Cumulative absolute absement for magnitude determination in earthquake early warning system using low-cost sensors

Through the utilization of P-Alert network data from Taiwan, this study endeavors to estimate earthquake magnitude (Mcaa) using the cumulative absolute absement (CAA) methodology across varying window lengths after the arrival of P-wave. It is differentiated that even the proximity of the nearest 12 stations to the epicenter results in robust magnitude estimations. Notably, the standard deviation between the estimated Mcaa and the moment magnitude (Mw) using 12 stations decreases with the increase in window length and is found minimum for 5s window length. For 3s window the variation between Mcaa and Mw is found ±0.385, whereas, for 5s window it is ±0.313. Consequently, the estimation of Mcaa remains reliable. The magnitude Mpd is alternatively deduced from Pd, utilizing the closest 12 stations situated near the epicenter. The standard deviation of the order of ±0.412 is observed between the estimated Mpd and Mw for 3s window, whereas for 5s window it is ±0.281. A difference is observed using Mpdand Mcaafor comparison with Mw. The standard deviation error decreases for Mcaaand Mpd with increase in window length. While Mpd performs better under a 5s window scenario, it tends to underestimate the magnitude of an earthquake with a magnitude of Mw 7.0. On the other hand, CAA surpasses Pd in magnitude estimation, though with a slightly higher standard deviation compared to Mcaa. As a result, Mcaa is considered a more reliable magnitude indicator.