• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Severe 2026 Severe Thread 🌩

Arcc said:
Wait, so we have two hatched levels now? Good grief.
Click to expand...
There’s more depending on whether is wind, tornado, or hail is my understanding… fox weather channel did a pretty in depth explanation tonight.. I think hail has 2.. wind has 3 .. and tornado has 4 or 5 categories ( I know there’s 5 strengths of tornados obviously just talking their shaded categories)… it’s different for sure.. what’s funny is they showed these sample maps with high categories around the ATL metro and I was like damn … if only I had this to track lol
 
shortstopsergio said:
There’s more depending on whether is wind, tornado, or hail is my understanding… fox weather channel did a pretty in depth explanation tonight.. I think hail has 2.. wind has 3 .. and tornado has 4 or 5 categories ( I know there’s 5 strengths of tornados obviously just talking their shaded categories)… it’s different for sure.. what’s funny is they showed these sample maps with high categories around the ATL metro and I was like damn … if only I had this to track lol
Click to expand...

That is insanity...
 
Arcc said:
That is insanity...
Click to expand...

It’s actually a good thing IMO. The new system separates maximum expected intensity from probabilities. It gives a lot more granularity, especially in high coverage low intensity events (i.e feeder bands in a hurricane) and low coverage high intensity events (i.e stout EML cases in the plains).

 
jmills152077 said:
It’s actually a good thing IMO. The new system separates maximum expected intensity from probabilities. It gives a lot more granularity, especially in high coverage low intensity events (i.e feeder bands in a hurricane) and low coverage high intensity events (i.e stout EML cases in the plains).

Click to expand...


And will do nothing more than confuse the public and be disregarded.

Think about it this way, right now we have two tornado warnings, regular and tornado emergency. When the TE is issued it grabs eyes because it is simple and shows extreme danger.

If we had Tornado warning levels one through ten, you water down the lower tiers( oh its just level three, it won't be bad)

Adding multiple threats of hatching to already expanded risk levels is just goofy. We don't need to expand the SS to Cat one though ten, nor do we need fifteen EF levels.

Hatching alone was plenty significant enough.
 
Arcc said:
And will do nothing more than confuse the public and be disregarded.

Think about it this way, right now we have two tornado warnings, regular and tornado emergency. When the TE is issued it grabs eyes because it is simple and shows extreme danger.

If we had Tornado warning levels one through ten, you water down the lower tiers( oh its just level three, it won't be bad)

Adding multiple threats of hatching to already expanded risk levels is just goofy. We don't need to expand the SS to Cat one though ten, nor do we need fifteen EF levels.

Hatching alone was plenty significant enough.
Click to expand...
Exactly. It's overkill and will just confuse the general public. The severe risk categories are already confusing enough. That's why WRAL just says levels 1 to 5. They probably only need 4 levels anyway. Just use low, medium, high and extreme. The ones they have now don't make sense when slight and marginal means the same thing, and something that is enhanced is actually greater than something that is moderate.
 
Arcc said:
And will do nothing more than confuse the public and be disregarded.

Think about it this way, right now we have two tornado warnings, regular and tornado emergency. When the TE is issued it grabs eyes because it is simple and shows extreme danger.

If we had Tornado warning levels one through ten, you water down the lower tiers( oh its just level three, it won't be bad)

Adding multiple threats of hatching to already expanded risk levels is just goofy. We don't need to expand the SS to Cat one though ten, nor do we need fifteen EF levels.

Hatching alone was plenty significant enough.
Click to expand...

I'd argue that SPC convective outlooks aren't intended for the general public - more so for local WFOs, media, emergency managers, etc...

Even when the public uses SPC outlooks, I bet the majority of the time they are focused on the categorical outlooks instead of the probabilities. The general public, in general, has extremely poor understanding of probabilities (the same reason people gripe when they get a thunderstorm in the summer with a 30% POP - even though the forecast itself may be perfectly fine), especially if they lack the framework to properly contextualize it.

A 10% probability of tornadoes within 25 miles of a point, while not a generational outbreak, is nothing to scoff at. However if someone hears that without understanding that 10% is significantly higher than the background probability on a random day, their first thought is probably that 10% "isn't that high"

A point brought up in the video I linked that I think some people are missing is being able to (mostly) separate categorical outlooks from probabilities and intensity allows the categorical outlook to focus on the expected impact of the event - which arguably makes it a much more useful product.

The example of a tornado warning doesn't hold up IMO. A tornado warning is a short fused product intended for widespread dissemination and immediate action - it has to be "short and sweet". A SWODY1 doesn't have the same impetus and can afford to be more granular.

Breaking the SS scale into "1 through 10" and the EF scale into "15 different levels" isn't what anybody is proposing or what is happening. The categorical outlooks aren't changing - all that's happening is changing SIG severe from a binary yes/no to a system that gives the forecaster the ability to better communicate the forecasted intensity. Previously, a hatched area could mean one or two EF2 tornadoes or it could mean multiple long track, violent tornadoes - two vastly different situations but previously they had the same treatment on the graphical outlook.

It's perfectly OK to have a forecast not intended for every member of the general public - there's a reason that SPC issues Public Severe Weather Outlooks during high impact days.
 
jmills152077 said:
I'd argue that SPC convective outlooks aren't intended for the general public - more so for local WFOs, media, emergency managers, etc...

Even when the public uses SPC outlooks, I bet the majority of the time they are focused on the categorical outlooks instead of the probabilities. The general public, in general, has extremely poor understanding of probabilities (the same reason people gripe when they get a thunderstorm in the summer with a 30% POP - even though the forecast itself may be perfectly fine), especially if they lack the framework to properly contextualize it.

A 10% probability of tornadoes within 25 miles of a point, while not a generational outbreak, is nothing to scoff at. However if someone hears that without understanding that 10% is significantly higher than the background probability on a random day, their first thought is probably that 10% "isn't that high"

A point brought up in the video I linked that I think some people are missing is being able to (mostly) separate categorical outlooks from probabilities and intensity allows the categorical outlook to focus on the expected impact of the event - which arguably makes it a much more useful product.

The example of a tornado warning doesn't hold up IMO. A tornado warning is a short fused product intended for widespread dissemination and immediate action - it has to be "short and sweet". A SWODY1 doesn't have the same impetus and can afford to be more granular.

Breaking the SS scale into "1 through 10" and the EF scale into "15 different levels" isn't what anybody is proposing or what is happening. The categorical outlooks aren't changing - all that's happening is changing SIG severe from a binary yes/no to a system that gives the forecaster the ability to better communicate the forecasted intensity. Previously, a hatched area could mean one or two EF2 tornadoes or it could mean multiple long track, violent tornadoes - two vastly different situations but previously they had the same treatment on the graphical outlook.

It's perfectly OK to have a forecast not intended for every member of the general public - there's a reason that SPC issues Public Severe Weather Outlooks during high impact days.
Click to expand...

But isn't this entire argument counter-intuitive? If the SPC forecasts are arguably not intended for the public, then why further complicate a system for those who already know how to read the maps?

For instance does the map *need* to differentiate EF2+ VS long track tornadoes? No. That is what the forecast discussion is for.

But what it could do is add even failure points and busts.
 
Arcc said:
But isn't this entire argument counter-intuitive? If the SPC forecasts are arguably not intended for the public, then why further complicate a system for those who already know how to read the maps?

For instance does the map *need* to differentiate EF2+ VS long track tornadoes? No. That is what the forecast discussion is for.

But what it could do is add even failure points and busts.
Click to expand...

The issue isn't being able to read the maps or not, it's about the previous hatching not really giving all that much information and the new hatching communicating different information.

Take 3/15 last year as a point of comparison.

day1probotlk_20250315_1630_torn_prt.gif

The hatched area (10% or greater chance of EF2+ tornadoes within 25 miles of a point) covers six states (seven if you count that little sliver of NC). Obviously the ceiling for the significant tornado threat is very different depending on where in the hatched area you are looking. Towards MS and west AL there was a much higher ceiling than the very fringes of the hatched area. Many would argue that the 30% area represents the highest ceiling/highest expected intensity, but that's not necessarily true. The old hatching system had such a "low bar" that you end up asking "what information is this really giving me if it covers such a large area?" In addition, the way the old system was arranged created significant restraints in the forecast process - introducing a hatched area often necessitated an upgrade in category when it may have not otherwise justified it. You had the colors on the map driving the forecast process, which is backwards.

Probability of ANY tornado and the probability of a SIG tornado are independent events. Think of a QLCS tornado event in the SE cool season where there is the possibility of prefrontal convection. There is probably a greater chance of a tornado within the QLCS due to greater dynamic support/low level shear/the fact that it's a giant line of storms, but the prefrontal storms, while likely having a numerically lower chance of a tornado within 25 miles (more spread out/do they even happen?) have a higher probability of a significant tornado - all to say they're not necessarily connected.

Compared to the new system:

day1_torn_202503151630.png

Everywhere there is a hatched area has a chance of a significant tornado. You could mentally "collapse" all of the hatching into one hatched area that is roughly equivalent to the old hatching system, but now the hatching gives information about how significant the tornadoes are likely to be. The hatching is no longer tied so tightly to the boundaries of the probabilities - arguably representing the actual nature of the situation/threat as opposed to following a fairly rigid set of rules.

There are a lot of decision makers between the level of a member of the general public and a meteorologist/someone with extensive meteorological knowledge - they can (with a little bit of practice) decipher these maps and benefit from the information given but would be lost looking at the discussion. The discussion text really isn't the forecast, it's the discussion of meteorological reasoning behind the forecast and is worthless to people without the requisite background. Having intensity information in the maps is a good "middle layer" between the surface level categories and the technical discussion.

I agree that it is a little cluttered visually and they could probably find a way to layer that information on the map. And I do agree that there is now a little more room for "busts", but I don't think that is enough reason to stick to a more flawed system. A forecast that is more useful is almost always a more specific forecast, which by the nature of the beast, increases the surface area for failures. However, I believe that is a worthy tradeoff here. Why do we forecast snow totals and not just whether or not a location will get snow? The binary yes/no forecast is a lot safer and less likely to fail, but is much less useful. The same applies here.
 
You must log in or register to reply here.
Back
Top