The First-Year Operation of the Plum Algorithm in the Earthquake Early Warning System of the Japan Meteorological Agency

Date: 4/25/2019

Time: 04:00 PM

Room: Puget Sound

To overcome technical difficulties in point-source earthquake early warning (EEW) approaches for earthquakes with large finite faults and multiple simultaneous events, the Japan Meteorological Agency (JMA) has begun to operate the Propagation of Local Undamped Motion (PLUM) algorithm (Kodera et al., 2018) in its EEW system since March 22, 2018. We report the first-year performance of PLUM, focusing on cases in which the system issued public warnings.

PLUM predicts ground motion without source characterization, assuming unattenuated plain wave incidence from observation points close (typically within 30 km) to a target site. The current JMA system employs PLUM and point-source algorithms and combines their results by taking the maximum. For the PLUM operation, JMA has incorporated ~400 additional stations that can only transmit observed intensities in real time.

From March 22 to December 31, 2018, the JMA system issued 13 warnings to the public (the predicted intensity threshold is 5-lower (5L) on the JMA scale, i.e., ~7 on MMI). For three out of the 13 cases, warnings were issued based only on PLUM ground motion predictions for earthquakes with the maximum observed intensity of 5L, showing that PLUM contributed to reducing the missing rate of strong motion. For the Mj 6.7 Hokkaido Eastern Iburi earthquake on September 6, 2018, which caused an observed intensity of 7 (maximum on the JMA scale; ~10–12 on MMI), ground motion predictions by PLUM exceeded the warning threshold ~3 s earlier than those by the point source algorithms. PLUM successfully calculated high-intensity ground motions in the early stage directly from observed intensities, while the point-source-based approaches took longer time to obtain accurate magnitude estimates. This indicates the high robustness of ground motion predictions using direct observation of wavefields.

Presenting Author: Yuki Kodera

Authors