Effects of Applied M-Filter on the Indonesian Tsunami Early Warning Application (Ina-TEWA) Performance
Madlazim, Tjipto Prastowo, Supardiyono

Universitas Negeri Surabaya

Abstract

A reliable method of detecting tsunami generation in the ocean for effective tsunami early warning requires rapid and accurate tsunami-discriminant calculations, which include rupture duration (Tdur), the duration exceeds 50 seconds (T50ex), and the dominant period (Td) of first P-wave arrivals extracted from seismic signals induced by earthquakes of varying magnitudes. In this context, this study is thus aimed to analyze the effects of M-filter implemented on the existing Tsunami Early Warning Application (Ina-TEWA) performance in terms of accuracy in determining such discriminants. A method of real-time experiments was developed in this study, where tsunami parameters obtained before and after M-filter implementation were analyzed for the number of false warnings issued. A total of 300 Indonesian events striking regions throughout the country with a range of magnitudes from 5.0 to 7.9 during 2018-2019 years were examined. The results showed that prior to M-filter installation 15 of all cases considered were falsely issued, confirming for a total of 95% assessment being adequately accurate. However, accuracy in the Ina-TEWA performance was found to improve as false alerts reduce to zero when M-filter was introduced, implying a 100% level of accuracy. Hence, the use of M-filter in the Ina-TEWA makes its performance better, providing with reduction in false warnings and hence improved accuracy.

Keywords: Keywords: M-filter, Ina-TEWA, improved accuracy, false warning, tsunami parameters

Topic: Physics