Fewer storm days but tornado-clustering a trend? Implications? Plausibility longterm?

[Rob H]

Experts inferring something from data doesn't mean it is so (see Tufte's analysis of the Challenger explosion for a great example). But I'm not throwing the authors under the wheels here because, this paper doesn't appear to be saying anything differently than what I said - it's plainly stated that due to the nature of the data there are likely guesses but they can't apply certainty. Unless we all get access to the study, it's possibly a little early to comment on anything besides the angle that The Smithsonian is pushing in that article, which really seems to be "climate change is making fewer tornadoes, but bigger tornadoes". Because that's what brings the readers in.

"What other causes could there be?" Occam's Razor suggests maybe we just have fewer tornado days and bigger tornadoes hitting all the extra structures that have been built lately. It's a pretty damn small data set we have to work with and reporting/damage estimation are still almost completely dependent on a) someone reporting a tornado and b) a tornado hitting something. The point isn't that we should have more tornadoes reported now, it's that tornado counts are fundamentally imperfect so it's difficult to compare data from 1980 to 2014 for a bevy of reasons.

 

[rdale]

Difficult - but not impossible Two papers by two respected meteorologists in one year should stand for something. But I think we've made our point(s)...

 

[Adam Lucio]

I think its simply the fact that back then you had fewer tornadoes being documented. There werent 90 thousand chasers, spotters, and bored locals on every tornado like there is nowadays. Back then, a couple tornadoes rolled through some towns or were easily visible from a populated area and they made the headlines, the "other tornadoes" werent witnessed like they are today.

 

[rdale]

I think you missed what this report says It notes that there are fewer tornado DAYS now than in the past. Since 90 thousand chasers are on every possible tornado producing storm, if everything was the same we should have MORE days where tornadoes were reported than less days...

 

[Shane Adams]

(My emphasis) “In effect, there is a low probability of a day having a tornado, but if a day does have a tornado, there is a much higher chance of having many tornadoes,” the researchers write.

I may be wrong, but I think what Adam was trying to say was, the reason that (fewer) tornado days in 2014 seem to have more tornadoes when they do occur, is because of the increased reporting, i.e. 90,000 chasers etc etc. I don't see where he was debating/questioning there are fewer tornado DAYS. Just that there's a reason many tornadoes on a given day in the past were not documented VS today.

 

[John Farley]

I guess I have to agree with the authors of the paper that it is hard to attribute it to reporting when the trend is opposite between tornado days and the number of tornadoes on days when at least one is reported. The effect of reporting should be the same for both - if more tornadoes are being seen and reported now than 40 years ago, then both the number of tornado days and the number of tornadoes on days whey they do occur should be going up. But these two indicators are moving in opposite directions. Also, remember that F-0 and EF-0 tornadoes were not included in the data. That is where the biggest reporting effect occurs, considering that something like half of all tornadoes are EF-0 and many of these do no damage. By definition, an EF-1 or stronger tornado has to do SOME damage, and thus is much more likely to be reported, probably even in times where spotters and chasers weren't out documenting every landspout and every dust whirl under rotating clouds.

 

[rdale]

(My emphasis) “In effect, there is a low probability of a day having a tornado, but if a day does have a tornado, there is a much higher chance of having many tornadoes,” the researchers write.

I may be wrong, but I think what Adam was trying to say was, the reason that (fewer) tornado days in 2014 seem to have more tornadoes when they do occur

But the fact that there are fewer tornado days overall doesn't mesh with the increased number of reporting people out there...

 

[Shane Adams]

But the fact that there are fewer tornado days overall doesn't mesh with the increased number of reporting people out there...

No, it doesn't. But the point I was making was, the increased number of tornado reports per tornado day does mesh with the increased number of people reporting. As for the fewer overall days themselves, I got nothing.

 

[rdale]

the increased number of tornado reports per tornado day does mesh with the increased number of people reporting.

That I'd agree with.

As for the fewer overall days themselves, I got nothing.

Me either - which is why I'll go with the conclusions from their research and say "something" is changing in the atmosphere "for some reason" and it has nothing to do with the number of spotters, 88D technology, or cell phones

 

[Patrick Marsh]

Man I wish I had seen this before now. By the way, if someone had a question with what we did, he or she could always ask me.

 

[rdale]

So, did you use any statistics when developing these "claims"?

 

[Patrick Marsh]

Yes. We used them. But carelessly.

 

[Rob H]

You're going to have to redo the study when a stalled front over KS next year leads to a 600 tornado outbreak sequence.

 

[Adam Lucio]

It can tie into there being fewer tornado days, because how many tornado days consist of ONLY weak tornadoes. Thats where I was going with that...its interesting to think about either way though. I'd be ok with years like 2013, quality over quantity. I'd rather see 5 Rozels than 20 bird farts.