PNSN PNW ShakeAlert Demonstration System Workshop

February 17th, 2015 at the University of Washington

Pacific Northwest Earthquake Early Warning Initial Stage

Meeting notes published February 18, 2015

The Pacific Northwest Seismic Network (PNSN) will start sending Earthquake Early Warning (EEW) Alert messages from the UW to a dozen or so regional beta testers via the Java application UserDisplay. This is a training exercise both for the PNSN and the beta testers. EEW PNSN research and development project is part of the USGS-coordinated West Coast EEW System (WCEEW) system under development. This information sheet provides a brief overview on how EEW works now in Washington and Oregon.

How does EEW work? (See also EEW Fact Sheet 2014-3083)

EEW is made possible by three main contributing factors:

1) Less destructive P waves travel faster than the more destructive S waves, and so will arrive first at any given location.

2) A dense seismic station network near the earthquake source can quickly detect seismic waves well before the more significant shaking will arrive at more distant population centers.

3) Data transmission to PNSN, processing at PNSN, and distribution to the end user is very fast relative to seismic travel times.

The amount of warning time increases rapidly with one's distance from the epicenter, and there will be a minimum distance from the hypocenter of the earthquake within which early warning is not possible. For a more detailed picture of the system, please read the USGS implementation plan for WCEEW.

What elements of EEW will the PNSN implement at this stage?

This is the first implementation of an EEW system in Washington and Oregon. We will provide alerts using the ElarmS algorithm of the CISN ShakeAlert suite. This algorithm uses data from the PNSN’s regional seismometers to rapidly detect an earthquake and estimate its location and magnitude. The ElarmS program sends XML-formatted “Alerts” to a Decision Module that determines whether the alerts are “reliable” and, if so, passes them to the end-user via the UserDisplay client app. The UserDisplay is configurable by the end-user to provide an estimate of time and level of shaking intensity at a specified locale.

ElarmS will only provide accurate alerts for earthquakes up to about magnitude 7. Future refinements of the system will allow alerts to be updated in the case that earthquakes grow to magnitude 9 and beyond.

We expect to send alerts for all earthquakes larger than M3 in the states of Oregon and Washington west of 121˚ W (approximately the Cascade range). This is the area where station density is sufficient to develop and transmit warnings before strong shaking arrives and where the most people live and work. Warnings for Vancouver Island and coastal northernmost California are also possible. Warnings for California are already available in prototype stage.

What are the goals of this release?

We are providing Earthquake Early Warnings to a restricted group of beta-users to obtain input from stakeholders to improve messaging, work out possible kinks in the delivery mechanism, demonstrate the reliability of the system, and to learn how organizations might reduce losses utilizing alerts from the future “operational” WCEEW system we are working to develop.

What is the track record of ElarmS alerts to date (in PNSN)?

Real-time and playback tests of ElarmS with Pacific Northwest earthquakes have shown our system to be fairly accurate and reliable, however, false alerts and missed events are possible. Details will be presented at the Feb 17^th workshop. Seismic network and telemetry improvements, continued software development, and stable operational funding will be required for the system to become robust enough to take actions based on the alerts it generates.

Who can participate in this project?

The PNSN has invited about 15 diverse organizations that have attended briefings and have agreed to assist the PNSN in the development of the WCEEW system. These partners understand that false alerts are probable in this development period and will not allow the experimental alerts to escape the test group. The size of the test group will be very limited for at least a year. However, we will be making a list of interested organizations that may be asked to join the development group once we are ready to expand it. Ultimately, the USGS and Canadian partners will need to be funded to run the WCEEW system in an operational manner for the warnings to be robust and available to all.

A summary of the West Coast EEW effort is online at this link.

In this workshop, we will discuss the most useful path forward to develop EEW, and we plan to forge partnerships with a half dozen companies to test the prototype that is under development.