Polar Ice

[songster]

Could the real extent be 12.5m ??

No.

An error of 10% means that a pixel "scoring" 15% could in principle be as low as 5% or as high as 25%. Similarly, one scoring 20%, and thus counted as "ice" for the purposes of extent calculation, could be as low as 10% - which should be not included as "ice" due to the 15% threshold. However, anything above 25% (i.e. could be between 15% and 35%) counts as ice no matter what the error is, while anything below 5% (range 0 to 15%) counts as open water irrespective of the error margin.

In practice, only a very small margin of ice around the edge of the pack falls into the "grey zone" which might be affected by error. Look at the Bremen University map at http://www.iup.uni-b...MSRE_visual.png and see how few pixels there are with a middle-blue range shading. (blue pixels cover 0-25%). For the extent calculation, these pixels, and only these pixels, have ~10% chance of being wrongly counted. So, while the per-pixel error margin is 10%, the error margin on the overall extent is much, much lower.

Edited May 27, 2010 by songster

[Gray-Wolf]

The sooner we shift over to 'volume' measurements the better!!!

That said the current 'volume' measure looks pretty grim even to one who understands/accepts what is occurring up there!!!

Is that line really vertical????

[songster]

That graph's a fortnight old. It hasn't been updated since the last time you posted it. Or the time before that, I suspect. Not sure why it's not being kept up to date though.

[Gray-Wolf]

I thought it was updated every 3 days.....I'll check it's not my P.C.!!!

EDIT: No you are correct, the 'continuously updated' plot seems stalled at the 13th (unlucky for some!) unlike the ice loss. I take it the line should be longer but just as vertical?

Edited May 27, 2010 by Gray-Wolf

[BornFromTheVoid]

Seems on the IJIS site that they've amended a few of the daily figures, so the rate of decline over the last few days has eased somewhat, and we're now 85,000km2 off 2004 but still 2nd from bottom in the series...

[Gray-Wolf]

I note from Modis that a large opening in the ice on the Alakan coast (in Beaufort sea) is spreading as the ice exits via the Bering straights. I'd imagine that we'll cross the lowest line on JAXA even with their 'amendments' over the weekend and then the Beaufort melt/Hudson Bay melt/Siberian shelf sea melt to take over reductions for early June as the Bering 'ice tounge' is now nearly gone.

It will be interesting to see whether ,once we are at the lowest level of the JAXA series, the ice loss runs parellel to the others through the faster 'summer melt' period (like 07' did) or whether it slows and re-joins the 'pack' (as it were).

I get the idea we'll see another ,lower, 07' like track.

[songster]

I thought it was updated every 3 days.....I'll check it's not my P.C.!!!

EDIT: No you are correct, the 'continuously updated' plot seems stalled at the 13th (unlucky for some!) unlike the ice loss. I take it the line should be longer but just as vertical?

Apparently so. Latest plot has the anomaly increasing by ~0.5km^3 in 11 days.

[Gray-Wolf]

With the anom already so great it has me quite concerned about the rest of the melt season (with the bulk of ice yet to melt). The current slowdown is a welcome respite from that early season collapse of the 'late spurt' ice.

[BornFromTheVoid]

Good to see the melt rate slowing down a little, 59,281.3km2/day for the last 10 days which is still well above average but not quite as exceptionally so as a few days ago.

A few weeks ago I copied the daily data from the IJIS site into excel and seperated the date and extent to analyse it and make graphs and whatnot. But it seems now that the daily extent figures for around this time in 2006 have been changed as they are higher than the figures I originally downloaded. Does anyone know if they have altered the formula they use for calculating ice extent or if perhaps this changing of data is commonplace with regards Arctic ice? I can understand them changing the figures from the day or a few days before because of problems with cloud cover and whatnot, but not for years before.

I haven't checked any of the other years but here are a few examples from 2006, all the figures from 21-5-06 to 11-6-06 have been increased on the current "Data Download" section.

Current Data Download___ What I Downloaded

05,24,2006,11339375___ 24/05/2006-11324531

05,25,2006,11297813___ 25/05/2006-11270781

05,26,2006,11250469___ 26/05/2006-11216094

05,27,2006,11222500___ 27/05/2006-11179531

05,28,2006,11199844___ 28/05/2006-11145156

[songster]

Good to see the melt rate slowing down a little, 59,281.3km2/day for the last 10 days which is still well above average but not quite as exceptionally so as a few days ago.

A few weeks ago I copied the daily data from the IJIS site into excel and seperated the date and extent to analyse it and make graphs and whatnot. But it seems now that the daily extent figures for around this time in 2006 have been changed as they are higher than the figures I originally downloaded. Does anyone know if they have altered the formula they use for calculating ice extent or if perhaps this changing of data is commonplace with regards Arctic ice? I can understand them changing the figures from the day or a few days before because of problems with cloud cover and whatnot, but not for years before.

I haven't checked any of the other years but here are a few examples from 2006, all the figures from 21-5-06 to 11-6-06 have been increased on the current "Data Download" section.

From the IJIS page, immediately below the graph...

<I>[update info.] The sea ice extent data on around June 1 and October 15 were updated in order to eliminate the apparent erroneous blips.　See “the method for calculating sea ice extent†below for the details.</I>

Basically, June 1 is when they switch from using the 'dry ice' algorithm to the 'wet ice' algorithm, to account for melt ponding. A couple of weeks ago they tweaked things to smooth the transition.

[BornFromTheVoid]

From the IJIS page, immediately below the graph...

<I>[update info.] The sea ice extent data on around June 1 and October 15 were updated in order to eliminate the apparent erroneous blips.　See “the method for calculating sea ice extent†below for the details.</I>

Basically, June 1 is when they switch from using the 'dry ice' algorithm to the 'wet ice' algorithm, to account for melt ponding. A couple of weeks ago they tweaked things to smooth the transition.

Thanks, don't know how I missed that

[BornFromTheVoid]

We're now bottom of the series, 43,000km2 below 2006.

Here are the average daily melt rates for May;

2010 - 65,972.77km2/day

2009 - 55,902.29km2/day

2008 - 47,137.10km2/day

2007 - 44,359.87km2/day

2006 - 42,963.71km2/day

2005 - 46,008.07km2/day

2004 - 37,147.16km2/day

2003 - 41,053.42km2/day

03-10 Average - 47,568.05km2/day

03-09 Average - 44,938.80km2/day

Hopefully the trend for anomalously high melt rate since the end of April doesn't continue, or a new record low extent will beckon.

Just note I haven't updated the figures yet to include the smoothing over between May 20th and June 11th, so all years other than 2010 should have a slightly lower melt rate, just by 1 or 2k, so the figures above still give a good representation of how things stand, if but a little flattering for 2010!

[Gray-Wolf]

Th 07' record melt didn't really set off (and depart from the rest of the plots) for another month yet. If the ice proves to be as thin as we suspect and we are as bereft of ice over 3m as it appears then we have the whole basin to go at over the next 3 and a bit months (unlike 07' which still had 20 odd percent of the old type perennial to work on).

I very much suspect that we will drop well below the lowest (06') prior to the onset of melt season proper and then the melt will continue on while the other lines ease up (late Aug) for another couple of weeks at least as the 'old perennial' that used to mark this 'slowdown' is no longer there so the new ice that has replaced it WILL continue to melt out (unlike the big chunks that would have just melted back a bit).

EDIT: OMG, it's worse than we feared!!!

Edited June 1, 2010 by Gray-Wolf

[Admiral_Bobski]

EDIT: OMG, it's worse than we feared!!!

:lol:

Brace yerself!!

[stewfox]

Brace yerself!!

Early July is when the ride starts

[J10]

I have updated my spreadsheet to reflect the new formula used by IJIS to iron out the apparent blips around June and October, This will appear here over the weekend.

[pottyprof]

Thank you J1

[J10]

Since my last update in late march, quite a lot has happened, the Ice peaked at quite a high level at 14,407,344

on the 31st March, which itself was very late in the year. Through April, the ice figures went above the long term average at times. However, Since then ice extent fell back sharply and since Mid May the ice extent has been below recent IJIS average figures. In the past week or so, ice drop has fallen back to around average (when compared to recent averages) but we are still over 300k below short term averages and over 850k below long term averages. We are also bottom of the IJIS extent and long term extent.

[cooling climate]

Since my last update in late march, quite a lot has happened, the Ice peaked at quite a high level at 14,407,344

on the 31st March, which itself was very late in the year. Through April, the ice figures went above the long term average at times. However, Since then ice extent fell back sharply and since Mid May the ice extent has been below recent IJIS average figures. In the past week or so, ice drop has fallen back to around average (when compared to recent averages) but we are still over 300k below short term averages and over 850k below long term averages. We are also bottom of the IJIS extent and long term extent.

Nice balanced report. The ice that has melted is still only peripheral ice and means very little at the moment.

I see that the ice in Hudson's bay is holding up well despite the mild winter although as posters have said many

times the difference between say -25c and -30c means very little when talking about ice formation.

[BornFromTheVoid]

Nice balanced report. The ice that has melted is still only peripheral ice and means very little at the moment.

I see that the ice in Hudson's bay is holding up well despite the mild winter although as posters have said many

times the difference between say -25c and -30c means very little when talking about ice formation.

Actually, Hudson ain't doing great tbh. It's 200,000km2 below average and almost the same below last year.

Thanks J1. May I ask where you get the long term average figures from? Been creating graphs similar to your yourself lately, think it would be handy to have.

[J10]

Actually, Hudson ain't doing great tbh. It's 200,000km2 below average and almost the same below last year.

Thanks J1. May I ask where you get the long term average figures from? Been creating graphs similar to your yourself lately, think it would be handy to have.

Data from: http://polynya.gsfc.nasa.gov/seaice_datasets.html, (doesn't appear to a direct link)

I have then attempt to merge this data with the figures from IJIS. http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm to create an going series back from 1979

[BornFromTheVoid]

Data from: http://polynya.gsfc.nasa.gov/seaice_datasets.html, (doesn't appear to a direct link)

I have then attempt to merge this data with the figures from IJIS. http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm to create an going series back from 1979

Found it there. Thanks very, I look forward to adding it in

[J10]

Found it there. Thanks very, I look forward to adding it in :lol:

I have attached the list in Excel format, which you may or not find useful.

[pottyprof]

Thank you J1

[Gray-Wolf]

http://www.sciencedaily.com/releases/2010/06/100602193423.htm

More on the Polyak paper and, for those with cash to burn.....

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VBC-4YKFMY0-2&_user=10&_coverDate=03%2F12%2F2010&_rdoc=1&_fmt=high&_orig=search&_sort=d&_docanchor=&view=c&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=74058b1ad94658ced1aaf5ea8c598906