Login to AccuWeather.com Premium Login to AccuWeather.com Professional Login to AccuWeather.com RadarPlus AccuWeather.com

Welcome Guest ( Log In | Register )

5 Pages V  « < 3 4 5  
Closed TopicStart new topic
> SPC Outlooks/Watches/Warnings - Disco
Hertz
post Jun 17 2011, 12:16 PM
Post #81




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





QUOTE(Chicago Storm @ Jun 16 2011, 11:46 PM) *
The ECMWF/NAM/GFS are on a whole different level compared to the UKMET/GGEM.

When forecasting, stick with the ECMWF/NAM/GFS/SREF (and other short term/hi-res models)...that's pretty much all you need.

Models like the UKMET/GGEM/NOGAPS/JMA are for "lol's".


As in, you're trying to say the models don't do as good a job overall?


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
CMeyers
post Jun 17 2011, 04:03 PM
Post #82




Rank: F5 Superstorm
***

Group: Member
Posts: 1,075
Joined: 29-May 10
From: Delton, MI
Member No.: 22,879





QUOTE(Hertz @ Jun 17 2011, 01:16 PM) *
As in, you're trying to say the models don't do as good a job overall?

Yes. That is exactly what he's saying.


--------------------
ChsChargers09 - Please be aware that there is a significant chance I was being sarcastic in the above comment. Thank you.
Go to the top of the page
 
+Quote Post
Hertz
post Jun 17 2011, 08:21 PM
Post #83




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





While far away from home, I must comment that I think severe probabilities in the SW half of Missouri (all three types) are FAR too high for tonight. It doesn't even mention that much shear and giving the capping, I think 2% tornado, 15% wind/hail would be entirely sufficient for that area given there's no current development (and the outlook depicts higher probabilities than I think are appropraite for all three hazards).

This was a long, long time ago (over 6 years), but this evening reminds me of this 01Z outlook over eastern Nebrasks/western Iowa:
http://www.spc.noaa.gov/products/outlook/a...50522_0100.html

This is 01Z from May 22, 2005. Similar to today nothing had developed in the mentioned area yet, and even the text in the outlook itself mentioned "strongly conditional" this time!

Again, unless there is a MAJOR intensifying storm that looks almost sure to cause an explosion (the evening of last October 25th would be an example of this), uncertain outlooks issued after peak heating should NEVER have hazard probabilities higher than 5%/15%/15% for tor/wind/hail respectively on them. It is true that despite my main attention being on tornadoes for such a long time, my main complaint with these two outlooks (this evening's and the one very similar case I remember to it) is that the wind/hail probabilities are/were entirely inapprorpriate for the situation. I guess I can agree that if the surrounding environment is extremely favorable for supercells, then using a 5% tornado risk is acceptable, but using 25%/30% risks for wind/hail is not warranted (given, these probabilities for wind/hail roughly match a 10% tornado probability - in that they're both the top level that will be used in a slight risk - 5% tornado would match up with 15% wind/hail in being mid/low-end slight risk level).

So in the end, it isn't really the tornadoes I'm thinking about in these two cases, but rather using such high wind/hail probabilities. Anyone agree/disagree with my consensus here, and why?

EDIT: Wow, they've even issued a severe thunderstorm wathc (with pretty high probabilities) for SW Missouri now. Must be thinking the cap won't prevent the small cluster currently over Kansas from expanding and moving through.

If this is the case, and maybe the same thing SPC was thinking when they issued the outlook (as there was at least a hint of activity over Kansas by issuance time) then I can back off a bit on today's outlook, although I'm going to leave in the details for the old one from May 22, 2005, because that evening there was truely no convection going on in the affected area and even the discussion itself mentioned "strongly conditional".

This post has been edited by Hertz: Jun 17 2011, 08:33 PM


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
Chicago Storm
post Jun 17 2011, 08:29 PM
Post #84




Rank: F5 Superstorm
***

Group: Member
Posts: 34,724
Joined: 14-February 08
From: The 630
Member No.: 13,697





QUOTE(Hertz @ Jun 17 2011, 08:21 PM) *
While far away from home, I must comment that I think severe probabilities in the SW half of Missouri (all three types) are FAR too high for tonight. It doesn't even mention that much shear and giving the capping, I think 2% tornado, 15% wind/hail would be entirely sufficient for that area given there's no current development (and the outlook depicts higher probabilities than I think are appropraite for all three hazards).

The area is highly unstable, on the edge of the cap, and will have some forcing crossing through the general area.

Short term/hi-res models all along have been showing development around or just after sunset in Kansas, growing upscale into an MCS that pushes into the area.

The SPC 1z outlook probs are fine.

Edit...bingo...



This post has been edited by Chicago Storm: Jun 17 2011, 08:29 PM
Go to the top of the page
 
+Quote Post
Chicago Storm
post Jun 17 2011, 08:41 PM
Post #85




Rank: F5 Superstorm
***

Group: Member
Posts: 34,724
Joined: 14-February 08
From: The 630
Member No.: 13,697





QUOTE(Hertz @ Jun 17 2011, 08:21 PM) *
This was a long, long time ago (over 6 years), but this evening reminds me of this 01Z outlook over eastern Nebrasks/western Iowa:
http://www.spc.noaa.gov/products/outlook/a...50522_0100.html

This is 01Z from May 22, 2005. Similar to today nothing had developed in the mentioned area yet, and even the text in the outlook itself mentioned "strongly conditional" this time!

They did say it was conditional...for W. NE/E. IA...only a small portion of the higher probs.

Even though it was conditional due to the cap, the environment was prime for a nice event in the case the avaliable forcing was enough to weaken the cap soon enough for development to occur before the environment became less favorable.
Go to the top of the page
 
+Quote Post
Hertz
post Jun 17 2011, 09:25 PM
Post #86




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





I admit I was wrong on today's probs...assuming the storms currently going in Kansas actually start moving. Radar over the last hour indicate they've been slowly expanding but almost stationary.

There is a LLJ though tonight as well so I pretty much concur 01Z probs are likely to turn out to be fine (and were fine at time of issuance).

In the '05 case though...I'd have more to discuss with that...but first are you familiar with the case I'm mentioning? Actually either way my thoughts are this...there were no good signs of initiation, peak heating had passed, I don't think there was a great LLJ - with a tornado watch already out in this case 5% tornado probs were basically a given and since it wouldn't make a lot of sense (except perhaps for a tropical system, which this clearly was not) to have sub-slight probabilities for other hazards when 5% tornado is out, 15% wind and hail would have been fine. But nothing had developed yet, and as far as I know there was no good LLJ signal, so hosting 25% wind/hail after peak heating seemed way inappropriate.

I actually consider this May 2005 outlook somewhat like the April 4, 2010 Arkansas case I discussed and you said you are familiar with - in the Arkansas case, there was good shear and instability but no clear focus; however peak heating had not yet been reached and the SPC did clear that slight risk at 01Z after peak heating passed and the cap had held (and there was still not a good focus). In the May 2005 case, again shear and instability were good, and this time there was a focus (a cold frontal boundary), but peak heating had already passed. There was no good signal of a LLJ in either case, so the favorables and unfavorables for severe would IMO basically balance out between the 01Z outlook in May 2005 and the midday outlooks in April 2010. Since the April 4, 2010 midday outlooks used 5%/15%/15% for the hazards, that's clearly the probabilities the 01Z I'm referencing from May 2005 should have used as well if it wanted to depict a conditional threat.

But I will repeat that I'm wrong about this evening...the probabilities used are warranted assuming the Kansas cluster takes off, and even if it doesn't, there's no way the SPC could be confident it wouldn't at that point (or even at this point).


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
Chicago Storm
post Jun 18 2011, 12:03 AM
Post #87




Rank: F5 Superstorm
***

Group: Member
Posts: 34,724
Joined: 14-February 08
From: The 630
Member No.: 13,697





QUOTE(Hertz @ Jun 17 2011, 09:25 PM) *
In the '05 case though...I'd have more to discuss with that...but first are you familiar with the case I'm mentioning? Actually either way my thoughts are this...there were no good signs of initiation, peak heating had passed, I don't think there was a great LLJ - with a tornado watch already out in this case 5% tornado probs were basically a given and since it wouldn't make a lot of sense (except perhaps for a tropical system, which this clearly was not) to have sub-slight probabilities for other hazards when 5% tornado is out, 15% wind and hail would have been fine. But nothing had developed yet, and as far as I know there was no good LLJ signal, so hosting 25% wind/hail after peak heating seemed way inappropriate.

I actually consider this May 2005 outlook somewhat like the April 4, 2010 Arkansas case I discussed and you said you are familiar with - in the Arkansas case, there was good shear and instability but no clear focus; however peak heating had not yet been reached and the SPC did clear that slight risk at 01Z after peak heating passed and the cap had held (and there was still not a good focus). In the May 2005 case, again shear and instability were good, and this time there was a focus (a cold frontal boundary), but peak heating had already passed. There was no good signal of a LLJ in either case, so the favorables and unfavorables for severe would IMO basically balance out between the 01Z outlook in May 2005 and the midday outlooks in April 2010. Since the April 4, 2010 midday outlooks used 5%/15%/15% for the hazards, that's clearly the probabilities the 01Z I'm referencing from May 2005 should have used as well if it wanted to depict a conditional threat.

Actually, there was a nice 20-40kt LLJ.

By the way, you're really overplaying the "after peak heating" and "after sunset" cards in this topic way too much.

This post has been edited by Chicago Storm: Jun 18 2011, 12:04 AM
Go to the top of the page
 
+Quote Post
Hertz
post Jun 18 2011, 09:22 AM
Post #88




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





QUOTE(Chicago Storm @ Jun 18 2011, 01:03 AM) *
Actually, there was a nice 20-40kt LLJ.

By the way, you're really overplaying the "after peak heating" and "after sunset" cards in this topic way too much.


Which day are you referring to the LLJ on - 4/4/2010 or 5/22/2005?

If there was a(n) LLJ on 4/4/2010 then it would actually surprise me that SPC dropped the slight from Arkansas without hesitation at 01Z that day.

Also could you clarify why you think I'm overplaying the "after peak heating"/"after sunset" stuff (barring LLJ's)?

As far as I know, without a(n) LLJ, severe storm development during dark hours is quite unlikely.


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
futureweatherman...
post Jun 18 2011, 07:15 PM
Post #89




Rank: F5 Superstorm
***

Group: Member
Posts: 11,919
Joined: 27-August 08
From: NW OH
Member No.: 15,447





QUOTE(Hertz @ Jun 18 2011, 10:22 AM) *
Which day are you referring to the LLJ on - 4/4/2010 or 5/22/2005?

If there was a(n) LLJ on 4/4/2010 then it would actually surprise me that SPC dropped the slight from Arkansas without hesitation at 01Z that day.

Also could you clarify why you think I'm overplaying the "after peak heating"/"after sunset" stuff (barring LLJ's)?

As far as I know, without a(n) LLJ, severe storm development during dark hours is quite unlikely.

Peak/Heating does not affect storm development as much as you think. Just look at today/yesterday, storms developed after peak heating and still were severe.


--------------------


QUOTE(SEMIweather @ Oct 17 2010, 02:10 AM) *
i was lclicking on it going pelasejk not nicki minaj m-please not micni minaj hughhhhhhhhhhhhhhh
Go to the top of the page
 
+Quote Post
Hertz
post Jun 18 2011, 09:26 PM
Post #90




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





QUOTE(futureweatherman12 @ Jun 18 2011, 08:15 PM) *
Peak/Heating does not affect storm development as much as you think. Just look at today/yesterday, storms developed after peak heating and still were severe.


Yes I am aware that happens often with LLJ's around, but I'd assume you generally need a LLJ to get anything more than isolated/weak thunderstorm development after peak heating. Many LLJ's actually tend not to produce the worst storms until the nighttime hours though...so if you have a decent/strong LLJ the possibility of storm development or intensification after peak heating must be honored, especially if models are depicting it. So this basically reduces my argument to I think severe storms don't form past peak heating without LLJ's...if you have a counterexample feel free to let me know though.

Generally on the days I've complained about the SPC maintaining a severe risk considerably greater than what was currently on the radar at 01Z, there was an LLJ involved, so it's not really that much of an argument, more a question. (Although, I'm pretty sure the LLJ in the May 2005 case was not that strong and it was also a bit east of where things were really primed to go surface-based, so my argument on that one outlook still holds - even if I'm wrong about the LLJ, I'd argue just the fact that the words "strongly conditional" and hail and/or wind probs >15% [or tornado probs >5% although they were only 5% in this case] in a 01Z outlook should be 100% mutually exclusive.)


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
futureweatherman...
post Jun 18 2011, 09:34 PM
Post #91




Rank: F5 Superstorm
***

Group: Member
Posts: 11,919
Joined: 27-August 08
From: NW OH
Member No.: 15,447





QUOTE(Hertz @ Jun 18 2011, 10:26 PM) *
Yes I am aware that happens often with LLJ's around, but I'd assume you generally need a LLJ to get anything more than isolated/weak thunderstorm development after peak heating. Many LLJ's actually tend not to produce the worst storms until the nighttime hours though...so if you have a decent/strong LLJ the possibility of storm development or intensification after peak heating must be honored, especially if models are depicting it. So this basically reduces my argument to I think severe storms don't form past peak heating without LLJ's...if you have a counterexample feel free to let me know though.

Generally on the days I've complained about the SPC maintaining a severe risk considerably greater than what was currently on the radar at 01Z, there was an LLJ involved, so it's not really that much of an argument, more a question. (Although, I'm pretty sure the LLJ in the May 2005 case was not that strong and it was also a bit east of where things were really primed to go surface-based, so my argument on that one outlook still holds - even if I'm wrong about the LLJ, I'd argue just the fact that the words "strongly conditional" and hail and/or wind probs >15% [or tornado probs >5% although they were only 5% in this case] in a 01Z outlook should be 100% mutually exclusive.)

LLJ enhances the tornado threat more than anything. However, storms can develop/intensify during the evening/overnight hours even without a potent LLJ. Take a look at Central MO. on the Mesoanalysis page. Despite weak 850mb winds, several severe thunderstorm warnings for large hail/damaging winds are in effect. These storms developed this evening and have continually intensified, due to high instability in place.

Also, storms have initiated again just east of Wichita, along the warm front, where there is good convergence. Although there is more of a LLJ in place just to the south of these storms, they developed in the early overnight hours tonight and are now severe thunderstorm warned.

What bothers me about this entire thread is that you continue to pick the needle in the haystack here. The SPC is an excellent forecasting office, and they do an excellent job on almost all days.
When dealing with conditional risks- if something does happen is always going to be the problem. If storms do manage to develop, then a risk level of 30% not entirely out of the question.


--------------------


QUOTE(SEMIweather @ Oct 17 2010, 02:10 AM) *
i was lclicking on it going pelasejk not nicki minaj m-please not micni minaj hughhhhhhhhhhhhhhh
Go to the top of the page
 
+Quote Post
Hertz
post Jun 19 2011, 10:00 AM
Post #92




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





With conditional risks, the SPC should take the probability of each hazard assuming something does develop, and multiply that by the percent chance they think that something will develop, to arrive at the percent chance they should display on the graphic.

For example, if probs would be 15%/45%/45% assuming development, but it is felt that the probability of development is 30%, then that equates to 4.5%/13.5%/13.5%, which would be indicated properly on an outlook of 5%/15%/15%.

The words "highly conditional" basically imply 40% or less condidence, so on the day I'm referring to, for the outlook of 5%/25%/25% to be warranted, the probabilities assuming development would have to be at least 12.5%/62.5%/62.5%, which would round to 15%/60%/60%, which would be a potential high risk! And that's assuming the confidence level was 40% - in reality with the words "strongly conditional" it's probably even less.

(If there's already some severe threat, but it is conditional that there will be an even higher severe threat, then the same rule above should apply except the probabilities on the graphic should be the average between the probabilities that already exist and the probabilities my formula would yield.)


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
CMeyers
post Jun 19 2011, 04:47 PM
Post #93




Rank: F5 Superstorm
***

Group: Member
Posts: 1,075
Joined: 29-May 10
From: Delton, MI
Member No.: 22,879





QUOTE(Hertz @ Jun 19 2011, 11:00 AM) *
With conditional risks, the SPC should take the probability of each hazard assuming something does develop, and multiply that by the percent chance they think that something will develop, to arrive at the percent chance they should display on the graphic.

For example, if probs would be 15%/45%/45% assuming development, but it is felt that the probability of development is 30%, then that equates to 4.5%/13.5%/13.5%, which would be indicated properly on an outlook of 5%/15%/15%.

The words "highly conditional" basically imply 40% or less condidence, so on the day I'm referring to, for the outlook of 5%/25%/25% to be warranted, the probabilities assuming development would have to be at least 12.5%/62.5%/62.5%, which would round to 15%/60%/60%, which would be a potential high risk! And that's assuming the confidence level was 40% - in reality with the words "strongly conditional" it's probably even less.

(If there's already some severe threat, but it is conditional that there will be an even higher severe threat, then the same rule above should apply except the probabilities on the graphic should be the average between the probabilities that already exist and the probabilities my formula would yield.)

That would be a completely useless method of making outlooks if the goal is to warn the public of potential dangers. Perhaps you'd be more in favor of an outlook setup that issues two sets of probabilities. First, the probability that storms form and second; the severe risks associated with any storm that forms.

Imagine a hypothetical here: Very warm and humid air-mass with extreme instability. However, significant cap is in place and most likely won't break. We'll say there's a 10% risk of storms developing at all. Thanks to the instability and other paramaters though there is a nearly 100% risk that anything that does develop becomes horribly severe. So, according to your little method the outlook would mention 10% severe probabilities. A percentage that wouldn't even cause the "typical" spc consumer to blink an eye. Nobody expects anything significant to occur and 90% of the time it wouldn't. But, perhaps today is the 10% and storms do pop. Within 30 minutes of initiation there are large tornadoes, high wind and baseball sized hail peppering the area. Nobody was warned. But, the spc can claim that they were actually "very accurate" because (as they'd claim) there was only a 10% chance of anything forming but they knew that if it did it would be severe, so, they were spot on with their 10% risk.

Obviously an extreme example, but the idea remains static in other instances. Multiplying the probability of storms forming by the probability a storm that forms will be severe provides no possibility for usefull warning at all.

If your big issue is that they don't properly account for the potential storms don't form in their severe outlooks then send them an email and request that instead of simply issuing one outlook they have two. The first outlook, which outlines likelihood of storm development would act as a "confidence" gauge for the second outlook. The second outlooks would, of course, be similar to the current outlook in that it would provide the severe probabilities (tor, wind, hail...).

Instead of trying to second guess everything they do, why don't you come up with a constructive idea and recommend it to them? The goal is to warn the public of the real danger. Not to hype things and not to pretend a threat doesn't exist. The way they do it now is pretty good, but there is room for improvement. I think a two part outlook would be a decent idea. Think about it, adjust it and mold it to your needs and then send the spc an email. This whole "hey, remember the 1z outlook in may of 2002? I think the 10% tor probabilites were waaaayyy to high, should have been 5%..." is getting old and really pointless.


--------------------
ChsChargers09 - Please be aware that there is a significant chance I was being sarcastic in the above comment. Thank you.
Go to the top of the page
 
+Quote Post
Hertz
post Jun 19 2011, 07:35 PM
Post #94




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





QUOTE(CMeyers @ Jun 19 2011, 05:47 PM) *
That would be a completely useless method of making outlooks if the goal is to warn the public of potential dangers. Perhaps you'd be more in favor of an outlook setup that issues two sets of probabilities. First, the probability that storms form and second; the severe risks associated with any storm that forms.

Imagine a hypothetical here: Very warm and humid air-mass with extreme instability. However, significant cap is in place and most likely won't break. We'll say there's a 10% risk of storms developing at all. Thanks to the instability and other paramaters though there is a nearly 100% risk that anything that does develop becomes horribly severe. So, according to your little method the outlook would mention 10% severe probabilities. A percentage that wouldn't even cause the "typical" spc consumer to blink an eye. Nobody expects anything significant to occur and 90% of the time it wouldn't. But, perhaps today is the 10% and storms do pop. Within 30 minutes of initiation there are large tornadoes, high wind and baseball sized hail peppering the area. Nobody was warned. But, the spc can claim that they were actually "very accurate" because (as they'd claim) there was only a 10% chance of anything forming but they knew that if it did it would be severe, so, they were spot on with their 10% risk.

Obviously an extreme example, but the idea remains static in other instances. Multiplying the probability of storms forming by the probability a storm that forms will be severe provides no possibility for usefull warning at all.

If your big issue is that they don't properly account for the potential storms don't form in their severe outlooks then send them an email and request that instead of simply issuing one outlook they have two. The first outlook, which outlines likelihood of storm development would act as a "confidence" gauge for the second outlook. The second outlooks would, of course, be similar to the current outlook in that it would provide the severe probabilities (tor, wind, hail...).

Instead of trying to second guess everything they do, why don't you come up with a constructive idea and recommend it to them? The goal is to warn the public of the real danger. Not to hype things and not to pretend a threat doesn't exist. The way they do it now is pretty good, but there is room for improvement. I think a two part outlook would be a decent idea. Think about it, adjust it and mold it to your needs and then send the spc an email. This whole "hey, remember the 1z outlook in may of 2002? I think the 10% tor probabilites were waaaayyy to high, should have been 5%..." is getting old and really pointless.


Your point about old outlooks is valid. I really shouldn't be referencing outlooks from that long ago, it just fit as an example in this specific case and the discussion ended up taking off from there - more so than I thought it probably would.

I think both my idea and yours make sense in a way and don't make sense in another. Your argument about my idea might be the reason for one of the most (in)famous about high risk busts of all time - a blue skies bust - May 7, 1988. Obviously I don't want to call a lot of attention to something 23+ years ago, but I could hypothesize that this particular day may have been an example of your extreme case described - a small chance of storms forming, but an extremely dangerous situation if they do. Of course the only reason I went back that far is that, given a high risk being a complete blue skies bust (not even any storms!) is an extremely unusual event (since most high risk busts simply involve strong storms that never quite took off as much as they could have), the conditions that day might have actually been as extreme as what you were hypothesizing, and I can't think of any other high risk day to hold this "blue skies bust" title.

But getting back on topic, with that out of the way, you do have a point - there are conditions where powerful thunderstorms could form, or no storms could form at all, with little opportunity for something in between (after all, if an updraft breaks through a very strong cap with a very high amount of energy, the result will be explosive, and the storm will be more than just mildly severe assuming there's any kind of shear/focus around). This would suggest that, in this type of case, my simple formula of multiplying would indeed not sufficiently warn of the strength of any storms assuming initiation occurs. However there is still a logic flaw:

-Cap-breakers can often be just one or two isolated very severe thunderstorms, in which the cap holds in 80% or more of places (no thunderstorms are observed), so since it cannot be predicted exactly where a storm is going to break the cap and exactly there it won't, it would be wrongful to say there's a 100% probability of severe in any single area even assuming at least one or two storms develop. So in this case, it would make sense to do your idea, with a "confidence level" outlook for any development, and then the normal type outlook that assumes development does occur. However, if it's a case in which if things start forming at one point, then they will start forming in all surrounding points too (i.e. the cap will be broken "simultaneously" in all alreas and a very strong squall line/bow/derecho/QLCS with rapidly evolve), then this would argue for probabilities on the "assuming development" outlook to be even higher than the level typically associated with modern day outlooks (in some cases, potentially even higher than it's current 60% maximum risk it will ever display).

So for this reason, I think FOUR outlooks may be needed to provide a full assesment of the situation:

1. An outlook showing the probability that thunderstorms will develop.
2. An outlook showing the expected coverage of thunderstorms assuming development. (This is actually very similar to the "enhanced thunderstorm outlooks" that the SPC already does.)
3. An outlook that shows the probability of each severe hazard assuming thunderstorms do develop, where it is also assumed that any random point in the potential severe area is affected.
4. The general severe weather outlook, which can be attained by multiplying the probability of each hazard in (3) by the product of the percentages in outlooks (1) and (2).

So, to make a random case up, let's say here's what each outlook shows in the affected area:

1. 20%
2. 30%
3. 30%/60%/45% (tor/wind/hail)
4. 30% * 20% = 6% (the product of the probabilities in (2) and (1)), so the outlook (tor/wind/hail) would be roughly 1%/2%/2% (which could be rounded to 2%/5%/5% so that no probabilities are considered to be negligible.

This outlook would clearly indicated that any storm that occurs is likely to be strongly severe, but it would equate to a "see text" in the total outlook due to low confidence that storms will develop and low expected coverage assuming they do develop. At the same time, however, outlook (3) would clearly indicate the possibility for very severe storms.

If the threat of development is less conditional, and/or the expected coverage assuming development is higher, then (1) and/or (2) would be greater percentages; therefore all products that yield the probabilities in the general severe weather outlook in (4) would also be higher. (And this statement can be extended to more than just a single example; any changes up or down in the probabilities assessed in (1), (2), and/or (3) would have the appropriate net effect on the general outlook in (4).)

So that's a lot of products, but that's what I think it would take to get a full and clear look at the situation, where the general severe weather outlook in (4) would yield the real probability of severe all things considered, but at the same time, outlooks (1) through (3) would guarantee that the SPC reader (or any forecasting service) has full warning of what may happen.

Basically, in short, since I agree with all of the issues you suggested about just showing lower probabilities in the outlook, I have come to the conclusion that more outlooks are needed to both show the real potential all things considered but also, at the same time, to give a much clearer picture of the situation. That way, the SPC can feel free to do 100% all things considered probabilities in its final severe thunderstorm outlook (4), because the nature of the potential it actually represents will be clearly shown in outlooks (1)-(3)

This post has been edited by Hertz: Jun 19 2011, 08:01 PM


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
Hertz
post Jun 19 2011, 07:43 PM
Post #95




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





I've seperated this into two posts because this is a general question and has nothing to do with what was discussed in the previous post. (Also my question in this post is for how SPC issues outlooks currently, since in reality it's highly unlikely they'd buy into my four-outlook idea, let alone this quickly.)

Basically, there is a storm going across the country this week that's already induced concern for tornadoes (some strong), and indications are tomorrow will only be worse than today. This main system doesn't reach my area (per forecasts) until at least Wednesday, but I am outlooked for severe both tomorrow and Tuesday as well. Therefore I'd like it if you'd let me know on what days (if any) this week you think I may be in a 5%+ area for tornadoes, and also if you think there's potential for me to have a day where my outlook will be even higher than 5% (i.e. 10% or greater).

(This could also be accompanied by just giving your predicted tornado outlook for my area each day.)

This post has been edited by Hertz: Jun 19 2011, 07:46 PM


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
Southern Indiana
post Jun 20 2011, 09:00 AM
Post #96




Rank: Tornado
**

Group: Member
Posts: 378
Joined: 21-April 11
From: Clark/Floyd County
Member No.: 25,607





QUOTE(Hertz @ Jun 19 2011, 08:43 PM) *
I've seperated this into two posts because this is a general question and has nothing to do with what was discussed in the previous post. (Also my question in this post is for how SPC issues outlooks currently, since in reality it's highly unlikely they'd buy into my four-outlook idea, let alone this quickly.)

Basically, there is a storm going across the country this week that's already induced concern for tornadoes (some strong), and indications are tomorrow will only be worse than today. This main system doesn't reach my area (per forecasts) until at least Wednesday, but I am outlooked for severe both tomorrow and Tuesday as well. Therefore I'd like it if you'd let me know on what days (if any) this week you think I may be in a 5%+ area for tornadoes, and also if you think there's potential for me to have a day where my outlook will be even higher than 5% (i.e. 10% or greater).

(This could also be accompanied by just giving your predicted tornado outlook for my area each day.)


The best thing I can tell you to do (and I have pseudo taught myself) is learn to read weather maps. That way you dont have to ask and wait for an answer.

Twisterdata.com provides GFS/NAM soundings for periods of time greater than 12 hours, and both Twisterdata and SPC provide RUC soundings (Twister gives you 12 hours where SPC to my knowledge is only the current and previous runs)

I will assume you are familiar with bulk shear, CAPE, EHI, Lifted Index, CIN and dew points... if you are not, read up on them, and find an idea of what ingredients need to be in place for certain kind of weather(in this instance rotating supercells that can produce tornadic weather).


--------------------
Clark & Floyd Counties,
Indiana 2013

Tornado Warnings: 0
Tornado Watches: 1
Severe Warnings: 2
Severe Watches: 1
Confirmed Tornadoes: 0
Go to the top of the page
 
+Quote Post
Hertz
post Jun 20 2011, 11:04 PM
Post #97




Rank: F5 Superstorm
***

Group: Member
Posts: 1,107
Joined: 14-May 10
From: Cleveland, OH
Member No.: 22,778





Sorry, I'm not personally that good or familiar with shear/SRH/helicity type graphics, or hodographs. I'm more tended twoards looking at mainly CAPE and synoptic system maps.

Therefore, I've not escaped today without a noteworthy risk of severe storms or tornadoes materializing, but do you personally see me outlooked above a 5% for tornadoes on any day this week, and if so what day(s)?

If 5% is the highest it is likely to get, then what days will have that risk level?

Finally, the Accuweather article still uses "strong" in describing tornado probability along the storm system tomorrow. This surprises me as SPC is just a 30% slight with mention of tornadoes, but not strong tornadoes. When should this system weaken enough to preclude any more intense tornadoes?

This post has been edited by Hertz: Jun 20 2011, 11:05 PM


--------------------
Let's hope this winter actually happens!

Severe Wx 2013 Cuyahoga Cty


Severe thunderstorm watches: 6/12-6/13 (night), 6/25, 7/10, 7/23, 10/31-11/1 (night)
Tornado watches: 11/17
Tornado warnings: 6/12, 7/10
Slight Risk days: 4/10, 5/21, 5/22, 5/28, 6/1, 6/12, 6/24, 6/25, 6/26, 7/10, 7/19, 7/23, 8/7, 8/27, 9/11, 9/20, 10/6, 10/31
Moderate Risk days: 11/17
High Risk days:
Strongest thunderstorm to date (at CLE airport): 57 mph on 6/25 & 11/17 (thunderless convection on 11/17)
Go to the top of the page
 
+Quote Post
Southern Indiana
post Jun 21 2011, 08:05 AM
Post #98




Rank: Tornado
**

Group: Member
Posts: 378
Joined: 21-April 11
From: Clark/Floyd County
Member No.: 25,607





QUOTE(Hertz @ Jun 21 2011, 12:04 AM) *
Sorry, I'm not personally that good or familiar with shear/SRH/helicity type graphics, or hodographs. I'm more tended twoards looking at mainly CAPE and synoptic system maps.

Therefore, I've not escaped today without a noteworthy risk of severe storms or tornadoes materializing, but do you personally see me outlooked above a 5% for tornadoes on any day this week, and if so what day(s)?

If 5% is the highest it is likely to get, then what days will have that risk level?

Finally, the Accuweather article still uses "strong" in describing tornado probability along the storm system tomorrow. This surprises me as SPC is just a 30% slight with mention of tornadoes, but not strong tornadoes. When should this system weaken enough to preclude any more intense tornadoes?


You probably know more about weather than I do, and it doesnt take a meteo degree to have a rudimentary understanding of a NAM, GFS, or RUC sounding.

If you already look at CAPE, then EHI and Lifted Index are no problem for you because those are just two different forms of Instability indicators, and CIN is the measure of the cap in place.

Shear is wind, nothing to extraordinary about being able to read that.

Stop asking us because all you do is doubt what we tell you and make some reference to an SPC outlook 20 years ago.


--------------------
Clark & Floyd Counties,
Indiana 2013

Tornado Warnings: 0
Tornado Watches: 1
Severe Warnings: 2
Severe Watches: 1
Confirmed Tornadoes: 0
Go to the top of the page
 
+Quote Post
jdrenken
post Jun 21 2011, 07:03 PM
Post #99




Rank: F5 Superstorm
***

Group: SuperModerator
Posts: 33,034
Joined: 26-March 08
From: Columbia, MO
Member No.: 14,521





This thread has run it's course. I will be watching in the other severe threads for any type of discussion related to what went on here.


--------------------
QUOTE
For the record...I AM THE MISSOURI MAULER!


It's a work in progress!

Have a question? Look at our FAQ first.






89.5 FM KOPN Weather Blog

If it is important enough to you, you will find a way. If it is not, you will find an excuse.
Go to the top of the page
 
+Quote Post

5 Pages V  « < 3 4 5
Closed TopicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 

RSS Lo-Fi Version Time is now: 21st October 2014 - 10:09 AM