Warmada et al., 2007; Rosana
et al, 2006; Milesi et al, 1994; Marcoux and Milesi, 1994)
The study area contains 134 metallic deposits and industrial mineral
occurrences of various sizes, but not all of metallic deposits are related to Au
occurrences. There are 107 metallic deposits occurrences containing Au (based on
attributed table of metallic deposits layer) which are used in this study (Figure
5.1).
Figure 5.1 : Distribution of Au occurrences in study area (Physiography frsom Van

Bemmelen, 1949 and Mineral deposits are modified from Setijadji,
Unpublished data and from previous research Nurcaho et al., 2012;
Syafrizal et al., 2011; Subandrio et al., 2010; Suparka et al., 2007; Utoyo,
2007 and Ismayanto et al., 2005)
65
5.2. Lithology
Lithology map is created from geological map, which were collect as

secondary data, that digital geological map is derived from digital version of
many 1:100,000 scale geological map (Figure 5.2).
Figure 5.2: Lithology map of the study area (lithology map is derived from digital
version of many geology maps sheets, scale 1: 100,000 by Geology Agency
of Indonesia, Unpublished data.)
Lithological and age data for each unit is the most important of geological
map, where the lithological unit and age of lithology is a potential for epithermal
gold deposit, Au-Ag low epithermal, porphyry Cu-Au and polymetallic (Section
2.2 ). According to the general characteristics of mineral deposits elsewhere and
study area, two criteria of lithology were considered as host rocks and heat source.
Metallic occurrences in Eastern Sunda Arc are numerous and dominated by

Cu-Au, magmatic hydrothermal systems associated with subduction-related
volcanic centers (Setijadji and Maryono, 2012). However, important mineral
districts are located within long-lived (overlapping) volcanic centers, such as in
66
Bayah dome region (Lower Miocene to Pliocene volcanic events). Gunung

Pongkor and some other Au-Ag epithermal deposits within the Bayah Dome are
dated to be young (Pliocene to Pleistocene). The west Java sector is dominated by
high-grade Au-Ag low sulfidation epithermal system of Upper Miocene to
Pliocene age, such as Gunung Pongkor, Cikidang, and Cibaliung (Setijadji and
Maryono, 2012).
The lithology units along the Southern Mountains Range comprise Tertiary
extrusive volcanic rocks and shallow intrusive rocks, as well as sedimentary rocks
(limestone and siliciclastic). Many epithermal-style alterations and mineralization
are associated with this volcaniclastic and intrusive rocks, including several gold
deposits along the Southern Mountain Range (e.g. in Cineam area, with alteration-
mineralization age of 13.5 – 8 Ma), and few prospect area for possible
mineralization of porphyry systems (e.g.in Cihurip area-Garut) (Basuki et al.,
2012). Cijulang area is located in the Southern Mountains of the West Java which
consists of Eocene-Miocene sediments and volcanic rocks such as Jampang
Formation (Old Andesite Formation, van Bemmelen, 1949). Lithologic units
exposing in the Cijulang prospect are of mainly calc-alkaline volcanic and
volcaniclastic rocks and comprise andesite lava, lapilli tuff, andesite and andesite
breccia (Tun et al., 2013). The host rocks of Cihar; Cibaliung; Bunikasih; Tutugan
area are andesite lava / flows and breccias, porphyritic with some mafic mineral
phenocrysts, Late Miocene-Pliocene in age (based on regional map by Alzwar et
al., 1992). Salopa area is occupied by Oligo-Miocene Jampang Fm. which is
composed of andesitic volcaniclastic breccias, tuffaceous sandstones and shales,
with dacitic dome or sub-volcanic intrusive rock (Basuki, 2012).
So, based on the characteristics of mineral deposits in study areas, all of

Tertiary to Pleistocene lithology unit is selected for host rocks (Figure 5.3) and
Tertiary intrusive rocks are considered as heat sources (Figure 5.4).
67
Figure 5.3: Tertiary- Pleistocene host rock (Extract from lithology map)
Figure 5.4: Tertiary intrusive rock (heat source, Extract from lithology map)
68
5.3. Geological Structure
In order to identify lineaments from the DEM, four shaded relief images
were generated. The first step is the production of four separate shaded relief
images with light sources coming from four different directions. The shaded
relief image created had a solar azimuth (sun angle) of 0°, 45o, 90 o, 315 o and
a solar elevation of 45° (Figure 5.5). The second step is to combine four
shaded relief image to produce one shaded relief image, for this purpose, the
combinations of the four shaded relief maps are computed by using GIS
composite technique, where four shaded relief images are overlaid to produce
one image with multi-illumination directions (0°, 45°, 90°, and 315°, Figure
5.6). After that the lineament is digitized using manual extraction techniques.
Azimuth 0o Azimuth 45o
Azimuth 90o Azimuth 315o
Figure 5.5: Four shaded relief images derived from DEM with different illumination
directions ( sun azimuth) of ,0°, 45°, 90°, and 315°, with a solar
elevation of 45°.
69
Figure 5.6: The Combined shaded relief DEM image created by combining serveral
shaded relief images with different illumination.
Figure 5.7: Lineaments map, interpreted from shaded relief DEM image
70
Figure 5.8: Rose diagram of lineament
Table 5. 1: Result of Rose Diagram

Statistical Summary
Calculation Method: Length
Class Interval: 10.0 Degrees
Min. Length Filtering: Deactivated
Max. Length Filtering: Deactivated
Azimuth Filtering: Deactivated
Data Type: Bidirectional
Population: 2
Total Length of All Lineaments: 16,635,627.55
Maximum Bin Populations: 275
Mean Bin Populations: 126.44
Standard Deviation of Bin Populations: 57.3
Maximum Bin Populations (%): 6.04
Mean Bin Populations (%): 2.78
Standard Deviation of Bin Populations (%): 1.26
Maximum Bin Length: 764,722.76
Mean Bin Length: 462,100.77
Standard Deviation of Bin Length: 141,059.12
Maximum Bin Length (%): 4.6
Mean Bin Length (%): 2.78
Standard Deviation of Bin Lengths (%): 0.85
Vector Mean: 46.5 Degrees
226.49 Degrees
Confidence Interval: 14.9 Degrees
(95 Percent)
R-mag: 0.11
71
The lineaments are created along valley where shown in hillshade image,
considered the geological structure, lineament feature of West Java, and also the
lithology of Java island (Scale 1:100,000). The result of lineaments (Figure 5.7)
that indicate some of lineaments are not shown in public geological map.
For evaluation and interpretation the orientation of the lineaments is used
RockWorks 15 software. The orientation of the lineaments is analyzed by rose
diagrams. These diagrams display length of lineaments for regular intervals. The
diagram is prepared by counting each lineaments of its length.
The result of rose diagram (Figure 5.8) indicates that the most lineaments
directions are found in two main directions as N15-70E and N20-65W,
respectively. Other minor directions are observed in N-S and E-W. Some previous
researches about geological structure that may control mineralization in research
areas are explained as follows:
These main directions correspond to the trend of the main structural features
of Java Island (Setijadji et al., 2006). A study of Landsat images on West Java
reveals that there are four major lineaments: N45oE, N10oW, N30oW and N55oW.
A high density of lineaments is found in the vicinity of Bogor and southwest
Cirebon, with a concentration of shallow and intermediate earthquake around the
Bogor lineaments (Bronto, 1989).
There are some comparation between the lineaments from shaded relief and
geological map of Arinem, Cupunagara, and Cibaliung of West Java. The
similarity between the lineaments from shaded relief and the structure from the
geological map show in the figures 5.9, 5.10, and 5.11. The lineaments (structure)
are mostly dominated by NE-SW and NW-SE in those areas. Therefore, the result
of lineaments map is accuracy when consider with the geological map, and
previous research. According to mineralization association with the structure
(lineaments), lineaments were separated into two directions NE-SW and NW-SE.
72
a b
Figure 5.9: a) Lineaments derived from shaded relief of Arinem districtDEM, b)

Regional geological map of the Arinem district, West Java (modified after
Alzwar et al., 1992)
a b
Figure 5.10: a) Lineaments in Cibaliung derived from shaded relief DEM, b)

Simplified geological map of Cibaliung area (modified after Angles et al.,
2002)
73
a b
Figure 5.11: a) Lineaments derived from shaded relief DEM, b) Geological map of
Cupunagara, Subang District, West Java (Aerial photo interpretation in
Utoyo, 2007)
5.4. Hydrothermal Alteration
Two hydrothermal alteration maps as clay and limonitic alteration are

derived from Landsat 5 TM. The image processing techniques were used software
defoliant methodology based on Carranza (2002) for analysis.
5.4.1. Clay Alteration
Clay alteration zones are made by using band ratio 4:3 and 5:7, vegetation
index (band ratio of 4/3), clay minerals index (band ratio of 5/7). The result of
DPC analysis on band ratio image 4:3 and 5:7 to map clay alteration is given in
Table 5.1. The DPC2 depicts the strong similarity between the input images and
represents zones in which the spectral responses of vegetation and clay alteration
cannot be differentiated. The negative sign of loading on ratio 5:7 suggests that
DPC1 should be negated to represent clay alteration zones.
74
Table 5.1: Directed Principle Components analysis of band ratio Landsat TM

images for clay alteration by the software defoliant technique.
Principle Eigenvectors Path/Row of
Components Band 4: Band 3 Band 5: Band 7 Landsat 5 TM
DCP1 0.88184 -0.47154
123/064
DCP2 0.47154 0.88184
DCP1 0.7667 -0.642
123/065
DCP2 0.642 0.7667
DCP1 0.7743 -0.63282
122/064
DCP2 0.63282 0.7743
DCP1 0.91572 -0.40183
122/065
DCP2 0.40183 0.91572
DCP1 0.83114 -0.55606
121/064
DCP2 0.55606 0.83114
DCP1 0.90691 -0.42133
121/065
DCP2 0.42133 0.90691
5.4.2. Limonitic Alteration
The DPC analysis on band ratio images 3:1 and 4:3 to map limonitic
alteration is described in Table 5.2. The DPC2 represent zones where the spectral
responses due to vegetation and limonitic cannot be differentiated. The DPC1
should thus map zones containing limonitic. The negative sign of loading on band
ratio 3:1 of sheet 1, 2, 3, 4, and 6 suggests that DCP1 image must be negated to
represent zones of limonitic alteration. All sheet of Landsat mapped limonitic
alteration zones in DCP1 but for sheet 5 mapped limonitic alteration zone in
DCP2 because DCP1 of sheet 5 cannot be differentiated.
75
Table 5.2: Directed Principle Components analysis of band ratio Landsat TM

images for limonitic alteration by the software defoliant technique.
Principle Eigenvectors Path/Row of
Components Band 3: Band 1 Band 4: Band 3 Landsat 5 TM
DCP1 -0.00232 1.00
123/064 (1)
DCP2 1.00 0.00232
DCP1 -0.00362 1.00
123/065 (2)
DCP2 1.00 0.00362
DCP1 -0.00012 1.00
122/064 (3)
DCP2 1.00 0.00012
DCP1 -0.00787 1.00
122/065 (4)
DCP2 1.00 0.00787
DCP1 0.00145 1.00
121/064 (5)
DCP2 1.00 -0.00145
DCP1 -0.0424 1.00
121/065 (6)
DCP2 1.00 0.0424
Figure 5.12 : Clay alteration map using the software defoliant technique
76
Figure 5.13: Limonitic alteration map using the software defoliant technique
5.4.3. Accuracy Assessment
The final map of clay and limonitic alteration are used reclassify tool for
separate zones of alteration and non-alteration, used mosaic tool to combination.
Compare with hydrothermal alteration references, which are previous research
(Harijoko, 2007; Suparka et al., 2007; Utoyo, 2007; ANTAM, 1994; Basuki et al.,
1994) and Normalized Difference Vegetation Index (NDVI) map, which derived
from Landsat 5 TM data.
Normalized Difference Vegetation Index (NDVI), one of the vegetation

indices, measures the amount of green vegetation. The spectral reflectance
difference between Near Infrared (NIR) and red is used to calculate NDVI. The
formula can be expressed as (Jensen, 2000);
NDVI = (band 4 (NIR) – band 3 (Red)) / (band 4 (NIR) – band 3(Red))

The NDVI values range from -1 to +1, where higher values are for green
vegetation and low values for other common surface materials. Bare soil is
represented with NDVI values which are closest to 0 and water bodies, could are
77
represented with negative NDVI values (Karaburun, 2010). Based on this

information, the result indicates that bare soil NDVI values range from 0.1 to 0.2,
sparse vegetation NDVI values fall between 0.2 to 0.6; while dense vegetation
NDVI values ranges from 0.6 to 1. Similarly, NDVI values less than 0.1 represent
water body, clouds or snow and areas without vegetation cover (Al-doski et al.,
2013).
Figure 5.14: The Normalized Difference Vegetation Index (NDVI) map, derived from
Landsat 5 TM data
The altered on rocks of Cupunagara consist of 5 zones, i.e: Argilite Silica,
Argilite, Propylite Pyrite, Propylite and Cloritize Andesite. The results of XRD
analysis of the altered rocks show various minerals i.e: quartz, calcite, albite,
haloysite, feldspar, pyrite, kaolinite, crystobalite and mica, which were formed in
the low temperature and pH. Mineralization is known by the present of pyrite,
which is formed in the epithermal condition. Consist of disseminated of pyrite and
calcopyrite, spotted, veinlets, vuggy and replacement of primary mineral.
Alteration zone in Cihurip have contains of quartz – biotite – magnetite ±
actinolite (potassic); quartz – chlorite – epidote ± actinolite± tremolite
(propylitic); quartz – sericite – clay (phyllic); quartz – chlorite – calcite
78
(subpropylitic) and quartz – clay (argillic) zones. Mineralizations generally have

occurred within argillic and propyllitic zone, another less within phyllic and
porphyry zone. (Figure5.14).
b
c
Figure 5.15: a) Clay alteration from Landsat 5 TM, b) Alteration zone of Cupunagara,
Subang, West Java (Utoyo, 2007), c) Alteration zone of Cihurip, Garut
Regency (2007)
79
For clay alteration zones (Figure 5.8) are potential in Bayah Dome, Ciemas,
Cibaliung, few alteration zones in Cianjur, Tasikmalaya districts, some alteration
zones show in northwest and northest of study areas, which are not metallic
district (Figure 5.12). Limonitic alteration (Figure 5.13) zones are also strong in
Western and Northwest of Bayah Dome and Ciemas districts, have some
alteration zones in Tasikmalaya and Cianjur districts.
The classified NDVI map indicates that the percentage of sparse vegetation
is higher than bare soil and dense vegetation. In term of NDVI, most of
hydrothermal alteration zones pertain to bare soil-sparse vegetation (Figure 5.14).
Several larger path of hydrothermal alteration are mostly mapped in southern part,
northwest, and northeast of West Java, where many mineral deposits occur
(Figure 5.1) is indicated to contain hydrothermal alteration.
Some areas of dense vegetation (Figure 5.14) are mapped as non-alteration
zones in the alteration maps and some of dense vegetation mapped as alteration
because of vegetation density in these areas are probably particularly low so that
the spectral response of ground surface are enhanced. There are also areas of spare
vegetation mapped as non-alteration zones, these areas probably where the
material sought is not present in sufficient abundance (Carranza, 2002).
The result of hydrothermal alteration maps are accurate when consider with
previous research. It shows good spatial coincidence between the hydrothermal
alteration units classified from the software defoliant technique and the
hydrothermal alteration from previous research. The agreements between the two
hydrothermal alteration maps are moderate. So based on references map previous
research, suggested that classified hydrothermal alteration map of the study area
can be used for predicting mineral potential mapping.
80
CHAPTER VI:
ANALYSIS AND RESULTS
6.1. Evidence Maps
Lithology, geological structure and hydrothermal alteration maps are

obtained from secondary and primary data as discussed in previous chapters. All
maps are presented using UTM system, zone 48S with Global Datum WGS 1984.
As results, the geological evidence features that are used as predictors of

gold potential in the study area consist of 6 evidence maps as follows: NE-SW
trending lineaments, NW-SE trending lineaments, host rocks, heat sources, clay
alteration and limonitic alteration.
a b
c d
e f
Figure 6.1: Evidences maps: a) Host rock, b) Heat sources, c) NE-SE trending
lineaments, d) NW-SE trending lineaments, e) Clay alteration, and
f) Limonitic alteration
81
6.2. Analysis
The weight of evidence method is statistical method and requires significant

number of mineral occurrences for modeling. Firstly, mineral deposits are split
into 2 subsets. First subset of 80 (75%) of the total 107 known mineral deposit
occurrences are used to generate the probabilistic models (called as training data).
The other subset of 27 (25% called as validation data) of 107 known mineral
deposit occurrences are used to validate the probabilistic models (called as
validation data). The mineral deposits that comprise the validation subset were
chosen randomly. A pixel size of 30 x 30 m was used in the probabilistic mapping
of Au mineralization potential. The extracted geologic features whose spatial
associations with the mineral occurrences are quantified with favourable rocks,
heat sources, lineaments and alteration zones.
6.2.1. Spatial Association of Au Deposit with Each Evidence Maps
The binary maps are created by using a statistical method (Bayesian

probability Model). The process of converting multiclass map to binary form can
either carry out subjectively, using correlation between the mineral occurrences
and evidence features. The weight of evidence map is calculated by formulation in
section 3.7 (Equation 3-7 and 3-8).
For each test domain the maximum contrast often gives the best measures of
spatial correlation with the mineral occurrence point in the case of large area and a
large number of mineral occurrences. In the case of small area and small number
of mineral occurrences the uncertainty of the weights could be large; therefore, C
can be meaningless. The studentized value of C, calculated as ratio of C to its
standard deviation, C/s(C), serve as an approximate test of the statistical
significance of the spatial correlation between the mineral occurrence points and a
test domain. The Studentized value of C is used to define the optimum cutoff.
Ideally it is nice to see a Studentized value larger than 1.5 or even 2. To define the
cutoff distance in this research are considered Studentized C, Contrast and mineral
occurrences.
82
6.2.1.1. Calculating Weights of Evidence of NE-SW Trending Lineaments
The resulting NE-SW lineaments is buffered at 250 to 6000m (Figure 6.2),

and crossed with the raster mineral occurrence point map to estimate weights of
evidence (Table 6.1). The Studentized C (5.690) is highest at 2000m and contrast
is highest at 5000 m. The optimum buffer is defined at 2000 m. The resulting
buffered lineament domain cover 18257 km2 (39.86%) of area, 59 (73.75%) out of
80 mineral occurrences are present in this distance (Figure 6.3). The weight of
evidence analysis reveals a strong correlation between the lineament domain and
mineral occurrences. The weights used in the final model are W+ = 0.671 (If in a
favorable lineament) and W- = - 0.83 (If not in a favorable lineament domain).
Figure 6. 2: Buffer distance s to the NE-SW lineaments

83
Table 6.1: Variation of weights and contrasts for cumulative distances from NE-
SW trending lineaments with respect to gold deposit
%
Distance N B  D % of
N{B} W+ s(W+) W- s(W-) C s(C) C/s(C) Occur
Buffer (m) Areas
rence
250 2505 12 1.012 0.289 -0.106 0.121 1.118 0.314 3.564 15.00 5.469
500 5202 17 0.628 0.243 -0.119 0.126 0.747 0.274 2.728 21.25 11.357
1000 10280 32 0.579 0.177 -0.257 0.144 0.836 0.228 3.660 40.00 22.444
1500 14614 50 0.674 0.142 -0.597 0.183 1.271 0.231 5.500 62.50 31.906
2000 18257 59 0.617 0.130 -0.830 0.218 1.447 0.254 5.690 73.75 39.860
2500 21293 63 0.528 0.126 -0.925 0.243 1.453 0.273 5.313 78.75 46.488
3000 23802 64 0.432 0.125 -0.877 0.250 1.310 0.280 4.683 80.00 51.966
3500 25938 71 0.450 0.119 -1.351 0.333 1.801 0.354 5.088 88.75 56.629
4000 27766 74 0.423 0.116 -1.660 0.408 2.083 0.425 4.907 92.50 60.620
4500 29317 77 0.409 0.114 -2.263 0.577 2.672 0.589 4.540 96.25 64.007
5000 30620 79 0.378 0.113 -2.586 0.707 2.965 0.716 4.139 97.50 66.852
>6000 45803 80 0.000 0.000 0.000 100 100
6 120
5.69
5 100
Contrast, Studentized C
Percent of occurrences
4 73.75% 80
2.965
3 60
2 40
Percent of Au occurrences
1 Contrast, C 20
Studentised C
0 0
250 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Buffer distance, m
Figure 6.3: Graph showing the studentized contrast, C and percent of Au occurrences
related with buffer at distance for NE-SW trending lineaments.
84
Figure 6.4: Binary predictor pattern of NE-SW trending lineamens with 2000 m buffer
zones.
6.2.1.2. Calculating Weights of Evidence of NW-SE Trending Lineaments
The NW-SE lineaments (Figure 6.1c) are created buffer distances from 500
to 5500 m. The variation of contrast for cumulative distances is given in Table 6.2
and Figure 6.5.
Most of buffer distance is more statically significant (Studentized value

greater then 2 (Bonham-Carter, 1994; Carranza, 2002)). The Studentized C
(5.822) of distance 1000 m is highest. However, at distance 1000 m wherein 36 of
the 80 gold occurrences (45%) are present within this distance and represents
8302 km2 (18.125%) of the total area (Figure 6.6). The optimum cutoff distance
from the NE-SW trending lineaments with respect to the Au occurrences is
therefore set at 1000 m. The weights used in the final model are W+ = 0.912 (If in
a favorable lineament) and W- = - 0.398 (If not in a favorable lineament domain).
85
Table 6.2: Variation of weights and contrasts for cumulative distances from NW-
SE trending lineaments with respect to gold deposit
Distance
N B  D
%
Buffer N{B} W+ s(W+) W- s(W-) C s(C) C/s(C) % of Areas
Occurrence
(m)
500 20 1.044 0.224 -0.196 0.129 1.239 0.259 4.789 25.000 8.833
4046
1000 36 0.912 0.167 -0.398 0.151 1.310 0.225 5.822 45.000 18.125
8302
1500 45 0.747 0.149 -0.517 0.169 1.264 0.226 5.604 56.250 26.690
12225
2000 50 0.605 0.142 -0.563 0.183 1.168 0.231 5.052 62.500 34.194
15662
2500 56 0.543 0.134 -0.682 0.204 1.225 0.244 5.017 70.000 40.720
18651
3000 61 0.498 0.128 -0.815 0.230 1.313 0.263 4.993 76.250 46.405
21255
3500 64 0.446 0.125 -0.892 0.250 1.338 0.280 4.785 80.000 51.250
23474
4000 70 0.458 0.120 -1.273 0.316 1.730 0.338 5.117 87.500 55.422
25385
4500 73 0.435 0.117 -1.543 0.378 1.979 0.396 4.999 91.250 59.103
27071
5000 73 0.382 0.117 -1.460 0.378 1.842 0.396 4.654 91.250 62.358
28562
5500 76 0.378 0.115 -1.942 0.500 2.320 0.513 4.521 95.000 65.181
29855
>6000 45803 80 0.000 0.000 0.000 100.000 100.000
7 100
90
6 5.822
80
5 70
60
4 Percent of Au
occurrences 50
45%
3 Contrast, C 40
2.32
2 30
20
1
10
0 0
500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Buffer distance, m
related with buffer at distance for NW-SE trending lineaments.
86
Figure 6.6: Binary predictor pattern of NW-SE trending lineamens with 1000 m buffer
7. zones.
6.2.1.3.Calculating Weights of Evidence of Host rocks
The variations in the contrast C for cumulative distances from host rock
(Figure 6.1a) are created buffer distances from 500 to 10000 m, shown in Table
6.3 and detail in Appendix C1.c).
The contrasts are mostly positive, which imply positive spatial association.
The positive spatial association is significant statistically within 1000 m, but
optimal within 0 m as indicated by the highest Studentised C (9.922). However, at
distance 0 m wherein 74 (92.500%) of the 80 gold occurrences are present within
this distance and represents 7130 km2 (15.567%) of the total area (Figure 6.7).
Favorable zones (Figure 6.8) for host rocks are created within itself. The weights
used in the final model are W+ = 1.791 and W- = -2.423.
87
Table 6.3: Variation of weights and contrasts for cumulative distances from host
rocks with respect to gold deposit.
Distance
% % of
Buffer N{B} N B  D W+ s(W+) W- s(W-) C s(C) C/s(C)
Occurrence Areas
(m)
0 7130 74 1.791 0.117 -2.423 0.408 4.213 0.425 9.922 92.500 15.567
500 11674 76 1.320 0.115 -2.703 0.500 4.024 0.513 7.842 95.000 25.487
1000 13529 77 1.185 0.114 -2.935 0.577 4.120 0.589 7.000 96.250 29.537
1500 15015 77 1.080 0.114 -2.888 0.577 3.968 0.589 6.742 96.250 32.782
2000 16290 77 0.999 0.114 -2.846 0.577 3.844 0.589 6.531 96.250 35.565
2500 17403 77 0.932 0.114 -2.807 0.577 3.739 0.589 6.353 96.250 37.995
3000 18403 77 0.876 0.114 -2.771 0.577 3.647 0.589 6.197 96.250 40.179
3500 19327 77 0.827 0.114 -2.737 0.577 3.564 0.589 6.055 96.250 42.196
4000 20200 77 0.783 0.114 -2.703 0.577 3.486 0.589 5.923 96.250 44.102
5000 21095 77 0.739 0.114 -2.668 0.577 3.407 0.589 5.788 96.250 46.056
10000 27853 79 0.486 0.113 -3.447 1.000 3.933 1.006 3.908 98.750 60.810
>10000 45803 80 0.000 0.000 0.000 0.408 4.213 0.425 9.922 100.000 100.000
12 100
9.922 99
10 98
97
8
96
95
6
92.5% 94
4.231 93
4
92
Percent of Au
2 occurrences 91
Contrast, C 90
0 89
Buffer distance, m
related with buffer at distance for host rocks.
88
Figure 6.8: Binary predictor pattern of host rocks.
6.2.1.4. Calculating Weights of Evidence of Heat source
The resulting heat sources domain map is buffered at distances of 500 m to

50000 m, and weights of evidence, contrast C, and Studentized value of C are
calculated (Table 6.4). The distribution of Au occurrences, relative to heat source
buffers, indicates that a only 11 (13.750 %) of occurrences fall within 500 m
(Figure 6.9), it means almost of occurrences are located far away from heat source
bodies. The maximum contrast value was defined at 30,000m, which cover 97.50
% of area, and 78 out of 80 occurrences are presented in this distance. However,
because at buffer distance 30,000 m from heat source quite far away and not
reasonable, 30,000 m could not be the optimum cutoff distance. The Studentized
C indicates that optimum cutoff distance is 4500 m meters. The optimum buffer,
yielding the maximum Studentized value of C, is defined at 4500m. The resulting
buffered igneous heat sources domain constitutes 17% of the area, and 39
(48.75%) out of 80 mineral occurrences are presented in this domain (Figure 6.9).
From these results, suggested that Au occurrences have strong spatial association
89
with areas within 4500 buffer from heat sources and consider as favorable zones
(Figure 6.10).
Table 6.4: Variation of weights and contrasts for cumulative distances from heat
sources with respect to gold deposit.
Distance
N B  D %
Buffer N{B} W+ s(W+) W- s(W-) C s( C) C/s(C) % of Areas
Occurrence
(m)
500 1249 11 1.625 0.303 -0.120 0.120 1.745 0.326 5.355 13.750 2.727
1000 2082 13 1.278 0.278 -0.131 0.122 1.409 0.304 4.638 16.250 4.546
1500 2914 19 1.322 0.230 -0.206 0.128 1.528 0.263 5.800 23.750 6.362
2000 3750 20 1.120 0.224 -0.203 0.129 1.322 0.259 5.111 25.000 8.187
2500 4582 23 1.059 0.209 -0.234 0.133 1.293 0.248 5.223 28.750 10.004
3000 5399 26 1.017 0.197 -0.268 0.136 1.285 0.239 5.375 32.500 11.787
3500 6203 32 1.086 0.177 -0.366 0.144 1.452 0.229 6.352 40.000 13.543
4000 7000 36 1.083 0.167 -0.433 0.151 1.516 0.225 6.734 45.000 15.283
4500 7792 39 1.056 0.161 -0.483 0.156 1.539 0.224 6.868 48.750 17.012
5000 8587 39 0.958 0.160 -0.461 0.156 1.420 0.224 6.339 48.750 18.748
6000 10153 45 0.934 0.149 -0.577 0.169 1.511 0.226 6.695 56.250 22.167
7000 11707 50 0.897 0.142 -0.687 0.183 1.583 0.231 6.848 62.500 25.559
8000 13280 51 0.790 0.140 -0.673 0.186 1.463 0.233 6.285 63.750 28.994
9000 14847 52 0.698 0.139 -0.659 0.189 1.356 0.235 5.781 65.000 32.415
10000 16396 57 0.690 0.133 -0.804 0.209 1.494 0.247 6.045 71.250 35.797
20000 30451 74 0.331 0.116 -1.499 0.408 1.829 0.425 4.309 92.500 66.483
30000 38602 78 0.146 0.113 -1.840 0.707 1.986 0.716 2.773 97.500 84.278
40000 42172 78 0.057 0.113 -1.155 0.707 1.213 0.716 1.693 97.500 92.073
50000 44139 78 0.012 0.113 -0.374 0.708 0.386 0.717 0.539 97.500 96.367
>50000 45803 80 100.000 100.000
8 120
7 6.868
100
6
80
4 Percent of Au occurences 60
48.75%
3 Contrast, C
1.986
40
Studentised C
2
20
1
0 0
30000
10000
20000
40000
50000
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
6000
7000
8000
9000
Buffer distance, m
related with buffer at distance for heat sources.
90
Figure 6.10: Binary predictor pattern of Heat sources with 1000 m buffer zones.
6.2.1.5.Calculating Weights of evidence of Clay alteration
The binary map cumulative distances 0 to 200 m from clay alteration zones
are created. The map is then crossed with the mineral occurrences map (of 80
occurences) to estimate weights of evidence with using formula above (Section
3.7). The results of analysis are shown in Table 5 and detail in C.1e of Appendix
C and Figure 6.11.
The optimum buffer that is yielding the maximum Studentized value of C

(6.326) is defined at 50 m. The resulting buffered hydrothermal alteration domain
cover 15.768% of the study area and 35 (43.75%) out of 80 mineral occurrences
are presented in this distance (Figure 6.11). The weights used in the final model
are W+ = 1.024 and W- = -0.404.
Based on the above relationships, clay alteration is created buffer at distance

within 50 m for favourable and outside of 50 m for non favourable (Figure 6.12).
91
Table 6.5: Variation of weights and contrasts for cumulative distances from clay
alteration with respect to gold deposit
Distance
N B  D
% % of
Buffer N{B} W+ s(W+) W- s(W-) C s© C/s(C)
Occurrence Areas
(m)
0 2114 12 1.183 0.289 -0.115 0.121 1.298 0.314 4.135 15.000 4.615
25 4343 22 1.068 0.214 -0.222 0.131 1.290 0.251 5.143 27.500 9.482
50 7222 35 1.024 0.169 -0.404 0.149 1.428 0.226 6.326 43.750 15.768
75 10869 39 0.722 0.160 -0.398 0.156 1.120 0.224 5.001 48.750 23.730
100 15324 44 0.498 0.151 -0.392 0.167 0.890 0.225 3.956 55.000 33.456
125 20600 46 0.246 0.148 -0.259 0.172 0.505 0.226 2.230 57.500 44.975
150 26705 49 0.049 0.143 -0.073 0.180 0.123 0.230 0.535 61.250 58.304
175 33641 54 -0.085 0.136 0.202 0.196 -0.287 0.239 -1.201 67.500 73.447
200 41410 56 -0.257 0.134 1.150 0.205 -1.407 0.244 -5.754 70.000 90.466
>200 45803 80 0.000 0.000 0.000 0.000 0.000 0.000 0.000 100 100
8 80
6.326
6 70
4 60
2 1.428 50
0 43.75% 40
-2 30
Percent of Au occurrences
-4 20
Contrast, C
-6 10
Studentised C
-8 0
0 25 50 75 100 125 150 175 200
Buffer distance, m
related with buffer at distance for Clay alteration.
92
Figure 6.12: Binary predictor pattern of clay alteration with 50 m buffer zones.
6.2.1.6.Calculating Weights of evidence of Limonitic Alteration
The binary map cumulative distances 0 – 200 m from limonitic alteration

zones are created. The map is then crossed with the mineral occurrences map (of
80 occurences) to estimate weights of evidence using formula above (Section 3.7).
The result of analysis was shown in Table C.1f of Appendix C and Figure 6.13.
The opimum buffer is defined at 150 m with highest Studentized value of C

(6.163). The resulting buffered hydrothermal alteration domain cover 18.566% of
the study area and 38 (47.5%) out of 80 mineral occurrences are presented in this
distance (Figure 6.13). The weights used in the final model are W+ = 0.942 and W-
= -0.44.
Based on the above relationships, clay alteration are created buffer at

distance within 150 m for favourable and outside of 150 m for non favourable
(Figure 6.14).
93
Table 6.6: Variation of weights and contrasts for cumulative distances from
limonitic alteration with respect to gold deposit
Distance
%
Buffer N{B} N B  D W+ s(W+) W- s(W-) C s (C) C/s(C) % of Areas
Occurrence
(m)
0 1987 11 1.062 0.317 -0.089 0.120 1.151 0.339 3.397 12.500 4.338
25 3154 14 0.935 0.268 -0.121 0.123 1.057 0.295 3.584 17.500 6.886
50 4146 22 1.115 0.214 -0.227 0.131 1.342 0.251 5.348 27.500 9.052
75 5141 26 1.066 0.197 -0.274 0.136 1.341 0.239 5.607 32.500 11.224
100 6188 31 1.057 0.180 -0.346 0.143 1.403 0.230 6.101 38.750 13.510
125 7307 33 0.953 0.174 -0.359 0.146 1.311 0.227 5.765 41.250 15.953
150 8504 38 0.942 0.163 -0.440 0.154 1.382 0.224 6.163 47.500 18.566
175 9785 41 0.877 0.157 -0.479 0.160 1.356 0.224 6.056 51.250 21.363
200 41410 43 0.794 0.153 -0.493 0.164 1.287 0.225 5.732 53.750 24.348
>200 45803 80 0.000 0.000 0.000 0.000 0.000 0.000 0.000 100.000 100.000
7 60
6.163
6 50
47.5%
5
40
4
Percent of Au occurrences 30
3 Contrast, C
Studentised C 20
2
1 10
0 0
0 25 50 75 100 125 150 175 200
Buffer distance, m
related with buffer at distance for Limonitic alteration.
94
Figure 6.14: Binary predictor pattern of limonitic alteration with 150 m buffer zones
6.2.2. Test of Conditional Independence
6.2.2. Test of Conditional Independence
The statistical validity of the resulting predictive maps is examined by

considering a contingency table based on all pairs of maps, using a chi-squared
test. In our case, the null hypothesis of the conditional independence is not
rejected at the 98% significant level for any of the six pairs of maps (Table 6.8)
Table 6.7 show example of using contingency table for testing conditional
independence between the binary map of host rocks (B1) and the binary map of
distance to heat sources evidence map (B2). Values in bold are observed on the
map, those in brackets are the expected values using right-hand of Equation 3.15-
3.18 (see Section 3.4.2 for more detail). Result of calculated chi-square value is
2.882. Tabled chi-square value at the 98% significant level with 1 degree of
freedom is 5.4; thus, the assumption of conditional independence between the two
binary maps is not rejected at 98% significance level.
95
Table 6. 7: Chi-square values ( χ2 ) for testing conditional independence between

host rocks (B1) and heat sources (B2) evidence map with respect to the Au
occurrences.
Evidence map B1 Present B1 Absent Total

B2 Present 38 (36.075) 1 (2.925) 39
B2 Absent 36 (37.925) 5 (3.075) 41
Total 74 6 80
The result of carrying out all the pairs of binary evidence maps are
summarized in Table 6.8. From the table, the calculated values χ2 are smaller than
the 5.4, thus the hypothesis for conditional independence is obeyed or accepted at
this probability level (Alder and Roessler, 1972), except the pair NE-SW and
Clay, NE-SW and limonitic, and clay and limonitic are bigger than 5.4. Thus,
three binary predictor patterns are not statistically significant. Based on the critical
χ2 used, they can still be combined to conduct map mineral potential according
due to equation 3.19 (see section 3.4.1), but the statistical validity of the resulting
posterior probability map must be examined by applying an overall test of
conditional independence (Carranza, 2002).
Table 6. 8: Calculated ( χ2 ) values for testing conditional independence between

all pairs of binary maps with respect to gold occurrences
Evidence
Heat NE-SW NW-SE Clay Limonitic
maps
Host 2.882 0.214 2.391 0.254 0.192
Heat 0.446 1.263 0.486 1.325
NE-SW 0.632 8.832 6.584
NW-SE 0.369 0.944
Clay 11.684
96
6.3. Results
6.3.1. Probabilistic mapping of Au Minerals Potential
The prior probability P {B} = 80/45803 = 0.0017466 and loge {D} = loge
(P {B}/1- P {B}) = - 6.3483 (equation 3-1, 3-4 and detail in Appendix E). The
values and binary predictor maps were used to generate regional scale posterior
probability maps of the area. Initially, a probability map is generated using the 6
maps and 4 maps (exclusive clay and limonitic alteration) with the values from
Table 6.10. The binary pattern of clay alteration and limonitic alteration show the
chi-square  2 more than 5.4 (with 1 degree of freedom and a probability level of
98%, table Chi-square value is 5.4), so clay alteration and limonitic alteration is
not strong relationship with Au occurrences (Table 6.8).
The binary predictor maps are integrated to generate a predictive map using
equation 3-14, which based on Bonham-Carter (1994), can also be written as
exp(-6.348+NE-SE +NW -SE +Host +Heat  Clay  Limonitic)

Pred.map =
(1  exp(6.348  NE-SE +NW -SE +Host +Heat  Clay  Limonitic)
According to the value of posterior probabilities, the favorability map of

Au potential is separated into favourable and unfavourable zones. A ratio less than
one ( Pposterior / PPr ior 1or Pposterior  PPr ior (0.0017466)) is indicated as unfavorable
zone and a ratio greater than one are indicated as favorable zone. The favorability
maps of Au minerals potential are shown in Figure 6.15 and6.16.
6.3.2. Validating of Predictive Maps and Determining best predictive map
The potential maps in Figure 6.15 and Figure 6.16 shown the different result
of the model validation. The results of the validation of both posterior probability
(potential) maps are given in Table 6.11.
The posterior probability map of gold occurrence in Figure 6.15 indicates

that about 16.27% is favorable zones and 83.73% is unfavorable zones. Due to
gold occurrences map reveals that 10 (12.5%) of 80 occurrences gold deposits are
97
presented in unfavorable zones and 70 (87.5%) are present in favorable zones of

the model deposits. Moreover, there have only one gold validation deposits are
located in unfavorable zone while other 26 gold deposits are located in favorable
zone; it means the statistical validity is acceptable (96.3% of validation deposits).
Based on the figure 6.16 clarified that about 21.62% of study areas is
favorable zone and 78.38% is unfavorable zone. For gold probability, there have
only 74 (92.5%) out of 80 gold deposits fall within favorable zones and there are
26 (96.3%) among of 27 gold validation deposits. According to model validation,
the statistical validity is acceptable.
Depend on the conditionally independent, high statistically significant

contrasts validation model; the posterior probability of gold potential map (Figure
6.16) has higher successful rate then the gold potential map (Figure 6.15).
6.3.3. Overall Test
The posterior probability map shown in Figure 6.15 was analyzed by an

overall test for conditional independence and the predicted number of deposits is
91 (calculated using equation 3.20, Table 6.9), this is higher than 13.5% greater
than the observed number and is very close to the accepted threshold of 15%
;therefore this map was accepted as statistically valid.
The predict of favorable zones in the map (Figure 6.16) are spatially
coincided with the zones that were known of mineral district in West Java
(Setijadji and Maryono, 2012; Bushki et al.,2012; Sunarie et al., 2011; Lubis et
al., 2012; Ismayanto et al., 2009; Harijoko et al., 2007; Warmada et al, 2007;
Milesi et al, 1994; Marcoux and Milesi, 1994) such as Bayah Dome, Ciemas,
Tasikmalaya, Cianjur districts, Pandeglang, Purwakarta and Bogor zone.
98
Table 6.9: Summarized of overall test conditional independence in posterior

probability map shown in figure 6.16
Probability values in map Area Predicted Number of deposit
0.000028036 20335 0.57
0.000051839 14104 0.73
0.0001709 4710 0.80
0.001937 1528 2.96
0.005712 2118 12.10
0.01024 1775 18.18
0.04502 1233 55.51
TOTAL 45803 90.85

99
Table 6.10: An appropriated the evidence maps base on conditionally independent and statistically significant contrasts for predict the Au
occurrences in research area
Distance
Binary predictor of Buffer N{B} N B  D W+ s(W+) W- s(W-) C s( C) C/s(C)
% % of
Ratio
Occurrence Areas
(m)
NE-SW lineaments 2000 18257 59 0.617 0.13 -0.83 0.218 1.447 0.254 5.69 73.75 39.86 1.85
NW-SE lineaments 1000 8302 36 0.912 0.167 -0.398 0.151 1.31 0.225 5.822 45 18.125 2.483
Host Rock 0 7130 74 1.791 0.117 -2.423 0.408 4.213 0.425 9.922 92.5 15.567 5.942
Heat Sources 4500 7792 39 1.056 0.161 -0.483 0.156 1.539 0.224 6.868 48.75 17.01 2.866
Clay alteration 50 7222 35 1.024 1.169 -0.404 0.149 1.428 0.226 6.326 43.75 15.77 2.775
Limonitic alteration 150 8504 38 0.942 0.163 -0.44 0.154 1.382 0.224 6.163 47.5 18.57 2.558
Table 6. 11: Inventory of Au occurrences in zones of different posterior based on figure 6.15, which are derived from all of predictor patterns
present in Table 6.9, and based on resulting from exclusion of clay and limonitic alteration binary predictor patterns.
Posterior probability map based on figure 6.15 Posterior probability map based on figure 6.16
Au Potential
% Area Model deposits Validation deposits % Area Model deposits Validation deposits
Number % Number % Number % Number %
Unfavorable
83.73 10 12.5 1 3.70 78.38 6 7.5 1 3.70
zones
Favorable
16.27 70 87.5 26 21.62 74 92.5 26 96.3
zones 96.3
Total 80 100 27 100 80 80 100 27 100

100
Figure 6.15: Posterior probability map of Au in research area, deveried from all predictor patterns present in table 6.10.
101
Figure 6.16: Posterior probability map of Au in research area, resulting from exclusion of clay and limonitic alteration binary predictor
patterns.
102
CHAPTER VII: DISCUSSION
7.1. Au Potential Map
Generally , WofE mineral potential modeling is successful if the

resulting mineral potential maps (1) predict the distribution of known mineral
occurrences, and/or (2) are useful for delineating or targeting regions of elevated
mineral potential for follow up exploration (Mihalasky 1999)
Based on the ratio of posterior probability to prior probability

( Pposterior / PPr ior ) (presented in Chapter VI, Figure 6.16), the resulting predictive
map is classified into three categories: low prospectively, moderate prospectively,

and high prospectively. Thus, if the ratio is less then 1 (Posterior P/Prior P <1)
than potential is unfavorable zones, if the ratio range from 1-2 is classified as low
potential, if the ratio range from 2-5 is classified as moderate and if the ratio is
greater than 5 is classified as high (Figure 7.1).
From the result of the Au prospectively map in Figure 7.1 indicated that
about 3649km2 (7.96%) of research area is low potential, 2844.75km2 (6.21%) is
moderate potential and 2907 (6.35%) is high potential zones. There have 6 targets
prospective target for further investigation are shown in Figure 7.1, the biggest
prospective gold deposits are located in Bayah dome (Taget 1) and other small
potential are located in Ciemas (Target 2), Garut (Target 3), Tasikmalaya (target
4) Bogor zone, Cianjur districts and Purwakarta (target 5), Pandeglang (Honjie
Igneous Complex, target 6). In contrast, there have no potential area of gold
deposit in Coastal Plan of Batavia.
1. Prospective Target 1:
The Bayah Dome (target 1) in the western part of the Sunda-Banda arc is
high potential zones by low-sulphidation epithermal vein sytems including
Gunung Pongor, Cikidang, Cirotan, Ciawitali Cikotok destricts and aother
deposits (Figure 7.3A).
103
The highest potential area are dominated with major NE-SW and NW-SE
lineaments, the intersection zones between lineaments, and also characterized by
minor N-S lineament (Figure 7.3B), which could provide fluid pathways for
mineralizing fluids at upper crustal levels and facilitate the movement of deep-
crustal fluids to higher crustal levels, as well as provide structural traps and the
appropriate physio-chemical conditions conductive to ore deposition. Lithologies
permissive high potential is dominaied by the intrusive intermediate, extrusive
intermediate lava/poymict/pyroclastic, sandstone, breccias and claystone- shale,
which the ranges of age are between Pleistocene-Eocene.
Mineralisations in Bayah dome are hosted by strike-slip faults cutting
Miocene to Pliocene volcanic and plutonic formation. Based on vein partterns, the
mineralization consists of two major styles: discordant gold-bearing quartz veins
or crustiform banding veins of the ‘‘Gunung Pongkor’’ type and breccias veins of
the ‘‘Cirotan’’ type (Marcoux and Milesi, 1994).
The prospective target 2 is located in southern part of West Java; the

potential zones are dominated in Ciemas destrict. This region is covering the
Cigaru local mine and there are several prospect areas such as Cimanggu,
Cijiwa, Cilubang, Cipaku, and Citugu (Figure 7.4A) and some exploration works
have been conducted by Hunamas and Mispec (Syafrizal et al., 2011; Basuki et
al., 2012).
The potential zones are located and parallel to major NW-SE lineaments
(Figure7.4B). The lineaments are not dense like target 1. The lithologies
associated with the potential zones are including Upper Miocene to Miocene
intrusive inetermediate, Lower Miocene extrusive intermediates lava, Miocene-
Pliocene clastic limestone, and Eocene sandstones.
104
The prospective target 3 is located Garut Renency, about 75 km to the

southeast of the major city of Bandung. These area targets have many gold
deposits like Cijulang, Arinem, Pakenjeng, Cihar, Cibaliung, Bunikasih and
Tutugan Area (Figure 7.5A).The potential zones are associated with NE-SW and
NW-SE lineament (Figure 7.5B) and the litologies is hosted by extrusive
intermediate pyroclastic, extrusive felsic lava, intrusive intermediate and the
ranges of age from Miocene-Upper Pleistocene.
4. Prospective Target 4, 5, and 6
The prospective target 4, 5, 6 are located in Tasikmalaya, Bogor, Cianjur,

Purwakarta and Pandeglang (Honjie Igneous Complex, target 6). The potential
zones are mostly associated with NE-SW and NW-SE lineaments (Figure 7.6B,
7.7B, 7.8B). There are many gold mineraliation such salopa, Cineam, Cipatu
(Tasikmalaya), Cibaliung, Cicurug, Cibeber (Honjie Igneous Complex), Awilega
Gunung Parang (Purwakarta), Gunung Pancar (Bogor). The litologies is hosted by
Intrusive intermediate, claystone-shale, extrusive intermediate polymict,
sandstone, extrusive felsic lava, intrusive felsic (Lower Miocene –Pleistocene).
60%
% of area 55%
50% % of Au occurrences
40%
30%
%
30%
20%
10% 7.79%7.50% 6.21% 6.35%
0%
Low potential Moderate potential High potential
Figure 7.1: Percentage of Potential class

105
Figure 7.2: Predictive map of gold in West Java and generate prospective target area for Au
106
Ciawital G.Pongkor
i
Cirotan Cikidan
Cikotok g
Figure 7.3: Gold potential target areas 1: A) Mineral potential map overlain by Au
occurrences as training and validation data, B) Mineral potential map
superimposed on shaded of topography, overlain by lineaments and Au
occurrences
107
Cigaru
Figure 7.4: Gold potential target areas 2 ; (A) Mineral potential map overlain by Au
occurrences as training and validation data ; (B) Mineral potential map
occurrences
108
Bunikasih
Cihar
Tutugan
Cijulang
Pakenjeng
ARINEM
occurrences as training and validation data: B) Mineral potential map
occurrences
109
Figure7.6: Gold potential target areas 4: A) Mineral potential map overlain by Au

occurrences as training and validation data: B) Mineral potential map
occurrences
110
Figure7.7: Gold potential target areas 5: A) Mineral potential map overlain by Au

occurrences
111
occurrences
112
7.1.2. Weight of Evidence Approach
The weight of evidence method uses the Bayesian approach of combining

datases, which is based on a probability framework. The approach using Bayesian
probability theory for mapping mineral potential provides a simple statistical
method for predicting mineral potential for districts where a number of
representative mineral occurrences are known. The method is practically suited
for modeling the spatial correlations between geological features and known
mineral occurrences which are important for mapping mineral potential.
In present study, the studentized C and contrast (C) is useful in determining

the cutoff distances to use for conversion of multi-class proximity maps, to curvi-
linear geologic features, into binary patterns. However, a good knowledge of the
mineral occurences is vital to determine the cutoff distances. For each test
domain, the maximum contrast (C) often provides the best measure of statial
correlation with the mineral occurrence points in the case of a large area and a
large number of mineral occurences. In contrast, in the cases of small area and a
small number of mineral occurrences, the uncertainty of the weights could be
large and contrast (C) can be meaningless. Therefore, the Studentized C is more
useful than contrast (C) to define the optimum cutoff.
Overall test and test condtional independence were implemented to validate

the result of the mineral potential map. Conditional independence (CI) is used
before combining map, to reject or reduce a problem for the invalidated binary
predictor patterns. The pairwise test of conditional independence is important
in determining which binary predictor patterns to combine to map mineral
potential, an overall test of conditional independence is important in
determining the statistical validity of the resulting posterior probability map.
The results of the gold potential map are getting from executing the
different six evidences maps such as Host rock (Tertiary- Pleistocene lithology
unit, heat source (Tertiary intrusive rock unit), NE-SW lineaments, NW-SE
lineaments, clay alteration and limonitic alteration. The number of mineral
113
deposits is used in this case consists of 107 known mineral deposits and divides
into 2 subsets (80 for model deposits and 27 for validation deposits).
All of evidence maps were created buffer at distance and optimum cutoff
distance are summarized in Table 6.10. The positive weight (W+) of host rock
evidence map and ratio of percentage of Au occurrences to percentage of total
area is greater than other evidence maps, following by Heat source, clay
alteration, NW-SE lineaments limonitic alteration, and NE-SW lineaments
evidence maps (Figure 7.9).
The host rock evidence map generates from lithology map is very important
control on emplacement of Au occurrences indicated to be the strongest
spatial predictor than other. The hillshade images derived from ASTER GDEM
was useful to extract lineaments feature and NE-SW lineaments is
important one.
The test for conditional independence shows that chi-square value for the
pair NESW and Clay, NESW and Limonitic, and Clay and Limonitic are
statistically significant. They do not reject the binary maps of these geological
features in the analysis because gold mineralization in the district is known to be
associated with them.
2.5
2 1.791
1.5
W+
1.056 1.024
0.912 0.942
1
0.617
0.5
0
NE-SW L. NW-SE L. Host R. Heat S. Clay A. Limonitic A.
Name of Binary map
Figure 7.9: Graph showing the positive weight (W+) of all evidence maps which used
to predict Au occurrences map.
114
CHAPTER VIII:
CONCLUSIONS AND RECOMENDATION
8.1. Conclusions
The method of WOFE is applied to obtain gold potential map based on

geological, structural, and remote sensing data or maps, and also, 107 known gold
mineralized locations in West Java. These exploratory maps are integrated to
produce the final predictive mineral potential map. The result of Au potential map
indicates that about 9902 km2 (21.62%) of the province as favourable zones. It
predicts correctly 74 (92.5%) of the 80 training data and predicts correctly 26
(96.35%) of the 27 validation data. Moreover, the result of Au prospectively map
indicats that about 3649km2 (7.96%) of research area is low potential, 2844.75km2
(6.21%) is moderate potential and 2907 (6.35%) is high potential zones (Figure
8.1). Bayah Dome is the highest potential zone, mineralization are associated with
major NE-SW and NW-SE lineaments, the intersection zones between lineaments,
and also characterized by minor N-S lineament and the lithologies of Bayah Dome
is dominaied by the intrusive intermediate, extrusive intermediate
lava/poymict/pyroclastic, sandstone, breccias and claystone- shale which the
ranges of age are between Pleistocene-Eocene. Potential areas of Au occurrences
in research area is associated with NE-SW and NW-SE structure/lineaments,
dominated surrounding the Tertiary extrusive volcanic rock and intrusive rock
and hosted in Miocene to Pleistocene lithology rock unit
The major prospective target areas for future exploration are located
in southern mountain of research area like Ciemas, Garut, Cianjur,
Tasikmalaya, and Bayha Dome is located western of Sunda-Banda arc. The small
prospective target areas for future exploration are located in Bogor zone,
Purwakarta and Pandeglang (Honjie Igneous Complex). These mineralization
zones are also confirmed following field checking or visit of the study area.
These target areas desirably include the variety of mineralization elements. This
research shows that the method of WoFE is an effective technique for the
exploration and evaluation of regional-scale mineral deposits.
115
Figure 8. 1: Predictive map of gold occurrences in West Java

116
8.2. Recommendation
There are some recommendations as follow:

 Structures in this research are considered as lineaments, which are generated
from ASTER GDEM. For future research should be considered as faults and
probably study of structure controls of mineralization in district or local
scale.
 Clay and limonitic alteration were generated from Landsat 5 TM data.
Future research should be considered other alteration types like quartz,
alunite, and illite and used high quality data as Landsat 7 ETM+ or ASTER
data. ASTER data has better spectral resolution in the SWIR and TIR
regions and higher spatial resolution in VNIR region.
 In this research, some of factors are not considered such as : Geophysical
(gravity anomaly data), geochemical anomaly (stream sediment, igneous
rock radiometric age dates, major and minor element data)
 Structures in this research are considerd as lineaments, which are generated
from ASTER GDEM. For future research should be considered as faults and
probably study of structure controls of mineralization in district or local
scale. One more, aeromagnetic data or gravity data should use to provide
structural information.
 The mineral potential maps in this area are small scale (regional scale),
which can be guide only exploration target to data collection; they are not
intended for delineating drilling targets. For future research should be
considered the large scale (local scale) and focus on the areas presents as
high potential zone in this research.
 The future study used another methodology for predictive mapping of the
mineral potential and compare to this Weight of Evidence Method like
fuzzy logic.
117
REFERENCES
ANTAM (1994) Exploration report on gold, silver ores and associated minerals,
Cikidang and surrounding area, Bayah district, Lebak Regency, West Java,
Indonesia. SPK. 719/2212/KOG/1994, 34p. (unpubl. report) (in Indonesian).
Abdullah, A., Akhir, J. M., and Abdullah, I. (2010) Automatic Mapping of
Lineaments Using Shaded Relief Images Derived from Digital Elevation
Model (DEMs) in the Maran–Sungi Lembing Area, Malaysia. The
Electronic Journal of Geotechnical Engineering, Vol.15, Bund. J, p. 949-
957.
Abrams, M. (2000) ASTER: data products for the high spatial resolution image on
NASA’s EOS-AM1 platform, p. 847–861.
Alzwar, M., Akbar, N., and Bachri, S. (1992) Systematic geological map,
International Journal of Remote Sensing, 21, pp. Indonesia, quadrangle
Graut 1208-6 and Pameungpeuk 1208-3, scale 1: 100.000, Geological
Research and Development Center , Bandung.
Barne, H.L. (1979) Geochemistry of hydrothermal ore deposits. John Wiley and
Sons, Newyork, 798p.
Basuki, N.I., Prihatmoko, S., and Suparka, E. (2012) Gold Mineralization Systems
in Southern Mountain Range, West Java, Proceedings of Banda and Eastern
Sunda Arcs 2012 MGEI Annual Convention 26-27 November 2012,
Malang, East Java, Indonesia.
Basuki, A., Sumanagara, A. D. and Sinambela, D. (1994) The Gunung Pongkor
gold-silverdeposit, West Java, Indonesia. Geochem.Explor, 50, 371–391.
Bonham-Carter, G.F. (1994), Geographic Information Systems for Geoscientists,
Modelling with GIS, Pergamon, Ontario, 398pp.
Bonham-Carter, G., Wright, D., and Agterberg, D. (1989). Weights of evidence
modeling with GIS: a new approach to mapping mineral deposits.
Geological Survey Canada Paper, 89-9:171–183.
Bronto, S. (1989) Volcanic Geology of Galunggung, West Java, Indonesia, PhD
Thesis, University of Canterbury.
118
Carranza, E.J.M. (2003) Practical Exercise: Weight of Evidence Modeling for

Mineral Potential Mapping. MREE Elective Module 9, 29p.
Carranza, E.J.M. (2002) Geologically-Constrained Mineral Potential Mapping.
PhD Thesis, Delft University of Technology, The Netherlands, 480 pp.
Claproth, R. (1989) Petrology and Geochemistry of Volcanic Rocks from
Ungaran, Central Java, Indonesia. Doctor of Philosophy Thesis, Department
of Geology, University of Wollongong.
Cook, D., and Gemmell, B. (1996). Ore Deposit Studies and Exploration Models.
4th edition, Manual 5, University of Tasmania.
Corbett, G.J. (2009) Geological models in epithermal – porphyry exploration:
Terry Leach’s legacy, www.corbettgeology.com.
Corbett, G.J. (2007) Controls to low sulphidation epithermal Au-Ag: Talk
presented at a meeting of the Sydney Mineral Exploration Discussion Group
(SMEDG) with powerpoint and text on SMEDG website
www.smedg.org.au.
Corbett, G.J. (2004) Epithermal and porphyry gold – Geological models in
Pacrim Congress 2004, Adelaide, The Australasian Institute of Mining and
Metallurgy, p. 15- 23.
Corbett, G.J. (2002) Epithermal Gold for Explorationists: AIG Presidents Lecture,
AIG online Journal April 2002, AIG website www.aig.asn.au
Corbett, G. J., and Leach, T. M. (1998) Southwest Pacific rim gold-copper
systems: structure, alteration, and mineralisation. Society of Economic
Geologists Special Publication 6.
Einaudi, M.T., Hedenquist, J.W., and Inan, E. (2003) Sulfidation state of fluids in
active and extinct hydrothermal systems: Transitions from porphyry to
epithermal environments: Society of Economic Geologists Special
Publication 10, p. 285–314.
El Khidir, S.O., and Babikir, I. A. A. (2013) Digital image processing and
geospatial analysis of landsat 7 ETM+ for mineral exploration, Abidiya
area, North Sudan. International Journal of Geomatics and Geosciences,
Vol. 3, No 3, P. 645-658.
119
Evans, A.M. (1993) Ore Geology and Industrial Minerals; An Introduction, 3rd
Ed. Blackwell Scientific Publication.
Fraser, S.J. and Green, A.A. (1987) A software defoliant for geological analysis of
band ratios. International Journal of Remote Sensing, vol. 8, no. 3, pp. 525-
532.
Fujisada, H., Bailey, G. B., Kelly, G. G., Hara, S. and Abrams, M.J. (2005)
ASTER DEM Performance. IEEE transactions on Geoscience and
RemoteSensing, 43, pp. 2707- 2714.
Hamilton, W. (1979) Tectonics of the Indonesian Region: U.S. Geological Survey
Professional Paper, 1078, US Geological Survey, Reston, 345
Hall, R. (2002) Cenozoic geological and plate tectonic evolution of SE Asia and
the SW Pacific: Computer-based reconstructions, model and animations.
Journal of Asian Earth Sciences, Vol. 20, p. 353–434.
Harijoko, A., Ohbuchi, Y., Motomura, Y., Imai, A. And Watanabe, K. (2007)
Characteristics of the Cibaliung gold deposit: Miocenelow-sulfidation-type
epithermal gold deposit in Western Java, Indonesia: Resource Geology,
57,114-123
Harris, J.R. (2006) GIS for the Earth Sciences, Geological Association of Canada
Special paper 44.
Hedenquist, J.W., Arribas, A., Jr., and Gonzales-Urien, E. (2000) Exploration for
epithermal gold deposits: Reviews in Economic Geology, v. 13, p. 245–277.
Hedenquist, J. W., Izawa, E.. Arribas, A., Jr. and White, N. C. (1996)
Epithermalgold deposits: Styles, characteristics, and exploration. Resource
Geology,Vol.57, no.1, p. 230-240.
Idrus, A., Fadlin, Putri, R.I. and Hatmanda, M. (2013) Styles and Characteristics
of Precious and Base Metal Mineralization at the Southern Part of West
Java Region, Indonesia, The 5th AUN/SEED-Net Regional Conference on
Geological Engineering 15-16 January 2013, Kuala Lumpur, Malaysia.
Ismayanto, A.F , Syafrizal, Notosiswoyo, S (2009) Characteristic of Hydrothermal
Mineralization at Gunung Subang Area, Cianjur District, West Java:
120
International Conference Earth Science and Technology Yogyakarta 6-7

August 2009,8pp
Katili, J. A. (1975) Volcanism and plate tectonics in the Indonesian island arcs:
Tectonophysics, 26, 165-188.
Kirkham, R.V., Sinclair, W.D. 1995: Geology of Canadian Mineral Deposit
Types. Geology of Canada No. 8, edited by O.R. Eckstrand, O. R., Sinclair,
W.D.,Thorpe, R.I
Koesmono, M., Kusnama. and Suwarna, N. (1996) Geologi Lembar
Sindangbarang dan Bandarwaru, Jawa Barat, Pusat Penelitian dan
Pengembangan Geologi, Indonesia.
Lindgern, W. (1933) Mineral Deposits, 4th Ed., McGraw-Hill, New York, 930pp.
Lubis, H., Prihatmoko, S., and Yerri Heryunanto, Y. (2012) Geology and
Exploration for Low Sulfidation EpithermalGold-Silver Mineralization in
Kerta, Banten, Proceedings of Banda and Eastern Sunda Arcs 2012 Mgei
Annual Convention26-27 November 2012, Malang, East Java, Indonesia
Malod, J.A., Karta, K., Beslier, M.O., and Zen, M.T.(1995) Fromnormal to
oblique subduction: tectonic relationships betweenJava and Sumatra. J.
Southeast Asian Earth Sci. 12, 85– 93.
Marcoux, E., and Milesi, J.P. (1994) Epithermal gold depositsin West Java,
Indonesia: geology, age and crustal sources: in van Leeuwen, T.M.,
Hedenquuist, J.W., James, L.P., and Dow, J.A.S., eds., Mineral deposits in
Indonesia, Discoveries of the Past 25 years: Journal of Geochemical
Exploration, v.50, p.393-408.
Martodjojo, S., 1984, Evolusi Cekungan Bogor, Jawa Barat, Doctoral
Dissertation, Department of Geology, Institut Teknologi Bandung,
Bandung, 355p.
Meinert, L.D., 1992: Skarns and Skarn Deposit. Geoscience Canada 19, 145-162.
Milesi, J.P., Marcoux, E., Nehlig, P., Sunarya, Y., Sukandar, A., and Felenc, J.
(1994) Cirotan, West Java, Indonesia: a 1.7 Ma hybrid epithermal Au–Ag–
Sn–W deposit. Geology. 89, 227– 245.
121
Milesi, J.P., Marcoux, E., Sitrous, T., Simandjuntak, M., Leroy, J., and Baily L.
(1999) a Pliocene supergene-enriched epithermal Au-Ag-(Mn) deposit.
Mineral deposite 34, p. 131-149.
Mitchell, P.A., Profett, J.M. and Dilles, J.H. (1998) Geological Review of the
Batu Hijau Porphyry Copper-Gold Deposit, Sumbawa Island, Indonesia.
Unpublished Newmont Nusa Tenggara Company Final report. PT Newmont
Nusa Tenggara, Batu Hijau project, Sumbawa, Indonesia, 164p.
Mitchell, A.H.G., and Balce, G.R. (1990) Geological Features of Some
Epithermal Gold Systems,Philippines. J. Geochem. Expl., 35, 241-296.
Mitchell, A.H.G., and Leach, T.M. (1991) Epithermal Gold in the Philippines:
Island Arc Metallogenesis, Geothermal Systems and Geology, 457pp.
Misra, K.C. (2000) Understanding Mineral Deposits: Department of Geological
Sciences, The University of Tennessee, Knoxville, U.S.A, p.701-702
Rosana, M.F., Haryanto, A.D., Yuniardi, Y., and Yuningsih, E.T. (2006) The
Occurrences of Base Metal Mineralization in Cikadu-Cisungsang Area,
Banten Province, Indonesia, Department of Geology, Padjadjaran
University.
Rosana, M. F. and Matsueda, H. (2002) Cikidang hydrothermal gold deposit in
Western Java,Indonesia. Resource Geology,vol. 52, no. 4, 341–352, 2002.
Sabins, F.F. (1999) Remote Sensing for Mineral Exploration. Ore Geology
Reviews, 14, p.157-183.
Sabins, F.F. (1987) Remote Sensing: Principals and Interpretation, 2nd ed, 450p.
Saro, L., and Hyun, J.H. (2011) Application of Artificial Neural Network for
Mineral Potential Mapping : Geoscience Information Center, Korea Institute
of Geoscience and Mineral Resources (KIGAM), 92, Gwahang-no,
Yuseong-gu, Daejeon 305-350, Republic of Korea.
Setijadji, L.D., and Maryono, A. (2012) Geology and Arc Magmatism of the
Eastern Sunda Arc, Indonesia. Proceedings of Banda Eastern Sunda Arcs
2012, MGEI Annual Convention, p. 1-22.
122
Setijadji, L.D. (2009) Gold Related Deposits in the Southern Mountains of East
Java, Indonesia. International Conference Earth Science and Technology,
Yogyakarta 6-7 August 2009, C12-1p.
Setijadji, L.D., Kajino, S., Imai, A., and Watanabe, K. (2006) Cenozoic Island Arc
Magmatism in Java Island (Sunda Arc, Indonesia): Clues on Relationships
between Geodynamics of Volcanic Centers and Ore Mineralization.
Geological Engineering Department, Faculty of Engineering, Gadjah Mada
University.
Setiawan, I., Sudarsono, and Indarto, S. (2010) Mineralisasi Polimetalik Emas dan
Logam Dasar di Pegunungan Selatan Jawa Barat: Kasus Daerah Pakenjeng.
Proceedings 39th Annual IAGI Convention and Exhibition, Lombok, 316.
Sharma, M. P. and Kujur, A. (2012) ASTER DEM Based Studies for Geological
and Geomorphological Investigation in and around Gola block, Ramgarh
District, Jharkhand, India. International Journal of Scientific & Engineering
Research, Vol.3, Issue 2, February-2012, p.1-5.
Simmons, S.F., White, N.C., and John, D.A. (2005) Geological characteristics of
epithermal precious and base metal deposits: Economic Geology, 100th
Anniversary Volume, p. 485–522.
Sinclair, W.D. (2007) Porphyry deposits, in Good fellow, W.D., (ed), Mineral
deposits of Canada: Geological Association of Canada Special Publication
5, p. 223–243.
Sillitoe, R.H., and Hedenquist, J.W. (2003) Linkages between
volcanotectonicsettings, ore fluid compositions, and epithermal precious
metal deposits: Society of Economic Geologists Special Publication 10,
p.315–343.
Smyth, H. R, Hall, R., Hamilton, P.J., and Kinny, P. (2005) East Java: Cenozoic
basins, volcanoes and ancient basement. Jakarta Proceedings, Indonesian
Petroleum Association Annual Convention, 30th, p. 251–266.
Soeria-Atmadja, R. Maury, R. C., Bellon, H., Pringgoprawiro, H., Polves, M. and
Priadi, B. (1994) Tertiary magmatic belts in Java. Journal of Southeast
Asian Earth Science, 9, p.13–27.
123
Sparkes, G.W. (2012)A Prospector’s Guide to Alteration and Epithermal Gold

Mineralization : Examples from Eastern Avalon.
Sunarie, C.Y., Rosana. M.F., and Watanabe, K. (2011) Gold Bearing Quartz Vein
of Tanggeung Prospect, Cianjur, West Java, Indonesia. Proceedings of the
1st Asia Africa Mineral Resources Conference 2011, Fukuoka, Japan
Subandrio, A.S, and Basuki, N.I . (2010) Alteration and Vein Textures
Associated with Gold Mineralization at the Bunikasih Area, Pangalengan,
West Java, Jurnal Geologi Indonesia, Vol. 5 No. 4 Desember 2010: 247-
261.
Suparka, E., Aziz, M., Abdullah, C.I., and Suparka, 2007. Mineralization of Cu-
Au Porphyry Deposits in Cihurip and Surrounding Area, Garut Regency,
West Java. Joint Convention The 36th IAGI Annual Convention and
Exhibition, Bali.
Syafrizal, Indriati, T., Heriawan, M.N., Hede, A., N., H., Hutabarat, Y., and
Saing, S., (2011) Base-metal and Gold Mineralization in Jampang Complex,
Cigaru Local Mine, West Java, Indonesia, Proceedings of the 1st Asia Africa
Mineral Resources Conference 2011, Fukuoka, Japan.
Tun, M.M., Warmada, I.W., Idrus, A., Harijoko, A., Verdiansyah, O., and
Watanabe, K. (2014) Fluid Inclusion Studies of the Cijulang High-
sulfidation Epithermal Prospect, West Java, Indonesia, 2014 3rd
International Conference on Geological and Environmental Sciences
Utoyo, H. (2007) Alteration and Mineralization of Ancient Caldera in
Cupunagara, Subang, West Java, Proceedings Joint Convention Bali 2007,
The 32nd HAGI, The 36th IAGI, And The 29th IATMI Annual Conference
And Exhibition.
van Bemmelen, R.W. (1949) The geology of Indonesia. V.F.A. Govermment.
Printing office. The Hague.732pp.
Verdiansyah, O., Bangun, P., and Rahmat , I. (2012) High Sulfidation Epithermal
Gold Occurrences in Cijulang Area, Garut , West Java, Proceeding of the
41st IAGI annual convention and exhibition, Yogyakarta17-20 September
2012, p 495-513.
124
Wagner. T., Anthony E., Jones. W., and Adrian J. B. (2005) Stable isotope-based
modeling of the origin and genesis of an unusual Au–Ag–Sn–W epithermal
system at Cirotan, Indonesia. Chemical Geology 219 (2005) 237– 260
Warmada, I.W. and Lehmann, B. (2003) Polymetallic Sulfides and Sulfosalts of

the Pongkor Epithermal Gold–Silver Deposit, West Java, Indonesia. The
Canadian Mineralogist Vol. 41, pp. 185-200 (2003)
Warmada, I.W., Lehmann, L., Simandjuntak, M., and Hemes, H.S. (2007) Fluid
Inclusion, Rare-Earth Element and Stable Isotope Study of Carbonate
Minerals from the Pongkor Epithermal Gold – Silver Deposit, West Java,
Indonesia, Resource Geology Vol. 57, No. 2: 124 – 135.
Washington, F.R., Franca-Rocha, Bonham-Carter, G.F, and Misi, A., (2003), GIS
Modeling for Mineral Potential Mapping of Carbonate-Hosted Pb-Zn
Deposits: Revista Brasileira de Geociências, Volume 33, 2003.
White, N.C. and Hedenquist, J.W., 1995, Epithermal gold deposits: Styles,
characteristics, and exploration: Society of Economic Geologists
Newsletter, no. 23, p. 1, 9-13.
Wright, D, F. and Bonham-Carter, G, F. (1996) VHMS favorability mapping with
GIS-based intergration models, Chisel Lake-Anderson Lake area:
Geological Survey of Canada, Bullein 426, p.339-401.
Yuningsih, E.T., Matsueda, H., Setyaraharja, E.P., and Rosana, M.F. (2011) The
Arinem Te-bearing gold-silver-base metal deposit, West Java, Indonesia.
Resource Geology: Rg10-36.
Yuningsih, E.T., Sutopo,B., Setyaraharja, E.P., Bangun, P., and Rosana M.F.
(2012) The Arinem Deposit: An Epithermal Gold-Silver-Base Metal
Mineralization System, West Java Province, Indonesia, Proceedings of
Banda and Eastern Sunda Arcs 2012 MGEI Annual Convention 26-27
November 2012, Malang, East Java, Indonesia.
Yulianto, I. Robert, H. Clements, B. and Christopher R. E. (2007) Structural and
Stratigraphic Evolution of the Offshore Malingping Block, West Java,
Indonesia
125
APPENDICES
126
A.GLOSSARY
Contrast ( C: W+ minus W-) : In binary themes is an overall measure of the spatial
association of an evidential theme with the training points. Contrast is the range in
weights-a measure of how much the prior logit could be up-weighted or down-weighted.
Mineral prospectivity/potential/favorability: Term referring to the chance or likelihood

that mineral deposits of the type sought is contained in a piece of land.
Posterior Probability: The probability that a unit cell contains a training point, given
states of information from the evidential themes. This measurement changes from
location to location depending on the values of the evidence, being larger than the prior
probability where the sum of the weights is positive, and the converse.
Prior Probability: The probability that a unit cell contains a training point before
considering the spatial evidence. Normally it is assumed to be a constant over the study
area equal to the training point density (total number of training points divided by total
study area in unit cells).
Conditional Independence (CI): The weights-of-evidence (WOE) model assumes
that evidential themes are independent, not correlated conditional on the locations
of training points. If this assumption is true, and if the prior probability is defined as
indicated below, the summation of area multiplied by the predicted posterior probability
over the study area should equal the number of training points.
Studentized contrast: contrast, C, divided by its standard deviation.

Training Points/Training Set: The set of spatial point objects whose locations are
to be predicted. In mineral exploration, these are the sites of known mineral deposits or
occurrences. Points are either present or absent. Size or
Other attributes of these points are not modeled. Must be a point theme. Cannot use the
label points of polygon coverage.
Unit Cell: A small unit of area used for counting, and for defining the probability of
occurrence of a point object. Each training point is assumed to occupy a unit cell. All area
measures are transformed to counts of unit cells. The number of unit cells with training
points are unaffected by changes in unit cell size, whereas the number of unit cells in an
area is affected. The unit cell should be small-normally as small or smaller than
the minimum spatial resolution of evidential themes. Weight values are relatively
insensitive to unit cell size if unit cell is small.
127
B. Useful ArcGIS operations, tools and tips

Buffering: For point/line/polygon use the “Buffer” tool. Cannot “buffer” a raster. But
you can “expand”, which effectively does the same thing. Spatial Analyst →Tools →
Generalization → Expand. You can expand the values of a class by a chosen # of
cells/pixels. See also “Multiple Ring Buffer” tool.
Calculate Weights Tool: make sure the “output coordinate system” in the
environments is set in order for it to work.
Converting polygon to raster: conversion tools → to raster polygon → to raster (or

“feature to raster” which will convert any of point/line/polygon).
Converting raster to polygon: Conversion tools from raster → raster to polygon. Make
sure that “simplify polygons” is checked.
Dissolve Tool: allows you to dissolve multiple polygons into one polygon. Data
Management Tools → Generalization → Dissolve.
Deleting data outside of study area: use the clip tool in “Analysis Tools” →
“Extract” →“Clip”
Density analysis tools: Line Density, Point Density

Joining tables: Data management →Joins → Add Join tool. Joins two tables based on a
common Field of your choice.
Merge tool: allows you to combine two layers into one layer. However, may do funky
things to subsequent manipulations/calculations if they are overlapping layers.
Projections:
 to change the projection of a .shp: Data Management Tools → projections
and transformations →feature → project
 to change the projection of a raster: data management tools→ projections
and transformations → project raster. These will transform a layer into a new layer
in the projection of your choice. You will have to import the new data layer.
Union tool: Analysis Tools → Overlay → Union. Combines two features into one.
Editing an attribute table:

 Field: a column in an attribute table. To add a new field
 to edit a field value: turn on editor → make edits to field value(s) → save edits,
stop editing
 to compute area for a polygon: make new field, r-click field title, calculate
geometry (editor does not need to be on)
128
 to compute area for raster, must know cell size beforehand, then: click field
title → Field calculator
 AREA= cell count times your known cell area.
When adding a new field:
Precision = # of digits
Scale = the # of decimal places
Float (when precision less than 6)/double (when precision greater than 6) =
allows decimal places
Value of ‘0’ = binary (either yes or not)
If no decimal places needed (ie scale = 0) then use either short integer or
long integer.
C. Statistics of Landsat TM data

Elementary statistic of spectral band of Landsat TM data of study area
which consist of 6 sheets
Path 123, Row 065
Band 1 2 3 4 5 6 7
Min 8 1 1 1 1 1 1
Maximum 255 255 255 255 255 170 255
Mean 81.05 33.47 30.75 42.13 37.38 130.61 16.43
Std. Dev. 39.76 28.08 34.07 37.38 40.1 4.85 19.66
129
Path 123, Row 064
Band 1 2 3 4 5 6 7
Min 1 1 1 1 1 1 1
Maximum 255 209 255 255 255 164 255
Mean 69.94 24.56 20.26 17.28 15.18 130.57 7.78
Std. Dev. 5.63 3.6 5.14 16.49 19.59 360 7.68
Path 122, Row 064
Band 1 2 3 4 5 6 7
Min 7 2 1 1 1 1 1
Maximum 255 225 255 255 255 177 255
Mean 96.73 36.74 33.99 31.82 37.64 134.84 18.75
Std. Dev. 18.63 10.89 15.9 23.54 38.25 5.13 18.91
Path 122, Row 065
Band 1 2 3 4 5 6 7
Min 1 2 1 1 1 1 1
Maximum 255 225 255 255 255 190 255
Mean 60.86 23.95 20.66 42.78 40.35 137.55 17.1
Std. Dev. 11.06 7.29 10.1 31.71 31.57 6.44 13.3
Path 121, Row 064
Band 1 2 3 4 5 6 7
Min 7 1 1 1 1 1 1
Maximum 255 255 255 255 255 209 255
Mean 79.38 30.49 27.6 21.89 22.22 136.08 12.74
Std. Dev. 40.19 26.96 33.8 30.72 39.61 3.455 22.37
Path 121, Row 065
Band 1 2 3 4 5 6 7
Min 2 1 1 1 1 1 1
Maximum 255 255 255 255 255 170 255
Mean 61.96 25.69 22.49 47.4 40.08 135.18 16.16
Std. Dev. 18.13 10.91 14.35 29.62 29.56 5.01 13.9
130
E. Example for calculation of weight and contrast

NE-SW trending lineament buffered distance 250 m, Table C.1a
Total areas = 45803 km2
Present areas = 2505 km2
Absence areas = 43298 km2
Total number of mineral for model deposits = 80
Number of mineral present = 12
Number of mineral absence = 68
So,
P {B/D} = 12/80= 0.15
P{B/D}= (2505-12)/(45803-80) = 0.055
P{B/D} 0.15
W   log e   1.012
P{B/D} 0.055
P{B/D}= (80-12)/80= 0.85
P{B/D}= (45803-2505-78+12)/(45803-80) = 0.945
P{B/D} 0.85
W   log e   0.106
P{B/D} 0.945
1 1 1 1
s(W  )     0.289
N{B  D} N{B  D} 12 2505  12
1 1 1 1
s(W  )      0.121
N{B  D} N{B  D} 80  12 45803  2505  80  12)
C  (W + )-(W - )=1.012-(-0.106) 1.118
s(C)  s 2 (W  )  s 2 (W  )  0.2892  0.1212  0.314
C/s(C)  1.118 / 0.314  3.564

131
Table C.1. Variation of weight of evidence for cumulative distances from (a) NE-SW trending lineaments, (b) NW-SE trending lineaments, (c) host rock,
(d) heat sources, (e) clay alteration and (f) limonitic alteration, with respect to the gold deposit
%
Binary Distance area Au
N{B} P{B/D} W+ W- C s(W+) S(W-) s(C) C/s (C) Occurren % Area Ratio
Predictor (m) N B  D Absent Total P{B/D} P{B/D} P{B/D} ce
250 2505 12 43298 80 0.150 0.055 1.012 0.850 0.945 -0.106 1.118 0.289 0.121 0.314 3.564 15.000 5.469 2.743
500 5202 17 40601 80 0.213 0.113 0.628 0.788 0.887 -0.119 0.747 0.243 0.126 0.274 2.728 21.250 11.357 1.871
1000 10280 32 35523 80 0.400 0.224 0.579 0.600 0.776 -0.257 0.836 0.177 0.144 0.228 3.660 40.000 22.444 1.782
1500 14614 50 31189 80 0.625 0.319 0.674 0.375 0.681 -0.597 1.271 0.142 0.183 0.231 5.500 62.500 31.906 1.959
(a) 2000 18257 59 27546 80 0.738 0.398 0.617 0.263 0.602 -0.830 1.447 0.130 0.218 0.254 5.690 73.750 39.860 1.850
NE-SW 2500 21293 63 24510 80 0.788 0.464 0.528 0.213 0.536 -0.925 1.453 0.126 0.243 0.273 5.313 78.750 46.488 1.694
trending 3000 23802 64 22001 80 0.800 0.519 0.432 0.200 0.481 -0.877 1.310 0.125 0.250 0.280 4.683 80.000 51.966 1.539
lineament 3500 25938 71 19865 80 0.888 0.566 0.450 0.113 0.434 -1.351 1.801 0.119 0.333 0.354 5.088 88.750 56.629 1.567
4000 27766 74 18037 80 0.925 0.606 0.423 0.075 0.394 -1.660 2.083 0.116 0.408 0.425 4.907 92.500 60.620 1.526
4500 29317 77 16486 80 0.963 0.640 0.409 0.038 0.360 -2.263 2.672 0.114 0.577 0.589 4.540 96.250 64.007 1.504
5000 30620 78 15183 80 0.975 0.668 0.378 0.025 0.332 -2.586 2.965 0.113 0.707 0.716 4.139 97.500 66.852 1.458
>6000 45803 80 0 80 1.000 1.000 0.000 0.000 0.000 100 100 1.000
area Au %
Binary Distance %
Predictor (m)
N{B} N B  D Absen Tota P{B/D} P{B/D} W+ P{B/D} P{B/D} W- C s(W+) S(W-) s(C) C/s (C) Occurre
Area
Ratio
t l nce
500 4046 20 41757 80 0.250 0.088 1.044 0.750 0.912 -0.196 1.239 0.224 0.129 0.259 4.789 25.000 8.833 2.830
1000 8302 36 37501 80 0.450 0.181 0.912 0.550 0.819 -0.398 1.310 0.167 0.151 0.225 5.822 45.000 18.125 2.483
1500 12225 45 33578 80 0.563 0.266 0.747 0.438 0.734 -0.517 1.264 0.149 0.169 0.226 5..604 56.250 26.690 2.108
2000 15662 50 30141 80 0.625 0.341 0.605 0.375 0.659 -0.563 1.168 0.142 0.183 0.231 5.052 62.500 34.194 1.828
(b) 2500 18651 56 27152 80 0.700 0.407 0.543 0.300 0.593 -0.682 1.225 0.134 0.204 0.244 5.017 70.000 40.720 1.719
NW-SE
3000 21255 61 24548 80 0.763 0.464 0.498 0.238 0.536 -0.815 1.313 0.128 0.230 0.263 4.993 76.250 46.405 1.643
trending
lineament 3500 23474 64 22329 80 0.800 0.512 0.446 0.200 0.488 -0.892 1.338 0.125 0.250 0.280 4.758 80.000 51.250 1.561
4000 25385 70 20418 80 0.875 0.554 0.458 0.125 0.446 -1.273 1.730 0.120 0.316 0.338 5.117 87.500 55.422 1.579
4500 27071 73 18732 80 0.913 0.590 0.435 0.088 0.410 -1.543 1.979 0.117 0.378 0.396 4.999 91.250 59.103 1.544
5000 28562 73 17241 80 0.913 0.623 0.382 0.088 0.377 -1.460 1.842 0.117 0.378 0.396 4.654 91.250 62.358 1.463
5500 29855 76 15948 80 0.950 0.651 0.378 0.050 0.349 -1.942 2.320 0.115 0.500 0.513 4.521 95.000 65.181 1.457
>6000 45803 80 0 80 1.000 1.000 0.000 0.000 0.000 100 100 1.000
132
%
N B  D
N{B} P{B/D} P{B/D} W+ P{B/D} P{B/D} W- C s(W+) S(W-) s(C) C/s (C) Occurren % Area Ratio
Predictor (m) Absent Total
ce
0 7022 72 38781 80 0.900 0.152 1.778 0.100 0.848 -2.138 3.916 0.118 0.373 0.354 10.502 90.000 15.331 5.871
500 11544 75 34259 80 0.938 0.251 1.318 0.063 0.749 -2.484 3.802 0.116 0.462 0.447 8.230 93.750 25.204 3.720
1000 13380 77 32423 80 0.963 0.291 1.196 0.038 0.709 -2.940 4.136 0.114 0.589 0.577 7.027 96.250 29.212 3.295
1500 14842 77 30961 80 0.963 0.323 1.092 0.038 0.677 -2.893 3.986 0.114 0.589 0.577 6.772 96.250 32.404 2.970
2000 16091 77 29712 80 0.963 0.350 1.011 0.038 0.650 -2.852 3.863 0.114 0.589 0.577 6.564 96.250 35.131 2.740
(c) 2500 17180 77 28623 80 0.963 0.374 0.945 0.038 0.626 -2.815 3.760 0.114 0.589 0.577 6.388 96.250 37.508 2.566
Host 3000 18158 77 27645 80 0.963 0.395 0.890 0.038 0.605 -2.780 3.670 0.114 0.589 0.577 6.235 96.250 39.644 2.428
rocks
3500 19063 77 26740 80 0.963 0.415 0.841 0.038 0.585 -2.747 3.588 0.114 0.589 0.577 6.095 96.250 41.620 2.313
4000 19921 77 25882 80 0.963 0.434 0.796 0.038 0.566 -2.714 3.511 0.114 0.589 0.577 5.965 96.250 43.493 2.213
5000 21529 77 24274 80 0.963 0.469 0.719 0.038 0.531 -2.650 3.369 0.114 0.589 0.577 5.724 96.250 47.003 2.048
10000 28347 79 17456 80 0.988 0.618 0.468 0.013 0.382 -3.419 3.887 0.113 1.006 1.000 3.863 98.750 61.889 1.596
100.00
>10000 45803 80 0 80 1.000 1.000 0.000 0.000 0.000 100 0 1.000
%
N B  D
N{B} P{B/D} P{B/D} W+ P{B/D} P{B/D} W- C s(W+) S(W-) s(C) C/s (C) Occurren % Area Ratio
Predictor (m) Absent Total
ce
500 1249 11 44554 80 0.138 0.027 1.625 0.863 0.973 -0.120 1.745 0.303 0.120 0.326 5.355 13.750 2.727 5.042
1000 2082 13 43721 80 0.163 0.045 1.278 0.838 0.955 -0.131 1.409 0.278 0.122 0.304 4.638 16.250 4.546 3.575
1500 2914 19 42889 80 0.238 0.063 1.322 0.763 0.937 -0.206 1.528 0.230 0.128 0.263 5.800 23.750 6.362 3.733
2000 3750 20 42053 80 0.250 0.082 1.120 0.750 0.918 -0.203 1.322 0.224 0.129 0.259 5.111 25.000 8.187 3.054
2500 4582 23 41221 80 0.288 0.100 1.059 0.713 0.900 -0.234 1.293 0.209 0.133 0.248 5.223 28.750 10.004 2.874
3000 5399 26 40404 80 0.325 0.118 1.017 0.675 0.882 -0.268 1.285 0.197 0.136 0.239 5.375 32.500 11.787 2.757
3500 6203 32 39600 80 0.400 0.135 1.086 0.600 0.865 -0.366 1.452 0.177 0.144 0.229 6.352 40.000 13.543 2.954
4000 7000 36 38803 80 0.450 0.152 1.083 0.550 0.848 -0.433 1.516 0.167 0.151 0.225 6.734 45.000 15.283 2.944
4500 7792 39 38011 80 0.488 0.170 1.056 0.513 0.830 -0.483 1.539 0.161 0.156 0.224 6.868 48.750 17.012 2.866
(d)
5000 8587 39 37216 80 0.488 0.187 0.958 0.513 0.813 -0.461 1.420 0.160 0.156 0.224 6.339 48.750 18.748 2.600
Heat
6000 10153 45 35650 80 0.563 0.221 0.934 0.438 0.779 -0.577 1.511 0.149 0.169 0.226 6.695 56.250 22.167 2.538
sources
7000 11707 50 34096 80 0.625 0.255 0.897 0.375 0.745 -0.687 1.583 0.142 0.183 0.231 6.848 62.500 25.559 2.445
8000 13280 51 32523 80 0.638 0.289 0.790 0.363 0.711 -0.673 1.463 0.140 0.186 0.233 6.285 63.750 28.994 2.199
9000 14847 52 30956 80 0.650 0.324 0.698 0.350 0.676 -0.659 1.356 0.139 0.189 0.235 5.781 65.000 32.415 2.005
10000 16396 57 29407 80 0.713 0.357 0.690 0.288 0.643 -0.804 1.494 0.133 0.209 0.247 6.045 71.250 35.797 1.990
20000 30451 74 15352 80 0.925 0.664 0.331 0.075 0.336 -1.499 1.829 0.116 0.408 0.425 4.309 92.500 66.483 1.391
30000 38602 78 7201 80 0.975 0.843 0.146 0.025 0.157 -1.840 1.986 0.113 0.707 0.716 2.773 97.500 84.278 1.157
40000 42172 78 3631 80 0.975 0.921 0.057 0.025 0.079 -1.155 1.213 0.113 0.707 0.716 1.693 97.500 92.073 1.059
50000 44139 78 1664 80 0.975 0.964 0.012 0.025 0.036 -0.374 0.386 0.113 0.708 0.717 0.539 97.500 96.367 1.012
>50000 45803 80 0 80 1.000 1.000 0.000 0.000 0.000 100 100 1.000
133
Binary Distance area Au % %

Predictor (m)
N{B}
N B  D Absent Total
P{B/D} P{B/D} W+ P{B/D} P{B/D} W- C s(W+) S(W-) s(C) C/s (C)
Occurrence Area
Ratio
0 2114 12 43689 80 0.150 0.046 1.183 0.850 0.954 -0.115 1.298 0.289 0.121 0.314 4.135 15.000 4.615 3.250
25 4343 22 41460 80 0.275 0.095 1.068 0.725 0.905 -0.222 1.290 0.214 0.131 0.251 5.143 27.500 9.482 2.900
50 7222 35 38581 80 0.438 0.157 1.024 0.563 0.843 -0.404 1.428 0.169 0.149 0.226 6.326 43.750 15.768 2.775
75 10869 39 34934 80 0.488 0.237 0.722 0.513 0.763 -0.398 1.120 0.160 0.156 0.224 5.001 48.750 23.730 2.054
(e)
100 15324 44 30479 80 0.550 0.334 0.498 0.450 0.666 -0.392 0.890 0.151 0.167 0.225 3.956 55.000 33.456 1.644
Clay
alteration 125 20600 46 25203 80 0.575 0.450 0.246 0.425 0.550 -0.259 0.505 0.148 0.172 0.226 2.230 57.500 44.975 1.278
150 26705 49 19098 80 0.613 0.583 0.049 0.388 0.417 -0.073 0.123 0.143 0.180 0.230 0.535 61.250 58.304 1.051
175 33641 54 12162 80 0.675 0.735 -0.085 0.325 0.265 0.202 -0.287 0.136 0.196 0.239 -1.201 67.500 73.447 0.919
200 41436 56 4367 80 0.700 0.905 -0.257 0.300 0.095 1.150 -1.407 0.134 0.205 0.244 -5.754 70.000 90.466 0.774
>200 45803 80 0 80 1.000 0.000 100 100 1.000
Binary Distance N{B} area Au %

Predictor (m) N B  D Absent Total
P{B/D}
P{B/D}
W+ P{B/D} P{B/D} W- C s(W+) S(W-) s(C) C/s (C)
Occurrence
% Area Ratio
0 1987 10 43816 80 0.125 0.043 1.062 0.875 0.957 -0.089 1.151 0.317 0.120 0.339 3.397 12.500 4.338 2.881
25 3154 14 42649 80 0.175 0.069 0.935 0.825 0.931 -0.121 1.057 0.268 0.123 0.295 3.584 17.500 6.886 2.541
50 4146 22 41657 80 0.275 0.090 1.115 0.725 0.910 -0.227 1.342 0.214 0.131 0.251 5.348 27.500 9.052 3.038
75 5141 26 40662 80 0.325 0.112 1.066 0.675 0.888 -0.274 1.341 0.197 0.136 0.239 5.607 32.500 11.224 2.896
(e)
100 6188 31 39615 80 0.388 0.135 1.057 0.613 0.865 -0.346 1.403 0.180 0.143 0.230 6.101 38.750 13.510 2.868
Limonitic
alteration 125 7307 33 38496 80 0.413 0.159 0.953 0.588 0.841 -0.359 1.311 0.174 0.146 0.227 5.765 41.250 15.953 2.586
150 8504 38 37299 80 0.475 0.185 0.942 0.525 0.815 -0.440 1.382 0.163 0.154 0.224 6.163 47.500 18.566 2.558
175 9785 41 36018 80 0.513 0.213 0.877 0.488 0.787 -0.479 1.356 0.157 0.160 0.224 6.056 51.250 21.363 2.399
200 11152 43 34651 80 0.538 0.243 0.794 0.463 0.757 -0.493 1.287 0.153 0.164 0.225 5.732 53.750 24.348 2.208
>200 45803 80 0 80 1.000 0.000 100 100 1.000

134
F. Example for calculation of Chi-square values

B1: Host rock
B2: Heat sources
B1Present B1Absent Total
B2 Present N {B1  B2  D} N {B1  B2  D} N{B2  D}
B2 Absent N {B1  B2  D} N {B1  B2  D} N {B2  D}
Evidence map B1 Present B1 Absent Total

B2 Present 38 (36.075) 1 (2.925) 39
B2 Absent 36 (37.925) 7 (3.075) 41
Total 74 6 80
4
(observed i -predicted i ) 2
χ 2 =
i=1 predicted i
Observed Predicted (O-E) [(O-E)-0.5] 2 [(O-E)-0.5]2/E

38 36.075 1.925 2.030625 0.056
1 2.925 -1.925 5.880625 2.010
36 37.925 -1.925 5.880625 0.155
5 3.075 1.925 2.030625 0.660
2.882
For other pairs of binary evidence maps are calculated same as this pairs,
summarized in Table 6.2
G. Example for calculation of posterior probability map

For detail explain about formula were shown in section 3.4.1
Total areas = 45803 km2

Total number of mineral for model deposits = 80
Total number of mineral for validation deposits = 27
N D 80
The prior probability, P{D}=   0.0017
N T  45803
So, log e O{D}  log e[ P{D}/ (1  P{D})  log e[0.0017 / (1  0.0017)]  6.348
135
Posterior Probability Map = exp (- 6.58 + NE-SW lineament + NW-SE

lineament + Host rocks + Heat source) / (1 + exp (- 6.58 + NE-SW lineament +
NW-SE lineament + Host rocks + Heat source))
This expression automatically converts the loge posterior odds calculated

according to equation 3.19 into posterior probabilities based on the relationship
between probability and odd, P = O/ (1+O).

358536

Uploaded by

358536

Uploaded by

MINISTRY OF RESEARCH, TECHNOLOGY AND HIGHER EDUCATION

UNIVERSITAS GADJAH MADA

MINERAL POTENTIAL MAPPING USING GEOGRAPHIC INFORMATION

This thesis is dedicated to my beloved parents, sister, brothers, cousins

I am deeply indebted to many people who have helped me in various ways

First of all, I would like to express my gratitude to the ‘ASEAN University

I would like to express my deepest gratitude to my Supervisor, Dr. Lucas

I am deeply grateful to Dr. Sugeng Sapto Surjono, Head of Geological

Finally, I would like to express my sincere gratitude to my parents, sister

The weight of evidence method is one of the most important data-driven

Keyword: GIS, weight of evidence method, mineral potential mapping, Bayah

APPROVAL SHEET ............................................................................................ ii

DEDICATION ...................................................................................................... iii

ACKNOWLEDGEMENT ................................................................................... iii

CONTENTS ......................................................................................................... vii

LIST OF TABLES .............................................................................................. xii

LIST OF FIGURES ........................................................................................... xiv

LIST OF ABBREVIATIONS ........................................................................... xix

CHAPTER I: INTRODUCTION ........................................................................ 1

1.1. Background ...................................................................................................... 1

1.2. Problem Formulation ....................................................................................... 2

1.3. Objectives of the Study .................................................................................... 2

1.4. Research Outcome ........................................................................................... 3

1.5. Scope of Research ............................................................................................ 3

1.5.1. The Study Area ...................................................................................... 3

1.5.2. Scope of Work ....................................................................................... 3

1.5.3. Research Limitation ............................................................................... 4

1.6. Previous Studies ............................................................................................... 4

1.7. Organization of the Thesis ............................................................................... 8

CHAPTER II: LITERATURE REVIEW......................................................... 10

2.1. Regional Geology of Western Java ................................................................ 10

2.1.1. Regional Geological Setting ................................................................ 10

2.1.2. Physiography of West Java .................................................................. 12

2.1.4. Quaternary Volcanism ......................................................................... 16

2.2. Style of Gold Deposits in West Java .............................................................. 17

2.2.1. Mineralization in Bayah Dome/Mountains.......................................... 18

2.2.2. Mineralization in Southern Mountains Range ..................................... 22

CHAPTER III: THEORETICAL BACKGROUND ....................................... 29

3.1. Gold Deposits ................................................................................................. 29

3.1.1. Hydrothermal Gold Deposit ........................................................................ 29

3.1.2. Porphyry Cu-Au Mineralization ................................................................. 30

3.1.3. Epithermal Gold ................................................................................... 32

3.1.3.2. Classification of Epithermal Deposit ................................................ 33

3.1. 3.3. Base Metal Veins ............................................................................. 38

3.1.4. Skarn Deposit ....................................................................................... 39

3.2. Remote Sensing (RS) ..................................................................................... 41

3.2.1. Landsat Images .................................................................................... 41

3.2.2. ASTER GDEM .................................................................................... 43

3.3. Software Defoliant Method ............................................................................ 43

3.4. Geographic Information System (GIS) .......................................................... 45

3.4.1. Mineral Potential Mapping using GIS ................................................. 45

3.5. Weight of Evidence Method .......................................................................... 48

3.5.1. Test Conditional Independence ........................................................... 51

1. Pairwise test of Conditional Independence ................................................ 52

3.6. Application of Weight of Evidence in Mineral Potential Mapping ............... 55

3.7. Extraction of Evidences Maps ....................................................................... 55

3.7.1. Lithology Map ..................................................................................... 55

3.7.2. Geological Structure Map .................................................................... 56

3.7.3. Hydrothermal Alteration Map ............................................................. 56

3.8. Hypothesis ...................................................................................................... 57

CHAPTER IV: RESEARCH METHODOLOGY ........................................... 58

4.1. Variable Data and Materials........................................................................... 58

4.2. Software ....................................................................................................... 58

4.3. Research Methodology................................................................................... 58

4.3.1. Research Preparation ........................................................................... 58

4.3.2. Spatial Data Processing and Analysis .................................................. 59

4.3.3. Predictive Modeling ............................................................................. 60

4.3.4. Model Validation ................................................................................. 61

CHAPTER V: SPATIAL DATASETS AND EVIDENCE MAPS ................. 64

5.1. Mineral Deposit Database .............................................................................. 64

5.2. Lithology ....................................................................................................... 65

5.3. Geological Structure....................................................................................... 68

5.4. Hydrothermal Alteration ................................................................................ 73

5.4.1. Clay Alteration ..................................................................................... 73

5.4.2. Limonitic Alteration ............................................................................ 74

5.4.3. Accuracy Assessment .......................................................................... 76