How SkepticalScience views global warming – one way only

Guest essay by Sheldon Walker

Most people have probably seen the SkepticalScience graph called “The Escalator”. If you haven’t seen it yet, then you can view it here:

the-ss-escalator

Source: http://www.skepticalscience.com/graphics.php?g=47

SkepticalScience claims that “Contrarians” inappropriately “cherrypick” short time periods that show a cooling trend.

But SkepticalScience uses a linear regression over the full date range (1970 to December 2014), to determine the “long-term global surface air warming trend of 0.16 °C per decade”.

We can create a global warming contour map, that shows the SkepticalScience view of the warming rate. Here it is:

graph-1-2

Because SkepticalScience uses a linear regression over the full date range, they can only get a single straight line, with a single fixed slope, as seen in part 2 of their escalator animation:

the-ss-escalator-pt2

It is impossible for them to show a slowdown or a speedup, if one existed. The method that SkepticalScience uses, guarantees that the global warming contour map of their results, will always be a triangle of a single colour.

We can also create a global warming contour map that shows what the warming rate actually did. Note that this contour map uses the same Gistemp global land and ocean temperature series that SkepticalScience uses. Here it is:

graph-2-2

Note that the SkepticalScience view of the warming rate agrees with what the warming rate actually did, when the trend length is greater than 26 years. However, when the trend length is less than 26 years, the SkepticalScience view of the warming rate looks completely bland, and is definitely wrong. Where are the El Nino’s and La Nina’s? Where are the slowdowns and speedups. Do they expect us to believe that global warming proceeds at a uniform constant rate?

I will take this opportunity to point out the recent slowdown in global warming. Look at Graph 2, between 2005 and 2010 on the X-axis, and between trend length 5 and 15 on the Y-axis. The large light-green area is the slowdown. Light-green means that the warming rate was between 0.0 and +1.0 degrees Celsius per century. The average warming rate for the whole graph, is the colour at the top of the triangle, which is yellow. Yellow means a warming rate between +1.0 and +2.0 degrees Celsius per century. So light-green is a slowdown compared to yellow.

This global warming contour map not only shows the slowdown, but it also suggests a possible reason for the slowdown. Look at the light-green areas above 1974, 1983, 1992, 1999 (this is a small one that looks as if it didn’t develop fully), and 2007-2008 (the recent slowdown).

It appears that there is a slowdown approximately every 9 or 10 years. This sounds like it could be a natural ocean cycle, like the PDO or AMO. I have read articles by scientists suggesting that the recent slowdown was caused by natural ocean cycles, and this global warming contour map certainly supports that view.

For anybody who would like to view more global warming contour maps, the website “mta-graphs.com” has 27 UAH contour maps, and 18 Gistemp contour maps. The UAH contour maps all cover 1980 to 2016 (because the UAH satellite series only started in 1979).

The Gistemp contour maps cover 1880 to 2016 (the big picture), 1970 to 2016 (the time of fairly constant global warming), and 1980 to 2016 (to match the date range of UAH).

For anybody who would like some general information about global warming contour maps, there is an article at:

https://wattsupwiththat.com/2016/12/04/new-website-provides-strong-evidence-of-the-recent-warming-slowdown/

Advertisements

192 thoughts on “How SkepticalScience views global warming – one way only

  1. The escalator graph includes the pause since 1998, and even Phil Jones has admitted that even a ten year pause was significant.

      • after there was a this pause , before has he claimed it was impossible it was a different story .
        But ha people only to find something with his data , which is also called doing good science, so that is OK

      • Ben Santer said

        The LLNL-led research shows that climate models can and do simulate short, 10- to 12-year “hiatus periods” with minimal warming, even when the models are run with historical increases in greenhouse gases and sulfate aerosol particles. They find that tropospheric temperature records must be at least 17 years long to discriminate between internal climate noise and the signal of human-caused changes in the chemical composition of the atmosphere.

        And we went beyond 17 years with no warming so I guess Ben thinks the models must be wrong.

      • Phil Jones did acknowledge that there was no statistical difference between three warming periods, ie., 1850 to 1880, 1910 to 1940, and 1975 to 1998.

        This is material because even the IPCC suggests that CO2 played no significant role post about 1950, so the fact that the rate of warming in the 1975 to 1998 warming period is no more than the rate seen in the earlier two warming periods undermines the claim that CO” is a strong driver of temperature.

        See: http://news.bbc.co.uk/2/hi/8511670.stm

        There was some suggestion by Santer that one needs at least 17 years worth of data to determine matters of significance. Periods shorter than 17 years are too highly impacted by natural variation.

        See: https://www.llnl.gov/news/separating-signal-and-noise-climate-warming

      • “Phil Jones did acknowledge that there was no statistical difference between three warming periods, ie., 1850 to 1880, 1910 to 1940, and 1975 to 1998.”
        There is one difference though. The one from 1975 to 1998 continued another 18 years.

      • 30 year warming periods are the modern norm.

        1850-1880. 1910-1940. 1975-2005 (not 98). Note 30-35 years in between:1880-1910. 1940-1975.

        A climate switch hit again in 2005. Cooling now til 2035-2040.
        Bank it. Invest based on it.

        The Climate Hustle window of opportunity ended in 2006 as the ClimateGate emails acknowledged

      • This is why President Trump is such a delight Nick. What YOU think simply doesn’t matter any more.
        It was worth it…just for that!

      • IIRC it was in a BBC TV interview in early 2010 , following the climategate exposure that Phil Jones stated that there was not significant warming since 1995.

        Note he was not cherrypicking 1998 to make that claim and there is now a detectable rise if the El Nino driven warming since 2011 is included or you use a dataset including the corrupted Karlised SST data.

      • “Phil Jones stated that there was not significant warming since 1995”
        That isn’t at all the same thing. You can have a substantial warming trend for a decade or more which is not statistically significant. Even though it matches predictions.

      • Tim, Ben said “at least 17 years” – that means the minimum period you need is 17 years, not that you can always reliably detect the trend given 17 years. I.e. 17 may be enough, but you may need more (for instance if you cherry pick the start and end dates to suit the argument).

      • dikranmarsupial writes

        Ben said “at least 17 years”

        And so when 17 year rolls around its no longer the models that are “discriminat[ing] between internal climate noise and the signal of human-caused changes in the chemical composition of the atmosphere.”

        Sorry dikranmarsupial, he said what he said.

        The models are fatally flawed for other reasons, but this is an example of a renowned climate scientist betting on them and losing.

      • @Nick Stokes

        “Phil Jones did acknowledge that there was no statistical difference between three warming periods, ie., 1850 to 1880, 1910 to 1940, and 1975 to 1998.”

        There is one difference though. The one from 1975 to 1998 continued another 18 years.

        Also, if you take the warming period from 1975 to present it is warming significantly faster than the earlier periods.

      • I think I was wrong to claim that 1975 – present was significantly different to 1910 – 1940.
        What I was looking at was the confidence interval for trend from 1975 – present was too small to contain the trend from 1910 – 1940.
        But the confidence interval for 1910 – 1940 can still contain the trend for 1975 – present.

        On the other hand 1975 – present is warming at a significantly faster rate than 1850 – 1880.

        To put some figures on these trends, using the SkepticalScience trend calculator and HadCRUT data,
        the trends in C per decade are

        1850 – 1880: 0.038 ± 0.085
        1910 – 1940: 0.129 ± 0.057
        1975 – 1998: 0.172 ± 0.078
        1975 – 2015: 0.171 ± 0.034

      • Do WUWT readers want to win public support for rational enviromental policy?

        The loony left cannot stand the light. Public Policy Debates would gain massive audience and FORCE the left to real debate instead of MSM talking points!

        PLEASE contact the Trump Transition Team and request Policy debates. https://www.donaldjtrump.com/c

        Presedential Public Policy Debates (PPPD) on CAGW and other issues like immigration with people like Pamela Geller, Robert Spencer and David Horowitz debating whomever the left wants to send to embarrass.

        Catastrophic Anthropogenic Global Warming debate would be excellent. The left CANNOT win those debates. So PLEASE email the Trump Transition Team to set these up.
        Edit Reply

  2. How strange … just two days ago I was wishing that somebody would address the escalator argument against skeptics and suggest why my gut has always felt that it was baloney, but I had not yet figured out how to argue why. … and here YOU have started the ball rolling on it.

    I will be interested to see how ensuing comments unfold to further enlighten me.

      • Wow, that was hilarious! His commentary during the video was priceless.

        Still, wasn’t there a recent paper purporting to show that polar zones reacted to climate much more strongly than the rest of the earth? Like by a factor of 4, or something? Regardless of the specific conclusions of that paper, this seems to be a fairly reasonable conclusion to me. And, the logical implication of this would be that to see a concurrent global response to the Greenland record, you’d have to reduce the Greenland record by whatever factor is appropriate. So, the wild swings in temperature from the Greenland data would necessarily be suppressed a bit. It would be interesting to see how that record would compare then to the modern one.

        …or, maybe I’m completely misunderstanding the basis of the paper, as well as its implications…

        rip

      • Is that vastly expanded view a graph of the temperature anomaly on the Greenland Ice Dome? Is that sensible to compare it to the global temperature. Splicing the instrumental record onto old ic ecore temperatures is a bit dodgy, isn’t it?

    • Same thing with me. I have been trying to figure out how to explain why the argument was problematic, but never quite figured out how to address it. I really appreciate this article.

    • In case it helps here’s a much less mathematical, more humanities-graduate-style objection to SkS’s claims about ‘how skeptics view temperature changes.’

      This appeared in a story we posted about the deni@l mechanisms preventing the MSM and bien-pensants from accepting that Trump was going to beat Hillary.

      The concept of the Escalator Graph, a standard tool in pejorative psychology, was first proposed by Bob Lacatena.

      The concept of pretending to know why people view the world differently from you—while religiously refusing to engage with them, empathize with them or ask them—was first proposed by retraction-prone “intellectual” Stephan Lewandowsky.

      Prof Lewandowsky’s methods were further vulgarized by his son and epigone (/’epəˌgōn/) John Cook, in defiance of the naysayers who didn’t believe it was possible to dumb them down any more.

      • Also, the escalator graph is ignoring the skeptics’ point. It’s not that there is no warming. Some may be using a slowdown to argue that or even for global cooling, but the bigger issue is a long term slowdown suggests the worst case scenarios of the warmists are overblown. It’s tough to see 6C of warming when 16 years gives you less than .1C.

  3. “Do they expect us to believe that global warming proceeds at a uniform constant rate?”

    Their GCMs essentially say so in their forward projections since they do not model natural cycles of variability. Their hindcasts have to tune-in aerosols in order to match trends, and not run too hot against the historical record. That tuning which in itself invalidates any claim of forward skill, means they far run too hot (linearly) for (future) projections. The Models are thus junk science built for a political purpose, that is to “get us to believe warming proceeds at a uniform constant rate” per CO2-GW theory.

    For all their complexity, weeks of number crunching, and massive file outputs the models are simply linear regression Rube Goldberg machines, where Lord Monckton’s irreducibly simple formulae work as well when realistic net feedbacks are used.

    • models are simply linear regression Rube Goldberg machines

      Good enough for government work (and spending), apparently.

    • “Their GCMs essentially say so in their forward projections since they do not model natural cycles of variability. ”

      wrong. Natural cycles are an emergent property of the results. That’s how they can evaluate that they can capture the frequency and amplitude of them correctly.

      What they cannot do is get the timing right.. because of the lack of good initialization data.

      • Mosher,
        show mw me where any GCM projects the ElNino and/or La Nina that almost certainly happens in the 2030-2040 time frame please. or 2020-2030. They don’t. They can’t.

      • Steven – with respect, how can natural cycles possibly emerge from models that do not have any inputs where solar IR output varies with sunspot cycles, do not calculate insolation as a function of orbital variations, and neither do they attempt to model ocean circulation nor do they include the well known ocean oscillations as inputs.

        Any model that purports to simulate actual climate changes must include as inputs, variations in heat received from the sun due to sunspot cycles and orbital cycles, and must account for the PDO and AMO, which obviously affect climate by varying the way they move heat from tropics to polar regions. Any model that doesn’t include those factors is (at the very best) going to simulate a kind of static world where nothing changes (except of course CO2, which is the only variable they consider.

        It’s quite conceivable that the idealized static world of a GCM, where nothing ever changes except CO2, might show some sort of cyclicity, and it’s possible that such a cyclicity might actually mimic the cyclicity we observe – within the current interglacial.

        A model that purports to simulate real climate should be able to simulate the switches between glacials and interglacials, and stadials and interstadials. Until a GCM or a swarm of GCMs can do that, how can they have any validity?

        I won’t go into clouds and albedo. Those are the Achilles Heels of all the GCMs. And volcanoes too. If GCMs need volcanoes to simulate the past, a volcano-free future cannot possibly be a valid prediction/projection of the future climate.

        I submit that, from consideration of this one point alone, the true purpose of GCM’s can be deduced. Ignoring for the moment the clouds/albedo issue, and ignoring for the moment variations in insolation, a GCM that does not include some randomly distributed volcanic eruptions of different magnitudes through the future IS NOT INTENDED to be a valid simulation of future climate. It is intended ONLY to demonstrate uninterrupted, monotonic warming consequent on emissions of fossil fuel-generated CO2; it is intended ONLY to alarm the reader; and is intended ONLY to promote progressive abandonment of fossil fuels in favour of so-called renewable energy.

      • Smart Rock,
        solar UV (and esp EUV) varies far more with solar cycles than IR (unless you mean incident radiation.
        Thanks for the long post response to Steven. I never would have had the patience to tap that long response out on my iPhone.

        Joel

      • Mosher writes

        That’s how they can evaluate that they can capture the frequency and amplitude of them correctly.

        They dont get the frequency or amplitude right. In fact they dont get anything much “right” about ENSO at all. Here is one that is supposed to be better at ENSO.

        http://journals.ametsoc.org/mwg-internal/de5fs23hu73ds/progress?id=fvptbACpNNZBZzpNikujljKGRCk77ZVnDsvwX8eHFzY,&dl

        Unless of course you believe having something like ENSO emerge at all is a win for the models.

        What they cannot do is get the timing right.. because of the lack of good initialization data.

        For example the timing of ENSO events is wrong and its not because of initialization data! They’re too frequent in this example paper. Wow, Mosh you’ve outdone yourself.

      • From SMART ROCK – “with respect, how can natural cycles possibly emerge from models that do not have any inputs where solar IR output varies”

        The simple answer to that is that the climate is a nonlinear dynamical system and such systems can produce
        periodic outputs from a non-varying input (see Hopf Bifurcation on wikipedia). They can of cause also be
        chaotic. So a sign that a climate model is working well is the appearance of El Nino like oscillations in the
        model. The best models do have oscillations similar to El Nino that arise spontaneously.

      • “… lack of good initialization data …”

        Why eg. the “calculate insolation”, is not reflected in the models?

        NASA: “Hal Maring, a climate scientist at NASA headquarters who has studied the report, notes that “lots of interesting possibilities were suggested by the panelists. However, few, if any, have been quantified to the point that we can definitively assess their impact on climate.” Hardening the possibilities into concrete, physically-complete models is a key challenge for the researchers.” (https://science.nasa.gov/science-news/science-at-nasa/2013/08jan_sunclimate)

        “… and how solar variations influence the Earth’s climate over long time scales REMAIN UNRESOLVED.”

        NOAA: “… our understanding of the indirect effects of changes in solar output and feedbacks in the climate system is MINIMAL …”

        IPCC 2013:
        “Moreover, it is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together. The best estimate of the human-induced contribution to warming is similar to the observed warming over this period …” .

        The best estimate?

        1951 … – http://www.eea.europa.eu/data-and-maps/figures/rate-of-change-of-global-average-temperature-1850-2007-in-oc-per-decade-5/image_xlarge – 1951 it’s a very specially year: “some comparatively short periods of negative changes”( http://www.eea.europa.eu/data-and-maps/indicators/global-and-european-temperature/global-and-european-temperature-assessment-6)

      • Nick- the model(s) can only be initiated using actual measurements at some point. The actual measurements always have an associated error, so every time the model with a different start point in that error range it will have a different output. And even if the model is run with the exact same inputs it will not necessarily follow the exact same progression due to emergent properties. The climate models can build plausible models of possible future climate states but there is no way to tell, a priori , which projection is the correct one.

    • I like that idea. Does anybody have an analysis where these ‘steps’ occur in data? Like I said it would be good to have periods where the steps go down. It looks like this will happen between now and 2030 but I just heard that my life expectancy has gone down and I may not make it.

      • There are no steps IN THE DATA.

        steps result when you CHOOSE to ASSUME a statistical MODEL and apply it to the data.

        Steps and Trends are properties of the models you choose to apply.

      • Mosh,

        There are no steps IN THE DATA.

        steps result when you CHOOSE to ASSUME a statistical MODEL and apply it to the data.

        Steps and Trends are properties of the models you choose to apply.

        I absolutely agree with your point that There are no steps IN THE DATA. I also absolutely agree with your point that there are no TRENDS IN THE DATA. As you astutely observe, Steps and Trends are properties of the models you choose to apply. OK, you’ve stated the patently obvious relationships/ dependencies between steps/trends and models. Can you enlighten us lesser mortals as to how you determine, or if you can determine, which models are appropriate?

      • Can you enlighten us lesser mortals as to how you determine, or if you can determine, which models are appropriate

        Indeed, Steven, a good question for you.

        My answer would be: if one has a bias, it shows up as selection of model that will end up showing data from certain angle.

    • Well their staircase is a totally crappy fit to what looks more like a triangular wave, with a slow climb and an abrupt reset. But a trendy straight line is the least credible fit of all.

      G

  4. “The escalator graph includes the pause since 1998, and even Phil Jones has admitted that even a ten year pause was significant.”

    exact quote and reference please!

    “It is impossible for them to show a slowdown or a speedup, if one existed.”

    IF being the operative word. If you want to show that there has been a slowdown or a speed up, then you need to perform an appropriate (i.e. accounting for the autocorrelation etc) test for a change in the rate of warming. However if you perform such a test, then you find the evidence is not statistically significant. Occam’s razor then suggests that a single linear model is preferable.

  5. So what does the warming rate do over time?
    What is the warming rate from 1970 to 2000?
    Same from 1970 to 2001?
    To 2002?
    To 2003?
    Same for every additional year till 2015?
    The warmimng rates will only indicate that as additional years are added into the mix, the rate of warming decreases over time

    • “The warmimng rates will only indicate that as additional years are added into the mix, the rate of warming decreases over time”

      Not really.

      GISS trend in degrees / decade from 1970 to selected years

      2000: 0.179
      2005: 0.182
      2010: 0.181
      2015: 0.174
      2017: 0.183

      For RSS 3.3 from 1979

      2000: 0.145
      2005: 0.161
      2010: 0.137
      2015: 0.124
      2017: 0.135

      All these differences are well well withing their confidence range.

      • “Start with La Nina and end all your trends with El Nino????? Give me a break.”

        I was using the dates Bryan A asked for. He said that if you started with the trend from 1970 to 2000, and then moved the end date forward you would see “the rate of warming decreases over time”.

        I was just demonstrating that this isn’t true. I could have gone through every year rather than a few sample years, but decided that would be a waste of time.

        I take it that you also objected to people claiming there had been no warming starting with the 1998 El Nino.

  6. By cherry picking 1970 as their start date, they start their trend at the end of the last hiatus in warming from 1945 to 1975. They also ignore the supposed “natural warming” trend, that no computer model can duplicate, from 1915 to 1945. Also 1915 to 1945 is nearly identical to 1975 to 2005. Pick your date range, pick your result! More here: https://wattsupwiththat.com/2016/08/22/virtually-indistinguishable-comparing-early-20th-century-warming-to-late-20th-century-warming/

    • Just a reminder, since you tend to use monthly figures. Today is 08/12/16, why do you stop your graph so early?

    • And if that is GISS, then its all highly suspect anyway.

      Particular around the 1940-1970 where the cooling period has been all but erased.

      Working with GISS fabrications is basically a total waste of time if you are seeking some sort of understanding or truth.

  7. How most honest scientists view it as 0.1c per decade is neither alarming or action needed to cut CO2 levels to help hardly change a mainly natural trend. Since the 1990’s the rate of warming has significantly fallen and has not been expected to do so by the climate models.

    A 0.1c per decade trend is not in the next decade or more going to suddenly become 0.3c or more. We don’t know this for sure, but the planet is not suggesting any reason why it would. No positive feedback, no missing heat, therefore after numerous decades, we now already have a very good idea what to except from what has already happened so far.

  8. “long-term global surface air warming trend of 0.16 °C per decade”

    …lies within the range of natural variability, 1SD of centennial variation over the last 6000 Holocene yrs. (Lloyd PJ, Energy & Environment 2015)

    • Robert Kernodle,

      Excellent comparison. Here’s another link showing where we are now – at the cool end of earth’s geologic temperature range:

      Next, this chart covers a shorter time frame (740,000 years):

      And this chart shows our current Holocene ‘climate’:

      Readers can draw their own conclusions about the SkS ‘escalator’…

    • Just three questions :
      – I thought that today’s global average surface temperature was around 14° (vz GISStemp), your’s is 8/9°?
      – When is “today”, meaning could you give us a date, please?
      -I have not been able to find what my ancestors looked like during the precambrian era, I do not know about yours, any idea?

    • Robert Kernodle,

      Can I assume from this graph that temperatures during the Cretaceous were about 12-13 C degrees higher than they are today?

      That would be amazing.

      • This is correct. But remember that we’re talking about some “averaged” global temperature. The equator doesn’t change much, but the poles were ice-free and temperate in the Cretaceous. That changed some 3 million years ago, when the planet returned to the current ice age (Yes, Virginia … we ARE living in an ice age!)

      • Thank you, thank you,

        This is BIG! This is HUGH!

        This alone destroys the Premise that a Warmer World is a Catastrophe. This knowledge alone destroys the Premise that we’re headed for disaster ESPECIALLY IN THE MINDS OF NON-SCIENTISTS.

        The scientists on this website need to know that when the General Public reads a news story about how 2015 and 2016 are the “hottest years ever”, they assume (or are deliberately led to believe) the world has never been so hot.

        The phrase “on record” isn’t processed the same way in scientific circles as it is in the minds of the General Public. When a lay person hears that the temperature has broken a record, they think it means it’s highest it’s ever been in the entire existence of the Earth – that’s why so many people are afraid of Global Warming. They don’t realize it’s just a statement about the last 200 years, and that we’re actually on the “cold” side of the Earth’s temperature experience.

        Later in January, when the Warmists will issue a story that 2016 was the HOTTEST YEAR EVER, while showing a graphic of the Globe with flames coming off it, they should be beaten down with the FACT that the Earth has endured millions of years where the temperature was at least 20̊ F warmer than it is now.

        It should be emphasized that the 200 year record means absolutely nothing in the grand scheme of things.

        What they’re doing is pure Fear Mongering!

        When Skeptics show graphs to refute Global Warming for the General Public, they should include the Prehistoric levels so that the people can make a comparison.

        Don’t use “deltas” and “anomalies.” Just show a graph that says the Cretaceous average temperature was around 77 degrees F and today it’s around 58 degrees. Use Fahrenheit in the US and Celsius elsewhere. Fahrenheit is easier to process for Americans.

        How many F degrees colder was the last major Ice Age in layman’s terms?

    • Sorry for where my former reply appeared in the sequence of comments. Duh, I’m just now figuring out that replies are tied to the immediately preceding comment, NOT to the entire discussion thread. I actually posted my “paltry attempt” before RH posted his much more entertaining video link to the counterargument for the alarmist escalator ploy.

      … new-kid-on-the-block dumb-dumb mistake. … I’ll get better. (^_^)

  9. It isn’t really a convincing graphic in my opinion Sheldon. The triangle thing doesn’t work for me. I do agree with your conclusions concerning a linear regression. Fait Accompli as the French might say; a forgone conclusion.

    Why not use polynomial regression? I’d think two should work to demonstrate rising and falling of the variable?

    • Not a good idea! You suggest a second order polynomial, but you have presumably not thought about any “predictions” (or any other term describing how to guess at possible future values). The safest model is the simple linear (first order) one. All others go wildly astray if extrapolated for more than a few x units. No model will reliably predict future temperature values for more than a few x units. In other words, all models are wrong, and useful ones are effectively nonexistent..

  10. “It is impossible for them to show a slowdown or a speedup, if one existed”

    The reason being of course looking at it a different way.

    If the graph showed a mirror image from the peak warmth with a cooling period for more than 30 years, but less than the length previously. It would still show an overall warming trend because it had not reached the same time interval. When the mirror image had reached the same time interval from the peak warmth, the trend would be a flat zero.

    This is why the graph from sceptical science is nonsense and doesn’t show anything it claims.

  11. A bit of off-topic news —

    Port Angeles, a small Washington State city spent more than $100,000 on three windmill-like turbines which are expected to generate $1.39 per day in electricity or roughly $42 per month.

    City councilwoman Sissi Bruch told The Peninsula Daily News that she was disappointed in the savings but did “appreciate” the fact it “would educate folks about wind power”.

    OH! SHE IS SO RIGHT!!!!!!

    Eugene WR Gallun

    • Poor Sissi Brunch, “generation” and “savings” are not synonymous. Unless the micro turbines reduce every demand peak the commercial power price formula probably reduces any “savings” to that of an art project.

    • I loved Port Angeles and I can see how there are people there who fell for the whole wind power riff. We had a windmill on the farm for our water well. It was nifty and had to be controlled when the winds grew. It seems like these things they use today are just too big! Same thing with those mirrors in the Mojave desert near Searchlight. That thing is scary and kills birds.

    • The turbines are expected to just barely power the safety lighting in the park. The Turbines cost $114,000.00. Of course, it wasn’t actually the city’s money, it was part of a grant. Funny how these projects only get built with other people’s money. A cost analysis was not part of the decision process, as it was meant to be both art and education. The city folk don’t think the turbines are esthetically pleasing , and I think the lesson learned is not the one that was intended.

      SR

  12. Those of us on the skeptical side of the coin allow the other side to decide on the terms and language. In particular average temperature.

                     The average of 1 and 99 is 50 and the average of 49 and 51 is also 50.

    Temperature records record daily highs and lows and they do so for a reason. Namely how hot and how cold it got during the day is important. Using averages loses a lot of valuable information. When Johnny Carson said it was really hot today and the audience responded, “How hot was it?” they weren’t asking about the average temperature. So how hot are the summer time temperatures? Summer goes from June 21st to September 21st and here’s a map of the United States that shows summertime high temperature trends back to 1895. Source is NOAA’s Climate at a Glance.

    For the contiguous United States as a whole summer time temperatures have trended downward for over 80 years. Here’s what that looks like:

    Any objective person who lives in the Midwest knows that we have enjoyed warmer winters and cooler summers for probably two decades or more. Climate change seems to have resulted in milder weather. The folks on the other side who say the new normal is extreme weather are going to have to express it in terms of Extreme Mildness.

    • Steve Case, you have pointed out one of the major issues with the consensus claims about dramatic and dangerous change: what dramatic changes, and what dangerous changes? When a Cat 5 hurricane is coming close, one does not need carefully derived synthesized data to prove it. Yet after over 30 years of asserting that dramatic dangerous global warming/ climate change is happening, nothing is happening that can be noticed without either fake news or stats analytics that fail to hold up under reasonable scrutiny.

      • Thanks for the reply. That the climate seems to be milder now than in past decades, can be demonstrated using the very data from the organizations asserting that the climate is becoming extreme.

        There are lots of bloggers who point out that NOAA’s data has been manipulated to favor the, …dramatic dangerous global warming/ climate change is happening,… point of view but even with those corrections/manipulations the signal that climate seems to be getting milder hasn’t been erased.

    • There’s an easy and objective method to verify your opinion. You may well be correct. All you need do is to compute the standard deviations over each year of the monthly means, and then examine these as a time series. For this you could choose to fit a linear model to to the SDs, or perhaps compute their cumulative sum relative to the average SD. This would disclose any abrupt change, such as the prominent one that occurs in Western European (an more widespread) data at about 1900, give or take a year or two.

      • The only opinion posted was a snark that the other side, when claiming the new normal is extreme weather, should consider the term “Extreme Mildness” Otherwise the post lays out facts as generated by NOAA’s Climate at a Glance webpage.

    • Steve,
      However, if you select endpoints of 1965 to 2015, you get a very different result! That is, in the last 50 years, about two generations and almost twice the time interval that differentiates climate from weather, you get increasing high temperatures. Similarly, you would have gotten the opposite result if you had a start point in the early-1900s. I’m no supporter of CAGW, but I consider this to be a transparent case of cherry-picking that only gives the ‘other side’ reason to criticize.

      • The Map was generated by asking the question for each state, “How far back can you go and still find a negative trend for the June – September Maximum temperatures?” As it turned out for most of the Mississippi and Ohio River valleys it was all the way back to the 19th century. For the contiguous United States, it was back to 1930.

        The flip side is to look at the Minimum temperatures, and they are all rising. So summer Maximums are falling and winter and summer Minimums are rising. Looks like if there’s been any climate change over the last century, it’s been toward a milder one. Hence if the other side is going to claim the new normal is extreme weather, they are going to have to call it extreme mildness.

        Did that address your point of transparent cherry-picking? Probably not, it merely explains what was done using all of the data back to 1895 including the highs and lows.

      • Clyde….”if you select endpoints of 1965 to 2015″…you can’t ignore the 1930’s heat….it obviously wasn’t man made

    • I notice that every one of your extrema values falls exactly on a sampled measurement point.

      Why is that ?? There are NO maxima or minima occurring in between the sampled values.

      That seems quite unlikely in the real world.

      G

      • I explained this to you on another thread. Not sure why you are so confused in the first place.

      • Well I see the pestilence is deeply rooted.

        What NOAA measured / observed / recorded / whatever was a set of discrete numerical values. Well I counted 120 distinct dots on the plot, apparently representing one plotted point per year. It is anybody’s guess how they arrived at each of those numbers.
        But I’m perfectly willing to accept that those 120 data points represent something credible, and that plotting those dots on a scatter plot is a legitimate enterprise.

        What is NOT legitimate, and is phonier than a three dollar bill, is having some idiot at NOAA join those dots with STRAIGHT lines, presumably with the idea that (s)he is communicating further information over and above the 120 dots.

        What that process actually does, is take a perfectly valid scatter plot of properly sampled measures of a real band limited continuous function; and convert it into a visual, that clearly is NOT a band limited continuous function since it has points of infinite curvature, each of which occur at the original data sampling points.

        So the red line plotted function has an infinitely broad spectrum, which CANNOT be sampled properly by even an infinite number of sample points.

        So the blue or black dots may in fact be valid. The red line graph is total bullshit, and demonstrates a complete lack of understanding of the general theory of sampled data systems.
        Nobody who draws such graphs should be considered to be any kind of credible scientist.
        It’s complete rubbish.

        Moreover it is not even necessary.

        If you chart those exact same 120 points in M$ Excel, a simple click with the mouse will connect those dots with a proper band limited continuous function that still goes through every measured point. It won’t necessarily be an exact replica of the original function but it will in all likelihood be a damn side better than the nonsense in that NOAA graph.

        G

        PS I actually downloaded the file purporting to contain that raw data. It is a totally useless file, that does not properly import into Excel, because it puts several numbers in a single data column, and you can only separate them by retyping the entire 120 pointsa of information.

        I’m not going to waste my time to unscramble a BS file created by some total lame brain.

    • Michael Jankowski December 8, 2016 at 3:06 pm
      I explained this to you on another thread. Not sure why you are so confused in the first place.

      ?????

      Link? Statement directed at who?

  13. I believe the triangle should be used with detrended data from 1850 to present. That gets rid of the effects of natural recovery from the cold depths of the Little Ice Age.

    One would then have a visual of the various periods of warming and cooling events. One could then compare them to attempt to explain differences according to assumed climate drivers. AMO, PDO, CO2, etc.

      • Francois, present means present: December 2016.

        Where the heck did you get “66 years ago?”

        The point is, cooling and warming trends should stand on their own. It doesn’t matter if a trend started at a higher or lower point (temperature), it matters what the period trend was. Only then could we hope to tease out causation.

        I have no intention of arguing this point. It is.

  14. However, when the trend length is less than 26 years, the SkepticalScience view of the warming rate looks completely bland, and is definitely wrong. Where are the El Nino’s and La Nina’s? Where are the slowdowns and speedups.

    That’s like saying cartographers are wrong to show a wooded area as a solid green, when the actual wood is made of lots of trees and parts that are not trees. The whole point of fitting a model to a noisy graph is to smooth out the noise and show an underlying signal.

    It appears that there is a slowdown approximately every 9 or 10 years. This sounds like it could be a natural ocean cycle, like the PDO or AMO.

    Have you considered it might be the solar cycle?

    • What if all that fluctuation (noise) IS the signal.

      Smoothing it eliminates information, and introduces false data i.e. NOISE.

      So DON’T smooth out experimental measured data values. That IS what you observed.

      G

    • No, the cartographers are wrong to show bare ground where there are trees. There is nothing wrong with solid green for a wooded area.

      I thought that the solar cycle was 11 years.

      • “What if all that fluctuation (noise) IS the signal.”

        “No, the cartographers are wrong to show bare ground where there are trees. There is nothing wrong with solid green for a wooded area.”

        The point is that like a map, a trend line is an abstraction of reality, and like a map it’s often more useful to know the broad outline of reality than the small scale detail. A map that showed the location of every tree in the country would be less useful than one showing where the woods are (as in not being able to see the woods for the trees).

        When it comes to temperature trends, it’s more useful to know the underlying trend, than every short term bump.

        “I thought that the solar cycle was 11 years.”

        I wasn’t sure if you’d established the exact length of these cycles. The solar cycle is supposed to have an effect on temperatures, but I imagine any effect on trends will depend on what length of trend you are talking about, and will probably be less important than El Ninos or volcanic activity.

      • abstraction of reality

        Either that, or a product of your imagination. Like a flat painting represents reality, it is just a painting.

        The trend is a selection that fits a predetermined bias to find a long linear trend. Worse, the linear trend directs search for errors that deviate from the imagined trendline.

        I’m not sure why cagwists don’t see this problem in the unreally good escalator.

      • @Bellman

        You said, “The point is that like a map, a trend line is an abstraction of reality, and like a map it’s often more useful to know the broad outline of reality than the small scale detail.”

        A global warming contour map does this by using a different colour for each 1 degree Celsius per century warming rate range.
        Light-green is 0.0 to +1.0 degrees Celsius per century.
        Yellow is +1.0 to +2.0 degrees Celsius per century.
        Light-orange is +2.0 to +3.0 degrees Celsius per century.
        Etc

        If I used a 0.1 degree Celsius per century warming rate range, then the contour map would be a mass of coloured speckles, and it would be very hard to understand.
        Using the 1 degree Celsius per century warming rate range gives you an abstraction which is easy to understand.

      • @Sheldon Walker

        You said, “The point is that like a map, a trend line is an abstraction of reality, and like a map it’s often more useful to know the broad outline of reality than the small scale detail.”

        A global warming contour map does this by using a different colour for each 1 degree Celsius per century warming rate range.

        I wasn’t complaining about your contour map, just the claim that a linear trend line was “definitely wrong” as it didn’t show every bit of noise. This becomes apparent in your graph as you move down to the base representing trends of just a few years. As the time intervals get shorter the trends become increasingly meaningless. Here for example is what rolling trends of just two years look like.

        The trends can switch between positive and negative trends of over 20 C / century. In your graph this shows up as alternating red and blue. Yet none of these trends is likely to be significant, all you are really seeing is that some years are warmer than others.

      • Well Bellman, if YOU have actually measured the exact location of every tree, then THAT would be the most interesting information to have. Specially since it is reality; and NOT ” an abstraction of reality. ” AKA unreality.

        You can’t ever have any more information than that which you have actually collected.

        You can’t simply create information out of emptiness.

        What you machinators do is invent something which can NEVER be observed, out of something else that has actually been observed and recorded.

        You may consider that to be new “information”; but it is ONLY information about what YOU DID with the reality to create a false image of something nobody ever observed or could observe because it never even happened.

        So don’t kid yourself. You certainly are not fooling us.

        G

      • @Bellman

        You said, “I wasn’t complaining about your contour map, just the claim that a linear trend line was “definitely wrong” as it didn’t show every bit of noise.”

        The linear trend line is fine, as long as people realise that it is an “average”, not an “absolute”. SkepticalScience are trying to ridicule skeptics for trying to find time intervals with a slope lower than the linear regression line. There is nothing wrong with doing this, and since the linear regression line is an “average”, you would actually expect to be able to find some. You would also expect to be able to find some time intervals with a slope greater than the linear regression line.

        In summary, SkepticalScience are wrong to treat the linear regression line as if it was an “absolute”, which can not be questioned.

        You said, “This becomes apparent in your graph as you move down to the base representing trends of just a few years. As the time intervals get shorter the trends become increasingly meaningless.”

        I am not sure that I agree that they are “increasingly meaningless”. I chose red to represent all trends greater than or equal to +5.0 degC/century, and dark-blue to represent all trends less than or equal to -5.0 degC/century. This means that all of the “strong” trends are handled well on the graph.

        Do you realise that all of the trends of greater length than the trends that you call “meaningless”, are just averages of the “meaningless” trends? So they are not really “meaningless”, are they.

        You said, “In your graph this shows up as alternating red and blue. Yet none of these trends is likely to be significant, all you are really seeing is that some years are warmer than others.”

        Don’t make the mistake of assuming that a trend which is NOT statistically significant, is not “real”. Statistical significance only tells you the probability that the trend happened by chance. The trend is still there, even if it happened by chance.

      • @Sheldon Walker

        The linear trend line is fine, as long as people realise that it is an “average”, not an “absolute”.
        SkepticalScience are trying to ridicule skeptics for trying to find time intervals with a slope lower than the linear regression line.
        There is nothing wrong with doing this, and since the linear regression line is an “average”, you would actually expect to be able to find some.
        You would also expect to be able to find some time intervals with a slope greater than the linear regression line.

        You say that the long term trend is an “average”, but the escalator graph shows why that is misleading.
        Look at all the straight lines in the escalator graph, they cover the entire series from 1970, but none of them is positive.
        If you could determine the overall trend by averaging them you would have to conclude that the complete trend was less than zero.
        The reason the underlying trend is greater than the sum of its parts is because of all those inconvenient upward steps between each “pause”.

        In summary, SkepticalScience are wrong to treat the linear regression line as if it was an “absolute”, which can not be questioned.

        No one should be suggesting a linear trend cannot be questioned.
        Linear trends are unlikely to be completely correct, but they can be useful even when not perfect.
        At the least this trend is the simplest explanation of what temperatures are doing.
        Other models are possible, but you should require strong evidence before claiming the linear trend is completely wrong.

        What the escalator graph shows is that it’s possible to cherry pick periods of zero-growth throughout the series,
        but it also demonstrates why this is misleading as it results in a discontinuous trend.

        You said, “This becomes apparent in your graph as you move down to the base representing trends of just a few years.
        As the time intervals get shorter the trends become increasingly meaningless.”

        I am not sure that I agree that they are “increasingly meaningless”.
        I chose red to represent all trends greater than or equal to +5.0 degC/century, and dark-blue to represent all trends less than or equal to -5.0 degC/century.
        This means that all of the “strong” trends are handled well on the graph.

        By “meaningless” I meant both that they are not statistically significant and that they have no predictive value.
        It’s just not credible that a trend of 20 C / century is a valid indication of what temperatures will do in the future.

        Do you realise that all of the trends of greater length than the trends that you call “meaningless”, are just averages of the “meaningless” trends?
        So they are not really “meaningless”, are they.

        To reiterate my first point –
        a trend is not an average of lots of small trends, it’s the line that “best fits” all the data.
        What’s really happening is that some years are above the trend line and some below it –
        this is what you’d expect with normal random variation.

        People have a tendency to see meaning in randomness – statistics is about trying to avoid that problem.

        You said, “In your graph this shows up as alternating red and blue.
        Yet none of these trends is likely to be significant, all you are really seeing is that some years are warmer than others.”

        Don’t make the mistake of assuming that a trend which is NOT statistically significant, is not “real”.

        Remember that the next time some uses the “no significant warming since year x” line.

        Statistical significance only tells you the probability that the trend happened by chance.
        The trend is still there, even if it happened by chance.

        A common mistake, but significance does tell you the probability that something happened by chance.
        It tells you the probability that under the assumption of the null-hypothesis you would have seen the same result.

        But you’re correct, any trend is a trend,
        it’s just not possible to tell if they are the result of some physical process or just random month to month variation.

        I would suggest that if you want to look at short term variation in temperatures, linear trends are not very helpful.
        Look at how the actual temperatures change over a short period of time, not the rate of change.

    • Either that, or a product of your imagination. Like a flat painting represents reality, it is just a painting.

      If a painting represents reality it isn’t just a product of your imagination. But yes, a painting can be a representation of reality, just as a statistical model tries to be a representation of some underlying reality.
      A painting might remove irrelevant detail from its subject, a linear model removes the irrelevant noise.
      So good analogy.

      The trend is a selection that fits a predetermined bias to find a long linear trend. Worse, the linear trend directs search for errors that deviate from the imagined trendline.

      I’m not sure why cagwists don’t see this problem in the unreally good escalator.

      I’m not sure what you are saying here exactly, but linear regression is a pretty important statistical tool. Getting rid of it just because it’s imaginary wouldn’t leave much room for analysis.

      • You say that as if ” analysis ” is something to be desired.

        ” Analysis ” is what the MSM do with the ” NEWS “. The talking heads never tell you what happened, they ” analyze ” for you what THEY want you to believe actually happened, after you already saw with your own eyes what really did happen.

        Stop with the analysis BS and start doing reality for a change.

        No wonder the world is in disarray. Everybody is fixated on analysis, which is simply a diversion from actually working on real problems that need fixing.

        G

      • You say that as if ” analysis ” is something to be desired.

        No wonder the world is in disarray. Everybody is fixated on analysis, which is simply a diversion from actually working on real problems that need fixing.

        Do you take issue with all this analysis?

        https://wattsupwiththat.com/?s=analysis

  15. It is a series of cut-off graphs–ie the old special pleading fallacy. Estimates of temperature exist (of varying quality and reliablity) for billions or millions of years. Picking any start date is almost always a distortion of reality. I think one has to always take into account the whole history, and the unreliability of the evidence, when making any claim.

  16. Sheldon, I am not sure quite what your point is. You lead off with two sentences about the “escalator graph” … and then never come back to it again!

    If your point is that linear fits never show changes in slope, well, that is pretty obvious.

    If your point is that the escalator graph is better/worse than a linear fit, you haven’t shown that.

    SkepticalScience does not use a linear regression to “force” a uniform trend. They use a linear fit because it is the simplest function that reasonably fits the data. If there was a significant, clear change in warming from a linear rise, then the residuals of the linear fit would show that. And then I suspect they would modify their fit. But the linear fit does work pretty darn well.

    If you want to show that a linear fit (two adjustable parameters) is NOT a good approximation to the data, then find like …
    * two horizontal lines (two adjustable parameters) that fit the data as well
    * a quadratic that fits significantly better.
    * a linear fit where the residuals show a clear signal.
    * some other function with more adjustable parameters that fits MUCH better.
    For example, breaking the time period into a linear rise until 1998 (or 2003) with a flat “pause” will not give a particularly better fit. (And don’t forget about von Neumann’s line “With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”)

    PS. For fun, I would encourage you to make a contour plot of the “escalator” graph You will find that the top looks about the same (a sea or yellow) while the bottom has some bizarre bright red triangles in a sea of light green. That doesn’t sound like it is any closer to the data than the linear fit.

    • “For fun, I would encourage you to make a contour plot of the “escalator” graph You will find that the top looks about the same (a sea or yellow) while the bottom has some bizarre bright red triangles in a sea of light green.”
      Yes, I think that would be interesting, and as you describe. The bottom edge would be light green (for zero) with just red dots at each jump. Each red dot would be he apex of an inverted triangle with sharp edges, with the red fading as you go up (still green between). The different triangles would merge, eventually to the same yellow.

      A prettier pattern, but no better as a descriptor. Both get the long term right. The escalator would have no blues – in fact, no negative trends.

      • Great. I want it even warmer and am doing my best to get us there. Although I have no faith in my contributions having any significant effect.

    • Think of it this way.

      If the world had been cooling generally for the last 40 years it would take more than 40 years just to get a positive trend with the same rate of change.

      If it had got to 38 years of global warming the same rate as the previous cooling and was still showing a slight cooling trend. Would you have said this shows the world is still cooling?

    • My main points are:
      – that the SkepticalScience escalator graph is inaccurate, and that a global warming contour map is much better at showing what happened to the warming rate.
      – that there is some justification for looking at shorter trends, which may be a slowdown or cooling trend.
      – I didn’t say this in the article, but I suspect that the contrarian view was made up by alarmists, who lied and blamed skeptics for it. Skeptics do look at shorter trends, but I have never seen them chain them together like that.

      I would like to see them use a global warming contour map, or a LOESS smooth, to show what has happened to the warming rate. If the LOESS smooth was done with a 10 year moving regression, and it came out straight, then people could have confidence that the warming rate was fairly constant. If it didn’t come out straight, then people could see what had happened to the warming rate.

      • “Skeptics do look at shorter trends, but I have never seen them chain them together like that.”

        I think this is misunderstanding the point of the SkepticalScience escalator graph.
        They are not saying that “skeptics” chain together a series of steps, that would be making the point that temperatures are warming.
        What they are saying is that at, as you say, skeptics look at the shorter periods and claim warming has stopped..

      • If the LOESS smooth was done with a 10 year moving regression, and it came out straight, then people could have confidence that the warming rate was fairly constant. If it didn’t come out straight, then people could see what had happened to the warming rate.

        I think this is what you are asking for – GISS temperatures with a LOESS (in red) using an alpha of 0.25.

        But I’m not sure what you will make of it. At this scale the LOESS is showing the meanders in the data, just as short term trend lines do. To me it really illustrates the key point that while there are ups and downs, the LOESS smoothing is never very far from the long term trend.

  17. Of course, Skeptical ‘Science’ uses the super manipulated GISS ‘data’ for their comparison graph.

    One of the first tasks for the new Trump team will be to hire Real Scientists to uncover the massive unbelievable temperature data manipulation (drain the scientific swamp) and to redo the temperature measured vs CO2 vs modelled analysis.

    GISS data manipulation

    http://notrickszone.com/2015/11/20/german-professor-examines-nasa-giss-temperature-datasets-finds-they-have-been-massively-altered/#sthash.ibiNW4TW.Saxx5o6a.dpbs

    From the publicly available data, Ewert made an unbelievable discovery: Between the years 2010 and 2012 the data (William: GISS) measured since 1881 were altered so that they showed a significant warming, especially after 1950. […] A comparison of the data from 2010 with the data of 2012 shows that NASA-GISS had altered its own datasets so that especially after WWII a clear warming appears – although it never existed.

    The old data showed regular cycles of warming and cooling over the period, even as atmospheric CO2 concentration rose from 0.03% to 0.04%. According to the original NASA datasets, Ederer writes, the mean global temperature cooled from 13.8°C in 1881 to 12.9°C in 1895. Then it rose to 14.3°C by 1905 and fell back under 12.9°C by 1920, rose to 13.9°C by 1930, fell to 13° by 1975 before rising to 14°C by 2000. By 2010 the temperature fell back to 13.2°C.

    But then came the “massive” altering of data, which also altered the entire overall trend for the period. According to journalist Ederer, Ewert uncovered 10 different methods NASA used to alter the data. The 6 most often used methods were:

    • Reducing the annual mean in the early phase.
    • Reducing the high values in the first warming phase.
    • Increasing individual values during the second warming phase.
    • Suppression of the second cooling phase starting in 1995.
    • Shortening the early decades of the datasets.
    • With the long-term datasets, even the first century was shortened.

    RSS Data Manipulation

    https://wattsupwiththat.com/2016/03/02/the-karlization-of-global-temperature-continues-this-time-rss-makes-a-massive-upwards-adjustment/

    Forget homogenization, that is so 2010. If the pause is bothering you and your belief is that there must be more warming, we only need to find it in the data, then what you need is “Karlization”, named after director of the National Climatic Data Center, (now NCEI) Tom Karl who pulled a fast one this summer trying to adjust the past down, so the present would be warmer. The sleight of hand on this was so obvious that even warm-oriented scientists such as Michael Mann and Ben Santer co-authored a rebuttal paper that said Karl was dead wrong and the pause was real. There is now a congressional investigationinto Mr. Karl’s apparently political actions disguised as science …

  18. If you go back to /79 and the lower step of their graph, the US was at the bottom of fifty year long cooling tend. That had began in about 1931. In /79 the average mean was even lower than it was in 1880. Their graph shows a warming of about 0.6, which transferred to the US historical temperature record would be about the same as the temperatures in the early /20’s.

    • But the further away we get from the data that is 50+ years old, the more accurate we can estimate it! :)

    • I have no idea what it would look like if you were to give us all the figures ’til today, but just a reminder : we are in the year 2016.

    • How come you have already a value for the 1880 five year mean, when the data only starts at 1880 ??

      But your five year mean would be a lot more credible if you simply erase the bogus fictional dotted lines, and just leave the little black square dots.

      The red curve MAY be a five year mean plot of the square dots, but it is NOT a five year mean of the fictitious dotted black line, which is total rubbish.

      G

  19. “This global warming contour map not only shows the slowdown, but it also suggests a possible reason for the slowdown.”

    What it does do is show the arithmetic that leads to it. You’ll notice that there are bands that go NW and NE, parallel to the edges. These were picked up in the earlier comments by people who saw a burning man and fire symbols.

    Those bands trace back to the bottom line, and to an event. An ENSO peak shows as a dipole; strong red (up) followed by blue (down). A dip has them in the other order. And if you look at a peak like 1998, it sends a blue shadow NE and a red NW. That reflects the fact that trends ending at a peak are positive, starting at a peak are negative.

    Dips send shadows the opposite way. And it gets blurred as shadows mix. A “slowdown” is where there are more blue shadows in the mix. You can trace the blue parts at the bottom that are responsible. The periodicity that you note is the periodicity of ENSO events.

  20. I clicked on the Skeptical Science link. I am now polluted, and certainly dumber having read some of the garbage there. How these clowns can ask for money to continue their bogus cause is laughable. I can certainly look at their list of donors as people I would avoid.

  21. The escalator chart is more an escalator of the adjustments applied by the NCDC to the actual temperature measurements. It is an example of the current topic called “fake news”.

    This is the real temperature and the only one we can rely on now. The lower troposphere temps from the satellites and prior to 1979 from the radiosonde weather balloons back to 1958. We will be back to 1959 temperatures in a few months.

    • Bill excellent data which shows warming has stopped since 1998 and the spikes of warmth and cold for that matter all ENSO related.

      Now the true test is coming and I think this prolonged solar minimum is going to change that temperature trend to one that is down.

    • Bill Illis on December 8, 2016 at 3:40 pm

      Hello Bill Illis,

      thanks for your nice chart, which however reminds me other radiosonde-satellite comparisons, where the atmospheric pressure level at which the balloons measured a given temperature gave an altitude other than that where the satellites measured a similar temperature.

      An exactly the same I experiece today evening when comparing, for the period Dec 1978 – Dec 2012, UAH6.0beta5 anomalies with those of Hada2 and… GISS land+ocean.

      NB: Dec 2012 is the end of the Hadat2 record in their yearly anomaly file
      http://hadobs.metoffice.com/hadat/hadat2/hadat2_monthly_global_mean.txt

      Following a communication of Roy Spencer this year, the average absolute temperature measured by UAH in 2015 was about 264K, i.e. -9 °C. Given a lapse of about 6.5 °C per altitude km, the NASA/NOAA satellites used by UAH should therefore operate at about 3.7 km altitude.

      According to the calculator
      http://www.csgnetwork.com/pressurealtcalc.html
      this altitude corresponds to a pressure level below 650 hPa.

      The linear trend calculated for Hadat2 for 1978-2012 is, in °C / decade, as follows:
      – 700 hPa: 0.153 ± 0.011
      – 500 hPa: 0.158 ± 0.012

      For the same period Dec 1978 – Dec 2012, UAH6.0beta5 anomalies give the linear trend
      – UAH6.0: 0.114 ± 0.009

      This is way below the radiosonde estimate: to obtain a similar temperature trend, you have to go up to a higher Hadat2 pressure level, somewhere between 300 and 200 hPa:
      – 300 hPa: 0.129 ± 0.015

      But for GISS land+ocean, you obtain
      – GISS L+O: 0.160 ± 0.069

      What means that GISS measures at the surface the same temperatures as Hadat2 at 500 hPa !!!

      And so does the comparison chart look like:

      Your comment is welcome…

  22. actually the escalator graph SKS uses is very incomplete: Bob Tisdale has a better version of it with the powerfull el nino’s in red. Odd that each step in it matches a strong el nino event.

    the simple truth SKS does not see in it is that the escalator graph points to another driver of the gradual long term warming we see: el nino, which is driven by the sun.

  23. Sheldon: There is a massive problem with this post: Uncertainty in the warming trend. If I pick selected monthly temperatures, I can find warming trends of about +0.25 and -0.25 (250 degC/century). You just can’t pick any point inside the triangle and say that the warming trend was X AND have the number X be meaningful.

    If I take a car on a 200 mile trip that takes 5 hours, I can say that my average speed was 40 mph. I could have been on a freeway averaging 60 mph for half the trip and on city streets averaging 20 mph for the other half; or I might have been on a freeway for almost the whole trip but stopped for a long lunch; or I might have been driving at exactly 40 mph for the whole trip. It is factually accurate to say that I averaged 40 mph, but that doesn’t convey very much useful information about what actually happened.

    When we use the word “trend”, we are usually talking about a rate of change that has or will persist for some period of time, something that is robust and not changing quickly. When you perform a linear regression over some period of time to get a warming trend, you are assuming that the deviations of the data point from a straight line are randomly distributed errors or noise. If so, in a replicate equipment, those deviations from linear behavior might occur at different times and produce a different trend. So a good linear regression will produce a central estimate for the trend and a confidence interval (typically 95%, 90% or 70% depends on your needs). The problem of properly calculating confidence intervals for autocorrelated data (data where this month’s deviation from linear behavior is tends to be similar to last month’s deviation) is complicated.

    If I did a linear regression of distance vs time data on my auto trip, the central estimate for my speed would be 40 mph for each trip above, but the confidence interval would be near zero for the trip at “constant” speed and much wider for the other trips – conveying a much more accurate picture of what really happened

    Your triangular graph can’t display confidence intervals. However, near the bottom of the triangle the confidence intervals are much wider than the difference between any two colors. This is certainly true for any color differences for periods shorter than 10 years and may be true up to 20 years. The differences in color in this part of your triangle are due to differences in WEATHER (including El Ninos), not due to differences in CLIMATE. Climate is often defined as a thirty-year average. If you want to say the climate is changing, you need two 30-year averages, 60-year of data. Even then you need to take into account the confidence intervals of those 30-year averages before claiming that climate has changed. Natural variability in WEATHER widens the confidence interval for average CLIMATE.

    The graph of the actual temperature vs time data accurately conveys more information to readers than any statistical summary of that data like your triangle. The triangle provides an ILLUSION of confidence in how the trend is changing at any one time, even though your mathematics is accurate. That’s why we hear the phrase: “Lies, dam lies, and statistics”.

    According to your graph, the 8-year trend around 1983 (1979-1987) is negative. If you plotted all of the data from 1970-2015 on a graph and a negative trend line 8 years long through the mean temperature for that period, people would laugh at your graph even though your analysis would be completely accurate. They would say that you were trying to fool them using statistics. And they would be right. If you properly calculated the confidence interval for the trend during this period and showed three trend lines (best estimate, upper bound and lower bound), they would say that those three lines provide better summary of what was happening in that period.

    The same thing would be true for the 8 year trend around 1996 where your triangle is bright red and orange. Put a single trend line on the data and three trend lines and ask yourself which picture accurately describes the data.

    Spend some time trying to understand the confidence intervals of the trends you plot. If you can’t properly deal with the mathematics of auto-correlation, try using annual temperature averages – there is little autocorrelation in that data. (You could probably get away with condensing down to the average temperature for each half year, the autocorrelation isn’t too bad on that time scale.) How long does a trend need to be before you can distinguish between 0.1 and 0.2 K/decade with reasonable confidence? If that period is 15 years (my guess), cut everything off the graph for periods shorter than 15 years. That would leave you only two small periods of green (slower than average warming): one around 2005 – which we call “The Pause” and one in the late 1980’s (caused by the 1982/3 El Nino being warm at one end and Pinatubo causing cooling at the other).

    The guys at SKS are misleading us with their graphs too. 1970 is a cherry-picked starting point. They are probably using a “PauseBuster” data set for their graph where some warming has been pushed from before 2000 by weighting changing sources of temperature data differently. The net weighting may or may not be justified. It is part of the uncertainty in the temperature record. Many (most?) climate scientists recognize that warming has slowed since 2000, but the magnitude of the slowing is uncertain.

    According to climate models, the upper part of the triangle should be light orange, not yellow. However, if you consider confidence intervals, we can’t be certain of that.

    Respectfully, Frank

    • “Your triangular graph can’t display confidence intervals.”

      It would certainly be hard to squeeze in more information. But you can plot the CI’s separately. I do that here (click on radio buttons to select). I also supporting masking trends not significantly different from zero, but the CIs are more general. Also you can click individual points to make it show trend and CIs.

      But even without that, the graph does help. Uniformity of colour (I would use finer gradations) at higher levels tells that you are getting some predictability.

      • Nick: Are you showing trends significantly different from zero or trends significantly different from the neighboring color?

        Thanks for the reply. Is you confidence interval corrected for autocorrelation? Given the autocorrelation associated with ENSO, I presume that correction yields only the equivalent of one or two data points per year.

        With periods of 10 years, the confidence interval is often 0.3 K/decade wide (HadCRU global). With periods of 15 years, 0.2 K/decade wide. With periods of 20 years, about 0.13 K/decade wide. Given the average warming rate of about 0.15 K/decade, these very wide confidence intervals. The wide confidence intervals span many different colors. Occasionally the lower limit for 10 and 15 year periods is occasionally greater than zero, but that happens in response to large ENSO events – in other words to what we know is noise.

        In this case, I think graphing the raw data is far more informative than the graph of the trends. If I want to know the trend for any period, I’d rather move the points around on the graph, than select a point from the triangle.

      • Frank,
        The pale colors are not significantly different from zero. The CI plots are just those 95% extremes. So if a given trend is 1±0.5, the upper plot shows 1.5 there, and the lower 0.5.

        “Is you confidence interval corrected for autocorrelation? “
        Yes. I describe that in the text, with links (more here). I use an AR(1) model with Quenouille correction.

        ” I’d rather move the points around on the graph, than select a point from the triangle”
        Yes, I do that too. The main use of selection is just to interpret the graph (or to cherrypick).

    • Hello Frank,

      This is part 1 of my reply.

      You have raised many interesting points, and I will try to answer them.

      Global warming contour maps are not directly concerned with “uncertainty in the warming trend”. Contour maps are more concerned with “the most likely warming trend”.

      To show you what I mean by this, please answer the following 2 quick questions:

      Question 1.
      ===========

      A person asks you to calculate the rate of global warming, based on 15 years of data. The person gives you the data.

      You calculate the slope using a linear regression, and get +1.5 degrees Celsius per century.

      You do a t-test on the results, and find that the slope is statistically significant.

      You tell the person the results, and they tell you that they have found a scientist who kept very accurate temperature records over the same 15 year interval that you used in your calculation. They are going to do a linear regression on the scientist’s data, and use that result for
      the rate of global warming.

      The person then makes you a special offer. You may guess what the rate of global warming will be from the scientist’s data, and if you are within 5% of the value, you will be given one million dollars.

      What value do you guess for the rate of global warming from the scientist’s data? Do you guess +1.5 degrees Celsius per century? Or something higher, or something lower?

      ===========

      Question 2.
      ===========

      In answering this question, assume that question 1 never happened. Do not use the result of question 1 to help you answer question 2.

      Question 2 is exactly the same as Question 1, except for one thing.

      When you calculate the slope using a linear regression, you get the same result, +1.5 degrees Celsius per century.

      But when you do the t-test, you find that the slope is NOT statistically significant.

      The person makes you the same special offer. You may guess what the rate of global warming will be from the scientist’s data, and if you are within 5% of the value, you will be given one million dollars.

      What value will you guess for the rate of global warming from the scientist’s data this time? The slope that you calculated is NOT statistically significant. Do you guess +1.5 degrees Celsius per century? Or something higher, or something lower?

      ===========

      My answer to the 2 questions above, is that I would guess +1.5 degrees Celsius per century for both questions. The fact that the trend in question 2 is NOT statistically significant, is not relevant to answering the question. The slope calculated for the trend in question 2 is still the best, or most likely, estimate of the true slope.

      Please tell me if you disagree with my answer.

      • Thanks for the replies. The questions are challenging.

        Sheldon says: “A person asks you to calculate the rate of global warming, based on 15 years of data.”

        I’d tell him 0.15 K/decade with an appropriate confidence interval. If I understand Nick’s response to my question, I would used his confidence intervals which are corrected for autocorrelation. I assume these are 95% confidence intervals. Then I would warn that the rate I reported isn’t particularly useful for discussing climate change because it varies significantly depending on the period chosen and climate refers to 30-year or longer averages.

        A t-test tells you when the trend is significantly different from ZERO. You provide a different color (suggesting a important difference in trend) every 0.1 K/decade. Nick shows even smaller differences in different color. One can also assess the likelihood that a trend is significantly different from 0.1 K/decade instead of 0.0 K/decade.

        I presume that is would provide an answer of 0.15 K/decade, independent of whether the confidence interval included or did not include zero (one definition of significant). However, if you try this on Nick’s trend viewer you will find the differences in trends using different sources (Had, BEST, GISS, NOAA) is often bigger than 5% – and the problem gets worse if you switch from global to land or to SST or to troposphere.

    • Hello Frank,

      This is part 2 of my reply.

      You said, “Your triangular graph can’t display confidence intervals.”

      I choose to not display confidence intervals. I could produce 2 triangular graphs, one showing the lower confidence value, and the other showing the upper confidence value. I believe that this would be a waste of time.

      I agree with you about the size of the confidence intervals near the bottom of the triangle.

      The distinction between weather and climate is arbitrary. Yes, 30 years is often used, but El Nino’s only last for 2 to 5 years, but they can dominate the “climate”.

      You said, “The graph of the actual temperature vs time data accurately conveys more information to readers than any statistical summary of that data like your triangle.”

      I disagree with this. People have trouble estimating the rate of warming from a graph of the actual temperature vs time. My global warming contour map shows a fairly accurate picture of how the rate of warming changes.

      You said, “Lies, dam lies, and statistics”.

      One of my favourite sayings, when I hear people say this, is “You can prove anything with statistics, even the truth”.

      You said, “If you plotted all of the data from 1970-2015 on a graph and a negative trend line 8 years long through the mean temperature for that period, people would laugh at your graph even though your analysis would be completely accurate.”

      Frank, people can laugh at me if they want to. Do you want me to hide the truth, or adjust it so that people will accept it, and not laugh at me.

      I believe in showing people the truth, whether they like it or not. If they don’t want to see the truth, then they shouldn’t look at my global warming contour maps. I try to act like a scientist, not a politician.

      Let me tell you why I have “confidence” in my global warming contour maps.

      When I calculate and draw a contour map (which is usually made up of over 97,000 linear regressions), it shows a logical and consistent pattern. The trends are consistent with the trends that are nearby, and the whole contour map tells a story which is usually consistent with what climate scientists say.

      Climate scientists tell us that there is more warming in the Northern Polar regions. My contour maps show that there is more warming in the Northern Polar regions. I could give you many examples of this.

      But Frank, there is more. I can draw the global warming contour maps for 14 independent temperature series, and find a similar pattern in all of them. And I am talking about totally different types of temperature measurements, Gistemp (surface measurements), and UAH (satellite measurements).

      Why would 14 totally independent temperature series show the same patterns, if the pattern wasn’t showing something that is real.

      The only alternatives are:

      – to believe that there is a conspiracy between climate scientists
      (I don’t believe that. I believe that most climate scientists are doing the best job that they can.)

      – the other alternative is to believe that there is a global warming fairy, who visits climate scientists when they are asleep in their beds, and inserts similar temperature sequences into their temperature series.
      (I don’t believe this alternative either)

      So I am left with the belief that global warming contour maps are showing, in a fairly accurate way, what has really happened to the warming rate, in the real world.

      • Sheldon: My favorite phrase description of statistics is “Meaning from Data”. If I just look at graph of temperature vs time, that conveys the most important point: temperature bounces around a lot, but has mostly gone up. So I’d add a linear regression line with slope and confidence interval. Then I’d look and ask if it has been warming less recently. I’d cherry-pick the point that shows the greatest difference between warming in early and later periods and plot those regression lines with confidence intervals. As I tried to carry out this strategy, I ran into some complications. These trends (K/century HadCRUT global) provide some information that is not readily apparent from the graph:

        75-end: 1.8 (1.6-2.0) starting before 1975 will produce slower warming rate, but much less forcing then.
        75-14: 1.7 (1.5-1.9) the recent El Nino added only 0.1 to the long term trend.
        75-98: 2.0 (1.5-2.5) or 75-07: 2.0 (1.7-2.3) most rapid rates of warming. Not meaningfully faster
        98-14: 0.5 (-0.1-+1.1) or 01-13: -0.2 (-1.0-+0.6) The Pause. Meaningfully different.
        98-end: 1.3 (0.6-2.0) The recent El Nino has ended the period that was meaningfully different. Next?
        74-84: 3.4 (2.0-4.8) Decade of rapid warming meaningfully faster than long term trend.
        84-06: 2.3 (1.9-2.8) Two plus decades of higher trend not meaningfully higher.

        I didn’t apply proper statistical test for the difference in two means (trends), so I’ve wimped out and used the term meaningfully different instead of statistically significant difference. I’m not sure which of these trend lines I would show.

        Having played around with the data some more and learned more, I might consider adding dots to your triangle to show regions where the difference from the longest term trend at the peak of the triangle is statistically significant. Nick’s trend viewer allow the lower and upper confidence intervals for the trend to be plotted and that points out periods when the trend might be significantly different from the long term trend. I’m all for showing that short-term trends differ significantly from long term trends. If looks like this is the case for some 10 year periods. So they shouldn’t be discarded. (This revises my above guesses that nothing shorter than 15 years is meaningful.) There remains the problem that 5% of trendsshould appear to be significantly different by chance.

        Sheldon asked: “Why would 14 totally independent temperature series show the same patterns, if the pattern wasn’t showing something that is real?”

        The pattern(s) is real. Should any conclusions be draw from these patterns? Is the warming rate different enough to want to assign a meaning to it – to say it must have been caused by sometime besides the random variations seen in the whole data set.

    • Frank,

      This is part 3 of my reply.

      Have a look at my website:

      mta-graphs.com

      Look at all of the UAH and Gistemp global warming contour maps. See the similar patterns that are in most of them.

      I have even made Gistemp ones for the same date range as the UAH ones, so that they are easy to compare.

      If you want to copy and paste some of the images, so that you can directly compare different contour maps, then right-click, choose “View page source”, scroll to the bottom, and you will see the links to the graphs.

      They look like this:

      Right-click the link and choose “Open link in new tab”

      If the link to the graph ends in “_d600.png” then you can usually get a bigger image by removing the “_d600”
      Leave the “.png” in the link to the graph.

    • Frank,

      This is part 4 of my reply.

      The last post turned the example of a link into the actual image. I will try to show the actual link here:

      The links look like this:
      “https://storage.googleapis.com/wzukusers/user-21443138/images/58420d289f48bI7aJiDY/North-Polar-Land-and-Ocean_d600.png”

      Right-click the link and choose “Open link in new tab”

      If the link to the graph ends in “_d600.png” then you can get a bigger image by removing the “_d600”
      Leave the “.png” in the link to the graph.

  24. Just recently, in a rather lengthy exchanged with LOUMAYTREES, he was going on about a cooling trend in relation to co2. First, from at least 2000 I have never said that there wasn’t warming. What I disagreed with is the cause. Second, if anybody does look, there seems to be an underlying warming trend. However, it is nowhere as dramatic or conform to any of the models. He wasn’t about to get me suckered me into that argument.
    The other issue is the adjusting the data. The actual warming may only 2/3 or 1/2 of what they are saying .
    The very biggest problem is that so much of focus has been on proving AGW, that in a downturn there won’t be any way of explaining it. Or making any kind of informed opinion on it that would make a difference. The ramifications are enormous.

    • Using satellite data at least before the very recent change showed there was no underlying trend noticeable. The change in the AMO easily removes any trend in global temperatures over this period.

      Any differences with surface data sets for example below were manufactured.

      • Maybe, maybe not… I don’t know. I’m using the information I had at hand. For some reason they hadn’t adjusted it until recently.
        I don’t know what they’ve done. No matter how you look at it, it’s bad science. More suited to a religious belief .
        I have an idea, but I don’t know how valid it is, I do think there has been some warming, I disagree with the cause (AGW). The early 1970′ s were scary. It was scary because food production was down and world reserves were close to running out. Whether it was weather or climate, we had back to back super harvests. Farmers were being urged to plant ” fence post to fence post “. Loans were easy and because of the superharvest and the ban on selling grain to the Soviets ( Afghanistan) a lot of farmers started going bankrupt. The reaction wasn’t immediate, but concerts to help farmers, farm aid. And in song John Melloncamp’s blood on the plow. There was a meeting in Switzerland that helped push a lot more American farmers into bankruptcy as food supplies became more stable. I still have the documents, I consider it so important that it is in a safety deposit box in a bank.
        What I know for sure is that they are lying, and hiding the information and process from the public eye. NOAA doesn’t want you to know or think anything different than what they are telling you.

      • There sure has been some warming, but by no underlying trend means after any natural contributions are taken into account.

        A strong El Nino warms the Arctic for example later significantly, so for the example above to reduce the temperatures around the strong 1997/98 El Nino. When it was supposed to be accounted for by more Arctic coverage was nonsense.

        They deliberately cherry picked warmed and cooled periods so perfect, it was obvious it was only bias confirmation by human tampering of data to support their agenda.

    • Thank you rishrac, for making my points from some time ago:

      1. IPCC AR5 (CMIP5) climate models did not reflect the known lower temperatures during 2000-2005 in their hindcasts. Had they, their out-year forecasts would have had to be cooler.

      2. IPCC AR5 (CMIP5) climate models ran so hot that AR5 had to adjust near-term (through 2035) forecasts downward because the models were so wacko hot.

      3. IPCC AR5 kept the wildly hot CMIP5 model out-year forecasts to appease the politicians. The arm-waving used was risible.

      It is my understanding that AR6 will simply accept AR5 warming conclusions and play games with negative assumptions about environmental impacts.

      Maybe President The Donald’s appointees to the international climate follies will wack some dicks.

      • You’re welcome. They don’t show the 95% certainty anymore on the models… even with the adjusted temperatures they are far below the lowest modeled numbers.

  25. Only two things missing form SkepticalScience
    Skepticism has its little more than a fanzine of climate doom
    and any science worth a dam .

    Remember its love child of ‘Real Climate ‘ and Mann , set up to give an ‘alternative’ view to the climate ‘scientists’ on how great climate ‘scientists’ are and how they are never ,ever wrong . And a love child of that pair is one not even a mother could love .

  26. John Cook”s SKS is really a psudo scientific political site dedicated to defending the AGW conjecture no matter what. Comments on their articles are sensored and many critical of the AGW conjectrure are deleted, At lease that had been by experiincing. Years ago I posted a comment over there that just demolished the AGW conjecture. They left it in because they did nto seem to understand it.

  27. The believers of AGW are going to face reality which is it never has existed, existed or will going forward.

    This is a theory in which the basic premises it was built on have failed to materialize , such as a decrease in OLR, a greater +AO evolving and the lower tropospheric hot spot missing in action.

    This theory has wasted so much time in research in the studying of this ill conceived theory (AGW) which should have been devoted instead to what really causes the climate to change and it is not CO2.

    I have wasted time studying this theory to some degree also.

    The historical climate record shows this period of time in the climate is not unique, yet this theory lives on but I think the cooling which is now happening may end it .

  28. Quote: Do they expect us to believe that global warming proceeds at a uniform constant rate?

    I can’t speak for the alarmists but I would certainly expect global warming or cooling to proceed in a very much smoother way than all these spiky graphs suggest. In particular a curve showing total heat content of the earth should be very smooth. After all it is not as though the energy provided by the sun fluctuates like the spikes on the graphs (does it?).

    My view is that whatever is being measured or calculated is far from a fundamental quantity.

    Standing by to be educated in 3.. 2.. 1..

  29. The second triangle graph looks very much like a fractal. But it would – climate is chaotic. Local disorder, global order.

    Anyway, if temperature data sets are stochastic, how can SkS overlay multiple trend lines on the graph? (AFAIK min & max data sets are not a signal, and cannot be treated as such, although that stuff is a bit advanced for me).

    • Karim Ghantous, December 8, 2016 at 10:49 pm … said:
      The second triangle graph looks very much like a fractal.

      Interesting you should say that, because looking at the tiny alarmist segment of the “escalator” and then at the “Holocene” epoch, and then at the geological-history, … I thought “embedded similarity”, you know, like in music, where overtones and undertones are embedded in the main frequency as self-similar, embedded cyclic patterns. [Am I straying into la-la-land-at-risk-of-being-deleted territory yet?]

  30. When the Earth is going through a period of solar induced warming there will be a step up in temperatures from one positive ENSO phase to the next.
    When the Earth is going through a period of solar induced cooling there will be a step down in temperatures from one negative ENSO phase to the next.
    We are likely at the transition point between upward stepping and downward stepping.

    • Right, except this is true with each and every possible cause, not just “solar induced”.
      Meaning opposing the “escalator” and the “single straight” is nonsense.

      What’s make the “escalator” relevant is just the fact that they ruled out the possibility of having a step 17y or more long. But it showed itself.

      We don’t know enough of climate to be sure of any future. Could be going downward, or resume upward, or whatever.
      What we do need, however, to falsify the sc4m, is that human keep dumping lots of CO2 in the atmosphere so. If CO2 stop going up despite that, the CAGW hypothesis is shown wrong. Likewise if temp stop going up, or even go down.
      We may be unlucky, however.

      • paqyfelyc

        Observing other parameters such as the timing of global albedo changes as per the Earthshine project relative to changes in solar activity should help to pin it down to solar changes.

        Earlier Earthshine data from before 2000 showed declining
        albedo until around 2000 and a slight recovery after 2000. I have now found data
        for the period to date:

        https://arxiv.org/pdf/1604.05880v1.pdf

        See Figure 2

        Overall, albedo has been roughly steady since 2000.

        A plausible interpretation is that after 2000 the solar influence on the jet
        stream tracks caused global albedo to recover from the pre 2000 low figures
        which is what caused the current ‘pause’ in the global temperature trend.

        To resume warming, global albedo needs to drop again and if albedo increases
        further then cooling should commence.

        The relatively weak solar cycle 24 has been able to cause the pause but it
        seems that we may need an even weaker cycle 25 to increase albedo further to
        tip the system into discernible cooling.

        It is possible that the thermal inertia of the oceans has been sufficient to
        delay the commencement of cooling which might otherwise have already
        commenced at the current level of albedo.

    • Stephen Wilde on December 9, 2016 at 4:02 am

      When the Earth is going through a period of solar induced warming there will be a step up in temperatures from one positive ENSO phase to the next.
      When the Earth is going through a period of solar induced cooling there will be a step down in temperatures from one negative ENSO phase to the next.

      Wow! Is this
      – a theory
      – a claim
      – a supposition ?

      Maybe you had this impression when comapring UAH and ENSO for the period 1997-2016…

      But a look at a chart comparing temperatures with ENSO signals for the period 1979-2016

      shows that you are a bit wrong (because you ignore for example volcanic forcings aka stratospheric aerosols). Some periods might let you indeed imagine it goes that way.

      And a look at a similar chart for 1871-2016

      should convince you that your assumption does not reflect reality. The temperature response to MEI over 150 years is simply too random.

      • you made the wrong comparison graph: the escalator says something else then your graph.

        you forget that la nina is not the opposite of el nino. even if this graph is ocean only, adding land surface to it doesn’t change it.

        el nino’s are discharges of warm water which then remain lingering around, while la nina’s are recharges of the warm pool where El nino originates from

        that means la nina’s do not discharge cold waters unlike el nino does.

        that’s a big difference that seems to be forgotten a lot in the ENSO debate. this means that each big el nino is able to ramp up the temperature if they follow in good sequence

        like this back to back la nina’s can hold a brake on the el nino.

        then to finish there is maybe a correlation with ENSO and PDO, but that’s not known in what way and how it works. that part is still heavily debated.

        i’m pretty sure that if we had accurate el nino data in the first warming period and cooling period we would see a similar escalator…

      • forgot to add: we do see since 1975 a sudden change from ENSO turning from more “la nina state” to more “el nino state” In the way ENSO works we now do see a discharge of the heat stored in the warm pool from the late 40’s till mid 70’s.

        Maybe that this correlates with the most recent warming is just coincidence?

      • Frederick,

        El Nino releases warmth to the atmosphere over and above that from incoming solar energy.
        La Nina reduces warming of the atmosphere by diverting incoming solar energy to the recharge process.
        In thermal terms as far as the air is concerned they are opposites.

        Over the period MWP to LIA I would expect that La Ninas became dominant relative to El Ninos and since the LIA it has been the opposite.

      • Even Judith curry did do a nice follow up on this by writing an interesting article about it: how El nino and La nina do “add up” and comparing it to the global temperature….

        article of Judith curry

        i agree that you can say that for the thermal terms of air they are “opposite”.

      • “forgot to add: we do see since 1975 a sudden change from ENSO turning from more “la nina state” to more “el nino state” In the way ENSO works we now do see a discharge of the heat stored in the warm pool from the late 40’s till mid 70’s.
        Maybe that this correlates with the most recent warming is just coincidence?”

        Not a coincidence when it can easily be seen the solar energy ramped up between the 1940’s and 1960’s. The atmosphere was unable to see this change quickly because numerous La Nina’s kept it away from the surface until later. Later in the 1970’s the Pacific Ocean shift change occurred and the warm AMO came with it.

        The difference in warmth recently compared to the warming up to the 1940’s was higher solar activity and the AMO/PDO both being positive at the same time with more El Nino’s of course.

      • The MEI is now useless for comparing with global temperature trends because they already assume global warming is causing the higher NINO 3 and NINO 4 SST’s. Unfortunately so they adjusted it recently to take this into account. It is very unscientific and very on the assumption side with no evidence El Nino’s having anything to do with humans.

  31. “Do they expect us to believe that global warming proceeds at a uniform constant rate?”

    Yes – because CO2 is increasing at a mostly uniform and constant rate since the industrial revolution. For their theory to hold true, the only reason for any increase in mean global temperatures must be mankind’s release of CO2 into the atmosphere. Which on its face value is laughable that the entire climate of the Earth hinges solely on CO2 levels.

  32. John Cook deleted my Escalator Gif from the escalator thread on SkS, then he deleted my SkS account, then he wrote a polite email asking me who created it? When I told him I did he told me not to bother trying to comment at SkS again.

    Touched a nerve I did

  33. Pseudo skeptical pseudo scientific site. I’ve met their climate warriors on facebook. They are really brainwashed and very aggressive. They deny logic, they deny physics, they deny the scientific method, they deny everything that’s against their religious doctrine. I quickly learned to put them on ignore, one cannot have a rational discussion with them :)

    • I have also had a less than worthwhile experience on their psudosceintific, political, AGW religious site. It is amayzing what they will allow in terms of comments if you start it off in such a way that you appear to be agreeing with them. I was able to post one comment that just blew away AGW if one bothered to read the whole thing. Maybe they have actually discovered and deleted it by now.

Comments are closed.