Preliminary study of tsunami simulations on megathrust earthquake scenarios in Pacitan Regency, East Java

During the period 1600 to 2007 there have been approximately 109 tsunamis in Indonesia with 90% caused by tectonic earthquakes. Pacitan located in an east java segment subduction area which is very prone to earthquakes and tsunamis. The coastal morphology of Pacitan has great potential for tsunami hazard. This research is focused on conducting numerical modeling of the tsunami generated by the hypothetical megathrust earthquake in the East Java segment with a magnitude of Mw 8.7. The method used is Cornell-Multi Grid Coupled applied shallow water equation with COMCOT 1.7 software. The nested grid system uses 4-layer consisting of 1-layer GEBCO, 2-layer bathymetry, and 1-layer integration of bathymetry and topography to produce tsunami modeling with high resolution. The results showed that the megathrust hypothetical earthquake scenario for the East Java segment caused a vertical displacement at sea level of -4.599 m to 7.019 m. The tsunami propagation northwards towards the Pacitan coast had a maximum amplitude of 33.16 m with a travel time of 23 – 29 minutes and then spreads in all directions. The farthest inundation range occurred in Pacitan Subdistrict as far as 4.189 km to the north of Ranuharjo Beach, while the maximum run-up of 21.82 m occurred in Soge Beach, Ngadirojo Subdistrict. The farthest affected area of inundation occurs on a gently sloping morphology with a slope of 0-8%, the highest run-up is on a moderate to steep sloping morphology with a slope of more than 25%. The results of tsunami simulation can be used for planning tsunami disaster mitigation in Pacitan.


Introduction
The activity of the Indo-Australian Plate subducting under the Eurasian Plate and the Pacific Plate in the ocean is a source of shallow earthquakes.During the period from 1600 to 2007, 109 tsunamis were recorded in Indonesia, with 90% of them caused by tectonic earthquakes.Tectonic earthquakes along the subduction area are the epicenters of these earthquakes, which in turn lead to tsunami disasters [1]. Figure 1 shows that Pacitan situated in a subduction area that is very prone to earthquakes and tsunamis.The coastal morphology of Pacitan presents a significant potential for tsunami hazards.Based on the results of tsunami hazard modeling in a hypothetical scenario of an Mw 8.5 earthquake, there are 16 1307 (2024) 012004 IOP Publishing doi:10.1088/1755-1315/1307/1/012004 2 subdistrict villages in Pacitan Regency that have the potential to be directly affected [2].According to the earthquake event relocations by the BMKG, there are seismic gaps along active plate boundaries that have not experienced major earthquakes for more than 30 years.This zone has the potential to be locked against plate shifts, resulting in energy accumulation and the potential for tsunamigenic megathrust earthquakes.When the deficit in the Central Java and East Java areas is released, it could trigger an earthquake with a magnitude of 8.8 [3].The southern region of Pacitan borders directly on the oceanic plate and the second plates and continental plates, which this area prone to earthquakes and tsunamis.Based on Figure 1, there are 7 Subdistricts in Pacitan that have the potential for a tsunami disaster.Geomorphologically, Pacitan is formed from two types of landforms, namely fluvial and marine has a flat topography with a height of less than 75 m [2].This categorization is based on information from the Department of Energy and Natural Resources of East Java Province regarding the seismic potential in the region [4].From Figure 1, it is crucial to mitigate tsunami hazards by creating inundation maps for vulnerable coastal areas.These maps are developed based on historical tsunami events and hypothetical scenarios [5].When creating a tsunami inundation map, a numerical model is employed to calculate the tsunami's propagation from the earthquake source to the coastal regions.The results provide information on runup and inundation values.
Tsunami numerical modeling utilizes shallow water equations due to their efficiency in simulating tsunamis through numerical schemes [6].Therefore, this research focuses on conducting numerical modeling of tsunamis by understanding the characteristics of tsunami-generating (tsunamigenic) earthquakes [7].In this study, the tsunami is generated by the South Java Megathrust earthquake.The research employs the Cornell-Multi Grid Coupled method, which applies long wave equations in shallow waters using COMCOT software to transform data for statistical analysis.The modeling process take into an account coastal morphological factors affecting tsunami waves and the simulation of inundation areas.

Data and Methodology
In this study, utilized Digital Elevation Model (DEM) and bathymetry data from the National Bathymetry Information System.This study employed six DEM datasets around Malang Regency, specifically DEMNAS with code v1.0, using vertical datum EGM2008 with a resolution of 0.27 arcseconds (approximately 8 meters).The bathymetry data covered the East Java region and used the Mean Sea Level (MSL) datum with a resolution of 6 arc-seconds (about 185 meters).We obtained Mean Sea Level (MSL) and Coastline data for the year 2021 from the Geospatial Information Agency.Additionally, parameter data for the tsunami-generating earthquake, specifically the East Java segment megathrust earthquake, were acquired from the Meteorology, Climatology, and Geophysical Agency (BMKG) and the National Seismic Centre (PUSGEN).The worst scenario simulation for a tsunami generating earthquake used is a hypothetical megathrust earthquake of magnitude Mw 8.7 obtained from BMKG sourced from the National Center for Earthquake Studies (PUSGEN).Recognized from Figure 2 in modeling the tsunami inundation, the hydrodynamic equation numerical modeling method is employed to discuss the governing equation and arrange it in a nested grid configuration.In deep waters, the nature of the equation is linear to shallow water which is nonlinear in nature, therefore it is necessary to set the boundaries of the moving boundary scheme and dispersion improvement.Tsunami modeling applies the Cornell Multi-Grid Coupled method to obtain inundation and run-up values [6].The longwave linear equation for the spherical coordinates used is: The value of the Coriolis force coefficient is influenced by the value of the earth's rotational speed multiplied by sin latitude. = Ω (4) The nonlinear equation is used in shallow water so that it can describe the fluid flow motion in the coastal zone as follows.
The value of F represents the subsurface friction in the X and Y directions which is written in the equation below.
This modeling uses a Nested Grid Configuration, where it is necessary to set a grid of count points that can be done flexibly even in stages from large to small grids in order to improve the balance between accuracy and efficiency.Based on Figure 3, the coarse grid setting is in the large domain layer (parent grid), while the finer grid is used for the smaller layer domain (sublayer grid) [6].

Integration Topography and Bathymetry Data
Tsunami modeling relies on detailed topography data for land and bathymetry data for the ocean to yield accurate results in hazard mapping [5].Bathymetry data provides information about water depth, but it operates on a large scale and lacks fine details.Conversely, topography data offers elevation values for the land but operates on a smaller scale with higher levels of detail.Due to these differences in data quality, our modeling approach involves integrating and combining topography and bathymetry data.These changes provide a more fluid and coherent explanation of the integration of topography and bathymetry data in tsunami modeling.The red line represents the coastline data for 2021, with an elevation of 0 meters, adjusted using the Indonesian Rupa Bumi (RBI) dataset of Pacitan Regency at a 1:50K scale as a reference for land area alignment.The bathymetry data is filtered 'after elevation' to include only elevations less than 0 meters, representing underwater depths.Conversely, the topographic data 'after elevation' includes elevations greater than 0 meters, representing land elevations.The integrated dataset successfully combines these two sources, ensuring positive elevations for land and negative elevations for the sea.This combined bathymetry and topography dataset has a resolution of 0.36 arc-seconds or approximately 11 meters.The integrated bathymetry and topography dataset serves as the foundation for subsequent grid creation processes.

Setting Grid
Tsunami modeling, employing the Cornell Multi-Grid Coupled method, utilizes a hydrodynamic model concept that incorporates a nested grid as the simulation medium.The principle behind this nested grid operation begins with the largest domain and progressively refines it to the smallest domain [6].In the 'Parent grid,' a single grid box contains nine 'Sub-level grid' boxes, each divided into three segments in both the x and y directions.This results in a resolution ratio of 3x from one layer to the next, cascading down to the lowest layer.The broader domain area serves to display tsunami modeling outcomes stemming from tectonic activities on a global scale.For this modeling, the largest domain employs data from the General Bathymetric Chart of the Oceans (GEBCO).Tsunami modeling involves four layers with varying grid sizes, as detailed below.COMCOT modeling requires grid setting of count points that can be done flexibly even in stages from large to small grids in order to improve the balance between accuracy and efficiency.A finer grid will be required for modeling in shallow water.The water depth varies within the computational domain, different grid sizes and time step sizes are used within the sub-regions.Nested Grid Configuration makes it possible to obtain more detailed information on coastal areas by using a finer grid [6] [10].
The layer arrangement based on table 1 is in accordance with the nested grid principle, namely the lowest layer (layer 1) has the largest domain with coarse resolution to the smallest domain with fine resolution (layer 4).The resolution setting for each grid layer is reviewed from the top layer first, in this modeling layer 4 as the top layer has a resolution of 0.0001 using bathymetric topographic integration data.Layer 4 will later become the focus of modeling analysis because its area covers each district.The resolution value in the lower layer is 3x that of the upper layer, so layer 3 has a resolution of 0.0003 using BATNAS data, layer 2 has a resolution of 0.0009 using BATNAS data, and layer 1 has a resolution of 0.0027 using GEBCO data.

Initial Conditions (Source Modeling)
The modeling involves a simulation of a megathrust earthquake, which comprises three earthquake source points.In Figure 1, the distance from the epicenter to the Pacitan coast varies, with the closest point being 195.65 km (subplot 1) and the farthest point reaching 381.49km (subplot 3).During the initial stages of the tsunami, there is a process of sea level fluctuation, involving both uplift and subsidence (vertical displacement).When the oceanic plate undergoes a rapid uplift process following a megathrust earthquake, it leads to significant fluctuations in the mass of water from the ocean floor to the sea surface [11].Shallow earthquake sources with substantial force release a vast amount of energy that displaces the water mass above them, resulting in increasingly significant vertical displacement differences [12].The three earthquake source points are oriented towards the south coast of Java.Earthquake strikes directed towards shallower bathymetric features generate taller tsunami waves, while those directed towards deeper bathymetric features produce smaller tsunami waves.

Tsunami Propagation (Oceanic Modeling)
The East Java segment of the megathrust earthquake has a magnitude of Mw 8.7, resulting in a tsunami magnitude of 4.862 and an energy release of over 12.8 x 10^20 ergs.This event has the potential for run-up with heights ranging from 24 to 32 meters.The propagation of the tsunami, modeled through oceanic simulations, spreads in all directions from the earthquake source due to the displacement of the Earth's plates.The output of the COMCOT processing generates 'zmax_layerxx' data, depicting the distribution of tsunami heights as flow depth and the extent of the tsunami's reach.Additionally, there is output data labeled 'ttt_layerxx,' which illustrates the propagation of tsunami travel time from the earthquake source to the coastline [5][10].The direction of the tsunami wave propagation in Figure 6 follows the dominant direction from the epicenter of the earthquake.The megathrust earthquake scenario has a strike direction of 275 o -285 o NE or heading north towards the south coast of Java.So, the modeling results show that the tsunami waves split into two heading north towards the south coast of Java and towards the Indian Ocean [10].
In this modeling using a recording time interval every 15 seconds, the modeling results show the value of z from z_01_000000-002292 with the resulting interval z_layerid_000003.
Table 3. Tsunami propagation snapshot at layer-1   3 presents the results of a snapshot of tsunami height recordings at layer 1.In this depiction, the blue areas represent subsidence, while the red areas indicate sea level rise (uplift).The propagation of the tsunami closely follows the underwater deformation pattern in the hypothetical earthquake scenario.Wave movement is directed northward towards the south coast of Java and partially southward into open waters.Figure 7 showcases the simulation at the 5 th minute, where the minimum wave amplitude decreased to -2.71 m, while the maximum wave amplitude reached 3.25 m.Moving to Figure 8, representing the 10 th minute, the subsidence area around the epicenter begins to experience sea level rise, with the minimum amplitude measuring -5.23 m and the maximum amplitude at 3.45 m. Figure 9, depicting the 15 th minute simulation, exhibits tsunami amplitudes ranging from -6.82 m to 4.81 m.By the 20 th minute (Figure 10), the maximum tsunami amplitude significantly rises to 14.17 m and begins to encroach upon the southern coast of Java, with the sea level at the earthquake's epicenter dropping to -9.72 m. Figure 11 illustrates the 25th minute, as the tsunami waves reach the coast of Pacitan, boasting a height of 16.45 m.Finally, Figure 12, at the 30 th minute of simulation, shows an amplitude peak of 20.96 m.The propagation of the tsunami from the 5 th to the 30 th minute reveals substantial changes in amplitude towards the coast, while in deeper waters, the tsunami amplitude diminishes.Travel time along the south coast of East Java spans from 22 to 33 minutes, while on the coast of Pacitan Regency, it ranges from approximately 23 to 29 minutes.

Inundation and Run-up Tsunami
The outcomes of the local-scale tsunami modeling are based on layer 4 (district domain), which allows us to assess inundation and run-up conditions along the southern coast of Pacitan Regency.In layer 4, run-up and inundation values are determined using the nonlinear shallow water equations.The output from the COMCOT modeling provides 'hmax_layerxx,' representing changes in sea level.From the layer 4 tsunami modeling results, we can observe the inundation and run-up values for each sub-district.In this context, inundation is calculated by measuring the farthest horizontal distance that the tsunami penetrates the land, relative to the coastline.On the other hand, run-up is determined by the vertical distance between the coastline, which has an elevation of zero, and the height of the tsunami on land.The variability in tsunami inundation and run-up across subdistricts in Pacitan Regency is primarily influenced by the region's morphological conditions.Ngadirojo Subdistrict, with its dominant bay area, experiences the highest run-up due to wave resonance and refraction, particularly in bay areas, where incident and reflected waves meet and amplify the wave height [13].The modeling results also indicate that the highest run-up occurs in areas with moderate to steep sloping morphologies, characterized by slopes exceeding 25%.In contrast, the farthest inundation area is found in Pacitan Subdistrict, characterized by a gentle sloping morphology with slopes ranging from 0-8% [14].These tsunami simulation results are valuable for planning tsunami disaster mitigation strategies in Pacitan.

Comparison of the Efficiency of Using Layers in Modeling
Tsunami modeling using the Cornell Multi-Grid Coupled method applies the Nested Grid Configuration principle making it possible to obtain more detailed information in coastal areas using a finer grid.In this modeling, it is important to compare the resolution and domain boundaries of each layer to determine the level of modeling accurary [6].The results of the 3-layer tsunami modeling have a maximum height value of 15.95 m, the COMCOT running process time is 1 hour 13 minutes with a total file storage used of 6.13 GB.The results show a maximum tsunami height of 17.82 m, COMCOT running process time of 2 hours with total file storage used of 16.2 GB.The difference in tsunami height values in 3-layer and 4-layer modeling is influenced by the number of domain layers used.Each box in the parent grid (large layer) will calculate 1 water level change value [6], the more sub-level grids there are, the more water level changes will be calculated, resulting in more detailed outcomes for 4-layer modeling that takes into account the domain boundary compared to 3-layer modeling.

Conclusion
In conclusion, tsunami simulation in Pacitan Regency due to the potential for a megathrust earthquake in the East Java segment with a magnitude of Mw 8.7 use the nested grid system uses 4-layer consisting of 1-layer GEBCO, 2-layer bathymetry, and 1-layer integration of bathymetry and topography to produce tsunami modeling with high resolution.The results showed that the megathrust hypothetical earthquake scenario for the East Java segment caused a vertical displacement at sea level of -4.599 m to 7.019 m.The tsunami propagation northwards towards the Pacitan coast had a maximum amplitude of 33.16 m with a travel time of 23 -29 minutes and then spreads in all directions.The farthest inundation range occurred in Pacitan Subdistrict as far as 4.189 km to the north of Ranuharjo Beach, while the maximum run-up of 21.82 m occurred in Soge Beach, Ngadirojo Subdistrict.The farthest affected area IOP Publishing doi:10.1088/1755-1315/1307/1/01200410 of inundation occurs on a gently sloping morphology with a slope of 0-8%, the highest run-up is on a moderate to steep sloping morphology with a slope of more than 25%.The results of tsunami simulation can be used for planning tsunami disaster mitigation in Pacitan.

Acknowledgment
The author gratitude to Geophysical Engineering Department for supporting research activities.The author appreciates the Meteorology, Climatology and Geophysical Agency for providing earthquake parameter data and assisting research methods, as well as the Geospatial Information Agency for providing DEM (Digital Elevation Model) and bathymetry data free of charge.Also, the authors thanks to the editors and reviewers for useful comments and corrections to improve the quality of the manuscript.

Figure 1 .
Figure 1.The research area (yellow rectangle) of tsunami modeling with overlay word imagery.There are 7 subdistricts on the coast that have potential for tsunami hazard, from left to right (Donorojo, Pringkuku, Pacitan, Kebonagung, Tulakan, Ngadirojo, Sudimoro).

Figure 2 .
Figure 2. The method used in this study.Integration of DEM and BATNAS as well as megathrust earthquake parameters are used to modeling the inundation and run-up of the tsunami Pacitan.

Figure 4 .
Figure 4. Comparison of data (a) Bathymetry and (b) Topography before integration, (c) Integrated bathymetry and topography data, coastline (in red) as a reference limit

6 Figure 5 .
Figure 5.The initial condition of sea level immediately after the earthquake occurred was a sea level change of -4.599 m to 7.019 m Figure 5 illustrates the maximum value, representing the vertical movement process in the rising oceanic crust.Conversely, the minimum value indicates a decrease in the oceanic crust's elevation.When the oceanic plate undergoes a rapid uplift process following a megathrust earthquake, it leads to significant fluctuations in the mass of water from the ocean floor to the sea surface[11].Shallow earthquake sources with substantial force release a vast amount of energy that displaces the water mass above them, resulting in increasingly significant vertical displacement differences[12].The three earthquake source points are oriented towards the south coast of Java.Earthquake strikes directed towards shallower bathymetric features generate taller tsunami waves, while those directed towards deeper bathymetric features produce smaller tsunami waves.

Figure 6 .
Figure 6.Tsunami propagation consists (left) Tsunami wave amplitude with maximum amplitude 33.19 m and (right) Tsunami travel time

Figure 13 .
Figure 13.Results of tsunami inundation and run-up modeling (left) Pacitan Subdistrict, (right) Donorojo Subdistrict According to Figure 13, in the Pacitan Subdistrict, the tsunami reached a height of 17.81 meters with an arrival time between 29 and 30 minutes.The inundation in the Pacitan Subdistrict extended furthest north, reaching 4.189 kilometers from Ranuharjo Beach.The highest run-up occurred at Telengria Beach, measuring 15.19 meters.In the Ngadirojo Subdistrict, the tsunami reached a height of 22.71 meters with an arrival time between 26 and 28 minutes.This subdistrict experienced inundation extending 3.98 kilometers north of Taman Ngadirojo Beach, while the highest run-up, at 21.82 meters, was observed at Soge Beach.The variability in tsunami inundation and run-up across subdistricts in Pacitan Regency is primarily influenced by the region's morphological conditions.Ngadirojo Subdistrict, with its dominant bay area,

Table 2 .
Comparison of the resolution of each grid layer