Tropical cyclones like typhoons may invoke imagery of violent winds and storm surges flooding coastal areas, but with the heavy rainfall these storm
Tropical cyclones like typhoons may invoke imagery of violent winds and storm surges flooding coastal areas, but with the heavy rainfall these storms may bring, another major hazard they can cause is landslides — sometimes a whole series of landslides across an affected area over a short time. Detecting these landslides is often difficult during the hazardous weather conditions that trigger them. New methods to rapidly detect and respond to these events can help mitigate their harm, as well as better understand the physical processes themselves.
In a new study published in Geophysical Journal International, a research team led by the University of Tsukuba developed a new method to identify and locate landslide events based on seismic surface wave data and applied this method to detect landslides in Japan associated with the transit of Typhoon Talas in 2011.
As study first author Professor Ryo Okuwaki explains, “Our surface-wave detection to locate seismic events was based on the AELUMA method, short for Automated Event Location Using a Mesh of Arrays. One hundred and three seismic stations across Japan were divided into triangular subarrays, and data from the days of the typhoon were analyzed to differentiate earthquake-related events from landslide signals.”
Using this method, multiple landslides that occurred during Typhoon Talas were identified, including one in the Tenryu Ward of Shizuoka Prefecture, about 400 km east of the typhoon’s track. In 2011, it took 3 days for this landslide to be detected, after the storm had cleared and conventional observation methods became possible. This shows the potential usefulness of applying this new method to more rapidly identify landslide events. The Tenryu landslide was much smaller than landslides previously identified based on globally recorded surface waves, and it was detected as far as 3,000 km from the epicenter using the new approach.
According to Professor Okuwaki, “We found that both small and large landslides may follow the same empirical scaling relationships. This allows previous research based on larger landslides to be applied to better understand the behavior of smaller landslides detected by using our novel method, which will have important implications in further research.”
This new method, based on data from a sparse seismic network, is a promising step forward for monitoring landslide occurrences down to a scale of about 100 meters across a broad region in real time, which may help with development of emergency alert technology in the future.