Note: $26.4M per year is needed to fully fund the operational costs of the West Coast system. What we have today is not a fully functional system!
Earthquake Early Warning
Pre-Conference Session
Partners in Preparedness Conference
April 17, 2018
Big earthquakes are not “points.” Length of the fault and larger faults are more complicated. We won’t know how large the earthquake will be, just based on the initial detection. The rupture moves around two miles per second.
How much warning will I get? It depends where you are. The ability to detect an earthquake depends on the number of stations. Detection requires four stations detecting the earthquake.
You can warn people:
- Drop, cover and hold
- Move away from hazardous areas
- Secure delicate medical procedures
- Slow/stop trains, traffic, aircraft
- Elevators go to next floor, doors open [Eric’s item]
- Close values, stop pumps, processes
- Equipment to safe mode
They were shooting for limited public alerts in 2018. Development has been underway for 10 years.
Alert generation is USGS. Delivery of the alert is with public and private partners.
Average cost of bringing a seismic station online is $58K. Optimum spacing is 20 kilometers and for urban areas, 10 kilometers.
PNW will be 51 percent complete
Target is a total of 1,115 stations
Institutional Users (Pilots)
Magnitude M3.5+
For Public Alerts:
- M4.5+
- They don’t want to over-alert
- Primary method is cellphones, using Common Alert Protocol (CAP)
- IPAWS mentioned and WEA
10 percent of adults don’t have phones. Then there is equity, low income and people with disabilities.
Fastest alerting is via 4G
Twitter system would be overwhelmed
They are engaging cell vendors and carriers
DataCasting being worked by CalOES
Data broadcast embedded in TV, but it needs a special receiver
- Rollout in 2018. Limited public warning.
- Several pilots are operating now and others will begin
- Target date — end of August 2018
- Scope of public alerting depends on mass notification technologies
- Full funding to complete and operate the system reliability not yet been secured. What's being done now is limited. Need more funding. To fully operate the system will require more funding — still really in a limited mode, for now and the foreseeable future.
Two numbers, capital cost, plus $26.4M to operate the system annually. [This is full funding]
There are some “proof of concept” things with machine to machine (M2M), but today there would be no public alerting.
They have satisfied the federal government standard for stringent operating guidelines when it comes to cybersecurity.
Mexico City Seismic Warning System — FEMA IX Briefing
Two earthquakes in 2017. One offshore and one closer to Mexico City. They looked at their warning system.
1985 earthquake killed 10,000 people
1986, they moved to create an organization to develop/run EEW system
Funding was originally provided by both state and federal governments. Now it is only state funded.
They don’t have a national warning system. Most large earthquakes are subduction quakes off the coast.
Most states don’t have the financial means to support a warning system.
Their early warning system was tied to radios hardwired into schools. The message was to get people to run out of buildings. Their messaging is not Drop, Cover and Hold
They have been unable to get the cell companies interested in partnering.
They are running their system on about $1.1M U.S. per year.
They had a false alert the day before the 8.5 mag earthquake. It was a technician’s error.
The 8.5 earthquake had two minutes' warning. The sirens sounded for one minute, then shut off. Some people re-entered buildings.
New Topic:
Universal Studios has been part of the pilot for years. Some M2M aspects. Also using an app — testing. The alert is also going over their radios. 15K-60K visitors on any one day, depending on time of year.
Working considerations:
- Shaking intensity
- Alert tone — still a big issue.
- Verbiage, internal vs. public facing
- Countdown vs. no countdown
- Information Security
Intensity shaking, looked at 4.5M, now looking at a lower level, to 2.5M — since people can feel it.
Early Warning — Next Steps:
- They are looking at PA systems (fixed and VoIP phones)
- Automatic generator starts
- Elevator/conveyance intervention
- Critical equipment shutdowns/broadcast shifts
- Utility shutdowns
- Ride interventions
- Elevator/conveyance interventions, varying ages of elevators/infrastructure. Expensive to add to old systems
- Banked high-rise elevators
- PA systems: Some PA systems are tied into fire alarms
- Latency in activation
- Conflicts in fire codes (verification pending)
- Policies and procedures. Cost justification process — not cheap!
- Emergency stops — on rides cause other issues, like rescues …
- Alerts go only to employees, but not to guests
- Many liability questions — false alarms?
- There is a bit of a research quandary. Most of the research is being done in foreign countries.
Mike Ripley, NBC Universal
Mike.ripley@nbcuni.com
818-777-5962
City of Los Angeles
Earthquake early warning — Hunter Owens, Senior Data Scientist, Information Technology Agency (Not an emergency manager)
Mayor proclaimed that in 2018 the city will have an earthquake early warning system.
City Team:
- Mayor’s Office of Public Safety
- Information Technology Agency
- USGS
- Amazon Web Services
- Apple and Google
- AT&T
50,000 users with the city of LA
Open source application
SMS
How do you get a message to 4M people quickly?
They are working on Early Warning Labs on the issue of elevator pilot
Hunter.owens@lacity.org
New topic: San Francisco:
City of San Francisco, Department of Emergency Management
Alicia D. Johnson, Resilience and Recovery Manager
They have:
- Outdoor warning sirens — 49 square miles (citywide) weekly tests
- Opt-in Public Notifications (90K people signed up, primarily text to mobile)
- Wireless Emergency Alerts
- Using Hazard Mitigation Grant Program
- Local cost match
- Warning is vital, but often not fundable by traditional sources
