A look behind the curtain at NOAA’s climate data center.
I read with great irony recently that scientists are “frantically copying U.S. Climate data, fearing it might vanish under Trump” (e.g., Washington Post 13 December 2016). As a climate scientist formerly responsible for NOAA’s climate archive, the most critical issue in archival of climate data is actually scientists who are unwilling to formally archive and document their data. I spent the last decade cajoling climate scientists to archive their data and fully document the datasets. I established a climate data records program that was awarded a U.S. Department of Commerce Gold Medal in 2014 for visionary work in the acquisition, production, and preservation of climate data records (CDRs), which accurately describe the Earth’s changing environment.
The most serious example of a climate scientist not archiving or documenting a critical climate dataset was the study of Tom Karl et al. 2015 (hereafter referred to as the Karl study or K15), purporting to show no ‘hiatus’ in global warming in the 2000s (Federal scientists say there never was any global warming “pause”). The study drew criticism from other climate scientists, who disagreed with K15’s conclusion about the ‘hiatus.’ (Making sense of the early-2000s warming slowdown). The paper also drew the attention of the Chairman of the House Science Committee, Representative Lamar Smith, who questioned the timing of the report, which was issued just prior to the Obama Administration’s Clean Power Plan submission to the Paris Climate Conference in 2015.
In the following sections, I provide the details of how Mr. Karl failed to disclose critical information to NOAA, Science Magazine, and Chairman Smith regarding the datasets used in K15. I have extensive documentation that provides independent verification of the story below. I also provide my suggestions for how we might keep such a flagrant manipulation of scientific integrity guidelines and scientific publication standards from happening in the future. Finally, I provide some links to examples of what well documented CDRs look like that readers might contrast and compare with what Mr. Karl has provided.
In 2013, prior to the Karl study, the National Climatic Data Center [NCDC, now the NOAA National Centers for Environmental Information (NCEI)] had just adopted much improved processes for formal review of Climate Data Records, a process I formulated [link]. The land temperature dataset used in the Karl study had never been processed through the station adjustment software before, which led me to believe something was amiss. When I pressed the co-authors, they said they had decided not to archive the dataset, but did not defend the decision. One of the co-authors said there were ‘some decisions [he was] not happy with’. The data used in the K15 paper were only made available through a web site, not in digital form, and lacking proper versioning and any notice that they were research and not operational data. I was dumbstruck that Tom Karl, the NCEI Director in charge of NOAA’s climate data archive, would not follow the policy of his own Agency nor the guidelines in Science magazine for dataset archival and documentation.
I questioned another co-author about why they choose to use a 90% confidence threshold for evaluating the statistical significance of surface temperature trends, instead of the standard for significance of 95% — he also expressed reluctance and did not defend the decision. A NOAA NCEI supervisor remarked how it was eye-opening to watch Karl work the co-authors, mostly subtly but sometimes not, pushing choices to emphasize warming. Gradually, in the months after K15 came out, the evidence kept mounting that Tom Karl constantly had his ‘thumb on the scale’—in the documentation, scientific choices, and release of datasets—in an effort to discredit the notion of a global warming hiatus and rush to time the publication of the paper to influence national and international deliberations on climate policy.
Defining an Operational Climate Data Record
For nearly two decades, I’ve advocated that if climate datasets are to be used in important policy decisions, they must be fully documented, subject to software engineering management and improvement processes, and be discoverable and accessible to the public with rigorous information preservation standards. I was able to implement such policies, with the help of many colleagues, through the NOAA Climate Data Record policies (CDR) [link].
Once the CDR program was funded, beginning in 2007, I was able to put together a team and pursue my goals of operational processing of important climate data records emphasizing the processes required to transition research datasets into operations (known as R2O). Figure 1 summarizes the steps required to accomplish this transition in the key elements of software code, documentation, and data.
Figure 1. Research to operations transition process methodology from Bates et al. 2016.
Unfortunately, the NCDC/NCEI surface temperature processing group was split on whether to adopt this process, with scientist Dr. Thomas C. Peterson (a co-author on K15, now retired from NOAA) vigorously opposing it. Tom Karl never required the surface temperature group to use the rigor of the CDR methodology, although a document was prepared identifying what parts of the surface temperature processing had to be improved to qualify as an operational CDR.
Tom Karl liked the maturity matrix so much, he modified the matrix categories so that he could claim a number of NCEI products were “Examples of “Gold” standard NCEI Products (Data Set Maturity Matrix Model Level 6).” See his NCEI overview presentation all NCEI employees [ncei-overview-2015nov-2 ] were told to use, even though there had never been any maturity assessment of any of the products.
NCDC/NCEI surface temperature processing and archival
In the fall of 2012, the monthly temperature products issued by NCDC were incorrect for 3 months in a row [link]. As a result, the press releases and datasets had to be withdrawn and reissued. Dr. Mary Kicza, then the NESDIS Associate Administrator (the parent organization of NCDC/NCEI in NOAA), noted that these repeated errors reflected poorly on NOAA and required NCDC/NCEI to improve its software management processes so that such mistakes would be minimized in the future. Over the next several years, NCDC/NCEI had an incident report conducted to trace these errors and recommend corrective actions.
Following those and other recommendations, NCDN/NCEI began to implement new software management and process management procedures, adopting some of the elements of the CDR R2O process. In 2014 a NCDC/NCEI Science Council was formed to review new science activities and to review and approve new science products for operational release. A draft operational readiness review (ORR) was prepared and used for approval of all operational product releases, which was finalized and formally adopted in January 2015. Along with this process, a contractor who had worked at the CMMI Institute (CMMI, Capability Maturity Model Integration, is a software engineering process level improvement training and appraisal program) was hired to improve software processes, with a focus on improvement and code rejuvenation of the surface temperature processing code, in particular the GHCN-M dataset.
The first NCDC/NCEI surface temperature software to be put through this rejuvenation was the pairwise homogeneity adjustment portion of processing for the GHCN-Mv4 beta release of October 2015. The incident report had found that there were unidentified coding errors in the GHCN-M processing that caused unpredictable results and different results every time code was run.
The generic flow of data used in processing of the NCDC/NCEI global temperature product suite is shown schematically in Figure 2. There are three steps to the processing, and two of the three steps are done separately for the ocean versus land data. Step 1 is the compilation of observations either from ocean sources or land stations. Step 2 involves applying various adjustments to the data, including bias adjustments, and provides as output the adjusted and unadjusted data on a standard grid. Step 3 involves application of a spatial analysis technique (empirical orthogonal teleconnections, EOTs) to merge and smooth the ocean and land surface temperature fields and provide these merged fields as anomaly fields for ocean, land and global temperatures. This is the product used in K15. Rigorous ORR for each of these steps in the global temperature processing began at NCDC in early 2014.
Figure 2. Generic data flow for NCDC/NCEI surface temperature products.
In K15, the authors describe that the land surface air temperature dataset included the GHCN-M station data and also the new ISTI (Integrated Surface Temperature Initiative) data that was run through the then operational GHCN-M bias correction and gridding program (i.e., Step 2 of land air temperature processing in Figure 2). They further indicated that this processing and subsequent corrections were ‘essentially the same as those used in GHCN-Monthly version 3’. This may have been the case; however, doing so failed to follow the process that had been initiated to ensure the quality and integrity of datasets at NCDC/NCEI.
The GHCN-M V4 beta was put through an ORR in October 2015; the presentation made it clear that any GHCN-M version using the ISTI dataset should, and would, be called version 4. This is confirmed by parsing the file name actually used on the FTP site for the K15 dataset [link]; NOTE: placing a non-machine readable copy of a dataset on an FTP site does not constitute archiving a dataset). One file is named ‘box.12.adj.4.a.1.20150119’, where ‘adj’ indicates adjusted (passed through step 2 of the land processing) and ‘4.a.1’ means version 4 alpha run 1; the entire name indicating GHCN-M version 4a run 1. That is, the folks who did the processing for K15 and saved the file actually used the correct naming and versioning, but K15 did not disclose this. Clearly labeling the dataset would have indicated this was a highly experimental early GHCN-M version 4 run rather than a routine, operational update. As such, according to NOAA scientific integrity guidelines, it would have required a disclaimer not to use the dataset for routine monitoring.
In August 2014, in response to the continuing software problems with GHCNMv3.2.2 (version of August 2013), the NCDC Science Council was briefed about a proposal to subject the GHCNMv3 software, and particularly the pairwise homogeneity analysis portion, to a rigorous software rejuvenation effort to bring it up to CMMI level 2 standards and resolve the lingering software errors. All software has errors and it is not surprising there were some, but the magnitude of the problem was significant and a rigorous process of software improvement like the one proposed was needed. However, this effort was just beginning when the K15 paper was submitted, and so K15 must have used data with some experimental processing that combined aspects of V3 and V4 with known flaws. The GHCNMv3.X used in K15 did not go through any ORR process, and so what precisely was done is not documented. The ORR package for GHCNMv4 beta (in October 2015) uses the rejuvenated software and also includes two additional quality checks versus version 3.
Which version of the GHCN-M software K15 used is further confounded by the fact that GHCNMv3.3.0, the upgrade from version 3.2.2, only went through an ORR in April 2015 (i.e., after the K15 paper was submitted and revised). The GHCN-Mv3.3.0 ORR presentation demonstrated that the GHCN-M version changes between V3.2.2 and V3.3.0 had impacts on rankings of warmest years and trends. The data flow that was operational in June 2015 is shown in figure 3.
Figure 3. Data flow for surface temperature products described in K15 Science paper.
Green indicates operational datasets having passed ORR and archived at time of publication. Red indicates experimental datasets never subject to ORR and never archived.
It is clear that the actual nearly-operational release of GHCN-Mv4 beta is significantly different from the version GHCNM3.X used in K15. Since the version GHCNM3.X never went through any ORR, the resulting dataset was also never archived, and it is virtually impossible to replicate the result in K15.
At the time of the publication of the K15, the final step in processing the NOAAGlobalTempV4 had been approved through an ORR, but not in the K15 configuration. It is significant that the current operational version of NOAAGlobalTempV4 uses GHCN-M V3.3.0 and does not include the ISTI dataset used in the Science paper. The K15 global merged dataset is also not archived nor is it available in machine-readable form. This is why the two boxes in figure 3 are colored red.
The lack of archival of the GHCN-M V3.X and the global merged product is also in violation of Science policy on making data available [link]. This policy states: “Climate data. Data should be archived in the NOAA climate repository or other public databases”. Did Karl et al. disclose to Science Magazine that they would not be following the NOAA archive policy, would not archive the data, and would only provide access to a non-machine readable version only on an FTP server?
For ocean temperatures, the ERSST version 4 is used in the K15 paper and represents a major update from the previous version. The bias correction procedure was changed and this resulted in different SST anomalies and different trends during the last 15+ years relative to ERSST version 3. ERSSTV4 beta, a pre-operational release, was briefed to the NCDC Science Council and approved on 30 September 2014.
The ORR for ERSSTV4, the operational release, took place in the NCDC Science Council on 15 January 2015. The ORR focused on process and questions about some of the controversial scientific choices made in the production of that dataset will be discussed in a separate post. The review went well and there was only one point of discussion on process. One slide in the presentation indicated that operational release was to be delayed to coincide with Karl et al. 2015 Science paper release. Several Science Council members objected to this, noting the K15 paper did not contain any further methodological information—all of that had already been published and thus there was no rationale to delay the dataset release. After discussion, the Science Council voted to approve the ERSSTv4 ORR and recommend immediate release.
The Science Council reported this recommendation to the NCDC Executive Council, the highest NCDC management board. In the NCDC Executive Council meeting, Tom Karl did not approve the release of ERSSTv4, noting that he wanted its release to coincide with the release of the next version of GHCNM (GHCNMv3.3.0) and NOAAGlobalTemp. Those products each went through an ORR at NCDC Science Council on 9 April 2015, and were used in operations in May. The ERSSTv4 dataset, however, was still not released. NCEI used these new analyses, including ERSSTv4, in its operational global analysis even though it was not being operationally archived. The operational version of ERSSTv4 was only released to the public following publication of the K15 paper. The withholding of the operational version of this important update came in the middle of a major ENSO event, thereby depriving the public of an important source of updated information, apparently for the sole purpose of Mr. Karl using the data in his paper before making the data available to the public.
So, in every aspect of the preparation and release of the datasets leading into K15, we find Tom Karl’s thumb on the scale pushing for, and often insisting on, decisions that maximize warming and minimize documentation. I finally decided to document what I had found using the climate data record maturity matrix approach. I did this and sent my concerns to the NCEI Science Council in early February 2016 and asked to be added to the agenda of an upcoming meeting. I was asked to turn my concerns into a more general presentation on requirements for publishing and archiving. Some on the Science Council, particularly the younger scientists, indicated they had not known of the Science requirement to archive data and were not aware of the open data movement. They promised to begin an archive request for the K15 datasets that were not archived; however I have not been able to confirm they have been archived. I later learned that the computer used to process the software had suffered a complete failure, leading to a tongue-in-cheek joke by some who had worked on it that the failure was deliberate to ensure the result could never be replicated.
Where do we go from here?
I have wrestled for a long time about what to do about this incident. I finally decided that there needs to be systemic change both in the operation of government data centers and in scientific publishing, and I have decided to become an advocate for such change. First, Congress should re-introduce and pass the OPEN Government Data Act. The Act states that federal datasets must be archived and made available in machine readable form, neither of which was done by K15. The Act was introduced in the last Congress and the Senate passed it unanimously in the lame duck session, but the House did not. This bodes well for re-introduction and passage in the new Congress.
However, the Act will be toothless without an enforcement mechanism. For that, there should be mandatory, independent certification of federal data centers. As I noted, the scientists working in the trenches would actually welcome this, as the problem has been one of upper management taking advantage of their position to thwart the existing executive orders and a lack of process adopted within Agencies at the upper levels. Only an independent, outside body can provide the needed oversight to ensure Agencies comply with the OPEN Government Data Act.
Similarly, scientific publishers have formed the Coalition on Publishing Data in the Earth and Space Sciences (COPDESS) with a signed statement of commitment to ensure open and documented datasets are part of the publication process. Unfortunately, they, too, lack any standard checklist that peer reviewers and editors can use to ensure the statement of commitment is actually enforced. In this case, and for assessing archives, I would advocate a metric such as the data maturity model that I and colleagues have developed. This model has now been adopted and adapted by several different groups, applied to hundreds of datasets across the geophysical sciences, and has been found useful for ensuring information preservation, discovery, and accessibility.
Finally, there needs to be a renewed effort by scientists and scientific societies to provide training and conduct more meetings on ethics. Ethics needs to be a regular topic at major scientific meetings, in graduate classrooms, and in continuing professional education. Respectful discussion of different points of view should be encouraged. Fortunately, there is initial progress to report here, as scientific societies are now coming to grips with the need for discussion of and guidelines for scientific ethics.
There is much to do in each of these areas. Although I have retired from the federal government, I have not retired from being a scientist. I now have the luxury of spending more time on these things that I am most passionate about. I also appreciate the opportunity to contribute to Climate Etc. and work with my colleague and friend Judy on these important issues.
A couple of examples of how the public can find and use CDR operational products, and what is lacking in a non-operational and non-archived product:
- NOAA CDR of total solar irradiance – this is the highest level quality. Start at web site – https://data.nodc.noaa.gov/cgi-bin/iso?id=gov.noaa.ncdc:C00828 Here you will see a fully documented CDR. At the top, we have the general description and how to cite the data. Then below, you have a set of tabs with extensive information. Click each tab to see how it’s done. Note, for example, that in ‘documentation’ you have choices to get the general documentation, processing documents including source code, data flow diagram, and the algorithm theoretical basis document ATBD which includes all the info about how the product is generated, and then associated resources. This also includes a permanent digital object identifier (doi) to point uniquely to this dataset.
- NOAA CDR of mean layer temperature – RSS – one generation behind in documentation but still quite good – https://www.ncdc.noaa.gov/cdr/fundamental/mean-layer-temperature-rss Here on the left you will find the documents again that are required to pass the CDR operations and archival. Even though it’s a slight cut below TSI in example 1, a user has all they need to use and understand this.
- The Karl hiatus paper can be found on NCEI here – https://www.ncdc.noaa.gov/news/recent-global-surface-warming-hiatus If you follow the quick link ‘Download the Data via FTP’ you go here – ftp://ftp.ncdc.noaa.gov/pub/data/scpub201506/ The contents of this FTP site were entered into the NCEI archive following my complaint to the NCEI Science Council. However, the artifacts for full archival of an operational CDR are not included, so this is not compliant with archival standards.
This paper first appeared on the Climate etc blog of Professor Judith Curry, who kindly gave republishing permission to the NZCPR.
Below are comments made by Professor Curry, including to David Rose of the Mail on Sunday, who broke the story, as well as her responses to specific criticisms received:
John Bates received his Ph.D. in Meteorology from the University of Wisconsin-Madison in 1986. Post Ph.D., he spent his entire career at NOAA, until his retirement in 2016. He spent the last 14 years of his career at NOAA’s National Climatic Data Center (now NCEI) as a Principal Scientist, where he served as a Supervisory Meteorologist until 2012.
Dr. Bates’ technical expertise lies in atmospheric sciences, and his interests include satellite observations of the global water and energy cycle, air-sea interactions, and climate variability. His most highly cited papers are in observational studies of long term variability and trends in atmospheric water vapor and clouds.
NOAA Administrator’s Award 2004 for “outstanding administration and leadership in developing a new division to meet the challenges to NOAA in the area of climate applications related to remotely sensed data”. He was awarded a U.S. Department of Commerce Gold Medal in 2014 for visionary work in the acquisition, production, and preservation of climate data records (CDRs). He has held elected positions at the American Geophysical Union (AGU), including Member of the AGU Council and Member of the AGU Board. He has played a leadership role in data management for the AGU.
He is currently President of John Bates Consulting Inc., which puts his recent experience and leadership in data management to use in helping clients improve data management to improve their preservation, discovery, and exploitation of their and others data. He has developed and applied techniques for assessing both organizational and individual data management and applications. These techniques help identify how data can be managed more cost effectively and discovered and applied by more users.
David Rose in the Mail on Sunday
David Rose of the UK Mail on Sunday is working on a comprehensive expose of this issue [link].
Here are the comments that I provided to David Rose, some of which were included in his article:
Here is what I think the broader implications are. Following ClimateGate, I made a public plea for greater transparency in climate data sets, including documentation. In the U.S., John Bates has led the charge in developing these data standards and implementing them. So it is very disturbing to see the institution that is the main U.S. custodian of climate data treat this issue so cavalierly, violating its own policy. The other concern that I raised following ClimateGate was overconfidence and inadequate assessments of uncertainty. Large adjustments to the raw data, and substantial changes in successive data set versions, imply substantial uncertainties. The magnitude of these uncertainties influences how we interpret observed temperature trends, ‘warmest year’ claims, and how we interpret differences between observations and climate model simulations. I also raised concerns about bias; here we apparently see Tom Karl’s thumb on the scale in terms of the methodologies and procedures used in this publication.
Apart from the above issues, how much difference do these issues make to our overall understanding of global temperature change? All of the global surface temperature data sets employ NOAA’s GHCN land surface temperatures. The NASA GISS data set also employs the ERSST datasets for ocean surface temperatures. There are global surface temperature datasets, such as Berkeley Earth and HadCRUT that are relatively independent of the NOAA data sets, that agree qualitatively with the new NOAA data set. However, there remain large, unexplained regional discrepancies between the NOAA land surface temperatures and the raw data. Further, there are some very large uncertainties in ocean sea surface temperatures, even in recent decades. Efforts by the global numerical weather prediction centers to produce global reanalyses such as the European Copernicus effort is probably the best way forward for the most recent decades.
Regarding uncertainty, ‘warmest year’, etc. there is a good article in the WSJ: Change would be healthy at U.S. climate agencies (hockeyshtick has reproduced the full article).
I also found this recent essay in phys.org to be very germane: Certainty in complex scientific research an unachievable goal. Researchers do a good job of estimating the size of errors in measurements but underestimate chance of large errors.
I have known John Bates for about 25 years, and he served on the Ph.D. committees of two of my graduate students. There is no one, anywhere, that is a greater champion for data integrity and transparency.
When I started Climate Etc., John was one of the few climate scientists that contacted me, sharing concerns about various ethical issues in our field.
Shortly after publication of K15, John and I began discussing our concerns about the paper. I encouraged him to come forward publicly with his concerns. Instead, he opted to try to work within the NOAA system to address the issues –to little effect. Upon his retirement from NOAA in November 2016, he decided to go public with his concerns.
He submitted an earlier, shorter version of this essay to the Washington Post, in response to the 13 December article (climate scientists frantically copying data). The WaPo rejected his op-ed, so he decided to publish at Climate Etc.
In the meantime, David Rose contacted me about a month ago, saying he would be in Atlanta covering a story about a person unjustly imprisoned [link]. He had an extra day in Atlanta, and wanted to get together. I told him I wasn’t in Atlanta, but put him in contact with John Bates. David Rose and his editor were excited about what John had to say.
I have to wonder how this would have played out if we had issued a press release in the U.S., or if this story was given to pretty much any U.S. journalist working for the mainstream media. Under the Obama administration, I suspect that it would have been very difficult for this story to get any traction. Under the Trump administration, I have every confidence that this will be investigated (but still not sure how the MSM will react).
Well, it will be interesting to see how this story evolves, and most importantly, what policies can be put in place to prevent something like this from happening again.
Response to critiques: Climate scientists versus climate data
by Judith Curry
Not surprisingly, John Bates’ blog post and David Rose’s article in the Mail on Sunday have been receiving some substantial attention.
Most journalists and people outside of the community of establishment climate scientists ‘get it’ that this is about the process of establishing credibility for climate data sets and how NOAA NCDC/NCEI have failed to follow NOAA’s own policies and guidelines, not to mention the guidelines established by Science for publications.
In this post, I go through the critiques of Rose/Bates made by NOAA scientists and other scientists working with surface temperature data. They are basically arguing that the NOAA surface temperature data sets are ‘close enough’ to other (sort of) independent analyses of surface temperatures. Well, this is sort of beside the main point that is being made by Bates and Rose, but lets look at these defenses anyways. I focus here more on critiques of what John Bates has to say, rather than the verbiage used by David Rose or the context that he provided.
The Data: Zeke Hausfather and Victor Venema
You may recall a recent CE post where I discussed a recent paper by Zeke Hausfather Uncertainties in sea surface temperature. Zeke’s paper states that it has independently verified the Huang/Karl sea surface temperatures.
Zeke has written a Factcheck on David Rose’s article. His arguments are that:
- NOAA’s sea surface temperatures have been independently verified (by his paper)
- NOAA’s land surface temperatures are similar to other data sets
- NOAA did make the data available at the time of publication of K15
With regards to #1: In a tweet on Sunday, Zeke states
Zeke Hausfather @hausfath @KK_Nidhogg @ClimateWeave @curryja and v5 is ~10% lower than v4. Both are way above v3, which is rather the point.
What Zeke is referring to is a new paper by Huang et al. that was submitted to J. Climate last November, describing ERSSTv5. That is, a new version that fixes a lot of the problems in ERSSTv4, including using ships to adjusting the buoys. I managed to download a copy of the new paper before it was taken off the internet. Zeke states that v4 trend is ~10% lower than v5 for the period 2000-2015. The exact number from information in the paper is 12.7% lower. The bottom line is that sea surface temperature data sets are a moving target. Yes, it is good to see the data sets being improved with time. The key issue that I have is reflected in this important paper A call for new approaches to quantifying biases in observations of sea surface temperature, which was discussed in this previous CE post.
Regarding #2. Roger Pielke Sr. makes the point that ALL of the other data sets use NOAA’s GHCN data set. Zeke makes the point that CRUT and Berkeley Earth do not use the homogenized GHCN data. However, as pointed out by John Bates, there are serious problems with the GHCN beyond the homogenization J
Regarding #3. John Bates’ blog post states: “NOTE: placing a non-machine readable copy of a dataset on an FTP site does not constitute archiving a dataset”
Victor Venema has a blog post David Rose’s alternative reality. The blog post starts out with a very unprofessional smear on the Daily Mail. He provides some plots, cites recent articles by Zeke and Peter Thorne. Nothing worth responding to, but I include it for completeness. The key issues of concern are in John Bates’ blog post (not what David Rose wrote).
The fundamental issue is this: data standards that are ‘ok’ for curiosity driven research are NOT ‘ok’ for high impact research of relevance to a regulatory environment.
Peter Thorne and Thomas Peterson
Thomas Peterson, recently retired from NOAA NCDC/NCEI, is a coauthor on K15. He tweeted:
Thomas Peterson @tomcarlpeterson 16h Buoys read 0.12C cooler than ships. Add 0.12C to buoys or subtract 0.12C from ships and you’ll get exactly the same trend.
Response: Well, in the new Huang et al. paper on ERSSTv5, it turns out that adjusting the ships to buoys results in a trend that is lower by 0.07oC. NOT exactly the same – in the climate trend game, a few hundredths of a degree actually matters.
In the comments on John Bates’ post, Peterson writes:
As long as John mentioned me I thought, perhaps, I would explain my concern. This is essentially a question of when does software engineering take precedence over scientific advancement. For example, we had a time when John’s CDR processing was producing a version of the UAH MSU data but the UAH group had corrected a problem they identified and were now recommending people use their new version. John argued that the old version with the known error was better because the software was more rigorously assessed. I argued that the version that had corrected a known error should be the one that people used – particularly when the UAH authors of both versions of the data set recommended people use the new one. Which version would you have used?
John Bates email response:
First, Peterson is talking about a real-time, or what we dubbed and interim CDR, not a long term CDR. So he is using incorrect terminology. In fact the UAH interim CDR was ingested and made available by another part of NCDC, not the CDR program. Of course, I never said, use the old one with the known error. But, what I actually did is check with the CDR program for what they did. And since, yes, this was fully documented I can go back and find exactly what happened as the trail is in the CDR document repository. The CDR has a change request and other processes for updating versions etc. This is done all the time in the software engineering world. As I recall, there was a very well documented process on what happened and it may actually be an example to use in the future.
So, Peterson presents a false choice. The CDR program guidelines are for non-real time data. We did set up guidelines for an interim CDR, which is what Peterson is referring to. So, customers were provided the opportunity to get the interim CDR with appropriate cautions, and the updated, documented CDR when it became available later.
Peter Thorne is coauthor on both Huang et al. ERSST articles. From 2010 to 2013, he was employed by North Carolina State University in the NOAA Cooperative Institute for Climate and Satellites (CICS). He has never been directly employed by NOAA.
Thorne published a blog post On the Mail on Sunday article on Karl et al. Excerpts:
The ‘whistle blower’ is John Bates who was not involved in any aspect of the work. NOAA’s process is very stove-piped such that beyond seminars there is little dissemination of information across groups. John Bates never participated in any of the numerous technical meetings on the land or marine data I have participated in at NOAA NCEI either in person or remotely. This shows in his reputed (I am taking the journalist at their word that these are directly attributable quotes) misrepresentation of the processes that actually occurred. In some cases these misrepresentations are publically verifiable.
Apparently Peter Thorne does not know much about what goes on in NOAA NCDC/NCEI, particularly in recent years.
Response from John Bates:
Peter Thorne was hired as an employee of the Cooperative Institute for Climate and Satellites NC in 2010 and resigned in may or June 2013. As such, Thorne was an employee of NC State University and not a government employee. He could not participate in government only meetings and certainly never attended any federal manager meetings where end-to-end processing was continuously discussed. As I discussed in the blog, my Division was responsible for running the ERSST code and the global temperature blend code from 2007-2011. We had begun more fully documenting that code including data flow diagrams and software engineering studies. In addition, my Division ingested and worked with the all the GHCN data and the ICOADS data. I developed extensive insight into how all the code ran, since I was responsible for it. Running of the ERSST and global temperature blend code was transferred to the other science Division in late 2010, prior to the arrival of Thorne at NCDC. Since I remained part of the management team the remainder of my time at NCDC/NCEI, I had deep insight into how it ran.
The key issue is this. John Bates is not a coauthor on any of these studies, and hence doesn’t have any personal vested interest in these papers. However, he is extremely knowledgeable about the subject matter, being the supervisor for the team running ERSSTv3 and the GHCN. He has followed this research closely and has had extensive conversations about this with many of the scientists involved in this research. Most significantly, he has collected a lot of documentation about this (including emails).
So this is not a ‘he said’—‘the other he said’ situation. Here we have a scientist that spent 3 years 2010-2013 at NOAA (but wasn’t employed by NOAA) and is coauthor of two of the papers in question, versus a supervisory meteorologist employed by NOAA NCDC for nearly two decades, that was formerly in charge of the Division handling the surface temperature data and the architect of NOAA’s data policies.
Regarding Thorne’s specific points:
1. ‘Insisting on decisions and scientific choices that maximised warming and minimised documentation’
Dr. Tom Karl was not personally involved at any stage of ERSSTv4 development, the ISTI databank development or the work on GHCN algorithm during my time at NOAA NCEI. At no point was any pressure bought to bear to make any scientific or technical choices. It was insisted that best practices be followed throughout. The GHCN homogenisation algorithm is fully available to the public and bug fixes documented. The ISTI databank has been led by NOAA NCEI but involved the work of many international scientists. The databank involves full provenance of all data and all processes and code are fully documented.
Response: Thorne has not been on site (NOAA NCDC) for years and not during the final few years when this took place.
3. ‘The land temperature dataset used by the study was afflicted by devastating bugs in its software that rendered its findings unstable’ (also returned to later in the piece to which same response applies)
The land data homogenisation software is publically available (although I understand a refactored and more user friendly version shall appear with GHCNv4) and all known bugs have been identified and their impacts documented. There is a degree of flutter in daily updates. But this does not arise from software issues (running the software multiple times on a static data source on the same computer yields bit repeatability). Rather it reflects the impacts of data additions as the algorithm homogenises all stations to look like the most recent segment. The PHA algorithm has been used by several other groups outside NOAA who did not find any devestating bugs. Any bugs reported during my time at NOAA were investigated, fixed and their impacts reported.
Response: Thorne left NOAA/CICS in 2013. As outlined in the original blog post, the concern about GHCN was raised by a CMMI investigation that was conducted in 2014 and the specific concerns being discussed were raised in 2015. I cannot imagine how or why Peter Thorne would know anything about this.
4. ‘The paper relied on a preliminary alpha version of the data which was never approved or verified’
The land data of Karl et al., 2015 relied upon the published and internally process verified ISTI databank holdings and the published, and publically assessable homogenisation algorithm application thereto. This provenance satisfied both Science and the reviewers of Karl et al. It applied a known method (used operationally) to a known set of improved data holdings (published and approved).
Response from John Bates:
Versioning of GHCNmv4 alpha – So, after I sent my formal complaint on K15 to the NCEI Science Council in Jan 2016, I pressed to have my concern heard but sessions were booked. I pressed on at the end of one of the meetings and specifically brought up the issue of versioning in additional to not archiving. Russ Vose Chairs the Science Council and Jay Lawrimore, who runs the GHCN code, were in attendance. I made my argument that the version in K15 was in fact V4 alpha and should have been disclosed as such whit the disclaimer required for a non-operational research product. I said that the main reason for changing the version number from 3 to 4 was the use of ISTI data per what Jay Lawrimore had briefed. Moreover, plots of raw, uncorrected ISTI vs GHCN 3 on the ISTI page (will find link after I send these thoughts) show that there are 4 decades in the late 1800 and early 1900s where there is a systematic difference of 0.1C between the two. The reason for this has to be explained before the data are run through the pairwise, and so since there is not GHCNv4 peer article doing this, provenance is lacking. The notion that the ISTI peer article does this is wrong. ISTI web site specifically says it is not run through pairwise and that is a later step.
I concluded, thus K15 uses GHCN v4 alpha consistent with the file name. Russ Vose then said, ‘no it’s version 3’. However, then Jay Lawrimore said, ‘John’s right, it’s version 4’. There was some awkward silence and, since the meeting was already over time, folks just left. So, contrary to Thorne I do meet with these folks and I discussed this very issue with them AND Jay Lawrimore who actually runs the GHCN data said I was right. Thus, Thorne is wrong.
5. [the SST increase] ‘was achieved by dubious means’
The fact that SST measurements from ships and buoys disagree with buoys cooler on average is well established in the literature. See IPCC AR5 WG1 Chapter 2 SST section for a selection of references by a range of groups all confirming this finding. ERSSTv4 is an anomaly product. What matters for an anomaly product is relative homogeneity of sources and not absolute precision. Whether the ships are matched to buoys or buoys matched to ships will not affect the trend. What will affect the trend is doing so (v4) or not (v3b). It would be perverse to know of a data issue and not correct for it in constructing a long-term climate data record.
Response: The issue is correcting the buoys to ships, and the overall uncertainty of the data set and trend, in view of these large adjustments
6. ‘They had good data from buoys. And they threw it out […]’
v4 actually makes preferential use of buoys over ships (they are weighted almost 7 times in favour) as documented in the ERSSTv4 paper. The assertion that buoy data were thrown away as made in the article is demonstrably incorrect.
Response: Verbiage used by David Rose is not the key issue here. The issue is the substantial adjustment of the buoy temperatures to match the erroneous ship values, and neglect of data from the Argo buoys.
7. ‘they had used a ‘highly experimental early run’ of a programme that tried to combine two previously seperate sets of records’
Karl et al used as the land basis the ISTI databank. This databank combined in excess of 50 unique underlying sources into an amalgamated set of holdings. The code used to perform the merge was publically available, the method published, and internally approved. This statement therefore is demonstrably false.
See response to #4.
What needs to happen next to clarify the issues raised by John Bates?
We can look forward to more revelations from John Bates, including documentation, plus more detailed responses to some of the issues raised above.
An evaluation of these claims needs to be made by the NOAA Inspector General. I’m not sure what the line of reporting is for the NOAA IG, and whether the new Undersecretary for NOAA will appoint a new IG.
Other independent organizations will also want to evaluate these claims, and NOAA should facilitate this by responding to FOIA requests.
The House Science Committee has an enduring interest in this topic and oversight responsibility. NOAA should respond to the Committee’s request for documentation including emails. AGU and other organizations don’t like the idea of scientist emails being open to public scrutiny. Well, these are government employees and we are not talking about curiosity driven research here – at issue here is a dataset with major policy implications.
In other words, with the surface temperature data set we are in the realm of regulatory science, which has a very different playbook from academic, ‘normal’ science. While regulatory science is most often referred to in context of food and pharmaceutical sciences, it is also relevant to environmental regulations as well. The procedures developed by John Bates are absolutely essential for certifying these datasets, as well as their uncertainties, for a regulatory environment.