From The Open Atmospheric Society http://theoas.org
FOR IMMEDIATE RELEASE – Sept 19th, 2014
Statement from the Open Atmospheric Society on the paper ‘Peak tornado activity is occurring earlier in the heart of “Tornado Alley” ‘ (DOI: 10.1002/2014GL061385)
The recent paper by Long and Stoy published September 10th, 2014 in Geophysical Research Letters and sent out as a press release⁴ by the American Geophysical Union makes claims that are not fully supported by the data.
The source of the data is from the NOAA Storm prediction Center (SPC) database of tornado reports. The OAS has no issue with the data, and believes it is correct and representative of actual tornado reports.
However, interpretation of the data by Long and Stoy may lead to a conclusion rooted in reporting bias.
Since the introduction of the nationwide WSR-88D Doppler radar network from 1992 to 1997, the ability to spot smaller scale tornadoes has increased dramatically, with warnings being issued based solely on radar signatures. Simmons and Sutter, 2004 stated:
The accuracy of tornado warnings has increased over the decades, and with the advent of Doppler weather radar, warnings can now generally be issued before a tornado actually touches down.
Further, in 2008, the Super Resolution upgrade² to the WSR-88D provided even better detection of the weakest tornadoes:
Super Resolution data should lead to increased tornado warning lead times. Simulations using Super Resolution data show that mesocyclone and tornado signatures can be detected at greater ranges than with legacy resolution data. In addition, other smaller scale features should be detectable in base products sooner or with greater reliability.
These improvements in technology, combined with greater awareness, greater saturation of news gathering, increased interest in both private and commercial storm spotting, plus an increased ability to rapidly report tornadoes has increased the count of weaker F0 and F1 Fujita Scale tornadoes in the last two decades.
In the paper Tornado Trends Over The Past Thirty Years³ they state:
The increase in reported tornado frequency during the early 1990s corresponds to the operational implementation of Doppler weather radars. Other non-meteorological factors that must be considered when looking at the increase in reported tornado frequency over the past 33 years are the advent of cellular telephones; the development of spotter networks by NWS offices, local emergency management officials, and local media; and population shifts. Changnon (1982) and Schaefer and Brooks (2000) both discuss these influences on tornado reporting. The growing “hobby” of tornado chasing has also contributed to the increasing number of reported tornadoes. The capability to easily photograph tornadoes with digital photography, camcorders, and even cell phone cameras not only provides documentation of many weak tornadoes, but also, on occasion, shows the presence of multiple tornadoes immediately adjacent to each other.
These improved detection and reporting reporting methods most certainly contribute to recent increases in tornado counts, and by extension earlier and more frequent tornado counts than would have been recorded 30 years ago.
The annual tornado trends chart is a result of the following methodology applied to the SPC observed tornado dataset by Harold Brooks, NSSL and Greg Carbin, SPC5.
“An inflation adjustment was developed to reflect the improved detection. The SPC simple linear regression equation is fit to the 1954-2007 annual tornado totals. This equation is then used to compute the delta, or difference, between the original/observed annual tornado total and the smoothed, or “adjusted” annual total represented by the point on the linear trend line for that year.
![linearplot[1]](https://theoas.files.wordpress.com/2014/09/linearplot1.png)
Figure 1: Tornado count per year with linear observed trend not adjusted for inflation (better detection). (Brooks and Carbin SPC 2008)
“Using 2007 as the “baseline” year, we apply each year’s delta value from 1954 to 2007 to the linear trend value of 1283.3 for 2007. Each year is thereby adjusted, or standardized, to the 2007 annual tornado numbers. (Note that applying the delta of -185.3 to the 2007 adjusted value of 1283.3 results in the original/observed “baseline” total for 2007 of 1098 tornadoes.) When these annual adjusted values are plotted, we see that the linear upward trend is removed from the data (see figure 2 below). Removal of this upward trend is desirable because the increase in tornado reports over the last 54 years is almost entirely due to secular trends such as population increase, increased tornado awareness, and more robust and advanced reporting networks. By removing the upward trend and making the broad assumption that 2007 represents something closer to reality for annual tornado numbers, we can attempt to answer the question, “what constitutes a normal year with respect to modern-day tornado reports?”
The answer becomes the adjusted average across the 54-year period, or 1283.3 tornadoes per year. This value is also the max trend line value at 2007 that was combined with the individual delta value for each year to adjust all annual totals in the data set.”
![adjplot[1]](https://theoas.files.wordpress.com/2014/09/adjplot1.png)
Figure 2: Tornadoes adjusted to 2007 baseline (detrended) to account for inflation (better detection) (Brooks and Carbin 2008)
In a note on September 17 to the OAS, Mr. Carbin added:
“I have not de-trended any more total annual data since the 2007 analysis. Obviously, we have experienced a few very quiet years since the remarkable year of 2011. The key point in the statement above is that we are expecting 2007 to represent something close to an average annual total. If we use the (E)F1 and stronger annual counts for the past 61 years, the annual mean is below 1000 tornadoes and there is a distinct upward trend in the annual counts). However, is we limit the annual EF1 tornado counts to the past 25 years (1989-2013) then we get an annual mean closer to 1200, and essentially no trend (see figure 3 below). Whatever period is used, one significant characteristic of the annual counts since 2000 is the profound variability about the mean. Some of the greatest annual standard deviations (both positive and negative) have occurred within a little more than the past decade. This is perhaps the most puzzling aspect of the record.”
![clip_image0083[1]](https://theoas.files.wordpress.com/2014/09/clip_image00831.jpg)
Figure 3: EF1+ Annual Count not detrended for inflation due to improved detection (Carbin 2014)
The NOAA SPC had shown in 2013 that there was no trend in strong tornados – EF3+ (see figure 4).
![Fig31_Tornadoes-600x361[1]](https://theoas.files.wordpress.com/2014/09/fig31_tornadoes-600x3611.jpg)
Figure 4; US Annual Count of Strong to Violent Tornadoes (F3+), 1954-2012 (Carbin, SPC)
We also note, in the AGU press release for the Long and Stoy paper, it was mentioned:
“If we take Nebraska out [of the data], it is nearly a two-week shift earlier,” noted John Long, a research scientist in the Department of Land Resources and Environmental Sciences at Montana State University in Bozeman, Montana, and lead author of the new paper. For tornadoes rated above F0, the lowest rung on the original Fujita scale of tornado strength, the shift is also close to 14 days, according to a preliminary analysis by Long and his colleagues that’s not included in the new paper.
It is the opinion of The OAS that this sort of methodology to remove a portion of a dataset to cite a result is unsupportable and without justification. Climatic scale detection of earlier onset of tornado activity cannot be dependent upon removal of a portion of the dataset. Other analyses of the same data set by the primary investigator for tornado climatology show that there are no trends in frequency or intensity of tornadoes, and by extension, suggest that the claims made by Long and Stoy are little more than artifacts of statistical methodology and an increase in the ability to spot, report, and categorize tornadoes that would have gone unnoticed and unreported thirty years ago.
###
For more information on The Open Atmospheric Society, or to become a member, see http://theoas.org
References:
1. Tornado Warnings: How Doppler Radar, False Alarms, and Tornado Watches Affect Casualties, Economic and Societal Impacts of Tornadoes 2011, pp 117-171 DOI 10.1007/978-1-935704-02-7_4 http://link.springer.com/chapter/10.1007/978-1-935704-02-7_4#
2. WSR-88D Build 10/Super Resolution Level II FAQs, NOAA Radar Operations center: http://www.roc.noaa.gov/wsr88d/buildinfo/build10faq.aspx
3. Tornado Trends Over The Past Thirty Years by Daniel McCarthy and Joseph Schaefer of the NOAA/NWS/NCEP/Storm Prediction Center, Norman, Oklahoma http://www.spc.noaa.gov/publications/mccarthy/tor30yrs.pdf
4. Tornadoes occurring earlier in “Tornado Alley”, AGU Newsroom 16 September 2014, http://news.agu.org/press-release/tornadoes-occurring-earlier-in-tornado-alley/
5. Inflation Adjusted Annual Tornado Running Total Trends by Greg Carbin, NOAA Storm Prediction Center2008 http://www.spc.noaa.gov/wcm/adj.html
==========================================================
[Added, should have been in original release] Prepared by OAS members Joe D’Aleo and Anthony Watts. Assistance from NOAA SPC Greg Carbin is gratefully acknowledged. For further information, comments, or other issues, write to contact “at” theoas dot org or use the contact form at http://theoas.org
Yes. This is debate that should have commenced in the public arena among climate scientists long before they got into bed with politicians salivating over purses and wallets. It will be a dog fight but it must be done. The specter of the science herd being driven by Gore’s dangling carrot on a stick will forever be a shameful half century in the annuls of history. And those that remained silent during that time bear at least equal shame, if not more.
You might have thought that Long and Stoy would have known all this, but there again, if they were only writing it for the benefit of msm headlines, they’d be sure no churno would check.
It is the opinion of The OAS that this sort of methodology to remove a portion of a dataset to cite a result is unsupportable and without justification.
==========
exactly. remove different states and your will get a different result. so what you have is no result at all. instead it is cherry picking.
“Some of the greatest annual standard deviations (both positive and negative) have occurred within a little more than the past decade. This is perhaps the most puzzling aspect of the record.”
============
it is only puzzling because of your assumptions. you are assuming the data follows a normal distribution. however dynamic systems are not normally distributed, they are a power law distribution.
outliers are common in power law distributions. it is the naive approach to statistics that is the source of the surprise.
First encountered this idea in an intro statistics course about 50 years ago:
http://en.wikipedia.org/wiki/Pareto_distribution
part 4. Applications — of interest
They say “for the past 61 years [..] there is a distinct upward trend [but for] the past 25 years [..] essentially no trend“. Yes, that’s naive. The behaviour isn’t linear so of course fitting a linear trend to different periods will give different results. Just fit linear trends to different parts of (say) a sinewave, and it’s easy to see why.
By manipulating data, statistics can be made to support your bias. Its a well known means of manipulating the public!
I’m reminded of the old phrase “There are three types of lies – lies, damn lies, and statistics.”
I don’t recall seeing much in the way of statistics. I saw a mean that was then adjusted. I suppose the variance is zero. Then there were statements about coming earlier if you cherry picked the data. Again, was the start date in 2010 statistically different than 2000 or any other year?
The exact quote is: “There are three kinds of lies: lies, damned lies and statistics.”
by Mark Twain in his autobiography.
Pleasure to read something about a measured trend in climatic observational data which doesn’t seem to need to invoke CO2 in its discussion. What could explain it – phase of the PDO, changing jet stream, SSTs at the equator?
what trend?
Forgive me if I didn’t read it right but weren’t they showing that tornados were happening earlier in the season but with a steady magnitude?
“The increase in reported tornado frequency during the early 1990s corresponds to the operational implementation of Doppler weather radars. Other non-meteorological factors that must be considered when looking at the increase in reported tornado frequency over the past 33 years are the advent of cellular telephones; the development of spotter networks by NWS offices, local emergency management officials, and local media; and population shifts. Changnon (1982) and Schaefer and Brooks (2000) both discuss these influences on tornado reporting. The growing “hobby” of tornado chasing has also contributed to the increasing number of reported tornadoes. The capability to easily photograph tornadoes with digital photography, camcorders, and even cell phone cameras not only provides documentation of many weak tornadoes, but also, on occasion, shows the presence of multiple tornadoes immediately adjacent to each other.
These improved detection and reporting reporting methods most certainly contribute to recent increases in tornado counts, and by extension earlier and more frequent tornado counts than would have been recorded 30 years ago.”
I think the quotes from the article explain it very well. While Long and Stoy made an “inflation adjustment” to account for the increased reporting, the fact remains it is reports of tornadoes that have increased, and we have no way of knowing if the number and timing of tornadoes has increased.
Here’s an interesting site for U.S. Tornadoes
http://www.tornadohistoryproject.com/
climate science continues to operate under the assumption that “climate” has an average and deviation that never changes. it is the naive approach to dynamic systems that prevents them from making reliable predictions.
Was the paper peer reviewed?
The basic failings of this paper were noticed immediately when it was released by everyone, novice and scientist alike.
How embarrassing this is to the entire process of review and it is an indication of the total collapse of scientific integrity in general.
Now you have a great example of what Anthony was trying to avoid when he pre-released the surfacestations paper.
We have profited immensely from the criticisms we received and when we do release, it will be much, much tighter.
Observational bias.
I don’t understand the article’s statement:
“The NOAA SPC had shown in 2013 that there was no trend in strong tornados – EF3+ (see figure 4)”.
If I eye-ball referenced chart #4 it appears the 1954-1975 linear trend of F3+ tornadoes is about 55; the 1976-2014 trend is about 35. A 36% decrease is a rather substantial change.
What am I missing here?
What you are missing is the ‘statistically significant’ part that was unsaid.
Trends may be real, but they have to be out of the norm to mean anything new is happening.
Please define statistically significant in the context of the F-3 and higher chart.
Who wrote this statement? Attribution to a “Society” raises the specter of the Royal Society, AAAS, etc. who are continually coming out with “position statements” without consulting the membership. Anthony, please let’s not make this mistake. Let the authors be known and stand behind their work.
I assume you have chosen this path because you think it may get these views into more journalists’ articles. You may in fact be right, but it opens the door to more “behind the scenes” work by a few “special” members of the Society as time goes on.
I repeat: Who wrote this and will he/she/they answer questions regarding their work?
Sure, Lance, no problem. This is the first one, so we are still sorting out the format and methods. Making such statements is listed in the goals:
“To offer statements and positions regarding atmospheric science as it relates to current news, where appropriate.”
The statement was written by Joe D’Aleo and myself, and we consulted Greg Carbin of the NOAA SPC and one other severe weather expert (in private practice) who provided an advisory role, but no written material.
Once we get our membership structure together, we can work on having more input and a better process that is more inclusive.
Thank you, Anthony, I do support the idea of the new Society providing an opening for persons of all views, and hope for your success.
It shows how important it is for all members to be very involved in who is on the Board, and to actively participate in whatever way they can.
I agree. We have a method we can put in place outside of the normal peer/publishing platform, We’ll work to get that implemented.
I would like to clarify that I thought the statement was excellent, and entirely within the directive of OAS – to issue timely analyses which are easily understood.
Allow me to share a few experiences, so that my remark can be taken in the spirit it was given.
In my view, scientists join organizations to showcase their own work, in general. They seek to be published and to talk about their own theories. It is just in the way of things.
This means that many of the administrative jobs and board positions are not well filled by the scientists who join the society, and the membership does not always find time for these meetings either. As a result, Boards get filled by people we don’t know that well, and who may have entirely different ideas about the goals of the Society. A scientific association should allow that people serve in their areas of giftedness – whether it is facilitating online meetings, building websites, presenting work, seeking new means of communicating, or whatever the case may be. Because administrative and tax issues are so uninteresting to scientists and members, it is easy to let any one who is willing take over these boring issues. But it does pay to remain engaged and interested in the Board and the direction of the organization. That is all I meant to say. And I do think most of us expect that AW will be a big part of the structure and future of the OAS.
Don’t know how far back the method goes, but the present procedure of counting a new tornado every time one lifts and resets (as explained by a local forecaster) or spins a brother at the same site looks to result in gross over counts. However it sure impresses the public who don’t realize how the counting is done. I’ve seen what should be a hand full of tornadoes here in Oklahoma turned into dozens by this method. In earlier days, when there was visual reporting only, these numbers would have been drastically lower. It is also difficult to find info on counting procedures (I wonder if that could mean anything), If anyone knows of a site with detailed primary info on count procedures, it would be appreciated.
I watched this happen this summer — tornadoes would appear out of the cloud, spin down to almost or barely touch the ground, and then “instantly” break up and dissipate. Two minutes later, the cloud would spawn another funnel that might only go halfway to the ground. Then another one that reached the ground. I watched easily four or five “tornadoes”, none of which persisted for any length of time on the ground (or possibly water — this was inside the waterway between Morehead City and Atlantic Beach and I couldn’t see the touchdown site) form in five or six minutes, and this cloud was identified as producing tornadoes by radar (which then buzzed all cell phone customers who were signed up for warnings with a real-time alert) fifteen minutes earlier which is how I came to be looking out the window to see it in the first place.
How many “tornadoes” was this? Some of them happened earlier in a rainstorm that was so dense that a tornado could have touched down 200 feet from my house and I couldn’t have seen it — I could barely see my own boat at the end of the doc less than 100 feet away. Nobody could have seen them if they came down on water or land and weren’t outside, right at hand. Radar did. Presumably there was at least one waterspout seen by humans on the far side of Beaufort. But the series I watched could have been counted as a single very weak tornado, no tornadoes at all, or six or more tornadoes (and possibly more — I couldn’t even see the cloud until the rain cleared and it could have been spawning transient funnels for ten miles back, and then there might have been other clouds).
Radar probably got them all, where humans might have missed seeing any of them, if we weren’t all alerted by screeching phones to go look out our windows. Or rather, to hide under our beds and not look out the windows, but human nature being what it is…:-)
rgb
Exactly. And who knows how the count procedures have evolved over the years (maybe like the temp adjustments) or how, or if, they are standardized. Looks like an area for some investigation.
Have they changed the definition? By my count of your description and my recollection of the definition of a tornado that required touching the ground, there were two tornadoes, F0s, pending damage data.
I would suggest that this would be a good time to give the appropriate agencies the benefit of the doubt. Lets go look at all of their justification for budget increases and see how much they claimed they could improve their data gathering. If we then assume that they did improve by that amount, we can then use that to adjust past records to see what they would have been if the current data gathering methods had been in use.
Well, and bear in mind that improved tornado spotting has a point. It saves lives and property. It is this point that matters. The “interpretative dance” of analyzing the data produced with ever improving detection and feedback-enhanced cell phone reporting (where the radar sees a single cloud capable of spawning a tornado, makes 50,000 phones nearby screech a warning, resulting in 100,000 eyes out actively looking for tornadoes and texting back pictures and movies to all of their friends and the news agencies) really is secondary to this primary purpose. Done badly, done well, even the unadjusted numbers aren’t particularly scary. Tornadoes happen. From the variance alone in the graphs above, there isn’t much good reason to think that there are any serious long term trends visible in the data, even before one thinks about improved reporting and positive feedback with cell phones and the internet.
rgb
Indeed. Back when I was a child and became aware of ‘twisters’ from the stories my grandparents told of the 1932 twister they (and my mother) survived, we were told that the US averaged 1000 tornadoes a year. 1000, or 1200, that’s the same order of magnitude of incidence.
I like the open honest discussion, that’s the way progress in science is made. I’m sick of hearing politicians say, “the debate is over”, “the science is settled”, they disgust me.
If we take Kansas out of the data….we never would have The Wizard of Oz. If we take Oklahoma out of the data…we never would have The Grapes of Wrath. Pick your past.
I’ll take the Wizard of Oz, so leave in Kansas. Never liked Grapes of Wrath, so OK is out. If we drop Missouri do we lose most all of Mark Twain?
Interesting.
In it’s first open publication we see.
1. No authors ( membership in the OAS is secret)
2. No data or methods.
3. No evidence of review
Now, we can grant that this is not a “paper” but it is a publication.
Granted other organizations, are only a bit more open ( not keeping member private)
Since this is a press release, who is the PR contact?
Who can I contact if I have questions about the release, in short what scientist will speak for this release?
And typos.. but I live in a glass house.
Other than those nits.. nice start
@Mosher, some valid points there,
1. The authorship and contact info should have been included, it is now. We have not even queried the membership yet and their wishes regarding open membership lists or not, so that issue is premature.
2. The is no data or methods because no new analysis was performed, the NOAA SPC analysis was used and referenced appropriately.
3. You make a good point, while 4 people had a hand in this, should such short statements like this be reviewed? Yes I think they should. There is a separate review facility in place outside of the Annotum platform for statements such as this. We’ll work to get that implemented so that membership can review such statements outside of the formal review/publishing environment. I think future statements will be better as a result.
One thing for everyone to note here: Valid criticism is OK and actually welcomed, we have the luxury of being able to learn and change based on member input.
Anthony,
Get a copy-editor, preferably an English major on board for a final grammatical review (no not Mosher).
Have all public statements reviewed by the editorial board (once established).
[No. ~ mod.]
“No” what?
Yes, yes this is the way to do it. Use the strictest scientific rectitude and cite from the study to show the errors. Sloppy science will find it more and more difficult to find a reception.
As a member (At least I hope I was accepted.) I hope that the OAS creates a standard for commentary wrt published lit. As an example, This commentary would reside well under an area of members comments on articles. This was the voice of OAS is that of its members.
As a possible member also, I note that the article is posted on the OAW website, here:
http://theoas.org/2014/09/19/statement-from-the-open-atmospheric-society-on-the-paper-peak-tornado-activity-is-occurring-earlier-in-the-heart-of-tornado-alley/
Sorry, In this WAY, the voice of OAS is that of its members.
Woah! Wait! Wait! Wait!
Nothing aggravates me more than a bad argument supporting a correct conclusion.
How does the detrending shown by figure 1 and figure 2 refute the argument of the original paper? I am not criticizing Brooks and Carbin – I haven’t read their paper and don’t know where they were going with it. I just can’t see how their chart applies here.
The overall argument (and a sound one) is that the trends in the Long and Stoy paper are artifacts of better observation, and not real. But to show that, the OAS critique looks at data that was detrended BASED ON THE ASSUMPTION OF NO TREND. This is circular reasoning at it worst and will now be my go to example when explaining the correct use of the phrase “begging the question”.
The quick seat-of-the-pants check is figures 3 and 4 – looking at larger tornadoes that are more likely to be noticed. The long and difficult, but right, way – is to develop estimates of detection rates by looking at tornadoes of differing magnitudes in populated vs unpopulated areas over time.
Consider this a lesson for the OAS – maybe some time for low-profile review and comments before putting it at the top of the WUWT page in the format of a press release. If I were a True Believer I would trot this out every time someone asks “But what about the OAS paper that says…?”
Climate shift? Perhaps their statistics do not reflect the actual process. They may want to limit their observation frame. After all, this is science, not philosophy, or faith.
So sometimes you’re in favor of adjusting data and sometimes you’re not
Do I see something not far off a ‘homogenisation’ of the raw data in the OAS comment here? OAS sounds like a great idea but it will need to guard against taking positions so easily or it may quickly look pretty partisan. It might end up with contributions only from those already on-side, and looking a bit like WATTS UP WITH THAT. I agree that the Long and Stoy paper doesn’t amount to very much, the flagged-up conclusions being very contrived, and it probably didn’t deserve the money and effort that was put behind it. But be fair, those comments are a bit out of order.
Chalk a partial point up for the Mosh: sometimes raw data just doesn’t make it.
yup. I’m skeptical of raw data. I’m skeptical of adjusted data. Skeptical of models too.
and on strange days im skeptical of skepticism as a method
@Anthony Watts; good start. pg
Professionalism of climate researchers .. no, I’ll be polite.
Like with cyclones, the relevant metric I think should implicate time on the ground and strength. Just number of occurrences seems irrelevant to me.
I’m not trying to be facetious, but if you are looking for a location to build your house or park your trailer, the number of tornadoes, regardless of time on the ground or intensity, that historically occur in that location is very relevant.
I think you are correct, jayhd. But if those tornadoes did not touch the ground or just barely, you would be mislead.
Detrending using a linear trend to a 54 year detection record, when the detectors are getting quantum upgrades over time is to me a huge source of error. A detrended gradual decline is the result of static detection sensitivity. A detrended rapid increase is the result of rollout of new hardware or software.
The detrending makes the assumption that 1954 on average is the same as 2007.
And then we look for differences???
Detrending is not subtraction. It is the ADDITION of a modeled quantity with error. Any statistical manipulation adds error. In this case it is the addition of an unknown and conceivably huge amount of error. What the paper is reduced to is analyzing the error they themselves added by the detrending.
My question was more basic. What happens if we take Texas out- or Oklahoma, or Saskatchewan?
I guess that if it is OK to adjust temperature measurements in the ’50s to be colder (and fit presumed trend line) then it is also OK to adjust tornado counts upward for that same period and similar reasoning.
Why would a change in the overall number of reports shift when they occur during the year? The technique finds the mean of the date of occurrence and the standard deviation of the dates of occurrence for each year in Texas, Oklahoma, Kansas, and Nebraska, both individually and as a group. The trend is then estimated for those calendar dates. Unless you have some reason to think that the increase in reports has preferentially happened earlier in the year, the increase in reports since 1954 has zero impact on their result. Long and Stoy don’t have to do any inflation adjustment because all they’re looking at is when tornadoes occur during each year, independent of all other years. If you asked the question “has the date when babies are most likely to be born during a year changed in the US over the last century”, the fact that there are more babies born now than 100 years ago wouldn’t change when during the year they are born.
As far as the Nebraska comment is concerned, they give the result for Nebraska in the paper (earlier by 4 days, p-value of the change is 0.03). The comment in the press release is based off of Table 1, which shows that the peak in TX, OK, and KS have all moved >10 days earlier. All the states show an earlier season, but Nebraska is the smallest change.
Your comments make me wonder whether you actually read the paper.
Now I feel dumb. Talk about missing the forest for the trees. The glaring flaw in the argument regarding whether there was any actual increase in the number of tornadoes blinded me to the fact that the original paper really wasn’t about that.
And I don’t think anyone else commented on it either.
At least for me, there is quite a lesson about the psychology of all this to be had here.
Holy cow!!! It is ever great to have an actual expert commenting! Harold, thanks so much for offering your input. Most people do not understand how difficult tornado research is, and what the assumptions are that must be made. We do our best with the data we have. It is not perfect, and all tornado researchers realize that.
Most people here do not even read the papers they are commenting on.
Thanks again.
Max wrote: “Most people here do not even read the papers they are commenting on. ”
No doubt some people don’t wish to pay GRL $35 for access, and that same claim would probably hold true for most MSM reporters who simply regurgitated this press release without reading the paper. Newsrooms of today don’t get GRL subscriptions.
But if researchers want their papers to be read, they should make them open access.
Making claims in the press release that can’t be easily researched by anyone reading the PR because the paper is behind a paywall does science a disservice in my opinion.
Bell, I commented on this above. The abstract even explicitly states this, and notes an average of 7 days earlier:
“Tornado frequency may increase as the factors that contribute to severe convection are altered by a changing climate. Attributing changes in tornado frequency to observed global climate change is complicated because observational effort has increased over time, but studies of the seasonal distribution of tornado activity may avoid sampling biases. We demonstrate that peak tornado activity has shifted 7 days earlier in the year over the past six decades in the central and southern US Great Plains, the area with the highest global incidence of tornado activity. Results are largely unrelated to large-scale climate oscillations, and observed climate trends cannot fully account for observations, which suggest that changes to regional climate dynamics should be further investigated. Tornado preparedness efforts at individual to national levels should be cognizant of the trend toward earlier peak tornado activity across the heart of “Tornado Alley”.”
I can’t see the entire paper to comment on the “outlier” issue, but I’m curious if a rare December tornado should be considered late season or early season. In any case, I very much doubt that removal of 3 outliers in 46 years of data had much effect on the results.
Then why remove it at all?
@Barry – So you did. The second comment, no less. I missed it somehow.
Now that I’ve seen the discussion between Harold and Anthony, I’m even more puzzled about what the OAS statement has to do with the original paper.
Dear. Mr. Brooks, thanks for your comments. You asked:
What we see in the data is an overall increase in reporting across all calendar dates. The new warning systems in place have had a great effect on tornado lead times, and an increase in storm chaser teams “itching to go” at the first sign of potential severe weather, combined with improvements in convective outlooks, NEXRAD signature detection, and other NOAA based and private improvements in storm chasing most certainly have an impact.
For example: https://www.kickstarter.com/projects/tornadochasers/tornado-chasers
There are now many more eyes, budgets, and reputations staked on catching the early season tornadoes where there was not 20 years ago. Plus, with the ability to interact with NOAA radar live on ceel phones, tablets, etc, the accuracy of spotting tornadoes has increased.
While they claim there is no effect on tornado reporting increases having a bias (mentioned in the last paragraph of the introduction) they did no analysis that I can see that specifically excludes that possibility. In fact, they removed some late season data, that if it had remained, might have negated some of the early shift they claim.
I don’t know how one can justify valid tornado reports as an “outlier” simply on a whim. Who decides if 4 STD is the limit? That seems like an arbitrary author choice. Either it is valid data or it isn’t.
We found no evidence that contradicts an increase in early season reporting due to these external reporting factors mentioned above. We would expect to see this evidence across the data set, as you note about table 1. f you have evidence that these external reporting factors have had no effect on the early reporting increase, I welcome seeing it.
The idea of removing Nebraska to make a point with layman readers who cannot get access to the paper should not go unchallenged. The layman certainly can’t, since the paper is not accessible unless you are a GRL subscriber.
For reference, here is Table 1 from Long and Stoy 2014:
It should be noted that the conclusion engages in quite a bit of wordsmithing:
Preaching for preparedness is a good thing, preaching that even though no other metrics or consensus support a link to AGW but somehow this one paper does via some statistical and outlier removal claims that in my view have not been fully examined, leads me to believe this paper is more speculative than factual.
There’s no way in the world three late-season clusters over 60 years would affect the results in any significant way. It can only impact three years out of the 60 in the series.
Actually, the paper is available for free. From the press release “A PDF copy of this article can be downloaded at no cost by clicking on this link: http://onlinelibrary.wiley.com/doi/10.1002/2014GL061385/abstract”
Defining outliers as exceeding some large value of standard deviation is a typical procedure. 4 standard deviations is more than is usually used. It represents discarding ~0.1% of the data. 3 SD is more usually done. They’re actually retaining more of the data than would typically be done.
My first cut, not using the circular variable approach, for Oklahoma has all tors having 1 day more of a change than (E)F1+ does, so a small effect. The authors indicate, in the press release, that they’ve carried out that work with their more sophisticated statistical approach than I could do in a couple of minutes, and also found a small change when thresholding for higher F-scale values. The thought that reporting increases would shift the distribution (not make it wider, but move the whole distribution) is a rather extraordinary claim.
Feel free to do your own analysis, but the objections you raised in this post are irrelevant to the paper.
Thanks Harold for your comment.
Two points.
That might be true for NOAA institutional access, but certainly isn’t true for the public at large. Try it from home on a browser not connected to any GRL accounts or NOAA. I think you’ll find the claim to be untrue. Perhaps they intended it to be open access, and it got bollixed, but I had to pay for it and others report to me they can’t download it for free, either.
Then why remove them if they have no effect? I find that sort of logic hard to support. If it has no appreciable effect, leave it in.
I fail to understand why you don’t see increased reporting as having an effect in early season.
And, when you did your analysis, did you remove those outliers or leave them in?
Anthony, I agree with you that it is very difficult for the “average” person to gain access to many of the research papers due to paywalling. Completely open access would be great.
However, there are far too many commenters who make very categorical statements without having read the paper (or probably not even looked at the abstract). This is not at all fair to the authors.
It might not be fair to the authors for such comments, I contend that news people who publish stories on such press releases without ever having read the abstract or article are a far greater problem and that is unfair to the wider public at large.
Anthony,
By snipping the section on Linear Models, you leave out the fact that the results in Table 1 are based on Circular Models, explained in the previous section of the paper (not shown). There were no outliers omitted from the circular model analysis, and in fact the authors indicate that the linear models show even larger shifts in the peak date (e.g., 22 days in OK).
Why would a change in the overall number of reports shift when they occur during the year?
The better the reporting, the more the “when” is extended — on both ends.
The OAS criticism does state clearly that improved reporting methods not only increase the number of reported events, and this affects earlier onset stats. Also that the timing of the season coincides with improved technology.
But the analysis shows that the width of the season hasn’t changed. Things aren’t extended on both ends. They’ve just moved earlier in the year.
jayhd wrote
September 19, 2014 at 10:25 am
I’m not trying to be facetious, but if you are looking for a location to build your house or park your trailer, the number of tornadoes, regardless of time on the ground or intensity, that historically occur in that location is very relevant.
The short answer is that almost anywhere west of the Rockies is safer than any location east of the Rockies, vis-a-vis tornadoes. Please see this great map showing widespread tornado incidence over the last 50+ years across the entire US east of the Rocky Mountains.
http://wattsupwiththat.com/2012/05/31/stunning-map-of-noaa-data-showing-56-year-of-tornado-tracks-shed-light-on-the-folly-of-linking-global-warming-to-severe-weather/
An even shorter answer is that your chance of dying in a tornado is about 1 in 60,000, according to some data, while dying in a car wreck is about 1 in 100, or on a bicycle about 1 in 4700.
~
Here’s another quick index card analysis of recent tornado activity by state:
Tornado Fatalities by year 2003-2013, by state, abstract
read
a = year; b = # of US states with tornado deaths, total dead; c = leading tornado fatality state # deaths,
a b c
2013 007, 055 OK 34, IL 8, TX 7
2012 010, 070 KY 23, IN 14, IL 9, MO OK 6
2011 015, 553 AL 245, MO 158, TN MS 32, NC 26, GA 16, OK 14, AR 12
2010 013, 045 MS 13, OH 7, AR 5
2009 008, 021 OK 8
2008 016, 126 TN 31, AR 21, MO 19, IA 13, KY 7, OK AL 6
2007 013, 081 FL 21, KS 14, AL GA 10, TX 9
2006 010, 066 TN 34, MO 13, NC 9
2005 009, 100 IN 24
2004 013, 035 IL 9, MO 7, FL 6
2003 009, 054 MO 19, TN 12, KS 8, GA 6
Summary
Over the period 2003-2013, 34 US states suffered tornado fatalities
read state, # appearances on yearly tornado fatality list 2003-2013
MO 10, GA 8, TN 7, OK 7, AL 7, AR 6, IL 6, MS 6, KS 5, KY 5, FL 5, LA 5, IA 4, TX 4, MN 4, IN 4, NC 4, WI 3, SC 3, NE 2, CO 2, MI 2, ND 2, OH 2, NJ, WY, PA, NM, NH, MT, NY, WV, MA, VA
Extracted from NOAA tornado fatality data:
http://www.nws.noaa.gov/om/hazstats.shtml
The bottom line is that tornadoes can occur almost anywhere in the US east of the Rocky Mountains when conditions are right, i.e. warm moist air from Gulf colliding with cool dry air from the north. Hot spots of tornado activity in the US vary from year to year. Consider the Palm Sunday outbreak of 1965
the http://www.tornadohistoryproject.com/outbreaks/the-palm-sunday-outbreak-1965
The main reason for the production of tornadoes is the production of lunar tidal bulges in the atmosphere, as the moon moves from over the equator towards the poles. On the Eastern, lee side of the Rockies the production from the two halves of the produced bulge, which every one acknowledges, the southern warm moist air mass clashing with the northern dryer colder air mass, always occurs just as the moon’s declination reaches culmination (Maximum North or South and heads back to the equator).
It is just an atmospheric game of crack the whip, Coriolis effects whip the cyclonic turbulence up to the point of tornado production.
There are cyclic patterns in the production of tornadoes due to these atmospheric tides, the patterns of declinational tides are not fixed to arrive the same dates each year, so the start dates of tornado production varies with the bulge generation arrival times.
unless you synchronize the data base to the lunar declinational patterns it is hard to see there are really patterns to the production rates.
http://research.aerology.com/wp-content/uploads/2010/03/6558days-cycle.jpg
This is a graph (click on to enlarge) of the past 3 cycles of tornado generation starting with the oldest being blue, then red the second cycle, and green the third cycle it is easy to see the increase in total numbers of reports between them. The Yellow spheres are the most recent cycle, with data on this graph posted up to 2010.
The “trend in tornado production” follows the change in the declinational angle of the moon at culmination. It is entirely possible to predict the time, and close location of the next set of outbreaks by looking at the current harmonic oscillations of the atmosphere.
If the weather service would look at this effect on the meridional circulation patterns in the global circulation, great advance in forecasting could be made.
That is really good! No Kidding. That’s 9 and a quarter years from where the moon starts to where it finishes then starts back for a total of 18.5 years. To put those 2 items together is awesome. I found the graph very interesting as well. If it were in 3-d it’d look somewhat like a dna molecule. I’m sure you’ve seen other features as well. Excellent.
As a side note. The Ute Indians in the southwest used a circular spiral that was 9 and quarters turn. Sometimes you can find it on their pottery. The purpose of that spiral, when set up relative to the horizon, was to gauge where the moon was in it’s cycle. You may already know that. I was looking at it from the perspective of the Indians who generally lived on the other side of the Rockies. There must be a connection, otherwise they wouldn’t have put so much time into it. I don’t know what it is.
If you did that, my hat is off to you.
likely the moon’s orbit also pumps the ocean circulations PDO/AMO as well, with an underlying 18.5 year harmonic imposed on the natural frequencies of the ocean basins/currents.
The ocean basins/currents each will have their own harmonic frequencies, and as they come in and out of phase with the lunar cycle, the ocean circulations are increased or decreased.
this has the effect of altering the mixing rate of the oceans, leading to complex patterns of alternating warming and cooling.
I would hope that whoever enters in the discussion has actually read the paper. This post is confusing to me because the authors analyze the date of peak activity in a year, not the total number of tornadoes reported in a year.
Dear Barry,
It is unclear if the authors of the OAS press release, or many of the people who provided comments here, read our paper. Our study did not investigate trends in the number of reported tornadoes over time.
Sincerely,
Paul C. Stoy
Montana State University
[Thank you for the courtesy of your reply; as you see appropriate, we request you address the specific questions many of our readers and reviewers brought up in the comments above. .mod]
In response to the moderator request, our study did not investigate changes in the number of tornado numbers over time. I further thank Barry for having read the paper.
Does tornado alley exist because of what happens in the atmosphere or is it because of what is taking place on and under the ground ? IMO tornadoes are electrically driven through the process of conduction. The ground becomes negatively charged as lightning strikes and draws positive charge from the top of thunderheads to the ground . Current flows from positive to negative and a funnel starts to form from the base of the clouds and if the negative charge is strong enough at ground level the funnel will touch down .
https://www.thunderbolts.info/tpod/2005/arch05/050330tornado-electric-discharge.htm
Is a tornado a 3 phase energy pathway? http://science.howstuffworks.com/environmental/energy/power3.htm
As population grows in tornado alley, so the need for more and more wireless comm’s and remote sensing takes place. http://globalmicrowave.org/
I suggest that this one graph s strongly indicative of no increase in tornadoes, indeed, a clear decrease post 1975 or so.
Ideas on this to consider and further research.
Clearly this chart eliminates the detection problem of smaller tornadoes. Apparently detection of strong tornadoes was equally good to today in the beginning of the chart.
Now consider that a weather system that produces strong F-3 to F-5 tornadoes, in general also produces many weaker tornadoes as well. This could be analyzed to estimate, with large error bars, the earlier part of the record before improvements in detection..
Also, it may be possible to more accurately analyze the observation dates of the F-3 to F-5 tornadoes, to see if there is any clear trend in their seasonal appearance.
Also this research could have been carried out to include the most recent complete season.
Nice. Needs more data to identify trends, or cycles. Vertical bars (y-axis) could be color coded to indicate month, date, or season in addition to year (x-axis) while spatial information could be shown on the z-axis.
Hey, I’m all for climate change. We didn’t hear a single tornado siren in Wichita in 2014.