[EAS] San Francisco Heat Alert

[Mike McCarthy]

I suggested they simply make SVR's include an entire span of the county
which an event will travel. Not a portion of a county first, and then
repeat for the latter portion under a new warning.  This occurs all the
time on SVR's where they warn the first impact portion of a given county
first and then the latter parts some portion of a hour later. This method
is based on projected time of arrival during the warning. Simply a poor
metric in this application.  A county should only be warned once for a
single event SVR.

(NOTE...I've seen where secondary development occurs over a different area
of a county and goes severe. If the warned area was different from this
new development in a partial county warning, that would warrant a new
warning for the new development.)

TOR's and FFW's are more challenging as those polygons are far more
precise and tightly defined than "broad" span SVR's. It is not uncommon
here to see multiple concurrent TOR's over a single county.

Much different than squall line SVR's  where the forecaster issues a broad
coverage warning covering a half dozen counties spanning >50 miles in
length. This is where the flooding aspect starts to happen...

More over, it may be time to revisit the radar indicated severe criteria.
Which is what a preponderance of warnings are based. I can't speak to the
percentage of ground verification on those warnings. But the prevailing
mood in the spotting community has increasingly been one of "They issued a
warning on 'that'?"

MM

On Thu, September 28, 2017 12:39 am, Botterell, Arthur at CalOES wrote:
>
> PS - Here's a question:  We're seeing push-back nationwide over what some
> folks call "message flooding" on EAS from NOAA.  As best I can tell, this
> is because NOAA has a workflow evolved over many years of Weather Wire
> operation, wherein for a moving weather system they feel obliged to issue
> updated FFWs or whatever to re-localize the alert every so often, even