Examination of Witnesses (Question Numbers
200-219)
PROFESSOR JOHN
BEDDINGTON, PROFESSOR
JULIA SLINGO
AND PROFESSOR
BOB WATSON
1 MARCH 2010
Q200 Chairman: He has said that this
afternoon.
Professor Beddington: He has explained
that to you. Well, taken together, I think that is an extremely
comprehensive inquiry. Could I also just concur while I have the
chance with the comments of Julia Slingo? You did not pose me
the question about whether what happened at UEA would undermine
the IPCC. Professor Slingo answered that and I completely concur
with her answer.
Q201 Dr Naysmith: I was just going
to ask Professor Slingo: one of the things that we have encountered
in previous sessions is that the Met Office has to be asked for
its data to be released. Is there any reason why there would be
any reason to withhold any data that the Met Office is making
available to CRU at East Anglia?
Professor Slingo: We are not withholding
any data that we have permission to release. You have to remember
that the observations that went into the CRU dataset are fundamentally
owned by the national met services, who, through the World Meteorological
Organisation, have requirements upon them to make some of that
publicly available. Those were the data we released in December.
Q202 Dr Naysmith: So it is really
just a formality?
Professor Slingo: It is a formality.
We wrote to 170 met services. We have so far had 58 have given
permission and you will see that we now have two-thirds, soon
to be three-quarters, of all the data that were used in the CRU
record publicly available on our website.
Q203 Mr Boswell: Are some of those
offices still very closely connected to their national ministries
of defence? Is that one of the constraints, that they are diffident
about publishing or releasing the data?
Professor Slingo: There are one
or two that we have had replies from where they have ... Let us
be clear; it is very few. Some governments see these data as having
commercial value, and I should say that one of the reasons that
we have now moved this whole construction of these datasets into
the auspices of the World Meteorological Organisation is to deal
with some of these issues about freedom of data. I think this
is a really important international issue that we must get sorted
out, because these datasets are absolutely crucial to the debate
around climate change.
Q204 Graham Stringer: Professor Beddington,
some of the written submissions we have suggest that it is accepted
that if a tobacco company pays for research into smoking, it tends
to find that there is no damage done—that is the historical
evidence. Is there a similar factor at work when tax dollars from
the United States or tax pounds from this country go into research
that the funding source affects the outcome? Having asked the
question, do you think research should be done into that area?
Professor Beddington: I do not
think I would pose it there. If you are seeing research which
is being funded, the control of the funder is to pose the question,
in a sense, "We would like this particular piece of research
done" but I do not believe there is any control whatsoever
on the answer, certainly not in UK science.
Q205 Graham Stringer: So you do not
think it is worth research being done?
Professor Beddington: I have not
seen that there is a glaring problem in this area. I think that
there are some areas of research which are problematic and where
there is considerable controversy and I think the key thing there
is to know that there is a controversy and that one does it, but
in other areas of concern, even when there is controversy, there
is an overwhelming burden of evidence and support from the scientific
community. So I think it important to say it is not a soccer match—that
is how I would probably put it—when you are in scientific
controversy; it is not necessarily equal sides. In some areas,
albeit that there are disputes about areas, the overwhelming base
of support of the scientific community is in one direction or
the other. On this? No, it is about even.
Q206 Graham Stringer: Professor Slingo,
were you here when Muir Russell gave his evidence?
Professor Slingo: No, I was not.
We were outside.
Q207 Graham Stringer: He basically
referred to the Wegman analysis of the argument about the hockey
stick between McIntyre and Mann and the reason I refer to that
is because you cite all the peer review that went on into the
different assessments done by the IPCC but Wegman, after 10 years
of argument, seemed to side with McIntyre, who was in a tiny minority
on this issue, certainly in terms of the statistical mistakes
that had been made in the original Mann paper. Does that not give
you some cause to worry about the peer review process at that
level?
Professor Slingo: Not at all,
no. The controversy around the original methods of Mann et
al has been fully addressed in the peer reviewed literature
and I think those issues are now largely resolved. As we have
already said, the unequivocal rise in temperature during the latter
half of the 20th century is now supported by many other variables
in the climate system, let alone surface temperature, as we have
already discussed. This goes back—and Bob Watson may like
to comment on this because this was around when you were in charge
of IPCC, I believe, Bob.
Professor Watson: Yes. I think
the key point is, a lot was made of the hockey stick but a much
more important issue is what has happened in the last half-century
of trying to say what has actually occurred within the climate
system, what has changed, and whether we can attribute cause and
effect. Therefore, the theoretical modelling that has been done
has tried to ask whether we can explain the observed changes,
especially since 1950, on natural phenomena alone—changes
in solar radiation, changes in volcanic activity—and the
answer is no. So one then asks the question whether we can explain
the observed changes on a combination of natural phenomena and
how we humans have affected the atmosphere, both with greenhouse
gases and with aerosols and changes in land surface albedo, etc
and the answer is yes, and that is when the IPCC came up with
the conclusion that it is very likely, greater than 90% probability,
that we humans are the major cause for the observed changes in
the Earth's climate since the Second World War. So the hockey
stick was an interesting idea of trying to look over the very
long temperature of, say, the last thousand years or so, the tree
ring data, and what has happened in the more modern instrumental
record. It is all part of science to have a paper like the Mann
paper and it can be challenged. I think it is now a fully resolved
issue. On the issue you mentioned of funding, I think the tobacco
issue is one of the worst examples where an industry did fund
research and the results were clearly distorted, but in other
areas—stratospheric ozone depletion, for example—the
private sector has done superb research and it was very honest
and had great integrity and was very consistent with research
funded both in the United States and within the UK. So I would
argue, having been a funder of research both in the US and in
the UK, that one tries to get all issues on the table to get to
the truth.
Q208 Graham Stringer: Can I go back
to Professor Beddington? I know you do not want to comment on
the University of East Anglia case but I will try and ask it in
a more general way, although it is obviously stimulated by East
Anglia. Do you think it is acceptable that climate scientists
or, for that matter, any other scientists refuse to publish the
computer programmes that have provided the basis for their published
papers?
Professor Beddington: I think
there is a question of timing. Computer programmes which actually
incorporate a methodology should be published and should be available.
I think that the methodology itself, which essentially is a mathematical
algorithm which specifies what the computer programme is supposed
to do, should be published and that should be generally available
because computer programmes have errors in them, and therefore
one of the reasons why one would want to publish them is to actually
get that out. That is a general principle. I would certainly support
that.
Q209 Graham Stringer: That leads
me nicely on to my last question. Is there a problem with scientific
software? We have had emails from Professor Darrel Ince and from
Professor Les Hatton saying that there are severe problems with
scientific software. Do you think that is a general problem in
UK or world science?
Professor Beddington: I would
probably ask Julia to comment in the context of climate change
science, and I think that there are some issues here. For example,
some of the coding in physics runs to a million lines and that
can be extremely difficult to actually guarantee that this is
correct but in fact, again, it is the weight of evidence, the
fact that a number of people are working on the same thing, and
you would expect when you get a set of people, a large number
of people, working on it that problems will start to emerge. Some
of these things are extremely complicated but in the context of
the climate change discussions, Julia could probably comment better
than I.
Professor Slingo: Yes. Around
the UEA issue, of course, we did put the code out at Christmas
time, before Christmas, along with the data because I felt very
strongly that we needed to have the code out there so that it
could be checked. If you think about the sorts of codes that we
use in climate modelling, we are literally talking of hundreds
of thousands of lines of code—I know because I have written
some of them—and of course, there will be errors in them.
At least for the UK the codes that underpin our climate change
projections are the same codes that we use to make our daily weather
forecasts, so we test those codes twice a day for robustness.
Q210 Graham Stringer: You do not
always get it right though, do you?
Professor Slingo: No, but that
is not an error in the code; that is to do with the nature of
the chaotic system that we are trying to forecast. Let us not
confuse those. We test the code twice a day every day. We also
share our code with the academic sector, so the model that we
use for our climate prediction work and our weather forecasts,
the unified model, is given out to academic institutions around
the UK, and increasingly we licence it to several international
met services: Australia, South Africa, South Korea and India.
So these codes are being tested day in, day out, by a wide variety
of users and I consider that to be an extremely important job
that we do because that is how we find errors in our codes, and
actually it is how we advance the science that goes into our codes
as well. So of course, a code that is hundreds of thousands of
lines long undoubtedly has a coding error in it somewhere, and
we hope that through this process we will discover it. Most of
the major testing is very robust.
Q211 Chairman: Can I just follow
up what Ian Stewart asked Professor Beddington? We heard from
the ex-Information Commissioner, Richard Thomas, earlier, who
indicated—and we will have to check the record on this—that
all data—and indeed Dr Naysmith enquired about methodology
as well—if in fact it was funded through the public purse
should be immediately available and should not require a Freedom
of Information request to get it. Is that not something which
you, as the Government's Chief Scientific Adviser, should instruct?
Professor Beddington: Yes. I obviously
was not here to hear that comment. I would be interested to comment
in detail when I have seen it. I think there are some issues to
do with timing. For example, the research councils, NERC in particular,
demand that any analysis that it is funding goes into its data
centre12[13]
but in terms of public bid, quite often scientists will want to
have the first cut at the data so that they can do their analysis,
because that is the way they will get their papers published,
but subsequently that should be available. I think that is probably
a reasonable way to go forward. I think the timing is therefore
going to be an issue. In terms of the comments by the previous
Commissioner on FOI, this is not an area I am well versed in,
but I am very happy to have a look at what he actually said.
Q212 Chairman: Can you see where
we are going with this?
Professor Beddington: Yes, I do.
Q213 Chairman: That is quite an important
issue because if all publicly funded research, the raw data, has
to be made available when it is, if you like, constructed, and
indeed the methodology, it actually saves people around the world
a lot of time and effort, does it not?
Professor Beddington: I think
we must judge it in the context of individual scientists who are
collecting data. There are some very large datasets which have
a significant value, scientists have spent their lives collecting
it, they then publish that and someone else does the analysis
before them. This is somewhat demotivating, but once they have
had the a chance to actually work—and there has to be some
sort of sensible limit to look at data that you have actually
collected—and publish on it, then it needs to be out into
the open. So I think there is a timing issue which does need exploration
for different sorts of subjects.
Chairman: We would just ask you, as the
Government Chief Scientific Adviser, if in fact you would reflect
on those comments and perhaps feed those into the Muir Inquiry
or indeed to make some statement yourself.
Q214 Dr Naysmith: I think the most
interesting series of experiments in this area is Gregor Mendel
and his peas, because of course he was right, but he did refine
his data that he published and people can still go back and do
the original experiment, can they not?
Professor Beddington: Sure. I
think the difficulty arises where you cannot replicate experiments.
If you have experiments that are replicable, then it is very straightforward
to actually do that but taking the climate case, there has only
been one history of our climate, so replication is much more problematic.
Chairman: And there is only one Ian Stewart!
Q215 Ian Stewart: Good evening. The
Chairman really teased out of you the answer to the first part
of the question I would have asked. However, I am conscious, Professor
Beddington, that you said that you would not wish to comment on
the CRU situation because of the investigation that is going on.
In relation to Freedom of Information, we heard earlier that CRU
went from two to four requests a year to 61, a very significant
change in requests, although Richard Thomas did say that the FOI
Commission did not actually anticipate higher requests than that
in general for the CRU. Do you have any sympathy for scientists
who find that they are bogged down by FOI requests and cannot
get on with their research?
Professor Beddington: I think
that is a real issue for institutions that they need to be thinking
about. Where you get an FOI request and it is a reasonably articulated
request and appropriate, you pretty much should answer that. The
volume of work is an issue for institutions but I certainly have
sympathy for the individual scientists who are actually being
in a sense expected to do all this extra work to provide their
information. I think it is an institutional process. If a particular
institution is overwhelmed—putting "overwhelmed"
in inverted commas—or has a very large number of Freedom
of Information requests to it, then it needs to be thinking how
it addresses that as an issue but it should not, I believe, be
a burden on the individual scientist.
Q216 Ian Stewart: Do you both agree
with that?
Professor Slingo: Yes.
Professor Watson: Yes.
Q217 Chairman: Could I move on to
Professor Slingo: on 24 February the Met Office issued a statement
calling for a programme to deliver a new global temperature dataset,
and you used the words "to augment current datasets, to refine
current datasets and to follow on from the pioneering work of
the University of East Anglia." Have you lost confidence
in the existing datasets?
Professor Slingo: Not at all,
no.
Q218 Chairman: So why are you calling
for that to be done?
Professor Slingo: What we are
actually proposing is a new assessment which looks at much higher
temporal resolution data, so in particular, if you think the CRU
dataset looks just at monthly means, which is very helpful if
you are interested in the global warming trend, the average changes
in our climate, but what is very clear now is that we need to
know much more than that; we need to know about extremes, we need
to know about heat waves, daily extremes of temperature, the sorts
of things we had in 2003, and whether those are changing as well.
Many of the impacts of climate change will be felt through changes
in extremes, probably more so than even just the average trend
of global warming. What we have proposed as an international initiative
under the World Meteorological Organisation is to create a new
dataset that looks at daily or even sub-daily temperature, and
following on from temperature then to do rainfall and other key
climate variables that impact on society, ecosystems, biodiversity,
and so forth. This was a really important initiative, one that
actually we had been thinking about for some considerable time.
We had an opportunity to present it to the Commission for Climatology,
which is the appropriate body in the WMO. They meet every four
years. It so happened that they were meeting this year in Turkey
and we decided in discussion both with CRU and many international
scientists that we should push ahead with this because we feel
it is badly needed.
Q219 Chairman: In your submission
to the Committee you say that there are numerous studies that
have tested the robustness of surface temperature records but
we have received, and our stuff is now published, many submissions
with contrary views. Why is there this confusion? Is that just
normal scientists taking different views? The data is all there.
Professor Slingo: Yes, as we have
shown since we have released the data. The robustness of the temperature
record, independent of how many stations you have, is not what
our initiative is about at all.
13 12 Note by witness: NERC requires that grant recipients
offer to deposit with it a copy of datasets resulting from the
research supported, after it is completed. NERC then selects those
it considers to be of long-term value, which are managed for long-term
re-use and re-purposing. The data would normally be offered to
one of NERC's designated data centres. Back
| 
