Alert exercise reveals some strengths, some weaknesses for communications during disaster – Sonoma West

Challenges with cell carriers, federal systems noted

In the wake of the October 2017 firestorm, questions were raised about the emergency alert system in the county and its ability to usefully warn individuals about impending disasters or other emergency situations. On Sept. 10 and 12 Sonoma County undertook an Alert Warning Functional Exercise. It was done through the Sonoma County Emergency Operations Center and took approximately three hours on each day. On Nov. 20 they released a comprehensive report on the exercise and their findings.

“In the event of an emergency, SoCoAlert, wireless emergency alerts (WEA), the emergency alert system (EAS) and other emergency messaging will direct citizens to our emergency information website (socoemergency.org) for more information,” stated the report. “This exercise was conducted to ensure local emergency public safety and emergency management organizations have a clear understanding of how alerts would perform in the varied threat hazards, topography, demographics and urban densities of Sonoma County.”

A Federal Communications Commission (FCC) waiver was required for for Sonoma County to conduct the “live code” test. The waiver outlined the parameters for “public outreach per the plan outlined in the request, a WEA message less than 90 characters and identified as only a test, and that any post-test analysis or reports consider customers’ privacy.”

The test was four phases, including pre-and post-test planning and assessment and well as an “All Call” (a test of the SoCoAlert system) and a live code test of WEA and EAS.

All Call test

On Sept. 10, SoCoAlert attempted 426,390 phone calls between 6:04 p.m. and 8:53 p.m. The system made 2,500 calls per minute.

“Response from the public was greater than expected. This is due partially to the message format. A number was provided in the message for more information. Many callers did not listen to the entire message and just called the number to ask why the system had been activated and if they needed to do anything. Additionally, six people called the Sheriff’s 9-1-1 dispatch stating that they had received the call, did not listen to the message and immediately called 9-1-1 to find out if they needed to do anything.”

“At the peak, there were seven call takers in the Emergency Public Information Hotline room answering calls. It is recommended that in the future if a message provides a source for additional information, specifically if it involves a phone number, that that number be up-staffed or a prerecorded message be played first to help clarify the intended message to callers,” stated the report.

In addition, it was noted that the county may need to consider acquiring phone numbers from the major carriers more frequently than its current once a year, in order to keep their ability up to reach people more up to date.

Feedback of the All Call test indicated citizens found that the sign up website for SoCoAlert was difficult to navigate and not mobile friendly. In addition, the phone number listed for assistance went straight to voicemail, something the report noted should not occur in the future.  

There was an issue with Caller ID as well, as the alert came from an 866 number. Landlines got a caller ID that stated “Code Red.”  It was recommended that in the future the phone number come from a 707 number and the Caller ID state “EMERGENCY.”

Live code Wireless Emergency Alert and Emergency Alert System Test

On Sept. 12, five geographical areas within the county took park in a WEA test. They were chosen to represent different threat hazards, topography, demographics and urban densities. Right after the WEA was activated, local TV and radio stations then delivered an EAS. Messages were delivered in English and Spanish and there was substantive “bleed over” during the WEA test to other geographical areas.

The targeted areas included Guerneville, selected for a river flood hazard scenario in a mountain area with a rural community in the unincorporated area; Glen Ellen/Kenwood chosen for its Wildland Urban Interface; Healdsburg chosen to test an “entire city boundary” and because it has multiple hazards with a high tourist population area; Penngrove, chosen to test a potential train accident/hazardous materials release scenario; and Roseland, selected for being an urban setting with a high population density with a significant Spanish speaking population.

In addition, there was a countywide EAS issued about 15 minutes after the other alerts had been sent out.

A public survey revealed that while citizens were appreciative that the county conducted the test, there was “confusion between the SoCoAlert All Call Monday night and the WEA and EAS test; most of the respondents that stated they did not receive a WEA message requested that the county conduct a whole county test to ensure their phones work; several hundred commented that they received the message in Spanish first and requested that it be issued in English first; some people were concerned that they received a WEA and were next to someone with a cell phone that did not receive a WEA or vice versa; and that WEA did not override the silent feature on cell phones.”

An additional issue identified during the test is the seeming lack of cooperation from the major cell carriers, both in relation to overriding the silent feature on cell phones and in creating appropriate geographic constraints to the messages sent out. According to the report, “it is almost impossible to target any area with confidence.”

Conclusions

The test revealed gaps and deficiencies in local warning systems, with public education rising to the top as an item of “critical importance.”

While the SoCoAlert system showed significant increases in participation from the community, protocols for more efficiently targeting sending messages will still need to be created.

“This exercise verified the need for improvement of the federal WEA system capabilities as it relates to conditions on the ground in Sonoma County. The exercise findings indicate that significant challenges remain regarding the effective use of the federal warning systems, including WEA and EAS. These challenges include incomplete and inconsistent alerting across telecommunication providers, significant bleed over when targeting specific geographic locations and the performance of the technology across various wireless devices. These shortcomings significantly conflict with the public’s expectations for service,” concluded the report. “Local government emergency managers will have to continue to take into account these shortcomings in developing and conducting alert and warning efforts. It is critical that local governments, Cal OES, FEMA and the FCC engage telecommunications providers to continue to improve the reliability and effectiveness of these systems.    

According to the report, “There remains no perfect solution to the complicated challenge of alert and warning. This exercise demonstrated that each of the tools available have specific strengths and weaknesses. Alert and warning officials cannot afford to rely on only one system to communicate in a time of disaster. Everyone, local government and the public, needs to understand how the various emergency warning systems work and how they connect to them.”

The findings outlined in the report will be used to draft an improvement plan for emergency communications in the county.