Application of waveform stacking to the Tainan and Ilan low-cost local early warning (EEW) arrays in Taiwan

Abstract:

Taiwan has been constantly threatened by large, damaging earthquakes. Nowadays, EEWS is a useful tool for seismic hazard mitigation， and it is practical in Japan, Taiwan, Mexico, and the southern California (Kanamori, 1997). There are two types of EEWS. Generally, the front-type warning has higher accuracy than the on-site one because of the use of more stations and longer time window. On the other hand, the on-site type warning has the shorter reporting time, but lower accuracy because of using only one single or couple stations. Based on the on-site process, we installed two local early warning arrays in Ilan and Tainan city of the newly established EEW network and adopted the waveform stacking method to enhance the coherent part of early arrivals.

We selected 15 events in 2013 of Central Weather Bureau earthquake catalog, and used the M_Pd method to estimate the magnitude. Since the EEW network is newly established and the accelerometers (named as Palert) are all installed on the building wall instead of the free surface, the present Pdattenuation relationships have to modify to account for the building effect to get the new regression equation.

We show that the accuracy of magnitude estimation (M_Pd) for the on-site EEW mode by using the array stacking method can be significantly improved. Since the station density of the Palert network is high, more arrays can be formed especially in the metropolitan areas.

Key words: EEW, EEW array, waveform stacking.