Arctic Ice July 2010
    By Patrick Lockerby | June 29th 2010 05:44 AM | 6 comments | Print | E-mail | Track Comments
    About Patrick

    Retired engineer, 60+ years young. Computer builder and programmer. Linguist specialising in language acquisition and computational linguistics....

    View Patrick's Profile
    Arctic Ice July 2010

    In about a week the National Snow and Ice Data Center - NSIDC - will be publishing its analysis of June's sea ice.  I expect them to report another record loss of sea ice.

    My prediction for July is that Arctic sea ice loss will accelerate.

    Measuring ice behaviour

    Ice is a material with mechanical properties.  In order to understand what it is doing at any moment and what it will do in future, one must understand its properties.  Sea ice is a special case of ice as a particulate material.  The fluidity of a particulate mass is a matter of basic physics which is independent of scale.  Particle size can range from microns to kilometers, it is all one to the laws of physics.

    In the absence of a force tending to produce clumping, particles on a fluid surface tend to spread out.  Imagine a swimming pool topped off with a foot thickness of polystyrene beads.  You have to remove a lot of beads before you can see water.  In the absence of air or water movement causing the beads to pile up, they will spread out.  It will be a very long time before removal of the beads reduces their distribution per unit of water area to 15%.

    Measuring ice by extent is like measuring those beads by their distribution.  An entirely arbitrary cutoff point of 15% is used for ice extent.

    For as long as the whole area of ice within the Beaufort Gyre is at greater than 15% distribution, that whole sea area will be included in measures of ice extent as if it was 100% ice covered.  Measures of extent do not discriminate between levels above 15% distribution.

    This map from NSIDC shows the major ice streams in the Arctic regions.

    The Beaufort Gyre and the Transpolar Drift are the two major ice motions in the Arctic.  The enormous forces generated within the pack by these motions are a key driver of ice formation and compaction.  In normal times, as tensions open a lead, it freezes over.  Compression fractures the new ice and squeezes it into the pack.  Ice sheets and slabs over- and under-ride each other.  Broken ice is pushed onto and under the pack.  Even during the summer, when the ice margins are melting, the main body of ice thickens and strengthens.

    Ice thickens by compression when its motion is obstructed.  Remove the obstruction and the ice will not thicken: it will just continue in its motion.  The Beaufort Gyre, the Transpolar Drift and other ice flows can only thicken the ice if there is an obstruction: an opposing force.

    At the point where the Beaufort Gyre and the Transpolar Drift thrust the ice against Ellesmere Island and the North coast of Greenland, the back-pressure of land will normally cause a great thickening of the ice.  Ice is forced ashore until the drag exerted by the ground overcomes the thrust exerted by the ice from seaward.  Thereafter, the ice piles up in huge ridges.  This ice is so thick and so consolidated that it can persist long after the sea ice has melted away.

    Along the coasts of Greenland, Ellesmere Island and Queen Elizabeth Islands the ice would normally be subject to enormous pressures.  Over the course of thousands of years the ice along those coasts built up as ice shelves.  They are gone.  They will not - cannot - return for as long as open passages permit the pressure to be relieved.  In their stead is an area of highly fragmented ice - shown in red below.

    The North West Passage is nearly open.  Just a few more days will see the end of the ice bridge that was blocking the main passage.  The broken ice in the passage will then more readily flow out and melt.

    Ice bridge break-up in Viscount Melville Sound
    MODIS aqua image June 28 2010

    The Nares Strait - the strait which opened in July 2007 and exported a substantial amount of the ice that year - has been open since last year.  Briefly, in March, it appeared that an ice bridge might form, but it did not happen.  In 2009 an ice arch formed in the Lincoln Sea, completely blocking the export of ice through the Nares Strait.

    With the Nares Strait entrance obstructed, two things were observed.  Firstly, the Nares Strait and Baffin Bay were soon swept clear of sea ice.  There was a lot of open water.  Secondly, the ice in the Lincoln Sea area thickened.  This year the Nares Strait is passing a stream of ice into Baffin Bay.  That rapidly melting ice would be thickening the main pack in a normal year.

    Arctic mosaic images for July June 26 2009 and 2010

    Edit:  thanks to Andrew Xnn for spotting my date error, above.

    Ice extent and error doubling

    Ice extent as defined by NSIDC is a count of grid areas of 25 x 25 km which contain at least 15% sea ice.  The method completely misses the loss of ice and consequent loss of mechanical strength within the Beafort Gyre.  The NSIDC method counts ice lost from the Beaufort Gyre as ice added to Baffin Bay.  This shows up in extent plots as a surplus of ice rather than a loss.

    Lesser amounts of ice are being lost all across the Arctic.  The reason, I suggest, is this.  Tightly packed ice pressed together thickens up.  Ice which is so fragmented and so interspersed with open water that it behaves as a fluid will simply vary in its dispersal patterns but will not usually consolidate.  Winds and currents which would normally consolidate the ice merely serve to fragment it more in most places.

    The Beaufort Gyre and the Transpolar Drift normally act as a giant compressor, forcing ice hard up against immovable shores.  This year they are acting more like a giant grinder, increasing fluidity.

    There is at least some thick ice.  The USCG Cutter Healy has been surveying the Arctic near Alaska.  The skipper reports plenty of rubble and ridges in one area of ice surveyed.
    The floe we have spent the last couple days in is a vast floe, miles and miles across, made up of a bunch of smaller flows that mixed up through the long Arctic winter. This resulted in pressure ridges and rubble piles over 30 feet thick, and quite a few areas where two floes got stacked up and are well over 10 feet thick.

    The USCGC Healy's position is shown on a frequently updated map.

    The aloftcam pictures are updated every hour.  This animation shows Healy set in a floe.

    Animation of aloftcam images from:

    To get a better impression of scale, watch lower left as the gangway is lowered and figures appear on the ice.  Lumps of ice can look small from a satellite, or even a webcam.  It takes a human figure to give a true sense of scale.

    I shall continue to monitor the Arctic ice.  I hope to continue posting updates with my observations and predictions.

    Edit: update #1 for July is now posted: Arctic Ice July 2010 - Update #1

    Related / further reading:


    nice comment and interesting blog. However, I have to admit, that I disagree with your wording of doubling the error. Although this statement is correct in pointing out difference between ice volume and ice extent, I think the extent measure shall be left as is, so extent as a measure remains intact. The reason I think is, that from information I have, extent is easier to measure and that measurement itself as a technical process is more reliable. I think that extent is good measure for mid term trend prediction, but since it is more prone to differences in ice configuration. Ideally, we would have good Ice volume measurements, but we don't have them yet.
    Nevertheless, good work.

    Patrice: thanks for the feedback.

    I do take your point about ice extent.  Elsewhere in my blogs I point out that I stick with NSIDC data because, for all that there are errors, at least they are consistent year - on - year.  This means that graphs from different years and months can be fairly compared with each other.

    I should have made my point more clear: ice which 'belongs' in the central pack is sitting in Baffin Bay and being used as 'evidence' of recovery.  But not by NSIDC, I hasten to add.

    Here is the latest NSIDC graph, which I quite forgot to include in my article.

    In passing, I want to add a point about shorefast ice.  When ice heaps up in solid masses against the coast it protects the coast from wave erosion during summer.  In many areas the shorefast ice no longer lasts the summer.  Many Inuit communities are at risk from coastal erosion and permafrost melting.

    Just saw I've made a typo: I think that extent is good measure for mid term trend prediction, but since it is more prone to differences in ice configuration... should be finished with: and cannot be used for asserting ice volume in short periods. In longer time, however, there is good correlation between extent and volume as different ice configurations average each other.
    Thanx, Patrick, for accepting my point. Can't wait to see your next blog about ice situation, it seems you know this quite well and I find some of your comments very valuable. BTW, I would like to ask you a question: It seems from GFS and ECMWF wheather prediction map that area of really hot air (20 deg C at 850 hpa level, which in mid lattitude areas with low altitude results usualy in 30 deg C + temperatures) has entered area of Laptev sea and it will be moved to East Siberian sea. How important is this for melting (and is it more change in currents and wind patterns that resulted in spotting several cracks in that area) ? I understand that long term above the average temp would on average result in decreasing amount of ice, but how these effects result in short term periods ?

    Another excellent post; basically a primer for guys like myself that find the subject fascinating, but have never seen a comprehensive explanation of the specific dynamics involved. Much appreciated.

    Sad to realize that the ice in the Baffin Bay this year is showing up as an increase in sea ice extent, but is actually representative of increased melting since it's further south than the main basin.

    I can imagine, that a 1000 years ago, when ice shelves were present, the Baffin Bay could become clear during warm summers. But now, it's going to be receiving more ice from the Nares Strait.

    I guess this also goes to show why there are not really any good computer models for arctic sea ice. The rate of melting is too dependent on blockages opening or closing and facilitating the transport more so than actual amount of warming from CO2.

    BTW, the following caption needs to be corrected to June: "Arctic mosaic images for July 26 2009 and 2010"

    Andrew: many thanks for spotting the error.  It is now corrected, with due credit.  :-)

    I am still working with a few hundred images trying to see patterns of ice loss over the years.  What I have observed so far is this:

    It appears that Baffin Bay, Lancaster Sound and Nares Strait contained warmer waters in June 2009 than they do currently.  Those waters were ice free.  It would seem that as ice flowed through Nares Strait continually from about September onwards, the melting ice lowered sea temperatures in the area generally.

    The continued flow of sea ice through the Nares Strait seems to be delaying a general rise of sea temperatures in the general area of Northern Baffin Bay.  Once the Kane Basin ice is gone I expect local sea temperatures to rise fairly rapidly.  This will be reflected in reports of increased open sea area.  In other words - less sea ice extent.

    The ice bridge in Viscount Melville Sound looks to be on its last legs.  I expect it to be completely fragmented by July 8th.  Neven has posted an animation which he is updating daily.

    Once that ice bridge is gone I expect to see open water appearing rapidly throughout the whole archipelago.

    I'm not quite ready to post a full update yet.  This is by way of a forecast.

    In Arctic Ice June 2010, on June 2nd I wrote:

    This next graph shows ice extent as of June 01 2010 with my annotations in red.  My projection for June ice loss is based on the current state of the entire Arctic ice and on past regional loss rates.  I consider my projection to be, if anything, over-optimistic.

    That projection was indeed on the optimistic side, as was shown by later NSIDC graphs.

    Here is my latest projection.  Now that the summer heat has finished melting most of the snow cover on the sea ice and coasts the Arctic taken as a whole will warm more rapidly.  This warming will accelerate the melting of ice, especially in areas such as Baffin Bay and the Kara Sea which have only remnants of highly fragmented mobile ice.

    By the end of the 3rd week of July, the NSIDC graph will look something like this:

    Adapted from NSIDC image dated July 03 2010
    Projected July 3-week ice loss in red.