An on‐site alert level early warning system for Italy

Abstract

An earthquake early warning (EEW) system is a real‐time seismic monitoring infrastructure that has the capability to provide warnings to target cities before the arrival of the strongest shaking waves. In order to provide a rapid alert when targets are very close to the epicenter of the events, we developed an on‐site EEW approach and evaluated its performance at the Italian national scale. The system is a P wave‐based method that measures in real time two parameters: the initial peak displacement (Pd) and the average period (τc). As output, the system provides the predicted ground‐shaking intensity at the monitored site, the alert level, and a qualitative classification of both earthquake magnitude and source‐to‐receiver distance. We applied the on‐site EEW methodology to a data set of Italian earthquakes, with magnitude ranging from 3.8 to 6, and evaluated the performance of the system in terms of correct warning and lead times (i.e., time available for security actions at the target). The results of this retrospective analysis show that for the large majority of the analyzed cases, the method is able to deliver a correct warning shortly after the P wave detection, with more than 80% of successful intensity predictions at the target site. The lead times increase with distance, with a value of 8–10 s at 50 km and 15–18 s at 100 km.

Comparing the Performance of the NEEWS Earthquake Early Warning System Against the CWB System During the 6 February 2018 Mw 6.2 Hualien Earthquake

“…When the tolerance range was lowered to a mere ±0.5 level, in a manner similar to Meier's suggestion (Caruso et al, 2017), the same improvement in classification performance was obtained for both the NEEWS and CWB EEW systems. In other words, if an alert was issued for a site that had an intensity above I i À 1, we considered it a TP; only if the measured ground motion was higher than one level below the threshold did we consider it an FP.…”

“…However, the classification performance of the CWB EEW system did not improve, even after we introduced the tolerance range. When the tolerance range was lowered to a mere ±0.5 level, in a manner similar to Meier's suggestion (Caruso et al, 2017), the same improvement in classification performance was obtained for both the NEEWS and CWB EEW systems.…”

“…The on-site EEW approach has recently been the subject of numerous studies worldwide (Allen et al, 2009;Böse et al, 2012;Carranza et al, 2013;Caruso et al, 2017;Emolo et al, 2016;Hsu et al, 2013;Kanamori, 2005;Nakamura, 1998;Odaka et al, 2003;Peng et al, 2015;Wu et al, 2013;Zollo et al, 2010). The on-site EEW approach has recently been the subject of numerous studies worldwide (Allen et al, 2009;Böse et al, 2012;Carranza et al, 2013;Caruso et al, 2017;Emolo et al, 2016;Hsu et al, 2013;Kanamori, 2005;Nakamura, 1998;Odaka et al, 2003;Peng et al, 2015;Wu et al, 2013;Zollo et al, 2010).…”

“…The National Center for Research on Earthquake Engineering (NCREE), Taiwan, has been developing an on-site EEW system since 2009 to provide increased lead times for regions with a higher likelihood of sustaining damage (i.e., regions near epicenter). The on-site EEW approach has recently been the subject of numerous studies worldwide (Allen et al, 2009;Böse et al, 2012;Carranza et al, 2013;Caruso et al, 2017;Emolo et al, 2016;Hsu et al, 2013;Kanamori, 2005;Nakamura, 1998;Odaka et al, 2003;Peng et al, 2015;Wu et al, 2013;Zollo et al, 2010). The NCREE implemented an on-site Early Earthquake Warning System (NEEWS) by first establishing several stations in elementary schools with a limited budget.…”

On-site earthquake early warning: a partially non-ergodic perspective from the site effects point of view

“…Different studies proposed empirical relationships for on-site EEW (Kanamori 2005;Wu & Kanamori 2008;Böse et al 2009;Zollo et al 2010;Picozzi 2012;Brondi et al 2015;Colombelli et al 2015;Caruso et al 2017). The data sets used for the calibration generally include recordings from multiple stations in a given region, such as (Caruso et al 2017) for Italy; (Wu & Kanamori 2005) for Taiwan; (Wu et al 2006) for California, or include data from different regions (Wu et al 2006;Zollo et al 2010). Generally, empirical relationships are calibrated without considering the station-to-station variability (Wang & Wu 2014), thus incorporating the response of the installation site into the ground motion aleatory variability.…”