Global Warming: Temperature Data Quality

ACADEMIC PAPERS
SPATIAL CORRELATIONS BETWEEN INDUSTRIALIZATION AND WARMING

McKitrick and Michaels 2004: In 2004 Pat Michaels and I published a paper examine spatial correlations between indicators of surface warming and industrialization in raw and homogenized temperature data sets. There has been a longstanding controversiy about this work ever since. 

Since I publish my data and code along with the papers, the results were heavily scrutinized. A blogger found a numerical error: I omitted to convert the latitude measure to radians (from degrees) prior to computing the cosine. This had only a small effect, but many climate scientists dismissed or ignored all the results largely on the basis of rumours that the results had been undermined. We published an Erratum right away, putting the before and after results side-by-side so people could see how small the changes were. The Erratum is available here.

There was also a comment on our work submitted to Climate Research. It was not peer-reviewed, instead the Editor just told us to write a response and, after editing, he published it with the comment. The author found that if he discarded half the data he got weaker results. Well surprise surprise. Here’s our response.
McKitrick & Michaels 2007: After our 2004 paper was published we extended the study to a global sample, and I built a new socioeconomic data base, adding new covariates and going to a complete coverage of the available land surface. The new results fully confirmed the old ones: the global land-based surface data set is heavily contaminated by extraneous ‘signals’ arising from land-surface modifications and variations in quality control. In a new paper we estimate the overall impact on measured warming over land and discuss the implications for understanding climate data. It’s high time for a much deeper re-examination of basic climate data sets.
ATMOSPHERIC OSCILLATIONS DO NOT EXPLAIN THE TEMPERATURE-INDUSTRIALIZATION CORRELATION

​In the 2007 IPCC Report, Pat Michaels and my 2004 paper was mentioned and dismissed as follows:
  • McKitrick and Michaels (2004) and De Laat and Maurellis (2006) attempted to demonstrate that geographical patterns of warming trends over land are strongly correlated with geographical patterns of industrial and socioeconomic development, implying that urbanisation and related land surface changes have caused much of the observed warming. However, the locations of greatest socioeconomic development are also those that have been most warmed by atmospheric circulation changes (Sections 3.2.2.7 and 3.6.4), which exhibit large-scale coherence. Hence, the correlation of warming with industrial and socioeconomic development ceases to be statistically significant.
There was no supporting evidence for the claim that our results were a spurious artifact of atmospheric oscillations. In other words it was a fabrication by the IPCC. I then spent several years trying to publish a response article. This yielded the paper:
In the 2012 IPCC report (AR5, chapter 2 page 34) the IPCC concedes they were fabricating evidence, as follows:
  • AR4 concluded that this correlation ceases to be statistically significant if one takes into account the fact that the locations of greatest socioeconomic development are also those that have been most warmed by atmospheric circulation changes but provided no explicit evidence for this overall assessment result.
PAPER  CONFIRMING CONTAMINATION OF SURFACE TEMPERATURE DATA
​
​In 2009 Gavin Schmidt published a paper in the International Journal of Climatology claiming the M&M2007 results, as well as those of de Laat and Maurellis who independently found the same things we did, were spurious. My rebuttal, coauthored with Nicolas Nierenberg, was published in The Journal of Economic and Social Measurement. 
Data/Code archive here. The paper provides a complete and thorough refutation of Schmidt's critique. Why JESM? First, because it is a journal that focuses on the critical evaluation of policy-relevant databases, and its editors and reviewers have considerable econometric depth, and this paper is fundamentally an application of econometrics to the evaluation of data quality. Second, we submitted the paper to the IJOC in April 2009, on the assumption that, having published Schmidt's paper, they were interested in the topic. Evidently their interest only extends to analyses that support IPCC views. After 10 months we found out that IJOC was rejecting our paper on the basis of some inane referee reports to which Nico and I were not given a chance to reply. We did anyway, and if anyone thinks the rejection by IJOC amounts to a knock against our paper, please read our response letter for some perspective. Whether or not the IJOC editors read it, they refused to reconsider our paper. Interestingly, we learned from the Climategate release that Schmidt's paper, which focuses on defending Phil Jones' CRU data against its various critics, was sent by the IJOC Editors to be reviewed by Phil Jones of the CRU. As you can imagine his review was shallow and uncritical, but evidently impressed the editors of IJOC. They didn't ask deLaat or me to supply a review, nor did they invite us to contribute a response. Every interaction I have had over the years with the IJOC has left me very unimpressed. 

UPDATE FEBRUARY 2014: The IPCC AR5 (chapter 2 page 34) summarized the issue as follows:
  • Subsequently McKitrick and Michaels (2007) concluded that about half the reported warming trend in global-average land surface air temperature in 1980–2002 resulted from local land-surface changes and faults in the observations. Schmidt (2009) undertook a quantitative analysis that supported AR4 conclusions that much of the reported correlation largely arose due to naturally occurring climate variability and model over-fitting and was not robust. Taking these factors into account, modified analyses by McKitrick (2010) and McKitrick and Nierenberg (2010) still yielded significant evidence for such contamination of the record.
ENCOMPASSING TESTS OF SOCIOECONOMIC SIGNALS IN SURFACE CLIMATE DATA  

​I published a paper in Climatic Change on the question of whether surface climate data are biased by non-climatic factors relating to socioeconomic development: Rather than try to settle the debate once and for all, I focus on why the various attempts to show the data are not contaminated do not disprove the results showing that they are. The problem has been that authors use non-overlapping data sets and different methods, and end up talking past each other. The way to settle matters, I argue, is to adopt an encompassing framework in which both types of results can be demonstrated on the same data set, where one arises in a restricted subset of another model, and the restrictions can formally be tested. I give two examples, one replicating a Parker-style equivalence between nighttime minimum trends in calm and windy conditions, then showing that this persists in a temperature data set that can be shown to be correlated with population growth. I also replicate the BEST-type results that rural trends are slightly greater than those of urban areas, and show that this result appears in a restricted subset of a larger model in which socioeconomic growth is significantly correlated with temperature trends. In both cases the restrictions necessary to yield the model that supposedly shows no data contamination are rejected. Data/code archive here.
​DOES A GLOBAL TEMPERATURE EXIST?
​In Taken By Storm Chris Essex and I explained in informal terms why there’s no such thing as a “global temperature.” We spelled out the theoretical argument in this paper: The implications are very important. The notion of a planet “warming” or “cooling” is ill-posed, at least in the context of the global warming discussion. Talking about a trend in a temperature average over a non-equilibrium system as if it is denotes “warming” (i.e. an increase in a number known to be a temperature) is scientific hand-waving and short cut-taking that conceals nontrivial physical problems. Every field (including economics) has to take short-cuts, but eventually when the discrepancies are piling up all around, someone has to go back and see where the wrong turn might have been. 
OTHER
A FIRST LOOK AT THE KARL ET AL. (2015) REVISED TEMPERATURE DATA SET
The team from NOAA released (June 4 2015) a new global temperature data set that shows an upward trend after 1998 of similar magnitude to the post-1950 trend, thus eliminating the "hiatus" observed in other data sets. The source of this change is the revision from the ERSSTv3b to the ERSSTv4 ocean surface data set. In order to help readers understand what i think the changes mean, or don't mean, I have prepared a commentary (June 8: Revised version posted):
BOOK CHAPTER ON GATEKEEPING
The process leading to publication of my paper on atmospheric oscillations took many bizarre twists and turns. In the end I went through seven journals, three of whom refused to review it and one which never replied to the submission. The journals that did review it obtained in total seven referee reports, six of which supported publication. The one who objected raised specific points to which I responded, but the editor would not acknowledge receipt of my responses. This episode has now been chronicled in a new book chapter:
BERKELEY EARTH STUDY REFEREE REPORTS
On September 8 2011 I was asked by Journal of Geophysical Research to be a reviewer for a paper by Charlotte Wickham et al. presenting the  Berkeley Earth Surface Temperature ("BEST") analysis of the effect of urbanization on land surface temperatures. This work is mainly associated with Richard Muller and his various coauthors.  I submitted my review just before the end of September 2011, outlining what I saw were serious shortcomings in their methods and arguing that their analysis does not establish valid grounds for the conclusions they assert. I suggested the authors be asked to undertake a major revision. 
   In October 2011, despite the papers not being accepted, Richard Muller launched a major international publicity blitz announcing the results of the "BEST" project. I wrote to him and his coauthor Judy Curry objecting to the promotional initiative since the critical comments of people like me were locked up under confidentiality rules, and the papers had not been accepted for publication. Richard stated that he felt there was no alternative since the studies would be picked up by the press anyway. Later, when the journal turned the paper down and asked for major revisions, I sought permission from Richard to release my review. He requested that I post it without indicating I was a reviewer for JGR. Since that was not feasible I simply kept it confidential. 
   On March 8 2012 I was asked by JGR to review a revised version of the Wickham et al. paper. I submitted my review at the end of March. The authors had made very few changes and had not addressed any of the methodological problems, so I recommended the paper not be published. I do not know what the journal's decision was, but it is 4 months later and I can find no evidence on the BEST website that this or any other BEST project paper has been accepted for publication. [Update July 30 2012: JGR told me "This paper was rejected and the editor recommended that the author resubmit it as a new paper."] 
   On July 29 2012 Richard Muller launched another publicity blitz (e.g. here and here) claiming, among other things, that "In our papers we demonstrate that none of these potentially troublesome effects [including those related to urbanization and land surface changes] unduly biased our conclusions." Their failure to provide a proper demonstration of this point had led me to recommend against publishing their paper. This places me in an awkward position since I made an undertaking to JGR to respect the confidentiality of the peer review process, but I have reason to believe Muller et al.'s analysis does not support the conclusions he is now asserting in the press. 
   I take the journal peer review process seriously and I dislike being placed in the position of having to break a commitment I made to JGR, but the "BEST" team's decision to launch another publicity blitz effectively nullifies any right they might have had to confidentiality in this matter. So I am herewith releasing my referee reports. The first, from September 2011, is here and the second, from March 2012 is here. 
A CRITICAL REVIEW OF SURFACE TEMPERATURE DATA PRODUCTS

This is my set of notes on the construction of various land and sea temperature archives and the associated data quality issues.
  • McKitrick, Ross R (2010) "A Critical Review of Global Surface Temperature Data Products"  SSRN Working Paper 1653928, August 6, 2010. 
AMPLIFICATION RATIOS
I was drawn into a dispute between Gavin Schmidt and Klotzbach et al. (see Pielke Sr.) over the latter paper's conclusion that the surface temperature record over land has a warm bias for the purpose of measuring global warming. I was cited in Klotzbach et al. as the source of a claim that the GISS model exhibits amplification over land of about 1.2. I should not have been cited, since all I did was report in an email to John Christy the average trop/surf trend in Gavin Schmidt's own GISS data pertaining to my 2007 surface temperature analysis. The information source, in other words, was Schmidt himself, not me; and in any case I did not provide it as a personal communication for the purpose of a journal article (which I did not know was being written). Phil Klotzbach and his coauthors have issued a correction on this point. In subsequent correspondence with Gavin Schmidt he reported to me that he had corrected an error his original IJOC archive and also that the GISS model classifies land differently than CRU so some of the 440 grid cells are actually over ocean in his model. He supplied me with the GISS landmask. I have recomputed the original results using the corrected data and the GISS landmask. The cosine-weighted amplification ratio over land is about 1.106 and over ocean is 1.602, where 'land' and 'ocean' are according to the GISS landmask applied to the 440 grid cells used in my 2007 paper. 
CRU DATA: In late 2009 I was contacted by several people, including a reporter, asking for supporting information regarding my July 2009 request for the raw data used in the CRU gridded temperature products. This has arisen in part because of Pat Michael's NRO article The Dog Ate Global Warming. The documentation of my request is here.