How NIWA added lots of warming in New Zealand – and got away with it – so far

Posted: October 24, 2012 by tallbloke in Analysis, Dataset, Legal, Measurement, methodology
Something is rotten in the state of Denmark New Zealand. A judge has disregarded as inadmissible expert evidence from a statistician who showed that the adjustment method NIWA claimed they used only gives a 0.3C/Century rise in temperature and ruled that NIWA (now a limited company) can adjust the temperature record as they see fit without having to demonstrate the use of a method based on any accepted science. here’s the graph:
NIWA are now claiming $118,000 from two named individuals in costs. This looks like a vindictive move.
NZCSET, the organisation which brought the review case to court are examining their options for appeal.
A comment on Jo Nova’s site summarizes the judgement.
Jaymez

It is a very disappointing judgement. I have read the entire judgement and at first pass it seems to me that Justice Venning found himself an escape hatch. Judges do not want their verdicts to be controversial or overruled, and you can bet he knew if he had found in favour of the ‘Coalition’ his ruling would have been attacked.

His escape hatch was two pronged:

1. Venning found that two of the key ‘expert witnesses’ for the Coalition, did not satisfy his definition of ‘expert witness’. That allowed him to simply disregard their arguments and evidence rather than look at them with as much weight as he applied to the testimony of NIWA’s experts which he appeared to accept without question.

[51] Section 23 of the Evidence Act 2006 provides that a statement of opinion is not admissible except as provided by ss 24 or 25. Opinion is defined in the Evidence Act at s 4 as: “A statement of opinion that tends to prove or disprove a fact.” I accept Mr Smith’s submission that there are substantial portions of Mr Dunleavy’s original and reply affidavits where he proffers opinions on matters in issue in the proceedings, particularly on scientific practices and the validity of the scientific practices of NIWA. Such evidence could only be admissible under s 24 or s 25. Section 24 is not applicable in the circumstances. Section 25 could only apply if Mr Dunleavy was an expert in the particular area of the science of meteorology and/or climate. He is not. He has no applicable qualifications. His interest in the area does not sufficiently qualify him as an expert. I also accept Mr Smith’s further point that Mr Dunleavy’s views are not capable of offering substantial help to this Court on the issue that it has to determine. To that extent I agree that substantial passages of Mr Dunleavy’s evidence are inadmissible.

[53] Similar issues (as to the limited nature of his expertise), apply to the evidence of Mr Dedekind. Although in his affidavit in reply Mr Dedekind purported to comply with r 9.43, Mr Dedekind’s expertise is in relation to computer modelling and statistical analysis.

In fact Vennings makes the following outrageous statement which reminds me of the same criticisms which were made of Ross McKitrick and Stephen McIntyre who were ‘unqualified’ to critique Michael Mann’s infamous Hockey Stick graph, with errors which had escaped the notice of thousands of supposedly qualified climate experts!

[54] I accept Mr Smith’s criticism of Mr Dedekind’s evidence to the extent that Mr Dedekind is not an expert in the application of statistical techniques in the field of climate science. Mr Dedekind’s general expertise in basic statistical techniques does not extend to any particular specialised experience or qualifications in the specific field of applying statistical techniques in the field of climate science. To that extent, where Mr Dedekind purports to comment or give opinions as to NIWA’s application of statistical techniques in those fields, his evidence is of little assistance to the Court.

These findings allowed Justice Venning to ignore or discount the Coalition’s evidence and instead preference whatever NIWA experts testified, even if they didn’t provide solid evidence; such as the hard copy of the adjusted temperature calculations, or the copy of the Australian Bureau of Meteorology Review. All they offered the court is testimony that their station data had been peer reviewed!

2. Venning found that it was not up to the court to make decisions between two competing scientific theories. So provided NIWA were using a credible theory then that was sufficient.

[45] I consider this Court should be cautious about interfering with decisions made and conclusions drawn by a specialist body, such as NIWA, acting within its own sphere of expertise. In such circumstances a less intensive or, to put it another way, a more tolerant review is appropriate.

[46] There is a further point. At times the witnesses have identified a difference of opinion about scientific methods applicable to climatology. There are a number of examples where the Court stated its reluctance to adjudicate on matters of scientific debate.
In SmithKline Beecham (New Zealand) Ltd v Minister of Health Ronald Young J said:
This Court’s function is not to rule on the science. The important point is that Medsafe, MAAC and Dr Boyd have considered all the Plaintiffs’ scientific propositions and have a credible view of the science by relevantly qualified scientists. They have considered and rejected on scientific grounds the Plaintiffs’ views on safety and efficacy and related matters.

[47] Unless the decision maker has followed a clearly improper process, the Court will be reluctant to adjudicate on matters of science and substitute its own inexpert view of the science if there is a tenable expert opinion: 

I guess had the ‘Coalition’ known this was the Justice’s position from the outset, then they would have known they were fighting an unwinnable case!

In fact Justice Venning made it pretty clear that not only would he not consider the intricacies of what method had been used to establish the temperature series and a warming trend, he was happy to take NIWA at it’s word on how it went about things:

[99] In summary to this point, the matters the Trust refers to and relies upon to support its argument that NIWA did not apply RS93 methodology to the 7SS are, at best, ambiguous and equivocal. Against that there is the express sworn testimony of Dr Wratt and Dr Mullan as to the application of the RS93. Dr Wratt deposes:41
The methods outlined [homogenisation methods] are the basis for the homogenisation used for the temperature series presented in the Salinger et al 1992 report, and for the NZ temperature series used in the Mullan et al 2010 report. Homogenised time series reported in Folland and Salinger 1993, produced by Salinger and colleagues in 1992 utilising the methods documented in Rhoades and Salinger 1993, [RS93] form the basis (with annual updates for new data) for the [7SS] used by NIWA between 1992 and mid-2010. …

[137] To the extent these matters involve differing contestable scientific opinions the Court cannot resolve them. However, I have to observe that Dr Wratt’s evidence appears credible and reasonable and, if necessary, I would accept it in preference to the evidence for the Trust on this point.

Vennings in the end really says he couldn’t care less how NIWA worked out it’s temperature trends, he would accept it since he has already made it clear he wouldn’t adjudicate the science!:

[157] It is unnecessary for this Court to resolve this scientific debate. On the evidence I accept it is more than arguable that the purpose of the NIWA review was to independently recalculate site change temperature adjustments from the underlying data and metadata and to document how those recalculated adjustments were produced and what their values were. I accept that NIWA could have recalculated the temperature adjustments in a different way yet still have arrived at a similar result which would strengthen the robustness and validity of the previous results.

In addition Venning accepted NIWA’s argument that they did not have a legal responsibility to provide New Zealand’s temperature record.

[38]. However, it has consistently been NIWA’s position that, while it has maintained a national climate database and has published the 7SS since 1992, it has not designated that as an official New Zealand temperature record. Dr Wratt explains that the reference to “the New Zealand temperature record” on NIWA’s website is used in a generic way to encompass a multitude of pieces of information which, together, compromise a record of New Zealand’s temperature.

[40]. NIWA contends that the Trust essentially overstates the extent to which NIWA’s activities have direct public consequences.

[44]….the Trust’s challenge is based on what it defines in its pleadings as “recognised scientific opinion”. A less intensive review is particularly apposite where the Court is not in a position to definitively adjudicate on scientific opinions. The Trust defines “recognised scientific opinion” as established scientific opinions and methods described in internationally recognised research journals. NIWA does not accept there is any such obligation, a matter to which I return shortly.

The Climate Coalition (or ‘Trust’) as it is referred to in the proceedings, raised valid point after valid point about the way data had been spliced, temperature stations had been selected and UHI had not been properly accounted for. In response, NIWA simply testified that they did handle each of those issues properly and gave some examples of how that might have been done without actually providing in evidence the actual adjusted data or the peer review of their work. Justice Venning simply accepted them at their word rather than take the legal course which was open to him, which would have been to assume an adverse position because they refused to provide evidence available to them.

If Justice Venning has acted appropriately under the law, it certainly doesn’t appear either logical or fair and will not stop people wondering what biases he may have allowed to come to bear, or what political pressure may have been brought to bear.

The Judgement in full:
http://www.courtsofnz.govt.nz/cases/nz-climate-science-education-trust-v-niwa-ltd/

Sceptical commentary and discussion:
http://www.climateconversation.wordshine.co.nz/2012/10/government-against-the-people/

The sceptics case is laid out in in various previous posts (NZ sceptics v. NIWA – summary of caseMore about the NZ temperature recordWhat warmingIncredible sham from NIWA and others) which summarise the Coalition’s case against the Seven-Station Series (7SS).

It seems the NZCSET could do with some help from a properly qualified bona fide card carrying climatologist to get their analysis validated and represented. Here’s hoping they can find someone willing to step up, and that they get leave to appeal.

UPDATE 28-10: NZ Farmers Weekly give an overview via climatescience.org.nz , who say:

 

A UK blogger who is a qualified engineer and a graduate of the History and Philosophy of Science, and interested in finding out how the solar system works and how Earth is affected by changes in it has analysed the judgement in the case taken in the New Zealand High Court by the New Zealand Climate Science Education Trust against NIWA.

NIWA seeks to make New Zealand legal history by claiming costs not from the New Zealand Climate Science Education Trust which took the case against it in the High Court, but from two private citizen members of that Trust.

Comments
  1. Doug Proctor says:

    Points to ponder as an appeal:\

    1. Mr Dedekind’s general expertise in basic statistical techniques does not extend to any particular specialised experience or qualifications in the specific field of applying statistical techniques in the field of climate science.

    The president of a statistical society or of a climate-based society would be able to determine if there is a “special” type of statistics needed for climatology. This, I doubt.

    2. This Court’s function is not to rule on the science.

    True. But if a tire manufacturer establishes braking distance on a high friction surface and gives that distance to a consumer on low friction surfaces, is that not a “correctable” act? If NIWA calculates a high temperature gradient and gives it to the government with the knowledge that this information will be used to create legislation that would not be brought in if the temperature gradient were shown to be much less, is that not a correctable act also?

    Not science. Engineering products.

    3. I accept that NIWA could have recalculated the temperature adjustments in a different way yet still have arrived at a similar result which would strengthen the robustness and validity of the previous results.

    The Coalition has (?) the recalculation under discussion that demonstrates the results would be different (?). Was the judge confused about the recalculation “ways” under discussion?

    4. NIWA contends that the Trust essentially overstates the extent to which NIWA’s activities have direct public consequences.

    Do the discussions leading to legislation include reference to NIWA’s temperature profile as demonstrating the need for the government to “act” on climate change?

    5. Justice Venning simply accepted them at their word rather than take the legal course which was open to him, which would have been to assume an adverse position because they refused to provide evidence available to them.

    Is this an opening for an appeal, that the judge did not insist that an order to provide something must be obeyed, and that a contempt of court position had been created by NIWA, and that NIWA were acting in a manner determined to thwart the ability of the Coalition to proceed? The problem is not that NIWA did not produce the data, but that NIWA failed to follow an order of the Court but was not held accountable for that failure.

    If these are errors, then the question is whether they were of a nature to potentially change the outcome of the trial. I see the expertise issue, the public consequences and failure to follow an order of the Court to be of serious enough potential impact to render an appeal reasonable.

  2. Ray Tomes says:

    As a New Zealander this case has been really important to me. I have noticed that similar disputes over past adjustments have also arisen in Australia, UK and US. There is a serious need for a proper independent scientific inquiry into all of these. The NIWA web site claims that the raw data shows the same ~1 degree increase increase. Someone is telling lies. NIWA https://www.niwa.co.nz/our-science/climate/information-and-resources/nz-temp-record/seven-station-series-temperature-data

  3. Nano Pope says:

    Thanks for keeping an eye on this travesty. Hopefully it will draw some attention to the issue, and not be a complete loss. Also, it might help others who are willing to go down this risky path.

  4. tchannon says:

    Some time ago the only New Zealand station which came out sensibly here from fingerprint tests or gross defects was Christchurch.

    Where is it in the list?.

  5. Richard C (NZ) says:

    Good stuff TB. My perspective as per Doug Proctor above:-

    #1 Mr Dedekind (Bob, and at least one other author) sent the ‘Statistical Audit of the NIWA 7-Station Review’ (pdf accessible from that title on web) to 3 independent professional statisticians for review. The final document was modified to accommodate their criticisms. By disallowing Bob’s expertise, J Venning disallowed the expertise of the professional statisticians, preferring the unsubstantiated word of a climate scientist (Dr Mullen). The R&S93 method that the ‘Statistical Audit’ followed rigourously is detailed in the Appendix, NIWA do not detail their method.

    #2 The Judicial Oath includes the words “without fear or favour”. J Venning exhibited both fear (caution) and favour (towards NIWA).

    #3 The trend graph in the post is not for the 7SS (that begins 1909), The linear 7SS trend by NZCSET after adjustment of the raw data using R&S technique was 0.34 C/century, NIWA’s linear trend is 0.91 C/century, graphed here:-

    Polynomial trend graphed here:-

    The warming in the record has been attributed to natural factors by NIWA scientists Salinger and Mullen 1999 (cited indirectly by the IPCC AR4, Australia and New Zealand section).

    #4 The NZCSET 7SS series is still valid statistically irrespective of the Judges decision. All that is needed is publication in a statistics journal (Bob et al considering publication). Once published, NZ will have 2 temperature series to choose from – take your pick. Even now there is no reason not to cite the ‘Statistical Audit’.

  6. tallbloke says:

    Nono Pope and Richard C: Welcome. Please put up a direct link to Bob Dedekind’s submitted assessment. I did find it a few months back but didn’t save a copy.

    There is clearly a miscarriage of justice here. We’ll try to help by publicising it as much as we can. Shame is occasionally effective on bureaucrats and politicians.

    If the judge wasn’t going to admit the expert evidence from NZCSET in forming his judgement, he should have said this right at the start to give NZCSet the opportunity to get their study validated by someone the judge would accept as an expert witness . The judge is therefore responsible for the time NIWA put in, and so, for the costs.

    It looks like entrapment from here.

    If we could get Bob Dedekind’s study openly validated by someone as expert as NIWA’s ‘expert witness’, quickly, would that help form the grounds for appeal?

    I’d rather help rustle up the funds to do that, than have to appeal for funds to hand to NIWA.

  7. Richard C (NZ) says:

    >”Please put up a direct link to Bob Dedekind’s submitted assessment”

    ‘Statistical Audit of the NIWA 7-Station Review’

    http://www.climateconversation.wordshine.co.nz/docs/Statistical%20Audit%20of%20the%20NIWA%207-Station%20Review%20Aug%202011.pdf

    Affidavit of Bob Dedekind

    http://nzclimatescience.net/images/PDFs/dedekind2.pdf

    16. I forwarded a penultimate draft of the documents to three separate consulting statisticians requesting that they “peer-review” the methodology and calculations. All three responded that they believed the documents to be correct. The minor improvements they suggested were incorporated into the text.

  8. Richard C (NZ) says:

    >”If we could get Bob Dedekind’s study openly validated by someone as expert as NIWA’s ‘expert witness’,”

    Already done by Bob D (3 consultant statisticians trumps one non-statistician climate scientist I would have thought). See #16 quoted from Bob’s affidavit in my last comment.

    As I said up-thread, by discarding Bob’s affidavit for lack of expertise the judge discarded 3 consultant statistician reviewers too.

  9. tallbloke says:

    Richard, did the three put their names to their reviews validating the affidavit, or were they ‘anonymous peer reviewers’?

    If the latter, even though I agree with you about their expertise, and that the judge was unnecessarily obtuse, I think NZCSET needs to move *very quickly* to get a named person well published in the climate science and temperature measurement fields to validate the study. That would help form grounds for an appeal I think.

    If payment for their time is needed, we would help with that.

    I would have thought that John Christy is the man for this job. developed Remote sensing of surface and tropospheric Temperature, has published many papers on surface temperature. He has done studies of surface gathered data in Africa and elsewhere. He has given senate testimony on several occasions, was an IPCC Author, and is an all round good guy.

    http://www.nsstc.uah.edu/atmos/christy2011/index.html

    UPDATE. I’ve emailed John Christy.

  10. Richard C (NZ) says:

    >”.did the three put their names to their reviews validating the affidavit, or were they ‘anonymous peer reviewers’.?”

    The reviews didn’t validate the affidavit, they validated the ‘Statistical Audit of the NIWA 7-Station Review’. They are anonymous in the ‘Statistical Audit’ but obviously Bob knows who they are.

    >”I think NZCSET needs to move *very quickly* to get a named person published in the climate science field to validate the study”

    For what purpose? For an appeal? An appeal has been lodged and I would assume that it would point out in no uncertain terms that J Venning overlooked the statistical peer-review. Adding another name from climate science doesn’t make the statistics any better because NIWA already say the ‘Statistical Audit’ is “too rigourous”. For publication? I’ve suggested that the reviewers become co-authors for submission to a statistical journal. Either that or the reviews are submitted along with the paper.

    >”If payment for their time is needed”

    NSZCS has bigger problems viz. a demand for NZ$118,000 costs from NIWA but I will relay your comment to Richard Treadgold here:-

    http://www.climateconversation.wordshine.co.nz/2012/10/niwa-says-it-wasnt-about-climate-change/#comment-130729

    Richard Treadgold says:
    October 25, 2012 at 8:02 pm

    There are no hearings regarding costs, just memoranda for Justice Venning. We have lodged an appeal, though. It’s low-key so far, I don’t have much information about it, and we still have funding problems.

  11. Richard C (NZ) says:

    >”I would have thought that John Christy is the man for this job”

    The contention is breakpoint analysis so John (or anyone) would first and foremost have to be a statistical breakpoint expert, is he?

    Judith Curry featured a breakpoint paper a while ago:-

    http://judithcurry.com/2012/06/21/three-new-papers-on-interpreting-temperature-trends/

    ‘Did the global temperature trend change at the end of the 1990s?’

    Tom Quirk
    Institute of Public Affairs
    Melbourne
    Australia

    http://ipa.org.au/library/publication/1339463007_document_break_paper_apjas_ipa.pdf

    Acknowledgements: I have benefited greatly from discussions with William Kininmonth and in particular, David Stockwell who introduced me to the Chow Break Test.

    From Wikipedia:-

    Kininmonth headed Australia’s National Climate Centre at the Bureau of Meteorology from 1986 to 1998,[2] He was Australia’s delegate to the WMO Commission for Climatology, was a member of Australia’s delegations to the Second World Climate Conference (1990) and the subsequent intergovernmental negotiations for the Framework Convention on Climate Change (1991–1992).[2]

    http://en.wikipedia.org/wiki/William_Kininmonth_%28meteorologist%29

    From Niche modeling (Stockwell):-

    A Ph.D. in Ecosystem Dynamics from the Australian National University in 1992, then consultant to WHO, Parks and Wildlife, Land and Natural Resources services until moving to the San Diego Supercomputer Center at University of California San Diego in 1997. There I helped to develop computational and data intensive infrastructure for ecological niche modeling mainly using museum collections data with grants from the NSF, USGS and DOT. I developed the GARP (Genetic Algorithm for Rule-set Production) system making contributions modeling of invasive species, epidemiology of human diseases, the discovery of new species, and effects on species of climate change. Judged by the US Immigration Service as an Outstanding Researcher, a category for those recognized internationally as outstanding in their academic field.

    http://landshape.org/enm/about-the-author/

    The Quirk paper is a “break point” analysis similar to what BOM used for ACORN-SAT and NZCSET used for the ‘Statistical Audit’ of NIWA’s 7SS

    To my mind, those three guys would be most helpful but I don’t think NZCSC approached them (maybe they did) and I doubt they will do so now either (maybe they will). I’m inclined to think that the problem is legal rather than statistical though. I got the impression that the claim was not well presented and introduced too much extraneous stuff at the expense of focus. J Venning described it as “prolix”.

  12. tallbloke says:

    According to judge Venning, whoever is going to be the expert witness first and foremost needs to be a climatologist.

    You didn’t answer my question about whether the three stats experts who reviewed Bob Dedekind’s work put their name to their reviews. Did they?

  13. tallbloke says:

    John Christy has replied. He needs to know if Tmax and Tmin datasets are available for the 7 stations. Can anyone confirm yes or no?

  14. Roger Andrews says:

    Here are some linear regression estimates of NZ warming between 1910 and 2010 using MY seven stations (Auckland, Wellington, Hokitika, Dunedin, Christchurch, Invercargill and New Plymouth).

    NCDC GHCNv3.2: 1.13C
    BEST: 0.66C
    GISS hom. adj: 0.40C
    Unadj. GHCN v2: 0.30C

    Quite a large range of variation.

    NIWA using its seven stations gets 0.96C, so it still comes second to NCDC.

  15. Richard C (NZ) says:

    >”You didn’t answer my question about whether the three stats experts who reviewed Bob Dedekind’s work put their name to their reviews. Did they?”

    I did answer:-

    “They are anonymous in the ‘Statistical Audit’ but obviously Bob knows who they are”

    Bob sought their reviews (#16 Affidavit up-thread) so there’s no problem attributing a name to each review. I would assume there was a cover letter for each review and that they signed off on each one but I simply do not know for sure, I’m not a party to the claim or the reviews.

    >”John Christy has replied. He needs to know if Tmax and Tmin datasets are available for the 7 stations. Can anyone confirm yes or no?”

    Yes, here at NiWA’s CliFlo database:-

    http://cliflo.niwa.co.nz/pls/niwp/wgenf.genform1?cdt=ls_mxmn&cadd=t

    He would need to obtain access by subscription but that is no obstacle. Quoting CliFlo the selection options are:

    Selected datatype(s)

    Statistics Calculated from Observations:
    Annual Statistics
    Monthly Statistics

    Totals, means and extremes for a year.
    Totals, means and extremes for a month.

    Mean Temperature
    Mean air temperature (°C) 02

    Maximum Temperature
    Mean daily maximum air temperature (°C) 03

    Minimum Temperature
    Mean daily minimum air temperature (°C) 04

    John Christy would also of necessity (whether climatologist or not), need to be competent with the breakpoint technique of Rhoades and Salinger 1993 (detailed in the ‘Statistical Audit’ Appendix) and the application of it. It is the application of that technique, specifically the overlap periods of breaks that is the core issue and basis of claim. As I pointed out up-thread, there are relatively local climatologists (and better) who are expert in breakpoint analysis who were not approached (to my knowledge).

    And I repeat, the word of a hundred climatologists will not improve the statistical review of three consultant statisticians in terms of rigour. NIWA has convinced the judge that such rigour is not required. The appeal has to reverse that IMO and the reviews obtained should be sufficient for that if the judge can be persuaded to take those reviews into consideration which he did not do in the hearing.

  16. Richard C (NZ) says:

    There’s daily data in CliFlo too.

  17. Bob D says:

    >John Christy has replied. He needs to know if Tmax and Tmin datasets are available for the 7 stations. Can anyone confirm yes or no?
    I believe they are. I used the Tmean to make sure I was comparing “like with like” when checking the NIWA Review, as that’s what they used.

  18. tallbloke says:

    Richard,
    Sorry, I missed the anonymity comment.

    And thanks for the data link, I’ll pass it on to John.

    To answer your question about his expertise with break points, he told me in email

    “I’ve glanced at the document and what I see appears to be a type of
    shift detection algorithm based on neighboring stations with a weighting
    factor related to the correlation. I remember working with Jim Salinger
    on the IPCC TAR, but don’t recall any work he did on homogenization.”

    and

    “I’ve published a few papers on breakpoint detection of surface
    temperatures (Alabama, California, East Africa). There are a number of
    ways to do it. Evidently, the problem here, as I glanced through the
    document, is that NIWA used a method that is now unreproducible, and
    which does not agree with the (relatively simplistic) method of Rhoades
    and Salinger.”

    So I’d say he’s well enough acquainted with the issues, personalities and techniques.

    I reminded him Jim Salinger’s PhD tutor was Phil jones. ;)

  19. tallbloke says:

    Roger A:
    Thanks, great summary of the range of trends and their provenance,

    Bob D, welcome, and well done for opening this can of worms. :)

    I understand why you used Tmean to be on the same sheet as NIWA.

    John Christy says
    “To do this right one would need TMax and TMin separately analyzed with
    various types of homogenization algorithms.”

    So if he’s willing, I suggest we let him have at it, and see what he has to say.
    We can always return to the issue of validation of your work, (which you already know is correct from your three reviewers) if necessary.

  20. Bob D says:

    We actually have quite a bit of other useful info for John, if he’s interested. You could put him in touch with me if you like, you have my email address now. We would love to have him on board, even if he’s just checking our workings.

    Basically, we have no doubt that we have applied the RS93 technique correctly, since NIWA have swarmed all over it and have grudgingly admitted that we applied the method correctly, although they feel we were “too rigorous”!

    They apparently prefer to ignore correlation-based station weightings and statistical significance tests of any type. They make no allowance whatsoever for balance between type I and type II errors – everything gets a free pass: they are happy to make site adjustments as small as 0.02°C!

    They also make no allowances for gradual changes such as sheltering or UHI in any of the stations chosen, even though they identify known problem stations (eg: Albert Park in Auckland, which we showed to be warming at 0.9°/century faster than other Auckland fringe sites) before ignoring the problems themselves (see the doc below for details, under the Auckland section)

    There’s one extra document you’ll need, it contains the bulk of the work:

    http://www.climateconversation.wordshine.co.nz/docs/Statistical%20Audit%20of%20the%20NIWA%207-Station%20Review%20Aug%202011%20SI.pdf

  21. tallbloke says:

    Thanks Bob, I’ll forward that to John along with your email address. Please keep me in the loop, and good luck.

  22. Bob D says:

    “…appears to be a type of shift detection algorithm based on neighboring stations with a weighting factor related to the correlation.”
    Sort of, yes. But it’s not like the GHCN algorithm, in that it doesn’t detect shifts. You have to know a priori that a site change occurred, and it then determines the magnitude, direction, and statistical significance of the adjustment.

  23. J Martin says:

    There is no conceivable mechanism that I am aware of that could require an adjustment of slope in a graph of temperature. It’s so obviously unjustified that any judge should have sufficient intelligence to see through it without the need to consult an ‘expert’.

    This decision and others, in particular the US high court declaration that co2 is a pollutant brings into question the competence of a number of judges and highlights both the decline of court standards and the atrophy of intellectual rigour which increasingly permeates the judiciary.

    Such decisions also run the risk of weakening the independence of the judiciary which brings with it the danger that the courts will become an ever enfeebled puppet to the wishes of government.

  24. Richard C (NZ) says:

    >”…they are happy to make site adjustments as small as 0.02°C!”

    BOM made no adjustments less than 0.3 C (their criteria) in their new ACORN-SAT series

  25. […] How NIWA added lots of warming in New Zealand – and got away with it – so far […]

  26. tallbloke says:

    This is going viral. Watts Up With That have picked up the story from here

    http://wattsupwiththat.com/2012/10/25/battle-of-the-credible-experts-in-new-zealand

  27. Ilma630 says:

    And now the latest Mann lawsuit is landed with a questionable judge too!

  28. omnologos says:

    Is “compromise a record” really in the original?

  29. tallbloke says:

    Omnologos: Good spot!

    There it is in the original pdf of the judgement from the link in the headline post

    “Dr Wratt explains that the reference to “the New Zealand temperature record” on NIWA’s website is used in a generic way to encompass a multitude of pieces of information which, together, compromise a record of New Zealand’s temperature.”

    Ah, the dangers of relying on spill-chuckers
    :) :)

  30. omnologos says:

    Seemingly that judge is clueless on more than just science

  31. tallbloke says:

    You never know. He might have slipped it in to let NIWA know what he really thinks of them.

  32. Richard C (NZ) says:

    I suspect that the names of the 3 consultant statistician reviewers may be in confidence to protect their respective businesses. Certain clients might look askance at their consultant’s participation in what they might perceive as a NIWA-busting project.

    Bob D could confirm this or not. I know that he has placed himself in exactly that position by the public knowledge of his ‘Statistical Audit’ authorship and court affidavit. He has risked losing clients by that publicity.

  33. Richard C, it is a pity your (NZ) Chartered Professional Engineers of NZ Act 2002 which binds the crown does not go a step further like the PE Act Qld. and make it a criminal offense for anyone who supplies an engineering service (which includes engineering data) to be not registered. It could have been argued that measurement of temperature and manipulation of the instrument output is an engineering service and should have been signed off by a professional engineer (who is required to adhere to a code of ethics).

    Now might be a good time to pressure for a change to the act. I understand that one of the so-called engineers responsible for construction of one of the building that collapsed and caused deaths in the Christchurch earthquake is to be prosecuted for identity theft of a qualified engineer in Australia.

  34. Steven Mosher says:

    your best bet would be to actually test the rs93 method. a test suite is actually available. i would not hold my breathe for the method compared to other approaches based on the description given here.

  35. tallbloke says:

    You can breathe easy Mosh, that’s exactly what Bob D did, as confirmed by the three expert statisticians who reviewed his paper, which found that by application of the method, the centennial trend was 0.3C.

    http://www.climateconversation.wordshine.co.nz/docs/Statistical%20Audit%20of%20the%20NIWA%207-Station%20Review%20Aug%202011.pdf

    The problem is that NIWA, although they couldn’t find anything wrong with the analysis, said Bob D had applied the method “too rigorously”, and the Judge believed them, despite NIWA showing no working to justify their stance.

  36. Richard C (NZ) says:

    TB, isn’t Steve Mosher suggesting a test of R&S against other breakpoint algorithms (the “suite”)? John Christy observes R&S is “simplistic”

    I don’t see the need given NIWA widely concur with the R&S approach, they just want latitude in application as Bob D describes up-thread.

  37. Richard C (NZ) says:

    It might be helpful to lay out the respective applications of R&S – NSCSET vs NIWA. From the “Statistical Audit':-
    ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    Basically, the R&S method for comparing a station with neighbouring stations involves the use of

    * Monthly data
    * Symmetric interval centred on the shift
    * A 1-2 year period before and after the shift
    * Weighted averages based on correlations with neighbouring stations
    *Adjustments only performed if results are significant at the 95% confidence level

    The NIWA method uses:

    * Annual data
    * Asymmetric intervals
    * Varying periods of up to 11 years before and after the shift
    * No weighted averages
    * No evidence of significance tests – adjustments are always applied.

    In the review reports, NIWA gives no reason for not implementing the R&S method correctly, nor is the matter discussed.

    # # #

    These differences are essentially what the court claim was about statistically I think.

  38. Slabadang says:

    What the judge really said!

    Its so nice to have a job where im allowed to leave my brain at home before I leave!
    Unacceptable …period!. A three year old can see whats wrong.

  39. tallbloke says:

    TB, isn’t Steve Mosher suggesting a test of R&S against other breakpoint algorithms (the “suite”)?

    He might be, Mosh is never very specific these days.

    John Christy observes R&S is “simplistic”

    He said “relatively simplistic” which has a slightly different connotation I think.

    “I don’t see the need given NIWA widely concur with the R&S approach, they just want latitude in application as Bob D describes up-thread. ”

    Well when you consider that GHCN’s assessment of the centannial trend of NZ’s temperature changes from 0.3 in version 2 to 0.95 in version 3, ‘latitude’ is much used by several other agencies as well as NIWA. John Christy describes Hadley’s latest historical SST’s as “a mystery”.

    My view (and Mosh would agree here) is that if they don’t show their working, their results are worthless.

  40. Bob D says:

    Richard C:
    “These differences are essentially what the court claim was about statistically I think.”

    Yes, pretty much. When challenged by the NZSCS, NIWA always responded that their 7SS was developed using internationally-approved best practice as published in peer-reviewed journals, and cited RS93 as that technique.

    They then (eventually) had to admit that they had “lost” all the calculations, history of adjustments, and even the source code used to make these adjustments, so couldn’t tell the world what they were or how they were made. Sharp questions were asked in Parliament and much foot-shuffling ensued.

    They were then given a taxpayer-funded budget to re-do the 7SS (now named the NZ7T) in 2010. When they produced their Review, as they called it, they mysteriously failed to mention (or justify) that instead of using RS93, as they apparently had in the original 7SS, they chose instead to use a crude, out-dated technique from Jim Salinger’s student thesis written in 1981.
    What we showed in court was that had NIWA used RS93, they would have achieved a far different result.

    Our analysis currently stands as the only analysis performed on the NZ data using RS93, the technique described by NIWA themselves as best practice. NIWA would like to claim that their 7SS was also performed using RS93, but they can produce no evidence for this.

    The main question that must be asked by the NZ tax-payer is this:
    Why on earth, when given a budget to re-do the 7SS in 2010, did NIWA not use RS93? Why revert to an old, out-dated and crude method from a student’s thesis in 1981 when the very technique they claimed as internationally best practice (RS93) was readily available?

  41. Bob D says:

    By the way, NIWA originally claimed that the 7SS came from the thesis, not RS93. They later changed their story in the court case to claim that RS93 had been used too (in 1992).

    Now what is doubly odd is that NIWA claimed the original thesis adjustments were lost in 1983 (see below). But how then could they have updated them in 1992? And why were the 1992 adjustments lost a second time? It seems a lot of results are lost.

    From http://www.climateconversation.wordshine.co.nz/2010/10/the-curious-case-of-the-missing-thesis/:

    The Minister responsible for NIWA, the Hon Dr Wayne Mapp, told Parliament on 18 February 2010 that the adjustments to the 7-station series were taken from the Salinger thesis. In a follow-up written answer (PQ1320) he explained more fully that “the adjustments used in the present “seven-station” series are consistent with those in the Salinger thesis. Some changes to the original adjustments have been necessary in the thirty years since the thesis was published.”

    In answer to PQ1193, Dr Mapp advised that the source material for NIWA’s Schedule of Adjustments were: “a list of the more than 30 sites used to develop the ‘seven-station’ series; raw unadjusted data for these individual sites from NIWA’s National Climate Database; the time series of adjusted monthly mean temperatures at the seven locations; and Appendix C from Dr Jim Salinger’s 1981 Ph.D. thesis.”

    In answer to other Parliamentary Questions, Dr Mapp described how the detailed calculations for the Salinger thesis had been recorded on the VUW mainframe, and were lost when the University changed its system in 1983. He defended the lack of 7SS peer-review by contending that the Salinger doctoral thesis had itself been reviewed by two supervisors. However, oddly, neither of them were climate scientists.

  42. tallbloke says:

    Bob D. thanks for your further comments. This helps us understand the context of the dispute.

    John Christy writes in email:

    “I looked at the website, http://cliflo.niwa.co.nz/pls/niwp/wgenf.genform1?cdt=ls_mxmn&cadd=t
    but it doesn’t seem to do what I want. At most
    you can get only 500 lines of data (daily). I would like to see monthly
    means of TMax and TMin for the longest stations available. The key
    proxy for the greenhouse effect is TMax as it is more closely tied to
    the deeper troposphere through better mixing than is TMin and thus
    TMean. TMin (and TMean) is subject to serious contamination. TMax is
    not the best proxy for the troposphere (where the greenhouse effect is
    truly measured), but it is better than TMin.

    My quick run through the BEST TMax data shows no trend since 1970 for a
    New Zealand average. However, the data only start in 1944.

    John C.”

    Please can someone supply me with longer monthly datasets of Tmax and Tmin to pass along to John. I don’t want to send him on wild goose chases.

  43. Bob D says:

    >Please can someone supply me with longer monthly datasets of Tmax and Tmin to pass along to John. I don’t want to send him on wild goose chases.
    Will do, if you can give him my email address I can help him directly with these numbers. CliFlo isn’t the easiest tool to use, and I’ll need to understand exactly what stations he’s looking for.

  44. tallbloke says:

    Hi Bob,
    I did pass your address along to John yesterday. I suggest you email him direct with reference to your role as principle analyst in the case.
    His email address is on this page: http://www.nsstc.uah.edu/atmos/christy2011/index.html

  45. Bob D says:

    Thanks TB.

  46. Rob R says:

    With regard to Tmax and Tmin for the 7 sites these are freely available from the NIWA cliflo database.

    With regard to the question by tchannon Christchurch is not one of the “7 sites”. Lincoln (near to Christchurch) is one of the 7 sites.

    The temperature series from Christchurch (Christchurch Gardens) is from two sites in the Hagley Park area. There is a ten year data gap and probably a site change between the two records. The second part of the Gardens record is derived from a facility that started as the official geomagnetic observatory. It was established in the gardens during the year 1900 to support the scientific objectives of the Antarctic research programme being conducted by Captain Robert Scott. The combined gardens record suffers from massive urban heat island influence. This can be readily detected by comparing the record from the Gardens with the temperature series from Wigram Airfield and Christchurch Airport (which are also probably influenced by UHI). So the Christchurch temperature record is just about as far from a “clean record” as it is possible to get in New Zealand.

  47. Richard C (NZ) says:

    In John Christy’s email “The key proxy for the greenhouse effect is TMax”

    I hope John is not assuming the court claim was/is a GHE issue.

  48. tchannon says:

    Fair enough, I’ve not looked at the stations etc.

    Ignore the data prior to about 1910, is irrelevant. (extending backwards is a different matter, of no interest here)

    Going on published data it is the only reasonably consistent data. The others are not.

    They are trying to bridge widely separated stations.

    A quick dig here.

    Taking what is claimed as Christchurch Airport in GHCN V1 (not a typo, published about 1990), deleting the early data we can have 1905 through mid 1950s without a break. Whilst bad but trivial to do the linear trend is zero (very slightly negative). Does UHI matter, does it cool with time?

    Then there is the real airport to 1990, also zero trend (slightly negative). Between the two means is a tiny difference. (the most recent data is no help to them)

    That is why.

    of course the stations they have chosen are wonderful pristine sites and datasets

    (Hagley Park is near here -43.528404,172.610323)

    I’ll drop out now, don’t want to waste your time on an irrelevancy, overloaded here anyway.

  49. […] I was distracted by work, first, Tallbloke’s Talkshop picked up the story, then WUWT repeated it. This is wonderful. Now we have John Christy looking at the […]

  50. tallbloke says:

    Who is the NZ minister for justice? I’m going to send him/her a letter advising that the eyes of the world are now on a country we had been led to believe has a well organised and fair judicial system.

  51. Rob R, not sure if you are from NZ but to clarify for others, Lincoln, south of Christchurch is in the country and hosting a university campus (centered on agriculture and rural things) On a recent visit had a coffee/lunch at the only open cafe in the little village area. It should be possible to have a class 1 site there. However, it seems a pretty recent establishment. It seems to have a long record as an agricultural college but It did not appear to be all that old from the buildings, trees etc. It seems that it became a College of Canterbury University in 1961 -maybe that is when it a had a growth phase. It became Lincoln University in 1990 and probably stayed the same since then. A google search indicates the site has 65ha and accommodates 650 students.

    Christchurch International airport would definitely have UHI. Beside the runways etc, there is a lot of industry in the surrounding area. For a start there are at least five companies with large parking lots for 1000’s (or 10000’s) of campervans, mobile homes etc for the tourists (did that, but last time hired a car at under $NZ40/day mileage included- the South Island is worth a visit)

  52. Tallbloke, just saw your last comment.
    A google search came up with this http://www.beehive.govt.nz/minister/judith-collins?page=1
    From the site this is the email address j.collins@ministers.govt.nz
    Go for it

    There is also an Ombudsman who deals with justice issues.
    http://www.ombudsman.parliament.nz/ and http://www.ombudsman.parliament.nz/make-a-complaint/complaining-about-state-sector-agencies-551307131224760#Making%20a%20complaint%20to%20the%20Ombudsman
    This maybe another avenue for NZCSET
    Publicity may change things to open up the “can of worms”

    [Reply] Thanks CF: Please can everyone crowdsource a suitable set of points to put to Judith Collins. Let’s put it together on this thread.

  53. Rob R says:

    Cementafriend

    Yep, definitrely from NZ. The official NIWA documentation for the Lincoln temperature series reveals that it is a composite of data from at least 6 different sites. In addition NIWA claim that there are at least 4 step-changes in the data from the first of these sites- thats about 9 different “splices” for the Lincoln temperature series. The series commences at the Agricultural College in 1881. You can trust the spliced record at your own peril. For details on Lincoln see the NIWA document: “Creating a Composite Temperature Series for Lincoln”, December 2010.

    1881 is the year that the first Christchurch Gardens station was closed. That was deliberate cost cutting by the NZ Government which only wanted to support 1 climate station in the Canterbury Region at that time.

    tchannon

    In the NZ climate database the temperature record for Christchurch airport begins in 1953 and continues to 2012. The temperature record from Wigram airfield commences in 1937 and extends to 1968. There has been massive industrial development in the vicinity of ChCh airport over the last 20 years so UHI is almost inevitable.

    In the NZ climatre database the temperature record for ChCh gardens recommences in 1905, but the full range of climate data was collected from 1901 and was reported in the Christchurch Press (newspaper). No matter what GHCN might or might not contain there are no truely clean temperature records available for Christchurch City. Even NIWA seem to recognise this.

  54. Solomon Green says:

    Solomon G.

    Having acted as an expert witness in more than four hundred cases, including tribunals and inquiries, I have found that it is not unusual for one’s qualifications and expertise to be challenged. In Hong Kong, the Channel Isles and England, where all my most recent cases have been based, it such challenges invariably take place at pre-trial hearings.

    To have accepted that Messrs Dunleavy and Dedekind were suitably qualified to provide expert evidence on matters that the court would have specified earlier and then to reject that evidence supposedly because their expertise was not considered relevant shows that Venning was either incompetent at case management or corrupt in his judgement (or both).

  55. Doug Proctor says:

    The science is interesting but, as I see it, not germane to the situation faced by the Coalition et al. What is needed is grounds for appeal of the ruling (and therefore punishment). Best would be if the decision were set aside, opening the ground for a second trial – which could be avoided with some openness, as who would want to risk defeat on either side.

    I am not a lawyer, but I can be anal, so I figure I might be on the right track when I say an appeal would have to relate to procedural (not conclusive) error. Denying statisticians because they are not climatologists, would strike me as one, in that the judge erroneously understood climatatology to use a fundamentally different style of statistical analysis not understood or in the expertise of a non-climatologically grounded statistician. If that basis for exclusion can be shown to be in error, then the plaintiff’s witnesses were incorrectly disallowed.

    NIWA also did not produce a statistical justification or even data (evidence). The judge would be correct in saying that they did not have to do so if he believed that only NIWA had the expertise required to understand what they did and why they did it. NIWA would be its own justifier: if I say a thing is so, it is so, because I am the recognized expert, not you. However, if the plaintiff’s experts were considered legitimate, then the claim of justification could only be settled by a review of the data by both sides and adjuicated by the judge: an adverserial position with referee of all cases in our system. So it comes back to the disallowing of the plaintiff”s witnesses.

    That the judge did not find an “adverse implication” in the refusal of NIWA to follow a court order stems, as I see it, in the irrelevance of the evidence when there was no recognized plaintiff witness to refute the evidence. A contempt of court situation did exist, in my opinion, but one that was moot. Again, back to the witnesses for the plaintiff.

    Are there procedural errors? An anal reading of the ruling suggests to me that the judge left some room for legitimate complaint. What does the Coalition’s solicitors say?

  56. […] Talkshop describes Manfred as an “expert” and a “statistician!”  (see How NIWA added lots of warming in New Zealand – and got away with it – so far). And Watts Up With That, a prominent denial blog, is describing him as a credible […]

  57. tallbloke says:

    The pingback above is a nice demonstration of how people on the other side of this debate prefer to attack the messengers, rather than take on the science contained in the message. Feel free to leave comments about the article there, rather than here. ;)

  58. Barry B says:

    An appeal against a High Court judgment in a Judicial Review proceeding is by way of rehearing in the New Zealand Court of Appeal. It is not confined to points of law or procedure.

    Special leave is required for new evidence to be admitted. It would be necessary to show that this evidence is directly relevant to a key point in issue and was not available to be presented at the High Court trial.

    The Appeal in the NIWA case relates to all three temperature records promulgated by NIWA:

    (i) the original adjusted 7SS 1853-2009 showing a warming trend of 1°C/century;
    (ii) the ‘reviewed’ adjusted 7SS (or NZT7) 1909-2009 with a trend of 0.91°C/century
    (iii) the raw 11SS, extended to 1931-2008 (actually 1955-94) with a trend of 1.35°C/century.

    Different climatological and legal issues arise in each case, so it is tantamount to three appeal cases. But legal advisers to NZCSET confirm that Venning J’s judgment is flawed in all three causes..

  59. patrick says:

    I could be wrong but I was under the impression that there are no climate stations operated by NIWA that were up to international scientific standards. I have used their data in my work in the past, but it could never be relied on as scientific fact. Might be something worth looking into.

  60. […] How NIWA added lots of warming in New Zealand – and got away with it – so far […]