SPC Forecasting changing?

[Shane Lee]

Has anyone else noticed a change in SPC forecasting?

First a disclaimer: I'm certainly not wanting to "call anyone out" at all. I very much appreciate the forecasters and the work they do to forecast a very dynamic atmosphere, so please don't misunderstand. They ABSOLUTELY save lives. I'm just wanting to ask: "Hey...a SLGT used to mean a fairly firm chance of severe weather and a "MDT" meant we were probably going to have a very bad day. HIGH risk scenarios used to mean that this set up is going to be historic....now it doesn't seem that way anymore. What changed?"

What brings this to mind is at least two MDT risk forecasts in recent memory when most all the severe weather happened outside the MDT risk area. Yesterday, 3 April 2014 was forecast to be a scenario firmly within the MDT risk criteria (I even heard it mentioned as a "high end MDT") in Arkansas and the lower Mississippi valley. Almost absolutely nothing happened at the center of the MDT in Arkansas after being in a tornado watch for around 15 hours. A few significant hail reports and three tornado warnings, two of which on the same storm. The lion's share of the storms happened in the wee hours of this morning after the MDT risk was pulled and the storms were in a squall line structure. The other event to which I'm referring happened last year...but I can't remember exactly when. The event was along the same lines, however....MDT forecast and almost nil severe weather. What causes these events to be forecast MDT all day...and then nothing happen? (I know, broad question)

Thank you, in advance, for answering my questions.

 

[Rob H]

It's the Broyles Bust in action.

Seriously though, high shear events are really hard to forecast. You get a 60kt LLJ and 65deg dewpoints and it's either apocalypse or what happened yesterday. A "risk" is still just that - there is no guarantee that any risk will verify. SPC has gotten a lot better over the last ten years, but Arklatex and Dixie Alley are still difficult places to assign risk. Relatively little research is done in those areas, and it is a challenge to observe storms there.

 

[Dan Robinson]

SPC's track record is remarkable and getting better every year. The very rare bust once or twice a year isn't an issue IMO. The forecasts are made with all of the available data, but the atmosphere sometimes doesn't follow.

Yesterday had all of the parameters in place. I think what happened was that the EML advection was not as strong into the area as forecasted, resulting in a relatively uncapped environment that allowed too many storms to go up too quickly. If the cap had been slightly stronger, that MOD risk would have verified easily. There's no way to see that type of thing coming without special soundings across the risk area. I'm not sure if that's done outside of the biggest events.

 

[Jason Boggs]

The morning convection screwed everything up further north. Down south into Northern TX along the dryline and away from the main forcing, the clouds cleared out early allowing instability to ramp up with discrete supercells.

 

[chrisbray]

Yeah, the day of the event I had a bad feeling about it; anytime you get that much early convection/cloud cover it can reallyscrew things up. I wonder if also the timing of the trough/wave was too late in the day/evening to really help? Seemed like the surface winds were pretty badly veered

 

[James Gustina]

I think they did the best they could considering this entire system was an extraordinarily difficult one to forecast. All the model variability and less than stellar upper-air observations made this a real tricky one from both a chasing and risk forecasting standpoint. Unlike the setups that will no doubt come along in May, these early season setups tend to be far more conditional (e.g. It's more of an "if" than a "when" on the cap breaking this time of year) which usually means everything has to be near perfect or it'll crap the bed.

The other reason is why it seems like high risks aren't the same anymore is probably because only the big high risk days get remembered. Most of the "busts" (excluding April 26, 2009. Bleh.) tend to be forgotten. Everyone remembers when a wedge goes tearing through a small prairie town, but no one remembers when they almost had a tornado outbreak.

 

[Mike Johnston]

To my knowledge, there have been no significant changes in the SPC's policies for several years. I believe the last one involved a minor adjustment to their criteria for when a Public Severe Weather Outlook could be issued.

Remember that a pretty significant rationale for the April 3 moderate outlook was the prospects for large hail. While this wasn't verified everywhere in the outlook area, it was verified in a number of locations. It's not always just about tornadoes. And, the tornado watches that were issued seem justified.

Also, there were many legitimate severe thunderstorms. And, in a monitoring of radar echoes that day, many storms seemed to be wound up and on the verge of displaying a hook. Many of them at least displayed a kidney bean shape with very high VIL's. I don't think the Moderate outlook was that far off.

 

[Rick Schmidt]

I agree with Mike J., and I think about that aspect frequently........ grades of risk area are meant for ALL kinds of severe weather. You could actually have a Moderate risk, and not have any tornadoes at all; but lots of severe thunderstorms with large hail and damaging winds. That doesn't happen very often, but it's certainly possible.

 

[Zack Hargrove]

This seems to be a classic case of short memory to me. The SPC has only gotten better over the past 10 years+

 

[Matt Hunt]

I was actually surprised with how high the tornado risk was after looking at models showing 850's out of the SW. It looked like more of a linear event to me, and that's pretty much what happened, or at least a broken line. But, they know more than me! Maybe that's part of the problem, as sometimes the more you know, the more you tend to over-analyze things. I chased a PDS tornado watch last year that resulted in 0 tornadoes. It happens. I haven't been chasing long enough to know how often it happens, though, and for the most part I'd say they get it right.

 

[Joshua Nall]

What causes these events to be forecast MDT all day...and then nothing happen?

I wanted to focus in on the "all day" aspect of your question. And others have said this, but it seems to me more times than not SPC will stick with their original risk level throughout the day, even when it looks like things aren't coming together as expected. Thursday I noticed part of the 2000Z wording was something along the lines of "we could still have a big tornado day if this and this and this happen"... then I believe they mentioned some of the reason why it was looking more and more like it wouldn't. Not that they don't back off of specific risk levels if very obvious limiting factors are noticed. I don't guess their goal is to provide weather enthusiasts with unbiased risk information.

 

[kevin-palmer]

Does anybody know when the SPC will start using the new risk categories?
http://www.washingtonpost.com/blogs/capital-weather-gang/wp/2014/01/15/weather-service-proposes-expanded-number-of-severe-weather-risk-categories-in-2014/

It seems like April 3 would have been an enhanced risk day with the new scale.

 

[Paul Knightley]

I think it's always best to go back and look at the definition of 'forecast': (e.g. http://www.collinsdictionary.com/dictionary/english/forecast?showCookiePolicy=true)

[h=4]verb[/h]

• to predict or calculate (weather, events, etc), in advance

[h=4]noun[/h]

• a statement of probable future weather conditions calculated from meteorological data

Forecasting the random chaos of a fluid like the atmosphere is always going to be inherently difficult. Forecasters must use a combination of numerical models (which estimate their starting points, and thus are going to be subject to ever increasing errors, perhaps very quickly in the case of hi-res mesoscale, convection-allowing models), their own judgement and skill, surface observations, historical analogues, etc etc. One can only assign areas of enhanced risk of severe weather, based on these and other factors - one can't make a perfect deterministic forecast of such.

Skill is measured not just on one event, but on the average outcome of forecasts - how many of us chasers get our forecast areas spot on each time?

Obviously, as a forecaster, I'm going to defend my profession! But I am pragmatic with it - a forecaster is her/his own harshest critic. If I get a forecast 'wrong', it really annoys me! (I would use a more suitable vernacular but I'm in (reasonably!) polite company!). But I also try to learn from it - why did it not go as expected? There's usually some kind of signal if you go back and look at the data - but sometimes it's just down to pure, random bad luck. Random chaos is rife in the atmosphere - the fact we can offer useful predictions at all is still an incredible feat of human ingenuity.

 

[Rob H]

I was actually surprised with how high the tornado risk was after looking at models showing 850's out of the SW.

Why did this surprise you and what don't you like about SW 850s? From my experience, taking the 850s out of context from the 850mb cyclone characteristics, moisture transport, and overall directional shear quality leads to a bad rule of thumb that I see a lot of people repeating.

 

[Skip Talbot]

How about these 850's?

Joplin.

Like Rob said, I've seen chasers dismissing setups just off the 850's, and it's a mistake. It might be more of an issue out west, say over the TX PH where your veered 850's are pulling hot dry air off the desert southwest, but with some kicking speed shear you can still get some pretty high helicity values with veered 850's. Yeah it can be a signal your directional shear might not be as a good, but you have to look at these winds in relation to the other heights and in the midwest, the moisture trajectory is often less of an issue.