Extended Interview

Newsmaker Interview
Phil Jones Defends Record of U.K. Climate Center

The release last fall of e-mails stolen or leaked from the Climate Research Unit (CRU) at the University of East Anglia in the United Kingdom has generated a fierce debate about both the process by which scientists around the world have carried out research on climate change, and what they have found. Phil Jones, who has been director of the unit since 2004, agreed to talk with Science about aspects of the controversy relating to the scientific record itself. This is a transcript of his 5 February interview with Eli Kintisch.

Q: Why are you speaking out now?

P.J.: It just seems like the right time. It was too difficult back in November and December.

Q: Let's pretend for a second that we threw out the CRU dataset. What other data are available that corroborate your findings about temperature rise?

P.J.: There's the two other datasets produced in the U.S. [at NASA and NOAA]. But there's also a lot of other evidence showing that the world's warming, by just looking outside and seeing glaciers retreating, the reduction of sea ice ... overall, the reduction of snow areas in the northern hemisphere, the earlier [annual] breakup of sea ice and some land ice and river ice around the world, and the fact that spring seems to be coming earlier in many parts of the world.

Q: Critics say that the fact that your work produces the same trends as the NASA and NOAA data sets is insignificant since you start with the same raw data.

P.J.: There are differences. The two American sets use a larger number of [temperature] stations than we do. They both use about 7200 stations and we use about 5000 stations. But we look at that data in different ways. We have different techniques of deciding whether the stations are used or not, and different ways of putting it all together. And if you then produce the time series for the global hemispheric averages, the results look pretty much the same. But also with the [National Climatic Data Center] analysis, you can look at just the raw station data, using no adjustments whatsoever. And that produces much the same result for the global hemispheric averages from 1880 to the present. That's including every station but making no adjustments to any of them.

Q: One of the real challenges is going from the raw data that's available to anyone to the final temperature sets that you release. Do you feel that you have released enough information so that someone could repeat that exercise?

P.J.: Yes, I feel we have. [Our papers] have been peer-reviewed; we've been doing this work for almost 30 years now. [NOAA] has something called the Global Historical Climatology Network (GHCN), and people can download the station data--it's essentially the same data [although] it may not be exactly the same. They could go and take that data, make their own choices about what stations to use. ... They could reproduce their own gridded temperature data. A lot of the people at the moment criticize what we do but [are not doing] anything constructive and new.

Q: One concern of your critics is whether there are adequate procedures in place to assure the quality of this data.

P.J.: That's the sort of work we've done in the past, and published in the papers.

Q: You've emphasized that you have a small staff. Wouldn't having more people available to check on these data be a useful thing?

P.J.: It could be useful, but then we've got to bring them up to speed in terms of what we're looking for. The national meteorological services are doing quality control on this data before it even reaches us.

Q: Have you ever encountered an example of raw temperature data being so unreliable that you had to exclude it?

P.J.: There have been examples of that. In terms of examples of data for a particular month at 1500 stations, I should let you know there is a handful of 10 or so stations where the number is incorrect, so we have a real-time quality control test.

Q: There's a story in the press about these [flawed data from] Chinese stations?

P.J.: The main point is that I did a second paper in 2008 and I got the same station data from the Chinese Meteorological Administration (CMA). For the same two sets of 42 stations and in a figure in that paper, which is in a press release, we showed that the average of the stations for the two networks--rural and urban--was exactly the same as the new data I got from CMA. So essentially the average of the two sites is the same for the same period, which was 1954 to 1983.

Q: When in your career did the pressure from outsiders to criticize or, as you put it, distort your work become significant?

P.J.: In 2007, [once] the blog sites started. I had responded to some of these people in years earlier, but had given up. ... I just didn't have the time to respond. They didn't seem to want to understand. I was trying to be helpful then, so I essentially gave up.

Q: One of those skeptics was Warwick Hughes, who wanted station data. What did you mean when you wrote in an e-mail "even if WMO [World Meteorological Organization] agrees, I will still not pass on the data? We have 25 or so years invested in the work."

P.J.: I'd rather not go there. It was an e-mail written in haste.

Q: When did the pressure grow most severe?

P.J.: In July 2009 we received 60 Freedom of Information requests in a few days--each request was for five countries' worth of data.

Q: [But] a sticking point with some of your critics has been how much of the data isn't available.

P.J.: We've been, with the Met office, canvassing the stations to see what versions of the data that we have we can release. A lot of countries have said yes, but a few countries have said no. But we've been putting up more of the data online on the Met office site. Up on the site now is 80% of the stations we use. You can download the data; you can download the program we use to produce the datasets. It's 80% of the data rather than a full 100%. When you do that, you find that 80% of the stations almost reproduce exactly the 100% of stations for the global averages. I've got a diagram here that shows that.

Q: You're a widely published scientist and with a lot more prestige than your critics. Why then were you reticent to share the data with these outsiders?

P.J.: I was pointing them to where they could get the same data and they could do their own analysis. They are intent on repeating the analyses of others. They are of course entitled to do whatever they want to, it's a free world, but they just don't seem intent on wanting to do their own work. They just seem intent on wanting to repeat what others have done. They're trying to slow us down, and waste time.

Q: When did you first become interested in creating and maintaining these datasets?

P.J.: This was in the early '80s. There was a lot of data available, magnetic tapes and things, and we just thought it was a good thing to have done at that time.

Q: In what way was your work pioneering?

P.J.: Our work was just bringing the data together more than some people had done in the 1960s and 1970s, just because technology had moved on a bit. So we could get access to more station data, we could use more sophisticated techniques to bring it all together. The results showed up in the first paper, we compared with the analyses that were used in the '60s and '70s and even one analysis that had done something similar in the 1880s. We weren't doing anything novel. But we were using the technology to do something better, to go down to the monthly annual timescale, whereas people in the past were just looking at 5-year averages.

Q: What was your background--what made you particularly interested in temperature records?

P.J.: I was a hydrologist before this. And I had just come down to Norwich a few years earlier, for my first job after getting a Ph.D.

Q: What was the best moment in your career at East Anglia?

P.J.: I don't really know. I've had several highlights. Some of them have been getting some of the papers published.

Q: Such as?

P.J.: Well maybe the first temperature paper. Some of the later ones--and some other papers that I've been involved with. ... Not particularly the ones on the temperature data, some of the other ones.

Q: Why were those papers particularly meaningful for you?

P.J.: I thought we'd done some good work and we needed to show these things.

Q: At what point did you start to think that your research wasn't just interesting from a scientific perspective but possibly also important for society?

P.J.: Probably when the first IPCC [Intergovernmental Panel on Climate Change] report came along in 1990.

Q: As you started to create these datasets and improve them, were there any surprise?

P.J.: Not really. Because the first paper, in 1982, actually shows that the climate had been cooling, slightly, from the 1940s and '50s. So it was after that time that we sort of saw that the temperatures were going up. But it wasn't just our work at the time. The other two groups were still doing similar things. The NASA group with Jim Hansen had published a paper even before us. And the CDC group had had papers over the years as well.

Q: Which of your findings do you think are the most important for society to take note of?

P.J.: The global hemispheric temperature records.

Q: How so?

P.J.: Well, the fact that the temperature has risen, is rising from the 19th century to where we are now, the first decade of the 21st century.

Q: Do you think scientists have managed to get that message across?

P.J.: I think we have, in terms of the three groups that have produced this data, that have almost identically agreed with one another. They have showed that the world has warmed up to the present from the beginning of the 20th century.

Q: One of the important details of your work is how you produce the gridded sets. If I understand the process, the computer code uses algorithms to produce the smooth gridded sets, taking into account in some cases biases between stations and other factors.

P.J.: Yes, the program is available on the Met Office Web site, available for anyone to look at and download and work with.

Q: Is that the so-called "Harry Code"?

P.J.: He was working on a completely different project. That code has nothing to do with the development of the CRU temperature data set. It was a different project.

Q: When did you start to feel that climate science was getting contentious and becoming an area that required a strong constitution?

P.J.: Probably around 2000. But I'm just scientist. But I'm not a savvy politician or anything like that. I've got no agenda here, telling government what to do. I'm just a scientist.

Q: Ben Santer of [Lawrence Livermore National Laboratory] says that some of these outsiders who criticize the work of the CRU dataset are "the forces of unreason." Do you think that's the case? Or are they simply people, like scientists, trying to make a name for themselves by finding errors in the work of others.

P.J.: What you're saying could be true. But I don't really know their motives. Because I tried to steer clear of them for the last couple of years in terms of the blogs. I just want to reiterate: I'd rather be known for my science than my e-mails. I don't want to discuss their particular motives or anything.

Q: You've been through quite an ordeal in the last few months. Did you ever expect this--that your science would lead to what's happened?

P.J.: I stand by the science have done. I'm 100% behind my science.

Q: Did you expect that it would get to the place that it's gotten to?

P.J.: No, but it's got to this because of an illegal act of hacking our e-mails, which I thought were personal things between scientists.

Q: What are your plans as a scientist now, apart from dealing with this Muir Russell review.

P.J.: That's the main thing I have to deal with now. I've really got nothing else to do at the moment. That's the main thing I'm doing.

Q: What have you learned in the last year?

P.J.: I think we need to be more open and proactive in the way we deal with people outside the science. And we're going to try to do that in the future. But as I've said before, I still stand by the science 100%.

Q: Do you have any other comments about any of the allegations that have been raised about your e-mails or science?

P.J.: I would just like you to be aware of the Met Office Web site that I pointed you to, where you can download the program, you can get access to the GHCN [Global Historical Climatology Network] dataset, and people can do their own work.

Q: It sounds to me like you believe that, if this had been available a year ago, you might have avoided some of the problems that have happened.

P.J.: Yes, but you could've gotten access to this GHCN and data many years ago and done this sort of analysis yourself. Anyone could have done it.