from: Tim Osborn <t.osbornatXYZxyz.ac.uk>
subject: Re: Remake of IPCC Figure -- redux
to: "Eugene.R.Wahl" <Eugene.R.WahlatXYZxyza.gov>
to replot the figure as you suggest is feasible for me.
Are you sure about this though? It has always struck me (and Keith, with whom I devised
this way of visualising the published results) as a rather ad-hoc approach and subject to
various lines of attack (e.g. should we really combine reconstructions that represent
rather different things [annual vs. summer, full NH vs. land], what does it mean if the
published uncertainty ranges overlap from multiple studies if some of those studies have
overlapping input proxy series and others have few overlaps?).
I don't want to put you off, and our IPCC chapter co-authors didn't seem put off despite
our (Keith and mine) prior expectations that they would. I just wanted to make sure that
you're clear about the possible criticisms.
Finally (I'm sure you know this anyway!), note that post-AR4 studies are not, of course(!),
represented here. e.g. Mann et al. 2008, Juckes et al. 2007, others? Is that a problem?
Thanks for your best wishes. Things are ok here in Norwich, September has been mild and I
even got up to the beach (north Norfolk) a couple of times! New students are now here and
so I'll be spending much time teaching during the next 2 months. I have a new PhD student
who will probably be studying blocking circulation patterns and whether they are altered by
If you ever want to pay us a visit (especially if you're coming to the UK for other
meetings) here in Norwich, then you'd be very welcome.
At 06:26 30/09/2008, you wrote:
Woops...here is a correction. I meant the 10% overlap lines on both the high and low
sides (rather than the 10% and 90% overlap lines). This can be 5% overlap on the high
and low sides if that makes better sense.
Also, re: the mid line...it occurs to me that there could be temperature "ties" for the
highest percentage of overlap for any given year, so in such a case, I'm imagining that
the mid line would be in the middle of such a range of overlap. Would that be a
sufficient tiebreaking rule to deal with all cases?
Eugene Wahl wrote:
I'm working with David Anderson, my supervisor here at NOAA-Paleo, and others higher up
in NOAA on paleoclimate information to go into the US Government's official overview
document on climate change research.
It is called the "Climate Change Science Plan Unified Synthesis Product", and is like a
kind-of US version of the IPCC SPM--Working Group 1. The goal is to have this ready for
the new government administration as early next year as possible.
We are trying to decide concerning a figure to summarize the high-resolution NH
paleo-climate record for the past 1000 years -- the last version of the document had MBH
99 superposed with the CO2 recored for the same time period. One thought that came down
to Dave and myself from others was to use the new Mann et al, 2008, Figure 3 over the
past millennium or some part of it. Dave and I think the IPCC WG1 Ch. 6 Figure 6:10c is
a better representation of the last millennium from a reconstruction standpoint,
especially representing the uncertainties involved. CO2 would just be left out. [We
also considered IPCC WG1 Ch. 6 Figure 14d as even better, as it shows the EMICs mostly
follow the middle range of the reconstructions' overlap, except that those that don't
have anthropogenic forcing deviate more and more cooler after the mid-20th century.
This coupling of the forcings-driven EMICs with the reconstruction record we find much
more powerful than a superposition of one or more reconstructions and the CO2 curve.
However, this suggestion will not go for the synthesis product, as it is considered too
abstract for that situation.]
Indeed, even Figure 6:10c as it is may be too complex for the target audience we want to
reach with the synthesis product.
In light of the foregoing, the question Dave and I would like to ask you is this: would
it be feasible to ask you to recreate Figure 6:10c with just three lines? These three
lines would be a heavy mid line of the temperature anomalies where the highest
percentage of overlap occurs for each year, and the 10% and 90% overlap lines (all
overlap percentages defined exactly as in the original IPCC graphics). It is our
understanding that you created Figure 6:10c, which is where this request comes from.
Alternatively, could you provide us with these three time series, or direct us how to
extract them from the the IPCC c. 6 archives we keep here? We understand that you are
quite busy and we want to minimize any impact on you in this request.
Please let us know what you think.
Here is hoping you are well, and that the weather there not yet to coolish. I get a
weekly commentary on nature and spirituality written from near UEA ("Word from
Wormingford" it is called), and often think of you-all there as I read it. Please also
say hello to Phil and Keith when you see them.