Buy

Books
Click images for more details

Support

 

Twitter
Recent comments
Recent posts
Currently discussing
Links

A few sites I've stumbled across recently....

Powered by Squarespace

Discussion > Emails relating to Rob Wilson's post "Large-scale temperature trends" 5/6/12

Hi everyone,

On 5th June, Rob Wilson wrote a guest post here on Bishop Hill, following Andrew Montford's talk at St Andrews University.

The Met Office and Reading University have received Freedom of Information requests regarding emails relating to this post and the seminar:

Request for emails exchanged between John Kennedy and Dr Rob Wilson
Request for emails exchanged between Dr Ed Hawkins and Dr Rob Wilson

Since all the emails were to do with items of interest to readers of this blog, Rob, Ed, John and myself have decided to make the entire email conversation available to everyone for your perusal. I therefore include them below. They are in two sets, one before Rob's post and one after. For my own convenience I have simply pasted in the last email in each chain, which includes all emails in reverse order below - so in each set, if you wish to read in chronological order, please start at the bottom and work up!

I have used XXX to remove information that would normally be redacted in an FOI release, eg: personal contact information, and names of people who are not directly involved in the conversation (except when we are talking about published papers)

There were also some attachments.

The first set, before Rob's post, is immediately below. The second set will be in my next post in this thread.

Cheers

Richard


=======================================
From: Kennedy, John
Sent: 30 May 2012 08:49
To: 'Ed Hawkins'; Robert Wilson
Cc: Betts, Richard
Subject: RE: AR4 projections

Hi Robert, Ed,

I suspect that the 1-way corrections might also include any data set that has been adjusted to make the past colder and the present warmer. GISS, NCDC, Ocean heat content, individual stations, sea level and so on... In all cases, the adjustments have been made for good reasons, as you point out, but a large number examples are out there in the blogosphere showing adjustments all going the one way.

HadSST3 is not one, but 100 data sets. Because of the uncertainties, some of those have a higher rate of warming, others have a lower rate of warming and that varies with the period considered. HadSST3's sin, in the eyes of some, is to hide real natural variability that's 'clearly evident' in the raw data. The obvious example of an adjustment that massively reduces temperature trends is the Folland and Parker bucket adjustments.

Concerning the flattening, I think it is easy to quantify - stick an OLS trend through it - but hard to understand. What does it mean? Is it significant in some sense of the word? What, if anything, caused it?

I think you might get some stick for saying that it "does not matter". Observations of what the climate is doing are clearly important, but an extra five or ten years of data is an incremental update of our understanding too. We've gone from 150 years to 160 years of data. There's been a bit of a rush in the literature to 'explain' the flattening, but given the difficulties of unambiguously detecting and attributing global temperature changes over multiple decades (IPCC: Most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations) I wouldn't expect a completely certain explanation of what caused a 10-year flattening, or given uncertainties in the measurements, precisely how severe the flattening really is. There are real difficulties in just getting the data needed to make these near-real-time assessments. The observed temperatures get updated regularly, but the estimated forcings don't.

John

John Kennedy Climate Monitoring and Research Scientist
[XXX]
http://www.metoffice.gov.uk
Global climate data sets are available from http://www.hadobs.org

--------------------------------------------------------------------------------
From: Ed Hawkins [XXX]
Sent: 28 May 2012 12:21
To: Robert Wilson
Cc: Kennedy, John; Betts, Richard; [XXX]
Subject: Re: AR4 projections


Hi Robert,
A nice summary - have made a couple of suggestions in the attached version of the document.

I do wish that some of these blogs had more tightly controlled comments without the personal attacks which [XXX] seems to have suffered from in his post. It would help so much if the interesting scientific questions which some have asked weren't diluted with so much rubbish which has to be waded through.

cheers,
Ed.


On 05/26/12 11:13, Robert Wilson wrote:
Hi Everybody.

Have been buried in marking and was not able to think about BH and temperature trends etc,.

However, [XXX] recent blog post on BH reminded me that it might be better to get something out sooner rather than later.


Please see attached.

I have tried to keep it as simple as possible and not be overtly prescriptive – i.e. I want the bloggers to look at the data and make their own decisions – while also ensuring that they have access to the relevant articles if they so want to read the more detailed studies.


Comments welcome. If you feel that something is missing, feel free to add anything and it can be a multi-author post.

Perhaps a part II could include model projects as well, although I still have a couple of weeks of exam related activities to wade through and also need to think about a NERC application.


Anyway, comments would be appreciated

Regards

Rob

From: Ed Hawkins [XXX]
Sent: 15 May 2012 17:51
To: Kennedy, John; Robert Wilson
Cc: Betts, Richard; [XXX]
Subject: Re: AR4 projections


Hi Robert et al.,

I have repeated my masked CMIP5 & HadCRUT4 comparison for DJF and 30-70N (see attached) - not much change in conclusion - the observations are within the ensemble range...

Happy to help with any blog article write up, and accessing CMIP5 data.

cheers,
Ed.


On 05/14/12 10:32, Kennedy, John wrote:

Hi Robert et al.


I haven't looked at quite such a detailed breakdown for the flattening. My impression is that it's more obvious in the SSTs than the land temperature, but that as the breakdown breaks it down further, the increased variability makes it harder to spot it, if indeed it is there.


I've focused mainly on global temperatures - as most discussion in the blogosphere has - or the simple divisions I mentioned before. I've attached plots of GISS global temperatures on which I've plotted running n-year trends for n between 2 and 16. They're marked red if the trend is positive, blue if they are negative. The black line is the GISS global average.


The running trends in the past 15 years turn negative at some point for trends up to 8 years in length. In the post 1970 warming this has happened at two other times both associated with large volcanic eruptions, although Pinatubo was associated with a 10-year "flattening". This kind of tallies with Ben Santer's paper (Separating signal and noise in atmospheric temperature changes: The importance of timescale).


http://www.agu.org/pubs/crossref/2011/2011JD016263.shtml


In his Figure 5 he shows trends in global tropospheric temperatures as probabilities for forced or control simulations. Again the recent period is similar in character to the periods following Pinatubo and El Chichon albeit not so marked.


Even the longest trends shown in my diagram are heading in the direction of flattening out. A flattening can also be seen in Ocean Heat Content in the 0-700m layer. http://www.nodc.noaa.gov/OC5/3M_HEAT_CONTENT/heat_content55-07.png


Whether this counts as a 'marked' or 'significant' flattening seems to depend on where your informal Bayesian prior lies. if you expect a lower rate of warming, or cooling, then you are more sensitive to any downward change in the trend. If on the other hand you think that the models are a somewhat reliable guide to what the climate can do then it's easy to rationalise any slow down as a fluctuation on the up and up, or as some inadequacy in the forcings used in the model projections. A number of papers have been written already explaining the slowdown so it has been acknowledged in the literature as something that needs an explanation.


Cheers,


John

John Kennedy Climate Monitoring and Research Scientist
[XXX]

Global climate data sets are available from http://www.hadobs.org


------------------------------------------------------------------------------

From: Robert Wilson [XXX]
Sent: 13 May 2012 13:44
To: Kennedy, John; Ed Hawkins
Cc: Betts, Richard; [XXX]
Subject: RE: AR4 projections

Hi John, Ed and Richard,

Sorry for my slow response. A busy marking period which does not allow much time to play.

The one thing that struck me when [XXX] spoke was that I was not really aware in my mind that the recent “flattening” in trend had been that marked. It certainly had not jumped out at me when doing my own palaeo work.


The attached PDF, compares extra-tropical (30-90) and tropical (30S-30N) land (both CRUTEM3 and 4) and SST (HADSST2 and 3) anomaly series for each season. I have not attempted to smooth the data in any way or add in linear trends.


Just eye-balling and trying to consider the year-to-year variability within the series, any flattening in the recent period is certainly not consistent between all these different seasonal/regional records.


As the Cohen paper says, winter ET NH land temperatures sows “possibly” a flattening, but we would need a few more years of data, before we could quantify if this was really a significant change in trend. Spring and summer certainly do not show any real flattening.


No flattening is observed in the tropical land temperatures and for the southern hemisphere, it is in fact perhaps summer temperatures that show a little flattening off in the recent period.


A similar story is observed with the SST data.

There is an interest “trend change” in the late 1980s in the ET NH SST data – especially in spring and in winter and with such marked changes, I find it difficult to understand why the sceptical community is jumping on a slight flattening in trend for only some seasons/regions when there is clearly quite large internally driven shifts in variability over the 20th century.


It would be interesting to include on these figures the CMIP5 projections. I somehow think that the recent trends will only come close to the edge of the 95% envelope in only a few situations.


I have at least another week of marking ahead of me, but then will have a delve into KMNI and see if I can access the relevant model data. I might then try and write up something for BH or “climate-lab-book” – maybe with some help?

I am doing this more for my own interest rather than trying to persuade any sceptics though.


Cheers

Rob


From: Kennedy, John [XXX]
Sent: 10 May 2012 11:35
To: Ed Hawkins; Robert Wilson
Cc: Betts, Richard; [XXX]
Subject: RE: AR4 projections


Hi Robert, Ed,


This question of whether observations are outside of projections is an interesting one because it depends on exactly what question you ask. I compared short term (2-15 years) trends from CMIP3 SRES A1B runs for the first two decades of the 21st century (which give a median warming rate of 0.2degC/decade a la IPCC) with observed temperatures trends from GISS, NCDC and HadCRUT data sets. If you don't mask the model data to look like the observations then HadCRUT falls outside the 95% range, but NCDC and GISS fall inside.


Rather than masking explicitly as Ed did, I looked at the Arctic (North of 65N) and the region 60S-60N separately. Over 60S-60N the three observational datasets agree much more closely, but all three now fall off the bottom of the spectrum of model trends for lengths of 14-15 years. For the Arctic all three observational estimates fall within the upper reaches of the spectrum of modelled trends with GISS actually scraping the 95% range on the high side. Overall the global warming lies within the model envelope, as Ed shows, but the way the world happens to be doing that warming seems to be unusual given the spectrum of models I could get hold of. The internal variability of those models is very different from model to model though, so it's not clear how to interpret such results. The other problem is that the forcings used in the projections aren't exactly the same as those used by reality.


It would be interesting to repeat this exercise with the new model runs from CMIP5.


I haven't done an analysis of SST that is similar to the Cohen paper, but his paper does show that in the MERRA analysis there was cooling over the globe in all seasons not just DJF.


John


John Kennedy Climate Monitoring and Research Scientist
[XXX]
http://www.metoffice.gov.uk
Global climate data sets are available from http://www.hadobs.org


------------------------------------------------------------------------------

From: Ed Hawkins [XXX]
Sent: 09 May 2012 15:33
To: Robert Wilson
Cc: Betts, Richard; Kennedy, John; [XXX]
Subject: Re: AR4 projections

Hi Robert,
I am disappointed if [XXX] is still claiming that the observations are outside of the projections, especially as he wrote this blog article:

http://www.bishop-hill.net/blog/2012/3/14/climate-hawkins.html

which referred to a post on my blog:

http://www.climate-lab-book.ac.uk/2012/on-comparing-models-and-observations/

which showed that exactly how you do the comparison is important. I have attached an updated version of the figure I made using HadCRUT3 and CMIP3 models, to using HadCRUT4 and the latest CMIP5 models. This demonstrates that the observations are within the uncertainties (but it is quite close - another 3-5 years of little warming and we will be outside). The important part is that masking the models in the same way as the data makes quite a difference.

I have only looked at the annual, global mean, but you may be interested in this paper
http://www.agu.org/pubs/crossref/2012/2011GL050582.shtml

which suggests that the lack of warming over land seems to be in winter only, and dominated by the NH extra tropics. It would be relatively easy to repeat what I have done for this region if we think it is useful? John may be able to tell us whether the SSTs show a similar signal?

As for getting the data - I could calculate some timeseries for CMIP5 models quite easily, but another possibility is to use the Climate Explorer (http://climexp.knmi.nl), hosted by [XXX] at [XXX]. There you can manipulate the CMIP3 and CMIP5 output, and download the raw data, although it is not always very intuitive.

A final alternative, which looks interesting is the DDC:
http://www.ipcc-data.org/ddc_gcm_intro.html
which hosts the global mean data from AR4, and gridded data from all the ARs!

I would be happy to host anything you write on my climate-lab-book blog, but you may get more impact over at BH!

cheers,
Ed.

On 08/05/12 23:27, Betts, Richard wrote:

Hi Robert,


Good to hear from you. Yes, there are definitely some people at Bishop Hill who simply do not want to sit down calmly and talks things over! However, there are also lots of other "lurkers" out there who actually do appreciate a voice of sanity, but stay quiet! I think it would be good if you came back at some point and gave some more insights on palaeoclimate issues (I struggle with that - not my area!)


[XXX] is *wrong* that the recent flattening is outside the error bars of the models, and I'm a bit surprised and disappointed that he said that because we clearly showed him that it wasn't when he came to visit. It is still within the bounds of internal variability - only just, I'll admit, but still within!


The AR4 data can be obtained from the British Atmospheric Data Centre (BADC). Ed Hawkins at Reading uses it a lot, he may be able to advise if needed. I'm not a direct "user" myself, although people in my team are, but I thought I'd put you in touch with Ed because he reads and comments at BH occasionally too.


Differences between CRUTEM3 and CRUTEM4 are pretty trivial, well within error bars, and obviously not fiddled like some people seem to want to think! :-) My colleagues John Kennedy and [XXX] may be able to help if you have any questions about CRUTEM.


Cheers


Richard

Professor Richard Betts Head of Climate Impacts
[XXX]
http://www.metoffice.gov.uk
Twitter: @richardabetts



------------------------------------------------------------------------------

From: Robert Wilson [XXX]
Sent: 01 May 2012 09:05
To: Betts, Richard
Subject: AR4 projections

Hi Richard,

As you saw, I had a bit of a run in at BH.

Not my most diplomatic interaction, but some private e-mails had really annoyed me and I guess that goaded me into my 2nd response. Oh well – hopefully my later posts calmed things down a little, although there are people at BH who don’t really want to list I fear.

A quick Q. I told [XXX] last week before all this Blog mess that I might do an odd guest post. [XXX] put up a graph showing the “flattening” of large scale temperatures over the last 10 years or so. I asked him if that was prevalent in all seasons and how different the trends were for the extra-tropics (N+S) and the tropics. He did not know and I thought it would be a relatively easy thing to post.

At the same time, he was concerned of differences between CRUTEM3 and 4. So I can compare both at the same time.

What would perhaps be interesting, is to place the different AR4 scenarios onto these plots as well. Is this possible? – for each season, extra-tropics (N+S) and tropics etc? If so, where could I get the relevant data? [XXX] said that the flattening was already outside the error envelope of the scenarios. Is that correct? I find that hard to believe.

Rob

-----------------------------------------------------------------------

Dr. Rob Wilson
[XXX]

http://www.st-andrews.ac.uk/gsd/people/rjsw/

".....I have wondered about trees.

They are sensitive to light, to moisture, to wind, to pressure.

Sensitivity implies sensation. Might a man feel into the soul of a tree

for these sensations? If a tree were capable of awareness, this faculty

might prove useful. "

"The Miracle Workers" by Jack Vance

-----------------------------------------------------------------------


-- ----------------------Dr Ed Hawkins - Research Scientist - NCAS-Climate EMAIL: [XXX] www.met.reading.ac.uk/~ed/BLOG: www.climate-lab-book.ac.ukTWITTER: @ed_hawkins Dept. of Meteorology, [XXX] -----------------------------------

-- ----------------------Dr Ed Hawkins - Research Scientist - NCAS-Climate EMAIL: [XXX] www.met.reading.ac.uk/~ed/BLOG: www.climate-lab-book.ac.ukTWITTER: @ed_hawkins Dept. of Meteorology, [XXX]
--
----------------------
Dr Ed Hawkins - Research Scientist - NCAS-Climate

EMAIL: [XXX]
WEB: www.met.reading.ac.uk/~ed/
BLOG: www.climate-lab-book.ac.uk
TWITTER: @ed_hawkins

[XXX]
------------------------------------------------------

Jul 24, 2012 at 11:30 AM | Registered CommenterRichard Betts

The second set of emails, after Rob's post, is here:

============================================
From: Robert Wilson [XXX]
Sent: 06 June 2012 21:58
To: Kennedy, John; Edward Hawkins
Subject: RE: AR4 projections

Hi John,

Indeed – funny how it went. Not sure if writing the odd guest post is really going to be useful as we seem to be demonised whatever we do.

Oh well – an attempt at outreach – probably failed – maybe some of the quiet lurkers spent a bit of time reading some of the papers and have an open mind.

As for FOI – well, I have nothing to hide. Some of the e-mails seemed inappropriate and I was quite polite about it. If the Uni does get an FOI, I hope they will see sense and not follow it through.

Thanks for the comments

Back to the day job

Rob


From: Kennedy, John [XXX]
Sent: 06 June 2012 17:16
To: Robert Wilson; Edward Hawkins
Subject: RE: AR4 projections


Hi Rob,

That went very weird very quickly. I hope the FOI thing was a joke.


I was thinking of chipping in, but I'm not sure it's worth the hassle.


John


John Kennedy Climate Monitoring and Research Scientist
[XXX]
http://www.metoffice.gov.uk
Global climate data sets are available from http://www.hadobs.org



--------------------------------------------------------------------------------

From: Robert Wilson [XXX]
Sent: 05 June 2012 19:46
To: Edward Hawkins
Cc: Kennedy, John; Betts, Richard; [XXX]
Subject: RE: AR4 projections

Hi Ed,

Indeed – just had a skim through

Have been busy with final degree classification meetings most of today.

Will try and find some time tomorrow to make some selective comments.

Rob

From: Edward Hawkins [XXX]
Sent: 05 June 2012 17:45
To: Robert Wilson
Cc: Ed Hawkins; Kennedy, John; Betts, Richard; [XXX]
Subject: Re: AR4 projections

Hi Rob,

Nice post. Have been away and travelling all day so not able to wade in.

Some rather predictable comments though from a quick glance on my phone. Several advocate not doing any corrections to raw data whatsoever, which could perhaps be discussed further. Many others link the science to policy decisions which is unfortunate - we need to separate these in my view. And some will just never believe anything that a climate scientist says!

Cheers,

Ed.

--------------

Ed Hawkins

NCAS-Climate


On 5 Jun 2012, at 08:45, "Robert Wilson" [XXX] wrote:

Morning All,My post is now up on Bishop Hill.http://www.bishop-hill.net/blog/2012/6/5/large-scale-temperature-trends.htmlAm in meetings most of today so will probably not have much chance to read through until this evening.Feel free to wade inJRegardsRob

-----------------------------------------------------------------------

Dr. Rob Wilson

[XXX]

http://www.st-andrews.ac.uk/gsd/people/rjsw/

".....I have wondered about trees.

They are sensitive to light, to moisture, to wind, to pressure.

Sensitivity implies sensation. Might a man feel into the soul of a tree

for these sensations? If a tree were capable of awareness, this faculty

might prove useful. "

"The Miracle Workers" by Jack Vance

-----------------------------------------------------------------------

Jul 24, 2012 at 11:32 AM | Registered CommenterRichard Betts

This thread is going to be moderated hard. Please be very careful about what you post, ensuring comments are on-topic and polite. Do not pester for responses to your questions.

Jul 24, 2012 at 12:02 PM | Registered CommenterBishop Hill

This is not to be rude. It is merely a preliminary observation.

I feel Rob Wilson has just peppered his emails with editorial commentary, which add nothing to the conversation other than slanting the views of the other person, just like a lot of blog comments :).

On openmindedness: rather than hoping to brainwash th lurkers, why not confront those to bother to ask questions and spend time writing comments head-on?

Sceptics want climate scientists to be open-minded enough to admit *in public* that the consensus position is not fully buttressed by the evidence presented to support it, highly plausible though it may be. Climate scientists want sceptics to be open-minded enough to accept the possibility that the consensus position may be right, the various percieved and real niggles notwithstanding. These are parallel, and mirror image argument tracks and they will not converge. Both merely refer to a state of mind after having looked at a case that has been presented.

As far as Rob Wilson's previous guest post is concerned, he makes the following comment:

As for temperature trends, in the same way that it does not really matter if the medieval period was warmer or cooler than today, it does not really matter if a particular seasonal time series shows an increase or flattening in temperatures. What MATTERS is that we need to understand the drivers of these changes.

This is a commonly made statement.

Whether the above is true or not, depends on two things:
1) Are there multicentennial change systems that exist in the climate? Do we know what they are?
2) Is the study of paleoclimate useful in trying to understand the drivers of climate change, as an independent verifier of causative factors conjectured and tested for in climate models prospectively, by providing us with actual number figures for the change (in the climate) and the causative factors (like CO2, etc), albeit those of a lower resolution?

I believe the answer to both these questions (a 'no' and a 'yes') invalidate Rob Wilson's emphatic statement. It is merely a misunderstanding and an inadvertent and forced tactical withdrawal in the argument landscape. The paleo goose is cooked so no one wants to go there. Just six years ago, the dominant paradigm was completely different with regards to the importance of paleoclimate in attributing change to CO2.

Jul 24, 2012 at 12:46 PM | Registered Commentershub

I appreciate John Kennedy's comment "Over 60S-60N the three observational datasets agree much more closely." I have wondered whether this metric (average surface temperature between 60N and 60S, perhaps dubbed AST60) would be superior to the usual one, due to the difficulties of obtaining accurate measurements in the polar regions and the controversy over e.g. the GISS extrapolation of land temperatures over large marine regions. Of course, one would have to get used to modified values for the usual "figures of merit" such as sensitivity, whether K/(W/m2) or K/(CO2 doubling). But some of the noise would go away. It would not seem difficult to evaluate models' results over this region rather than globally.
.
The metric would seem to have the additional advantage that it is more relevant to the majority of human experience. As the Arctic already experiences changes quite different from the global average (or 60N-60S average), I think that it is more meaningful to have a separate metric for that region.
.
Of course, there will be those who have reflexive (polemical) reasons for favoring/opposing a metric with a smaller numerical value. But from a more scientific perspective, what is your opinion of this?

Jul 24, 2012 at 12:46 PM | Registered CommenterHaroldW

Rob Wilson's comment:

"If the Uni does get an FOI, I hope they will see sense and not follow it through."

Are we meant take that as tongue-in-cheek or indicative of intent?

Jul 24, 2012 at 1:24 PM | Unregistered CommenterGixxerboy

"Yes, there are definitely some people at Bishop Hill who simply do not want to sit down calmly and talks things over! However, there are also lots of other "lurkers" out there who actually do appreciate a voice of sanity, but stay quiet!"
One of the reasons (imo) the CA is so influential is that the discussions stay on topic and are usually restrained. I think this style carries a lot of weight with the lurkers, however the debate turns out. By the way can anyone come up with a better term than 'lurker' for those that read but don't usually post.

Jul 24, 2012 at 1:38 PM | Unregistered CommenterEddy

More emails here...
http://www.warwickhughes.com/blog/?p=1688#comments

Jul 24, 2012 at 1:47 PM | Unregistered CommenterMarch

[Snip - O/T. You can discuss the emails or Rob's post, but please don't use this thread to vent frustrations with the climate debate as a whole]

Jul 24, 2012 at 1:52 PM | Unregistered CommenterScottish Sceptic

From: Kennedy, John [XXX]
Sent: 06 June 2012 17:16
To: Robert Wilson; Edward Hawkins
Subject: RE: AR4 projections


Hi Rob,

That went very weird very quickly. I hope the FOI thing was a joke.


I was thinking of chipping in, but I'm not sure it's worth the hassle.


John

Unless the Met Office people grow up from this _demonising_ attitude, where every BH commenter is seen not as an individual but as part of some Evil Skeptics group, I'm afraid much of the dialogue will be meaningless.

So far I can only sense some good understanding of this in Richard Betts. Hopefully the revolutionary idea of talking to people as if they were people, will start spreading.

Jul 24, 2012 at 2:13 PM | Registered Commenteromnologos

Omnologos

Demonising? C'mon, get real.

Jul 24, 2012 at 2:23 PM | Registered CommenterBishop Hill

"Hi John,

Indeed – funny how it went. Not sure if writing the odd guest post is really going to be useful as we seem to be demonised whatever we do. (...)
Back to the day job

Rob"

Jul 24, 2012 at 2:25 PM | Registered CommenterMartin A

Who is M. Ackroyd who originated this FOI? (http://www.whatdotheyknow.com/user/m_ackroyd)

Jul 24, 2012 at 2:34 PM | Unregistered CommenterJack Cowper

M Ackroyd is me.

Jul 24, 2012 at 2:39 PM | Registered CommenterMartin A

Bish - as noticed by Martin, I wasn't the one inaugurating that particular page of the dictionary 8-)

Jul 24, 2012 at 2:41 PM | Registered Commenteromnologos

"If the Uni does get an FOI, I hope they will see sense and not follow it through."

Are we meant take that as tongue-in-cheek or indicative of intent?
Jul 24, 2012 at 1:24 PM Gixxerboy

Please let me have copies of all emails received by Dr Wilson that
he considers would be embarrassing to other commenters on the
Bishop Hill blog. Please also let me have copies of all emails
received by Dr Wilson as a result of his article which he
considered were insulting.


St Andrews University refused the FOI request on the grounds:

[1] That the University does not hold the emails (for the purposes of the FOISA) because:
- The postings Dr Wilson made to the blog were an act of his own volition, unconnected with his employ at the University;
- These postings are considered to be a personal matter concerning Dr Wilson;

[2] Even if they were held by the Unversity ( for the purposes of the FOISA), personal data and confidentiality would preclude the emails being released.

The University has been asked to review their decision.

Jul 24, 2012 at 2:46 PM | Registered CommenterMartin A

O/T for this thread but the FOI'd emails on Warwick Hughes' blog (linked by March above) show what to expect in the resubmitted Gergis paper.

Jul 24, 2012 at 3:02 PM | Unregistered Commenterredc

Martin - point [1] sounds quite puzzling. If overturned, it might be interpreted as to warrant editorial control by Universities about anything written by their employees?

Methinks somebody's got the FOI principle very wrong.

Jul 24, 2012 at 3:08 PM | Registered Commenteromnologos

Jul 24, 2012 at 3:08 PM omnologos

Not sure that I understand your point

If a university lecturer posts stuff on a blog on a subject unconnected with their work (model train collecting, say) then, even if done on a university computer, it is not FOI-able.

If it's done as part of their job, then related information held by their employer is FOI-able.

Jul 24, 2012 at 3:45 PM | Registered CommenterMartin A

This part is rather interesting

"I looked at the Arctic (North of 65N) and the region 60S-60N separately. Over 60S-60N the three observational datasets agree much more closely, but all three now fall off the bottom of the spectrum of model trends for lengths of 14-15 years. For the Arctic all three observational estimates fall within the upper reaches of the spectrum of modelled trends with GISS actually scraping the 95% range on the high side. Overall the global warming lies within the model envelope"

This is quite clearly an example of post hoc analysis bias. He is quite subjective in his choice of what to look at and when he finds conformation bias, he is comforted. He did not a prior, set out to examine where the various models correlate best and worst with the various temperature series; for instance he could have blindly wrote a program to generate a difference spectrum of (model minus data series) for all global areas. Instead, he looks for what he wants to find.
This is by no means a criticism of his scientific ethics or his personal honesty, it means that he is quite human. The double-blind trial is the cornerstone of biomedical research for this very reason, people are people. It is very easy to disregard data you 'don't like' and to concentrate on the 'good' stuff.
He did not a prior, have any reason to believe that the two banded sets; (North of 65N) and (60S-60N) generated by the models would have intrinsic properties distinct from reality in different ways. He finds that one is bumping the top of the CI and the other is below the bottom; then he states 'Overall the global warming lies within the model envelope'. This is rather like finding the front tire of your bike is flat and that the rear is over inflated, but being happy that they are, on average, at the required pressure.
I believe that this paragraph of Kennedy sums up my dissatisfaction with the field, and this dissatisfaction is based on the fundamental philosophy of science.
The philosophy is rather simple; formulate an hypothesis, test the hypothesis against observation, reject the hypothesis of it fails to match observation.
With the usage of complex computer models we have a problem; either a computer model is a 'hypothesis' and the various runs of this program constitute an experimental output that can be tested against reality and then be REJECTED or they are not.
If a model is a complex hypothesis, then Kennedy should reject the hypothesis, because the model has failed to match reality; as he states

'but all three now fall off the bottom of the spectrum of model trends for lengths of 14-15 years. For the Arctic all three observational estimates fall within the upper reaches of the spectrum of modelled trends with GISS actually scraping the 95% range on the high side"

If a model is not a hypothesis, in the classical sense of being falsifiable, then what exactly is it? How can a model be deemed useful if it can tested against reality and found wanting?

Jul 24, 2012 at 3:48 PM | Unregistered CommenterDocMartyn

Richard, you declared here:

"Obviously it would be much better if everyone archived their data and made it available to other researchers."

I then asked some vital questions:

"[W]hat makes one a researcher? How do you define 'researcher[s]'"? (Confer my comment below; you can skip this question, if your definition of a researcher includes people of the general public.)
I think other crucial questions (for instance for Rob and you) were/are whether or not this requirement for "better science" ("make data available") also applies for example to code, parametrizations, homogenizations or other "data"? And also whether or not we, the public around the world, can see what was done exactly, when was it done, where was it done, for what reason was it done...?

With respect to adjustments you, Richard, wrote candidly to Don Keiller here:

"I agree that adjustments, however well justified, should be open, transparent and reproducible, and the data available for people to check."
Can you confirm that you mean with "people" the "general public" (in and outside the UK)? (If you mean in fact "the general public" you can scratch my question from above: "How do you define a researcher")

Are the unadjusted data already available or is the release not definite yet?

Thank you for your time. I'm looking forward to your responses here and/or elsewhere to my (partially vague) questions.

Jul 24, 2012 at 3:52 PM | Unregistered CommenterSeptember 2011

Martin - it just seems like the University believes the point is about "volition" making statements "a personal matter". So the opposite of that would be that anything connected to work, could be a "non-personal matter" thereby subjected to University control.

Jul 24, 2012 at 4:18 PM | Registered Commenteromnologos

So it all appears to be happening in the Arctic, where there is very little actual data and a great deal of extrapolation?

Jul 24, 2012 at 4:19 PM | Unregistered CommenterNW

Hopefully on-topic: "Opinion: Scientists’ Intuitive Failures" by Nisbet and Scheufele

Jul 24, 2012 at 4:21 PM | Registered Commenteromnologos

1:38 PM/ Eddy,

I must admit I do not like the term "lurker"... it sounds either like a dog or somebody doing something they shouldn't. Personally I would much prefer "spectator".

Jul 24, 2012 at 5:13 PM | Unregistered CommenterDavid Porter