Gulf Stream does not determine Europe's temperature?
Michael Tobis
Scientist. It can currently be seen at
http://www.americanscientist.org/template/AssetDetail/assetid/51963/page/1
Richard Seager describes work he did with David Battisti at the
University of Wshington which shows that turning off the Gulf Stream
doesn't affect Europe's temperature very much. It also calls into
question Broecker's favorite awakening-the-beast scenario.
In addition to the intrinsic interest of this argument, interested
non-professionals might want to conside the description of how climate
models were used in obtaining this result. Prognosis is not the only
use of climate models. This article illustrates how such models are
used to advance understanding.
mt
Eric Swanson
Reading Seager's 2002 paper does provide some insight into his
experiments.
http://rainbow.ldeo.columbia.edu//papers/qj.pdf
He describes 2 modeling efforts, one with the GISS model and the other
with the NCAR CCM3. The GISS experiment used a computed sea-ice model,
while the work with the CCM3 relied on proscribed sea-ice. Given that
the feedback of the sea-ice on albedo would be rather strong in the
latitudes of the North Atlantic sub-polar gyre, the differences between
the two experimental results is understandable. The GISS experiment
showed much greater cooling over Europe than that using the CCM3.
That the Gulf Stream does provide considerable warmth to the cold air
which flows over it is not really in question. But, the Gulf Stream
is a wind driven current, whereas the Thermohaline Circulation branches
off and sinks because of differences in density due to salt and
temperature. That the THC occurs is the result of both salt and
temperature and the sinking temperature is very close to the freezing
point of the surface water. The result is that the NADW is only a few
degrees warmer than the freezing point.
If more fresh water is added to the surface layers than has happened in
the recent past, the result may be a weakening or total shutdown of
the THC. If Seager is correct, this alone would not cool Europe
greatly, as he claims that the quantity of heat moved northwards as the
result of the THC is a small term. Most of the flow in the Gulf Stream
would likely continue across the
mid latitudes of the North Atlantic. However, freshening the surface
where the THC has historicly appeared may lower the temperature at
which the surface water would become dense enough to sink. Since this
process already occurs close to freezing, any further cooling would
thus lead to the formation of sea-ice in areas where it would not
otherwise appear.
Seager's claim is that there is a strong interseasonal transfer, with
heating of the oceans in summer providing most of the energy that warms
the atmosphere later during winter. Increasing sea-ice area blocks
this process until the sea-ice melts. This is the so-called "positive
feedback", as the sea-ice has a much larger albedo than that of the
ocean water. Seager’s results with the CCM3 model can not capture
this change in albedo, as the sea-ice area was proscribed in this
experiment. He notes this problem, stating:
“More problematic for certain regions is the treatment of sea
ice. In CCM3 we held the ice covered, but when it was allowed to vary
in the GISS model removal of OHT caused a large expansion of seasonal
ice cover in the Kamchatka region and in the Norwegian and Barents
Seas, cooling the air above and to the east. The thermodynamic sea-ice
model in the GISS GCM probably overestimates the increase in sea-ice
extent...”
I think his focus on Ocean Heat Transport (OHT) alone misses the real
issue, which is, what will happen to the sea-ice as climate warms?
While I am not one of the professional atmospheric scientists, I have
other concerns regarding his explanation of the atmospheric
circulation. For example, AIUI, the low pressure storms in mid
latitudes are not the CAUSE but the RESULT of the flows of large air
masses, primarily between the tropics and polar latitudes. The flows
of warm air poleward and the returning cold air masses rotate as they
move due to the Coriolis effect, thus the boundaries between warm and
cold masses are moving in opposite directions. That’s where low
pressure storms form. He appears to be saying that the storms cause
the flows, based on his interpretation of the math. He focuses
particularly on the “Icelandic Low”, which is found by averaging
surface pressure over a period of time. What this actually means is
that the storm tracks tend to follow similar paths across Iceland, not
that the pressure is always lower over Iceland, AIUI. From this
description, the storms tracking near Iceland result in the lower
pressure which appears in the time series analysis, which is not the
same as saying that the lower pressure found in the analysis causes the
storms.
After just one quick reading, I find it difficult to accept his
conclusions based only on his model results. But, hey, what do I know?
William M Connolley
On Sat, 10 Jun 2006, Michael Tobis wrote:
> A fascinating and accessible article appears in the American
> Scientist. It can currently be seen at
>
> http://www.americanscientist.org/template/AssetDetail/assetid/51963/page/1
Interesting. This is based on the QJ 2002 paper, which we discussed at work.
I've added that link to
and
http://en.wikipedia.org/wiki/Thermohaline_circulation#Impacts_on_global_climate
which you might find interesting (and which can both probably do with some work)
-W.
William M Connolley | w...@bas.ac.uk | http://www.antarctica.ac.uk/met/wmc/
Climate Modeller, British Antarctic Survey | (01223) 221479
If I haven't seen further, it's because giants were standing on my shoulders
--
This message (and any attachments) is for the recipient only. NERC is subject
to the Freedom of Information Act 2000 and the contents of this email and any
reply you make may be disclosed by NERC unless it is exempt from release under
the Act. Any material supplied to NERC may be stored in an electronic
records management system.
James Annan
>
> On Sat, 10 Jun 2006, Michael Tobis wrote:
>
>>A fascinating and accessible article appears in the American
>>Scientist. It can currently be seen at
>>
>>http://www.americanscientist.org/template/AssetDetail/assetid/51963/page/1
>
>
> Interesting. This is based on the QJ 2002 paper, which we discussed at work.
> I've added that link to
>
> http://en.wikipedia.org/wiki/Shutdown_of_thermohaline_circulation#Thermohaline_circulation_and_fresh_water
>
> and
>
> http://en.wikipedia.org/wiki/Thermohaline_circulation#Impacts_on_global_climate
>
> which you might find interesting (and which can both probably do with some work)
I thought that this argument had been largely debunked, and also the 2C
cooling cited on that wikipedia page seems low. Of course, it depends a
bit on precisely which area and what processes are being considered.
James
William M Connolley
On Sun, 11 Jun 2006, James Annan wrote:
> >>http://www.americanscientist.org/template/AssetDetail/assetid/51963/page/1
>
> I thought that this argument had been largely debunked, and also the 2C
> cooling cited on that wikipedia page seems low. Of course, it depends a
> bit on precisely which area and what processes are being considered.
? Which argument - the one they are putting forward, or the one they are
opposing ?
James Annan
>
> On Sun, 11 Jun 2006, James Annan wrote:
>
>>>>http://www.americanscientist.org/template/AssetDetail/assetid/51963/page/1
>>
>>I thought that this argument had been largely debunked, and also the 2C
>>cooling cited on that wikipedia page seems low. Of course, it depends a
>>bit on precisely which area and what processes are being considered.
>
>
> ? Which argument - the one they are putting forward, or the one they are
> opposing ?
Their argument.
http://www.realclimate.org/Rhines_hakkinen_2003.pdf
and eg Vellinga and Wood say a THC shutdown gives a localised cooling of
about 8C:
http://www.ocean.washington.edu/people/faculty/luanne/classes/pcc586/papers/vellingawood_thc2002.pdf
IMO, how one chooses to split up and attribute the energy transport in a
nonlinear system is less interesting than in predicting the overall
change to a specified perturbation. But I guess that's just my POV.
Also, even without the THC, of course western europe would still have a
maritime influence and relatively mild winters (compared to, say,
Moscow). If that is Seager and Battisti's point, then that's fair enough.
James
William M Connolley
On Mon, 12 Jun 2006, James Annan wrote:
> http://www.realclimate.org/Rhines_hakkinen_2003.pdf
Well I don't understand their point about the latent heat (point (i)). Do you?
Apart from that can't see anything wrong. Did Seager ever respond? And did the
promised longer write-up of Rhines ever emerge?
James Annan
>
> On Mon, 12 Jun 2006, James Annan wrote:
>
>>http://www.realclimate.org/Rhines_hakkinen_2003.pdf
>
>
> Well I don't understand their point about the latent heat (point (i)). Do you?
No, but I don't have the Seager paper to hand. It seems like part of the
debate rests on what is meant by "due to the THC" - in a nonlinear
system, you can't just decompose it easily. Also, what is a
"significant" change in the climate?
>
> Apart from that can't see anything wrong. Did Seager ever respond? And did the
> promised longer write-up of Rhines ever emerge?
Dunno. A search for Seager on realclimate throws up a handful of
relevant comments. My meta-analysis (ie what I think others think about
it) is that they somewhat overstated their claims. I think pretty much
everyone else agrees that turning off the THC would have a significant
effect for western europe. But there is some underlying truth that there
are other factors which also help to keep the climate on the warm side.
James
Eric Swanson
> William M Connolley wrote:
> > On Mon, 12 Jun 2006, James Annan wrote:
> >
> >>http://www.realclimate.org/Rhines_hakkinen_2003.pdf
> >
> >
> > Well I don't understand their point about the latent heat (point (i)). Do you?
>
> No, but I don't have the Seager paper to hand. It seems like part of the
> debate rests on what is meant by "due to the THC" - in a nonlinear
> system, you can't just decompose it easily. Also, what is a
> "significant" change in the climate?
Thanks for posting the link to the paper by Rhines and Hakkinen. In my
earlier comment (#2 above), I posted a link to Seager et al. Looks
like my off-hand comment about sea-ice may have been on the mark.
I think that the discussion of THC changes due to increased transport
of fresh water misses an important point. The THC actually occurs at
rather small scales (as I've noted before), below the grid spacings
which have been used by many AOCGCM experiments. Consider that the
experiment by Vellinga and Wood used a 1.25x1.25 deg spacing for the
ocean. In the typical "hosing" experiment, there is a tendency to
spread the fresh water inputs over large areas, thus minimizing the
local impact of the addition. If the actual process which initiates
THC sinking is sensitive to strictly local conditions, these hosing
experiments may understate the impact of changes in fresh water
transport into the Sub Polar Gyre of the North Atlantic. In fact, the
main locations of the THC sinking in the North Atlantic are near the
East Greenland Current, which transports fresher waters from the Fram
Strait to the south and into the eastern side of the Labrador Sea.
The climate response to events such as the Great Salinity Anomaly may
be representative of a component of natural variability which has
been somewhat neglected in the debate about AGW. Adding even greater
amounts of fresh water into such a natural cycle might well push the
cycle much beyond it's previous range, perhaps reaching some "point of
no return" (OK, "tipping point") with unpleasant consequences.
> > Apart from that can't see anything wrong. Did Seager ever respond? And did the
> > promised longer write-up of Rhines ever emerge?
>
> Dunno. A search for Seager on realclimate throws up a handful of
> relevant comments. My meta-analysis (ie what I think others think about
> it) is that they somewhat overstated their claims. I think pretty much
> everyone else agrees that turning off the THC would have a significant
> effect for western europe. But there is some underlying truth that there
> are other factors which also help to keep the climate on the warm side.
One would think Seager had seen the two papers referenced, however, he
did write the new paper in the American Scientist to which MT pointed.
Kooiti MASUDA
confusing.
(This is my thought just after reading his paper in American Scientist.
I feel I must read his paper in QJRMS, but not yet.)
The meaning of "the Gulf Stream is responsible for Europe's mild
winters"
is not self-evident.
Seager refutes <i>his</i> version of that statement.
But I do not think that his interpretation of the statement is shared
by all.
At least <i>my</i> version is somewhat different.
It seems certain that, by "mild winters", Seager meant higher seasonal
average near-surface air temperature in winter.
Let us admit this concept, though the term may have many other
meanings.
Still a question remains: higher than what?
He is mainly concerned with the difference across the Atlantic,
i.e. that Europe has warmer winter than the east coast of North
America.
He says that similar difference exists across the Pacific,
i.e. that the west coast of North America has warmer winter than East
Asia.
Thus, the fact that Kuroshio does not go so far north as the Gulf
Stream does
is not important for <i>these</i> differences.
Since the prevailing wind direction is westerly in this zone,
it is likely that the difference comes from whether the upwind region
is an ocean or a continent. So far, so good.
I want to compare west coast with west coast, east coast with east
coast.
I browsed a few station data, though I have not yet made a serious
study.
Britain seems to be a little warmer than British Columbia (western
Canada) at
the same latitude. (The difference is ca. 2 deg. C in the annual mean,
and the difference in winter is subtler.)
In this sense, the North Atlantic is warmer than the North Pacific.
And the probable cause is the difference of shapes between the Gulf
Stream
and Kuroshio (if morphologically speaking), or the thermohaline
circulation
at least in part (if mechanistically speaking).
Admittedly, this is not so a large part of warmth of Europe in winter.
(I also compare Nova Scotia (eastern Canada) and eastern Hokkaido
(Japan)
around 44 deg. N. So far, I do not find a remarkable difference in this
pair.
But this is just preliminary browsing and not conclusive.
Perhaps I should compare Labrador and Kamchatka.)
In other words, winter warmth can be decomposed into
annual mean warmth and the annual range of temperature.
The Gulf Stream does not matter to the latter,
but it matters somewhat to the former.
I intuitively agree with Seager that many people have overestimated the
effect
of possible shutdown of the thermohaline circulation to European
climate.
But I feel that Seager's argument has gone to the opposite extreme,
perhaps
not by its actual content but by its mood.
P.S.
Though the annual range of temperature is certainly influenced by
land-sea configuration and prevailing winds, it cannot easily be
explained
by just these factors. (Why the maximum exists in Eurasia near its east
end,
while it exists in North America near its center? ....)
Only the comparisons among near-coast locations can be made so
straightforward
that we can safely say that the range would surely be smaller if the
prevailing winds come from the ocean.