2014 U.S. Tornadoes and Recent Trends

[vortex95]

http://www.spc.noaa.gov/climo/online/monthly/newm.html

 

Oct avg is 61 (69 prelim) and Nov avg is 58 (27 prelim).

Prelim number for Dec so far is 27 and avg is 24.  Although

an isolated tornado or two this weekend along the Gulf

coast is possible, that essentially appears to be it for the

year.  The 27 prelim count this month is about 10 less for

a final count when you remove duplicate reports from

12/23.  Yearly avg is 1253.

 

I estimate the final count for 2014 will be ~835.  That is

an extraordinary deviation from the mean.  67% of avg

6% lower than 2013 and 8% lower than 2012!  This is

the lowest count since 1988 when 702 tornadoes occurred

(1988 had a big CONUS drought as in heat and lack of

rainfall and that keep the tornado count down).  1989

is second with 856.

 

Here is a graphic that shows "detrended" tornado counts over the

years.  This helps to account for the better detection and documentation

over time.  Not perfect, but better than just the raw numbers.  Look

at 2014...about two standard deviations below avg!

 

http://www.spc.noaa.gov/wcm/2014/1953-2014-detrend-thru-nov.png

 

Look at the last 15 years...I see more well below avg than well above. 

Another way of looking at it, there are more wild swings in the past 15

years.  True, but the door swings both ways and unlike say precipitation

(too wet or too dry) the lack of tornadoes is a good extreme regardless

of how you look at it.  So with tornadoes, the more wild swing hypothesis

due to climate change cancels things out more or less in the long run.  Yes,

there will be big years and devastating outbreaks, but that is part of normal

climatic variation.  As we have seen, for some reason the very occurrence

even a single tornado is somehow "unusual" and makes national headlines

routinely, even if it hits nothing or does minor damage with no harm to anyone. 

When the avg is 1253 in this country, most of them "weak", this can be

considered the norm, not unusual.

 

Another item, this is the third year in a row where U.S. tornado fatalities have

decreased.  It is not just regression to the mean either due to the very high

toll in 2011.  Last three years the numbers have been 70, 55, and 46.  This

has more due to with random chance (or "luck") than climate, as we know all it

takes is one badly placed intense tornado to make it a well above avg year (i.e.

Joplin), but to level the playing field on an argumentative and prevailing public

mindset level, a decrease is a decrease, and that is a good thing.   Contrast if it

went up 2012-2014, then the klaxons would probably be sounding and this would

be utilized as "proof", even though as I said, this is random chance, either way. 

Mesoscale event placement is so far below mean large scale climate it is basically

irrelevant to the issue.  Another way of looking at it is that fatalities have resumed

the typical in the last 3 years to what has been occurring in most years going back

to 1975.  Two main things at work here...1) watches/warnings for tornadoes have

improved considerably in the last 40 years, and 2) the number of people in this

country has increased considerably, from about 216 million in 1975 to 316 million

in 2014.  So you have one item that decreases the risk of fatalities and one item

increasing the risk of fatalities.  You can crunch the numbers all you want here, but

overall it at worst appears to be status quo in the long term mean.

 

One more item, although the number of total tornadoes per year is going up

overall (and that is more a better detection/documentation fact than anything), the

number of F3/EF3+ has shown a slight decrease in trend.  These are the ones that

do the most damage and cause the most fatalities, and what we should be most

concerned with.

 

http://www1.ncdc.noaa.gov/pub/data/cmb/images/tornado/clim/EF3-EF5.png

 

Bottom line there is reasonable evidence currently that tornadoes in this

country both from a meteorological and fatality point of view are not getting

worse.  The U.S. represents only 6.6% of the land on the globe, but is the tornado

capital of the world, and has a very extensive tornado database, so looking at

tornado trends here is better than anything else we currently have to measure

overall activity. 

 

Questions/comments welcome.

[Sunny and Warm]

Thank you for this interesting report and graph.  A question about the data.  How have you accounted for the probable fact that we are better able to detect and count tornadoes today versus yesteryear when we had lesser technology and fewer people and/or chasers?  TIA

[vortex95]

Thank you for this interesting report and graph.  A question about the data.  How have you accounted for the probable fact that we are better able to detect and count tornadoes today versus yesteryear when we had lesser technology and fewer people and/or chasers?  TIA

I did say that in "better detection and documentation" in the 6th paragraph above.  The "detrended" graphic tries to account for this somewhat.  Either way, unquestionably we detect and document many more tornadoes today, esp. the brief, weak ones, than even 20 years ago, simply due to better technology and awareness.  I suspect the actual annual tornado count for the U.S. is more like 1400-1500 as there are large areas of the West and even the Plains where they are missed due to remoteness.  So the sub-850 count likely in 2014 is even more of an anomaly compared to 25 years ago.  In 1989 there were 856 tornadoes, and that was pre-NEXRAD, pre-NWS Modernization, and storm chasing was still in its early stages overall.  Video devices were less common and not nearly as conveniently portable then, so many were missed or went undocumented.

[andyhb]

Another item, this is the third year in a row where U.S. tornado fatalities have

decreased.  It is not just regression to the mean either due to the very high

toll in 2011.  Last three years the numbers have been 70, 55, and 46.  This

has more due to with random chance (or "luck") than climate, as we know all it

takes is one badly placed intense tornado to make it a well above avg year (i.e.

Joplin), but to level the playing field on an argumentative and prevailing public

mindset level, a decrease is a decrease, and that is a good thing.

 

It's also important to remember that these years are still deadlier than a number of other more active years over the past 25 years. I think the concern regarding a "minimum" in tornado fatalities is a valid one, as increasingly effective warnings/technology/etc. is at least partially offset by population growth and the fact that there are simply more people in the path of these things nowadays.

 

This is also coupled with the fact that it seems there is a growing amount of people out there who would rather get one last glance at what is coming towards them rather than seeking shelter. I'm surprised the case of a strong tornado tracking down a clogged interstate hasn't reared its ugly head yet, but we won't like what happens when it does.

[vortex95]

It's also important to remember that these years are still deadlier than a number of other more active years over the past 25 years. I think the concern regarding a "minimum" in tornado fatalities is a valid one, as increasingly effective warnings/technology/etc. is at least partially offset by population growth and the fact that there are simply more people in the path of these things nowadays.

 

This is also coupled with the fact that it seems there is a growing amount of people out there who would rather get one last glance at what is coming towards them rather than seeking shelter. I'm surprised the case of a strong tornado tracking down a clogged interstate hasn't reared its ugly head yet, but we won't like what happens when it does.

Yes, this is correct.  The growing population is likely offsetting our improvement in watches/warnings

simply b/c there are just more of us "in the way" so to speak, and there are a host of social issues when

it comes to people actually acting on these watches/ warnings. I would argue missed or inadequate

watches/warnings are not the real problem anymore,  If anything, too much warning and hype, primarily

from non-govt agencies, has made a significant part of the public indifferent and apathetic.  There is a

fallacy mindset that "more is always better".  That is not the always the case.  There comes a point of

diminishing returns and when you "cry wolf" too often, it is a fact that some will inevitably get to the point

they will get sick of it all and tune it out. Of course, this is nothing to do with climate or changes to it, it is

a social issue.  The point is progressively more population and infrastructure will unfortunately lead to

higher death tolls and damage as time goes on, even if the climate stays exactly the same since storms

and their extremes are part of the normal climatic variation.  I think this sometimes gets conveniently

pushed aside in some discussions/debates, but it is a huge factor in escalating weather tolls. Let me put

it this way, in 1950, the U.S. population was about 151 million, in 2010, it had nearly doubled to 309

million.  That is a non-trivial increase, and when you pack a lot of that population in small regions, like

California or the East Coast, that makes it even more vulnerable to very high tolls in individual events.

[WxKevinM]

One more item, although the number of total tornadoes per year is going up

overall (and that is more a better detection/documentation fact than anything), the

number of F3/EF3+ has shown a slight decrease in trend.  These are the ones that

do the most damage and cause the most fatalities, and what we should be most

concerned with.

 

http://www1.ncdc.noaa.gov/pub/data/cmb/images/tornado/clim/EF3-EF5.png

 

 

Great post, I'm kind of curious on peoples thoughts about the slight decrease in violent tornadoes (F3/EF3+). Could that trend, although slight, be a product of better construction where buildings are able to withstand more? I would think that with an increased population creating more human impacts, that number should trend even or slightly positive.

[andyhb]

Great post, I'm kind of curious on peoples thoughts about the slight decrease in violent tornadoes (F3/EF3+). Could that trend, although slight, be a product of better construction where buildings are able to withstand more? I would think that with an increased population creating more human impacts, that number should trend even or slightly positive.

 

Likely related to the increase in scrutiny of damage ratings.

[vortex95]

Great post, I'm kind of curious on peoples thoughts about the slight decrease in violent tornadoes (F3/EF3+). Could that trend, although slight, be a product of better construction where buildings are able to withstand more? I would think that with an increased population creating more human impacts, that number should trend even or slightly positive.

 

Don't assume better construction across the board as time goes on.  I've read about and seen presentations on tornado

damage surveys from engineers who are also meteorologists, and all too often they seem to find building practices questionable

or not up to standards.  How new homes are anchored to their foundations are a biggie.  Are anchor bolts every 3 feet or 5?

What kind of anchor bolts are being used?  How about the roofs?  Often losing the roof on a home, it is declared a total loss. 

Keep the roof, and that is a big difference.  So how is the roof anchored?

There is a mindset that still exists that if a home is swept off its foundation, that must mean a violent tornado. 

Well, if an home is not anchored well or properly to its foundation, even a weak tornado can sweep it off

its foundation.  One of the signs of this is look at the home itself.  So it was moved off its foundation, what is

the structural condition of the home?  If it is mostly intact, then you probably do not have a violent tornado. 

Also, look at smaller erect things near and around the home, such as the mailbox.  Are those left standing? 

What about any trees in the vicinity, are they flattened and stripped of bark, or still standing?  Point is that

when you get strong wind impacting a large surface area, moving it isn't has difficult as you may think (look

at tractor trailers), so a smaller object actually stands (no pun intended) a better chance of surviving or not

being moved in a weaker tornado.  In a violent tornado, often everything gets scoured, ripped apart, not only

by the wind, but also from numerous pieces of small and large debris you find in a more intense tornado.

[vortex95]

It's also important to remember that these years are still deadlier than a number of other more active years over the past 25 years. I think the concern regarding a "minimum" in tornado fatalities is a valid one, as increasingly effective warnings/technology/etc. is at least partially offset by population growth and the fact that there are simply more people in the path of these things nowadays.

 

This is also coupled with the fact that it seems there is a growing amount of people out there who would rather get one last glance at what is coming towards them rather than seeking shelter. I'm surprised the case of a strong tornado tracking down a clogged interstate hasn't reared its ugly head yet, but we won't like what happens when it does.

I read some documentation from SPC personnel awhile back about the 5/3/99 OK outbreak.  You look at the shear number of

tornadoes, how strong a lot where, and populated areas impacted, esp. the long tracked F5 that devastated Moore.  There

were 40 fatalities in OK.  Now, think about the incredible non stop coverage local TV stations gave, it was still daylight, and

most tornadoes were quite visible from some distance.  Some areas could see F5 Moore tornado from a distance that would

end up hitting their immediate area over 30 minutes later.  One thing a lot of people have is confirmation bias.  They need to

see "it" before actually acting, despite credible warning sources.  Well, unlike Joplin, you really did not have this problem.  Plus

the live TV coverage showing the Moore tornado in progress would further negate this problem.  Yet still, there were 40

fatalities.  It was said this toll was likely the minimum that could be expected with an outbreak like this esp. when large

violent tornadoes move through populated areas.  The warning process was outstanding, and yet still.  This is evidence

that despite excellent warnings, people still lost their lives, so we see how good the system works, yet it still isn't perfect

despite our best efforts!  As much as I don't like saying it, when a large violent tornado moves through a densely populated

area, chances are very high currently that there will be multiple fatalities.

 

The curiosity/gawking issue is unquestionably a factor.  Sometime people will freeze in awe when a tornado is bearing down

on them as the "fight or flee" instinct is conflicted.  Plus, having easy access to video compels a lot not take cover as

quickly as they should.  A lot of this is human nature and it is not a "fault" per se.  Many of us are naturally

amazed by something so imposing and powerful!  Yet common sense and plain erring on the side of caution

should prevail here.  The times I have been storm chasing in the Plains, the one thing I swear by is visibility! 

If you can see it, then that makes things so much less complicated.  If visibility is poor, give that storm that has

a TVS on it a wide berth even with Radarscope handy.  Things can change so fast you can be caught

completely off guard!  Darkness?  That is quite unnerving as well, even with frequent lightning to illuminate

the tornado.  It is still hard to determine motion, let alone discern rotation in the clouds for possible satellite

tornadoes around the main tornado!

[Quincy]

Let me put it this way, in 1950, the U.S. population was about 151 million, in 2010, it had nearly doubled to 309

million. That is a non-trivial increase, and when you pack a lot of that population in small regions, like California or the East Coast, that makes it even more vulnerable to very high tolls in individual events.

I'd be interested to see what the rate of increase with population was in Midwest and Plains cities during the same time. (% rate and population increase in hundreds of thousands/millions) California does not see many deadly tornadoes, but the East Coast sees them a bit more regularly. One could argue that the PHL-BWI-DCA area is at highest risk due to their population density and relative frequency of tors/sig tors.

[nwburbschaser]

One could argue that the PHL-BWI-DCA area is at highest risk due to their population density and relative frequency of tors/sig tors.

Although the population density is nowhere near the East Coast , I think the Madison-Chicago-Milwaukee area is at a significant risk for a high end tornado.  If you look at the last 30 years, all three of those cities had an EF-5 tornado within a decent proximity. Oakfield, Barneveld, and Plainfield were in that region.  

[Hoosier]

I'd be interested to see what the rate of increase with population was in Midwest and Plains cities during the same time. (% rate and population increase in hundreds of thousands/millions) California does not see many deadly tornadoes, but the East Coast sees them a bit more regularly. One could argue that the PHL-BWI-DCA area is at highest risk due to their population density and relative frequency of tors/sig tors.

 

 

That PHL-BWI-DCA corridor has a high ceiling in terms of death/destruction - what could happen - but the odds of it happening there are not as great as in/near the cities of the Midwest/Plains.  Not trying to minimize things as any tornado is capable of taking lives, but the violent EF4/EF5s produce a disproportionate amount of fatalities and those don't occur nearly as often in or around the I-95 corridor.

 

IMO, places like St. Louis, Chicago, etc. have a higher threat of a big disaster given where they are located and that they've had multiple violent tornadoes in their metro areas since 1950 (and also going back through records prior to 1950). 

[Hoosier]

a few maps

 

 

(E)F2 or greater tornadoes

 

 

 

(E)F4 or greater tornadoes

 

 

 

tornadoes that produced at least one fatality (any strength)

 

[FHSC Sam]

So, here's something somewhat more off-topic, but not by a whole lot. 

We've been seeing the decrease in tornadoes across the States, but my question is, is it part of a pattern that will reverse in time, or is there an underlying cause, such as climate change resulting in a shift or decrease in prime setups in the spring? 

I'm still more enty level to severe weather and atmospheric sciences, so I'm still learning how to figure this out.

Thanks

[wxtrackercody]

So, here's something somewhat more off-topic, but not by a whole lot. 

We've been seeing the decrease in tornadoes across the States, but my question is, is it part of a pattern that will reverse in time, or is there an underlying cause, such as climate change resulting in a shift or decrease in prime setups in the spring? 

I'm still more enty level to severe weather and atmospheric sciences, so I'm still learning how to figure this out.

Thanks

The link between climate change and tornadoes across the USA is unclear at best. On one end of the spectrum, a warming atmosphere equates to more available moisture for tornado outbreaks; on the other end, the decreasing temperature differential between the mid- and upper-latitudes is expected to cause a weakening of the jet stream--reduced wind shear for tornado outbreaks. The best hypothesis I've seen so far is that the numbers of days featuring tornadoes should decrease, but there should be more tornadoes on the days that see them.

[FHSC Sam]

The link between climate change and tornadoes across the USA is unclear at best. On one end of the spectrum, a warming atmosphere equates to more available moisture for tornado outbreaks; on the other end, the decreasing temperature differential between the mid- and upper-latitudes is expected to cause a weakening of the jet stream--reduced wind shear for tornado outbreaks. The best hypothesis I've seen so far is that the numbers of days featuring tornadoes should decrease, but there should be more tornadoes on the days that see them.

Which does make sense. I've noticed that, while there have been less significant tornadic events, the few we've seen have been much more intense. I've been watching storm report trends from years past, and that does seem to make sense. 

Either way, I'm thinking 2015's storm season should provide more information, and we can watch for trends like the ones you mentioned.

[vortex95]

So, here's something somewhat more off-topic, but not by a whole lot. 

We've been seeing the decrease in tornadoes across the States, but my question is, is it part of a pattern that will reverse in time, or is there an underlying cause, such as climate change resulting in a shift or decrease in prime setups in the spring? 

I'm still more enty level to severe weather and atmospheric sciences, so I'm still learning how to figure this out.

Thanks

 

There is no real way to tell either way at this time.  The thing to keep in mind that the decrease in the last 3 years could be due to

climate change, or it may just be part of normal climatic variation.

 

There have been different overall patterns that suppressed tornado activity in the U.S.  2012 was largely due to drought conditions

over much of the central U.S.  2013 had frequent winter storm activity in the E with the mean trough position here, and that

mitigated tornado activity significantly during the prime of the season.  2014 didn't have a lot of significant E Coast storms,

just persistent cold in the E with again the mean trough position here for a lot of year and that shut down the Midwest/Plains for

tornado activity.  Drought means lack of moisture and thus higher LFCs which is not conducive for tornado formation and

with persistent NW flow due to the mean trough position in the E, this prevents good moisture return from the Gulf of Mexico,

so again lack of moisture and higher LFCs preclude tornado formation.  Plus when it is just plain widespread cool/cold wx E of

the Rockies, that shuts things down pretty well also.

 

The point of the above is that it wasn't just one kind of pattern than has lead to near record lows in U.S. annual tornado

count in the last 3 years when accounting for count inflation over time.  So you see it is not as simple as linear more warming

equals X result.  This I think is overlooked and is an important detail.  There are feedbacks both positive and negative in a

changing climate, and it is entirely possible that these feedbacks largely cancel things out in terms of significant changes in

climate over the long term.  That's the key...long term.  A problem is that it is all too easy for human perception to freak

out when a year like 2011 for tornadoes occurs, and just "oh no, if this is the new normal, we are in trouble!".  It is just like

after the 2004-05 hurricane season in the Atlantic.  The same knee jerk psychological reaction occurred.  I agree it was

shocking and impact was huge, but when dealing with climate, single seasons or individual events are *not* proof of anything,

either way.  Look what has happened since 2005...the longest period on record in the CONUS w/o a Cat 3 landfall.  So you

can't let the here in now, wham bam events influence the actual science here.

 

Another item I should point out, the lack of "extremes" is an extreme within itself.  In other words, when storms show

a distinct lack of occurrence, that is an extreme b/c it is a deviation from the average, but just in the other way.

Of course, such does not sell so the mainstream media often ignores this completely, but you have to remember the door

swings both ways here.

 

[vortex95]

Which does make sense. I've noticed that, while there have been less significant tornadic events, the few we've seen have been much more intense. I've been watching storm report trends from years past, and that does seem to make sense. 

Either way, I'm thinking 2015's storm season should provide more information, and we can watch for trends like the ones you mentioned.

You have to remember that there is more "stuff" out there for tornadoes to hit all the time, so

there will be a tendency for increasing big time damage frequency.  By default, we

underestimate many, many tornadoes, b/c so many simply do not hit anything to reveal their

true intensity.  Also, now we have portable Doppler radar in use that is starting to let us rate

tornadoes based on that, even if they do not hit anything.  So with time, there will be more and

more higher rated tornadoes due to this, and it has nothing to do with climate change.

 

You have to be careful with actual storm reports.  The shear volume of reports is higher by

default since detection of documentation of events gets better with time, and is much better

than even only 20 years ago.  Wx awareness and spotters/chasing is at an all-time high,

video devices are everywhere, and social media makes it very easy to verify and document svr

wx reports quickly and efficiently now.  Since hail and straight line wind are far more common

than tornadoes, the number of reports with time is subjected to even greater variability. 

What happens in 2015 will likely not tell us much more than we already know (or don't know). 

Say it is yet another year of suppressed tornado activity.  Does 4 instead of 3 years in a row

of near record low activity mean anything in the big picture for climate?  Not really.  Too

short of a time frame.  If 2015 is very active, the same applies, meaningless really in the big

picture.  The thing we do know for sure, and what should keep individuals or organizations

clanging the alarm bell if it is active, ups and downs in annual activity occur as part of the

overall climate, regardless of any changes in it.  And when you look at things, so far, the

atmosphere hasn't done anything so far off the charts or deviant that can't be expected.  So April

2011 we had the all-time record tornado outbreak in terms of number of tornadoes.  37 years

before we had a similar outbreak.  Not as many tornadoes, but you have to remember,

detection and documentation, esp. for the smaller, brief tornadoes, was not as thorough

then.  We think 2004-2005 were bad in the U.S. for hurricane landfalls, esp. FL, and it was,

but when you look back at the first part of the 20th century, it was far worse in terms of

frequency and intensity.

[audioguy3107]

What's also interesting about the bottom two maps is how hard the northern part of Alabama has been hit with strong tornadoes.  I've always felt the general public only considers the great plains as tornado alley while those in the meterology field do recognize a "mini" tornado alley through central and northern Mississppi and Alabama.  On the bottom fatality map, northern Alabama is really blanketed.  It just goes to show how dangerous tornado outbreaks can and have been in that area.  A lot of folks may not realize it until seeing it represented on the map.