Pielke Jr. – U.S. Tornado damage continues to fall, 2018 activity near record lows

For those that are sure there’s global warming driving tornadoes and other severe weather events, here’s some inconvenient news. Dr. Roger Pielke Jr. has updated his tornado loss data via his Twitter account. He writes:

2017 update to our normalized US tornado losses based on our 2013 paper:

  • Past 5 yrs have 2nd lowest normalized tornado damage of any 5-yr period since 1950 (1997 #1)
  • 2016 had least
  • 2015 2nd least
  • 2017 3rd least
  • 2018 near record-low tornadoes

Normalized US tornado damage to 2017 ($) values:

Average annual damage

  • 1950-1983 (34 yr) = $7.6 billion (median= $5,5B)
  • 1984-2017 (34 yr) = $3.9 billion (median= $2.9B)

There are legitimate debates about tornado incidence, but clearly recent years have seen depressed levels of tornado incidence, accounting for the depressed levels of observed damage.

Here is data from 2012-2017 on tornadoes from NOAA WCM (here: ). There is legitimate debate about this data & inflation-adjustments, but it is safe to conclude that overall 2012-2017 tornado incidence was below long-term average. Lucky us, once again.



0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
July 25, 2018 12:15 pm

It just goes to show how variable the weather is and why it so foolish to read anything into the ups and downs over our tiny timescales.

R. Shearer
Reply to  Trebla
July 25, 2018 12:43 pm

Yes, all kinds of trends can be found in random data. At least there doesn’t appear to be any need for increased concern.

Reply to  R. Shearer
July 25, 2018 2:15 pm

Within a few years tornadoes will become “a very rare and exciting event.”

“Children just aren’t going to know what tornadoes are.”

Reply to  Dave Burton
July 25, 2018 4:24 pm

In other words, the weather will become dangerously mild.

Kristi Silber
Reply to  Dave Burton
July 25, 2018 6:15 pm

Dave, Your link is about snowfall, not tornadoes.

Reply to  Kristi Silber
July 26, 2018 12:28 am
Reply to  Kristi Silber
July 26, 2018 9:59 am

Do have to work extra hard not getting it?

July 25, 2018 12:21 pm

I thought Roger Pielke Jr. had taken so much flack he’d given up commenting on anything to do with the climate … apparently not.

Reply to  commieBob
July 25, 2018 12:31 pm

Here’s what he said in 2015:

So to be clear: I am no longer conducting research or academic writing related to climate, I am not available for talks, and on the climate issue I have no interest in speaking with reporters or giving testimony before Congress. This is not a problem for me or the issue. The field is rich and full of smart people doing interesting work, and the topic will for a long time remain scientifically important and politically significant. I’ve been privileged as a scholar to have been able to participate in the debate, along with great colleagues with whom I’ve collaborated and learned from, and to have our views been heard at the highest levels of both science and policy. As a policy scholar I could not ask for any more. link

I was worried that I was misremembering.

Bryan A
Reply to  commieBob
July 25, 2018 2:37 pm

You had me worried there for a moment. At least he didn’t state that he would be no longer blogging about it or posting comments regarding the work of others.

He just isn’t going to be authoring any more academic papers, talking to reporters or Congress or at meetings.

So Blog Posts are OK as they fall outside the exclusion limits

July 25, 2018 12:25 pm

It also shows the foolishness of using dynamic metrics (items whose values do not remain consistent) to measure things. How different the world would be if we measure distances in $$, or time in inches?
Even measuring structural damage is problematic as structures and building materials do not remain constant nor even consistent.

Thomas Homer
Reply to  rocketscientist
July 25, 2018 12:46 pm

I agree with what you’re saying but your note reminded me of two famous anecdotes:

Distance in $$: “… if you had a stack of thousand-dollar bills in your hand only 4 inches high, you’d be a millionaire. A trillion dollars would be a stack of thousand-dollar bills 67 miles high.” [ Ronald Reagan ]

Time in inches: “… length of wire that is as long as light can travel in one nanosecond. The length is a very portable 11.8 inches.” [ Grace Hopper – Computer scientist & US Navy rear Admiral ]

Reply to  Thomas Homer
July 25, 2018 2:01 pm

Those are proxy metrics 🙂

Reply to  rocketscientist
July 25, 2018 2:38 pm

I used the nano sec one frequently in my research, if I wanted to have two laser pulses from a laser arrive at the same place at the same time then I had to make sure that they travelled exactly the same distance. With nano sec pulses I worked on the basis that a foot difference in distance travelled was a nano sec difference in arrival time.

Reply to  Phil.
July 25, 2018 7:19 pm

The speed of light makes for a very useful constant (at least for many of our frames of reference) from which many other reliable proxies can be created.
And, the specific wavelengths for light can be used as well. I believe the meter is defined by light travel distance.
We can even use it to measure time.
And, in the instance you mention we use it to determine inertial reference used for attitude (orientation) and position measuring by creating fiber-optic gyroscopes.

Reply to  rocketscientist
July 26, 2018 10:00 am

I thought a meter was defined as a set number of wavelengths of light from a particular (argon?) laser.

Greg Strebel
Reply to  MarkW
July 27, 2018 8:12 am

Originally AFAIK it was 1 ten millionth of the distance from the equator to the pole(s). Ten thousand km distance = 111 km per degree of latitude (90 degrees per polar quadrant) = 60 nautical miles per degree => 1 nautical mile per minute. Subsequently defined as a prescribed number of wavelengths of light produced by a defined electron orbital transition, which deemed more immutable than the geography-based definition.

Peter Morris
Reply to  rocketscientist
July 25, 2018 3:25 pm

My favorite proxy metric is the bowl of cereal. I was watching the news years ago and they were talking about a new stadium, and gave its volume in bowls of breakfast cereal. I laughed for a long time at that one.

R. Shearer
Reply to  Thomas Homer
July 25, 2018 2:53 pm

They no longer make thousand-dollar bills, at least in the U.S. after 1969.

Tom Schaefer
Reply to  rocketscientist
July 25, 2018 12:53 pm

Indeed. Le’s say he is normalizing for inflation into some constant dollars. Two problems:
1) Real estate has been running above inflation for a few decades (with a notable adjustment in 2008-2009).
2) The amount and DENSITY of real estate in the areas affected by tornadoes has been rising. Imagine what the damage totals would be if the high population cities and sprawl of the mid Atlantic was tornado alley.

I would like to see the numbers recast as a percent of total real estate value to largely remove these two factors….

“The report says the total value of U.S. housing stock grew to a total of $29.6 trillion in 2016, marking an all-time high. Zillow’s report shows that housing stock showed an increase of $1.6 trillion from 2015, a 5.7% increase in value to reach that record level.”

So if 5 billion in tornado damage occurs in 2019, it will be on a base of >30 trillion in housing (more if one were to include other structures)- or .015% or total real estate. Let’s focus on termites and leaking basements.

Bryan A
Reply to  Tom Schaefer
July 25, 2018 2:45 pm

What is needed is a mandate that Insurance Companies MUST allow homes to be rebuilt to be more robust and withstand the next Hurricane or Tornado that way they won’t have to pay to rebuild year after year, decade after decade

Reply to  Bryan A
July 25, 2018 10:06 pm

Then you must do the same wrt homes destroyed by floods, wildfires, earthquakes, and mudslides. Then you need to mandate that insurance companies must provide coverage for the cardboard boxes we’ll be living in, since no one will be able to afford the cost of a house.

Reply to  Bryan A
July 26, 2018 10:02 am

Such a mandate will make insurance more expensive. Anyone who wants could buy such a policy now.

AGW is not Science
Reply to  MarkW
July 26, 2018 1:54 pm

Actually no, you couldn’t. An insurance company will only insure your property’s “Insurable interest,” which would not cover “upgrades.” You might be able to secure limited coverage for mandated upgrades to meet new building code requirements, but without that coming into the mix the best you’ll get is “replacement cost” coverage for “like kind and quality.”

As to such mandates making insurance more expensive, yes and no. The long term driver of insurance prices is the cost of related losses; so replacing housing NOT built to survive local weather events with housing that IS built to survive such events, as opposed to rebuilding over and over the same stupid thing they built in the first place, will ultimately REDUCE insurance costs.

Reply to  rocketscientist
July 25, 2018 9:58 pm

We have very heavy traffic in this area. Distance by car is often given in units of time, not miles. A house two miles away in the same subdivision could be six minutes away. Going to a shop two miles down a major thoroughfare could be twenty minutes. So when someone asks how far away something is, it’s usually better to tell them in minutes – and you have to adjust that by whether it’s rush hour or not.

Go Home
July 25, 2018 12:34 pm

What bothers me is that I go to the tornado page on WUWT several times a week, to check on this years progress, only to find that NOAA has not updated it regularly and most recently for nearly a whole month. Not sure why. The conspiracy nut in me thinks they do not like posting it when the numbers for year to date are at record lows and are just waiting for the next outbreak to repost. THat way they would not make any headlines that would counter the usual narrative. But they would never do that would they.

July 25, 2018 12:48 pm

1950-1983 (34 yr) = $7.6 billion (median= $5,5B)
1984-2017 (34 yr) = $3.9 billion (median= $2.9B)

…from 1950-1983 there was a lot less infrastructure to damage….and cost wise, it was twice as much

US population 1983……234 million 1950……150 million
Us population 2017……325 million

…those 100 million more people had to have suburbs, shopping, etc

from 1950 to 2017 US population doubled…..so did the infrastructure to support them

Adam Gallon
Reply to  Latitude
July 25, 2018 1:59 pm

Which is why such data is levelized.

Bryan A
Reply to  Adam Gallon
July 25, 2018 2:41 pm

and why Tornados tend to level things as well

Reply to  Adam Gallon
July 25, 2018 2:56 pm

“……and provides an indication that maximum damage levels have the potential to increase should societal change lead to increasing exposure of wealth and property.”

nothing to do with how many or the strength of tornadoes

July 25, 2018 12:50 pm

In addition to the normal variability of weather events giving rise to tornados and consequent damage … there is another significant reason for the falling damage numbers …

Tom Schaefer
Reply to  Kenji
July 26, 2018 6:27 am

Excellent point! I love those things when I’m building stuff. I had a 4 m2 sail area tree house made from 9 – 1/2 inch 8x4ft plywood, 4 – 4X4in corner posts, and 8 2X6in edge boards survive a cat 2 hurricane.

Joel O'Bryan
July 25, 2018 1:10 pm

I think it was Roger Pielke, Jr himself that said measuring climate change impact in dollars makes us dumber. I think the context was in regards to claims about SLR and endangering infrastructure/property in S. Florida.


Joel O'Bryan
Reply to  Joel O'Bryan
July 25, 2018 3:28 pm

Here Roger, Jr criticizes Seth Borenstein for using economic damage from hurricanes as an indicator of climate change;


Kristi Silber
Reply to  Joel O'Bryan
July 25, 2018 6:23 pm

And it was a good point he made, especially in the case of tornadoes, which have such localized effects! One tornado in the right place could effect the financial damage of a score of them in sparsely populated areas.

July 25, 2018 1:22 pm

Interesting are the “freak” years: 1953, 1965, 2011 and, not so much but way over normal, 1974.

The “freak” years show just how much the record can be skewed by an outlier year. Picture a climate-length 30-year trend 1981-2011 or 1987-2017. This exercise demonstrates why strict so-called 30-year averages or 30-year trends are not valid for policy making when confronted by data that contains such outliers.

2016 looks like an outlier on the low side.

Always better to look at all the reliable data that one can find.

john Minich
Reply to  Kip Hansen
July 25, 2018 4:56 pm

I think this shows another reason to understand the differences among mean, median, and mode.

Kristi Silber
Reply to  Kip Hansen
July 25, 2018 6:34 pm

It depends on the data being considered. Tornadoes and hurricanes are discrete, relatively rare events, while temperature is always there, and is much more amenable to applying 30-year means. That doesn’t contradict using all the available reliable data.

Tornado damage is a meaningless measure of climate change, since location is all-important.

R. Shearer
Reply to  Kristi Silber
July 25, 2018 7:02 pm

You don’t know what you are talking about. There are over 1000 tornadoes annually just in the U.S. People actually are impacted more today, because there are many more people.

Climate change has nothing to do with it because climate has always changed and always will along with all the natural weather events from which it is comprised.

Kristi Silber
Reply to  R. Shearer
July 30, 2018 11:16 am

I should have stated it differently: tornadoes are relatively rare (or uncommon, anyway) if considering the global average, not just the U.S. According to NOAA, Canada has the second-most number of tornadoes/year, around 100. The count of more than 1000/year included EF-0 tornadoes, the very weak ones. If considering only EF-1+, the *record* is about 900 annually. In addition, the “disparity between tornado records of the past and current records contributes a great deal of uncertainty regarding questions about the long-term behavior or patterns of tornado occurrence.”

Then there was the paper suggesting the number of “‘extreme outbreaks’—clusters of a dozen or more twisters rated strong enough to have caused at least moderate damage to structures” has been rising since 1965, but it’s paywalled, so I don’t know how well-documented the past outbreaks have been. It was the subject of an WUWT article in which Anthony asserts that it is the result of reporting bias. (While that may be true, it is still conjecture.)

“Climate change has nothing to do with it because climate has always changed and always will along with all the natural weather events from which it is comprised.”

Yes, climate has always changed (sometimes slowly, sometimes quickly). Does climate have nothing to do with the incidence of tornadoes?

Kristi Silber
Reply to  Kip Hansen
July 25, 2018 7:36 pm

“Always better to look at all the reliable data that one can find”

Pielke presents the tornado count from 2012-2017. What he doesn’t show is that the data from 2008 and 2011 were extremely high, which skews the mean upwards.
comment image
…Oh, but he says the yearly graphs are reporting the median. Why does that make sense? And where did he get that info?

What’s the “inflation adjusted” part mean? Oh, I guess that refers to the local reports that over-count the incidence?

And did you notice that because the data corresponding to the graph for that year is adjusted by 0.85, some of the yearly totals are below the minimum for the whole time period?

It’s all kind of odd to me. Not something that means a lot to me without explanation.

At any rate, “it is safe to conclude that overall 2012-2017 tornado incidence was below long-term average” is not, to my mind, a safe conclusion at all, since the “long-term average” is based on 10 years.

Reply to  Kristi Silber
July 25, 2018 9:48 pm

Silber displays the logical fallacy ‘Argumentum ad Ignorantiam’ extremely well.

e.g. “What’s the “inflation adjusted” part mean? Oh, I guess that refers to the local reports that over-count the incidence?”

Where Silber admits ignorance then fills in it’s inflation conundrum with it’s fictitious speculations.

Silber starts off their zero knowledge display with “ Pielke presents the tornado count from 2012-2017. What he doesn’t show is that the data from 2008 and 2011 were extremely high, which skews the mean upwards.“; where Silber assumes that a few “high years” skews the presentation.

If Silber had bothered to check weather history, it would have discovered that tornadoes are in a long term decline. In spite of massive population increases and commensurate structure build-outs along with incredible technology leaps. Allowing comprehensive tracking, and identifying tornadic storms, versus history’s reliance upon people actually seeing a funnel and reporting that funnel.

comment image

comment image

Kristi Silber
Reply to  ATheoK
July 30, 2018 12:11 pm

That’s what I get for asking questions! Attacked for making an argument. Where’s the logic in that?

My comment about Pielke, Jr.’s presentation was based on the information available to him at the site where he got the info he presented. He chose to present only part of the information available, that which best supported his argument. That is usually called cherry-picking, but it was a weak argument in the first place, comparing 5 years of data with the 10 year average it’s part of.

I also said that I didn’t understand what was being presented. Evidently ATheO thinks that’s an admission of not understanding anything related to the subject.

I don’t see a long-term decline in that first graph he posts (I’ve seen both already, and knew about the decline anyway – no need to check history just to know that! ATheO wishes to think me an idiot, though, as it confirms his erroneous ideas of others. Sorry to disappoint you). Both graphs support my idea that the incidence of tornadoes is variable enough that it’s hard to see a strong signal through the noise, whether positive or negative.

“…tornadoes are in a long term decline. In spite of massive population increases and commensurate structure build-outs along with incredible technology leaps.” Gee, I didn’t know that the annual tornado numbers bore a relation to population or structures. I guess I could learn something from ATheO after all. Mind if I quote you?

“Allowing comprehensive tracking, and identifying tornadic storms, versus history’s reliance upon people actually seeing a funnel and reporting that funnel.” This is an important point (even if grammatically atrocious). Before we were able to track tornadoes using technology, couldn’t some tornadoes have been counted multiple times by separate people? Might one expect this to be the case especially for the strongest tornadoes? Are they generally visible from a greater distance, and do they last longer/move further/touch down in more places? (Serious questions – but don’t worry, ATheO I don’t expect to learn from you.)

Clyde Spencer
July 25, 2018 1:23 pm

A half-degree increase in average temperatures seems like a small price to pay for lowered risk of tornado damage. Wouldn’t you say, Toto?

Jacob Frank
July 25, 2018 2:05 pm

The tornadoes are obviously hiding in the bottom of the ocean deniers

Bryan A
Reply to  Jacob Frank
July 25, 2018 2:42 pm

Or running up and down De-nile

July 25, 2018 2:08 pm

NOAA June 1, 2018:
The U.S. is having an incredibly quiet tornado season – so far
Low number of tornadoes also brings fewest fatalities on record

Kristi Silber
July 25, 2018 6:09 pm

Is there any evidence that the mainstream climate scientist community is confident overall tornado incidence would increase due to global warming? Or was that just a silly alarmist idea?

I suggest it’s more likely the latter. Even the Union of Concerned Scientists says, “Tornadoes are relatively small, short-lived phenomena and scientists don’t have robust enough data to determine whether and how climate change may be affecting tornado frequency, intensity, or the geographic range where tornadoes are most likely to form.”

Then there’s this rather interesting article, suggesting that, “researchers looked at increasing trends in the severity of tornado outbreaks, measuring severity by the number of tornadoes per outbreak. They found that these trends are increasing fastest for the most extreme outbreaks. While they saw changes in meteorological quantities that are consistent with these upward trends, the meteorological trends were not the ones expected under climate change.

…”‘The fact that we don’t see the presently understood meteorological signature of global warming in changing outbreak statistics leaves two possibilities: Either the recent increases are not due to a warming climate, or a warming climate has implications for tornado activity that we don’t understand. This is an unexpected finding.'”

(I find it very odd indeed when people suggest that data which doesn’t conform to climate change predictions is suppressed. I find very little evidence to support this, judging by the numbers of papers I come across that raise questions about or critique climate models. What doesn’t get published in peer reviewed journals is unprofessional attacks on individuals, and bad science. I’d appreciate it if people here will point out specific examples where my perceptions are wrong.)

Reply to  Kristi Silber
July 25, 2018 6:46 pm

Kristi Silber

(I find it very odd indeed when people suggest that data which doesn’t conform to climate change predictions is suppressed. I find very little evidence to support this, judging by the numbers of papers I come across that raise questions about or critique climate models. What doesn’t get published in peer reviewed journals is unprofessional attacks on individuals, and bad science. I’d appreciate it if people here will point out specific examples where my perceptions are wrong.)

Two weeks after we pointed out here at WUWT that the 2016 and 2017 NSIDC Regional Sea Ice Extents set all-time, record-breaking summertime highs for the Bering Sea, Hudson Bay, Sea of Okhotsk, and Gulf of St Lawrence, the NSIDC deleted those datapoints and reset ONLY the 2016 and 207 sea ice extents to 0.0 for those regions for those dates.

Kristi Silber
Reply to  RACookPE1978
July 30, 2018 5:32 pm

It appears that things have changed since then, if I’m looking at the right data. Whole columns are zeros for some months in some areas; Bering is all zeros for August only. Maybe there were errors. It would be strange indeed if those areas had record highs when the other areas in the Arctic are nowhere near record. Did you notice that?

I can see why you’d be suspicious, but I’d say the evidence as it stands is a little weak. To me it seems that if it were true, and there really was a record but they are now “cooking the books” for all regions so that the rankings are much lower (in the tens and teens out of 40), other agencies would notice and see that something is off. Too blatant. Two years of records in a row would be obvious outliers, given the trends.

Tom Abbott
July 25, 2018 9:26 pm

Well, there have definitely been fewer tornadoes in this part of Tornado Alley this year.

I would also go out on a limb and say July has been below normal in temperature this year. I did have one 109F day at home a few days ago, but the next day was back down to about 90F. Normally, we get several weeks of above 100F weather around this time of year, but not this year. We can’t buy a high-pressure system. Usually, one will sit on top of us around this time of year. Sometimes it will sit there for a week or two and sometimes longer. But not this year, and we are running out of summer. August 15 is about the hottest part of our year and then it starts cooling off.

If this is CAGW then we want lots more of it.

Clyde Spencer
Reply to  Tom Abbott
July 26, 2018 9:16 am

Tom Abbott,

“…but the next day was back down to about 90F.” In other words, it got colder in unison with the oceans becoming more acidic.

Tom Abbott
Reply to  Clyde Spencer
July 26, 2018 2:27 pm

Something like that. 🙂

I think what happened was a high-pressure system was over Texas and Oklahoma and things got real hot and then a storm front came through Oklahoma and cooled things off. We’re down below 90F today and for the rest of the month, it looks like, which is very unusual for this time of year. It’s also unusual for us to be getting rain this time of year.

I like CAGW! 🙂

Doug S
July 25, 2018 11:23 pm

One of the very first things I noticed about the climate models was the omission of the Kinetic Energy term on the planet in question. I posted this at realclimate about 8 years ago and they did respond in a kind way. They explained that as a Bachelor in Physics, I wasn’t used to seeing the Climate equations expressed in terms restricted to radiation transfer only. It’s still my opinion that the Kinetic energy present on the planetary body must be considered when calculating the climate conditions. And, I’m out of my depth of expertise but so willing to be shown my errors.

Bloke down the pub
July 26, 2018 2:29 am

Tornadoes are largely driven by the temperature difference between the tropics and the poles. Cagw theory predicts a polar amplification such that the poles would warm faster than the tropics, thereby reducing the temperature difference between the two. If cagw alarmists were smart enough, they’d realise that a reduction in tornado damage is actually proof of their theory but then , their only aim is to scare people into taking action, no matter how stupid it is.

D. Cohen
Reply to  Bloke down the pub
July 26, 2018 3:47 am

Absolutely, and this same observation could explain the drop in hurricanes hitting the east coast of the US.

By the way, if the poles are warming much more than the equator, wouldn’t that increase the area available for humans to live comfortably, not to mention more food from increased growing seasons, etc.? Sounds like global warming produces much more good than bad. (And if the polar ice caps do melt significantly, well it’s a safe bet that would at the very least expose more mineral resources to exploit, especially in Antarctica.) The assumption that global warming is automatically bad does not stand up to any reasonable sort of analysis.

Ernest Bush
Reply to  D. Cohen
July 26, 2018 11:05 am

Or the drop in hurricanes on the East Coast could be explained by the North Atlantic cooling over the last few years in those areas where hurricanes develop. Also, there is a huge amount of African dust in the area off the coast of Africa where hurricanes would develop. Meteorologist Joe Bastardi comments about this frequently in his free daily forecast.

Today he commented on the huge drop in sea surface temperatures across the North Atlantic in general from last year. He also noted that the MSM is focused on the heat in the West while ignoring the fact that the temperatures are falling across most of the U.S. with the Midwest about to get a lot cooler. He has pronounced that summer will end over a majority of the U.S. land mass over the next few days.

July 26, 2018 9:10 am

Might not be such a great idea to turn back the clock…

July 26, 2018 11:12 am

I see the left side of Jesus’ face in those graphs, which surely is a sign of the eminent second coming.

July 26, 2018 4:16 pm

Still not sure why tornados are quantified in terms of financial effect, even normalized dollars. It requires that in given season on average they hit the same distribution of high value and low value targets. Wouldn’t it just be better to monitor frequency and power?

%d bloggers like this: