There has been an increasing number of studies on the tsunami hazards in the South China Sea region since the 2004 Sumatra earthquake. Many of them are carried out based on tsunami scenario simulations, which adopt seismic source models constructed from scaling relations between seismic magnitude/momentum and rupture parameters. Various sets of scaling relations have been proposed on the basis of different earthquake catalogues. In this study, we perform synthetic tests to evaluate the impact of scaling relations on the generation and propagation tsunami waves in the South China Sea. Results show that the range of the affected coastline can be significantly different for an earthquake of the same magnitude using different scaling relations. Additionally, the maximum tsunami wave height near major cities may vary as large as two times. Thus, it is worth further research on the choice of scaling relations for tsunami hazards assessment and the building of early tsunami warning system.
The Manila subduction zone was identified (Kirby et al., 2006) as having high potentials to generate hazardous tsunamis in the USGS (the United States Geological Survey) tsunami source workshop (Liu et al., 2007). As shown in Fig. 1, seismic events are very active near the Manila Trench. Tsunami from the Manila Trench is a potential threat to coastal countries around the South China Sea (SCS), such as China, Vietnam, Philippines et al. Although studies on tsunami scenarios have been performed (Liu et al.2007, 2009; Wu & Huang, 2009, Megawati et al., 2009, Nguyen et al., 2014), the rupture model and mechanism of the Manila trench is far away from being enough.
Eyes were turned to the Manila subduction zone after the 2004 Sumatra tsunami. Basic geometry and orientation parameters were provided by Kirby et al.(2006). But the parameters cannot be employed directly to performing numerical simulations, for the rupture width, the focal depth and the dislocation were not provided. Liu et al.(2007, 2009) provided the source parameters of an earthquake scenario Mw8.0 based on the geometry and orientation parameters of Kirby et al.(2006). The rupture width was identified as 35 km by squaring the region of an earthquake Mw7.3 and its aftershocks in 1999. The dislocation was calculated from the scaling relation of Wells & Coppermith (1994). However, the slip of the sub-fault E6 is the largest among the six sub-faults, which conflicts with geodetic data (Li et al., 2016). Based on geodetic data from Yu et al.(1999), a complex rupture model was provided by Megawati et al.(2009) by splitting the Manila fault into 33 pieces of elements. An earthquake, whose moment magnitude reaches Mw9.4, is probably to dilacerate the whole Manila Trench and brings extreme tsunami. Wu & Huang (2009) also provided a rupture model of an earthquake Mw9.35 by simply comparing the geometric similarity of the Manila fault to the fault of 1960 Chilean earthquake, the 1964 Alaska earthquake, and the 2004 Sumatra earthquake. Nguyen et al.(2014) redesigned the subfaults' geometry and orientation parameters. In fact, except the geometry and orientation parameters of the Manila Trench that proposed by Megawati et al.(2009), the rest models of earthquake source parameters do not meet the scaling relations (SRs) of megathrust fault. Apart from providing all parameters of an earthquake source, they also studied the tsunami impact on the eastern coast of Vietnam (Nguyen et al., 2014), Singapore (Huang et al., 2009) and Taiwan (Wu & Huang, 2009). Liu et al.(2007, 2009) attempted to establish an early tsunami warning system in the SCS so as to mitigate disasters to coastal land areas around the SCS.