5
2. METHODOLOGY
Time and Location
This research was conducted from January to June 2013 including the data preparation, data processing, and developing Method. The processing of this data
was  being  conducted  in  the  Tsunami  and  Earthquake  Centre  Agency  for Meteorological, Climatological and Geophysical, Jakarta.
The  study  area  is  located  at  the  Bangka  Island  between  the  latitudes  of  - 1.30°  and  -3.30°  and  longitudes  of  105°  and  107°,  bordered  by  the  following
areas:  Bangka  strait  in  the  west,  Karimata  strait  in  the  east,  Natuna  Sea  in  the north  and  java  Sea  in  the  south.  The  study  area  includes  5  regency  totally  and
partially. The total area is 11.676,05 km
2
or 1.167.605 Ha. Figure 3 illustrates the location of the study area.
Materials
The data in the research came from several sources. Some information were collected by conducting interview, questionnaire and discussion with expert from
Indonesia  agency  for  Meteorological,  Climatological  and  Geophysical  BMKG and National Nuclear Energy Agency of Indonesia BATAN.
Figure 3 Location of study area
6 The basic materials which are used and considered in this research are listed
as follow:   Earthquake  Catalog  from  1973  to  2012,  source  Indonesia  Agency  for
Meteorological, Climatological and Geophysical   Microtremor Data From year 2011, source National Nuclear Energy Agency of
Indonesia   Digital Elevation Model SRTM resolution of 90m produced by NASA source
downloaded from http:www.cgiar-csi.org   Geological  map  from  the  year  1994  and  1995  At  scale  1:250.000,  source
Geological Research and Development Centre, Bandung.   Administration  boundary  map  scale  1:250.000  sources  BAPPEDA  Bangka-
Belitung and Geospatial Information Agency BIG.
Methods
Hazard is a dangerous phenomenon, substance, human activity or condition that may cause loss life, injury or other health impacts, property damage, loss of
livelihoods  and  services,  social  and  economic  disruption,  or  environmental damage Guzey, et.al, 2013 Determining  earthquake hazard areas is a procedure
for estimating the total earthquake damage effect from ground shaking and related phenomena  by  taking  into  account  the  effects  of  local  site  condition.  The
subsurface  and  topographic  condition  can  amplify  or  reduce  the  ground  shaking acceleration at a site with respect to what would be expected for feasibility site of
nuclear  power  plant  ground  at  that  location.  In  this  research,  the  determining hazard  areas  associated  with  the  damage  in  case  of  earthquake  method  was
presented for related to Nuclear Power Plant site selection. In this study, important factors  closely  related  to  earthquake  hazards  such  as  seismicity,  geology,  slope,
earthquake  intensity,  soil  distribution  and  amplification  factors  data  were compiled for spatial  database using GIS, and  ranked by relative susceptibility of
earthquake hazards.
To  produce  earthquake  hazard  map  areas  for  Bangka  Island,  the  Simple Additive  Weighting  technique  was  used.  After  the  data  was  obtained  from
multiple  sources,  the  SAW  procedure  was  followed  by  manipulating  the  data using  the  raster  calculator  function  of  Spatial  Analyst.  Figure  4  gives  a  general
overview of the study.
Data Collection
GIS analytical techniques include: overlays, distance calculation, buffering etc.  The  first  step  in  GIS  process  is  collection  data.  Data  collection  refers  to  the
process of identifying and gathering the data required for specific application. The data  in  the  research  came  from  several  sources.  Data  collection  Category  is
divided  into:  1  primary  data  obtained  from  field  observation  and  interviewing some  stakeholders  who  are  chosen  based  on  knowledge,  expertise  and  position
and  2  secondary  data  which  are  collected  from  government  institution`s documents and comprehensive literatures.
7
Figure 4. General Framework of this study
Data Preparation
In  the  preparation  phase,  all  necessary  geometric  thematic  editing  was done  on  the  original  data  sets  and  a  topology  was  created.  In  the  next  step,  all
vector  layers  were  converted  into  raster  format  and  the  spatial  datasets  were processed  in  GIS  software.  In  the  following,  the  practical  preparation  steps  for
creating criteria maps are described. In this study, important factors closely related to earthquake hazards such as seismicity, geology, slope, earthquake intensity, soil
distribution  and  amplification  factors  data  were  compiled  for  spatial  database using GIS, and ranked by relative susceptibility of earthquake hazards Hereby, six
raster datasets were prepared for the weighted overlay analysis with the following steps:
8 1.  Calculate Earthquake Intensity
Earthquake  intensity  is  a  measure  of  the  degree  of  damage  caused  by  an earthquake at a given place UNDP, 1994. It describes the effect of an earthquake
on  the  surface  of  the  earth  and  integrates  numerous  parameters  such  as  ground acceleration,  earthquake  duration  and  subsoil  condition  Munich  Re,  2000.  It
depends upon the strength of the earthquake, the distance of the location from the hypocenter and local subsoil condition. For historical event, for which there are no
instrumental  records,  but  only  information  concerning  the  damage  they  caused, the  intensity  are  either  estimated  directly  from  data,  or  taken  from  catalogues.
There  are  many  empirical  formula  concerning  the  relationship  of  intensity, magnitude  and  hypocentral  distance  have  been  proposed  by  many  investigator.
Earthquake  intensity  in  the  study  area  was  determined  using  an  official Earthquake and Tsunami center BMKG Modified Mercalli Intensity MMI scale
map.  The  MMI  scale  is  divided  into  12  continuous  categories  Wood  and Neumann  1931.  The  lower  degrees  of  the  MMI  scale  generally  deal  with  the
manner in which the earthquake was felt by people. The higher degrees are based on observed structural damage.
For  estimation  of  earthquake  intensity  in  this  study  was  used  empirical formula from the attenuation formula of Modified Mercalli intensity proposed by
Dowrick et.al 1998. Dowrick et.al developed Modified Mercalli MM intensity attenuation  relationships  from  observed  intensities  in  New  Zealand  earthquakes.
The MM intensity MMI scale measures the earthquake effects at a site in terms of the effect it has on the natural and built environment. The model calculates the
intensity using the Dowrick attenuation function and they are predict the intensity on the Modified Mercalli Intensity MMI scale as a function of the distance from
the earthquake.
The  estimated  earthquake  intensity  calculated  for  each  point  from  the earthquake parameter and described as follows below:
a The  first  step  is  to  set  the  range  area  interest  to  be  estimated  of  the  initial
calculation, in this study used the rectangular area of 2 x 2 degree per range area.
b Set calculation grid, we set the calculation grid at 0.25 degree.
c Area sources are geographical areas within which an earthquake of a given
magnitude is equally likely to occur at any time or location, d
Calculate  the  distance  between  two  locations  from  epicenter  to  certain location each of grid point using the Spherical formula:
∆ = acossinlat
1
.sinlat
2
+coslat
1
.coslat
2
.coslong
2
−long
1
.R 1
Where:       lat1    = Latitude of epicenter lat2    = Latitude of location x
long1 = Longitude of epicenter long2 = Longitude of location x
R       = E arth’s radius mean radius = 6,371km
∆       = Distance between epicenter and location x in km e
Measure the hypocenter distance  :
9
2 Where  R  is  hypocentral  distance,  h  is  epicentral  depth  and  ∆  Distance
between epicenter and location x in km. f
Calculate  estimation  of    maximum  intensity  in  location  x  using  Dowrick Empirical formula:
I = 1.41 M - 1.18 ln R - 0.0044 R + 2.18
3 Where R is hypocentral distance, and  M is magnitude we replace Ms with
M. Seismic intensity are  estimated using an  empirical  attenuation  relationship,
which  is  for  describing  the  attenuation  of  intensity  with  distance.  To  derive maximum seismic intensity every location based on seismic intensity attenuation
formula  used  the  visual  basic  program  to  calculate  the  hypocentral  distance  and intensity.  All  the  spatial  data  were  processed  using  GIS  Tools  software  used  to
plot  contour  of  seismic  intensity  at  the  area  interest.  The  analyses  used  in  this research  are  descriptive  analysis.  The  descriptive  analysis  were  using  super
imposed  method  of  maps  to  describe  the  characteristic  and  facts  of  destructive earthquake  based  on  distance  to  epicenter,  seismic  intensity  value  in  the  study
area. Earthquake intensity classification can be description in Table 1.
Table 1 Earthquake Intensity Classification FEMA,2001
MMI Description potential Damage
I, II, III, IV, V VI, VII
VIII, IX X, XI, XII
None to very light damage Moderate to Heavy damage
Heavy damage Very Heavy damage
2.  Microtremor
Microtremor is also called ambient noise. Noise is the generic term used to denoted ambient vibration of the ground and floor caused by sources such as tide,
turbulent wind, effect of wind on trees or buildings, industrial machinery, cars and trains,  human  footsteps,  oceanic  wave,  volcanic  tremor,  etc.  Microtremor
measurement  is  one  of  the  practical  methods  to  estimate  the  effect  of  ground motion  characteristics  due  to  an  earthquake.  Although  there  are  uncertain  things
geophysically, this method can be a very useful tool in identifying seismic ground motion amplification for earthquake hazard assessment, because it is very simple
and  economical  in  operation.  The  purpose  of  this  study  is  to  evaluate  the characteristics  and  usage  of  microtremors.  Those  parameters  are  predominant
period, classification of soil conditions and amplification ratio in site location.
10 a. Microtremor Measurements
Microtremor  measurements  were  carried  out  by  using  very  high  sensitive seismometers  with  servo  system.  The
Seismometer  short  period  type  DS-4A  3 components  dan  TDL  303  portable  digital  seismograph,  with  sampling  rate
frequency100 Hz, can be observing ground motion for the period ranging from 0.1
Hz to 50 Hz. More than 19 points microtremor measurements were conducted in Bangka  Island.  For  each  point  of  measurement  20  minutes  of  ambient  noise
recorded. We have conduct 2 to 3 records in each site for verifying the stability of microtremors.  Figure  5  shows  the  locations  of  microtremor  measurements
represented on administration map of Bangka Island.
b. Microtremor Processing In  processing  data  we  used  Geopsy  software,.  This  software  contains
information of recording time, the amount of data and other supporting data. The result  is  a  spectrum  at  each  point  will  then  be  analyzed  to  obtain  HVSR  peak
value A and predominat frequency fo. The data processing to obtain the HVSR at each site was performed in the following way:
1.
Fourier  transformation:  firstly,  Fourier  spectra  of  the  selected  segments  of two  horizontal  and  the  vertical  components  are  calculated  using  the  fast
Fourier  Transform  FFT  algorithm.  As  the  Fourier  spectra  of  the  two horizontal components looked alike, their horizontally combined spectra were
calculated to  obtain the maximum Fourier amplitude spectrum as a complex vector in the horizontal plane, while the one vertical component provided the
vertical motion spectra.
2. Smooting of the spectra: Secondly, digital filtering has been employed on the
combined horizontal and vertical spectra applying a logarithmic window with a  bandwidth  coefficient  equal  to  15  .  This  filtering  technique  is  applied  to
reduce the distortion of peak amplitudes.
3. Calculation of transfer functions: the smoothed combined horizontal spectrum
are  divided  with  the  vertical  spectrum  using  equation  1  given  below,  which provided  the  soil  response  in  term  of  amplified  periods  of  the  investigated
portion 20.48s of records.
R f =
√
FNS f  FEWf  FUDf                                     4 4.
Where Rf is the horizontal to vertical spectral ratio and FNSf, FEWf and FUDf is the Fourier amplitude spectra in the North South componentsNS,
East West EW and Up Down Directions respectively. 5.
Normalizing  the  data  set:  After  obtaining  the  HV  spectra  of  the  tree segments, the average of the spectra are obtained as the HV spectrum fror a
particular site as a relatively non-biased response. From this spectrum we can determine  the  value  of  the  dominant  frequency  fo  and  peak  spectral  ratio
HV A at the measurement site microtremor. Based on the relationship T = 1fo then we will get the value of a dominant period in the measurement site
Sesame, 2004
11 Based  on  the  microtremor  measurements  analysis  HVSR  determine  the
predominant frequency fo and amplification factor Am. In  addition  to  assigning  a  predominant  period  class  to  each  recording  site,
we also classify each site to the four classes as rock, hard soil, medium soil, and soft soil as defined by Molas  Yamazaki 2005. See also Table 2, which shows
the  approximately  corresponding  site  classes  defined  by  the  Japan  Road Association 1980 and the approximate correspondence with NHERP site classes
2000.
Table 2. Site Class Definition Based on Predominant period Molas et.al,2005
Description Natural Period
Rock Hard Soil
Medium Soil Soft soil
T  0.2 sec 0.2 ≤ T  0.4 sec
0.4 ≤ T  0.6 sec T
≥ 0.6 sec Soil  amplification  is  a  main  factor  influencing  the  distribution  of  damage  and
causalities  in  urban  areas  when  large  earthquake  occur.  Thus,  it  is  important  to define  how  the  seismic  waves  are  affected  by  recent  and  non
–consolidated geological  deposits  to  better  quantify  the  ground  motions  at  surface.  The  higher
amplification factors of a soil type under a certain frequency of seismic wave, the higher  the  degree  of  hazard  for  the  structures  of  that  frequency.  Therefore,  the
ranks of the seismic hazard have been differentiated into four classes based on the amplification factors which are given in Table 3.
Figure 5  Location of Microtremor measurements
12 Table 3  Ranking of ground shaking hazard based on amplification factors
Kamal,et.al,2006 Amplifications
Ranks 1.0
– 2.5 2.5
– 3.5 3.5
– 4.5 4.5 - 7
Very low hazard Low hazard
Moderate hazard Relative High Hazard
3.  Slope
Slope is defined by a plane tangent to a topographic surface, as modeled by the  DEM  at  a  point  Burrough,  1986.  Slope  presents  the  percent  change  in
elevation over a certain distance. The output slope can be calculated as either the percent or degree slope. In this study, percent of slope was chosen.
Slope  is  an  important  factor  while  considering  the  ease  of  engineering construction  and  susceptibility  land  sliding  cause  by  earthquake.  Slopes  are
particularly  vulnerable  to  bedrock  failures.  Keefer1984,1993  noted  that  more than  90  percent  of  earthquake  induced  slope  failures  on  rock  slopes  were  rock
falls  and  rock  slides.  The  physical  characteristics  of  the  rock  masses  underlying steep  slopes  are  of  fundamental  importance  in  evaluating  their  susceptibility  to
slope  failure.  Therefore,  the  slope  layer  will  only  contribute  to  determine  the earthquake  hazard  areas  in  ease  of  engineering  constructions,  since  steep  slopes
interfere with excavation processes. The slope map was prepared in degrees using DEM  of  the  study  area.  Afterwards,  the  slope  values  were  subdivided  into  four
main  classes  according  AGS  Sub-Commite  Australian  Geomechanics  2002 Slopes between 0 and 9 were assigned as the flat class, slopes between 10 and
15  were assigned as moderately steep class, slopes between 16 and 20 were assigned  as  very  steep  class  and  slopes  greater  than  20  were  assigned  as  the
extremely steep class. Table 4 shows the degrees and description of slope class.
Table 4 Slope Classification Andrian, 2009 Percent of Slope
Information Hazard Categories
– 9 10
– 15 16
– 20 20
Flat Moderate Steep
Very Steep Extremely Step
Very Low Moderate
High Very High
4.  Fault Distance A fault is a break in the rocks that make up the Earth’s crust, along which
rocks on either side have moved past each other. Fault have an important role in determine  earthquake  hazard  areas,  because  the  faults  have  a  main  effect  to
movement of earth layers after earthquake.
13 Table 5 Classification fault based on distance Mohsen et.al., 2012
Distance Earthquake hazard
– 30 km 30
– 50 km 50 km
Very Damaging effect Damaging effect of moderate
Negative effect of poor Determining  earthquake  prone  areas  of  useful  measures  to  reduce  the
severity  of  the  damages  is  considered,  it  is  hereby  be  limited  use  of  high  risk areas. Construction of some buildings for nuclear plant in these zones can prevent
by identifying risk zones in cities in low risk areas of vital arteries decided. Fault as  seismic  sources,  including  plate  movement  and  withdraw  are  major  factors,
thus,  away  from  the  faults  could  be  considered  as  one  of  main  parameters  of determine  earthquake  hazard  areas.  Classification  fault  for  significant  level  of
earthquake hazard based on distance from the fault defined by Mohsen et.al,2012 shown in Table 5.
5.  Lithology
Lithology is the description of rock composition and texture. The geological formations type and condition are closely related to the landslide after earthquake
occurrence. Lithology data was extracted from geological map. This data had been prepared  for  common  geological  functions  without  take  into  account  the  special
needs  of  the  earthquake  hazard  evaluation.  The  geological  units  were  regrouped based on lithological attributes rather than their stratigraphic content and age.
Geology  units  strongly  influences  slope  stability  and  it  is  clear  that  there exists  an  associated  between  types  of  lithology  material.    However,  this
association  may  be  strong  or  weak  largely  depending  upon  the  type  of  lithology material.  Lithology  is  one  main  factors  influencing  the  type  and  the  intensity  of
the  earthquake  ground  shaking.  Based  on  the  geological  map  scale  1:  250,000 published by Geological Research and Development Centre of Indonesia, various
rock formations in the study area have been grouped to prepared the lithology data layer. The lithology classification for determine earthquake hazard areas is shown
in table 6
Table 6 Classification of Lithology Bealand, 1996 Lithology Age
Hazard Level Quarternary  2.5 Myrs
Tertiary 2.5 ≤ Age  ≥ 75 Myrs Basement Rock  75 Myrs
Least Hazardous Quite hazardous
Not very hazardous
Simple Additive Weighting SAW
After the data was prepped and ready to analyze, it was time to implement the steps of the SAW method in Spatial Analyst. This process was almost entirely
carried out using the raster calculator. Once the layers were reclassified, the next step  was  to  standardize  the  values.  In  this  method,  all  of  the  layers  are
concurrently  considered  in  assigning  weight  values, and all classes of each layer
14 are  also  concurrently  considered  while  assigning  rank  values.  As  a  result,  six
weight values were assigned to the six layers. Simple  Additive  Weighting  which  is  also  known  as  weighted  linear
combination or scoring  methods is  simple and most often used in  multi attribute decision technique. The method is based on the weighted average. An evaluation
score is calculated for each alternative by multiplying the scaled value given to the alternative  of  that  attribute  with  the  weights  of  relative  importance  directly
assigned by decision maker followed by summing of the products for all criteria.
Simple  Additive  Weighting  SAW  is  calculated  using  the  following formula:
A
i
=
Ʃ
W
j
X
ij
5 Where X
ij
is the score of the ith alternative with respect to the jth attribute and  W
j
is  the  normalized  weight.  The  weights  represent  the  relative  importance  of  the attributes. The most preferred alternatives is selected by identifying the maximum
value of A
i
, i = 1, ….,m. The  SAW  methods  can  be  implemented  using  GIS  having  overlay
capabilities.  The  GIS  based  Simple  Additive  Weighting  method  involves  the following steps Malczewski, 1999:
1. Definition of the set of evaluation criteria map layers and the set of feasible
alternatives,, 2.
Standardization of each criterion map layer, 3.
Definition of the criterion weights; that is, a weight of  “relative importance” is directly assigned to each criterion map; that is, multiply standardized map
layers by corresponding weights, 4.
Construction of the weighted standardized map layers, 5.
Generation  of  the  overall  score  for  each  alternative  using  the  overlay operation on the weighted standardized map layers,
6. Ranking  of  the  alternatives  according  to  the  overall  performance;  the
alternative with highest score is the best alternative. Simple Additive Weighting method is the most often used method in multi
attribute decision rules. The method can be operationalized using any GIS system having overlay capabilities. The overlay techniques allow the evaluation criterion
map layers input maps to be aggregated in order to determine the composite map layer output map.
It  should  be  emphasized  that,  there  are  two  strong  assumptions  implicit  in the SAW method; the linearity and additively attributes. The linearity assumption
means that the desirability of an additional unit of an attribute is constant for any level  of  that  attribute.  In  many  spatial  decision  situations  these  two  assumptions
are  very  difficult  to  apply.  Because  of  the  complementarities  between  different attributes, the SAW method may lead to false results.
15
3. RESULTS