Comments 1 to 19:

1. hank at 02:58 AM on 20 October, 2014

   Aside — you'll find many 'septic' sites posting a partial quote from Al Gore's 2007 Nobel acceptance speech (one way to identify that sort of misattribution and spin is to look for their copypaste repeats of the partial quote).

   Here it is in context:

   "Last September 21, as the Northern Hemisphere tilted away from the sun, scientists reported with unprecedented distress that the North Polar ice cap is "falling off a cliff." One study estimated that it could be completely gone during summer in less than 22 years. Another new study, to be presented by U.S. Navy researchers later this week, warns it could happen in as little as 7 years."

   The US Navy research:  http://scholar.google.com/scholar?q=maslowski+2007+arctic&hl=en&as_sdt=0%2C5&as_ylo=2007&as_yhi=

2. Tom Curtis at 10:50 AM on 23 August, 2016

   Elsewhere, victorag is arguing that a "pause" exists, basing the argument on Fyfe et al (2016).  He ignores Werner et al (2015) (discussed here).  Werner et al first determine the locations of the break points in the piecewise trends in GMST for a forced number of breakpoints up to eight.  Breakpoints only appear around 1998 or later if you force seven or more breakpoints:

   

   They further apply statistical tests to determine the statistical support for each number of breakpoints, finding the best statistical support for three breakpoints, with potential support for two, four, or five breakpoints.  There is little statistical support for seven or eight breakpoints, and hence little statistical support for a 21st century slowdown:

   

    Fyfe et al (2016) do not include Werner et al in their references, and therefore ignored this evidence.  Further, their rejection of statistical tests showing no change in underlying trend because of their supposedly extended baseline, which is true of the IPCC AR5, but not true of all such statistical tests.  Using their preferred intervals for the "big hiatus" and the "slowdown" and updated NOAA (Karl 2015) data, we can see that the Jan 1972- Dec 2000 interval has a trend of 0.171 +/- 0.061 C/decade.  The Jan 2001- Dec 2014 interval has a trend of 0.078 +/- 0.140 C/decade.  That is, the trend for the second interval includes the trend for the first interval in its uncertainty, and therefore the null hypothesis that the trend has not changed cannot be rejected.  (Note that it can be rejected using HadCRUT4, but that is because HadCRUT4 has limited coverage, particularly in the Arctic and North Africa, as can be checked by looking at the HadCRUT4 Krig data.)  (See also Cahill et al (2015), also ignored by Fyfe et al.)

   In short, Fyfe et al insist on a slowdown because (as they say) their "exploration of an alternative baseline period is motivated by ΔF, the estimate of anthropogenic radiative forcing" rather than because of any statistical evidence of a slowdown.  They have either mistated or ignored the statistical evidence that, based on the temperature series, no slowdown exists (Werner et al), or at least that no slowdown has been demonstrated (statistical uncertainty on the trends of their chosen intervals).

   Worse for victorag is that even if we accept a slowdown in the trends, it was actually predicted by the CMIP5 computer models.  Fyfe et al show the following graph:

   

    The black line is the running fifteen year trend, while the grey shaded area is one standard deviation from the mean of the CMIP5 running fifteen year trend.  The clear dip in the predicted running fifteen year trend is easilly seen.  More importantly, the GISTEMP running fifteen year trend skirts the botton of the 1 SD shaded area, showing that it is easilly within the 2 SD prediction zone.  And if the data is within the prediction range of the model ensemble, the model ensemble is not falsified by the data.  (Further discussion here.)

3. victorag@verizon.net at 11:49 AM on 23 August, 2016

   With regard to Fyfe et al., see the very interesting blog post and discussion at Ed Hawkins' blog. See also the Guest Post by Fyfe et al. on the same blog, followed by many very interesting comments as well.

   I won't comment further until I receive a reassurance that my comments will not be deleted or censored, as I see no point in wasting my time. 

   Response:

   [JH] All commenters on this site are required to abide by the SkS Comments Policy.

4. victorag@verizon.net at 02:10 AM on 24 August, 2016

   (My apologies again: the links I provided got lost when I re-pasted this post. Please delete post #4. Thank you.)

   For reasons already stated, I will be brief, so as not to waste my time with long explanations that might get edited or deleted. Maybe it's better that way anyhow.

   Regarding the difference between "eyeballing" and statistical analysis: there are many ways to manipulate results using statistics, whereas one's eye sees the data directly. Sorry, but where the picture is clearly there for all to see, I'll trust my eye, thank you.

   Regarding statistical smoothing:

   Smoothing creates artificially high correlations between any two smoothed series. Take two randomly generated sets of numbers, pretend they are time series, and then calculate the correlation between the two. Should be close to 0 because, obviously, there is no relation between the two sets. After all, we made them up.

   But start smoothing those series and then calculate the correlation between the two smoothed series. You will always find that the correlation between the two smoothed series is larger than between the non-smoothed series. Further, the more smoothing, the higher the correlation. (From the blog of Dr. William M. Briggs (PhD in Mathematical Statistics))

   Tom Curtis at #119: Regarding scattergrams: the scattergram offered by Tom lacks sufficient detail to be very useful. Here's another that does, compiled by Danley Wolfe from raw data available to all at:

   Mauna Loa: ftp://aftp.cmdl.noaa.gov/products/trends/co2/co2_mm_mlo.txt

   NASA GISS: http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts.txt

   Wolfe does nothing to "massage" his data, it's directly transcribed from the two sites referenced above.

    

5. victorag@verizon.net at 02:35 AM on 24 August, 2016

   #2 Tom Curtis. I'm not qualfied to distinguish between the claims of Werner and Fyfe, but I do find myself agreeing with Fyfe when he comments on the value of a purely statistical analysis:

   Statistical analysis is a vital tool in any climate scientist’s toolbox. However, even the application of sophisticated statistical tools can shed more heat than light, particularly in arguments that focus on limited aspects of statistical significance rather than on broader physical understanding. . . 

   These results illustrate the dangers of relying solely on a statistical test to tell us whether there is, or is not a physically-based change in warming rates.

   Regardless of which view is the correct one, such disputes shed light on the problems faced by climate scientists intent on proving their thesis beyond reasonable doubt. It's all too easy to fall back on the notion that it's now up to skeptics to provide the evidence they themselves have been unable to provide.

6. MA Rodger at 03:53 AM on 24 August, 2016

   Victor Gauer @6.

   I was wondering how long you would take to introduce Danley Wolfe's fabrication into SkS. His MLO CO2 data is as stated but is unadjusted for the annual cycle. And the GISTEMP data isn't as Wolfe states. It is not "NASA GISS global mean (absolute) land temperatures" as Wolfe claims. It is the NASA GISS LOTI anomaly with 14ºC added. (In detail, its origins appear to be a 'collection' of LOTI data. It is certainly not the LOTI data as published in May 2014.) The steep slopey line Wolfe draws on to his graph has no basis other than arbitrarily joining up the two flat unslopey lines, which are themselves not the cherry-picked OLS results that Wolfe claims. When I have a moment I shall illustrate graphically how dreadful this nonsense from Mr Danley Wolfe truly is.

   And after a quick google, Dr William M Briggs appears not to be a good expert to rely on uncritically. Specifically here, what does he mean by "smoothed"? Is it a concern for anything under discussion here?

7. victorag@verizon.net at 04:23 AM on 24 August, 2016

   #7 MA Rodger

   If Wolfe's scattergram doesn't suit you, why not produce one that does? But if you do, please use raw data, not data that's been statistically modfied to produce the desired result. 

8. Tom Curtis at 05:37 AM on 24 August, 2016

   victorag @8, I have responded to your comments on CO2 correlation on the original (and appropriate) thread.  In the comment I have shown that:

   Using the data as downloaded (Mauna Loa monthly plus BEST LOTI) increases the correlation fractionally.

   And that Wolfe cherry picks a restricted temperature data set which artificially deflates the correlation. 

   I have also shown a new scatter plot satisfying your stricture that the data as downloaded be used (except for interpolation of missing months).

    

   MA Rodger, regardless of Briggs merits or otherwise, correlation is covariance divided by the product of the standard deviations of the data.  Smoothing reduces the standard deviation, and therefore must increase the correlation.  This was something I was quite aware of which was why my primary analysis used monthly data to obviate any issue of inflating the correlation by smoothing.  As I went out of my way to avoid artificially inflating the correlations, I take exception to Victor's suggestions that my mathematics has been manipulation rather than analysis.

9. victorag@verizon.net at 05:40 AM on 24 August, 2016

   Here's another study, dating from 2014, thus prior to Fyfe et al., in which the "hiatus" is taken seriously — only in this case the authors are unable to account for it: Application of the Singular Spectrum Analysis Technique to Study the Recent Hiatus on the Global Surface Temperature Record

   Some excerpts:

   From the abstract:

   Global surface temperature has been increasing since the beginning of the 20th century but with a highly variable warming rate, and the alternation of rapid warming periods with ‘hiatus’ decades is a constant throughout the series. . . 

   Henceforth, MDV [multidecadal variability] seems to be the main cause of the different hiatus periods shown by the global surface temperature records. However, and contrary to the two previous events, during the current hiatus period, the ST [secular trend] shows a strong fluctuation on the warming rate, with a large acceleration (0.0085°C year−1 to 0.017°C year−1) during 1992–2001 and a sharp deceleration (0.017°C year−1 to 0.003°C year−1) from 2002 onwards. This is the first time in the observational record that the ST shows such variability, so determining the causes and consequences of this change of behavior needs to be addressed by the scientific community.

   From the Discussion section:

   After the maximum warming rate associated with MDV was reached by approximately 1990, ST showed a distinct peak from 1992–2001, with an unprecedented increase of its warming rate from 0.0085°C year−1 to 0.017°C year−1, almost doubling in one decade. After this warming rate peak, the ST shows a pronounced decline, 0.017°C year−1 in 2001 to 0.003°C year−1, in 2013. This type of quick fluctuations in the ST warming rate has no precedent in the observational record (Fig. 3a). . . 

   Therefore, the very recent strong changes observed in the warming rate associated with the ST appear to be a global phenomenon that had not occurred before (at least not during the last 160 years). It could not be attributable to MDV or any other form of climatic variability (such as solar cycles), as the different contributions are effectively separated by the SSA analysis (Fig. 2). This unprecedented modification of the ST behavior should be more deeply studied by the scientific community in order to address whether a change in the global climate sensitivity [21] has recently occurred.

   Here's an article on these findings from the European Commission's website: Last decade's slow-down in global warming enhanced by an unusual climate anomaly

10. Tom Curtis at 05:45 AM on 24 August, 2016

    victorag @6, I quite agree that:

    "Statistical analysis is a vital tool in any climate scientist’s toolbox. However, even the application of sophisticated statistical tools can shed more heat than light, particularly in arguments that focus on limited aspects of statistical significance rather than on broader physical understanding. . ."

    You, however, cannot consistently do so.  First, that is because it means Fyfe et al's acceptance of evidence of a change in forcing leading to a slow down logically precedes their acceptance of the existance of a slow down.  Logically, therefore, it is not an attempt to plug up a theory and make it resistant to contrary observations (as you have elsewhere suggested).  Either the evidence of reduced forcing that justify belief in a slow down is sound (in which case your attempt to impugn it is ill motivated), or it is not (in which case it cannot be relied on to infer the existence of a slowdown, where the statistical evidence is inadequate to justify that inference).

    Further, your argument against the temperature effect of CO2 is entirely statistical.  (Bad, and cherry picked statistics, but statistical never-the-less).  But if you accept the principle above, you cannot rely on entirely statistical arguments in judging the effect of CO2.

    In short, your defence of Fyfe et al based on that quote is, for you, a matter of tactical convenience only as there is no evidence you apply that principle more generally.

11. victorag@verizon.net at 05:58 AM on 24 August, 2016

    #9 Tom Curtis

    I responded to your last post, with the new scattergram, on the other thread. To my eyes your result isn't that different from Wolfe's.

12. victorag@verizon.net at 06:17 AM on 25 August, 2016

    #10 Tom Curtis: "Further, your argument against the temperature effect of CO2 is entirely statistical."

    I never said I rejected statistics. My point, in agreement with Fyfe's point, is that it is not always the best tool in all circumstances and can sometimes distort the physical reality behind the raw data. 

13. Rovinpiper at 21:34 PM on 16 September, 2017

    I don't understand the concept of the warming of the entire climate system. If the oceans, for instance, are accumulating energy is that reflected as an increase in their temperature? 
    
    The first figure shows energy accumulation. I think the units are Zettajoules. Is that right? How is energy accumulation measured?

14. scaddenp at 07:53 AM on 17 September, 2017

    Yes, the oceans are warming, and the energy accumulatation is calculated by integrating temperature change over depth. 

15. Rovinpiper at 21:01 PM on 17 September, 2017

    Okay, so that means that temperature is measured and the energy is calculated from the temperature, mass, and specific heat of the media in question, right?
    
    It's strange to me that the energy accumulation is below zero at c. 1971. Negative energy isn't possible, is it? Am I right to conclude that it has to be an artifact of the comparatively huge uncertainties in the measurements at that time?
    
    The energy content of ice is not something that I'd ever thought about before, but it makes sense. Earth can get much colder than 273 K, so much of the ice has considerable warming up to do before it melts.

16. Eclectic at 22:10 PM on 17 September, 2017

    Rovinpiper @15 , the IPCC chart you mention has the basic purpose of showing the alteration in accumulated energy — so you must expect a "negative" level as you go back in time.   If you were thinking of the tiny "negative dips" around 1980 and 2000 etc, then you will have noticed that they are insignificant compared with the uncertainty bounds of the measurements.

    Possibly a real "negative dip" could occur for a year or two if there were some very major volcanic eruption that shot enough fine [reflective] particles into the upper atmosphere.  But that would be brief, and global warming would soon resume.  When you think it through, you will see that the warming greenhouse effect of [primarily] CO2 is causing a heat inflow into the planet [an inflow averaging approximately 2 watts per squ.meter or in other words roughly 7,000 horsepower per squ. mile] and this is going on 24/7 and year round . . . and will continue until Earth reaches a new higher equilibrium temperature (many years after we have achieved zero net CO2 emissions in [hopefully] year 2050 or 2060 — or maybe 2080, the way our politicians are dawdling over the emissions problem! ].

    And you will recognise that a genuine pause is impossible, under the present conditions of ongoing CO2 emission.  That's why any "pause" can only be a Myth !

17. scaddenp at 07:05 AM on 18 September, 2017

    Further to that, note that I said OHC was calculated as temperature change integrated over depth. ie it is always with respect to a baseline. The methodological paper describing OHC say this right at the start: "We use the term “ ocean heat content” as opposed to “ocean heat content anomaly” used by some authors because “ocean heat content” is an anomaly by definition. OHC is always computed with a reference mean subtracted out from each temperature observation. Otherwise the OHC computation depends on the temperature scale used."

    Thus you obviously get "negative" OHC when temperature is less than baseline. I strongly recommend you have a read of the paper. This was calculation done in 2012, but you can follow results from same methodology published here.

18. Rovinpiper at 20:34 PM on 18 September, 2017

    I strongly recommend you have a read of the paper.

    Thanks. I'll do that.

19. braintic at 16:22 PM on 20 January, 2019

    As this presentation shows, even the UAH data shows that there was no pause in warming:

    UAH Comparison by Decade

    Response:

    [PS] Fixed link

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.