[EAS] concerning the request for new weather Event Codes

Thu Jul 7 19:22:07 CDT 2016

Hi Dave, let me try to answer the main question you had. The use of
the polygon versus county-based short-fused warnings began around
2007. In time we likely will have different methods of displaying the
polygons and the threat. The polygon can be drawn in any shape and the
size typically varies between 300 and 1200 km² but depending on the
circumstances and threat it can be smaller or larger. For example, we
may draw a larger polygon out West if we expect the current
severe/Flash flood thunderstorm to spawn another entirely new
thunderstorm several miles down the road as they often do or if
movement is uncertain.

Any warning is actually a higher level than any watch. The reason for
this is that the preparations are done with the watch, but the actual
action of going to a basement, getting away from a window or pulling
over on a roadside to avoid flooding or turning around (hiking or
driving) occur with a warning because that means your area is likely
going to be hit by the impact. The size and scope of a Watch will lead
to some areas false alarmed and in fact locations may never even see a
drop of rain in a severe thunderstorm or tornado 'watch'. We are
predicting nature through science here and unless it's within an hour
of time it's not going to show up on radar in most situations. Yes we
do look out the window! The Wireless Emergency Alerts only carry
'warnings'.

You are correct that the polygon warning can result in the same county
(especially large ones) having several warnings at one time or
throughout a period of time. However, I can assure you that these are
often different storms especially in the West. The only exception is
when two warnings are issued that cross an area of responsibility
between NWS offices. This could happen with the county based too, but
we make every effort to not enter a new county if it's avoidable.  Of
course in operational mode a forecaster may not know if a storm is
going to completely fall apart and a storm may decide to continue to
track into another county. The forecaster 'does' know of any existing
warning (viewed on maps) within their area responsibility and outside
their area. The nature of any thunderstorm or short-fused warning for
that matter is never the same for any given event. However, without
this years of observing thunderstorm movement on radar and evolution
in different parts of the country it's completely understandable how
it may appear differently on paper.

The current definition of a thunderstorm warning includes hail 1 inch
size in diameter or larger and/or wind speeds of 58 mph or higher. Of
course, this is just a baseline trigger and in reality the storms can
vary greatly and have much different impact in any given location. If
the verification shows half-inch Hail and 50 mph winds and there's no
damaging impact such as a tree knocked over...then it is considered
overwarned in that polygon area.
Also there is no magic button that says the storm will be "severe" and
it takes skill and practice.

Unfortunately, it does not include quantity or impact from lightning
and lightning is expected to occur in any given location inside the
warning polygon (or outside).  We can actually verify this as well
with high resolution detection systems but this is after the
occurrence. Therefore, you may know the obvious that it's raining hard
or that you hear thunder but distinguishing between a damaging
thunderstorm and just a weak to moderate event is what you won't know
and why the 'warning' is used as notification. It's not a perfect
system with humans involved as well as computers involved, but we aim
to keep the false alarms down (POD up) and also do not consider that
there may be no one in the area of threat so it's not population
centric even though a major population center may be given more weight
which can lead to a rarer tornado emergency or flash flood emergency.

The tornado warning indicates that either a tornado is spotted or
radar indicates it is touching the ground or will in very short
notice. We know however that any tornado is going to vary greatly in
wind speeds, size, duration and magnitude. If this information is
known or indicated that message may also contain it. On the contrary a
tornado watch only indicates that atmospheric conditions are favorable
over a large area such as Half of the state or all of San Bernardino
County. So the watch versus the warning all depend on what you're
trying to accomplish for notification. Having grown up in areas that
we have thunderstorms and thunderstorm watches it is true that they
are often very effective but it would be unnecessary for hundreds of
people to cancel or stay in a home with windows closed in many
instances for a watch unlike the warning which is a much higher
imminent probability of occurrence in a much shorter timeframe and
lead time.

For example, but on a different scope, a hurricane watch may be the
time where you actually start to evacuate and board up windows. A
hurricane warning for those who didn't leave or thought there was too
much uncertainty would be the time where you have to hunker down.

The same would be true for heavy rain flash flood events which are
often even more localized then a thunderstorm wind related event. We
are working on trying to develop something to distinguish the average
(predicted) severe thunderstorm, which follows the definition listed
above, and an extreme thunderstorm warning which may be something more
common in the Midwest or sometimes referred to as a derecho. Those
events like discussed before are not spotty and can be widespread and
continuous, and end up affecting almost the entire original watch
area. The current extreme wind warning that was referenced is only
used for the eye wall of a hurricane. This stemmed from the inability
to let people know in past events that a category 4 eyewall (example)
was imminent and their current shelter may not be sufficient.

I am sure there are better ways but this is how it's currently done. I
also know that social scientists are involved as well to see how
effective the services are.

Check out this recently released extensive report on what we're talking about.

http://www.ppi.noaa.gov/wp-content/uploads/Risk-Communication-and-Behavior-Best-Practices-and-Research-Findings-July-2016.pdf

In case someone missed the prior discussion, each office actively
works to reduce the length of any 'warning' or the actual issuance of
a polygon in consideration of interruptions and the two minute limit.

Alex Tardy
Warning Coordination Meteorologist
NOAA/National Weather Service
San Diego,CA
Web: weather.gov/sandiego
@NWSSanDiego

> On Jul 7, 2016, at 11:49 AM, Dave Kline <dkline at tvmail.unomaha.edu> wrote:
>
> Alex,
>
> Thank you for weighing in on this and giving us the perspective from the NWS viewpoint.
> I have some comments that I will preface by saying these are just my personal opinion.
>
> When was it that NWS began alerting based on polygons?
> I know they have been doing it for a while, but I don't recall really hearing much about it in the beginning.
> It seems that with NWS using polygons and EAS using counties (or 1/9 of counties, which is possible but not used widely, if at all) that this can lead to issues of multiple warnings for a single county from one storm. Depending on the size of polygons (do they differ or are they a fixed size across the nation?) and the size of the county, can this lead to more alerting for some areas and less for others? I am not familiar with the ins and outs of the NWS polygons. Perhaps some more info about those is needed for broadcasters.
> Maybe there is a way that broadcasters can better match up their EAS local areas with the polygon structure since most local EAS alerts are weather related.
>
> I believe that a Thunderstorm Watch (SVA) actually has more value than a warning (SVR).
> This is very similar to Tornado Watches. In both instances we are given notice that we could be in for some ugly weather.
> With both of these watches, the word gets out. But unlike a Tornado Warning which can happen rather suddenly, and can cause major property damage, injuries or death, A T-Storm warning kind of tells us what we already know. It's raining real hard, the wind is blowing very hard.
> There might also be hail with the T-Storm so I wonder if a separate Hail Warning might be more useful that the more generalized T-Storm Warning. We can get flooding with a thunderstorm, and we have a separate alert for Flash Floods. And like with flooding or hail a T-Storm can spin up a Tornado and we have a warning for that. NWS wants to add high winds to the alerting menu, so with that and a hail warning, why do we need T-Storm warnings that can, at least in the heartland, move across several counties and trigger several alerts for the same storm?
> My thinking is that a one time issued T-Storm Watch for a wide area, would alert folks to the possibility that today is likely to bring severe weather. Heed the watch and prepare for what may come or suffer the consequences of doing nothing. Whereas the T-Storm Warning tells us that the bad weather is happening now. It is a bit late to do much about anything when it is issued.
>
> I think broadcasters are kind of shying away from the T-Storm Warning just because it does seem like over alerting as the storm moves.
> Tornado and Flash Flood Warnings on EAS carry the most value in our area and offer the least amount of disruption to programming. Though if a station is not carrying up to the minute coverage of a storm as it passes through the listening/viewing area, the audience will tend to turn to someone who is. From my perspective, folks in the heartland are pretty weather savvy. At least they pay attention to what is going on in the immediate area. Hopefully, folks across the US are also well aware of, and interested to follow weather that affects them.
>
> My last comment is about EAS alerts on cable that blanket all channels. When that happens around here, the EAS alert covers up what the local TV stations are doing to cover the storm. I know this is a cable issue and not an NWS problem, but it is a real issue in these parts at least. Cable companies must find a way to not interrupt those stations that are carrying detailed information about severe weather that is happening right now. Cable, which is already operating essentially as a government sanctioned monopoly in most areas and charging exorbitant rates already, should be able to step up and fix their warning systems so as to not interrupt the much better information coming from local sources.
>
> Thank you for your time and consideration.
> Dave
> ----------------------------------------
> Dave Kline   UNO-TV/Mav Radio/KVNO
> University of Nebraska at Omaha
> 6001 Dodge St. Omaha, NE  68182  CPACS 200
> Voice 402-554-2516   Fax 402-554-2440
> dkline at tvmail.unomaha.edu
> ----------------------------------------
> Why did the chicken cross the Möbius strip?
> ----------------------------------------
>
>> On Jul 7, 2016, at 12:27 PM, Alexander Tardy - NOAA Federal wrote:
>>
>> Good discussions that should continue in each region.
>>
>> There are strong concerns of "messaging flooding" but there also is not a clear understanding of the actual "storms". Many in the NWS are trying to seek new improved solutions for an old limited EAS system. Unfortunately, it is not just a matter or the same "storm"
>> going from one location to the next. The last thing we want to do is message flood and overwarn so that the public we are serving
>> are not taking proper action to avoid the dangers. Each major event there are often extensive surveys of the public and officials to document how effective the warning or watch messages are (the meteorology and the warning itself). With 122 weather forecast offices there is definitely one nearby to voice local concerns about broadcasting needs and requirements.
>>
>> Thunderstorm 'Watches' are issued for large areas such as half of a state. They are issued in advance, up to 12 hours. The 'warning' message is designed
>> to be imminent danger for a much smaller location (defined polygon) and to take action (shelter) or avoid flooded roads while travelling. There may
>> be better methods but as of today the short duration warning message that goes to the Wireless Emergency Alert may be your only notification
>> that you are driving into a washed out road, that a wall of water is coming down the mountain while it is sunny, or that a tornado is a few blocks from your house versus just a strong thunderstorm.
>>
>> If one "warning" was issued for an entire medium to large county valid for 5 hours (the window of daily thunderstorms) would that be effective? This has
>> been suggested but likely will not  meet anyone's mission.
>>
>> It is not uncommon to have 2-7 different (separate) thunderstorms and are impacting individual events in a large county and busy summer day. All of these "events" are likely occurring at a different time of the day. Entirely different scenario would be a winter storm.
>>
>> The other issue is that out West the storms have very little movement (summer monsoon) so they do not travel from one area to another. Rather they live for 1 to 2 hours (move little) and an entirely new storm (cloud) may form nearby or tens of miles away in another location.  MIdwest and Eastern US thunderstorm events can often move rapidly (30-70 mph), cover much more square miles (larger and organized versus sporadic) which results in the same "storm"
>> having additional new "warnings" issued downstream over a large swath of land (many counties). The only way to visualize what is actually occurring is on a map
>> and not a daily series of text product messages. Out West it looks like pepperoni on the pizza, and scientifically it is quite random though thunderstorms
>> tend to favor high deserts, mountains and valley convergencing areas (similar to how high wind favors areas used for wind farms).
>>
>> Try these links in real-time or in past viewing mode for any location in the US.
>>
>> https://nwschat.weather.gov/live or view old events at https://nwschat.weather.gov/vtec (substitute vtec with lsr) or follow your local weather.gov page as
>> suggested.
>>
>> One thing for sure is that efforts are made to validate every "storm" whether it be on the radar (proactive before it hits the ground as rain, flood or wind), using social media, emergency management and the public, and 'following' the storm in a survey. The lead time before damaging impact has increased since the warnings are often a forecast/prediction (not just a reaction to a report) but this can also lead to false alarms. NWS needs the broadcasters as a voice of the life and property
>> saving messaging.
>>
>> Alex Tardy
>> Warning Coordination Meteorologist, Manager
>> Emergency Preparedness and Partner Collaboration
>> Education and Outreach Coordinator
>> Media and Public Information Officer
>> Cell: 858-442-6016  Office: 858-675-8700
>> Skywarn Program Manager
>> NOAA/National Weather Service
>> 11440 W. Bernardo Court, SanDiego, CA
>> weather.gov/sandiego
>>
>> Register_as_a_Weather-Ready_Nation_Ambassador (click here)
>> Facebook page for NWS San Diego
>> Twitter @NWSSanDiego
>> http://www.youtube.com/NWSSanDiego YouTube
>> We need precipitation reports! http://cocorahs.org/
>>
>>> On Wed, Jun 22, 2016 at 3:41 PM, Mike McCarthy <towers at mre.com> wrote:
>>> They still issues both watches. Awareness of the watches has never been
>>> greater...which is a good thing.
>>
>>> Message flooding of warnings is the problem which there needs to be a
>>> balance. The root problem is the use of timed polygons. It is not uncommon
>>> for the same county to have multiple warnings for the same storm simply
>>> because the projected polygon time doesn't cover a whole county.
>>
>>> Particularly on SVR's and FFW's which are broad swath warnings covering
>>> only a portion of a given county. If a whole county is going to be
>>> impacted, then issue the warning for the whole county once. No two...or
>>> even three times.
>>
>>> Or in the case of tornadoes, multiple spotted and/or radar indicated
>>> tornadoes are warned in different parts of the county and/or times.  A
>>> couple years ago, Kankakee County here had 4 concurrent TOR's in progress
>>> in a county less than 750 square miles.  That evening saw one of our
>>> stations issue 6 TOR's in a span of only a couple hours and upwards of 10
>>> the whole event.
>>
>>> MM
>>
>>>> On Wed, June 22, 2016 4:35 pm, Tim Stoffel wrote:
>>
>>>> It used to be in the olden days, they would issue a severe thunderstorm
>>>> watch for an area, and then warnings for specific areas. This was in the
>>>> pre-internet days, when you would learn of this via radio and TV. I think
>>>> they still issue tornado watches, but you don't (at least not here) hear
>>
>>> __________________________________________________________
>>> The EAS Forum Discussion List is hosted by the BWWG (Broadcast Warning Working Group). http://eas.radiolists.net
>>> Please invite your friends to join our Forum! The sign up is at: http://lists.radiolists.net/mailman/listinfo/eas
>>> ___________________________________________________________
>>
>> __________________________________________________________
>> The EAS Forum Discussion List is hosted by the BWWG (Broadcast Warning Working Group). http://eas.radiolists.net
>> Please invite your friends to join our Forum! The sign up is at: http://lists.radiolists.net/mailman/listinfo/eas
>> ___________________________________________________________
>