Tornado outbreak forecasted 4 weeks in advance

rdale

EF5
Mar 1, 2004
6,984
500
21
49
Lansing, MI
skywatch.org
There's nothing "misleading" about it at all.. Did you read the story?

"From April 28 on, the ERTAF team highlighted the likelihood of an active period of severe weather three to four weeks into the future. The prediction was especially notable given the pre-season expectation of below-average frequencies of U.S. tornadoes due to the presence of weak El Niño conditions in the tropical Pacific Ocean. "

If you look at their forecasts, most of them during the tornado season were for a normal or below normal tornado count. This one they hit hard would be above average, and it was. It's not like they took the ST route and said "OUTBREAK" every time the GFS 18Z 360hr run had a high CAPE :)
 
  • Like
Reactions: John Farley

Todd Lemery

Supporter
Jun 2, 2014
490
482
21
54
Menominee, MI
It would be nice if they were on to something, I’ll wait before I rely on it for future forecasts. I could say now that I predict next May 22nd through May 29th of next year will be a tornado outbreak and then if it worked out say I nailed it. The truth is that it will take a few times in a row of hitting it right to move the needle for most people. I hope they are on to something, but it will take several years of replication to know one way or the other. If they can do it several times in a row, sign me up!!
 
To clarify, it's great to be able to assemble a forecast based on the data they note in the article, but I've also seen some very sharp people here on ST note similar patterns well in advance. Once someone can forecast a specific day, location ( +/- 200 miles) and severe weather type that far in advance with a better than 70% success rate, I'll be impressed.
 

rdale

EF5
Mar 1, 2004
6,984
500
21
49
Lansing, MI
skywatch.org
The truth is that it will take a few times in a row of hitting it right to move the needle for most people. I hope they are on to something, but it will take several years of replication to know one way or the other.
They have done it for several years, and if you notice that more than 3/4ths of their forecasts were verified this year. I think they are on to something :)
 
Jul 5, 2009
892
633
21
Newtown, Pennsylvania
If this is reliable, it would be great for us chase vacationers that only have a week or two to chase but have flexibility over when to head out to the Plains. I have been looking at the ERTAF page the past couple years, but use it as just one data point. Normally once I am inside my “window” of potential chasing, I make a go/no-go decision one week at a time. So at that point I have the usual models and SPC outlooks to rely upon. I always make the decision based on “week 1”; I wouldn’t head out for a bad “week 1” just because “week 2” looked good. But many others need to make earlier chase vacation decisions, and some years I do too, so getting a 4-week outlook would be great in those situations.

Or, it could lead to complete hopelessness about the season and result in missing some awesome mesoscale accidents 😒
 
Spotting a 'pattern' on a model/set of models a few weeks in advance is not the same as making a reasoned assessment of the likely flow regime based on a number of indices and teleconnections. The formation of analogues based on the various teleconnections can also help. It is possible to forecast sub seasonal patterns like this, and is increasingly being done, and being successful (including in the seasonal range).
 

Jeff Duda

Resident meteorological expert
Staff member
Oct 7, 2008
3,075
1,615
21
Broomfield, CO
www.meteor.iastate.edu
There's nothing "misleading" about it at all.. Did you read the story?

"From April 28 on, the ERTAF team highlighted the likelihood of an active period of severe weather three to four weeks into the future. The prediction was especially notable given the pre-season expectation of below-average frequencies of U.S. tornadoes due to the presence of weak El Niño conditions in the tropical Pacific Ocean. "

If you look at their forecasts, most of them during the tornado season were for a normal or below normal tornado count. This one they hit hard would be above average, and it was. It's not like they took the ST route and said "OUTBREAK" every time the GFS 18Z 360hr run had a high CAPE :)
I agree with both of you to some extent.

There is a difference between making a forecast of "there will be 50% more tornadoes than climatologically expected for this one-week period" and "outbreak on day XXXX". If climo for a week in May is 59 tornadoes (which, without loss of generality, implies 8 or 9 tornadoes per day during that period) and 12-14 tornadoes occur each day spread throughout the country, it is unlikely that any given day would be classified as an outbreak, but the former statement would verify.

ERTAF fits in very well with the "forecast funnel" framework of weather forecasting, so what Victor and his team do is completely appropriate. They are making multi-week (subseasonal) forecasts which cover a large geographical area, which is about all the science can offer right now. An outbreak, on the other hand, is much more focused spatiotemporally, and thus fits into a different part of the forecast funnel than does an ERTAF forecast.

However, I was at the HWT spring forecasting experiment in Norman the same week Victor was, and I can personally vouch that he specifically mentioned "May 20th" as a big day (we were there the week of 6-10 May, and he said it on the Thursday or Friday of that week). At the time that was basically a direct read of this chart, which I recall having shown a maximum right on 20 May. However, given that chart is based on GEFS forecasts, that particular utterance was within the 10-15-day window where medium range NWP can capture it.

Furthermore, ERTAF forecasts only go out 3 weeks, so the headling saying "nearly 4 weeks in advance" is definitely a journalistic stretch. It was 3 weeks in advance at best. ERTAF did hit that particular week as "above average", but it missed the previous week that also verified as "above average". So they did hit one period that happened to be active, but there were misses elsewhere during the experiment.
 

John Wetter

SN President
Staff member
Dec 11, 2005
855
46
11
Maple Grove, MN
www.WxChaser.com
ERTAF had a break-out year this year, that is no doubt. I think this is a fantastic new area of research and really pulls back out look to a more global view as systems move and interact. I was watching it very closely this year and found it to be a great resource for longer term planning and to compare and contrast with the SPC chiclet plots.
 
  • Like
Reactions: rdale