Friday, December 21, 2012

The latest from "Trend Central"



This is fun. Commentor Hoo is diligently calculating to show that the graph of world temperature over the past 16 years (above) shows a warming "trend".

Hoo:
And with that, I will repost the numbers for the temperature trend obtained from HadCRUT4 from various starting dates until today: 
1995–2012: +0.109±0.119 °C/decade
1996–2012: +0.107±0.131 °C/decade
1997–2012: +0.058±0.136 °C/decade
1998–2012: +0.052±0.153 °C/decade
1999–2012: +0.095±0.162 °C/decade
2000–2012: +0.056±0.179 °C/decade
We are talking about the time dependence of the average global temperature anomaly ΔT as a function of time t. Specifically, ΔT(t) = ΔT(t0) + β(t−t0). Here t0 is a starting point in time and β is the linear slope. 
... the HadCRUT4 data set (among others) yields a positive β even if you cherry-pick the starting point? The values of β for various starting points are quoted in my previous comment.

Hoo's evidence for global warming in that graph:

"ΔT(t) = ΔT(t0) + β(t−t0)".

:-/


Global warming science in real time, as it happens.


24 comments:

  1. Dr. Egnor,

    I indeed calculated the trend for the data shown in that specific graph and I found a positive warming trend that I quoted in one of the previous threads, through which you are running like a mad hare.

    Dr. Curry agrees with me that the temperature anomaly shown in the figure has a positive warming trend, albeit not as large as 0.2 degree per decade.

    You are the only one here who insists that there is none.

    You can go ahead and crunch the data yourself, as I did. Use your considerable statistical skills, Dr. Egnor.

    Hoo

    ReplyDelete
  2. Dr. Egnor,

    Here is Dr. Curry again, in case you forgot.

    Has there been any warming since 1997 (Jonathan Leake’s question)? There has been slight warming during the past 15 years.

    Do you agree with her?

    Hoo

    ReplyDelete
  3. Hoo,

    It's futile arguing with Michael. He doesn't understand mathematics. He doesn't understand science. He's the perfect example of an intelligent person who has gone into a field (medicine, and neurosurgery in particular) which relies on rote learning and following treatment protocols, without encouraging much in the way of developing critical reasoning skills.

    To which is added the directed reasoning bias produced by his conservative and religious ideologies.

    AGW is based on the well known and well understood physical properties of greenhouse gases, which have been known since the 19th century, set on a mathematical basis by Arrhenius over a hundred years ago.

    He attacks temperature graphs because he doesn't understand science, or doesn't want to.

    ReplyDelete
  4. bachfiend: ...he doesn't understand science...

    To be precise, Dr. Egnor doesn't understand YOUR BRAND OF science which smells like s**t here in Canada from down under Australia

    ReplyDelete
    Replies
    1. Pepe,

      Idiot. There's only one brand of science. And then there's Michael's motivated reasoning, which starts with the conclusion, and looks for anything supporting it. You should get your sense of smell tested, or at least change your diapers a bit more often.

      Delete
  5. I would assign Hoo the task of squaring the circle. He has talent for such endeavour!

    ReplyDelete
    Replies
    1. Thank you for your consideration, M. Pépé. What was your profession before you retired due to old age?

      Hoo

      Delete
  6. 1995–2012: +0.109±0.119 °C/decade
    1996–2012: +0.107±0.131 °C/decade
    1997–2012: +0.058±0.136 °C/decade
    1998–2012: +0.052±0.153 °C/decade
    1999–2012: +0.095±0.162 °C/decade
    2000–2012: +0.056±0.179 °C/decade

    In every case the error range exceeds the measured value. So what, exactly, is your point, Hoo?

    The signal you allege is indistinguishable from the noise.

    ReplyDelete
    Replies
    1. Mr. Matteo,

      May I ask what is your degree of familiarity with statistics?

      Or you could perhaps simply discuss the implications of a data analysis showing a number +0.107±0.131 °C/decade at 2 standard deviations. According to this, what are the probabilities of warming and cooling, respectively? Is there a significant difference in the probabilities? If you were a betting man, what odds would you put on a cooling trend?

      This are fairly simple questions that anyone who has a basic understanding of statistics should be able to provide.

      Thank you,

      Hoo

      Delete
    2. Hoo:

      Matteo is looking at a graph that obviously shows no significant warming. He is a sensible intelligent guy, and he points that out.

      Your reply is arrogant and dissembling. The fact is that no honest reputable scientist would take that data set and claim significant warming. You can piddle with statistics all you want-- it makes it worse, not better, because it demonstrates quite clearly how ideologues like you in the global warming movement manipulate the numbers to mislead the public.

      There ain't no significant warming in the data graphed above.

      Are we to conclude that the whole Potemkin village of AGW is based on the same dodgy statistics-- taking data that obviously don't show any consistent changes to speak of and by cherry-picking statistical analyses coming up with the breathless announcement-- THERE'S A TREND. WE'RE RIGHT!

      @Pepe, Matteo, and Torch: now you can see even more clearly why the Climategate email crowd fought frantically for years to evade Freedom of Information Act requests to release their actual data, even breaking the law and conspiring to destroy the data rather than release it.

      They had been massaging the data with statistical tricks, and they understood that if the public and skeptical scientists ever saw the actual data (like the graph above) they'd realize that AGW was fabricated out of whole cloth.

      Delete
    3. Dr. Egnor,

      I have asked you several times already to state whether you agree with Judith Curry's assessment. Here are her words:

      "Has there been any warming since 1997 (Jonathan Leake’s question)? There has been slight warming during the past 15 years."

      Do you agree with her assessment?

      Hoo

      Delete
    4. Dr. Egnor: Are we to conclude that the whole Potemkin village of AGW is based on the same dodgy statistics-- taking data that obviously don't show any consistent changes to speak of and by cherry-picking statistical analyses coming up with the breathless announcement-- THERE'S A TREND. WE'RE RIGHT!

      Dr. Egnor,

      You have been reluctant to crunch the data yourself. You merely look at a graph with a naked eye and declare no warming trend. But an eye is easily mistaken. You have to work with numbers. This is why we have statistical analysis.

      There is nothing sinister in statistical analysis. If it scares the shit out of you, do not fear. I hold your hand and guide you through the numbers in this comment.

      Enjoy!

      Hoo

      Delete
  7. Hmmm...the temp in 2012 is exactly what it was in 1997. Is there really more to understand than that? There were ups and downs, but for every up there was an equal down.

    I may be a layman but that's how I see it.

    The Torch

    ReplyDelete
  8. The Torch,

    The significance is that it started with an El Niño, which causes warmer than average global temperatures and finished with a La Niña, which would be expected to produce cooler than average global temperatures (and a La Niña that is warmer than previous la Ninas).

    What it means is that; 1. You can't take such a short period, particularly when you 'cherry-pick' and start 'warm' and finish 'cool'. Another reason is the the standard deviations are greater than the variations in the trend. 2. There are other factors driving global temperatures, some producing a short term cooling effect.

    ReplyDelete
    Replies
    1. bach:

      The 'cherry-picking' point can be applied to pro-AGW arguments in spades.

      From 1940 to 1980 there was cooling, not warming. Everyone agrees with that. So the period of warming that set off this hysteria-- 1980-1996-- was a period of emergence from a cooling period, and of course it will show warming. Furthermore, the entire period since 1800 has been an emergence from the Little Ice Age-- a natural period of cooling followed by a natural period of warming.

      In your global warming scare tactics, you've cherry-picked years in which we just began to emerge from natural cooling, and built your hysteria on data manipulations of the expected warming.

      The whole global warming panic is built on cherry-picking "start cool and finish warm" of natural variation, and the interval of warming is now the same as the interval of stasis (15 years). If the stasis is too short to be meaningful, why is the warming long enough to presage the apocalypse?

      Delete
  9. Mr. Torch,

    Hmmm...the temp in 2012 is exactly what it was in 1997. Is there really more to understand than that? There were ups and downs, but for every up there was an equal down.

    Let us take a closer look. The data shown in he opening post are taken from HadCRUT4 set, which can be found here.

    Click on the monthly Global (NH+SH)/2 time series and you will see data in the text format. Each line shows the year and month; the next number is the global temperature anomaly.

    In 1997, we have (Jan through Dec):
    0.206, 0.323, 0.346, 0.287, 0.288, 0.400, 0.369, 0.436, 0.475, 0.553, 0.490, 0.507.
    In 2012 (Jan through Nov):
    0.288, 0.208, 0.339, 0.525, 0.531, 0.506, 0.470, 0.532, 0.515, 0.524, 0.512.

    What do we make of these numbers? First, there are a lot of fluctuations. The numbers range from 0.2 to 0.5 in both years and change in a more-or-less random fashion from month to month. These fluctuations reflect random natural fluctuations—for the most part.

    But look more closely and you can spot a trend. In 1997, only two months had a temperature anomaly above 0.500. In 2012, seven had. You can also calculate the average numbers and get 0.390 for 1997 and 0.450 for 2012. The change is +0.060 degrees over 15 years, or +0.040 degrees per decade.

    This is not a very sophisticated statistical analysis, but it gives you an idea. A more serious analysis will include not just the end years 1997 and 2012, but all of the data in between. Roughly speaking, a statistician draws a straight line through the data points, finds how much the points deviate from the straight line and then finds the best straight line, with the smallest amount of deviation. This is how the numbers cited in the opening post were obtained. From 1997 to 2012, the best trend for the temperature anomaly was +0.058 degree per decade, not too far from the result I obtained here from direct comparison of the two years.

    The uncertainty (standard deviation σ=0.068) reflects the level of noise. We can discuss statistical implications of that if you are interested.

    Hoo

    ReplyDelete
    Replies
    1. @Hoo:

      It's pretty funny that you beat this dead horse of statistical milking-- every month will show either a slight increase or a slight decrease, and the real question is whether the net increase or decrease is meaningful.

      It is not.

      By the way, how does the Hockey Stick look now, when you graft this period of stasis?

      How does deletion of proxy data that does not fit the warming scenario (the Nature trick to hide the decline) beginning in 1960 affect your confidence intervals? Is deleting unfavorable data just at the inflection point of a graph consistent with meticulous statistical analysis? Or do you only use statistical manipulation when the resulting "trend" favors your hypothesis, and discard rudimentary scientific integrity when it is to your benefit?

      What is the statistical correlation between the fact that the period of stasis from 1996 to now corresponds to a very large epoch of CO2 release (a third of the total human CO2 emissions)? Why was most of the warming from 1800 to now prior to the largest CO2 rates of emission, which only became substantial in the mid-20th century, and now do not correlate with substantial warming (from 1996 to 2012)?

      What influence does the constant variation in the number and location of temperature measuring stations (many cold Siberian stations went off-line in the 1980's with the decline of the USSR) and the fact that the Southern Hemisphere had very few temperature measuring stations until relatively recently, thereby changing substantially the sampling of climate around the world.

      What statistical manipulations have you performed to adjust for these major changes in the sources of your data?

      Do you stand by your Hockey Stick graph? Why isn't it still used by the IPCC? Would you still color the period 1996-2012 bright red to scare the public, or pick a neutral color, reflecting the stasis?

      Delete
    2. Dr. Egnor:

      It's pretty funny that you beat this dead horse of statistical milking-- every month will show either a slight increase or a slight decrease, and the real question is whether the net increase or decrease is meaningful.

      It is not.


      Dr. Egnor,

      The first paragraph in your comment is a very thoughtful assessment of the role of random fluctuations, with which I fully agree.

      The second paragraph is a naked assertion not backed up by anything resembling an argument. This is highly unfortunate.

      Note that I have patiently explained, in very simple terms, how one arrives at the conclusion that there has been a warming trend even in the last 15 years. I have provided a back-of-the-envelope calculation showing the presence of the warming trend. Anyone can do a more sophisticated analysis using a statistical calculator and see that there is a warming trend.

      Judith Curry says that there has been a warming trend.

      Yet, despite all of this, you stick fingers in your ears and shout "La-la-la, I can't hear you! Global warming is a fraud!"

      You look like a dork, Dr. Egnor.

      Hoo

      Delete
    3. Dr. Egnor: How does deletion of proxy data that does not fit the warming scenario (the Nature trick to hide the decline) beginning in 1960 affect your confidence intervals? Is deleting unfavorable data just at the inflection point of a graph consistent with meticulous statistical analysis? Or do you only use statistical manipulation when the resulting "trend" favors your hypothesis, and discard rudimentary scientific integrity when it is to your benefit?

      Dr. Egnor,

      You mischaracterize the reason why tree-ring proxies are discarded. Not because they do not fit the global-warming theory, but because they diverge from other proxies and from direct measurements. I have explained this to you on a number of occasions. For some reason, you don't seem to understand this explanation. It is not clear to me that nth time will be a charm.

      Hoo

      Delete
    4. Dr. Egnor: What is the statistical correlation between the fact that the period of stasis from 1996 to now corresponds to a very large epoch of CO2 release (a third of the total human CO2 emissions)? Why was most of the warming from 1800 to now prior to the largest CO2 rates of emission, which only became substantial in the mid-20th century, and now do not correlate with substantial warming (from 1996 to 2012)?

      These are all good questions. We can discuss them once you agree that there has been a slight warming trend in the last 15 years. As shown by simple data analysis. And as a greed to by Dr. Curry.

      Hoo

      Delete
    5. @Hoo:

      You'll enjoy my most recent post. It clears up the Curry issue.

      Delete
    6. @Hoo:

      On the tree-ring data deletion, if the tree ring data is unreliable and does not reflect temperature increases, why was it left in the data set for the pre-1960 epoch?

      After all, if tree ring data is insensitive to temperature increases akin to those of 1980-1996, then including it in the pre-1960 data may erroneously suppress previous elevations of temperature.

      But of course previous elevations of temperature would have demolished the warmist claim that the 1980-1996 elevation was unprecedented and man-made.

      In order to make the warming appear unprecedented, they left the tree ring data in the pre-1960 data and removed it from the post 1960 data.

      Fraud, in other words.

      Delete
  10. The warming trend in the last 15 years has been relatively small, somewhere between 0.05 and 0.10 degrees per decade. The upper limit corresponds to what it has been on average over the 20th century. So this value is nothing to sneeze at. With all of the CO2 being pumped into the atmosphere, we were expecting to see a faster warming trend, closer to 0.2 degrees per decade. That has not happened yet.

    This is what Judith Curry calls "a pause in global warming." Note that by a pause she does not mean a complete stop. She defines a pause "to mean a rate of increase of temperature that is less than 0.17 – 0.2 C/decade." She, in fact, affirms the existence of a warming trend, she just thinks it is too small.

    Hoo

    ReplyDelete
  11. Dr. Egnor,

    I have asked you several times to state whether you agree with Judith Curry's assessment:

    "Has there been any warming since 1997 (Jonathan Leake’s question)? There has been slight warming during the past 15 years."

    Do you agree with it?

    Hoo

    ReplyDelete