SPC Tornado Data - is Tornado Alley moving east?

[ChristofferB]

This thread is more about an interesting (and somewhat mathematical) discussion than an actual question, so bare with me.

I got a question on social media earlier today about whether Tornado Alley is moving east. I have heard it does and as I am currently in the process of learning Machine Learning I thought it would be fun to analyze the SPC data. At first, just for fun, I just checked the overall mean latitude & longitude for ALL tornadoes between 1950-2017 (excluding zero values as well as Hawaii, Alaska & Puerto Rico). It turned out that Diggins, MO, is the geometrical "Tornado Capital" of the USA (@ 37.169/-92.833).

Doing a linear regression on the data (year vs longitude) I found a positive slope 0.0134 * yr + -93.3613 (where year 1950 = 0). Converting longitude to miles, it suggests that the mean longitude is shifting apprx 2 miles per year to the east.

Now, the correlation here is almost non-existent (0.02! For non-mathematicians: a correlation of 1 is max and 0 is the least - so it is as random as it gets) so there is no validity in these numbers at all. This doesn't surprise me considering the spread of the data in longitude. Although the data is based upon 64.000 tornadoes the data cannot be used for this really given the geography of the States (5 tornadoes in California tilt the data more towards the west than 10 in Kentucky does to the east).

But, how would one determine/calculate whether "tornado alley" is shifting more towards the east? There are some ways of breaking down this question:

- Is Tornado Alley defined as the place where tornadoes appear during "season" (i.e. April to June/July)? I.e. should all tornadoes outside of these months be excluded?

- Would a linear regression of the mean longitude per year be more useful? Or, perhaps a sliding 5 year average?

- Should certain states be excluded completely (even considering months). I.e. is a tornado in New York state or California a sign of Tornado Alley shifting? Not really...

I am using Weka as data mining tool by the way and although I have done my fair share of studies in math and statistics, it was some while ago - and I am just getting started with Machine learning.

Any thoughts on this?

 

[rdale]

https://www.sciencedaily.com/releases/2018/10/181017172846.htm

 

[Quincy Vagell]

Just some speculation (I haven't done thorough research on this or read the entirety of Gensini's study)...

With respect to climate change, if one assumes surface temperatures are increasing and sea surface temperatures in the Gulf of Mexico have also increased, that would probably effect tornado frequency across the Gulf Coast states and vicinity. The warmer air combined with greater moisture transport would, while making some assumptions, allow for a longer tornado season, better ingredients moving farther inland into areas such as Arkansas/Tennessee, higher STP values (Gensini) over an increasingly large area, etc.

On the flip side, recent droughts over the past decade (or longer) across the western United States, when coupled with an assumed surface temperature rise, would likely yield larger T/Td spreads, particularly across the High Plains, closer to drought areas and farther removed from the Gulf of Mexico, where south to southeasterly low-level wind components would have little to no significant change with respect to surface dew-points. Even if you assume that dew-points have stayed the same, a slight increase in temperatures would steal yield larger T/Td spreads and in return, less favorable boundary layer moisture quality.

I also wonder if the 2011 tornado season was enough of an outlier to shift the geographical area of "tornado alley" farther east. It is hard to deny the fact that over the past 10 years or so, there has been generally less tornado activity across the High Plains and more activity over the Midwest/Mid-South/Gulf Coast vicinity), but 2011 sticks out like a sore thumb.

Considering that the atmosphere is a fluid and weather patterns are always shifting (even if it's not a large or otherwise noticeable amount), it's completely reasonable to think that tornado alley can and should shift a bit over the years, rather than remaining completely fixed. However, if such speculation and studies are correct, there may be a larger, more permanent? shift taking place.

This is an intriguing topic and I'd be interested in hearing the thoughts from others. I've skimmed over some of Gensini's study, but maybe I ought to sit down and read it all to get more information.

 

[ChristofferB]

Thanks for some really interesting feedback, that study was the one I had heard of before. It is interesting that they would include STP in the mix as well - and not only tornadoes. I think it is a matter of what knowledge you actually want to extract: "Is Tornado Alley moving east?" is a very broad question (with the definition of "Tornado Alley" being very ambigious, I mean if it would shift 200 miles east, would that not then be the new Tornado Alley?) but if the questions is: "Will tornadoes, or the potential for tornadoes, occur more often towards the east compared to now, in the future?" there is something juicier to work with.

Then again, is 50 tornadoes on one day (April 27th 2011) more significant than 50 tornadoes spread over 50 days? Should an EF-5 weigh more than 1 EF-0? How about, 5 EF-0's?

STP parameter is an interesting choice but what is the correlation of STP and tornado? I mean, we don't know what makes a tornado form (and not form) even during the best circumstances. Maybe tornadoes that don't form when an STP is calculated actually don't form for a specific reason?

The most interesting would be to look at it from a storm chaser's perspective: "Will tornadoes be as accessible in the future?", for example. If tornadoes will be more common in the east/dixie and less common in the Plains the answer would very likely be "No". Any average Tornado in Kansas or East Colorado is much more accessible and watchable than any average tornado in Alabama due to terrain, road network, trees vs fields but also (and correct me if I'm wrong) that Dixie tornadoes tend to be faster and more HP in general.

Btw, I know @Dan Robinson started putting together data about "tornado accessibility/beauty" or the likes of it, right Dan?

I will continue to look into this data and see what I can find but, as you can see, there is a big question about what the Big Question actually is.

 

[adlyons]

Thanks for some really interesting feedback, that study was the one I had heard of before. It is interesting that they would include STP in the mix as well - and not only tornadoes. I think it is a matter of what knowledge you actually want to extract: "Is Tornado Alley moving east?" is a very broad question (with the definition of "Tornado Alley" being very ambigious, I mean if it would shift 200 miles east, would that not then be the new Tornado Alley?) but if the questions is: "Will tornadoes, or the potential for tornadoes, occur more often towards the east compared to now, in the future?" there is something juicier to work with.

Then again, is 50 tornadoes on one day (April 27th 2011) more significant than 50 tornadoes spread over 50 days? Should an EF-5 weigh more than 1 EF-0? How about, 5 EF-0's?

STP parameter is an interesting choice but what is the correlation of STP and tornado? I mean, we don't know what makes a tornado form (and not form) even during the best circumstances. Maybe tornadoes that don't form when an STP is calculated actually don't form for a specific reason?

The most interesting would be to look at it from a storm chaser's perspective: "Will tornadoes be as accessible in the future?", for example. If tornadoes will be more common in the east/dixie and less common in the Plains the answer would very likely be "No". Any average Tornado in Kansas or East Colorado is much more accessible and watchable than any average tornado in Alabama due to terrain, road network, trees vs fields but also (and correct me if I'm wrong) that Dixie tornadoes tend to be faster and more HP in general.

Btw, I know @Dan Robinson started putting together data about "tornado accessibility/beauty" or the likes of it, right Dan?

I will continue to look into this data and see what I can find but, as you can see, there is a big question about what the Big Question actually is.

Reading Victor's paper, he highlights a few of the issues you found with the report data being very noisy and low correlation. Because of the reporting bias triggered by population densities its going to be too much of a control issue in their experiment. They dedicate a whole section to talking about environmental analogues like STP and how its much more effective to use reanalysis data to track potential tornado environments rather than reports. Victor's study is robust with several significance tests validating the correlations. Quincy's hunch about climate changing the moisture and temperature is also something that is under investigation but I am fairly confident is being realized in these studies.

 

[V. Gensini]

As lead author of the study, I'm happy to answer any serious questions about the research.

 

[ChristofferB]

Victor: That would be great. Let me dig through the data as well as read your paper and I will surely have some questions.

My super simple first approach was to do a linear regression on the longitudes of each year's tornadoes and see if there was a trend, but as mentioned - the correlation was non-existent.

 

[Michael Wilkinson]

This thread is more about an interesting (and somewhat mathematical) discussion than an actual question, so bare with me.

I got a question on social media earlier today about whether Tornado Alley is moving east. I have heard it does and as I am currently in the process of learning Machine Learning I thought it would be fun to analyze the SPC data. At first, just for fun, I just checked the overall mean latitude & longitude for ALL tornadoes between 1950-2017 (excluding zero values as well as Hawaii, Alaska & Puerto Rico). It turned out that Diggins, MO, is the geometrical "Tornado Capital" of the USA (@ 37.169/-92.833).

Doing a linear regression on the data (year vs longitude) I found a positive slope 0.0134 * yr + -93.3613 (where year 1950 = 0). Converting longitude to miles, it suggests that the mean longitude is shifting apprx 2 miles per year to the east.

Now, the correlation here is almost non-existent (0.02! For non-mathematicians: a correlation of 1 is max and 0 is the least - so it is as random as it gets) so there is no validity in these numbers at all. This doesn't surprise me considering the spread of the data in longitude. Although the data is based upon 64.000 tornadoes the data cannot be used for this really given the geography of the States (5 tornadoes in California tilt the data more towards the west than 10 in Kentucky does to the east).

But, how would one determine/calculate whether "tornado alley" is shifting more towards the east? There are some ways of breaking down this question:

- Is Tornado Alley defined as the place where tornadoes appear during "season" (i.e. April to June/July)? I.e. should all tornadoes outside of these months be excluded?

- Would a linear regression of the mean longitude per year be more useful? Or, perhaps a sliding 5 year average?

- Should certain states be excluded completely (even considering months). I.e. is a tornado in New York state or California a sign of Tornado Alley shifting? Not really...

I am using Weka as data mining tool by the way and although I have done my fair share of studies in math and statistics, it was some while ago - and I am just getting started with Machine learning.

Any thoughts on this?

It would be interesting to look at the median longitude also. It would keep those California tornadoes from weighing so much more than a Mississippi tornado.

 

[beaudodson]

It would be better to define tornado alley as including a larger geographic area.

Decades ago, someone decided to label Nebraska south to Texas as tornado alley.

It never should have been defined as that.

Sent from my SM-N960U using Stormtrack mobile app

 

[Jeff House]

The original Tornado Alley might be the most chasable. However, I believe impacts to property and especially life are more severe in the South or Dixie Alley. Storms are less textbook, visibility is lower, storms happen at night, and population is more dense in Dixie. Therefore I agree Tornado Alley is misleading. People in the South need to know the impacts.

 

[Trey Thee]

I think there are several problems with the argument that tornado alley is moving east.
First, I recently analyzed SPC data for Oklahoma going back to 1950 and if anything the trend has been slightly higher in recent decades. I attribute this not to a real change in storm frequency but due to reporting and better data. The summary from the SPC data was that tornadoes, at least in Oklahoma have significant up and down years and have since 1950.
My point is data reporting, especially since, say 1980 and especially 1990 with wide spread deployment of radar may look different then prior data because we have a much fuller and complete picture today.

As for your data analytics, I would probably just throw away outliers, California, Maine, Delaware, Vermont are outliers. California is a big state and since 1950 they've probably averaged 7 or 8 tornadoes a year? Don't let the outliers skew your data is my point. The longitudinal limitations above make a lot of sense IMO.

 

[Warren Faidley]

Non-scientific disclaimer......

I remember years ago when more and more chasers started reporting and recording tornado events, the number of reported tornadoes in eastern Colorado, for example, climbed much higher. When Doppler came on line it also effected the count, or at least the number of radar-warned storms. It's almost impossible now days for a tornado in the Plains to go unnoticed. Having said that, over the past 30 years, my average chase destinations have moved from LBB to near DDC and Colorado. It would be interesting to compare severe thunderstorm data with tornado data and see how they correlate, e.g., has everything moved east or just tornadoes.

 

[rdale]

Severe thunderstorm reports are worse than the tornado database (everyone estimates winds at 60mph and everyone measures hail at 1") so that's why Gensini used "environments" and not just raw reports.

 

[Marshall Stoner]

It would be better to define tornado alley as including a larger geographic area.

Decades ago, someone decided to label Nebraska south to Texas as tornado alley.

It never should have been defined as that.

Sent from my SM-N960U using Stormtrack mobile app

The plains seem to have a highest prevalence of supercells due to frequent high helicity near the dryline and very steep mid-level lapse rates. The issue is the higher LCLs cut back on the number of storms that actually produce tornadoes. The original plains tornado ally would more accurately be called "supercell ally" or "big hail ally". Supercells are less frequent as you go further east, but when you do get them they are more likely to produce tornadoes, especially in the deep south where very low LCLs are common.

 

[Jeff Wright]

We have had an active season in Alabama to be sure. Thank you Vortex SE.