Wednesday, May 23, 2012

4597.txt

date: Thu Oct 1 12:52:08 1998
from: Keith Briffa <k.briffaatXYZxyz.ac.uk>
subject: comments on Thematic bid
to: J.Lowe@rhbnc.ac.uk,njs5@cam.ac.uk,p.j.valdes@reading.ac.uk,sfbtett@meto.gov.uk

Dear all,
I am passing on some comments on the proposal made by 'a modeller' via John Lowe. By now you should have the draft of the proposal that Neville will circulate to the ESTB . There was not time to do much more than some small corrections and to put in a GANTT chart. Neville says the size of the overall text is not a problem now and if it gets further , obviously, things will need changing again . Neville also tells me that I should not now attend the ESTB meeting , which is a relief because I have to go to a PAGES meeting in Belgium next week anyway.
So below are my replies and below that is John's message . I don't have time to cut and paste the appropriate comments and answers so you best just list the whole thing off.
Keith


John,
Thanks for the comments . I have been through them and made some small changes. I don't think that anything more is warranted at this time - especially given the fact that Neville needs it now. It is a bit long but he says this is not vital at the moment and I would rather over emphasise some points now rather than cut lots of stuff out before getting everyone else's say so.

As for your modeller's comments ( I will call him/her M for now)

1. I don't what M is saying here. What fuxes precisely? We are not proposing to use palaeo-data to take the place of , or validate every possible feature or process within the GCM. We wish to produce data and methods with which to test the realism of important features or outputs of an existing model and aid in the interpretation of predictions made with it, and to show how it needs inprovement.The project is about finding out what we can do and can't do with the existing Palaeodata along these lines, and about going and getting the best data to do more. Some data will allow us to compare seasonal variability e.g. tree-ring data show interannual patterns of summer changes in high latitudes, or moisture patterns across North America that can be directly compared with model output. These may not show the full amplitude of regional mean variability over centuries . Other data may only have decadal resolution, but if they capture low-frequency changes with good fidelity ,such as in the behaviour of the NAO over centuries, they would be invaluable.

2. If M thinks we are trying to drive a GCM with palaeo-data , then this is not sensible. Possibly such data could be used to constrain experimental runs to different extents, an issue I know is of interest to some groups (such as Hans von Storch in Germany) but again this would need to be carefully formulatd in terms of the spatial , seasonal and timescale limitations of the various proxies. As for the UK trying to do what the interbnational community are doing -the project is intended to allow UK workers to participate in what is clearly a priority need. Our community has much to offer this effort but it needs the sort of initiative we are proposing to enable it to deliver. Also, there is no restriction on using non UK data. On the contrary, the project should support initiatives in gathering, screening, and exploiting good data from anywhere, provided they match the criteria for joint data/model experiments.
I don't understand the reference to NOAA.

3. The whole point is that the UK has already got one of the most advanced predictive tools in the world i.e. The Hadley Centre Unified Model. This is not to say that it is good enough. Hence, the continuing work to improve it. Meanwhile, it is being used to inform the world about future warming and patterns of coming climate change. We can help reduce the uncertainties in these predictions. By all means someone can go ahead and develop limited-area models where these contibute to our understanding of the GCM, but such development for its own sake ( or for predicting local changes) would be ruled out of this work and anyway, it would need to be driven by a GCM.

4. There is no need to ' approach .. CLIVAR..to establish what the UK could do '. The proposal represents exactly part of what CLIVAR has said needs doing - in its published Implementation Plan. This is clearly stated in the document. Also, who does M think CLIVAR is/are anyway? Surely it is a group of internationally collaborating scientists that include people at the Hadley Centre and members of the UK palaeo-community. As for 'distinctive', where else in the world is there a formal prgramme of coordinated model/proxy data projects?

5. Such a model falls within the scope of this proposal, provided it can be justified and shown to be testable with palaeo-data . Incidently, the Hadley Centre has such a model. Restricting the whole theme to such an end is ,however, going to tell us nothing about the validity of future predictions of global changes , and ( as I said above) must anyway be driven by a GCM.

6. I do not understand this question.

7. Again,(see points 1+2 ) , what inputs does M mean? We are not trying to drive a model , except with additional external forcings gleaned from palaeo-data. True, there may be problems avoiding circularity in distinquihing cause and effect in such work, but this is a different matter.

I am happy to discuss these points further , but only if we get to a stage where this is worth doing. The proposal might not get past ESTB and ASTB. Frankly, the points raised by M do not in my judgement suggest changes in the current draft. If even several of the rest of you think so then we can look at them again later.


I am passing these comments on to Simon , Paul , Nick and Neville to give them the chance to comment but I don't think they should go wider for now.

thanks again for your thoughts and help

Keith





Dear Keith

I have been through the PRESCIENT document (I am not too taken by this
name, but cannot think of a better one at present - how about PRISM -
Palaeoclimatic Research and Integrated earth System Modelling ??).

Some minor details:

1st line, para 2, change 'demands' to requires', as the word 'demands'
could be read as verb or noun, so that the sentence threw me for a
while.

para 3, l.6: "..resolution than AVAILABLE hitherto." ??

under 'General Scientific Context', para 1, sentence 2. I found the
construction of this to be cumbersome. How about "The reality of
climate change on many timescales is now widely appreicated, as is the
need to quantify the different 'natural' and anthropogenic factors that
are believed to 'force' the climate system, an understanding of which
holds the key to successful prediction of future climate changes." Or
something like that ?

Same section, para 2: "To meet these challenges ...." Which ?
The sentence follows on from consideration of "..the only prospect...".
Are you referring to (a) the prospect of the challenges of model
construction, (b) identifying the important parameters to be included
in the model model, or (c) something else ?

Section headed "Background and Potential of Current Pal.. Research"
para 1: palaeoceanography (not -oo- as in your sp.)

para 2. grammatical blip in l. 8 "UK scientists have, and will, make
...." is grammatically incorrect because if you take out the "and will",
the construction is wrong. It needs to be "...have made, and will
make..."


General

It is getting tighter, especially in emphasising the need for REAL
co--operation, from design stage onwards, between the modelling and
palaeo- communities. It is, however, too long, since the max ed = 6
pages at font size 10.

In my view there is a little repetition or over-emphasis of certain
points, and I would recommend omission (or serious reduction) of the
following sections:
1st para under 'Research Programme' - this does not seem necessary to
me.

Much of the penulttimate para (with bullet points) at the end of the
same section, from "Within the above themes....." down to "climate and
vegetation interactions, etc."

I would expand, a little, or emphasise in bold, the final sentence in
this section "Each proposal should test a hypothesis...." This seems
to me to be vital to what is being proposed.


Under "Relevance to UK and Internat...."
Could the first para here be cut right down ?

Under DELIVERABLES
include a statement about giving appropriate advice to UK government
offices and to international collaboration programmes concerned with
global climate change and its environmental consequences. The
deliverables should indicate to whom the deliverables are being
delivered.

___________________________________________________-


Finally (and I hope you do not object), I took the liberty of showing
the document to a numerical modeller, IN STRICTEST CONFIDENCE - who
wishes to remain anonymous for the time being. I simply wanted to see
how a modeller would repond to the document,. I will precis his
comments. You may be able to rebuff some of these outright; on the
other hand you might find them of value, and use them to tighten up
arguments, with the help of UGAMP and Hadley personnel. At least you
will be prepared if the same comments come from ESTB members or from
external referees.

His main comments were as follows:

1) The perception from reading the document is that the fundamental
premise is flawed because the palaeo-comunity will not be able to
deliver time-series data at scales appropriate to represent fluxes in
the models, which are based on 'real time' and can represent seasonal
fluxes (not the annual to decadal scales aimed for AT BEST in the
palaeo-data).

2) Why is the UK community (apparently) trying to do what the
international community is now collaborating to do, on a huge scale ?
Can the UK palaeo-data community hope to provide data-sets that are
anywhere near adequate for 'driving' global models ??? Surely
globally-spread data are going to be needed, and will the UK community
seek to procure them ? Or will we simply work in tandem with NOAA
etc. ??

3) If the UK community wants its own stand-alone modelling programme,
why does it not go for a limited-area model, capable of dealing with
transient climate states (which is a major need), cf. developments under
way in Germany and France, rather than attempt to spread its expertise
(perhaps unwisely) over the globe ?

4) At the very least, an approach should be made to CLIVAR or other
global modelling groups to establish what the UK community could do that
would be distinctive AND meet a perceived need. Links to, and an
established dialogue with, the international collaborative programmes
should be essential first steps, clearly spelled out in the document.

5) Given the perceived importance of the North Atlantic in the global
climate machine, and given the UK's known lead in studying the N
Atlantic region (and recent investment by NERC in studying the region,
such as NEAPACC), why not focus on developing a nested regional climate
model for the N Atlantic which will link in to an established GCM ?

6) What is the numerical stochasticity which should be used for for
palaeo-conditions, and how will that be defined ?

7) Are UK scientists able to provide inputs at the same scale of
resolution and detail for all spatial fields included in the model AND
provide the validation data-sets at the same time ?

There are other queries, concerning the degree of 'determinism' in the
models or 'palaeo-models' and (mis-)match between palaeo-fields and
model-fields, that I am not equipped to deal with.

I therefore have some concerns about how the document might be receivd
by external (overseas) referees. I assume it will go to quite a
number of external referees, in view of the sum of money being sought.
These seem to me to be good questions (though some may need to be
clarified a bit), and if we have good answers to them, then fine. We
can go to ESTB with a spring in our step. It would be dangerous to
press forward, however, if we feel uneasy about how well we could
respond, because we could well be slaughtered by external referees.

Sorry to end so negatively, and I hope my pet modeller has got it all
wrong, (he is not a climate modeller) and that ESTB can be convinced
that our case is a sound one. On reflection, and after receiving
these comments from the modeller, it does look as though what he have
produced is somehting which makes the best of what the UK currently has
to offer (a climate modelling centre with a sophisticated global model
and a palaeo-community engaged in diverse research programmes that have
not traditionally been tied into the modelling efforts). Is the
combination of these two teams what the international community NEEDS
right now to help solve some of the pressing problems ? The external
referees will tell us ! Do Hadley personnel have a handle on this
????

You have done a sterling job with the document. I just wonder
whether we are attempting to bite off more than we can chew, and whether
referees might also think so. I cannot help feeling that a nested,
regional model might provide a much tighter focus, where we can test, at
various timescales, the degree to which, for example, the North Atlantic
circulation pattern is active or passive in the overall process which
leads to abrupt climate shift (and model effects on the UK or NW
Europe). Could we lead the way in developing the most sophisticated,
integrated regional model, which can be incorporated into GCMs ?
Could it be that other major bodies could also develop similar models
for other parts of the global system (Southern Ocean; China Monsoon
system; Arctic Ocean; etc.), and that this is one way of eventually
leading to very sophisticated global models ? Or am I talking twaddle
??


John

No comments:

Post a Comment