Record temperatures and random chance

This is something I have been thinking about for a while but has been brought to a head by this article: Extreme weather in 2019 broke over 120,000 records in US: report.

Extreme weather incidents in 2019 set more than 120,000 daily records across the U.S. according to data from the National Centers for Environmental Information.

The 122,055 records set include record daily high and low temperatures as well as record rain and snow.

CNN reported on the record events, noting heat waves that hit Alaska and raised temps to 90 degrees Fahrenheit, a winter storm that brought snow to Hawaii, and rain that flooded the Midwest and hurt crops.

The report comes after a year of devastating climate events, from Hurricane Dorian to record melting of ice sheets. Other countries also saw record-breaking heat and cold, and raging fires severely damaged the Amazon rainforest.

Note that these records are not just about heat, as in global warming, but about cold, and rain and other climate events that ended up breaking whatever had been the previous record. My question therefore is this, and I will stick to just record high temperatures.

If there are 365 days in a year, and a weather station has been in existence somewhere for say 120 years, how many days during any normal year will turn out to produce a record high temperature just by random chance?

As far as I can tell, there should be approximately three such record temperatures a year for each and every weather station. And if the station has only been running for 60 years say, then there should be around six days during the year that should produce a record high temperature. A new station should have 365 records a year and after a year 183. Indeed, the same temperature on any given day would be described as equalling some previous high and be seen as evidence of a warming trend and be described in that way.

This looks obvious to me. So is this right or wrong, and if it’s wrong, why is it wrong?

And of course, if it is right, then these weather cranks should shut up every time some record is broken somewhere since a new record is in itself meaningless.

This entry was posted in Global warming and climate change policy. Bookmark the permalink.

49 Responses to Record temperatures and random chance

  1. stackja

    The MSM relies on meaningless weather records cranks.
    Recently we had Greta.
    More credible, of course, is Jo Nova.

  2. Bruce of Newcastle

    The report comes after a year of devastating climate events,

    What rubbish.
    One graph blows this propaganda away.
    The data says extreme weather is getting less devastating.
    Many similar studies of accumulated hurricane energy, tornado numbers, rainfall and the like all support that, if anything, global warming causes better weather. But that message is very inconvenient to the climate activists’ cause.

    The graph is from this article:

    Why Climate Advocates Need To Stop Hyping Extreme Weather (14 Dec)

  3. Sydney Boy

    Bruce,

    Another graph: BOM cyclones

    I love trotting this one out to all the climate catastrophists who like to claim cyclones are becoming more frequent.

  4. Petros

    Yes but multiply by every weather station where there is a new record and voila a big scary number.

  5. yarpos

    Notice that records are set for maximums and minimums yet CNN only talks maximums. The distribution would be interesting.

    Modern weather stations are capturing temps that occur instantaneously vs old mate wandering out to look at the thermometer periodically, so snagging tecords should be easy.

    When looking at absolute records globally heat records are still generally a far way in the past , while cold records are relatively recent. Just an odd counter narrative observation.

  6. Biota

    But weather is not a purely random event with a standard distribution. ENSO, AMO, PDO, Milancovitch cycles, Indian Ocean Dipole to name some influences. Then there is local geography and the well documented siting issues such as beside aircon exhaust and tarmac. This is why weather forecasts don’t do so well beyond 24 hours.

  7. Kneel

    “This looks obvious to me.”

    You are correct, much in the same way that a “1%” event – also described as a “100 year event” – should happen, well, 1% of the time. That’s more than once a century.

    And too, don’t forget that temps, rainfall etc aren’t random – they have some amount of auto-correlation, so that we tend to see a multi-day “heat wave” or “cold spell”, days of rain, droughts etc, and almost never see a record hot day followed by a record cold day at the same place (which, were it truly random, we would see more often). Even a small amount of auto-correlation changes the stats significantly – both the “result” and the required methods.

    Consider too that many of the weather variables are not independent – in Aus at least, drought causes high temps, not the other way around, for obvious physical reasons. The data also seem to indicate that the droughts are broken by floods, with longer droughts broken by bigger floods.

    I’m not convinced that most of these weather variables are Gaussian distributed either – many intuitively would be otherwise, maybe binary, certainly asymmetric “tails”. So OLS may be inappropriate as a trend indicator.

  8. JohnJJJ

    Every day I break a record of some kind in my life. Yesterday it was the red wine/white wine/fruit cake ratio.

  9. If Carbon Dioxide was warming the planet, then it’s the minimum temps that should be breaking records more so than maximum temps (according to the Gold Standard of Climate Science, the IPCC assessment Reports).
    Most of the records should be at or near the poles and at winter. Much of the rest of the records should be at the higher altitudes during winter time.
    The fact that the above isn’t happening and hasn’t happened despite rising levels of CO2 in the air disproves the theory that CO2 drives climate. The theory has been disproved many times.
    But that’s neither here or there because $$$trillions$$$ are at stake, not to mention a NWO and central control.
    It’s a scam pure and simple.

    Merry Christmas you lot.

  10. Professor Fred Lenin

    Watched Marine le Pen being interviewed by a smart ass Engish lefty ,who introduced her as leader of an extreme right group ,he spewed bullshit poll results to create a gotcha moment She laughed and said
    “You are English and believe the polls ? “Hahahaha ,squashed him like a fly.
    Thats what you do with climate figures , how many lying records were broken in this period ? Much more relevant question .

  11. Crossie

    The 122,055 records set include record daily high and low temperatures

    When I was in high school we were taught that the tilt of the earth on its axis influenced winter and summer temperatures. The greater the angle of the tilt the higher or lower the summer and winter temperatures respectively.

    The only things school children are taught these days is respect for Dreamtime and LGBQWERTY so no wonder any claim about weather and climate, no matter how erroneous, sounds legitimate to them. They have been conditioned to accept lies as truth and to stay within the herd at all costs.

  12. Just as the survival rate falls to over a long enough timeline, the variance of a time series will increase until t -> ∞.

    This is so obvious to anyone who is competent in modelling time series data.

    “The maximums and minimums are breaking records more and more”

    No. Nothing is happening.

  13. Roger

    I understand overseas tourist numbers visiting the GBR are down because of the alarmist hype that the reef is dying.

    I wonder what effect the “Australia is burning” headlines will have on overall tourist numbers?

    The tourist sector ought to be at the forefront in exposing the lies. Where are they?

  14. Crossie

    oger
    #3272454, posted on December 26, 2019 at 10:05 am
    I understand overseas tourist numbers visiting the GBR are down because of the alarmist hype that the reef is dying.

    I wonder what effect the “Australia is burning” headlines will have on overall tourist numbers?

    The tourist sector ought to be at the forefront in exposing the lies. Where are they?

    We have just come back from a European holiday and a transatlantic cruise. These were the comments were were getting most often – the Reef is dead and the rest of the continent is burning that there is no point in going. I would try to explain about both the Reef and the fires but mostly got dubious looks.

    The tourist sector is doing its best but the nongs in charge of Tourism Australia are not very effective or not interested in telling the truth because then they will be called bad names on Twitter.

  15. Leo G

    Extreme weather in 2019 broke over 120,000 records in US: report.

    Not true. A “record weather event” is not synonymous with an “extreme weather event”. Extreme events are usually defined by the frequency distribution of the parameter, and normally at the upper 5% level.

  16. Crossie

    How could the media and (potential) overseas visitors not think the Reef is at least in trouble if not dead when Malcolm gives some unknown entity 400 million dollars to supposedly study its demise?

    Has anyone checked recently what is being done with that money? How is it being spent? Who is doing what with it?

  17. struth

    The Australian tourism industry has , for many years now, been controlled by government.
    Accredited by the very same government to operate means no rocking the government’s climate hoax boat.
    Tourism Australia has done more to kill tourism in Australia than any other organisation.

  18. classical_hero

    The bs about record ice sheet melt is rubbish. You can see from this graph that 11-12 was the greatest in recent times.
    https://images.app.goo.gl/BBNogYPjY4z8iFB76

  19. Arky

    So, when they really should be worried is if they could say; “This year no minimum records were broken”.
    That would be a worry.
    But they can’t. Because the temperature really isn’t doing much, certainly not what their models said it would.

  20. Entropy

    Modern weather stations are capturing temps that occur instantaneously vs old mate wandering out to look at the thermometer periodically, so snagging tecords should be easy

    To be fair, they had this thing called max/min thermometers.

  21. Tel

    To be fair, they had this thing called max/min thermometers.

    Yes … mechanical devices that depended on the expanding Mercury to push a little flag along the tube. Therefore they would respond in several minutes to a change in temperature, but not in a few seconds like what an electronic thermometer can do.

    Mind you … some of the most extreme measurements ever seen in Australia were a century ago, so it makes you wonder just how high the max would have been if they had electronic thermometers back in 1909.

    http://joannenova.com.au/2014/08/hiding-something-bom-throws-out-bourkes-hot-historic-data-changes-long-cooling-trend-to-warming/

  22. Roger

    Tourism Australia has done more to kill tourism in Australia than any other organisation.

    AGW alarmists are poised to give them a good run.

  23. I_am_not_a_robot

    The global and local temperature records are the product of false precision.
    Even the most modern max-min mercury or alcohol thermometers claim an accuracy limit of ± 0.2°C, the older Six’s versions ~ ± 0.5°C.
    “… in science and engineering, convention dictates that unless a margin of error is explicitly stated, the number of significant figures used in the presentation of data should be limited to what is warranted by the precision of those data …” (Wiki).

  24. I_am_not_a_robot

    The global and local temperature records are the product of false precision.
    Even the most modern max-min mercury or alcohol thermometers have an accuracy limit of ± 0.2°C, the older Six’s versions ~ ± 0.5°C.
    “… in science and engineering, convention dictates that unless a margin of error is explicitly stated, the number of significant figures used in the presentation of data should be limited to what is warranted by the precision of those data …” (Wiki).
    Besides the fact that many historical records have been incrementally ‘adjusted’ multiple times, the global record as presented to the public showing no uncertainty.
    As a result the public fall victim to precision bias viz. thinking greater (false) precision implies greater accuracy.

  25. I_am_not_a_robot

    The global and local temperature records are the product of false precision.
    Even the most modern max-min mercury or alcohol thermometers have an accuracy limit of ± 0.2°C, the older Six’s versions ~ ± 0.5°C.
    “… in science and engineering, convention dictates that unless a margin of error is explicitly stated, the number of significant figures used in the presentation of data should be limited to what is warranted by the precision of those data …” (Wiki).
    Besides the fact that many historical records have been incrementally ‘adjusted’ multiple times, the global record as presented to the public showing no uncertainty.
    As a result the public fall victim to precision bias viz. thinking greater (false) precision implies greater accuracy.
    (I could illustrate some of those points but links are not working for me).

  26. Professor Fred Lenin

    The workers in America are slaves to the capitalists Soviet workers are the masters , remember that ?
    Yeltsin and Scheverdnadze believed that untill they went to a US supermarket and saw the slave opressed workers buying trollysvfull of food Soviets never saw in a lifetime and a truckdriver buying a cake fit for a politburo member .
    Keep repeating the crap and supress the truth and everyone will believe it.

  27. Leo G

    Even the most modern max-min mercury or alcohol thermometers have an accuracy limit of ± 0.2°C, the older Six’s versions ~ ± 0.5°C.

    Distinctions between precision and accuracy are very blurred for meteorological meaurements of temperature.
    What is the “true value” of temperature for a station used to monitor weather or climatic variation when the station is representing a large geographic area? If you surveyed within such an area (even for such an area of a few hundred sq km in a city) you might find a 95% certainty range within, say, +/- 5 degree C. This suggests that the station for that area might have a temperature precision of +/- 0.2 degree C but an accuracy of +/- 5 degree C.
    The new age meteorologist doesn’t survey the representative area, though, but uses a distant station (which likely has the same accuracy problem) and uses historical series from that as the basis for “correcting” the accuracy.
    The “true value” of temperature in that context is an ill-defined notion.

  28. I_am_not_a_robot

    On record temperatures and the temperature record in general for instance according to the BoM the highest temperature recorded in Mildura was 50.7C on 7 Jan 1906.
    I don’t know if that was recorded in Celsius or Fahrenheit in either case (50.7C converts to 123.26F) reading less than a one degree increment in 1906 would have been dicey more likely impossible.
    It’s ridiculous.

  29. 2dogs

    Has there ever been an explanation of the physics behind how incrased CO2 causes more extreme weather?

    The only discussions I can find on this topic are from the late 1990s and early 2000s that mention “polar amplification”, a process by which increased CO2 would result in less extreme weather.

  30. Tel

    The new age meteorologist doesn’t survey the representative area, though, but uses a distant station (which likely has the same accuracy problem) and uses historical series from that as the basis for “correcting” the accuracy.
    The “true value” of temperature in that context is an ill-defined notion.

    As I’ve said elsewhere … the surface of the Earth never reaches perfect thermodynamic equilibrium, but for that matter very few things do, so it’s a question of making an ESTIMATE of what the result would look like if it was allowed to reach equilibrium. The notion of what you are attempting to do is well defined … but the methodology of how to achieve that is a bit of a personal opinion.

    https://s3.amazonaws.com/jo.nova/guest/aust/bom-audit/gillham-chris/2019/very-hot-days/australia-hot-days-60.gif

    When I look at that chart and the systematically adjusted past temperatures that “just happen” to end up increasing the overall warming slope, I find it incredibly difficult to think of any way that could be legitimate. That’s not to say no possible adjustment could ever be legitimate but it’s fraught with difficulty to claim that you have improved the data quality be tweaking it. The chances are overwhelming that you will end up building in your own personal beliefs.

  31. Bruce of Newcastle

    Even the most modern max-min mercury or alcohol thermometers have an accuracy limit of ± 0.2°C

    Hehe, in undergrad chemistry I was taught to read an alcohol or mercury thermometer to 0.1 C, based on 1 C gradations. But if I wanted to do an experiment I’d get a handful of thermometers out of the draw in the lab and dump them into a beaker of boiling water. Then I’d pick out the one reading 100 C and use that (since most experiments were temperature controlled around 95 C or so).

    The bunch of thermometers would read anywhere from 98 to 102 C in the beaker of boiling water. Scientific-grade thermometers. Thermocouples would be the same, but you could calibrate their electronics.

  32. I_am_not_a_robot

    @ Tel 5:16 pm,
    An old Soviet joke: “The future is certain; it is only the past that is unpredictable”; practitioners of CC™ ‘science’ have taken it to a new level.
    As Prof Humlum @ climate4you ruefully comments in his ‘temporal stability of temperature’ section that obviously the past doesn’t change so an unstable record cannot be correct all the time and ‘a temperature record which keeps on changing the past hardly can qualify as being correct’.

  33. Mark M

    2dogs
    #3272689, posted on December 26, 2019 at 4:59 pm
    Has there ever been an explanation of the physics behind how incrased CO2 causes more extreme weather?

    >> They have no idea. Here is one example:

    2017: It is the second consecutive year Australia’s cyclone count has been chronically low.

    But he said “basic physics” governed that [global warming] would increase the intensity of cyclones in the future.

    “For each degree of warming there is about seven per cent more moisture in the atmosphere, so there is more moisture now than, say, 50 years ago,” he said.
    “In reasonably simple terms, the warmer it gets, the more steam comes off the water and the warmer the atmosphere, the more moisture it can hold.”
    He said it naturally followed that the amount of rain each cyclone brought would be higher, and thus the cyclone more intense.

    It does not, however, explain this season’s anomaly.

    “Being perfectly honest, [global warming] is a factor in most of our climate science these days but in terms of tropical cyclones you couldn’t put this season down to [global warming],” he said.

    https://www.news.com.au/technology/environment/climate-change/cyclone-blanche-is-latest-to-cross-land-in-second-consecutive-quiet-season-in-australian-history/news-story/220bd07cbd24d1db32cfd2175d3ec2ac

  34. David Brewer

    “Has there ever been an explanation of the physics behind how incrased CO2 causes more extreme weather?”

    Three answers:

    1. The explanations, such as they are, are simply that this is what comes out of climate models.
    2. There has never been an “engineering grade” explanation of the physics of global warming, full stop. See here:

    …my recommendation has consistently been that people who are worried about the impact of increased CO2 need to provide an “engineering quality” exposition of how doubled CO2 leads to (say) 3 degree C and thence to problems….

    Such an exposition would probably be 1200 or 2000 pages, not 10 pages….The main area of scientific uncertainty is in cloud feedbacks. In an engineering quality report, there might be several hundred pages on this topic, describing precisely what is known and what is not known and how the scientific uncertainties might be reduced. In AR4 [the IPCC 4th Assessment Report], this important topic was covered in less than two pages.

    I’ve raised this issue with climate scientists on a number of occasions. To date, I haven’t encountered a single climate scientist that remotely comprehended what was missing, while professionals from other fields often understand the sort of document right away…

    Typically, if a climate scientist responds, they provide a link to some little article on climate sensitivity that is not remotely equivalent to an engineering quality exposition, with the citation merely showing that the climate scientist doesn’t have a clue about the form of communication employed in the professional world….

    Unfortunately, some climate scientists – Gerry North for example – have even sneered at the idea of such an exposition.

    3. See also “The Extreme Weather Meme” here.

  35. Ian of Brisbane

    When you change the historic data to allow new records to be set, they will be set.

  36. mem

    In the original IPCC membership make-up statisticians and astrophysicists were excluded from participating. Yet these were possibly the most valuable areas of expertise to include if you were really interested in understanding long term climate shifts and contributing components. I have come to the conclusion that it has never been about science or truth. It is about progressing a scam to hoodwink the public, influence politicians, gaining political power. and line the pockets of a bunch of international financial racketeers. At the same time it benefits China and some other countries that make most of the wind farm and solar components whilst vaguely espousing the Paris target claptrap. Australians are being treated like, and acting as, patsies whilst being duped by bodies such as the Australian Climate Council, the BOM and its propaganda arm the ABC. The climate has not changed dramatically in the past thirty years and all the finger wagging and hyperventilating about a half of a degree of temperature one way or the other, fails to impress me. My belief is that most thinking Australians understand that it is clap trap yet their votes have been hijacked by politicians that refuse to listen to their electorates.

  37. Lutz

    If I read the NOAA data right for the last 365 days there were more days with a record low high temperature and not much between record low max and min. It would also be nice to know by what measure these were declared high or low. Usually we are talking about minuscule difference of 0.1 or less.

    High Max High Min Low Max Low Min
    18369 27091 23701 18111

  38. mem
    #3272845, posted on December 26, 2019 at 9:33 pm

    In the original IPCC membership make-up statisticians and astrophysicists were excluded from participating.

    Scandalous.

  39. StubbornlyRational

    22,055 records were set. Over 365 days. This means approximately 334 records per day. That might sound impressive to someone who is numerically illiterate, as the author of “The Hill” article apparently is. But note that the actual page on the NOAA website contains the following disclaimer:

    =======
    This tool provides simplistic counts of records to provide insight into recent climate behavior, but is not a definitive way to identify trends in the number of records set over time. This is particularly true outside the United States, where the number of records may be strongly influenced by station density from country to country and from year to year. These data are raw and have not been assessed for the effects of changing station instrumentation and time of observation.
    ========

    Suppose 334 records were set yesterday. What, precisely does that mean? Note, we have no idea from the Hill article OR the NOAA website how many stations there are! And note, all records for highest high temp, highest low temp, lowest high temp, lowest low temp, high precipitation, high snow depth, etc are lumped into the same pile, EVEN THOUGH THEY MEAN COMPLETELY DIFFERENT THINGS!

    It gets even funnier. Suppose that there has been no trend at all in the weather at a particular station, and the station has been in business for 50 years (30 is actually the NOAA minimum.). Then 2% of such stations will show an “all time record” on any given day. But there are at least 6 different “record categories”.

    How many stations are there? We aren’t told!!! We’re just given a raw number. So it may well be (and probably is) the case that 120,000 “new all time records” is not a surprising number at all. We can see that, if there are 1000 statiions being tracked, 334 records per day is actually not at all surprising *if our climate is completely stable* and there is no climate change at all!

    The fact that The Hill author Brooke Seipel fails to include any of the fundamental information required to evaluate the 120,00 number indicates either (a) she’s terminally stupid and numerically illiterate, (b) completely disdainful of her audience, or (c) both. She does manage to get in a plug for Greta Thunberg, named “person of the year” by a defunct magazine that falsified its own prize for the sake of climate ideology.

    As I mentioned above, each daily record means something different. Consider just two. The “High Max” record was set 18420 times, while the “Low Max” was set 23646 times over the past 365 days. This means that new record highs were detected less frequently than daily highs that were the coolest ever. Does this sound like “global warming”? Not really, but, as the NOAA points out, you really cannot tell from these data.

  40. Retired stats geek

    I am sympathetic with your point of view but I don’t think your math is correct. If you could share your reasoning that would help.

    Let’s focus solely on temperature highs. While daily temperatures are not independent, if they were, the probability that a new days temperature was higher than the previous N days max temperature is 1/N. In the case of 120 years of daily readings N=43,800 and 1/N=0.000023. Or over a year, the number of days exceeding the max would be approximately 1/120=0.0083.

    Again daily temperature readings are hardly independent, so maybe your analysis took these correlations into consideration? The closest I can come to your “3 days per year” is to make a little more reasonable assumption that annual maximum temperatures are independent so that the probability that this years max temperature breaks the record for the previous 120 years is 1/120. Now if you multiply this by 365 you get 3.03 days but this isn’t kosher since the uncorrelated random variable we are concerned with is the annual max temperature and not individual days. It’s easy to see why multiplying by 365 is wrong since it assumes that the probability of exceeding the annual max is the same on a winters day as a summer day.

    Can you share how you arrived at 3 days per year/

  41. mem

    I hear from the loony leftists and finger wagging climate preachers that the Boxing Day Cricket test may have to be changed to later in the year commencing next year because of the catastrophic warming caused by climate change. I am sitting here with a jumper on at 10.00 am in the morning of the 27 December wondering how stupid they are to imagine that by this time next year it would be so hot as to destroy our cricket These people are nuts!

  42. Leo G

    Every day I break a record of some kind in my life. Yesterday it was the red wine/white wine/fruit cake ratio.

    That’s the climacteric cycle for you.

  43. Leo G

    “This year’s extreme weather broke more than 120,000 daily records across the US — from temperatures to precipitation and snowfall records, according to the National Climatic Data Center.” – CNN

    In fact, the records “broken” were historical same-day-of-the-year records not daily records and were not necessarily extreme weather events (which in the case of temperature would imply statistical historical all-days-of-the-year and/or successive days criteria).
    Of course, CNN itself often reports like a broken record.

  44. Zatara

    “This year’s extreme weather broke more than 120,000 daily records across the US — from temperatures to precipitation and snowfall records, according to the National Climatic Data Center.” – CNN

    Records are made to be broken. Celebrate the human spirit that enables us to keep doing so.

    Even if we have nothing to do with it.

  45. JC

    “This year’s extreme weather broke more than 120,000 daily records across the US — from temperatures to precipitation and snowfall records, according to the National Climatic Data Center.” – CNN

    lol.. okay, so it was a little rough but Lomborg’s point remains. There are much less weather related deaths than even 80 years ago.

    How dare he!

  46. David Brewer

    Retired stats geek: The article is talking about daily records – the record for January 1, the record for January 2 etc. So the probability of breaking a daily record, for somewhere with 120 years of records, is 1/120, or 0.0083, EVERY DAY.

  47. Retired stats geek

    Thanks David that clears it up. In retirement I find that I don’t read things as carefully as I once did.

Comments are closed.