How “good” are real‐time ground motion predictions from Earthquake Early Warning systems?

Abstract

Real‐time ground motion alerts, as can be provided by Earthquake Early Warning (EEW) systems, need to be both timely and sufficiently accurate to be useful. Yet how timely and how accurate the alerts of existing EEW algorithms are is often poorly understood. In part, this is because EEW algorithm performance is usually evaluated not in terms of ground motion prediction accuracy and timeliness but in terms of other metrics (e.g., magnitude and location estimation errors), which do not directly reflect the usefulness of the alerts from an end user perspective. Here we attempt to identify a suite of metrics for EEW algorithm performance evaluation that directly quantify an algorithm's ability to identify target sites that will experience ground motion above a critical (user‐defined) ground motion threshold. We process 15,553 recordings from 238 earthquakes with M > 5 (mostly from Japan and southern California) in a pseudo‐real‐time environment and investigate two end‐member EEW methods. We use the metrics to highlight both the potential and limitations of the two algorithms and to show under which circumstances useful alerts can be provided. Such metrics could be used by EEW algorithm developers to convincingly demonstrate the added value of new algorithms or algorithm components. They can complement existing performance metrics that quantify other relevant aspects of EEW algorithms (e.g., false event detection rates) for a comprehensive and meaningful EEW performance analysis.

Earthquake Early Warning ShakeAlert System: Testing and Certification Platform

“…This is in contrast to Meier (2017), who compares predicted shaking to the observed ground-motion records only at locations with seismic stations. Here, we use the ShakeMap as ground truth for the events used in the test suite because available ground-motion observations may be limited, particularly for older historic events.…”

“…We compute a threshold-based assessment to determine if an end user would undertake a correct action given a particular MMI threshold for their site. The methodology we apply is based on that developed by Meier (2017). The method compares observed with predicted ground motions at a site and classifies each site into one of four categories for a given MMI threshold as follows.…”

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

“…Based on the metrics suggested by Meier (2017), but adjusted to the conditions native to Taiwan, for the classification performance of the CWB EEW system, the TP, FN, and FP rates were 0%, 83.3%, and 16.7%, respectively. Based on the metrics suggested by Meier (2017), but adjusted to the conditions native to Taiwan, for the classification performance of the CWB EEW system, the TP, FN, and FP rates were 0%, 83.3%, and 16.7%, respectively.…”

Earthquake Early Warning ShakeAlert System: West Coast Wide Production Prototype

“…For example, users may decide to initiate action when an alert has larger uncertainty and/or at a lower alert threshold to ensure they receive a timely alert and have enough time to initiate an action before peak shaking arrives. In general, users who choose a lower alert threshold will receive longer warning times (Meier, 2017;Minson et al, 2017).…”