IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Vulnerability Assessment of Tsunami-Affected Inundated Area Using

Geospatial Analysis Based Tsunami Run-Up Simulation
To cite this article: Rina Febrina et al 2021 IOP Conf. Ser.: Mater. Sci. Eng. 1062 012036

View the article online for updates and enhancements.

This content was downloaded from IP address 110.137.37.100 on 21/02/2021 at 17:30

International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Vulnerability Assessment of Tsunami-Affected Inundated Area

Using Geospatial Analysis Based Tsunami Run-Up Simulation

Rina Febrina1*, Lusmelia Afriani2, Rr M I Retno Susilorini3 and Tri Prastio1

1
Departement of Civil Engineering, Universitas Malahayati, Bandar Lampung 35153,
Indonesia
2
Departement of Civil Engineering, Universitas Lampung, Bandar Lampung 35153,
Indonesia
3
Departement of Infrastucture and Environmental Engineering, Soegijapranata
Catholic University, Semarang 5023, Indonesia

*Corresponding e-mail: [email protected]

Abstract. A tsunami can cause considerable loss of life, extreme property damage and
infrastructure destruction. A tsunami run-up simulation can be used to forecast wave generation
caused by an earthquake. This study aims to assess the tsunami-affected inundated area on the
west coast of Lampung province using geospatial analysis-based tsunami run-up simulation. The
parameters used in the assessment were applied elevation and slope computed from Aster
GDEM, coastline distance determined from a vector map and tsunami inundation obtained by
Hills and Mader equation and vegetation density calculated from ALOS ANVIR-2 image. The
AHP process was applied for weighing the parameter through a pair-wise comparison using four
iterations of the normalised matrix. Five classes of vulnerability were defined and analysed for
the tsunami inundation mapping. Weighted overlay through spatial analyst in GIS was used to
draw the final map of tsunami vulnerability. The assessment result indicates that 11.26 square
kilometre of the study area falls under the high tsunami vulnerability zone and 6.6 square
kilometres of the building area comes under the inundation area. The results presented in this
study may facilitate improvements in basic data for city planning related to evacuation processes
and management strategies during disasters.

Keywords: tsunami, property damage, infrastructure, vulnerability

1. Introduction
Tsunami based disasters are natural, highly destructive and unavoidable phenomena. Vulnerability
mapping assessments of a tsunami are observed as an effective way of reducing the impact of the
disasters and have been successfully applied for disaster mitigation measures [1-4]. The assessments
include the mapping of evacuation covering routes and buildings, the construction of tsunami barriers,
disaster risk management and regulation and disaster education for the public [5]. Several pieces of
formative information such as reliable and accurate spatial and temporal characteristics and impact of a
tsunami with damage potential at different magnitude scales are required for the assessments. The
availability of the information has created more parameters which are vitally important in developing a
realistic description of vulnerability for a spatial and temporal assessment. The probability of risk,
exposure, capacity measurement, and vulnerability are parameters used to determine a disaster risk [6-
8]. However, it was shown that external pressure imperilled a person, community, or a system.
Vulnerability is explained as a process and condition refers to that produced by economic,

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

environmental, physical, and social factors. Moreover, vulnerability enhanced the susceptibility of
hazards impact community [9,10].
Mapping the distribution of damage area because of disasters and assessing the vulnerability area via
remote sensing and GIS are seen to be a potentially effective way of mitigating the disastrous impact of
a tsunami. Geo-reference data were used to make a decision, while GIS analysis was employed for the
assessment through spatial multi-criteria. On top of that, geographical references and criterion attributes
were needed [11,12]. Previous studies used the GIS method for tsunami and an overlay of official land-
use mapping [13]. In addition, while GIS can be used to analyse the risk of tsunami, remote sensing data
and analysis of the physical built-up of infrastructures, such as building, contour of inundation area, and
highest level of tsunami can be used to predict vulnerability [5],[14].
The aims of this study are to map tsunami vulnerability areas on the west coast of Lampung Province
and to evaluate the tsunami impact by mapping possible inundation areas using a geospatial analysis-
based tsunami run-up simulation. Vulnerability is calculated based on the run-up, elevation, slope,
coastline distance and vegetation density using tsunami run-up simulation generated by software L-
2008. Furthermore, the Soil Adjusted Vegetation Index (SAVI) is applied. The vegetation mapping is
used as a Normalised Difference Vegetation Index (NDVI). In addition, the analytical hierarchy process
is used to estimate the weights of the parameters.

2. Research Methodology

2.1. Data
The ASTER Global Digital Elevation Model (ASTER GDEM) version 2 provided the elevation data.
ASTER GDEM was used as a platform for data collection via a space-borne observing optical device,
while the 2nd version of ASTER GDEM was used to enhance the resolution of global digital elevation
and the accuracy of elevation. In addition, ASTER GDEM was also used in advancing algorithm and
reprocessing a total of 1.5 million scene data, including 250,000 additional scenes acquired after the
previous release. The data were posted approximately 30 m at the equator. Subsequently, the ALOS
ANVIR-2 image was analysed for the vegetation density mapping. Moreover, run-up information was
used based on the tsunami run-up simulation (L-2008). The protocols of analysis were started from the
data collection itself. It proceeded with the surface analysis of DEM data, vector data processing, and
inundation processing which continued with the AHP process and finally the raster overlay processing
through the GIS approach as in Figure 1.

2.2. Study area

The study was conducted on the west coast of Lampung Province as in Figure 2, non the island of
Sumatra. It is an area which is highly exposed to a wide range of potentially devastating impacts from
earthquake and tsunami threats.

2.3. Simulation of tsunami

2.3.1. Source modelling

The initial profile of a tsunami centre can be obtained by considering it as a seabed deformation with
fault parameters, determined on the mechanism solution data of an earthquake centre. The parameter
setting for the source modelling is shown in Figure 3. A11 (km) and A12 (km) refer to the fault length,
and Aw1 is the fault width. The positions x_eq fault, y_eq, and z_eq are defined as the longitude, latitude
and depth of the earthquake, respectively. Furthermore, the source modelling was conducted using the
L-2008 software. The output of the source modelling is the vertical displacement that occurs in the
seabed. Subsequently, this deformation was used in the ocean and run-up modelling.

2
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Figure 1. General framework of study

3
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Figure 2. Map of study area

Figure 3. Parameter setting for source modelling

4
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

2.3.2. Ocean modelling

Output data of source modelling and bathymetric data were needed for ocean modelling. In addition, the
appropriate bathymetric data with a study area was needed to get a good simulation result. In this ocean
modelling, the parameter simulation as an input software tsunami L-2008 was maximum time.
Maximum time was needed to vine the simulation of 3600s. Subsequently, the data save internal of 25s
was added and resulted in 1 output per 25s.

2.3.3. Run-up Modelling

In this study, the height of the tsunami modelling (run-up modelling) was demonstrated by digitising
the predetermined area. Digitising was conducted at various points along the coastline of the tsunami-
affected area. On observation, every affected area recorded the highest run-up value.

2.4. Spatial analysis

2.4.1. Elevation classification

Elevation data are required to calculate vulnerability and tsunami inundation. A digital elevation model
was created using an elevation map. The data from GDEM which were downloaded from DEMNAS
website were used to derive a set of physical vulnerability parameters. The elevation was then classified
into five vulnerability categories as in Figure 4 based on the height of the surface. A low elevation for
example indicated a high tsunami vulnerability wave as in Table 1.

Table 1. Vulnerability class based on elevation, slope and coastline distance [15], [16]
Elevation (m) Slope (%) Coastline distance (m) Vulnerability class
<5 0–2 0-545.65 High
5 - 10 2–6 545.65 - 1261.80 Slightly high
10 - 15 6 – 13 1261.80 - 2187.40 Medium
15 - 20 13 – 20 2187.40 - 3154.65 Slightly low
>20 >20 >3154.65 Low

5
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Figure 4. Tsunami vulnerability map based on parameters of

elevation

2.4.2. Slope
In this study, a slope is identified based on the maximum changing of the z-value rate from each cell of
the image. The third-order finite difference method and GDEM were applied to create a slope map. The
range of the slope percentage was identified as zero to almost infinity. The slope map was categorised
into five categories according to the value of tsunami vulnerability, as described in Table 1. The
vulnerability map shown in Figure. 5 was made by applying slope classification, which is based on the
tsunami vulnerability class.

2.4.3. Coastline distance

Coastal areas were divided into five classes of vulnerability using a multi-buffering calculation.
Buffering distance was determined using the possibility range of tsunami when reaching the ground.
The maximum tsunami height which was obtained from historical reports was used to calculate the
buffering distance below [13],[17]:

4 𝑌0
𝑙𝑜𝑔𝑋𝑚𝑎𝑥 = log 100 + 𝑙𝑜𝑔 ( ) (1)
3 10

where 𝑋𝑚𝑎𝑥 is the maximum reach of the tsunami over land, and 𝑌𝑜 is the tsunami height at the coast.
Besides, the buffering distance was determined based on the range of possible tsunami flows flowing

6
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

toward the inland. The score of the 𝑋𝑚𝑎𝑥 in this study was based on the simulation of a tsunami using
software L-2008. Coastline distances in the vector map of Lampung through GIS processing were used
to generate a vulnerability map based on coastline distance.

Figure 5. Tsunami vulnerability map based on parameters of slope

7
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Figure 6. Tsunami vulnerability map based on the parameters of

coastline distance

2.4.4. Inundation
The simulation of a tsunami was demonstrated using ArcGIS 10.6 software with the Raster calculator
functions [21] as shown below:

𝑋𝑚𝑎𝑥 = (𝐻𝑠)1,33 𝑛−2 𝑘 𝐶𝑜𝑠 𝑆 (2)

where 𝑋𝑚𝑎𝑥 is the maximum inundation (m), Hs is the maximum tsunami run-up (m), 𝑘is constant
(0.06), manning coefficient of the land is 𝑛, the slope of the land is 𝑆, and 𝑋𝑚𝑎𝑥 in the vector map of
Lampung through GIS processing to generate a vulnerability map based on inundation.

2.5. ALOS AVNIR-2 Image Processing

SAVI was used for mapping vegetation density. Briefly, the digital numbers of ALOS AVNIR-2 were
converted to reflectance values before generating the synthetic NDVI and SAVI images [19,20]. A
vegetation density map was created as below:

2.5.1. Reflectance Calculation

A radiance conversion from digital number values is required before calculating reflectance. Calculating
radiance is a fundamental step in relating image data from multiple sensors and platforms into a common

8
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

radiometric scale [21]. The general equation for DN to radiance conversion is demonstrated in equation
(3) [22].

𝐿𝜆 = 𝐺𝑟𝑒𝑠𝑐𝑎𝑙𝑒 × 𝑄𝐶𝐴𝐿 + 𝐵𝑟𝑒𝑠𝑐𝑎𝑙𝑒 (3)

where Lλ is spectral radiance at the sensor’s aperture (W/m2/sr/µm), Grescale is rescaled gain, 𝑄𝐶𝐴𝐿
is Digital Number (𝐷𝑁), and Brescale is rescaled bias [23]. Conversion of reflectance is obtained using
Equation (4),
𝜋 × 𝐿𝜆 × 𝑑2
𝜌𝜆 = (4)
𝐸𝑆𝑈𝑁𝜆 × 𝑐𝑜𝑠𝜃𝑠

where ρλ is unitless planetary reflectance and 𝐿λ is spectral radiance at the sensor’s aperture.
Meanwhile, 𝑑2 is the earth-sun distance in astronomical units from a nautical handbook, 𝐸𝑆𝑈𝑁λ is mean
solar Exoatmospheric irradiances, and θs is solar zenith angle in degrees.

2.5.2. NDVI Calculation

The concept of NDVI is about the first channel causing considerable absorption of incoming sunlight
and the second channel creating considerable reflectance [24-26]. The differences in wavelength are
calculated using NDVI. The range of wavelength used is between -1 and 1, in which 0.5 indicates dense
vegetation and < 0 indicates no vegetation, including water [27]. NDVI is calculated as

(𝑁𝐼𝑅 − 𝑉𝐼𝑆)
𝑁𝐷𝑉𝐼 = (5)
(𝑁𝐼𝑅 + 𝑉𝐼𝑆)

where 𝑁𝐼𝑅 is Near Infra-Red band, and 𝑉𝐼𝑆is a visible band of the red band. ALOS AVNIR-2 band
3 is red, and band 4 is 𝑁𝐼𝑅.

2.5.3. SAVI Calculation

SAVI can be defined as an algorithm developed to generate vegetation indexes by eliminating soil factor
[28]. A SAVI based model is capable of characterising the forested and non-forested area [20].
Therefore, the model is created to reduce soil brightness, such as shadow [29]. NDVI is calculated as
[30],

(𝑁𝐼𝑅 − 𝑅)
𝑆𝐴𝑉𝐼 = 𝑥(1 + 𝐿) (6)
(𝑁𝐼𝑅 + 𝑅 + 𝐿)

where 𝑁𝐼𝑅 is Near Infra-red band, 𝑅 is red band, and 𝐿 is the soil calibration factor, in which L is
equal to 0.5 [31]. The secondary backscattering effect of canopy transmitted is reduced using a
correction factor (𝐿).𝐿value of 1 is optimal in semiarid environments. The SAVI generated map is shown
in Figure 7(a).

9
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

(a) (b)

(c)
Figure 7. (a) SAVI Map; (b) Vegetation density; and (c) Vulnerability map based on vegetation density

3. Results and discussion

3.1. Tsunami run-up simulation

A scenario simulation was demonstrated in the central earthquake location on the ocean plate of the
West coast of Lampung. Tsunami modelling parameters which are length, width, maximum and average
displacement were created for measuring time, amplitude, and wave height of the tsunami in one area.
The input process of the tsunami model was generated using the bathymetric input data of the earthquake
location on the Western region of Sumatera island, and topography data required were used in the ASCII

10
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

XYZ format. The results of the input parameters decision for source modelling are shown in Table 2.
Subsequently, the mechanism data of earthquake were prepared by deciding on the design modelling,
while the grid creation of the bathymetric value, simulation design of tsunami and the wave height of
tsunami modelling were then determined. The initial profile of a tsunami centre can be calculated by
considering the profile as a seabed deformation with fault parameters which can be determined by using
the mechanism solution data of the earthquake centre. Deformations of the seabed as outputs of centre
modelling were assessed using tsunami L-2008. Furthermore, it was applied in the ocean and run up
modelling. The vertical displacement (deformation of the seabed) as an output of source modelling was
needed in the ocean modelling using L-2008 application and bathymetric data. The bathymetric
coordinate boundaries of this study were between 4°51'23.43"S to 6°43'50.50"S and 102°3'3.30"E to
105° 2'29.63"E. Moreover, the level of the earthquake was 7.8 SR, and the distance of the coast to the
epicentral was 241 km. In this study, the input parameter was observed at 739.66 km x 82.41 km of
length and width fault, resulting in a fault area and slip of 60955.38 km2 and 11.036 m respectively.

Table 2. Earthquake input parameters for source modelling

Parameter Research Model
Al1
369.83 369.83
Length (km)
Al2
369.83 369.83
Aw1
82.41 82.41
Width (km)
Aw2 0 0
Depth (km) 28 28
Strike (Degree) 312 58
Dip (Degree) 223 64
Slip (m) 11.036 11.036
Mw (Richter) 7.8 7.8
Lat. -7.082527° -7.082527°
Centre Fault Coordinate
Long. 103.307943° 103.307943°
Type Fault Normal Fault Normal Fault

The result of the model design was vertical displacement in Figure 8, shown as deformation in the
seabed. The movement of deformation which occurred in the seabed was followed by a shift in the sea
level. The source modelling results illustrated variation in the maximum vertical displacement and event
and a minimum vertical displacement value of 3.35 m and -3.35 m, respectively. The maximum vertical
value showed an increase in the seabed vertical movement, but the minimum vertical displacement
showed a reduction in the seabed vertical movement. The rapid vertical movement of the seabed was
detected as a response to the earthquake and had caused the rise and fall of the sea level. The ocean
modelling results showed the propagation of tsunami waves in all directions. In this study, the timing
and height that were taken for the propagation of a tsunami wave from the epicentre to the beach (target
area) were 28.56 min and 2.65 m, respectively shown in the Figure 9. In this study, the target location
was Krui on the west coast Lampung. The run-up modelling result from the calculation fault of the
length and width is shown in Figure 10. The run-up score was affected by the distance of the epicentre
to the affected area, bathymetric, the morphology of the beach, the slope of the beach, and the wave
height of the tsunami [32]. Enhancement of the height of the tsunami wave was comparable with the
shallowness of the approaching land.

11
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Figure 8. Vertical displacement

(a) (b)

12
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

(c)
Figure 9. (a) Cross-section of plot A-B (241 km); (b) Cross-Section of ocean bottom A-B; (c) Cross-
section of maximum tsunami height A-B

This phenomenon was followed by refraction, reflection, and diffraction and had brought about a
higher tsunami wave. In this study, the height of the run-up and inundation were analysed. The length
of the inundation affected the tsunami wave which then affected the population.

Figure 10. Tsunami height based on calculation

3.2. AHP processing

This study classified all the parameters of tsunami vulnerability into five classes: Number 1, 2, 3, 4, and
5 in which representing low, slightly low, medium, slightly high, and high vulnerability, respectively.
All the parameters were assessed using a weighting and scoring system in order to determine the limiting
factors of each parameter. The weighting system was also used to determine the different classes of

13
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

tsunami vulnerability. The AHP was applied to assess the weight of each parameter using pair-wise
comparisons. The AHP method created a scaled set of preferences and described the essential parameters
in relation to other parameters [33-37]. In addition, assessed the natural hazards of vulnerability. The
sequencing number of 1 to 9 as tabulated in Table 3 was used as a comparison scale to create a pair-
wise comparison matrix. The number depended on the relative weight of each parameter and the relative
weight was produced by pair-wise comparison in Figure 11.

Table 3: Nine-point comparison scale by Saaty [38]

Intensity of importance Definition
1 Equal importance
3 Weak importance of one over another
5 Essential or strong importance
7 Demonstrated importance
9 Absolute importance
2,4,6,8 Intermediate value between the two adjacent judgements

Prior to the spatial analysis, a consistency level computation was required. Consistency Ratio (CR))
is a procedure that determines the index of consistency. It shows that the probability of the matrix
findings was randomly calculated [35]. In other words, AHP allows inconsistency by using the
consistency ratio computation. The tolerable consistency ratio is less than or equal to 10 per cent [39].
𝐶𝑅 is the ratio between the consistency index (𝐶𝐼) and random consistency index (𝑅𝐼), and can be
expressed using equation (7)

𝐶𝐼 (𝜆 𝑚𝑎𝑥 − 𝑁)
𝐶𝑅 = 𝑎𝑛𝑑 𝐶𝐼 = (7)
𝑅𝐼′ (𝑁 − 1)

where 𝜆𝑚𝑎𝑥 represents the largest eigenvalue, and 𝑁 is the size of the comparison matrix. In this
study 𝑁 = 5. 𝜆𝑚𝑎𝑥 was calculated from the sum of all the parameters, and then multiplied by its
eigenvector. 𝑅𝐼 is based on the random consistency index (𝑅𝐼 = 1.12 for five parameters). Normalised
matrix (Fig. 10) was calculated from the pair-wise comparison in four iterations, and showed that 𝐶𝐼
was 0.252, while 𝐶𝑅 was 22.5%. The pair-wise comparison revealed that run-up was the most important
factor (50.18%) followed by elevation (19.52%), slope (13.10%), coastline distance (9.53%) and
vegetation density (7.67%). 𝐴𝐻𝑃 results showed that those classes were the most important parameters
in deciding vulnerability. Spatial multi-criteria were analysed using the 𝐴𝐻𝑃 method from which the
vulnerability map was created by using weight factors for each parameter. It was found that the run-up
parameter showed the highest weight. 0.99% as the data provided in table 4. The area was identified as
being slightly high while a high class of tsunami vulnerability was found in the coastal area which has
a lower elevation as in Figure 12(b).

14
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

Pair-wise comparison matrix

𝐶𝑜𝑎𝑠𝑡𝑖𝑎𝑙 𝑃𝑟𝑜𝑥𝑖𝑚𝑖𝑡𝑦

𝑉𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛
𝐸𝑙𝑒𝑣𝑎𝑡𝑖𝑜𝑛

𝑅𝑢𝑛𝑑 𝑈𝑝
𝑆𝑙𝑜𝑝𝑒
𝐸𝑙𝑒𝑣𝑎𝑡𝑖𝑜𝑛 1.00 2.00 3.00 0.33 3.00
𝑆𝑙𝑜𝑝𝑒 0.50 1.00 2.00 0.25 3.00
𝐶𝑜𝑎𝑠𝑡𝑖𝑎𝑙 𝑃𝑟𝑜𝑥𝑖𝑚𝑖𝑡𝑦 0.33 0.50 1.00 0.20 3.00
𝑅𝑢𝑛𝑑 𝑈𝑝 3.03 4.00 5.00 1.00 5.00
𝑉𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛 [ 0.33 0.33 0.33 0.20 1.00]

Normalized Matrix
𝐶𝑜𝑎𝑠𝑡𝑖𝑎𝑙 𝑃𝑟𝑜𝑥𝑖𝑚𝑖𝑡𝑦

𝑉𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛
𝐸𝑙𝑒𝑣𝑎𝑡𝑖𝑜𝑛

𝑅𝑢𝑛𝑑 𝑈𝑝
𝑆𝑙𝑜𝑝𝑒

𝐸𝑙𝑒𝑣𝑎𝑡𝑖𝑜𝑛 0.1924 0.2553 0.2647 0.1667 0.2000

𝑆𝑙𝑜𝑝𝑒 0.0962 0.1277 0.1765 0.1263 0.2000
𝐶𝑜𝑎𝑠𝑡𝑖𝑎𝑙 𝑃𝑟𝑜𝑥𝑖𝑚𝑖𝑡𝑦 0.0641 0.0638 0.0882 0.1010 0.2000
𝑅𝑢𝑛𝑑 𝑈𝑝 0.5831 0.5106 0.4412 0.5051 0.3333
𝑉𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛 [0.0641 0.0462 0.0294 0.1010 0.0667]

4th iteration
0.1952 0.1952 0.1952 0.1952 0.1952 19.52%

0.1310 0.1310 0.1310 0.1310 0.1310 13.10%

0.0953 0.0953 0.0953 0.0953 0.0953 9.53%

0.5018 0.5018 0.5018 0.5018 0.5018 50.18%

[0.0767 0.0767 0.0767 0.0767 0.0767] [ 7.67% ]

Figure 11. Pair-wise comparison and normalised matrix

3.3. Spatial Analysis for Vulnerability Mapping

The weighted overlay was applied using spatial analysis in GIS after creating a parameter map. In
addition, the weighted overlay tool was used to solve multi-criteria problems. The weighted overlay
method is a protocol that applies a standard scale of values for various parameters. The overall steps of

15
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

this process are shown in the figure 12(a). The level of vulnerability starts from 1 (low vulnerability) to
5 (high vulnerability). Furthermore, the statistic of vulnerability map was calculated. The vulnerability
classes as in table 4 were created using the subtraction of maximum and minimum values. Moreover,
the number of the class was also used to divide the vulnerability classes. Finally, tsunami vulnerability
was mapped into five classes as show by figure 12(b). Results showed that the high vulnerability areas
were mostly found in the coastal area with the sloping coast type.

Table 4. Vulnerability classes

Vulnerability Class Area (square kilometre) Percentage (%)
High (5) 11.26 0.42%
Slightly High (4) 15.28 0.57%
Medium (3) 266.18 9.93%
Slightly Low (2) 112.218 4.19%
Low (1) 2276.06 84.90%

(a)

16
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

(b)
Figure 12. (a) Spatial analysis step (b) Tsunami vulnerability map of west coast Lampung

3.4. Tsunami inundation and impact assessment

The simulation of tsunami wave results found that the maximum wave height was 2.65 m, inundation
area was 26.54 km2, and vulnerability level was 11.26 km2. While the building area was estimated at
approximately 6.6 km2 along the West coast of Lampung as in Table 4 and Figure 12, the inundation
areas were predicted as high vulnerability and slightly high vulnerability. Meanwhile, a significantly
different result of the inundation area (22.09 km2) was recorded from the previous study [40].
Enhancement of the inundation area in 2019 was 4.45 km2 compared with the BNPB result in 2016.
This might be due to the changing vegetation around the area as vegetation plays an essential role as a
tsunami barrier that reduces the impact of tsunami destruction [41-44]. In addition, SAVI values of West
coastal Lampung showed the state of vegetation of the inundation area as follows: 6.02 km2 (22.68%)
of low vegetation, 11.85 km2 (44.65%) of slightly low vegetation and 8.67 km2 (32.67%) of medium
vegetation as in Figure 7 and 12.

Table 5. Vulnerability class of building area

Inundation area
Vulnerability class 3 2 1
5 4
Building area (km2) 0.91 5.69 3.26 2.69 3.88

17
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

(a) (b)
Figure 13. (a) Map of possible inundation area in west coast Lampung (b) Tsunami vulnerability
class of building area in west coast Lampung

4. Conclusion
The identification of tsunami vulnerability areas on the west coast of Lampung Province and evaluation
of the tsunami impact by mapping possible inundation areas using a geospatial analysis-based tsunami
run-up simulation was studied. The simulation was carried out using the L-2008 software. It resulted in
a tsunami wave of 2.65 m in height at a timing of 28.56 min when it reached the coast area. There were
five parameters that were used to create a tsunami vulnerability map and predict the possibility of an
inundation area along the west coast of Lampung, including the run-up, elevation, slope, coastal line,
and vegetation. In addition, the analytical hierarchy process (AHP) was used to estimate the weight of
the parameters of AHP was calculated using a pair-wise comparison matrix. The results showed 𝐶𝐼 and
CR values of 0.252 and 22.5%, respectively. The pair-wise comparison revealed that run-up was the
most important factor (50.18%) which affected vulnerability and inundation area, followed by elevation
(19.52%), slope (13.10%), coastline distance (9.53%) and vegetation density (7.67%). The final map of
tsunami vulnerability was constructed using the weighted overlay through a spatial analysis in GIS.
Furthermore, the classes of vulnerability were identified into five classes with a series of numbers from
1 (low vulnerability) up to 5 (high vulnerability), and the tsunami inundation mapping was analysed.
The assessment result indicates that 11.26 square kilometres of the study area falls under the high
vulnerability zone due to tsunami and 6.6 square kilometres of the building area comes under the
inundation area. Subsequently, our study has shown that the total number of inundation area along the
west coast of Lampung was 26.54 km2. Therefore, the result of this study can be used as a reference by
the local government to facilitate improvements in basic data for city planning related to evacuation
processes and management strategies during disasters.

References
[1] Birkmann J 2006 Measuring Vulnerability to Hazards of Nature Origin (Tokyo: UNU press) p 1
[2] Birkimann J and Wisner B 2005 Measuring the Un-Measureable (Germany:UNU Press) p 40
[3] Papathoma M and Dominey-Howes D 2003 Nat. Hazard Earth Syst. Sci.3(6) 733-747

18
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

[4] Papathoma M, Dominey H, Smith D T M and Zong Y 2003 Nat. Hazard Earth Syst. Sci.3(5) 377
[5] Papathoma M and Dominey-Howes D 2003 Nat. Hazard Earth Syst. Sci. 3(6) 734
[6] United Nations Disaster Relief Coordinator (UNDRO) 1991 Mitigation Natural Disaster:
Phenomena, effects and Options, a Manual for Policy Makers and Planners (United Nations
Publication, New York)
[7] Pelling M 2003 The Vulnerability of cities: Natural Disaster and Social Resilience (London:
Earthscan)
[8] Bollin C, Cardenas C, Hahn H and Vasta K S 2003 Natural Disaster Network: Disastere Risk
Management by Communities and Local Goverment (Washington DC: Inter American
Development Bank)
[9] Schmidt-Thome P 2006 Natural and Technological Hazards and Risks Affecting the spatial
Development of European Regions (Finland:Special Paper 42 Geological Survey)
[10] Mahendra R S, Mohanty P C, Bisoyi H, Srinivasa K T and Nayak S 2011 Ocean Coast. Manag.
54(4) 302
[11] Caver S J 1991 Int. J. Geo. Info. Syst. 5(3) 321
[12] Jankowski P 1995 Int. J. Geo. Info. Syst.9(3) 251
[13] Sinaga T P T, Adhi N, Lee Y W and Yongcheol S 2011 KSCE J. Civ. Eng. 15(3) 537
[14] Strunz G, Post J, Zosseder K, Wegsceider S, Muck M, Riedlinger T, Mehl H, Dech S, Birkmann,
Gebert N, Harjono H, Anwar H Z, Sumaryono, Khomarudin R M, and Muhari A 2011 Nat.
Hazards Earth Syst. Sci. 11(1) 67
[15] Lida K 1963 Magnitude, Energy and Generation Mechanisms of Tsunami and a Catalogue of
Eartquakes Associated with Tsunamis” (Honolulu:Proceeding of Tsunami Meeting at the 10th
Pacific Science Congres) pp 7-18
[16] Van Zuidam R A 1983 Guide to Geomorphological Aerial Photograpic Interpretation and
Mapping (Enschede:International Institute for Geo Information Science and Eart Observation)
[17] Bretschneider C.L and Wybro P.G 1976 Coast. Eng. Proc. 1(5) 59
[18] Hills J G and Mader C L 2006 Ann. N. Y. Acad. Sci. 822(1) 381
[19] Huete A R 1988 Rem. Sens. Environ. 25(3) 295
[20] Araujo L S, Dos Santos J R and Shimabukuro Y E 2000 Int. Arch. Photogramm. Remote Sens.
Spat. Inf. Sci. 33 77
[21] Chander G and Markham B 2003 IEEE Trasns. Geosci. Remote Sens.41(11) 2674
[22] Bouvet M, Chander G, Goryl P and Sater R 2007 IEEE Int. Geosci. Remote Sens. Symp. 2673
[23] Data Users Handbook Data Products, Chapter 11 2011 (National Aeronautics and Space
Administration)
[24] Tucker C J 1979 Remote Sens. Environt. 8(2) 127
[25] Jackson R D, Slater P N and Pinter P J 1983 Remote Sens. Environ. 13(3) 187
[26] Tucker C J, Newcomb W W, Los S O and Prince S D 1991 Int. J. Remote Sens.12(6) 1133
[27] Hansen M C, Defries R S, Townshend J R G and Sohlberg R 2000 Int. J. Remote Sens. 21(6-7)
1331
[28] Gong P, Pu R, Biging G S and Larrieu M R 2003 IEEE Trans. Geosci. Remote Sens. 41(6) 1355
[29] Huete A R, Hua G, Qi J, Checbouni A and Van Leeuwen W J D 1992 Remote Sens. Environ.
41(2-3) 143
[30] Huete A R 1988 Remote Sens. Environ. 25(3) 295
[31] Jensen J R 2000 Remote Sensing of the Environment An Earth Resource Perspective (Prentice
Hall, Upper Sadle River)
[32] Peta Sumber Bahaya Gempa Indonesia 2017
[33] Rahmawati N I, Santosa B J, Setyonegoro W, Bambang S 2017 Jurnal Sains dan Seni 6(20) 2337
[34] Latief, H., Sunendar H, Yuhsananta. P and Riawan E 2006 Laporan Program Penelitian dan
Pengembangan IPTEK 155
[35] Saaty T L 1970 J. Math Psyc.15(3) 234
[36] Saaty T L 1980 The Analytic Hierarchy Process, Planning,Priority Setting, Resource Allocation

19
International Colloquium on Computational & Experimental Mechanics (ICCEM 2020) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1062 (2021) 012036 doi:10.1088/1757-899X/1062/1/012036

(New York:McGraw Hill)

[37] Saaty T L 2003 Eur. J. Oper. Res. 145(1) 85
[38] Forman E H and Selly M A 2001 Decision by Objectives (World Scientific Publishing Company)
[39] Saaty T L 1996 Decision Making for Leaders: The Analytical Hierarchy Process for Decisions
in a Complex World (Pittsburgh:RWS Publications)
[40] Harada K, Imamura F 2002 Proc. 12th Int. Offshore Polar Eng. Kitakyushu pp 652-658
[41] Alongi D M 2008 Estuar. Coast. Shelf Sci J. 76(1) 1
[42] Pranata Y A dan Johnny G P 2014 Jurnal Teknik Sipil 13(1) 25
[43] Husrin S, Jaya K, Aprizon P, Joko P, Yudhi C, Aditya H 2015 Procedia Earth and Planetary
Science 12(2005) 20
[44] Susanto D, Faida L R and Sunarto S 2019 Jurnal Ilmu Kehutanan 13(1) 4

20