At AGU 2013: Are tornadoes getting stronger? No.

This AGU session asked that question directly and I covered this previously. Now that the video of the entire session is available, I thought it worth revisiting. Of course, the answer is “no”, even when they try to make a hockey stick out of it. Watch the video, especially at just before the 20 minute mark, where in Q&A the presenter (James B. Elsner) remarks that he’s “not even sure if the slope is completely real“.

Contradicting that. Elsner also said “If I were a betting man I’d say tornadoes are getting stronger

Commentary follows.

Comments:

I was going to write up something on tornado climatology myself, but in doing research on the topic, I decided to let NOAA’s Nataional Climatic Data Center speak to the issue, since they do a good job of refuting this hockey stick nonsense even before it was conceived.

From: http://www.ncdc.noaa.gov/climate-information/extreme-events/us-tornado-climatology/trends

NCDC writes:

Today, nearly all of the United States is reasonably well populated, or at least covered by NOAA’s Doppler weather radars. Even if a tornado is not actually observed, modern damage assessments by National Weather Service personnel can discern if a tornado caused the damage, and if so, how strong the tornado may have been. This 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. Improved tornado observation practices have led to an increase in the number of reported weaker tornadoes, and in recent years EF-0 tornadoes have become more prevelant in the total number of reported tornadoes. In addition, even today many smaller tornadoes still may go undocumented in places with low populations or inconsistent communication facilities.

With increased National Doppler radar coverage, increasing population, and greater attention to tornado reporting, there has been an increase in the number of tornado reports over the past several decades. This can create a misleading appearance of an increasing trend in tornado frequency. To better understand the variability and trend in tornado frequency in the United States, the total number of EF-1 and stronger, as well as strong to violent tornadoes (EF-3 to EF-5 category on the Enhanced Fujita scale) can be analyzed. These tornadoes would have likely been reported even during the decades before Doppler radar use became widespread and practices resulted in increasing tornado reports. The bar charts below indicate there has been little trend in the frequency of the stronger tornadoes over the past 55 years.

EF1-EF5 Tornado Counts

EF3-EF5 Tornado Counts

=======================================================

Elsner has an 18 year data set, which is derived from  measuring the length and width of a tornado’s damage path to give an indication of its strength, which is driven by the storm’s peak wind speed. It is difficult if not impossible to measure that speed directly, even TOTO (TOtable Tornado Observatory for which I built the corona electrometer instrument) has had only occasional luck in doing so. Putting TOTO directly in the path of an oncoming tornado required an astronomical amount of luck, and so the project was abandoned.

So, like Mann’s hockey stick, Elsner’s data is a proxy, not the actual measurement, and the hockey stick slope we see (that Elsner says isn’t even sure is real) at the end of his graph is created by data from one year: 2011.

tornadoes_kinetic_energy_trend

It’s the old familiar sensitivity to endpoints graphing problem.

But note in the second graph from NCDC, the year 1974, when the April super outbreak occurred. There were many more F3+ tornadoes in 1974 than in 2011.

If Elsner had the same path data for an 18 year period from 1957-1975 and plotted it, he’d get an even bigger hockey stick shape at the end. Statistical artifacts from short period data shouldn’t be used to draw conclusions.

UPDATE: Paul Homewood has a detailed analysis here: http://notalotofpeopleknowthat.wordpress.com/2013/12/13/are-tornadoes-getting-stronger-2/

About these ads

28 thoughts on “At AGU 2013: Are tornadoes getting stronger? No.

  1. It’s the old familiar sensitivity to endpoints graphing problem.
    ============
    if the trend changes based on the choice of endpoints, it is not a trend.

  2. I’m sure the fact that Elsner says he isn’t even sure is real will be lost when Dana get’s his hands on this graph.

  3. a true trend is a property of the data, not of the statistical method. so, if the trend changes when you change the endpoints, you are not seeing a true trend. a true trend is unchanged by a change in the statistical method, because the underlying data has not changed, so the true trend cannot change.

  4. A man in need of an independent statistical expert review. He is mesmerized by wriggles and sees catastrophic omens in them without regard to the vagaries of statistical analysis limitations.

  5. http://notalotofpeopleknowthat.wordpress.com/2013/12/13/are-tornadoes-getting-stronger-2/

    b) Until 1994, it was the mean widths of tornadoes that were measured. After 1994, it has been maximum width. This means that the period prior to 1994, which Elsner shows on his graph is not just dubious, it is totally incompatible with the post 1994 period.

    Was he aware of this change of practice? If he was, why has he produced a positively misleading graph.
    =================
    Not only misleading, but a nonsense. How can you say one tornado is stronger than another by looking only at the maximum width. Nothing says the maximum is representative of the total area.

    Two tornadoes can touch down for 1 mile. One cuts a swath 1 mile wide over the whole length. The other cuts a swath 1/10th of a mile wide for 9/10th of a mile, but in the last 1/10th it cuts a swath 2 miles wide.

    By Elsner’s reasoning, after 1994 the second tornado would be twice as strong as the first. But before 1994 the first would have been 3 times stronger than the second.

  6. The warmists seem to be increasingly clutching at straws in desperation as their AGW theory is starting to unravel. I thought that everyone is in agreement that there has been no warming for 18 years. How then can AGW be causing any change either in the frequency, intensity or in fact anything else associated with tornadoes over the last 18 years?
    Here is a link to a photograph (which some sources say has been faked) demonstrating the impact of static global temperatures!

    http://www.irishmirror.ie/news/cairo-snow-picture-snow-covered-pyramids-2929265

  7. TOTO (TOtable Tornado Observatory
    good name.

    Where are all the models showing that tornado forming conditions should be more common with global warming? Cold dry polar air meets warm moist tropical air — too warm, fewer suitable cold fronts.

  8. So that’s where the missing heat has been going … into making fewer yet more intense tornadoes.

    (/sarc)

  9. Having seen the results of several large tornadoes, I have to wonder how accurate you can estimate the strength of a tornado based on the “width of damage”:
    1 – Width will depend on the materials damaged – their weight, consistency, etc
    2 – Building practices will impact damage analysis
    3 – A person’s training and/or bias will impact estimated width

    Obviously, you can make an estimate that is within a certain error range, but the error range may be larger than the ability of the model to produce a meaningful result. Too many uncontrolled variables, and too much bias.

  10. As Paul Holmwood points out, it’s a serious omission to not have included 1974 and before when there was much more tornado activity. Also, someone remarked in questions that, if we are not drawn by the “smoother” , the data does not show any increase before 2007,2008.

    I have previously shown that the pre-1974 global cooling period was when tornado activity was far higher and suggested that it is not temperature itself but the change of temperature that we should be looking at. The post ’74 global warming period was marked by less tornado activity.

    At least on the decadal scale, warming ended somewhere around 1997-2005 depending upon what metric we choose. Certainly there has been a beginning of actual cooling since 2005.

    Prof. Elsner’s scale would seem to be another confirmation that we are moving into the next period of cooling and with that we are moving into a time of increased tornado activity. This is not a result of “global warming” (being the tag attached to _warmer_ surface temps, rather than _warming_ ) but a result of global cooling.

    It is most notable the his index is low and flat precisely across the 1975-2000 period when there was a sustained “global warming”. If his new metric means anything, it is clear proof that there is LOW tornado activity during (caused by ?) global warming.

  11. The last figure shows “Integrated kinetic energy trends”. What does that mean? Yearly accumulated? A rolling average of a few years? And the gray shaded area means what? A standard deviation (SD)?
    It is extremely strange the last data points wouldn’t explode that putative SD. That would mean there have to be very few points in the recent data added to a much larger number of other points. If the mean changes, the new yearly numbers have to be a large fraction of the total. My conclusion is the gray shaded area is not informative of anything other than deception. Although they never define what the shaded area is, they let us assume.
    Caveat emptor.

  12. “If I were a betting man…”
    “Not sure about the slope of the line..”
    This isn’t science. Why did he even waste our time with this?

  13. “Pamela Gray says:
    December 15, 2013 at 9:21 am

    A man in need of an independent statistical expert review. He is mesmerized by wriggles and sees catastrophic omens in them without regard to the vagaries of statistical analysis limitations.”

    —————————————–

    Something akin to reading tea leaves, or tossing sticks in the air and then interpreting the pattern they took when they landed. Since I spent 3 hours this morning blowing 35cm of fresh global warming, maybe I’ll throw some hockey sticks in the air and write a paper on the increased frequency of CO2 emissions correlated with the anthropogenic removal of solid state hydrogen oxides, using my distributed hockey stick statistical proxy. Should be a Nobel prize winner.

  14. Reading through “Hiding the Decline” it is easy to see that AGW opinion makers are not making an honest effort. They are pushing an agenda for power unrelated to their being accurate. They are not after an honest debate. That Elsner is playing both sides of this appears more as a survival tactic. Reading what Montford has documented on how Mann & gang treat those who dare to disagree gives insight on how a Lysenko-like cult can hijack a science.

  15. The good records go back to 1950. Why start with 1954?

    However, I don’t even see going back to 1950 showing a significant increase of tornadoes of any category stronger than F1/EF1, and only minimal increase of F1/EF1 tornadoes.

  16. I have heard Dr. Roy Spencer explain how global warming is likely to be weakening northern hemisphere major tornadoes. The most critical ingredient for those is a tornado-favoring wind sheer that is powered mainly by horizontal temperature gradient. Since the Arctic south of 80 degrees north and the near-Arctic have warmed faster than the tropics, horizontal temperature gradients that favor major tornadoes and other major northern hemisphere midlatitude windstorms of extratropical types has weakened.

  17. andrewmharding said in part on December 15, 2013 at 10:04 am:

    “I thought that everyone is in agreement that there has been no warming for 18 years. How then can AGW be causing any change either in the frequency, intensity or in fact anything else associated with tornadoes over the last 18 years?”

    18 years now? Not the 17, 16 or 15 that I have heard a lot lately? 18 years ago was in late 1995. Even the most-hiatus-showing of the major global termperature indices, RSS TLT, shows a rising trend from 1995.83 to 2013.84. http://woodfortrees.org/plot/rss/from:1995.83/to:2013.84/plot/rss/from:1995.83/to:2013.84/trend

    In this same thread, ferdberple said in part on December 15, 2013 at 9:18 am:

    “a true trend is a property of the data, not of the statistical method. so, if the trend changes when you change the endpoints, you are not seeing a true trend.” At risk of sounding like a broken record, I don’t think a true trend is shown by a 15, 16, or 17 year trend with a century-class El Nino near the starting endpoint. Instead, I see the hiatus as starting in 2001, a little over 12 years ago. However, I do expect it to continue for about 20 more years.

  18. @Donald; Wrong. The starting point for the halt in warming is now, and moving back to wherever the trend become positive. Sorry, you don’t get to cherry-pick 2001 as the start date.

  19. To Pamela Gray:

    You wrote: “He is mesmerized by wriggles…”

    This is a common failing of the male half of the race, though it’s
    usually wriggles of a profoundly different sort.

  20. Donald L. Klipstein says:
    December 15, 2013 at 8:16 pm
    18 years now?…..I don’t think a true trend is shown by a 15, 16, or 17 year trend with a century-class El Nino near the starting endpoint.

    It is actually 17 years and 3 months for RSS since September 1996. And if you want to avoid the El Nino, it is still an even 14 years since December 1999. As for statistically significant warming at the 95% level, that goes back 21 years for RSS to December 1992.

  21. Yeah, delete the El Nino and back out some adjustments and the Pause goes back 33 years. Can’t have it both ways.

Comments are closed.