Riken-led team to model damage from compound disasters

The Yomiuri Shimbun

A team of researchers from Riken and other institutions plans to model the impact of a compound disaster such as a Nankai Trough earthquake and rain-related disaster.

The team will utilize Fugaku, the world’s fastest supercomputer, at Riken’s Center for Computational Science in Kobe and so-called digital twin technology to create virtual models of municipalities. The technology is designed to create digital versions of buildings, towns and other real-world objects, as identical to the real thing as if they were twins, reflecting data on such factors as the movement of people and traffic volume.

According to government estimates, there is a 70% to 80% chance that a magnitude-8 to -9 Nankai Trough earthquake will occur within 30 years. If it happens, the government estimates there will be up to 230,000 casualties and economic damage equivalent to more than ¥170 trillion.

Rain-related disasters have increased due to the effects of global warming and other factors, prompting concerns that an earthquake combined with heavy rains could result in a compound disaster.

The University of Tokyo, Kyushu University and the Japan Agency for Marine-Earth Science and Technology are also participating in the research.

The team will create digital representations of towns based on residential maps and topographic data, among other sources.

By evaluating the impact of a maximum-scale earthquake in a digital model, the team aims to predict how individual buildings would be affected and in which areas soil liquefaction would occur. It will also consider the impact of heavy rains on earthquake-damaged land, assessing the risks of landslides and flooding, among other risks.

The economic impact will be calculated based on corporate and other data held by private credit research companies to evaluate how a disaster might affect things such as evacuation behavior.

By identifying in advance buildings and bridges with insufficient earthquake resistance and areas with soft ground, appropriate investments could be made, thus minimizing the extent of damage in the event of a compound disaster and supporting efforts toward swift recovery and reconstruction.

The team will first make digital representations of urban areas in the Kansai region near the expected epicenter of a Nankai Trough earthquake.

The team is said to be able to make predictions for areas with about 10 million residents with the Fugaku supercomputer, which has the capacity to perform the massive calculations necessary for such models.

The team is considering including an outbreak of an infectious disease in the study in response to the COVID-19 pandemic.

Based on the expected evacuation behavior of affected residents, the team could potentially forecast the spread of infections and the number of seriously ill patients.

The information will also be used to create a hazard map of the Tokyo metropolitan area, based on simulations of an earthquake with a focus directly under the capital.

“If we know that a certain bridge will be out of service, we can make evacuation and recovery plans based on that assumption,” said team leader and disaster prevention expert Satoru Oishi of the Riken research institute. “We want to eliminate the idea of unexpected damage and help create a society that can recover from a complex disaster at an early stage.”