from: Chick Keller <ckelleratXYZxyzp.ucsd.edu>
subject: Some thoughts on climate change proxy temperatures in the last
to: "Michael E Mann" <mannatXYZxyzginia.edu>, <rbradleyatXYZxyz.umass.edu>, "Phil Jones" <p.jonesatXYZxyz.ac.uk>, <k.briffaatXYZxyz.ac.uk>, tom crowley <tomatXYZxyzan.tamu.edu>, "Jonathan Overpeck"<firstname.lastname@example.org>, Tom Wigley <wigleyatXYZxyzr.edu>, Mike MacCracken <mmaccracatXYZxyzcrp.gov>
Two points here:
1. I read with some consternation Wally Broecker's latest piece in
Science (23Feb. 2001). First you can all take up some other topic
since Wally says only Boreholes and treeline changes are accurate
enough to do low frequency trends. What does he mean by "only two
proxies can yield temperatures that are accurate to 0.5�C"? and do you
agree that tree rings and sediments, etc are not sufficiently accurate
to exhibit correct low frequency trends?
2. Here are some references to recent Holocene time-frame records you
probably have seen, but just in case... I found them interesting
without knowing how good or representative they are.
Surprisingly they were given me by one who cited them as examples of
evidence for a MWP and LIA. I read them differently but they caused me
to consider one question I hadn't heard discussed (see below).
Based just on these four, one comes to the following tentative
conclusions and observations:
*MWP was a generally warm time interspersed with coolings and not well
synchronized hemispherically or globally.
*LIA was global and capable of better (but not completely) synchronized
large amplitude variations
*20th Cent. was the only time when all records agree (tree ring
problems with CO2?)
MIGHT THIS RELATIVE UNIFORMITY BE USED AS A CHARACTERISTIC OF 20TH
CENTU WARMING THAT SETS IT APART FROM PREVIOUS CLIMATE CHANGES?
*Borehole inversion is too smoothed to be of much use but it does
indicate a larger temp amplitude if it weren't smoothed. ~1.5�C
And this brings up my question. How one averages these records.
One way would be to note that the temperature amplitude (1000 - 1950)
for each is ~1.5�C. Thus you could conclude that hemispheric/global
climate varied ay over a degree Celcius (although with regional
Another way would be to average the records. The resulting temperature
amplitude would be smaller because extremes would cancel since
variability is large and each region's extremes occur at different
Thus, if people simply looked at several records they would get the
impression that temperature variations were large, ~1.5�C. Imagine
their surprise when they see that the ensemble averages you publish
have much smaller amplitude.
Comparison of amplitudes is given below (although difficult to do since
amplitude depends on averaging so these are very approximate).
Approximate Temperature Amplitudes for period 1000-1950
Mann et al 1999 ~0.5
Jones et al 1998 ~0.8
Crowley and Lowery ???
Briffa 2000 ???
Dahl, Jensen ~1.5
Huang, et al ~0.8 (500 yrs only)
Overpeck et al ~1.3 (400 yrs, polar only)
Bradley & Jones(93) ~0.7 (600 yrs only)
(Not surprising that the contrarians take great exception to Michael's
<bold><underline>This is important in the current debate even, it would
appear, with people like Wally.</underline></bold> I have been looking
for what the real issue is between researchers like yourselves and
skeptical scientists. Politics and agendas aside, I think it is close
Anyone looking at the records gets the impression that the temperature
amplitude for many individual records/sites over the past 1000 years or
so is often larger than 1�C. They thus recognize that natural
variability is unlikely to generate such large changes unless the sun
is having more effect than direct forcing, or there is some fortuitous
but detectable combination of forcings. And they see this as evidence
that the 0.8�C or so temperature rise in the 20th century is not all
The community, however, in making ensemble averages gets a much smaller
amplitude ~0.5�C. which they say shows that reasonable combinations of
solar direct plus volcanos and internal variability with the help of
THC can indeed explain this AND the 20th century warming is unique.
Thus, the impass--one side pointing to large temperature variations in
many records around the globe and the other saying "yes, but not
synchronous and so averaged hemispherically no big deal.
But, just replying that lack of synchronous events (sometimes by a few
decades) is the reason might not be enough. It seems to me that we
must go one step further. We must address the question: what forcings
can generate large amplitude temperature variations over hundreds of
years, regional though they may be (and, could these occur at different
times in different regions due to shifting heat inertia patterns)? If
we can't do this, then there might be something wrong with our
rationale that the average is low amplitude even though many regions
see high amplitude. This may be the nubbin of the disagreement, and
until we answer it, many careful scientists will decide the issue is
still unsettled and that indeed climate in the past may well have
varied as much or more than in the last hundred years.
(Also, I note that most proxy temperature records claim timing errors
of +-50 years or so. What is the possibility that records are
cancelling each other out on variations in the hundred year frame due
simply to timing errors? as in hitting or missing C&L's triple warming
peak 1000-1200 AD)
Referendes to proxy temp records
<excerpt>(1) Bodri, L. and V.Cermak Climate change of the last
millennium inferred from borehole temperatures: Regional patterns of
climatic changes in the Czech Republic - Part III, Global and Planetary
Change, 21, 225-235. 1999
As with other borehole data the record is incredibly smoothed. It has
essentially three warming features.
from 1000 to after 1500 there is a broad warming pulse;
1900-1950? rapid warming <underline>Total amplitude ~1�C (1.5�C if not
I don't know what to make of the more than 500 year warming pulse.
Most records show warming either in the 1100's or 1200's but usually
The rest of the record looks reasonable given the smoothing.
</excerpt>(2) Filippi , M.L., Lambert, P., et al, Climatic and
anthropogenic influence on the stable isotope record from bulk
carbonates and ostracodes in Lake Neuchatel, Switzerland during the
last two millenia, Jour. of Paleolimnology, 21, 19-34, 1999
<excerpt> Graph actually begins at 805 AD (all dates are advertised as
Starts out warm but already cooling which it does till about 1150.
warms till 1242, second peak 1298 then cools to minimum at 1500
warms significantly to 1600 then cools to about half of 1500 max and
essentially stays that way till 1850 when cools to 1500 level again and
into 1950s and still warming. <underline>Total amplitude
</excerpt>(3) Naurzbaev, M.M. and E.A.Vaganov,Variation of early
summer and annual temperature in east Taymir and Putoran (Siberia) over
the last two millennia inferred from tree rings, JGR 105, 7317-7326,
<excerpt>moderately cool 800-950,
rapid warming to max 1000 dip ~1050, recovers till ~1180
cools fast to minimum ~1250,
</excerpt> warms to max ~1400
cools to 1450 slight cooling till 1700
<excerpt>warms to ~1780
rapid cooling to ~1830
rapid warming till ~1930 <underline> Total Amplitude
</excerpt>(4) Wilson, A.T., Hendy, C.H. and Reynolds, C.P., Short-term
and New Zealand temperatures during the last millennium, Nature 1979,
Used stalagmites (delta 18 O proxy)
</excerpt> This is a strange record, but the authors compare it
favorably with the
<excerpt> central England record.
1100 starts and warms in two pulses one at 1250, min at 1300, big max
at 1400, followed by dive to minimum 1450
rises to max 1500
drops to min 1600
rises a bit 1700 and into 1850
drops to minor min1880 rises after that <underline> Total Amplitude
Charles. "Chick" F. Keller,
IGPP.SIO.UCSD - Attn: Chick Keller
9500 Gilman Drive
La Jolla, CA 92093-0225
(858) 822-1510 office
(858) 456-9002 home
Is the noticeable increase in surfers off Scripps Beach a possible
indication of global warming?