SPC Outlook Question

[Robert Dewey]

I have had this question since I can remember looking at SPC outlooks...

Why is it that SPC continues a SLGT risk, even if there is a good number of SVR's present? Using last night as an example, it would seem the area in and just downstream of the severe convection would have an elevated risk of SVR's compared to other areas within the SLGT risk ... so wouldn't it make sense that those areas be highlighted with a higher risk?

Perhaps that would be a new product idea ... A "real-time" outlook that would show where the enhanced SVR threats will be in the short term. For example, the daily outlook may show a SLGT, but since severe convection has fired, an "enhanced MDT" would be issued just downstream/in the path of the severe convection.


Disclaimer: I realize that alot of people like to accuse others of "bashing" the SPC ... Please read the post before you answer, and understand that this is a legitimate question

 

[Kevin Scharfenberg]

Robert,

Not speaking for SPC nor my own employers...

As you know, the SPC outlooks are driven by probabilities now. So it seems natural to want to keep driving the probabilities to smaller scales (temporally and spatially) as the event unfolds. However that very quickly gets into stormscale analysis -- is that developing cell going to grow into a supercell, a wind machine, or just be a hailer? Should I bump up the tornado probabilities now, or just the hail numbers, or both? There's simply no way SPC could (or should) do this manually for every developing storm in the CONUS most severe weather days. It's a warning scale problem, right now handled by deterministic warnings. Over the next few years my group at NSSL, with help from the NWS and many other agencies, is hoping to start trying these concepts in real time. Eventually there will be a continuum of hazard probabilities from the climo scale to the outlook scale, to the watch scale, to the warning scale, and even to smaller/shorter scales, but this takes some pretty sharp planning and a lot more science.

 

[rdale]

"Disclaimer: I realize that alot of people like to accuse others of "bashing" the SPC ... Please read the post before you answer, and understand that this is a legitimate question"

So you're taking a cheap shot ;>

SPC outlooks are for planning purposes... Once the storms start - watches and warnings take over as the products of note. Probably not the best use of resources to have a met go back and keep reissuing DY1's when the most important role during sevwx is to stay on top of the storm through MCD's & additional watches.

 

[Robert Dewey]

Got it... I kind of figured the outlooks were for "general purpose" but it was just something that was always in the back of my mind.

I suppose the categorical outlooks are primarily based on number of expected reports, so even if convection fires the previous outlook would suffice, provided there aren't any huge forecast changes (i.e. May 3rd, 1999 ... shortwave was stronger than anticipated by any of the numerical guidance).

 

[Rich Thompson]

Got it... I kind of figured the outlooks were for "general purpose" but it was just something that was always in the back of my mind.

I suppose the categorical outlooks are primarily based on number of expected reports, so even if convection fires the previous outlook would suffice, provided there aren't any huge forecast changes (i.e. May 3rd, 1999 ... shortwave was stronger than anticipated by any of the numerical guidance).
[/b]

If we (SPC) anticipates the change in the threat *before* it happens, or there is no appreciable outlook in an area in need of watches, we'll amend the outlook. Otherwise, the outlook is for planning purposes and doesn't make much difference once you're well into the event. SLGT risk will be upgraded to MDT risk on occasion, but it requires us to develop an internal concensus in favor of the upgrade. The outlooks yesterday had a 30% SIG hail area in the right place, and we just underestimated the tornado threat.

I will say that we make some effort to be consistent within the HIGH risks, such that you'll almost always see PDS tornado watches in HIGH risk areas driven by the tornado probabilities. That can be good when it works out, and it can be not so good when the HIGH risk turns out to be overdone. Think of it this way - the watch decisions are largely independent of the outlook in SLGT risks (and most MDT risks), while the watch decisions have pretty much been made in advance when a HIGH risk is issued.

Rich T.

 

[Kiel Ortega]

SPC outlooks are for planning purposes... Once the storms start - watches and warnings take over as the products of note. Probably not the best use of resources to have a met go back and keep reissuing DY1's when the most important role during sevwx is to stay on top of the storm through MCD's & additional watches.
[/b]

I don't think it's a waste of resources. It's not the SPC's job to track specific storms (though they may "quote" a storm in their MDs) or issue any warnings/wx statements. Their resources are for quasi-synoptic/mesoscale time scale forecasting, so issuing DY1 outlooks while an event is occuring is exactly what they should be doing to give guidence for the shorter time scale forecasts.

 

[Donald Van Dyke]

I don't think it's a waste of resources. It's not the SPC's job to track specific storms (though they may "quote" a storm in their MDs) or issue any warnings/wx statements. Their resources are for quasi-synoptic/mesoscale time scale forecasting, so issuing DY1 outlooks while an event is occuring is exactly what they should be doing to give guidence for the shorter time scale forecasts.
[/b]

It was my understanding though that the MDs (not the Day 1 outlooks) give guidance for the shorter time scale forecasts within the time scale of the Day 1 outlook.

 

[Jonathan Racy]

I'd be interested in knowing how long it takes to make a DY1 outlook, and if that time would be better spent in the MCD area (or if the same forecaster even does both?)
[/b]

DY1+ preparation time scales are highly dependant on the weather situation, internal and WFO collaboration and the forecaster doing the forecast. I'll speak for myself and say that I can take anywhere from an hour to 3 hours to prepare a forecast. That includes: 1/analysis of current weather; 2/viewing the suite of model data and drawing of composite charts; 3/collaborating with WFO/other SPC forecasters as needed; 4/drawing the TOR/HAIL/WIND probabilities, converting them into the categorical SLGT, MDT, HIGH; 5/drawing/collaborating the general thunderstorm lines (and their associated enhanced temporal scales); and, 6/putting it all into words.
Often there are multiple areas of the country to worry about simultaneously.

SPC has 3-4 people on shift 24-hrs a day. At least two forecasters are dedicated to a 24-hr weather watch for the mesoscale discussions and/or watch responsibilities. That also depends on the weather situation (mdt-high risks often see extra staffing).

 

[Carrie Halliday]

DY1+ preparation time scales are highly dependant on the weather situation, internal and WFO collaboration and the forecaster doing the forecast. I'll speak for myself and say that I can take anywhere from an hour to 3 hours to prepare a forecast. That includes: 1/analysis of current weather; 2/viewing the suite of model data and drawing of composite charts; 3/collaborating with WFO/other SPC forecasters as needed; 4/drawing the TOR/HAIL/WIND probabilities, converting them into the categorical SLGT, MDT, HIGH; 5/drawing/collaborating the general thunderstorm lines (and their associated enhanced temporal scales); and, 6/putting it all into words.
Often there are multiple areas of the country to worry about simultaneously.

SPC has 3-4 people on shift 24-hrs a day. At least two forecasters are dedicated to a 24-hr weather watch for the mesoscale discussions and/or watch responsibilities. That also depends on the weather situation (mdt-high risks often see extra staffing).
[/b]

Thanks Jonathan (and Rich) for explaining some of the methods that SPC uses to get things done. It is appreciated, and takes away some of the questions I am sure many of us had.

 

[Kiel Ortega]

So Kiel's request that the DY1 be updated constantly would probably be a little too time-consuming?
[/b]

I never said "constantly updating" (way to take what I said completely out of context and proportion!). I meant that even during an event (because as I interpreted what was said: "it's useless to continue to issue DY1 outlooks once storms are going") the DY1 outlooks should still come out at their normal times because that's what the SPC is there to do...issue the outlooks along with MDs.

 

[Kiel Ortega]

"the DY1 outlooks should still come out at their normal times "

I really don't think I took what you said out of context... SPC issues their DY1 outlooks at the same time every day, whether skies are completely clear across the entire US or a 1000-mile long line of supercells is in progress. RDewey's post that started this thread wondered about the possibility of a real-time DY1 constantly updated during storm events, which I suggested might be a waste of resources. You said "I don't think it's a waste of resources" as "issuing DY1 outlooks while an event is occuring is exactly what they should be doing."
[/b]

Ahhhh...okay, I was misinterpreting (and not completely responding) to his post (it was mostly a response to somethings after that). The way the conversation had gone, I thought we were talking about the current way the SPC does outlooks and MDs and not the hypothetical way that was presented. I guess I was out of the context of the thread, sorry about that. :mellow: :unsure: