You are here

Tsunami Hazard in South China Sea is Likely Greater than Previously Thought

19 Oct 2016

The 2011 Mw 9.0 Tohoku earthquake came as a surprise to many. (Source: Douglas Sprott/

Scientists from the Earth Observatory of Singapore (EOS) have successfully assessed the probability of tsunami hazards for the South China Sea coasts with greater accuracy. They found that these coasts face a greater risk of tsunami hazards than previously thought.

Published on 31 Aug 2016 in the Journal of Geophysical Research: Solid Earth, the study explains how EOS scientists used an improved numerical modelling approach, one constrained by updated geophysical data from the region, to conduct the assessment.

The risk of large tsunamis in the South China Sea region primarily comes from the Manila trench that lies to the west of Luzon in the Philippines. The Manila trench system is interesting in that it shares many of the characteristics of the source areas that brought on the 2004 Sumatra-Andaman earthquake, as well as the Japan trench of the 2011 Tohoku-Oki earthquake. Both disaster events had caught many people by surprise.

Locations of historical tsunamigenic earthquakes or tsunami affected areas in the SCS. Current available data show that both the Philippines and China were more frequently affected by tsunamis historically. Major cities are marked in red dots. The main tsunami threat in the SCS comes from the Manila Trench megathrust, along which no earthquake larger than Mw 7.8 has been observed since 1560s.

Recognising the fact that tsunami hazards are frequently underestimated or unrecognised due to inadequate historical tsunami records, the EOS research team worked on this study in collaboration with Associate Professor Robert Weiss from Virginia Polytechnic Institute and State University (Virginia Tech).

Lead author Dr Lin Lin Li explained, “This study considered the complexity of how the sea floor is displaced when two tectonic plates move along a fault plain in an earthquake”. Known as a slip distribution, this element of tsunami modelling was commonly oversimplified for the purpose of uniformity, despite the reality that earthquakes rupture heterogeneously. 

Examples of slip distributions of Mw 8.0 earthquake. The slip distributions are first generated by the hybrid model [Gallovič and Brokešová, 2007] in a rectangular area (the length and width of the area are equal to the rupture area calculated by scaling relations for each earthquake magnitude) and then projected to the chosen rupture area.

Associate Professor Adam Switzer, leader of the study, said, “Using the heterogeneous slip approach, we ran nearly 30,000 scenarios with earthquake magnitudes measuring between M 7.0 to 9.0, together with earthquake return periods that were derived from the latest seismic, geodetic and tectonic data along the Manila megathrust.”

He was surprised by the effects of heterogeneity, adding that the study clearly showed that by applying the commonly used uniform slip approach in the relatively small and confined South China Sea, it led to significant underestimates of the tsunami hazard particularly for southern China, Taiwan and central Vietnam.

According to Dr Li, “what was particularly alarming were the results for the area around Hong Kong, Macau and Shenzhen in southern China, where the effect of non-uniform slip predicts a wave amplitude 20-60 % larger than the assessment based on uniform slip models for a 500-year return period.”

The findings also revealed that near the Manila trench in western Luzon and south Taiwan, the chances of a tsunami greater than 1-m in height, occurring in the next 100 years, are around 50-60 %.”

Further away in Guangdong, Hong Kong and Macau, these areas were discovered to have a 20-50 % chance of getting hit by 1 m-high tsunami waves in the next century, with the likelihood diminishing across the South China Sea to 20-30 % in Central Vietnam and to less than 5 % for the rest of the region, which includes Singapore and Brunei.

Due to the shape and gradient of the continental slope, southern China, including coastal cities like Guangdong, Hong Kong (pictured) and Macau, are in the direct path of tsunamis. (Source: Roger Wagner/

Assoc. Prof Weiss explained, “We found that, given the geophysical constraints of the Manila trench system, the worst-case earthquake scenarios are estimated to be at M 9.0. However, an earthquake measuring M 8.0 would likely generate 6-m tsunami waves locally in the Philippines, and 2-3-m waves in southern China, should it rupture at the shallow portion of the trench. Based on the geodetic information, such an event would have a return period of 122 years.”

Assoc. Prof Weiss went on to add that “studies like this highlight the importance of developing a tsunami warning system in the South China Sea”. Such a programme would complement existing tsunami warning centres, such as the Pacific Tsunami Warning Center based in Hawaii and the Japan Meteorological Agency based in Tokyo, which mainly focus on the tsunamis generated by Pacific-rim earthquakes.

Looking to the future, Dr Li noted that “fortunately, the regional tsunami warning centre based in Beijing had begun operations earlier this year in March”. This centre is supported by the Intergovernmental Oceanographic Commission and operated by the National Marine Environmental Forecasting Centre of State Oceanic Administration in China.

This paper was highlighted on AGU’s Journal of Geophysical Research webpage, as well as on their Eos blog, which draw special attention to notable journal publications. Author of this piece, Ms Wudan Yan, gives a concise summary of the tsunami risk assessment conducted by Li et al. She explains the researchers’ methodology and the findings from their study. You may read the full article here and here.

Thumbnail source: Roger Wagner