Influence of the Warning Area Division on the Effect of the Propagation of Local Undamped Motion (PLUM) Method

Abstract

It is possible to obtain a detailed estimates of the expected vibration by dividing the early warning area, to achieve an optimized early warning effect. This paper studies the influence of the warning area division on the warning effect of the the Propagation of Local Undamped Motion (PLUM) method. Japan is divided into grids with side lengths ranging from 30 km to 120 km, and each grid is considered an early warning area. Using the data recorded by the National Research Institute for Earth Science and Disaster Resilience (NIED), we studied the effects of different grid sizes on PLUM warnings by simulating the "311 earthquake" in Japan. The simulation results obtained with the grid divisions are compared with the simulation results obtained using predefined warning areas, and the influence and applicability of the warning area division method on the warning effect are analyzed. The calculation results show that an increased grid size leads to an expanded warning range, but at the same time, unnecessary warnings are sent to some areas. Under the conditions of the "311 earthquake" in Japan, the prediction scores and precision of the 60 km and 70 km grid are higher, and the mean warning times is longer than those of other grid sizes; the overall warning results of these grids are the closest to those obtained with the predefined warning areas in Japan.