Watch boxes, convective outlook risk areas, and issuance of watches

[Jason Bolt]

There are a couple questions I've had for a while:

It seems to be a daily occurance to see multiple SVR warnings and even TOR warnings w/ confirmed touchdowns in areas that are not in even a slight risk area and are under no WW. How does this happen with such consistency? I guess I am asking if there is continually severe weather in an area with not even a "see text" box, should the criteria be changed?

Secondly, once 5 or 10 SVR warnings have been issued by the local NWS is there a mechanism to go ahead and issue a watch? Wouldn't that give counties ahead of the storm some kind of advanced notification as a watch is intended to do?

Next, are there firm guidelines in place regarding the boundaries of watch boxes? Take SVR watch 520 as an example: http://www.spc.noaa.gov/products/watch/ww0520.html
Red Willow county (McCook) is drawn nearly entirely inside the watch box but is omitted from the list of counties. On the other side of the watch box, there are two counties that are almost entirely outside of the watch box yet they are included. How is this decided?

Finally, like today, once a couple TOR warnings are issued inside a SVR box, is there consideration to upgrading the watch? Does the frequent occurance of severe weather with no watches issued damage the public perception of severe weather forecasting?

 

[mike scantlin]

its still a new science. i mean, if a storm shows low-level rotation, a tor warning is issued. that doesnt mean that it will produce a tornado, nor does it mean the environment is ripe enough for tornadoes. it just means that on that radar scan, there was enough low-level rotation to cause some concern. how many clasiic supercells have you been on that were spinning like crazy in a prefect environment, that didnt produce a tornado? how many times do you hear of crap storms producing tornadoes? tornado/severe predictibility still has a LONG way to go.

they only issue convective outlooks at certain times each day. i think once around 2am, again late morning, again around 3pm, and again around 7.

the public perception of meterology in general is messed up. i cant tell you how many times a day someone i know asks for a precise forecast, and when i dont have an answer, goes off about how none of us know anything. or how about the "thats the best job in the world! you get paid whether you're right or wrong! haha!" and im not even a meteorologist.

theres always going to be the "we had no warning" people and there will always be the people who get in the cellar if theres a warning within 100 miles. i think there should be classes in school or something that most of the public will see that explains the watch/warning system accurately and thouroughly. but thats probably a crazy idea. im done. now.

 

[Andrew Pritchard]

Next, are there firm guidelines in place regarding the boundaries of watch boxes? Take SVR watch 520 as an example: http://www.spc.noaa.gov/products/watch/ww0520.html
Red Willow county (McCook) is drawn nearly entirely inside the watch box but is omitted from the list of counties. On the other side of the watch box, there are two counties that are almost entirely outside of the watch box yet they are included. How is this decided?

Regarding the odd shape... the SPC only issues the general outline, then I believe the local NWS offices are allowed to discuss which counties in their area they believe warrant the watch area.

 

[John Robinson]

SPC usually deals with *organized* severe weather events in watches and, to some extent, in outlooks. There can be days in the warmer months when quite a few warnings are issued on pulse-type storms. These situations rarely call for watches as there is no organized forcing mechanism.

SPC does go by certain guidelines in issuing watches and in deciding their risk categories in outlooks. You can see the definitions here...
http://www.weather.gov/directives/sym/pd01005012curr.pdf
or here...
http://www.spc.noaa.gov/products/

Before SPC issues a watch, they hold a conference call with the WFOs that will be affected. The discussion during the call may include whether a watch is necessary and what type of watch to make it. One thing that is always discussed is the counties/parishes that will be included in a watch. SPC proposes which counties/parishes to include and the WFOs often ask that certain other counties or parishes be added or subtracted. This is why the "box" is only an approximation of the actual watch. The counties/parishes actually included in the watch appear in the lists issued by SPC and the WFOs. The map of the actual areas included often has irregularly-shaped edges because of the additions and subtractions that have been agreed upon, so the actual watch is rarely ever really a parallelogram, rectangle, or square.

Whether to issue a watch after several warnings have been issued is an age-old dilemma. If it looks as though there is some organization to the severe weather system and a number of additional warnings may be needed, then a watch might be issued. On the other hand, if it looks like there may be only a few more warnings, then a watch may not be issued. SPC and the WFO may call each other to discuss whether a watch is needed in such situations. There are definitely two schools of thought here. Some people would rather have a watch. Others would say that the horse has already left the barn and that a watch issuance after a number of warnings have been issued simply clutters things up, especially if the warnings are already doing a good job of handling the situation.

SPC does, at times, upgrade the type of watch. In fact, some of the springtime watches will even mention in the body of the text that a severe thunderstorm watch that is being issued relatively early in the day may require an upgrade to a tornado watch later. Keep in mind, though, that a couple of tornado warnings here and there, or even a weak, isolated tornado touchdown do not meet the criteria for a tornado watch issuance.

As far as public confidence in severe weather forecasting goes, people should realize, as noted above, that SPC does not strive to issue watches for a bunch of pulse-type events. That's simply not their goal or their purpose.

A couple of SPC forecasters are members of this board, so they may have more to add from their perspective.

 

[Jared Guyer]

Really not too much to add beyond John's post, but a few additional comments...

Amendments to the existing Day 1 Convective Outlooks are occasionally made. Such amendments typically only occur when the SWODY1 varies considerably from the short term expectations, but it's ultimately a judgement call by the Outlook and Lead Forecasters. In the short term, the expectation is that our users are more likely consulting existing watches and mesoscale discussions, and of course the latest warnings/statements from local NWS offices.

The official tornado watch criteria for the SPC is a reasonable expectation of 2 or more tornadoes (was 3 until a couple of years ago), or the possibility of at least 1 significant tornado. Even if a Severe Thunderstorm Watch has been issued, conditions are perpetually monitored for an appreciable change. Even if a number of tornado warnings are issued or an isolated tornado report comes in, a Tornado Watch upgrade may not occur if it's believed the environment/situation has not appreciably changed and/or if it will remain short-lived/localized. If the tornado threat may continue for another couple of hours, and previous expectations have clearly been exceeded, then certainly a Tornado Watch upgrade may be prudent.

The key goal for SPC is forecasting organized severe storms, ones that have a degree of sustainability and pose a threat across a sufficient geographic extent. The goal is to not to necessarily have a Watch precede every warning or severe report, but rather to catch the bulk of the organized severe, especially those storms that may pose a "significant" threat for strong tornadoes, very large hail, and widespread damaging winds.

As John mentions, the "box" is a legacy product. In this era of "Watch by County", the counties are the official determination of a Severe Thunderstorm or Tornado Watch. SPC tries to have the "box" fit the included counties as best as possible, but there'll always be discrepancies given the constraints of a parallelogram.

Jared

 

[Rich Thompson]

There are a couple questions I've had for a while:

It seems to be a daily occurance to see multiple SVR warnings and even TOR warnings w/ confirmed touchdowns in areas that are not in even a slight risk area and are under no WW. How does this happen with such consistency? I guess I am asking if there is continually severe weather in an area with not even a "see text" box, should the criteria be changed?

Secondly, once 5 or 10 SVR warnings have been issued by the local NWS is there a mechanism to go ahead and issue a watch? Wouldn't that give counties ahead of the storm some kind of advanced notification as a watch is intended to do?

Next, are there firm guidelines in place regarding the boundaries of watch boxes? Take SVR watch 520 as an example: http://www.spc.noaa.gov/products/watch/ww0520.html
Red Willow county (McCook) is drawn nearly entirely inside the watch box but is omitted from the list of counties. On the other side of the watch box, there are two counties that are almost entirely outside of the watch box yet they are included. How is this decided?

Finally, like today, once a couple TOR warnings are issued inside a SVR box, is there consideration to upgrading the watch? Does the frequent occurance of severe weather with no watches issued damage the public perception of severe weather forecasting?

Jason,

As John and Jared mentioned, the watch process normally originates at the SPC, but there's plenty of give and take between the NWS offices and the SPC regarding the need for a watch, the counties/parishes included, and the type of watch. The most difficult cases are those that are thought to be "marginal" at the beginning. It takes a bunch of convincing to completely change a forecast, and a few warnings isn't always enough. The months of June and July are notorious for these loose clusters of "pulse" severe storms from the Gulf coast into the Appalachians and OH valley. The difference between 100 penny hail/tree down reports or almost nothing can be quite subtle.

In my experience, personal perception can vary widely between people experiencing the same events. You mention that these events are missed "daily", which may be true, but I also suspect that you're really referring to just a couple of events that stick in your mind. Do you have specific days and areas that got you motivated to make this post? If so, it would help if we could look at the specific events to provide an explanation for what happened (or didn't happen).

Rich T.

 

[Jason Bolt]

Thanks for the insight guys. Rich, I really don't have any certain events in mind. I wish I did; I'm really trying to learn about the technical and analytical side of things. I think it is just the public safety/EM side of me coming out. I'll try to come up with some examples to dissect.