Slower decay of landfalling Hurricanes in a warmer world — really?

Reposted from Dr. Judith Curry’s Climate Etc.

Posted on November 17, 2020 by curryja

by Frank Bosse

A recent paper published in “Nature” made some excitement in the media, see here or here.

In the paper by Li & Chakraborty (L&C 2020 thereafter), the authors find a statistically significant increase of the decay time when a North Atlantic hurricane makes a landfall due to warmer SST in a warming environment. They also undertake some model-related research about the impact of this observations.

The key point of thepaper is the finding that warmer SSTs lengthen the decay time of hurricanes after landfalls.

In L&C 2020, this is shown by figure 1f:

Fig.1: The reproduction of Fig.1 f in L&C 2020. The ordinate reflects the decay time τ in hours.

In the legend the authors state: “We note that the τ time series echoes the SST time series with Pearson correlation r = 0.73.”

The authors describe the way they found the relation, which declares an increase of the decay time of more than 40 hours per 1K SST increase:

“We average τ for all the landfall events in a given year and apply a 3-year smoothing, twice in a row, to this time series.”

They made a regression with strongly smoothed time series, a procedure that is normally frowned on.

In the supplementary data (freely available) one can download an Excel sheet where the raw data used can be found.

For the deduction of the increasing τ with increasing SST (in the area 10°N…35°N ; 100°W…75°W ,  the authors take advantage of the data for 71 landfalling hurricanes during 1967 to 2018.

The SST for each event is determined as follows:

“We average the SST in time over the hurricane season, June–November, …”

The result “R=0.73”, see Fig.1, of the linear regression implies that 53% of the variation of τ comes from the variation of the SST.

I had a look at the raw data and a few questions arose:

  1. The use of the average SST of the whole hurricane season for a single event?

The actual named hurricane develops over a few days in an actual environment, not the average SST of the actual whole season. It makes a difference if the landfall happens during July or November, the average SST difference is 1.8 K in this case, which is much more than the range of the abscissa in Fig.1 .

The use of the seasonal average SST  for all hurricanes during that season, rather than the actual SST applicable to each hurricane, has the potential to produce highly misleading results. The average SST applicable to each hurricane might have little relationship with the average SST during the whole season

  1. The use of the average of all τ in a year if more than one hurricane is involved?

Every hurricane event is a discrete event. In the raw data many years have only one hurricane per year, these events are not averaged of course.

  1. Applying a double 3 years smoothing before making the regression shown in Fig.1. ?

The authors state:

“this approach lessens the effects of non-climatic factors and random noise”.

However, the whole research is about the point:”To what degree impact the actual SST the decay time of landfalling hurricanes?” There will be some other influences and it’s not appropriate to smooth over several years partly out to elicit a strong climate related signal. Applying a 3-year smoothing to both decay time and SST data twice in a row is unjustifiable.

I decided to recalculate the regression shown in Fig.1 but I used the actual SST for every hurricane from the monthly ERSSTv5 data for the described area. I included every hurricane because this is the physical approach: It’s not justified at all to use an average in some years and in some years not, as that gives radically different weightings to hurricanes depending on how many are included in the raw data in each year.

I also use the unsmoothed data to avoid spurious correlations due to the applied smoothing.

This is the result:

Fig.2: The regression of the decay time on the SST without the data-preconditioning in L&C 2020.

There is only a tiny non significant trend in the raw data- p=0.1, so the slope does not reach the standard 95% confidence level.

On twitter  Ryan Maue questioned the raw data selection; that issue is beyond the scope of this post.

The outcome of L&C 2020 is very overconfident when it comes to the dependency of the decay time on the SST. The R²=0.53  found in LC 2020 vanishes to an insignificant 0.04 if one uses the physical data, without the applying of unjustified averages and smoothing actions prior to the regression.

This means:

The SST impact on the decay time is negligible, other influences accounting for almost all variability in the decay time.

The peer review process of “Nature” for L&C 2020 lasted more than 8 months, it makes wonder if there was no reviewer with some fundamental skills in statistics involved.

However, this must be the case unfortunately: In the “methods-statistical significance” section the authors mention a test for autocorrelation and there is written: “(which we test using the Dublin–Watson test)”. This must be a typo, the name of the test is “Durbin– Watson”.

One should hope that the peer review process of “Nature” would be improved soon to avoid overconfident, obviously flawed papers like L&C 2020.

61 thoughts on “Slower decay of landfalling Hurricanes in a warmer world — really?

    • Yep, the original ‘The Sky is Falling’ just doesn’t get much traction anymore. I mean, even AL Gore’s favorite, Global Warming has been proven to be a hoax!

    • It would appear warming SST and unchanged atmospheric temperature would lead to a ramped up heat engine for hurricanes. But, as they claim the atmosphere is warming faster than SST, then the heat engine is weaker, based on temperature differences.

      With weaker heat engines, there would likely be more category 1’s and 2’s which would be slow and ponderous and slow to fall apart because of internal energy and momentum. Without selecting a hurricane category, it would be very hard to wrap one’s mind around changes in hurricane dissipation. Just a widening of the categories without considering changes in the average category, makes this hard to examine.

    • Oh no ResouceGuy, there’s an endless number of alarmist angles to spin, LOL. I just received a 200k grant to study the effect of climate change on presidential elections. My conclusion is that the more CO2 we emit the greater the chance of a contested election. Now all I have to do is play with the numbers to fit my conclusion. Under the RCP 8.5 scenario, I project we will be in a civil war by the end of the century, however more funding is needed to study this emerging problem more deeply (ha ha).

  1. Yet more proof that peer review is not a valid process for determining the worth of papers. What PR means nowadays is you have cited a paper of mine and I agree with your conclusion as it fits my pre-conceived notions, so publish. – it is pal, not peer, review. The fact that so many junk papers get through with either bad data or bad maths would ring alarm bells anywhere else. But in “climate science”, it is the norm.

    • “peer review is not a valid process for determining the worth of papers”

      There are many ways to objectively determine the content of a paper. Here are two.
      1) If the paper is on any topic in Inorganic Chemistry:
      Burn the paper and weigh the ashes. This is how we all learned to determine the inorganic content of a paper.
      2) If the paper involves thermodynamics:
      Burn the paper in a calorimeter, measure the heat produced.


  2. My impression is, there are “scientists” first looking for “where can I blame climate change” and than start to specify their research with assumptions not fitting the problem to get the chossen result.

    • They don’t just use assumptions. They torture the data with “unusual” maths (and dataset modification) to get the answer they want.

    • This is a spectacularly bad example. I’ve seen shotguns produce patterns with higher correlation than the raw data here. Of course, no one would publish a shotgun pattern . . . or would they?

    • The scales on the graph appear to be chosen to make the graph look “normal”.
      The hour scale has has 60 hours packed into a half the size of the horizontal scale.
      The temperature scale also span less than a degree K exaggerating the slope effect.

      Another problem is that the five point on the right end of the graph have error bars 2-5 times as great as the cluster on the left. The whole claim could fail if one or two were dropped.

      The reviewers might have come from the Music or Women’s Gender Studies.

  3. It would seem to me that you must take a hurricane and compare it with others that landfall in the same area. Since topography plays an important part in the time it takes a land falling hurricane to decay, I don’t think it makes much sense to compare a southern Florida land falling hurricane to one where there are higher elevations and a rougher topography.
    And that is just one factor. Dry air and wind shear are other important factors in how fast a hurricane decays. Without comparing each hurricane with others having the same surrounding atmospheric conditions you cannot draw any sensible conclusions on decay times.

    • Tom: this would bring the “slope” more down. My intention was to show that it would be possible to make proper assumtions with the available data in their own sheet.

      • frankclimate, Tom in Florida mentions “wind shear” as a factor. I seem to remember articles that ascribe the different wind shears through the hurricane formative areas as varying with ENSO state and strength. I wonder what the decay time of a hurricane is relative to ENSO state?

        • Ron

          Fast I bet
          La Niña decreases wind sheer so hurricanes this year should disperse slower

          Interesting to look at, at end of season, add to this chart

    • Also their dataset contains storms that move tangential to the coast and storms moving perpendicular to the coast. That’s a huge difference in how they should decay.
      Never mind. Put everything into a bag then shake vigorously. And the answer you want just falls out.

    • A critical assumption made by the authors of L&C2020 is that the term “landfall” is a rigorously scientific term critical to a hurricane’s life cycle.

      To this end, all hurricane “decay times” are based upon when a hurricane makes “landfall”; i.e. a hurricane’s decay time is based upon when the center of a hurricane passes over land.

      A hurricane’s center that brushes sandy coastal land and heads back out over the oceans is equivalent to a hurricane heading inland and removing all warm ocean water influence.

      Coupled to this assumption is a very selective time frame of hurricanes. A time frame choice that is just as false as starting all temperature or sea ice trend graphs from low points.

      Not only should hurricanes making landfall be compared to other hurricanes landfalling in the same location; they should only be compared to hurricanes making full landfall into similar topography and weather systems.

      From whatever perspective chosen, L&C2020 is as dodgy a bit of research as they come.

  4. Applying smoothing before linear regression is huge red flag and is almost always a major “Don’t do this,” especially on sparse data sets like the hurricane data set used by LC2020. There are less 50 data points in Fig 1, with large vertical error bars in the 5 right most data points. Those 5 pts with more than double the error range of the other pts are what pushes the slope upwards. The fact they used it (smoothing on a sparse data set and then LR) and the reviewers let them get away with it brings statistical competency discredit to the authors and the reviewers.

    • May have posted some of this before but you don’t need a degree in climate science (don’t have), good stat background (have) and hurricane experience (6 decades) does help, to know nonsense. Maybe smoothed it but least it wasn’t a log transformation which I run into too often. I call them bad or skewed choke shotgun graphs. Complex and difficult about jet stream enough for rejection consideration. Tropics and arctic at odds, other stuff in the way.

      “….in summary, we have shown that over the past 50 years the value of τ [ the decay timescale, for North Atlantic landfalling hurricanes has increased by 94%……If a landfalling hurricane interacts with the jet stream, the increased wind shear may cause its intensity to decay rapidly. The overall effect on τ will be mediated by the details of the interaction, which are complex and difficult to study. Future studies may shed light on the effect of such extratropical interactions on the decay of landfalling hurricanes.”

  5. More total BS from the Krazy Klimate Klowns – this time from the Red Cross. Where do they find these false-alarm hysterics?


    GWPF calls for the withdrawal of Red Cross report based on unethical practices

    London, 20 November: A new report from the Red Cross, which alleges a 35% increase in the number of climate and weather-related disasters since 1990, was today slammed for misleading the public.

    According to GWPF director, Dr Benny Peiser, climate-related disasters have actually declined by 15% in the last 20 years.

    UN data shows that “climate-related” disasters have declined over the past 20 years (2000-2019), Source: Roger Pielke Jr.

    “The authors of the Red Cross report have used the EM-DAT dataset which shows a significant decline in climate-related disasters since 2000. It is generally acknowledged that the dataset is unreliable before 2000.

    “In fact, the Red Cross’s own report documents a pronounced decline in climate and weather-related disasters since that time, so they have grossly misled the public,” Dr Peiser said.

    Figure 1.1 shows significant decline in climate and weather related disasters since 2000. World Disasters Report ‘Come Heat or High Water’, Geneva 2020

    Any credible assessment of the link between global warming and natural disasters should also highlight the remarkable decline in the number of people dying from climate disasters, showing a sharp and continuing decline in the last 100 years, despite the rise in global temperatures.

    Global annual number of deaths from natural disasters by decade. Source Our World in Data

    This is the second major report that has led to allegations of deception. Just six weeks ago the United Nations Office for Disaster Risk Reduction (UNDRR) was publicly humiliated when its report was accused of using the same unethical tricks to make claims of increasing climate disasters.

    Remarkably, UNDRR is one of the sponsoring organisations of the Red Cross report.

    “This is another intercontinental shambles” said Dr Peiser.

    “The Red Cross should withdraw its report or face accusations of using unethical practices to mislead the public.”

    The report, World Disasters Report 2020: Come Heat or High Water is published by the International Federation of Red Cross and Red Crescent Societies, Geneva, 2020.

    • “The Red Cross should withdraw its report or face accusations of using unethical practices to mislead the public.”

      Well, yes, it should withdraw it, but it won’t. The MSM will publish their own alarmist slant on the report as it stands, ignoring all criticism, so the public will never that know they have been mislead yet again and the report will go into the archives as fact.

  6. They are simply looking for some part of the cash-flow from the UNFCCC’s Climate Aid Fund. When they put their grant/money request package together to send to the UNFCCC funders, they will just reference this scam paper part of their “evidence” of need. That’s how the climate aid game is played.

    They are all hungrily licking their chops at the wet-dreams of Dementia Joe making illegal payments to the UNFCCC climate fund. As his master-puppeteer Hussein Obama did in 2016 and just before he left office in 2017 with 2 x $500 million payments from US State Dept funds appropriated for other issues, they expect Dementia Joe to illegally transfer funds to the Climate Scam as well without Congressional approval.

  7. At what point do these things become crimes, something that has actual consequences?

    Maybe a journal gets issued a “stop publishing” order for 1 year as a penalty for such shoddy misleading garbage?

  8. There are at three problems with this paper.

    Ryan Maue showed that the data Is a cherrypick. The upper outliers are all storms that made landfall then went back out out to sea to restrengthen.

    Frank Bosse showed the math is junk.

    And there is no evidence of the result in any ACE metric. Ifvthe paper were correct, there should be an ACE increase.

    As Heisenberg said about a quantum paper, “so bad not even wrong”.

  9. Truly difficult to comprehend ~ there is so much crap publishing in the field of climate change. Peer review means does it pass the green smell test.

  10. These are sim players messing around with scenarios. The use of actual data is tangential to their purpose so they torture it until it spits out something even if it’s wrong. Remember what was said about torture, prisoners end up telling you what you want to hear.

    Climate scientists have turned into data hounds working with computers and numbers with no idea what the numbers really mean. Someone needs to use AI to develop software that questions people using stats software about the assumptions underlying each tool in order to make sure they are meeting the underlying requirements.

  11. ” obviously flawed papers like L&C 2020″
    This is not any sort of social science. The assertions are not flawed. They are not faulty.
    They are wrong.
    Wrong, Wrong, WRONG. Not faulty, not flawed, Wrong.

    There, now that we have that sorted out…..
    You Do Not smooth a data set then do a regression on what is left.
    “a procedure that is normally frowned on.”
    Frowned on and then some.

    “this approach lessens the effects of non-climatic factors and random noise”
    Ohh My Gawd. Speechless.
    Not faulty, not flawed, just plain Wrong.

      • Criminal, but here in canada CNR

        “Criminally not responsible”

        The poor dears didn’t mean to butcher all those people (data) they simply don’t know right from wrong

        The social science take on crime, coming soon to a country near you

  12. This appears to be Mother Nature displaying her anger….again. She is becoming less tolerant of man’s CO2…this is clear…it’s not nice to make Mother mad…better stop it….or else…..she knows who’s been bad….she even knows your address…..expect more ‘caines….bigger and longer lasting until man learns his lesson.

  13. The news in Sweden was interested in the hurricanes in US and this Nature article.
    They are very keen to pic up all news about the coming catastrophic climate.
    I hope they can be informed of this development in the research. (Erica Bjerström SVT)

  14. “…the average SST difference is 1.8 K in this case,”

    Umm, a difference would be 1.8 C

    1 K is 1 C above absolute zero.
    0 C is the freezing/triple point of pure water.
    1 C is 274 K.
    -1 C is 272 K.
    K is the scale, C is the unit.
    The English version is R-ankine and F-ahrenheit.

    Elsewhere tells me that K & C are interchangeable.
    That is just flat wrong.
    It’s similar to psia, psig, psid and vacuum.

    You had better understand the differences and which one goes in the equations or your answers will not only be wrong, but fatally so.

    Is this sloppy science just caving to millennials?

    • …the average SST d i f f e r e n c e is 1.8 K in this case.
      This is a true sentence. Your comment has no content at all.

      • An SST difference from 0 K to 1.8 K means a solid sea surface content not liquid.
        That would certainly make for a well defined surface.

        The distinction I make might not mean much to a pimply faced twenty something PhD camped at a keyboard in some academic cube farm, but in the world of the real it matters.

        • Take care, you write absolute BS! The SST difference 0f 1.8 K is not only the difference between 0k and 1,8K. The difference of 1.8K is also the difference between 302 K and 300,2 K as it’s in the case in question! Stop kidding!

    • Wen I was told while ago that the difference of °Kelvin values gets the unit K (without deg),
      I was puzzled. After pondering a couple of years, I found out that this is probably because K can be negative, not so °K. In that sense, Fig.1 is not correctly labelled.

    • Indeed, the November Hurricanes develop when the SST is near 300K ( =about 27°C). They should have a tau of far below 10 hours, according zi fig.1.
      In the real live they had a tau of up to 57 hours. The first time I read L&C 2020 I thought it was a joke-paper. It wansn’t. 🙁

  15. If you torture data enough with statistical methods supplied in computer libraries you can prove anything you like. Most “scientists” do not bother reading up on the proper use of any statistical method – I mean the computer must know what its doing, right?

    This is rank amateurism – so about average for climate-style “scientists”. Mann would be proud of them.

Leave a Reply

Your email address will not be published. Required fields are marked *