Seismic Activity and Tsunami Potential in Bali-Banda Basin

Transcription

1 Seismic Activity and Tsunami Potential in Bali-Banda Basin Suci Dewi Anugrah 1 and Bambang Sunardi 2 1 Tsunami Mitigation Subdivision Indonesian Meteorological Climatological and Geophysical Agency, Jl. Angkasa 1 No. 2 Kemayoran, Jakarta Pusat sucirahman@yahoo.com 2 Research and Development Centre Indonesian Meteorological Climatological and Geophysical Agency, Jl. Angkasa 1 No. 2 Kemayoran, Jakarta Pusat b.sunardi@gmail.com Abstract. Since the year of 1975 to 2011, Bali-Banda Basin had been shaked by 83 times earthquakes above 5.0 Mw. All of the earthquakes were occurred in the sea, however only one earthquake generated tsunami in 1992 at Flores. The objective of this study is to determine the most potential areas for tsunami hazard by analyzing and conducting the tsunami modeling based on the earthquake historical data at the area.the NEIC earthquake catalog from and Global CMT catalog were used to analyze seismotectonic parameters and earthquake recurrence time at the area 6 o -12 o S and 114 o -126 o E. To determine tsunami hazard in this region, we use three tsunami simulation models which have high tsunami run up potential.the result shows that seismic activity with the magnitude more than 5 Mw tends to get higher exponentially. The recurrence time of earthquake with magnitude 6.5 is about 7-35 years, magnitude 7 is about years and magnitude 7.5 is about years. Tsunami wave distribution from tsunami simulation will be the discussion materials to decide tsunami potential along Bali-BandaBasin. Keywords: Seiscmic activity, tsunami potential, Bali-Banda Basin 1 Introduction As a consequence of having surrounded by Indo-Australian plate, Pacifik plate and Philippine plate which subduct beneath the Eurasian plate, made Indonesia as a very seismically active region. Many big earthquakes and tsunami events are recorded in this area. After the Sumatera Tsunami 2004 event, many research studied about the tsunami potential due to the earthquake subduction along the Indian Ocean. Sumatera and Java are the threatened area due to the earthquake. Beside those two islands, Sulawesi, Papua and some small islands in the eastern part of Indonesia like Bali, Nusa Tenggara and Flores has also a potential tsunami threat due to another

2 earthquake mechanism. Generally, the tsunami potential of these areas is caused by some small fault segments that surrounded the islands. This research studies the seismic activity and tsunami potential in the east part of Indonesia namely Bali-Banda Basin which is located along the north part of Bali in the west to the northeast of Flores island in the east. A history of an earthquake generated a big tsunami event was recorded in this region which is known as The Flores Tsunami 1992 (Yieh, et al., 1993). This even had caused substantial casualties and property damage. About 2080 people were dead while half of that was caused by tsunami. Most of the tolls are the people of Maumere, an area that are located in the north part of Flores, and people of Pulau Babi, and island in the north of Flores. Further tsunami potential studies in this area, based on the historical fact, are needed to understand tsunami potential in the future. The main reason of this study is the lack of research of seismic activity related to the tsunami potential in this region. Another research background is the presence of some fault segments in the Bali-Banda Basin as a source of earthquake which can generate a very local tsunami that impact to the closest islands along the Basin. In relation to the Tsunami Early Warning System which is run by Meteorological Climatological and Geophysical Agency, it needs a lot of effort to give some knowledge to the people in the area that has a potential experience of a very local tsunami, if the tsunami early warning system could not give a warn due to the very fast tsunami arrival time. 2 Tectonic and Geology Bali-Banda Basin 2.1 Geographic and Tectonic Aspects of Bali-Banda Basin The Bali-Banda Basin is located at the eastern part Indonesia. It is a prolongation of Sunda Arc. At the north of Bali-Banda Basin, a series of fault segments extend from east to west. The fault segments are known as Sunda back arc. Silver, et al., (1983) stated that the Sunda back arc structure is dominated by two large north directed thrusts, the Wetar and Flores thrusts and more minor thrusts. Those thrusts may represent early stage of subduction polarity reversal of the arc. In addition, the Wetar thrust has negative relation to the volcanism system. The Bali-Banda Basin is also dominated by the collision between the eastern Sunda Arc and Indo-Australia Plate. Collision zone is located in the south of Sumba, where Indo Australian plates grind the front of the arc (forearc) with a reverse fault mechanism. This collision has occurred since 3 million years ago. Irsyam, et al., (Australia 2010), reviewed the existence of the all the fault segments and subduction zona in Indonesia, which stated that there are four major segments located around the Bali to Flores, beside the Indo-Australia subduction zone ( Fig 1). The faults are known as Sumba fault, Wetar fault, Timor fault, and Flores fault.

3 Fig.1. Fault segments around the Bali-Banda Basin (Irsyam, et al., 2010) 2.2 Geological Conditions and Bali-Banda Basin Structure The basement underlying the Bali Basin is still a matter of controversy. Curray et al., (1977) and Hamilton, (1979) stated it was formed by a transitional crust that has a thickness between oceanic and continental crust; Ben-Avreham and Emery (1973) inferred that the oceanic crust is the basement material of the basin as a western extension of the Flores Basin; McCaffrey and Nabalek (1987) argue d that the basement of the basin was formed by the continental crust having some thickness and origin as that beneth the Sunda Shelf. Cretaceous rocks that reflect sunda exposure can be traced to the northern of basin under Kangean-Sepanjang ridge. Based on the gravity and seismic refraction data, the north part of the basin represents a Paleocene extensional tectonic regime, while the back arc fold thrusts zone formed since Neogene time was associated with both the Australian margin- Banda Arc collision as well as subducting of the oceanic plate in the Sunda trench in the south of Bali (Prasetya, 1992). The Bali Basin is a narrow (100 x 200 km), half circle shape, gradually getting deeper bathymetry to the east with the maximum depth of 1.5 km. Between the Bali Basin and Flores Basin an elongated Lombok Trough exists from north of Lombok to the central of Sumbawa. It is a short and narrow trough. The dimension of the trough length and width is about 100 x 50 km. The maximum bathymetry of the through is about 1.5 km whereas the southern ridge are between 1000 to 1300 m. The trough is underlain by oceanic crust as a continuation of the Flores Basin. The Flores Basin is an E-W turned to SE oriented. The Basin is deeper than Bali Basin with the maximum bathymetry of about 5 km. The eastern part of the basin is bounded by the east Salayar Ridge extending onto South Sulawesi. In the western part of the ridge, the basin is composed of Neogene volcanic and sedimentary rock dipping to the west and southwest. The eastern most part is the Banda Basin. It is near the triple junction area between three major plates, namely the Eurasian, Pacific and The Indo-Australian Basin which

4 are converging since Mesozoic times. The basin exhibits an elongated shape with ENE-WSW direction, parallel to the Banda Arc. The dimension of the basin is about 800 km long and 150 km wide. The basin can be divided into two parts which are separated by the volcanoes fracture zone: the Wetar Basin and The Damar Basin. The depth of the two basins is about 4500 m and 5000 m respectively. The nature of the basement underlying the Banda Basin is oceanic crust and has been interpreted as a Cretaceous-Eocene basin related to the Celebes and Sulu Basin, but a Neogene back arc origin was also considered (Hinschberger, et al., 2001). 3 Method 3.1 Seismic Analysis A series of seismic catalog in the Bali-Banda Basin was analyzed to know the seismic activity and its characteristic of the area. The characteristics refer to the earthquake recurrence period, the possible of the biggest magnitude, and the earthquake mechanism. The Gutenberg-Richter relation (1954) is used to find out the earthquake characteristic of an area. The Gutenberg-Richter relation is simply formulated as follows: The n(m) is a number of earthquake with a magnitude of M, while the a and b parameters referred to a seismic activity representing a seismic rate of a certain period. The first parameter is depended to the observation period and the dimension of the field observation, while the second one can be estimated statistically (Utsu, 1965). One of the formulas to estimate the b parameter is : (1) (2) whereas is a magnitude average and M min is a minimum magnitude. The value of a can be a estimated by using the formula : (3) The cumulative distribution can be calculated as follow: Number of earthquake occurrence/year can be estimated by divided the value of a with the observation period (T): (4)

5 (5) The number of earthquake cumulative frequency every year or seismicity index is: (6) Therefore, one can calculate the probability of once or more earthquake occurences with a bigger than M magnitude of T period as follows: The average of return period can be estimated by: (7) tahun (8) 3.2 The Tsunami Potencial Analysis The back arc thrust reactivation at the Bali-Banda Basin causes many shallow depth earthquakes in this area. Historical seismicity recorded a Seririt Earthquake on the year of 1976 with a magnitude of 6 Mw. The earthquake was located in the north of Bali Island, killed more than 500 people. This event might be a proof that the back arc thrust reaches to the north of Bali, and also as a reason about the presence of tsunami potential in the north of Bali. The Flores Tsunami 1992 was also recorded as a big tsunami in this study domain caused by the back arc thurst in the Bali-Banda Basin. The presence of the back arc thrust in the Bali-Banda Basin might be considered as a source of the shallow earthquake generated tsunami to the north coast of Bali-Flores Island. The tsunami potential analysis was also carried out in this study by using a deterministic method of tsunami numerical modeling. A series of tsunami modeling based on earthquake and tsunami catalog were done in this study such as the Flores Tsunami Three tsunami simulations in this region had been done in this study to prove the tsunami potential at the basin of Bali-Banda. The Mamuru Nakamura Tsunami Software was used to conduct the tsunami simulations. 4 Data In this study we used a series data of earthquake event in the area of 6 o -12 o S and114 o -126 o E. The CMT Harvard earthquake catalog from the year of 1973 to 2011 and the NEIC earthquake catalog of the period were used to obtain the data. Seismic analysis was performed with the ZMAP software (Wiemer, S., 2001).

6 For tsunami modeling, we used bathymetry data from ETOPO2 (can be downloaded from with 2 resolution. The bathymetry grid interval is 2.5 x 2.5 km. Modeling of tsunami could be applied for near-shore regionrun-up specifically by using a non-linearshallow water equations (Satake, 1995). The open boundary condition is used atthe edge of the computational area ( Nakamura, 2006). The finite-difference is a method that is used to calculate tsunami run-up. Three tsunami simulation with different scenario as listed in the Table 1 were done in this study. Some earthquake parameters data as well as a set of bathymetry data are needed in this simulation. The parameters are the fault dimension, strike, dip, slip, moment magnitude ( Mw) and the location of the earthquake source. Modeling of tsunami propagation was applied for near-shore region by using a non linear shallow water equation. The open boundary condition is used at the edge of the computational area (refer Nakamura, 2006). Run-up calculations using a code that is modeled by Nakamura (2006) using the finitedifference method was applied to simulate for the 1771 Yaeyama tsunami, southern Ryukyu Island, Japan. The bathymetry data were obtained from ETOPO2 provided by National Geophysical Data Center. The grid interval of bathymetry was 2.5km x 2.5km. Table 1. Fault parameters for each model used for tsunami simulation 5 Results and Discussion 5.1 Seismic Activity in Bali-Banda Basin Fig 2a shows seismic density in the region of Bali-Banda Basin and its surrounding obtained from NEIC catalog. The contrast red colour zone shows that Sumba, Sumbawa, Flores and some small islands in the south part of Banda Basin are the most active seismic region in this study area.

7 Fig.2. Earthquake density based on the seismic catalog and earthquake occurrence (>5 Mw) of the Bali-Banda Basin Based on both the Global CMT seismic catalog, there are 83 earthquakes that occurred in this area during with magnitude more than 5 Mw. The curve of seismic even with the magnitude more than 5 Mw tends to get higher exponentially (Fig 2b). The earthquake frequency tends to get higher in the period of after the tsunami earthquake event of 1992 in Flores. Fig.3 shows that most of the events are located in the sea. Four earthquakes have a magnitude more than 6.9 Mw are located at the eastern part of the basin. The biggest one was the 7.7 Mw of the 1992 Flores Earthquake. This earthquake generated a big tsunami. The other ones have a magnitude of 7 and 7.5 Mw. Those tsunami earthquakes were located in the basin of Sumba. Although those big earthquakes were a shallow earthquake but they did not generate a tsunami. Fig.3. The Earthquake epicenters of Bali-Banda Basin (M>5 Mw) during Seismotectonic The measurement of b-value resulted that the values varied spatially. It is showed that the b value are relatively low at southeast of Bali, north of Lombok, and at the

8 north of Flores. The b value represents a stress rate of an area. A low b value is correlated with a high stress rate. The result of the b value as seen in the Fig.4 is appropriate to the earthquake density as it was showed in the Fig.2a. Fig.4. Spatial variations of b-value and a value Bali-Bandabased on NEIC catalog The value of a represents an earthquake activity rate of an area. For this study area, the values vary approximately from 5 to 9. Two blocks are obviously seen having about 9 of a value. Both areas have a high rate seismic activity. 5.3 Earthquake Recurrence Time The earthquake with a magnitude 6.5 Mw in the region of Bali-Banda Basin has a recurrence time about 7 to 35 years as showed in Fig.5a. Fig.5b shows that the range of recurrence time of the bigger earthquake magnitude 7 Mw is about years, while Fig.6 shows that the earthquake with a magnitude 7.5 has a recurrence time range about years. Fig.5. Earthquake recurrence time with magnitude 6.5 and 7 Mw The short recurrence time is correlated with a high a and b value. The short recurrence time represents a very active seismic activity. For this study domain we

9 find the area of north Flores and the South of Sumba and Sumbawa as the region with a high seismic activity. Fig.6. Earthquake recurrence time with magnitude 7.5 Mw 5.4 Bali-Banda Basin Tsunami Potencial The length dimension of the back arc thrust in the Bali-Banda Basin is more than 1000 km. Due to its length it is possible to generate an earthquake with a magnitude more than 7 Mw in that region. It can be proved by some earthquakes recorded with a magnitude more than 7 Mw. However, an earthquake potential with magnitude more than 7 Mw in the north of Bali Island is less than the earthquake potential in the north of Lombok-Flores Island. The biggest earthquake magnitude recorded in the Bali Basin was only 6.5 Mw. Three scenarios of tsunami simulation in the Bali-Banda Basin had been done in this study based on seismic history, earthquake characteristic and considering the tectonic setting in that area. 5.5 Tsunami Numerical Simulation 1 We conducted a Flores Tsunami simulation based on CMT Harvard solution. In this case, the tsunami was caused by an earthquake with a magnitude of 7.7 Mw, 20 km depth and the earthquake parameters were 80 o of strike, 40 o dip, and 95 o slip. The earthquake was a thrust fault with the length and width dimension is about 100 km and 35 km respectively. We obtained 6 m of dislocation based on computational calculation. The simulation result shows that some areas have a tsunami wave height more than 3 meters which were happened at Lewobunga, Pamana Island, Lato beach, Babi Island and Nebe. The tsunami reaches the beach at 4 minutes after the earthquake. Fig.7 shows the tsunami maximum height and its propagation at t (time) = 4 minute for tsunami case model 1.

10 Fig.7. Distribution of maximum tsunami height (a) and tsunami propagation at minute 4 (b) for simulation model 1. Fig.8 shows the tsunami wave height comparison between our tsunami simulation and tsunami filed survey conducted by International Tsunami Survey Team ITST (Imamura, 1995). The simulation shows a closed result to the field survey. Fig.8. The comparison of tsunami wave height between field survey and numerical simulation The tsunami risk in those areas is relatively high considering a very high seismic activity. The possibility of local tsunami occurrence, a less 10 minutes tsunami arrival time, is another reason to enhance the risk in those areas. 5.6 Tsunami Numerical Simulation 2 The second simulation took a part in the Bali Basin region. This simulation referred to the earthquake event on Januari 6 th We assumed that the earthquake has a magnitude 7 Mw, 20 km depths, and the fault parameters are based on the CMT Harvard: 132 o strike, and 13 o dip. The fault dimension is about 50 km long and 20 km wide. A computational calculation results a dislocation value about 2.5 m. The simulation results that some areas experience a high tsunami wave more than 3 meters. The wave reached the coast at 4 minutes after the earthquake. Fig. 9 shows the wave height distributions and the tsunami propagation at 10 minutes after the earthquake.

11 Fig.9. Distribution of maximum tsunami height (a) and tsunami propagation at minute 10 (b) for simulation model Tsunami Numerical Simulation 3 The third simulation took a part in the region around Sumbawa Island. This simulation referred to the earthquake event on November8 th We assumed that the earthquake has a magnitude 7 Mw, 20 km depths, and the fault parameters are based on the CMT Harvard: 90 o strike, and 25 o dip. The fault dimension is about 50 km long and 20 km wide. A computational calculation results a dislocation value about 2.5 m. The simulation results that the tsunami height reached more than 3.5 meters in some areas. The wave come the coast at 4 minutes after the earthquake. Fig. 10 shows the simulation model 3 of the wave height distributions and the tsunami propagation model at 4 minutes after the earthquake when tsunami reached the coast for the first time. Fig.10. Distribution of maximum tsunami height (a) and tsunami propagation at minute 4 (b) for simulation model 3

12 6 Conclusion 1. Considering the trend of seismic activity during in the region of Bali- Banda Basin, the frequency of a big earthquake with magnitude more than 5 Mw is predicted to get higher along the Bali-Banda Basin. 2. Based on the earthquake density rate, the Flores to Banda Basin is the most seismically active region compared to the Bali Basin. 3. The Bali- Banda Basin earthquake recurrence times vary for each magnitude. For magnitude 6.5 Mw, the recurrence times are about 7-35 years, while for the magnitude 7 and 7.5 Mw are about and years respectively. 4. According to the both tsunami history and tsunami simulation the north part of Flores Island is the most tsunami risk experienced area compared to the north of Bali, Sumba and Sumbawa. 5. Due to the distance of the earthquake source to the coast, it is predicted that the tsunami at the north coast of Flores will be a very local tsunami. Therefore, it is important to educate people in that area to understand the very local tsunami characteristic and prepare whether the Tsunami Early Warning System will run well if the tsunami occurred. 6. Although the historical tsunami did not record a tsunami event in the north coast of Bali, Lombok, and Sumbawa Island, but the length dimension of the back arc in the north of the islands is considered as a source of an earthquake generated tsunami with a magnitude more than 7 Mw. Due to the tsunami simulation this area is also predicted will have a very local tsunami experience. References Ben-Avraham, Z and K.O. Emery (1973), Structural framework of Sunda Shelf, Bull. Am. Assoc. Pet Geol., 57, pages Curray, J.R., Shor, G.G., Raitt, R.W., and Henry (1977), Seismic refraction and reflection studies of crustal structure of the eastern Sunda and Western Banda Arcs, Journal of Geophysical Research., 82 No 17 pages Gutenberg, B., and Richter, C.F (1954), Seismicity of the earth and associated phenomena, Princeton Univ, Press, Princeton N, J., 2nd edition. Hinschberger, F., Malod, J, A., Dyment, J., Honthaas, C., Réhault, P., and Burhanuddin, S (2001), Magnetic lineations constraints for the back-arc opening of the Late Neogene South Banda Basin (eastern Indonesia), Tectonophysics., 333, pages Imamura, F., Gica, E., Takahashi, T., and Shuto, N (1995), Numerical simulation of Flores Tsunami, interpretation of tsunami phenomena in northeastern Flores Island and damage at Babi Island, Pageoph., 144, pages Irsyam, M., Sengara, W., Aldiamar, F., Widiyantoro, S., Triyoso, W., Natawidjaja, D.H., Kertapati, E., Meilano, I., Suhardjono., Asrurifak, M., and Ridwan, M (2010), Ringkasan Hasil Studi Tim Revisi Peta Gempa Indonesia 2010.

13 Mamoru, N (2006), Source Fault Model of the 1771 Yaeyama Tsunami- Southern Ryukyu Island Japan Inferred From Numerical Simulation, PureAppl. Geophys., 163, McCaffrey R and Nabale, J (1987), Earthquakes, gravity, and the origin of Bali Basin: an example of an ascent continental fold and thrust belt, Journal Geophysics Res., pages Mogi, K. Magnitude-frequency relationship for elastic shocks accompanying fractures of various materials and some related problems in earthquakes, Bull. Earthquake Res. Inst. Univ. Tokyo., 40, pages , Prasetya (1992), The Bali-Flores Basin, Proceedings Indonesian Petroleum Associatio. Satake, K., and Imamura, F (1995), Introduction to Tsunamis: , Pageoph, nos 3/4. Silver, E.A., Reed, D., and McCaffrey, R (1983), Back arc thrusting in the eastern Sunda Arc, Indonesia: a consequence of arc-continent collision, Journal of Geophysical Research., Vol 88, no 89, pages Utsu, T (1965), A method for determining the value of b in a formula of log N=a-bM showing the magnitude frequency relation for earthquakes, Geophys. Bull. Hokkaido Univ., 13 pages Wiemer, S (2001), A software package to analyze seismicity: ZMAP, Seis. Res. Lett., 72, pages Yeh, H., Imamura, F., Synolakis, C., Tsuji, Y., Liu, P., and Shi, S (1993), The Flores Island Tsunamis, EOS., Vol. 74, No 33.