How Often Can Earthquake Early Warning Systems Alert Sites With High‐Intensity Ground Motion?

Abstract

Although numerous Earthquake Early Warning (EEW) algorithms have been developed to date, we lack a detailed understanding of how often and under what circumstances useful ground motion alerts can be provided to end users. In particular, it is unclear how often EEW systems can successfully alert sites with high ground motion intensities. These are the sites that arguably need EEW alerts the most, but they are also the most challenging ones to alert because they tend to be located close to the epicenter where the seismic waves arrive first. Here we analyze the alerting performance of the Propagation of Local Undamped Motion (PLUM), Earthquake Point‐Source Integrated Code (EPIC), and Finite‐Fault Rupture Detector (FinDer) algorithms by running them retrospectively on the seismic strong‐motion data of the 219 earthquakes in Japan since 1996 that exceeded Modified Mercalli Intensity (MMI) of 4.5 on at least 10 sites (Mw 4.5–9.1). Our analysis suggests that, irrespective of the algorithm, EEW end users should expect that EEW can often but not always provide useful alerts. Using a conservative warning time (tw) definition, we find that 40–60% of sites with strong to extreme shaking levels receive alerts with tw > 5 s. If high‐intensity shaking is caused by shallow crustal events, around 50% of sites with strong (MMI~6) and <20% of sites with severe and violent (MMI ≥ 8) shaking receive alerts with tw > 5 s. Our results provide detailed quantitative insight into the expected alerting performance for EEW algorithms under realistic conditions. We also discuss how operational systems can achieve longer warning times with more precautionary alerting strategies.

Commentary: The Role of Geodetic Algorithms for Earthquake Early Warning in Cascadia

“…Moreover, Ruhl et al [2019] initially claimed that warning times of over 100s would be common for sites that experienced strong to severe shaking in large Cascadia megathrust earthquakes, but then published a correction (available online) which fixed errors that had previously caused the warning times to be significantly over estimated. In contrast, recent evaluations of seismic EEW algorithms in great subduction zone events in Japan by Meier et al [2020a] showed warning times for severe to violent shaking were typically in the realm of ~30 seconds or less even for M>8.0 subduction megathrust ruptures.…”

“…general result. To date, the best estimates of warning time for seismic algorithms that estimate location, magnitude, and potentially rupture length for earthquakes up to M9 come from comprehensive evaluations using Japanese data and fall in the 0-40 s range [Meier et al, 2020a].…”

“…Both theoretical (Minson et al., 2018) and retrospective studies of past earthquakes (Meier, Kodera, et al., 2020) have shown that it is very difficult to provide long warning times for high intensity shaking even for algorithms that track the growing rupture length. Alerting at low predicted intensity levels in M8‐9 subduction zone earthquakes does allow many sites that eventually experience severe‐to‐violent (MMI 8–10) shaking to potentially achieve 10–40 s of warning with seismic algorithms (Meier, Kodera, et al., 2020). However, the amount of warning time is not constant for a particular user and instead depends strongly on a user’s location relative to the particular earthquake.…”

The transformer earthquake alerting model: a new versatile approach to earthquake early warning

“…All methods degrade with increasing PGA levels, particularly for Japan. This degradation is intrinsic to early warning for high thresholds due to the very low prior probability of strong shaking (Meier 2017;Minson et al 2019;Meier et al 2020). Furthermore, shortage of training data with high PGA values results in less well constrained model parameters.…”

“…Propagation based methods, like PLUM (Kodera et al 2018), infer the shaking at a given location from measurements at nearby seismic stations. Predictions are more accurate, but warning times are reduced, as warnings require measurements of strong shaking at nearby stations (Meier et al 2020).…”