I posted on how Europe might end up cooler rather than warmer as a consequence of global warming and its affect on ocean currents, click here for the post.
Following the post, a reader of this blog, Phil Done, suggested in response to a comment from Ian Mott, that Mott really should read up on the phenomenon at Wikipedia.
Mott, who likes working things out for himself, has had a read at Wikipedia and done a bit of a general google and emailed me his findings as follows:
A check of the first pages of google sites dealing with the claimed ice age that would be produced by the collapse of the ‘Atlantic Conveyor’ reveals some interesting stuff. Most carry vague descriptions of how this would take place and seem to indicate that it will be caused by a change in northern salinity levels due to melt water from the Greenland Ice Sheet that will prevent this less dense water from submerging and thereby altering the flow pattern. Most carry the claim that evaporation from the gulf stream currently make this body of water very saline and more dense than the rest of the ocean. All point to the disruption of this salinity level by fresh melt water as the primary agent of disturbed flow pattern.
A good example is http:www.enviroliteracy.org/article.php/545.htm which has a curious link to a graphic called “The Atlantic Thermohaline Circulation” at www.clivar.org/publications/other_pubs/clivar_transp/pdf.files/ . Now the most curious thing about this graphic is that the Gulf Stream is shown as flowing due east from New York to Portugal before heading north past the UK. The generally accepted route up the US east coast appears to have been an inconvenient fact to be ignored for the sake of the story. Even more curious is a “cold, saline bottom current” heading north past New Zealand, through the shallows of Vanuatu into the far north pacific where it surfaces between Alaska and Kamchatka where, curiously, it is supposedly warmed in this sub-arctic clime for the journey south.
The brightest spot was at www.awitness.org/column/global_warming_ice_age.html that rightly pointed out that the mini-iceage of the 1400’s was no such thing and that the other common example provided 12000 years ago was actually a slight pause in the middle of a period of glaciation with little relevance to this scenario.
But it is the actual numbers involved that reveal the truth. According to www.encyclopedia.com/html/G/GulfS1tre.asp the initial speed of the Gulf Stream is 6.4km/hour over a width of 80km. This slows further north as it widens so for the sake of this analysis we assume the Gulf Stream/North Atlantic Current has a median flow of only 3.6km/hr, is 100km wide and about 500m deep. At this speed the entire trip from Florida to Iceland will only take 70 days. And even assuming zero rainfall, the maximum evaporation is only likely to be 70/365 days x 2000mm evaporation = 383mm evaporation per cycle. And this means the 500 metre thick water column is left to absorb the salt reserves of 0.383 metres of evaporated water before it heads south again. As normal ocean salt level is 3.5% then the 499.617 remaining metres of the water column absorbs 3.5% of 0.383 metres of water, a total of 13.4 millimetres of salt that is added to the 17.5 metres of salt in the column. This is an increase of only 0.07657 of 1%.
And as for the claimed impact of fresh water on the current, we have a total of 180 km3 of water flowing past any given point in the north atlantic each hour. This amounts to 1.577 million km3 each year which disperses into the approx 160 million km3 North Atlantic (40 million km2 x average depth of 3.926km). This north atlantic data is derived from the whole of atlantic data from http://en.wikipedia.org/wiki/atlantic_ocean .
The entire ice volume of Greenland is only 2.44 million km3 so a complete melt over a highly improbable twenty years would add only 0.122 million km3 to the annual current flow of 1.577 million km3, for a total of 1.699 million km3 pa. The existing salt in the NA Current, at 3.5% of volume, will be 0.055195 million km3. And this will only reduce the salinity level of the combined current and melt water by one 14th to 3.25%.
A slightly less improbable 100 year total melt, but still very rapid in climatic terms, would involve only 0.0244 km3 per annum. This would produce a combined flow of 1.6014 million km3 with a salinity of 3.45%. This range is only half the variation normally observed within a 1000 metre ocean profile. See www.windows.ucar.edu/tour/link=/earth/water/salinity_dept
In summary:
The North Atlantic currents are horizontal cycles that flow in a clockwise direction due to the rotation of the Earth. They are doing the same thing as the water in a northern hemisphere bathtub and for the same reason. They are, in most part, not vertical cycles with surface water flowing north and sea floor water flowing south so any theory based on a disturbed flow due to lower salinity from ice sheet melt water is a theory that ignores the primary determinants of current flow, the rotation of the earth.
The theory of increased salinity in the Gulf Stream/Nth Atlantic Current due to evaporation ignores the fact that precipitation also takes place in the same regions. Ocean salinity maps indicate that highest salinity is actually in the middle of the North Atlantic (Sargasso Sea with high air pressure and low rainfall), not in the northern regions.
The volume of water in the Gulf Stream/Nth Atlantic Current is of such magnitude that a complete melt of the Greenland Ice Sheet over a period of only 100 years, contracting North at 27km/year, would dissipate the fresh water to such an extent that salinity levels would only drop from 3.5% to 3.45%. This variation is well within the normal range of ocean salinity levels. To suggest that a minor change in the chemical composition of such a large body of water could override the influence of factors of such magnitude as the rotation of the Earth, itself, is pure fantasy.
Consequently, we must conclude that the Pentagon was rather charitable in describing the Atlantic Conveyor/Ice Age scenario as an extreme scenario and very low probability event. It is, in fact, a highly improbable event that not only extrapolates known effects to improbable extremes but also excludes mitigating factors that are of thousands of orders of magnitude greater.
I am not endorsing Mott’s conclusions, but posting them for general discussion.

Jennifer Marohasy BSc PhD has worked in industry and government. She is currently researching a novel technique for long-range weather forecasting funded by the B. Macfie Family Foundation.