SPC Outlook Percentages To Change 2-14

prob_to_cat_day1_seetext.jpg


Mike
 
In addition to the new 10% tornado probability, I see that they have added a 60% probability area for the Day 2 outlook. A HIGH risk can be issued on Day 2 should there be a 60% hatched area, the way I read it. In addition, MDT risks may be issued on Day 3 outlooks if there is a 45% hatched risk area. Hopefully as the models get better, the confidence in a forecast of an event 2-3 days out can be great enough to go for a HIGH (MDT) on a Day 2 (Day 3) outlook.

Also, the PWOs will be issued for moderate risk Day 1 Outlooks only when there is at least a 15% tornado or a 45% wind probabilitiy forecast.
I do like the change to only issue PSWOs for the higher-end MDT risk events. It seemed like there 20 PSWOs issued in 2005 (I doubt there were that many, but my point is that they seemed much more commonplace than in years past) with the change of policy to issue PWSOs for all MDT risk events. The upcoming change seems like a good compromise between highlighting days with an enhanced threat of severe weather (not just for those 0-3 days per year with HIGH risks), and keeping the products "high-end" enough to evoke the desired, "you REALLY should pay attn to the skies today" response.

EDIT: Apparently, I wasn't looking closely enough to notice that they got rid of the 25% and 35% in favor of a 30% risk. In addition, I see that there can be a SLT risk with 45% unhatched hail. IIRC, previously, 25% hail was the max for a SLT risk, with 35% risk hail requiring a MDT risk, so that's quite a change!

Assuming the probability forecasts for similar events stay the same this year (i.e. an event like 6-9-05 will see similar probs this spring as was forecast last spring), I don't expect to see as many MDTs, that's for sure!
 
Assuming the probability forecasts for similar events stay the same this year (i.e. an event like 6-9-05 will see similar probs this spring as was forecast last spring), I don't expect to see as many MDTs, that's for sure!

Agreed! Unless they are planning on adjusting the probabilities to account for different numbers of reports (it does say there the reasoning was to account for MORE reports in recent years...so I'm unsure). That's what I was inquiring about if anybody knew in my initial post.

We shall see how it goes. I do like the PWO change though. Last year was probably too many (especially in the spring, they were up like every day). The change should isolate the dangerous events more.
 
Question for the SPC guys...these probabilities are absolute, right? In other words, if the forecaster believes there will be a 60% chance of a severe report within 25 miles of a point given initiation, but only a 40% chance of initiation, the probability forecast would be 0.6*0.4=0.24 or 24%, right? If so, it seems to me it will be very rare to get a 60% contour on a day 2 outlook. How often will anyone forecast an outbreak AND be highly certain about initiation at day 2?
 
Kevin,

The SPC probabilities are unconditional, meaning that they inherently include both the prob of tstms and the prob of severe given a tstm.The SPC are subjective in that we don't explicitly produce conditional severe probabilities.

A few examples might help:

1. Strong cap, strong instability along the dryline. If a storm goes, it will be severe (large conditional severe prob of, say, 80%), but the chance of intitiation is small (only 10%). The unconditional severe probability would be .8 X .1 = .08. That situation would be outlooked as a 5% area.

2. Weak cap, weak instability along cold front. Storms are likely (80%), but chance for severe storms is small (10%). The net result would be the same as the first example.

I'm glad to see that the "prob to cat" conversion table can be deciphered!

Rich T.
 
Back
Top