[EAS] Good article on EAS in AK

[Sean Donelan]

The article is very good, but of course the journalist used a layperson's 
terminology for some of the details.  The article includes links to the 
source materials, including the after action reports and emails between 
the various parties.

The deep details are covered in the source materials.

Overall, Alaska's stakeholders did a pretty good job. Yes, multiple things 
failed. But Alaska had more redundancy in their processes than many other 
states. In spite of the problems, some warning reached nearly all of the 
impacted population.

Not perfect, some stuff needs fixing. And not unique to Alaska.

On Wed, 27 Jun 2018, Adrienne Abbott wrote:
> Rich--
> Just got a chance to read the article and even though the reporter did a
> thorough job of looking into the situation, I have several questions...Was
> the NOAA connection to IPAWS disabled because NWS is not capable of issuing
> CAP EAS messages? It's one thing to say that the NWS office "disabled" the
> IPAWS connection, but it's an entirely different thing if NWS is incapable
> of sending CAP messages. Yes, NWS can send an EAS message, but the content
> of a SAME EAS message is very limited compared to what you would receive in
> a CAP EAS message. No NWS office is issuing CAP EAS messages at this time.
> While NOAA is working on a CAP platform, it has been a very slow process.
> Nevada has volunteered to be a test site for NWS CAP EAS activations when
> NOAA is ready to roll out a product.

NOAA -> IPAWS -> EAS has been blocked since the rollout of IPAWS.

Only NOAA -> IPAWS -> WEA is enabled.

WEA is triggered at the NOAA national gateway based on specific NOAA 
codes in the NOAA Weather Wire messages. As far as I know, local weather 
forecast offices cannot directly send IPAWS CAP messages.

> Also, why did it take so long for the people involved here to notice that no
> EAS messages had been issued, including a CAP EAS message from EMNet? Why
> didn't anyone call EMNet directly to find out why they hadn't issued a CAP
> EAS activation? Why wasn't there a back up to EMNet? Why doesn't Alaska have
> a plan for using another state or the FEMA IPAWS office to issue CAP EAS
> activations?

The EOC checklist is usually the EAS was "sent." Very few EOCs actively 
monitor whether the EAS was "received" beyond a few TV's in the state EOC 
tuned to local TV stations or a weather radio. Most EOCs seem to find out 
about WEA messages when a personal cell phone sounds the alert. I've only 
seen one EOC with an old Sage grey box still hooked up as a confidence 
check.

Tsunami Warnings aren't typically originated by state EOCs. See the NOAA 
after action report for the gory details about what happened with the 
alternate WFO. As seen in California, Tennessee, Oregon and other states, 
state EOCs don't have much practice originating alerts.

Also refer to the on-going debate about coordinating which warnings 
should be originated at the local, regional, state or federal level.

Tornado, Tsunami -> NOAA
Volcano, Wildfire -> local county?
Inbound missile -> who knows?

Every roundtable held by the FCC and FEMA seems to have emergency managers 
surprised by who is expected to originate which kind of warnings.

As far as I know, the FEMA IPAWS office never originates "actual" alerts 
on behalf of anyone. The FEMA IPAWS office does a great job of helping 
originators debug problems, fix their alert problems, and eventually the 
originator sends the message themself. But I haven't seen an actual 
(non-test) alert directly from FEMA besides the EAN.

Note: I saw the weird Puerto Rico alerts with the fema.gov email address 
as the sender, but I'm assuming Puerto Rico just forgot to change their 
alert software template.

> As an example, our IPAWS CAP provider recently developed a problem and we
> had two CAP EAS test failures within a 12-hour period. In both cases, the
> CAP EAS originators reported their test failures to me. I called our
> provider and the problem was fixed within hours, even though it was a
> weekend.

I think I recall in your situation the computer software threw up 
an IPAWS error message while sending the test message. I wouldn't be as 
certain how quickly originators notice lack of receipt of the test 
message without that computer error message. A missing RMT gets noticed 
quickly by receipients because its scheduled, but other missing tests or 
missing actual alerts, maybe a few days or never.

Again, see the source materials for the gory details about what happened 
from the vendor.