[EAS] Cell Phone Warning Failure

[Sean Donelan]

On Thu, 29 Jun 2017, Botterell, Arthur at CalOES wrote:
> Very true, Sean... in fact, we added a function to a telephone-dialer 
> alerting application at Contra Costa County in CA that let us preview 
> how many phone calls alerting a specified geographic area would entail, 
> and I recall referring to it frequently.  We were counting from our 
> reversed-911 database, but a dip into Census data wouldn't be hard 
> either.

The reason I found the academic example interesting was most subscription 
based notification vendors only have a 10%-20% sign-up rate.  So "mass 
notification systems" generally only notify a small, self-selected group. 
Subscription groups don't complain as much about getting alerts, because 
they signed up to get those alerts.  And can unsubscribe in most cases.

The academic use of the census data and GIS data for non-subscription 
alerts emphasized the much larger reach of the WEA/EAS alert systems at 
the time of the alert.  Even if the alert polygon was "tiny," the 
checkmark next to EAS included the census population from the entire EAS 
alert area.  If that wasn't what the alert originator intended, they could 
uncheck the EAS box before sending the alert. Likewise, WEA alerts reach 
about 2-miles to 20-miles from a tower, even if the polygon is smaller.

Although that wasn't the purpose of the academic study, I thought having 
an audience size feedback loop could help alert originators make better 
choices.