2021 06 03 CSIRO answers / archived / read-only

 
  • Part of: Points on Climate Change
  • Introduction
  • Summary
    • The government has repeatedly stated it relies on CSIRO for advice on climate change. We asked CSIRO for logical points to support their assertion that the current rate of temperature rise is unprecedented for 10,000 years.
    • By logical point we mean a statement that is supported by text quoted from a specific page in the cited reference. References should ideally provide the raw and intermediate data and associated logical reasoning for review and potential duplication of the calculations or analysis steps.
    • The following analysis of CSIRO's letter shows that it contains no scientific logical points and relies instead upon unsubstantiated assertions, irrelevant citations and 3 allegedly flawed papers.
    • Conventions used in this document;
      • Direct quotes written by CSIRO are in green.
      • Direct quotes from a paper cited by CSIRO are in blue.
  • Examples of unsubstantiated assertions which CSIRO has not addressed by 2021 06 03. These need to be addressed by a new Office of Scientific Integrity which would police unsubstantiated assertions in documents intended justify government spending.
    • The conclusion that the rate of temperature change today is unprecedented in the past 10,000 years is supported by many studies and multiple lines of evidence.
    • Without links to Logical Points in some of the studies and lines of evidence this is an unsubstantiated assertion.
    • There are no peer-reviewed published studies that contradict this conclusion, to our knowledge.
    • Without listing the databases searched and the search terms used this is an unsubstantiated assertion.
    • The most recent reconstruction of global temperatures during the Holocene (Kaufmann et al., Scientific Data, 2020) is consistent with that of Marcott et al. (2013)
    • Without links to Logical Points in both studies this is an unsubstantiated assertion.
    • In relation to claims made in blog posts regarding the integrity of the Marcott study, and the questions raised regarding this research, it should be noted that there are currently 265 [>640, as of October 2020] published papers in the peer-reviewed literature citing Marcott et aI (2013),
    • Citations in no way validate a reference - Appeal to the Majority - Argumentum ad populum
    • CSIRO reaffirms the scientific rigour of the Marcott et al. (2013) study and the conclusion that the rate of recent warming of global mean temperature is unprecedented in the past 10,000 years.
    • Marcott 2013 did not contain the conclusion that the rate of recent warming of global mean temperature is unprecedented in the past 10,000 years. It did not contain any "conclusions" at all. Marcott 2013 did have "our results indicate" but nothing about measured centennial rates of temperature change during the 10,000 years
    • and there is no published study that draws the temperature reconstruction into question for the period for which sufficient proxies are available (~11,300 BP to 1800 CE).
    • Without listing the databases searched and the search terms used this is an unsubstantiated assertion.
  • The allegedly flawed papers
  • Marcott 2013
    • The conclusion drawn by CSIRO and by the Marcott study is based on a comparison of recent temperatures, as measured by instruments, to past temperatures inferred from proxy data. Neither Marcott, nor CSIRO, make any conclusions based on the "uptick" part of the proxy temperature record. This approach is used because the best information we have about recent temperatures are those made by instruments measuring temperature (thermometers measure temperature more accurately than tree rings, ice cores, or plankton preserved in sediments).
    • Marcott's Gain Function Plots show no response is possible for events shorter in duration than 300 year periods such as the current 40yr warming from 1970 as alleged in Mann 2008.
      • Mann 2008 shows about 0.9deg in 40years which is a small fraction of a 300 year period.
      • Q Does CSIRO understand Gain Function Plots? The ratio of output to input by frequency band.
        1. Yes
      • Q Is CSIRO aware of the Gain Function Plots in Marcott 2013?
        1. Yes
      • Q Is CSIRO aware of Marcott's statement about the Gain Function Plots: The results suggest that at longer periods, more variability is preserved, *with essentially no variability preserved at periods shorter than 300 years***,
      • This means that any temperature changes whose duration is less than 300 years will not be seen.
      • See Appendix - Marcott 2013 Gain Function Plots
        1. Yes. CSIRO note that the statement that little variability is preserved in frequency space is not the same thing as a statement about the duration of detectable events: a temperature spike (say a rapid rise and fall of temperature over 100 years) contains substantial energy at periods longer than 300 years, and therefore would have energy that would pass through this filter. The signal will be damped on passing through the filter, so whether the signal is detectable after filtering depends on the magnitude of the input signal. The recent observed temperature rise is very large compared to inferred variability of the early Holocene (about 50 standard deviations), so this signal would likely be detected after filtering.
        • In addition, CSIRO note that none of the individual proxies used by Marcott et al. show evidence of a past change in temperature to rival the rise in measured temperatures over the past century; it is unlikely that a global temperature increase and decrease with a magnitude to rival the observed rise over the past century would have been missed by every one of the individual proxy records used. Furthermore, there is no scientific evidence in any other climate variable to support the supposition that there may have been periods of rapid warming in the past that were missed by Marcott’s reconstruction method.
        • Finally, no physical mechanism has been identified to drive such a large and rapid rise (and fall) of temperature in the past
        • A potentially more significant issue with the analysis of Marcott et al. (2013) and Kaufmann et al. (2020) is the influence of seasonal bias in the proxy records used (i.e. many of the proxies may reflect summer, rather than annual mean, temperatures). Bova et al. (2021) show that correcting for this seasonal bias results in lower temperatures in the early- to mid-Holocene, leading the authors to conclude that modem annual mean temperatures exceed those of the past 12,000 years.
      • Q Can CSIRO see that, on the basis of the Gain Function Plots, Marcott's process cannot detect any changes with periods of less than 300 years?
        1. Yes see answer at 3 above.
        1. No see answer at 3 above.
    • Marcott 2013 contains illogical reasoning comparing short term minimums and maximums to long term means.
      • Q If a daily minimum and maximum temperature are known, can the day's temperature be declared unprecedented by comparing it to the annual mean temperatures for 100 years? Can CSIRO see that this would be absurd?
        1. CSIRO agree that trends cannot be detected by comparing a daily record to a 100-year mean.
      • Our results indicate that global mean temperature for the decade 2000–2009 (34) has not yet exceeded the warmest temperatures of the early Holocene (5000 to 10,000 yr B.P.). These temperatures are, however, warmer than 82% of the Holocene distribution as represented by the Standard5×5 stack, or 72% after making plausible corrections for inherent smoothing of the high frequencies in the stack (6) (Fig. 3). In contrast, the decadal mean global temperature of the early 20th century (1900–1909) was cooler than >95% of the Holocene distribution under both the Standard5×5 and high-frequency corrected scenarios. Global temperature, therefore, has risen from near the coldest to the warmest levels of the Holocene within the past century, reversing the long-term cooling trend that began ~5000 yr B.P.
      • Q Can CSIRO see that the above statement contains this absurd logic and is misleading?
  • Lecavalier 2017
    • Proxy data was used when more accurate thermometer data was available from multiple Arctic sites.
      • From CSIRO Letter 28 October 2020 to Senator Malcolm Roberts: The conclusion drawn by CSIRO and by the Marcott study is based on a comparison of recent temperatures, as measured by instruments, to past temperatures inferred from proxy data. Neither Marcott, nor CSIRO, make any conclusions based on the "uptick" part of the proxy temperature record. This approach is used because the best information we have about recent temperatures are those made by instruments measuring temperature (thermometers measure temperature more accurately than tree rings, ice cores, or plankton preserved in sediments).
      • We agree with CSIRO's opinion and have replaced the proxy data in Lecavalier 2017 by instrumental temperatures from GHCN Arctic stations after 1937.
      • Dr Lecavalier provided his very useful datasets by personal communication which enabled us to duplicate some analysis and to splice our All Arctic dataset instead of the short 2009 proxy. We would be happy to provide this data to CSIRO subject to Dr Lecavalier's permission.
      • Fig. 3.(C) Rate of temperature change based on the Gaussian filtered Agassiz temperature reconstruction |100%
      • Rate of temperature change based on the Gaussian filtered Agassiz temperature reconstruction with our All Arctic series replacing the short 2009 ice core. There is nothing unusual in the 100 year trends and recent trends are not unprecedented.
      • See Appendix - All Arctic Thermometer Series
    • In Lecavalier 2017 Error estimates were not carried through the Gaussian filter or the 100yr running means. This is a fundamental omission. The plot as drawn by Lecavalier implies no uncertainty.
      • Fig. 3.(C) Rate of temperature change based on the Gaussian filtered Agassiz temperature reconstruction
      • Shows the error estimates of the filters included. Filtering makes the plot look cleaner but the error estimates can grow rapidly and should not be omitted. This shows the possibility that recent rates of change of temperature may actually be cooling.
    • There was no mention of Heteroscedasticity in the Methods.
      • The paper presents data from ice cores drilled from the Agassiz Ice Cap in the Canadian Arctic. From the ice cores, past temperatures can be inferred by a complicated process with varying degrees of uncertainty. Raw temperatures at 25yr intervals were available from two cores: 1997 and 1999 (apparently used but not mentioned) . The 1984 and 1987 cores yielded 1yr data and the short 2009 core yielded 1yr data but with higher variance than 1984 and 1987.
      • Lecavalier's raw data showing the difference in variance between the cores.
      • There was no mention of Heteroscedasticity in the Methods. The sigma bounds on the 25yr temperature would likely be significantly greater if heteroscedasticity had been considered.
      • The plot of centennial rates of temperature change does not contain sigma bounds. If ARCH methods were used then the sigma bound would likely have been very wide and the impact much reduced.
    • Do you expect the apparent warming of the last 100 years to continue?
        1. Yes. As anthropogenic greenhouse emissions have caused the earth to warm, further emissions will cause further warming.
      • Is the Lecavalier 25yr temperature data serially correlated?
        • Durbin Watson statistic: 0.072 < 1.76 Serially correlated Ljung Box lags=10 Max pValue 0.0 <= 0.05 Serially correlated Serially correlated OLS model inapproprite
          1. CSIRO are not aware of any published work that contradicts the conclusions reached by Lecavalier et al. (2017) and we would encourage anybody who has such information to publish their analysis in the peer reviewed literature.
      • Does serial correlation effect the interpretation of least square means?
        • An ARIMA model was fitted to the data and showed good residuals and error bounds on the forecast. Therefore these ice cores give very little indication of what the future temperatures of the Agassiz Ice Cap might be.
          1. See answer 9 above, ( CSIRO are not aware of any published work that contradicts the conclusions reached by Lecavalier et al. (2017) and we would encourage anybody who has such information to publish their analysis in the peer reviewed literature.)
    • What length and rate of warming would CSIRO consider normal when exiting a Little Ice Age and 3 solar minimums?
      • End of MWP Medieval Warm Period LIA Little Ice Age SM Sporer solar minimum MM Maunder solar minimum DM Dalton solar minimum
        1. The answer to this hypothetical question depends on the forcing and feedbacks.
  • Kaufman 2020
    • Does CSIRO stand by Kaufman 2020 as supporting a recent unprecedented rate of change of temperature?
      1. Yes.
    • Did CSIRO look in the public domain for prima facie evidence of issues with Kaufman 2020?
      1. Yes.
    • Is CSIRO aware that the recent uptick occurs only in the Antarctic latitude band and no other latitude band?
    • Is CSIRO aware that Kaufman 2020 disagrees with PAGES 2k 2013 Antarctic reconstruction which has no uptick?
    • Is CSIRO aware that only one of the 16 proxies in the Antarctic latitude band showed the uptick and that was a borehole and the other 15 more reliable methods (including ice cores) did not?
    • Is CSIRO aware that the original borehole LawDome Dahl-Jensen1999 paper did not show an uptick?
      1. Answer to questions 14-19, referring to “the recent uptick in temperature in the Kaufman et al. (2020) paper”: We did not refer to a recent uptick in the proxy temperature record in our correspondence. As discussed above, and in previous correspondence with you, the papers cited by CSIRO use the instrumental record, not proxy data, to determine recent temperatures. The recent increase in global mean temperature shown in black in Figures 2 and 3 of Kaufman et al. (2020) is from the instrumental record, not proxy measurements.
    • Is CSIRO aware that the error in importing the Dahl-Jensen borehole data was to miss the first data point and to load the remaining data points in the reverse time order?
    • We agree that when erroneous upticks are removed from Marcott 2013 and Kaufman 2020 they are consistent with the conclusion of no unprecedented recent temperature changes.
    • What due diligence did CSIRO do on this paper before submitting it to the Senate and misleading the Senate?
    • Is CSIRO concerned that Kaufman 2020 passed peer-review in Nature?
  • Specific questions and CSIRO answers
    • Processed
        1. Does CSIRO understand Gain Function Plots (the ratio of output to input by frequency band)?
        1. Yes
        1. Is CSIRO aware of the Gain Function Plots in the Marcott 2013 paper referenced in CSIRO's 28th October 2020 letter to me?
        1. Yes
        1. Is CSIRO aware of the statement about the Gain Function Plots in Marcott's 2013 paper referenced in CSIRO' s 28th Octo her 2019 letter to me that said: "The results suggest that at longer periods, more variability is preserved, with essentially no variability preserved at periods shorter than 300 years"?
        1. Yes. CSIRO note that the statement that little variability is preserved in frequency space is not the same thing as a statement about the duration of detectable events: a temperature spike (say a rapid rise and fall of temperature over 100 years) contains substantial energy at periods longer than 300 years, and therefore would have energy that would pass through this filter. The signal will be damped on passing through the filter, so whether the signal is detectable after filtering depends on the magnitude of the input signal. The recent observed temperature rise is very large compared to inferred variability of the early Holocene (about 50 standard deviations), so this signal would likely be detected after filtering.
        • In addition, CSIRO note that none of the individual proxies used by Marcott et al. show evidence of a past change in temperature to rival the rise in measured temperatures over the past century; it is unlikely that a global temperature increase and decrease with a magnitude to rival the observed rise over the past century would have been missed by every one of the individual proxy records used. Furthermore, there is no scientific evidence in any other climate variable to support the supposition that there may have been periods of rapid warming in the past that were missed by Marcott’s reconstruction method.
        • Finally, no physical mechanism has been identified to drive such a large and rapid rise (and fall) of temperature in the past
        • A potentially more significant issue with the analysis of Marcott et al. (2013) and Kaufmann et al. (2020) is the influence of seasonal bias in the proxy records used (i.e. many of the proxies may reflect summer, rather than annual mean, temperatures). Bova et al. (2021) show that correcting for this seasonal bias results in lower temperatures in the early- to mid-Holocene, leading the authors to conclude that modem annual mean temperatures exceed those of the past 12,000 years.
        1. Does CSIRO believe the process used in Marcott's 2013 paper referenced in CSIRO's 28th October 2020 letter to me can detect any changes with periods of less than 300 years?
        1. Yes see answer at 3 above.
        1. Can CSIRO see that, on the basis of the Gain Function Plots, the process used in Marcott's 2013 paper referenced in CSIRO's 28th October 2020 letter to me cannot detect any changes with periods of less than 300 years?
        1. No see answer at 3 above.
        1. If a daily minimum and maximum temperature are known, can the day's temperature be declared unprecedented by comparing it to the annual mean temperatures for 100 years? What is CSIRO's response to a claim that doing this would be absurd?
        1. CSIRO agree that trends cannot be detected by comparing a daily record to a 100-year mean.
        1. Does CSIRO agree that the statement in Marcott's 2013 paper referenced in CSIRO's 28th October 2020 letter to me that: "Our results indicate that global mean temperature for the decade 2000-2009 (34) has not yet exceeded the warmest temperatures of the early Holocene (5000 to 10,000 year B.P.). These temperatures are, however, warmer than 82% of the Holocene distribution as represented by the Standard 5 x 5 stack, or 72% after making plausible corrections for inherent smoothing of the high frequencies in the stack (6) (Fig. 3). In contrast, the decadal mean global temperature of the early 20th century (1900-1909) was cooler than >95% of the Holocene distribution under both the Standard 5 x 5 and high-frequency corrected scenarios. Global temperature, therefore, has risen from near the coldest to the warmest levels of the Holocene within the past century, reversing the long-term cooling trend that began ~5000 year B.P." Contains the logic that if a daily minimum and maximum temperature are known, the day's temperature can be declared unprecedented by comparing it to the annual mean temperatures for 100 years and is misleading?
        1. No, the two statements contained in question 7 are not equivalent. The paper compares the instrumental temperature record to the best estimate of past temperatures (from proxy records); it does not compare observations from a single day to a long-term average. While the time resolution of the proxy record of temperature changes over the past 10,000 years cannot fully resolve short term variability, there is no evidence to support an inference of undetected rapid changes in the early to mid-Holocene, nor is there evidence of a forcing mechanism that could have driven such a change (see answer to question 3).
        1. Does CSIRO expect the apparent warming of the last 100 years to continue? And if so, on what specific logical scientific points does CSIRO expect it to continue?
        1. Yes. As anthropogenic greenhouse emissions have caused the earth to warm, further emissions will cause further warming.
        1. In relation to the Lecavalier 2017 paper referenced in CSIRO’s 28th October 2020 letter to me, is the Lecavalier 25 year temperature data serially correlated?
        1. CSIRO are not aware of any published work that contradicts the conclusions reached by Lecavalier et al. (2017) and we would encourage anybody who has such information to publish their analysis in the peer reviewed literature.
        1. Does serial correlation effect the interpretation of least square means?
        1. See answer 9 above.
        1. What length and rate of warming would CSIRO consider normal when exiting a Little Ice Age and 3 solar minimums?
        1. The answer to this hypothetical question depends on the forcing and feedbacks.
        1. Does CSIRO stand by the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me as supporting a recent unprecedented rate of temperature change?
        1. Yes.
        1. Did CSIRO look in the public domain for prima facie evidence of issues with the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me?
        1. Yes.
    1. Is CSIRO aware that the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me occurs only in the Antarctic latitude band and no other latitude band?
    1. Is CSIRO aware that the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me disagrees with PAGES 2k 2013 Antarctic reconstruction, which has no uptick in temperature?
    1. Is CSIRO aware, in relation to the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me, that only one of the 16 proxies in the Antarctic latitude band showed the uptick in temperature?
    1. Is CSIRO aware, in relation to the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me, that the one of the 16 proxies in the Antarctic latitude band that showed the uptick in temperature was a borehole?
    1. Is CSIRO aware, in relation to the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me, that while one of the 16 proxies (a borehole) in the Antarctic latitude band showed the uptick in temperature 15 other more reliable proxies (such as ice cores) did not?
    1. Is CSIRO aware, in relation to the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me, that while one of the 16 proxies (a borehole) in the Antarctic latitude band showed the uptick in temperature, the LawDome Dahl-Jensen 1999 paper did not show an uptick in the original borehole?
    1. Answer to questions 14-19, referring to “the recent uptick in temperature in the Kaufman et al. (2020) paper”: We did not refer to a recent uptick in the proxy temperature record in our correspondence. As discussed above, and in previous correspondence with you, the papers cited by CSIRO use the instrumental record, not proxy data, to determine recent temperatures. The recent increase in global mean temperature shown in black in Figures 2 and 3 of Kaufman et al. (2020) is from the instrumental record, not proxy measurements.
    1. Is CSIRO aware, in relation to the recent uptick in temperature in the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me, that the error in importing the Dahl-Jensen borehole data was to miss the first data point and to load the remaining data points in the reverse time order?
    1. Yes, CSIRO are aware that an error was made in ingesting the borehole data and this has now been corrected by the authors in the relevant online database. As noted above, this has no impact on the recent observed rise in temperature in the instrumental record.
    1. What due diligence did CSIRO do on the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me before submitting that letter in answering my questions from October Additional Senate Estimates? Is CSIRO aware that in submitting its response to my questions for senate estimates that CSIRO misled the Senate?
    1. CSIRO rejects the assertion that it has misled the Senate.
    1. Is CSIRO concerned that the Kaufman 2020 paper referenced in CSIRO’s 28th October 2020 letter to me passed peer-review in Nature?
    1. No.
    1. Did CSIRO read the North Report 2006 referenced in CSIRO’s 28th October 2020 letter to me before citing it?
    1. Yes.
    1. Why did CSIRO include a report published in 2006 (North Report) among references published after Marcott’s 2013 paper in CSIRO's 28th October 2020 letter to me?
    1. Answer to questions 24-27: CSIRO included this information due to Senator Roberts interest in past temperatures, and believed that the other papers cited would be useful.
    1. Does CSIRO stand by their statement, in CSIRO’s 28th October 2020 letter to me, that the North Report 2006 was cited among the references that were published after Marcott’s 2013 paper, and as supporting the conclusion that recent rates of warming are unprecedented?
    1. Is CSIRO aware that the North Report 2006, referenced in CSIRO’s 28th October 2020 letter to me, states on pi 18: “Very little confidence can be assigned to statements concerning the hemispheric mean or global mean surface temperature prior to about A.D. 900 because of sparse data coverage and because the uncertainties associated with proxy data and the methods used to analyse and combine them are larger than during more recent time periods”?
    1. Does CSIRO explain the citing of the North Report 2006 in CSIRO’s 28th October 2020 letter to me to be incompetence, negligence, or a deliberate attempt to mislead the Senate?
    1. In CSIRO’s 28th October 2020 letter to me the IPCC AR5 WG1 is cited. The Summary for Policy Makers contains no reference to rates of temperature change in 10,000 years. There is no reference to Holocene or 10,000. Most citations are for only the last 1,000 years. Can CSIRO explain the inclusion of this irrelevant citation?
    1. CSIRO cited Chapter 5 of the IPCC AR5 WG1 report (not the Summary for Policy Makers), which includes an assessment of the literature covering past changes in temperature.
    1. Can a paper covering only the last 2,000 years support claims of what is unprecedented over 10,000 years?
  • Answer to questions 29-33: CSIRO included this information due to Senator Roberts interest in past changes in temperature, and thought references covering other periods in Earth’s history would be useful.
    1. Does CSIRO agree that PAGES 2k (2013) cited in CSIRO’s 28th October 2020 letter to me covers only the last 2,000 years and cannot support claims of what is unprecedented over 10,000 years?
    1. Why did CSIRO cite PAGES 2k (2013) in CSIRO’s 28th October 2020 letter to me?
    1. Does CSIRO agree that PAGES 2k (2017) cited in CSIRO’s 28th October 2020 letter to me covers only the last 2,000 years and cannot support claims of what is unprecedented over 10,000 years?
    1. Why did CSIRO cite PAGES 2k (2017) in CSIRO’s 28th October 2020 letter to me?
Loading page ...

Print options

Expand or collapse list branches

Want checkboxes? Change the list style

List style:

Display or hide list attributes

x

Expand & collapse ec

Import im

Examples and details

Word count wc

Current selection
Words: #{js-wc-sel}
Characters with spaces: #{js-cc-space-sel}
Characters without spaces: #{js-cc-sel}
The whole list
Words: #{js-wc}
Characters with spaces: #{js-cc-space}
Characters without spaces: #{js-cc}

List view options oo

  • hc
  • hf
  • sd
  • List style
  • om

Any email, forwarded to this address, will appear in beginning of this list.

Send an email to yourself and add the sender to Contacts for future use.

  • The email subject becomes the list item's text.
  • The email body becomes the list item's note.
  • All attachments from the email are attached to the list item (PRO only).
  • In the subject, you can also add #tags, ^due dates, and @assignees with Checkvist's smart syntax.

You can also set up voice integration on mobile devices