Mapping of Tsunami disaster evacuation pathways based on Tsunami altitude scenario using Network Analyst Method (case study: Palu City, Central Sulawesi)

Puta and Chernovita/ JAGI Vol 4 No 1/2020 304 Mapping of Tsunami disaster evacuation pathways based on Tsunami altitude scenario using Network Analyst Method (case study: Palu City, Central Sulawesi) I Made Edy Kusuma Putra *, Hanna Prillysca Chernovita 1 1 Universitas Kristen Satya Wacana Jl. Dr. O. Notohamidjojo, Kel. Blotongan, Kec. Sidorejo, Kota Salatiga 50715, Indonesia *Corresponding author: e-mail : 682016023@student.uksw.edu


Introduction
A natural disaster is an event or series of events caused by natural phenomena or natural factors that can cause environmental damage, material loss, or human victims (Kamadhis UGM, 2007). Indonesia is a disaster-prone island country, such as earthquakes, tsunamis, and land or landslide movements. This disaster is caused because Indonesia is located on 3 continent plates namely Eurasian plate, Indo-Australian plate, and Pacific plate which continue to move in collide/converging, away/divergent, and cross-pass/transform to cause tectonic earthquakes and trigger the occurrence of tsunami wave (BNPB, 2012). Catastrophic tsunami will ruin what is passed and swallow many casualties. According to (BNPB, 2011) , tsunamis are a series of gigantic waves of waves arising from a shift in the seabed due to earthquakes. One of the Indonesian areas that is a few times the tsunami disaster is the city of Palu and its surroundings.
Palu City is located on the plains of Palu Valley and Palu Bay, with an average altitude of 0 -700 meters above sea level, located at position 0 º,36 "-0 º,56" South latitude and 119 º, 45 "-121 º, 1" east longitude. The area of Palu city is 395.06 km 2 . The city of Palu consists of 8 sub-districts and 48 village areas. The population of Palu was 379,782 inhabitants in 2017 (BPS, 2018). In the area of Palu and surrounding areas, several pieces of fault are potentially generating earthquakes that are quite strong and potentially tsunami. The fault is a hammer-Koro fault that extends from Palu to the south and southeast through the northern South Sulawesi to the south of Bone to the Banda sea (Daryono, 2011).
The city of Palu and surrounding areas have been struck by tsunami disasters, based on the Indonesian Tsunami catalog of 416-2018416- (BMKG, 2019 and the Tsunami Trail recording of the Palu Bay 2018 (Pribadi et al., 2018), can be seen also in table 1. The first occurred on 14 May 1921 in central Sulawesi, the earthquake Force 6.3 Richter Scale (SR) with a tsunami wave height of 1 Meter (m). On 1 December 1927, there was an earthquake with an apparent magnitude of 6.3 SR and followed by a tsunami that was sourced in the city of Palu. The tsunami altitude reached 15 m, the number of victims was 15. The 3 m wave altitude of tsunami occurred at Parigi on 20 May 1938 with the 7.6 SR earthquake force, resulting in a casualty of 50 people.

Open Access
Eleven years in the aftermath of the 20/05/1938 earthquake and tsunami struck Parigi, with the 7.6 SR earthquake force with a height of 3 m tsunami wave, the victim inflicted a 50 person. On 14 August 1968, the earthquake and tsunami rocked the Bay of Tambu Balaesang Donggala with a wave height of 10 meters and a tsunami runoff reached 500 meters to the mainland of the coastline. 160 people died, 40 people were declared lost and 58 were severely injured. On 1 January 1996, the Toli-Toli earthquake occurred with the earthquake strength of 7.9 SR and a tsunami with a height of 3.4 from the coastline engulfed a total of 9 people. On 4 May 2000, there was an earthquake and tsunami centered on Banggai, with the strength of the earthquake 7.6 SR and a tsunami high of 6 m from the coastline. The casualties were 50 people. On 28 September 2018, there was an earthquake centered on Donggala with the strength of the 7.4 SR earthquake and caused a tsunami in the Donggala and Palu areas. The high tsunami wave of the coastline reached 11.3 m and had a victim of 2,037 people.
According to the fact, the city of Palu requires mitigation of the creation of a tsunami disaster evacuation route to provide information to the community to save itself from the tsunami disaster. An evacuation route is made using the geographic system of Geography (GIS), as it can generate geography-based maps. GIS can assist in the manufacture of evacuation routes by viewing the tsunami runoff. The runoff of tsunami hazards is made using the tsunami wave altitude scenario.
The objective of this study is (1) to create a tsunami hazard map using the tsunami altitude scenario of the coastline. (2) Determine the evacuation point and evacuation route of the tsunami disaster in Palu based on the tsunami hazard map. The benefits of this study are (1) seeing the tsunamiprone areas.
(2) Provide information on evacuation points and evacuation routes so that people can save themselves during the tsunami. The scope of this research is the creation of a tsunami evacuation route, with the case study of Palu City.

Literature Studies
Research related tsunami evacuation route is done under the title of the use of GIS to mapping the evacuation route of the tsunami disaster in Tonggolobibi village of Sojol district of Donggala which was written by Nurfaida, DKK. The purpose of this research is to mapping the evacuation route of the tsunami, in line with the tsunami evacuation route using the system that is insyaAllah (SIG) and the determination of Shelter for the tsunami disaster in the village Tonggolobibi Sojol District Donggala. Research using the method of network analyst in GIS application is ArcGis 10.1. The preparation of the tsunami evacuation route uses the same parameters of the base map of which altitude, the road network, and the village administration map as well as the data of village monographic data. The results obtained are two evacuation routes, the first path of the width of the road 4, 56 meters, the length of the line 2225, 5 and the journey around 14 minutes. The CHNRD line has a road width of 2, 90 meters, road length of 2643, 9 and when it takes about 18 minutes (Nurfaida, 2016).
Another research titled Tsunami Insecurity Map and the plan of evacuation route in Desa Parangkritis Sub-district of Kretek District, Bantul regency of the special region of Yogyakarta, which was written by Gentur Handoyo, et al. This research aims to create tsunami-prone areas and draft tsunami evacuation routes. The study used overlay analysis in determining tsunami-prone areas and the evacuation route determination was determined based on the runoff of the tsunami area and based on BNPB rules in 2012. The parameters used are the topography altitude, the distance from the river, the distance from the coastline, the topography slope, the use of land, the tsunami runoff, the road network, the presence of hills, and high buildings. The result of this research is the main route for the evacuation of the Parangkritis village is Parangkritis Street, Jalan Parangkritis-Depok, and Depok Street. The evacuation route goes directly to the 12 temporary evacuation sites (TES) (Handoyo et al., 2014).
Conceptually, the above research has similarities with this research, namely to determine the evacuation route of the tsunami. This research focuses more on determining tsunami evacuation pathways based on tsunami run-ups. This research uses the cost-distance analysis in ArcGis software to determine tsunami-prone areas. For determining tsunami evacuation routes using network analyst.

Tsunami
A simple process of tsunami occurs due to a large amount of water volume shifting. Water moves in any direction forming tsunami wave tide, at the time of the water back to the initial front then there will be a tidal tsunami surge. And at the boarding time, the tsunami waves will beat and ruin all that he goes through and rescatter the oceans. Tsunami is caused by several factors, namely (1) The Underwater earthquake, (2) Submarine Volcano, (3) Landslide and (4) Falling Meteor (Tjandra, 2017). The wave height can be interpreted as a vertical distance between the highest sea-level conditions and the lowest sea-level conditions in a wave (Khoirunnisa et al., 2017). According to Abdurahman, et al. (2013) in Tjandra, 2017, based on the distance of the tsunami source to the territory of the tsunami is divided into 3 namely: (1) Tsunami with a distant source. Its run is more than 1,000 kilometers, for example, the Aceh tsunami of 2004, (2) the regional Tsunami radius of its run is no more than 1,000 kilometers from its source, (3) The local Tsunami is not more than 100 kilometers from its source. The following is a history of tsunami events in Palu.  (Pribadi et al., 2018).

Tsunami Evacuation Route
The main strategy in the making of tsunami disaster mitigation is to create an evacuation route to quickly evacuate the public from a tsunami disaster. The tsunami evacuation route for 2 is: The horizontal evacuation route is a higher information delivery away from the disaster in security zones. The vertical evacuation route is the provision of information to be directed towards the building or nearest place, with the provisions where the Evakuasinya must be earthquake resistant, have a room that is spacious and easy to reach. This research focuses on creating horizontal evacuation lines using network analyst methods.

Geographic Information System (GIS)
The Geographic Information System (GIS) is an informational system consisting of hardware components, software, geographic data and human resources that work together to input, store, repair, process, update, analyse, integrate, manipulate and display all geographical data (Adil, 2017). The components that build an information system are computer systems, software, spatial data, data management, analysis, and humans that operate GIS. Based on the GIS data structure is divided into 2 spatial data models, namely raster data models and vector data models. Raster data is the data that is got from remote sensing systems whose objects are in the form of pixels. Vector data is the shape of the Earth's surface, which is in the form of lines, dots, and areas. Each spatial data has its weaknesses and strengths, depending on the purpose, data availability, the accuracy that will be made and easy analysis (Irwansyah, 2013).

Network Analyst
The process of mapping the tsunami evacuation route using the closest facility method in network analyst in Arcmap application. Analysis using the closest facility method will consider the road network, road width, and road length. In making a tsunami evacuation route using the closest facility is required to have a starting point and endpoint of the tsunami evacuation route. The starting point is the tsunamiprone area and the endpoint is the evacuation site.

Satellite imagery interpretation techniques
Image interpretation is an activity on satellite imagery or aerial photographs to identify objects in them and extract information and assign value to them. 3 activities are done to recognize objects that are detection, identification, and analysis. Detection is the observation of an object. Identification is to characterize a detected object by using an existing caption and the analysis is to collect further information. 2 divided interpretation techniques are visual and digital interpretation techniques (Al Amin, 2017).

Geometry correction
Geometry correction is the process for Meproyeksikan imagery to have geographical coordinates. Geometry correction has 3 objectives, namely: (1) perform rhetorician or image improvement to return to its geographical coordinates, (2) perform a redistribution or image match with a geometrid corrected image and, (3) perform registration or image matching with a map that has corrected geometry to produce a specific projection (Candra and Suprianto Ahmad Afif, 2015). In geometry correction, there are 2 types of methods used are Rectification method (image to map) and the method of registration (image to image). To reduce geometric distortion is done by the rectification of geometry. Rectification is the process of repairing imagery to the actual position caused by a position shift. The accuracy of the geometry correction is highly dependent on the number and spread of the Ground Control Point (GCP). The terms of the GCP range are as follows: (1) The perimeter side of the image, (2) the center of the area, (3) The border area, (4) spread evenly and, (5) adjusting the terrain condition. The GCP (Ground Control Point) or ground control point is a location point that can be identified in real Rungan (on the ground), which serves as an allied point between the map coordinate system and the photo coordinate system. An ICP (Independent Check Point) or an accuracy test point is as quality control of an object by comparing the coordinate of the model to the actual coordinates. ICP requires that the point of the join should spread evenly throughout the area of the object to be tested. The geometry accuracy component is divided into 2 namely: horizontal and vertical accuracy test. In this study will be tested in horizontal accuracy (Al Amin, 2017).

Research materials
The materials used in this study are: 1. Palu City Satellite imagery Satellite imagery used to acquire/edit land use data, coastlines and road networks 2. Contour Map of Palu City Contour Map of Palu city is used to obtain tilt data 3. Surface roughness coefficient 4. Palu city administration and land use Data

Research tools
The tools used in this study are: 1. Laptop set Asus P452LJ 209 with a specification processor Inter Core i5 and RAM 4 GB 2. ArcGIS software version 10.3 used for data processing and map creation 3. Set Microsoft Office 2013.

Stages of research
Research stages can be seen in Figure 1.  Broadly, there are 4 stages of research that is done, the first is the phase of data collection and preparation, the data in the improvement is secondary data. Basic map of Administration map, land use map, line map. Beach and road network maps are getting from (www.tanahair.indonesia.go.id) with a scale of 1:50,000, the Elevation Data map model (DEM) in Can of (www.tides.big.go.id/DEMNAS) on 6 october 2019 and satellite imagery data in a can from the Universal Maps Downloader (UMD) application.
The second stage is the data preparation phase, satellite imagery in geometry correction using the corrected road network geometry, the testing done is horizontal accuracy test. Then the basic map (land use map, the line will do the editing process by performing the satellite image interpretation manually with digital on-screen, with an analysis scale 1:5,000. After the editing process is completed, the next will be calculated on the surface roughness of land use data using Berryman numerical modeling. Here is the table of surface roughness values. Where: Hloss = loss of tsunami altitude for 1 m of the propagation distance H0 = The initial altitude of the tsunami on coastline n = coefficient of surface roughness S = slope surface Considering the data on the surface roughness, land slope and the initial altitude of the tsunami of coastline, the result is a map of tsunami hazard. Furthermore, the determination of evacuation points. Further, determine tsunami evacuation pathways using a road network with the Analyst network method. The fourth stage is the presentation of the results of (1) Tsunami hazard map of Palu City (2) tsunami evacuation route of Palu City.

Geometry Correction
The geometry correction method used in this study is a rectification (image to map). This method uses maps from RBI as a ground control point (GCP) Scale of 1:50,000. The map used is a road network map to pull the image back into an actual position. The geometry corrections are used on the road network map of RBI that has been corrected with geometry. In this study, the coordinate system used is with the Datum WGS 84 projection System 50 S. Selection of ground control points is done by looking at the position of clear natural objects, not easily changing or shifting in a short time, must be on the surface of the ground and have access to GCP locations such as crossroads or T-junction roads, sewers, and corner buildings. The results of the accuracy-test can be 5.506. Based on the results of the standard accuracy test that is still allowed in the map performance is scale 1:25.000.

Land-use of Palu City
The use of Palu city land is divided by coefficient's surface roughness as in the table 2. Result of interpretation of land use of Palu water body, shrub/shrub, forest, plantation, vacant land, farmland, settlement, Pond/dam, and road network. There is no mangrove in Palu city. Dominating land use at the research site is forest. The land-use map of Palu city can be seen in

Tsunami Hazard Analysis
Tsunami Hazard Analysis is achieved using equation (3.1). In the creation of a Tsunami hazard map, There are several inputs namely surface roughness map, slope map, shoreline and the scenario of tsunami altitude 15 m based on the Indonesian tsunami catalog in 416-2018, bmkg 2019. The tsunami hazard analysis also uses the tools cost distance to measure the loss of tsunami wave height of each pixel in uniform starting from the coastline. Based on the calculations obtained the most widespread subdistrict is the North Palu sub-district with a total area of 8.643528 km 2 . Mantikulore district with an area affected 8.078757 km 2 District Ulujadi 2.201102 km 2 , West Palu sub-district 7.971213 km 2 . The district of the Taweili tsunami-affected by 5.210359 km 2 and East Palu Sub-district as much as 1.422311 km 2 . District Tatanga and Palu Selatan are not exposed to the tsunami runoff because it is far from the beach. Can be seen in Fig 4.

Number of inhabitants of Palu
Population Data is used to determine evacuation sites. Not all of the inhabitants of Palu City were evacuated because some were not impacted by tsunami runoff. Palu City District, the most populated of the population, was 71.52 inhabitants in 2017. It can be seen in Table 3. The data used to determine evacuation sites is land use data and tsunami hazard data. Analysis of the determination of evacuation places is a region or place that is not flooded with the tsunami. The existence of buildings and buildings is assumed by many people who activity the premises.

Mapping Tsunami evacuation pathways
The mapping of tsunami evacuation pathways is the result of this study. Planning tsunami evacuation lines using the gathering points and evacuation points that have been in the previous analysis. These two points are determined based on the outcome of tsunami hazard data. The evacuation route planning takes care of road type and width. The evacuation route was made as many as 93 lines divided into six sub-districts, namely Ulujadi District, West Palu Sub District, East Palu Sub District, Kecamtan Mantikulore, Taweili and, North Palu subdistrict.
Ulujadi District There are 13 evacuation routes, the longest line 2,717 m with a travel time of 17.5 minutes by running. Shortest path 96 m with a travel time of 0.6 minutes by running. The calculation of mileage obtained based on human average runs according to Weiner (1968)

Conclusion
The tsunami runoff with the scenario of tsunami altitude of 15 meters is the first step to see the extent of the potential of the city of Palu against the tsunami disaster. The most impactful area is the North Palu sub-district with a total area of 8.643528 km 2 . A possible mitigation is to create a tsunami evacuation route. The creation of a tsunami evacuation route is made using the network analyst method to produce 93 evacuation routes divided into six sub-districts namely North Palu Sub-district, Mantikulore, Ulujadi, West Palu East Palu, and Taweili. District Tatanga and Palu Selatan are not exposed to the tsunami runoff because it is far from the coastline so there is no need for tsunami evacuation routes. The longest evacuation route is 4,297 meters with a travel time of 27.6 minutes with running and the shortest evacuation route of 96 meters with a travel time of 0.6 minutes by running.