Predict Summer 2011 Minimum Arctic Sea Ice Extent (millions sq km)

[Clifford]

Based on the JAXA AMSR-E dataset.

What do you anticipate the minimum 2011 Arctic sea ice extent to be?

Vote EARLY !!!

http://www.ijis.iarc...aice_extent.htm

http://www.ijis.iarc...extent/plot.csv

[skierinvermont]

Well I'm entering a prediction of 6.25 million sq km on behalf of JB

I'll make a forecast later today. Last year I predicted 5.2.

[Clifford]

As an optimist...

I decided that the Arctic Ice is slowly thickening. And thus 2011 will just edge out 2009.

so that put me at the 2002 level, between 5.5 & 5.75 (noting, of course, that JAXA is missing some of the 2002 data).

We've discussed PIPS2 which seems to indicate thickening ice. Previous objections to the accuracy of the map are noted. However, due to lack of additional information, this will have to stand in for my projections. There haven't been any recent PIOMAS predictions.

Conditions in the Bering Sea and Bering Strait seem to be turning neutral. However, it is noted that PIPS2 seems to be indicating low amounts of ice just north of the Bering Strait.

[skierinvermont]

As an optimist...

I decided that the Arctic Ice is slowly thickening. And thus 2011 will just edge out 2009.

so that put me at the 2002 level, between 5.5 & 5.75 (noting, of course, that JAXA is missing some of the 2002 data).

We've discussed PIPS2 which seems to indicate thickening ice. Previous objections to the accuracy of the map are noted. However, due to lack of additional information, this will have to stand in for my projections. There haven't been any recent PIOMAS predictions.

Conditions in the Bering Sea and Bering Strait seem to be turning neutral. However, it is noted that PIPS2 seems to be indicating low amounts of ice just north of the Bering Strait.

There is PIOMAS up to december... we are so far below the volume of 2002-2008 that there is no chance we have "recovered" in the last two months because it takes years for ice to thicken. I am fairly confident that there is less volume than there was in 2007 and probably 2008 based on the loss of multi-year ice and PIOMAS. On this basis I will go with 5.0 million sq km. It could be anywhere from 4.3-5.6 based on the weather this summer.

[BethesdaWX]

Mine is between 5.51-5.75.

[skierinvermont]

Speaking of sea ice, I'd like to point out that Bethesda's claim that winter sea ice has been recovering since the PDO went negative is going to look pretty silly in a month. Unless we see a big bump this will be the worst or 2nd worst winter for sea ice in the satellite era. Of course now that the PDO has gone negative and sea ice has continued to decline, skeptics are changing their minds and deciding it's the AMO. Of course, by the time the AMO does go negative in 10 or 20 years, we will probably have lost another million+ sq km so this whole argument will be moot.

December

January

February....

[BethesdaWX]

PIOMAS?!?!?

You've shown just how desperate you are Andrew.

How about actual OBS

NORSEX

NSIDC

Really, everything has the increase in ice overall since 2007. Forget the Minimums, overall ice extent is up, period. There is no refuting this, even with PIOMAS.

[Clifford]

The Cryosat-2 has been whizzing around Earth for nearly a year now with promises to give accurate estimates of sea ice volume.

http://www.esa.int/esaLP/LPcryosat.html

There seem to be lots of positive notes about the satellite, except that we don't seem to have any comprehensive reports from it.

The Icesat-1 was supposed to have the capabilities to estimate sea ice (and I think it was the ones that PIOMAS had used for calibration. Its lasers went out on February 2010, and the satellite has now been decommissioned.

http://icesat.gsfc.nasa.gov/icesat/

I imagine there are efforts to calibrate the instruments on Cryosat-2... as it can be problematic to revise values at a later time, but it leaves us without good data.

[Clifford]

January

February....

Great charts...

Except that they only provide single data points.

The JAXA chart at the top of this topic gives the same info, but one can look at progression of the ice.

With the JAXA chart, it is obvious that 2011 has had more ice than 2006 for the entire month or February, and has followed 2005 very closely.

It had met the 2010 curve earlier in the month, but is now slightly below the 2010 levels. However, by observing 2010, it is apparent that the sea ice extent for 2010 peaked slightly after the other years, but perhaps this late-forming ice also was quicker to melt.

2009 also increased the ice extent throughout March, and then at the beginning of the melt season, it ended up with the highest extent of the decade.

Judging by recent temperature anomalies,

2011 will likely follow an increasing trend for summer ice like 2008, followed by 2009, rather than a decreasing trend in 2010.

[John S2]

My prediction is between 2007 extent and 2008 extent. The ice is in a more fragile state than in 2007, but against that there were particularly favourable synoptics in 2007 which caused 2007 to be a major outlier in the general downward trend. If synoptics are favourable I think we would go below 2007.

[skierinvermont]

Great charts...

Except that they only provide single data points.

The JAXA chart at the top of this topic gives the same info, but one can look at progression of the ice.

They are averages for the entire month. It is certain at this point that the combined DJF will be the lowest or perhaps 2nd lowest extent on record. It's nice to look at the progression, but it doesn't make the least bit of difference to my point.

February probably will average a little higher than 2006, and about the same as 2005, perhaps lower. But DJF will probably the lowest on record, because December was so terrible.

[skierinvermont]

PIOMAS?!?!?

You've shown just how desperate you are Andrew.

How about actual OBS

NORSEX

NSIDC

Really, everything has the increase in ice overall since 2007. Forget the Minimums, overall ice extent is up, period. There is no refuting this, even with PIOMAS.

Ice extent charts have nothing to do with volume. PIOMAS has been validated against IceSat observations, and submarine draft observations. It has performed fairly successfully in the past. I see no reason why it would suddenly stop being successful.

Posting dozens of smileys doesn't make your argument any more convincing. In fact, I don't see any evidence contained in your post that PIOMAS is wrong whatsoever.

And as I have reminded you, I am not on a first name basis with people that call me "Mr. Homo!!!" and "***sucker!!!" I find you obnoxious, rude, childish, offensive and irrational.

[BethesdaWX]

Bingo,

Our volume measurements not only have HUGE bars of error, but the way we measure will not be adequate until the New Cryosat data is ready. You constantly harangue UAH/RSS for minimal bars of error, yet you support PIOMAS/GISS/Hansen?!?!

Now do you understand why I'm in total awe?

This is why I don't post volume data, I'mm all for extent, concentration, area...you know, things we can measure accurately?

When CRYOSAT comes out with the data, I'll be using that for Volume.

[skierinvermont]

Bingo,

Our volume measurements not only have HUGE bars of error, but the way we measure will not be adequate until the New Cryosat data is ready. You constantly harangue UAH/RSS for minimal bars of error, yet you support PIOMAS/GISS/Hansen?!?!

Now do you understand why I'm in total awe?

This is why I don't post volume data, I'mm all for extent, concentration, area...you know, things we can measure accurately?

When CRYOSAT comes out with the data, I'll be using that for Volume.

Those are not error bars on the graphs.

[BethesdaWX]

Those are not error bars on the graphs.

Never said they were...but the window of error on the volume measurements we use are Huge. You Rip UAH constantly for minor errors, and Praise PIOMAS/GISS, that says alot.

[Clifford]

This is why I don't post volume data, I'mm all for extent, concentration, area...you know, things we can measure accurately?

The University of Illinois that does some of the sea ice concentration studies has changed their scale to emphasize the fine differences in concentration between 80 and 100%.

There seems to be very little stability of these readings in the middle of the Arctic in the winter, and it seems to be strongly influenced by surface storms, and perhaps also atmospheric conditions.

Lower concentrations, especially around the edges, are more representative of the actual ice conditions.

As far as PIOMAS,

Errors of +/- 2 meters for 1 meter ice... just seems significant.

They only report a single volume value, although the submarine studies were more localized, and I believe showed significant localized thickness errors which were all averaged out.

[skierinvermont]

As far as PIOMAS,

Errors of +/- 2 meters for 1 meter ice... just seems significant.

They only report a single volume value, although the submarine studies were more localized, and I believe showed significant localized thickness errors which were all averaged out.

I posted the charts... even localized error was usually quite small. I do not believe the error bars are 2 meters.. please post evidence. It looks like 95% of the time PIOMAS was within .5m of the submarine reading so that is a decent ballpark error bar for localized readings (+/-.5m).

Moreover, getting every single local reading exactly correct is not essential to getting the mean correct. The model had the mean nearly exactly correct according to multiple submarine trips, and according to IceSat.

So again, I ask, if PIOMAS has been accurate in the past, why would it suddenly stop being accurate?

[Clifford]

I posted the charts... even localized error was usually quite small. I do not believe the error bars are 2 meters.. please post evidence. It looks like 95% of the time PIOMAS was within .5m of the submarine reading so that is a decent ballpark error bar for localized readings (+/-.5m).

Moreover, getting every single local reading exactly correct is not essential to getting the mean correct. The model had the mean nearly exactly correct according to multiple submarine trips, and according to IceSat.

So again, I ask, if PIOMAS has been accurate in the past, why would it suddenly stop being accurate?

Looking at the charts you posted.

1987 (winter), PIOMAS was as much as 2-3 meters low, 4-7 meter ice

1996 (summer), PIOMAS was about 1 meter high for about half of the measurements, 1-3 meter ice.

1997 (summer), PIOMAS was as much as 2 meters high, 1 meter ice.

If I could integrate it in my head, I'd say it would average out to be about 25% high for all of the summer 1996 and summer 1997 measurements.

And, perhaps 25% low for the Winter 87 measurements.

Several of the other measurements have significant positive and negative errors.

For example, summer 93, it had errors in the range of +/- 1 meter with 1 to 3 meter ice.

I think I remember seeing a paper that indicated an East/West error. I'll see if I can find that again.

On the chart, it lists a single correlation statistic of 0.76.

I haven't analysed stats enough lately to say if a 0.76 correlation is good or not for a model. I suppose it would depend on the correlation of other similar models.

If the error is 25%. That would mean that if the ice extent is 4 million km², the model could read either 3 or 5 million km², or the equivalent in thousands of km³.

Anyway, on a year like 93, if it averages out to being close to the correct volume (using the data that we have available), is that good enough? Or is the +/- 1 meter error an indication of a systematic error that could cause problems for the model (as in 87, 96, and 97)?

[Clifford]

Here are some other diagrams...

Not posted by the University of Washington who made PIOMAS

http://stratus.ssec....model_jgr10.pdf

http://cimss.ssec.wi...09_400_Wang.pdf

http://cimss.ssec.wi...ter_guc.ppt.ppt

Validation of the model with 96 to 99 submarine and Buoy information. Compared to submarine and One-Dimensional Thermodynamic Ice Model (OTIM), which doesn't seem to be providing additional information.

The PIOMAS readings were almost double the control submarine and buoy readings.

Depending on whether or not PIOMAS is always high, or if it is just sometimes high, the downward trend in sea ice indicated by PIOMAS may not be accurate.

If PIOMAS shows twice as much ice as there actually was in 96-99, and then they matched the model to the IceSat in 2003-2007, then their trendline could be significantly off.

Here is another article that discusses draft to sea ice thickness calculations.

http://www.arctic.noaa.gov/detect/Rothrock2008.pdf

They indicated that the loss of ice seemed to be levelling off around 2000 (not including later data sets). It would indicate that the Wang article oversimplifies their draft/thickness calculations, but it doesn't seem to be enough to account for the PIOMAS discrepancy.

[superjames1992]

I'm figuring something around 2009's minimum. I voted 5.26-5.50, but it could very well be 5.01-5.25, or I could be totally wrong.

[skierinvermont]

Looking at the charts you posted.

1987 (winter), PIOMAS was as much as 2-3 meters low, 4-7 meter ice

1996 (summer), PIOMAS was about 1 meter high for about half of the measurements, 1-3 meter ice.

1997 (summer), PIOMAS was as much as 2 meters high, 1 meter ice.

If I could integrate it in my head, I'd say it would average out to be about 25% high for all of the summer 1996 and summer 1997 measurements.

And, perhaps 25% low for the Winter 87 measurements.

Several of the other measurements have significant positive and negative errors.

For example, summer 93, it had errors in the range of +/- 1 meter with 1 to 3 meter ice.

I think I remember seeing a paper that indicated an East/West error. I'll see if I can find that again.

On the chart, it lists a single correlation statistic of 0.76.

I haven't analysed stats enough lately to say if a 0.76 correlation is good or not for a model. I suppose it would depend on the correlation of other similar models.

If the error is 25%. That would mean that if the ice extent is 4 million km², the model could read either 3 or 5 million km², or the equivalent in thousands of km³.

Anyway, on a year like 93, if it averages out to being close to the correct volume (using the data that we have available), is that good enough? Or is the +/- 1 meter error an indication of a systematic error that could cause problems for the model (as in 87, 96, and 97)?

You are looking at the peaks and troughs. Most of the time even individual locations it is within half a meter.

The mean (which is all that really matters) is quite accurate. So again, if the mean has been accurate in the past, why would it suddenly stop being accurate?

[ORH_wxman]

I'm leaning toward something in the 5.3 mill sq km range, but I haven't looked at the summer pattern yet (or tried to look what the summer pattern might be). The Big -AO certainly kept a lot of ice in the interior of the arctic basin this winter so it should be in better shape...but the PAC side is still vulnerable despite this. It will take more -AO winters to make it less susceptible to summer patterns than prior to 2007.

I'll probably look a lot closer once winter is over up here. Been pretty busy this winter.

[BethesdaWX]

You are looking at the peaks and troughs. Most of the time even individual locations it is within half a meter.

The mean (which is all that really matters) is quite accurate. So again, if the mean has been accurate in the past, why would it suddenly stop being accurate?

Dude, just stop.

How can you claim the PIOMAS mean "accurate" when we do not even know really what it is? What are you looking at to claim the Mean accurate? PIOMAS cannot "veify" itself.

If you are going to support PIOMAS and the huge "potential error", you shouldn't be hammering UAH for a +/- 0.05C/decade potential error, case closed.

[skierinvermont]

Dude, just stop.

How can you claim the PIOMAS mean "accurate" when we do not even know really what it is? What are you looking at to claim the Mean accurate? PIOMAS cannot "veify" itself.

If you are going to support PIOMAS and the huge "potential error", you shouldn't be hammering UAH for a +/- 0.05C/decade potential error, case closed.

submarine data. PIOMAS showed nearly the exact same mean depth as half a dozen+ submarine voyages did.

[BethesdaWX]

submarine data. PIOMAS showed nearly the exact same mean depth as half a dozen+ submarine voyages did.

Submarines cannot get the proper coverage needed to make a good verification. Either way, they're not the same

http://nsidc.org/noaa/scicex/index.html

[skierinvermont]

Submarines cannot get the proper coverage needed to make a good verification. Either way, they're not the same

http://nsidc.org/noa...icex/index.html

There were plenty of voyages to conduct verification. PIOMAS showed the same ice thickness as the submarine voyages did.

[BethesdaWX]

There were plenty of voyages to conduct verification. PIOMAS showed the same ice thickness as the submarine voyages did.

One SCICEX voyages 1 per month will not gauge the proper # because the entire arctic isnot covered by the voyage at the time/abilities of the submarine voyage. Its almost worse than PIOMAS.

They don't show the same thing regardless......................http://nsidc.org/noa...icex/index.html

[Clifford]

You are looking at the peaks and troughs. Most of the time even individual locations it is within half a meter.

The mean (which is all that really matters) is quite accurate. So again, if the mean has been accurate in the past, why would it suddenly stop being accurate?

Did you see my reference to the Wang article?

http://stratus.ssec.wisc.edu/papers/ice_thick_model_jgr10.pdf

According to Wang (2010), PIOMAS didn't match the sub data very well.. For the data analysed, 96-99, PIOMAS was reading high. Of course, Wang et al, wasn't involved with the UW, (independent is good), except that they had their own model that they were plugging (so there is a disincentive to make PIOMAS look good).

As mentioned, I tried to estimate the area under the curves from the data you presented, and it appeared as if PIOMAS was off by about 25% on several of the years. If you have the raw data files, we can verify.

I find it odd that PIOMAS hasn't been updated since December 31, 2010.

Even PIPS shows some difference in Ice volume in the ice volume between December 2010 and now!!!

The Alaska and Bering Strait weather diagrams have looked very hot lately (and thus the Bering strait ice hasn't done well lately).

That hot spot seems to be moving Eastward... which may give the Bering Strait a break before the spring melt.

However, a purple dot just showed up in the Gulf of St. Lawrence. Perhaps that is a sign of things to come.

The Gulf of St. Lawrence has also looked very blue in the NOAA Sea Temperature Maps.

http://www.osdpd.noaa.gov/data/sst/anomaly/2011/anomnight.2.24.2011.gif

At this point, we just have to wait and see what happens.

The 90's and 2000's were hard on the Arctic ice. Is it beyond a "tipping point"? Time will tell.

One article I read suggested that there is now seasonal ice where there once was year-round ice, and where there once was seasonal ice, it often doesn't freeze.

[Mencken_Fan]

My vote: 4.51-4.75

Polar cap is gonna melt like a sumamabich this year.

[iceicebyebye]

< 5 m sq km