From JSCHLOER@rzmain.rz.uni-ulm.de Fri May 17 13:36:22 EDT 1996
Article: 26368 of sci.geo.meteorology
Path: news2.digex.net!howland.reston.ans.net!blackbush.xlink.net!rz.uni-karlsruhe.de!news.uni-ulm.de!news
From: JSCHLOER@rzmain.rz.uni-ulm.de (Jan Schloerer)
Newsgroups: sci.environment,sci.geo.meteorology
Subject: Glacial Tropics: Warm or Cool ?
Date: 17 May 1996 16:56:23 GMT
Organization: University of Ulm, Germany
Lines: 258
Distribution: world
Message-ID: <4nib3n$q3q@rigel.rz.uni-ulm.de>
NNTP-Posting-Host: main01.rz.uni-ulm.de
X-News-Reader: VMS NEWS v1.25
Xref: news2.digex.net sci.environment:101930 sci.geo.meteorology:26368

Outline of the glacial tropics puzzle.  Repost, some pieces added.

In case anyone misses the deleted remark about glacial trade winds:
they may well have been stronger, but this need not point to a warm
equatorial ocean.  It may also indicate a cold poleward ocean.  Trade
winds off the coast of Venezuela probably were stronger during the
Younger Dryas, a time of cold in the North Atlantic, than before or
thereafter.  At least, if high surface-ocean biological productivity
was, during the Younger Dryas, caused there as it is now by upwelling
of nutrient rich water induced by trade winds  (Konrad A. Hughen,
Jonathan T. Overpeck, et al., Nature 380, 7 March 1996, 51-54).

Off topic, this is a remarkable paleorecord, extending back to
about 15,000 years ago.  During deglaciation, sediments off the
coast of Venezuela indicate a pattern of climatic variation which,
down to the decadal time scale, closely resembles the pattern
seen in the GRIP Greenland ice core, 6000 or so km away.

Yes, the preamble is too long :)

Jan Schloerer
jschloer@rzmain.rz.uni-ulm.de

----



Glacial Tropics:  Warm or Cool ?

               Last revision:  May 1996


Were the glacial tropical oceans warm or cool ?   Outline of a major
unsolved problem in climatology.  A confusing detective story, one
cannot yet tell the victim from the culprit.  Also, an aside to the
endless dispute about the merits and shortcomings of climate models.
If models and data disagree, the models need not always be wrong.
Nor are the models always "tuned" until they show what they are
expected to show.  This isn't my field, corrections and amendments
are welcomed.  Abbreviations:

    LGM      Last glacial maximum, about 18,000 years ago
    CLIMAP   Climate: Long-Range Investigation, Mapping and Prediction

                              **

Around 1990, a summary of general circulation model paleostudies
for the last 20,000 years looked like this [Crowley & North, p 91]:

    Region                      Years ago      Model-data agreement

    Northern high latitudes    18,000 (LGM)        very good
    Low latitudes (tropics)    18,000 (LGM)           poor
    Low latitudes (tropics)     9,000              very good
    Southern high latitudes    18,000 (LGM)           fair
    Ocean circulation          18,000 (LGM)     insufficient work

The second line of the table exhibits the model-data discrepancy
addressed here.  The headaches spring from the LGM tropical ocean.

                              **

In the 1970s the CLIMAP Project investigated sea surface temperatures
during the LGM.  They looked at frequencies of warm- and cold-loving
plankton species whose shells sink to the ocean bottom and can be
retrieved from deep sea cores.  The data indicated that, except for
eastern boundary currents, most of the tropical ocean surface remained
rather warm and was only 1-2 o C cooler than it is today [CLIMAP]
[Crowley & North, p 59].

At the same time, changes in mountain vegetation and a lowering of
the snowline by about 1000 m indicate that at elevations greater than
2000 m the tropics were probably 5 to 6 o C cooler during the LGM
than they are today [Crowley & North, p 49-50] [Stute 95].

Now, if climate models are forced to stick to the warm tropical ocean
surface derived by CLIMAP, then they stubbornly refuse to fit the land
data for the last glacial maximum: the climates modelled over tropical
land areas, especially those for higher elevations, are too warm
[Crowley & North, p 75] [Anderson].  This may well be a case of poor
model performance, indicating a gap in understanding of basic climate
processes.  But it ain't necessarily so  -  the warm tropical CLIMAP
LGM ocean has come under fire.

                              **

In the southwestern United States, LGM annual mean surface temperature
was probably about 5 o C lower than its present-day value, while CLIMAP
says that the surfaces of the adjacent oceans were only 2 o C or so
cooler than today [Stute 92].  Another study strongly suggests that in
equatorial northeast Brazil the LGM was 5 o C cooler than the current
climate.  Again, CLIMAP claims that the neighboring stretch of the
Atlantic was only about 2 o C cooler than it is now [Kerr] [Stute 95].
These discrepancies are hard to reconcile.

Two ice cores from the Peruvian Andes indicate that last glacial
maximum conditions up there were considerably cooler than the modern
Andean climate.  This may also hold for parts of the tropical Atlantic,
where most of the moisture locked in the Peruvian ice probably came
>from  [Thompson] [Broecker 95].  Cores drilled from submerged coral
reefs suggest that, during the LGM, the sea surface near Barbados may
have been 5 o C cooler than it is now [Guilderson] [Anderson].

CLIMAP is on the defensive now, but the above results do not yet
prove that CLIMAP's LGM tropical ocean is indeed too warm.  The corals
>from  Barbados are not a fail-safe thermometre.  Apart from temperature,
other factors like growth rate also affect the composition of coral
remnants [Anderson] [Villiers].  The oxygen thermometre, used for the
Peruvian ice cores, can be tricky as well: the oxygen 18 content of
the ice is affected not just by the temperatures of the moisture source
and of the transporting air, but also by changes of the air's path,
that is, by changes in atmospheric circulation [Thompson].

Still other paleodata tend to support CLIMAP.  The oxygen isotope
composition of fossil plankton shells (planktonic foraminifera)
argues for a warm glacial tropical ocean.  Several lines of evidence
suggest that the surface of the glacial western tropical Pacific
was not much cooler than it is now [Anderson] [Broecker 86]
[Crowley & North, p 59] [Linsley].

                              **

On the other hand, the land data from the southwestern US [Stute 92]
and from Brazil [Stute 95] seem fairly convincing.  They are based
on noble gases dissolved in ground water which was isolated from
the atmosphere long ago.  Solubility depends on temperature, there
is no biology to complicate the interpretation.  A map in [Stute 95]
displays these and several other results which suggest that the land
surface of the American tropics and subtropics was 5 or more o C
cooler during the LGM, with snow-lines and vegetation zones lowered
by 1000 or more metres.  Some data, like lowland pollen records, are
problematic, yet the overall picture looks consistent: fairly uniform
cooling.

Might the ocean surrounding tropical America have been warm, and
the American tropics have been cool during the LGM ?  Climatologists
do not love this idea.  They point out that, except for its lowest
layer, the equatorial troposphere can hardly maintain large temperature
gradients [Pierrehumbert] [Sun].  These days, at least, this is borne
out by observations: at a given time, the temperature of the equatorial
middle troposphere shows little variation [Pierrehumbert].  Thus the
widely observed lowering of snow-lines is probably characteristic
for much of the glacial tropics.

One way out of this dilemma would be stronger cooling at high than
at low elevations, that is, technically speaking, a steeper lapse
rate in the LGM tropics.  Alas, this way out seems to be blocked by
the noble gas data which indicate strong cooling in low-lying land
areas of equatorial Brazil and of the southwestern United States.
But then there might be as yet unknown factors that could explain
the discrepancy between cool land and warm sea.

                              **

Beyond the question whether the models or CLIMAP or both are wrong,
the puzzle has several more implications.

It is crucial for climate change forecasts whether and how far the
tropics are "thermostated" [Pierrehumbert].  Several paleoclimate
reconstructions suggest that, over the past few million years,
equilibrium temperature changes near the equator were often only
a small fraction of those at higher latitudes [Shabalova] [Covey].
How universal is this rule ?  Are there exceptions ?

If CLIMAP's warm LGM tropical ocean should eventually turn out to be
about right, then explaining the glacial tropical climate need not
but could require iconoclastic ideas [Covey].  For instance, one
proposal for reconciling lowered snow lines with a warm ocean surface
in the glacial tropics involves a negative (yes, negative) feedback
>from  atmospheric water vapor [Sun].

On the other hand, if CLIMAP's LGM tropical ocean should be much
too warm, it might be hard to explain why the tropical sea surface
got that cool without invoking relatively high sensitivities to
greenhouse gases [Crowley] [Kerr].  Estimates of the LGM climate's
sensitivity to greenhouse gases depend, among others, on LGM mean
global surface temperature which is uncertain and currently
estimated as 3 to 5, maybe 6 o C cooler than today [Barron]
[Crowley & North, p 79].  Much of this uncertainty stems from
the tropical ocean.

                              **

Were the glacial tropical oceans warm or cool ?  The case is open.
The pitfalls of some methods are not yet fully understood, especially
those of techniques involving remnants of living things [Anderson].
Do "the seemingly invincible oxygen isotope[s]" [Broecker 95] from
the shells of planktonic foraminifera really signal a warm ocean
surface ?   Have the Barbados corals fooled their excavators ?
[Villiers]   Were some of CLIMAP's deep sea core tops damaged, or
have CLIMAP researchers miscounted some microfossils ?  [Emiliani]
There may be a long way to go.

                              **

[Anderson]   David H. Anderson and Robert S. Webb,
    Ice-age tropics revisited.    Nature 367 (1994), 23-24
[Barron]   Eric J. Barron,   Lessons from past climates.
    Nature 360 (1992), 533
[Broecker 86]   Wallace S. Broecker,   Oxygen isotope constraints
    on surface ocean temperatures.   Quaternary Research 26 (1986),
    121-134
[Broecker 95]   W.S. Broecker,  Cooling the tropics.
    Nature 376 (1995), 212-213
[CLIMAP]   CLIMAP Project Members,   The surface of the ice-age Earth.
   Science 191 (1976), 1131-1137    (Preliminary results.  The final
   CLIMAP LGM ocean map appears, e.g., on page 59 in Crowley & North)
[Covey]   Curt Covey,   Using paleoclimates to predict future climate:
   How far can analogy go ?   Climatic Change 29 (1995), 403-408
[Crowley]   Thomas J. Crowley,   Pleistocene temperature changes.
    Nature 371 (1994), 664
[Crowley & North]   Thomas J. Crowley,  Gerald R. North,
    Paleoclimatology.   Oxford University Press 1991
[Emiliani]   Cesare Emiliani, Pleistocene paleotemperatures, Science
    257 (1992), 1462.    Cesare Emiliani, Tropical paleotemperatures.
    Science 268 (1995), 1264
[Guilderson]   Thomas P. Guilderson, Richard G. Fairbanks,
    James L. Rubenstone,   Tropical temperature variations since
    20,000 years ago: modulating interhemispheric climate change.
    Science 263 (1994), 663-665
[Kerr]   Richard A. Kerr,   Chilly ice-age tropics could signal
    climate sensitivity.   Science 267 (1995), 961
[Linsley]   Braddock K. Linsley,   Oxygen-isotope record of sea
    level and climate variations in the Sulu Sea over the past
    150,000 years.   Nature 380 (21 March 1996), 234-237
[Pierrehumbert]   R.T. Pierrehumbert,   Thermostats, radiator fins,
   and the local runaway greenhouse.   Journal of Atmospheric
   Sciences 52 (1995), 1784-1806
[Shabalova]   M.V. Shabalova  and  G.P. K"onnen,   Climate change
   scenarios: Comparison of paleoreconstructions with recent
   temperature changes.   Climatic Change 29 (1995), 409-428
[Stute 92]   M. Stute,  P. Schlosser,  J.F. Clark,  W.S. Broecker,
    Paleotemperatures in the southwest United States derived from
    noble gases in ground water.   Science 256 (1992), 1000-1003
[Stute 95]   M. Stute,  M. Forster,  H. Frischkorn,  A. Serejo,
    J.F. Clark,  P. Schlosser,  W.S. Broecker,  G. Bonani,
    Cooling of tropical Brazil (5 o C) during the last glacial maximum.
    Science 269 (1995), 379-383
[Sun]   De-Zheng Sun  and  Richard S. Lindzen,
    Water vapor feedback and the ice age snowline record.
    Annales Geophysicae 11 (1993), 204-215
[Thompson]   L.G. Thompson, E. Mosley-Thompson, M.E. Davis, P.N. Lin,
    K.A. Henderson, J. Cole-Dai, J.F. Bolzan, K.B. Liu,
    Late glacial stage and Holocene tropical ice core records from
    Huascar'an, Peru.    Science 269 (1995), 46-50
[Villiers]   Stephanie de Villiers, Bruce K. Nelson, Allan R. Chivas,
   Biological controls on coral Sr/Ca and Delta18O reconstructions
   of sea surface temperature.   Science 269 (1995), 1247-1249


This file is archived at:
   http://grumbinescience.org/radix/climate/
    


Jan Schloerer                     jschloer@rzmain.rz.uni-ulm.de
Uni Ulm     Klinische Dokumentation     D-89070 Ulm     Germany