I wrote this nine days ago :
stevengoddard says:
June 28, 2010 at 10:16 pm
In three days, the slope of the Arctic extent graph will begin to drop off.
Mark it on your calendar.
Anyone who doubts the correctness of the PIPS based methodology I have been using, should take a look for themselves.
http://ocean.dmi.dk/arctic/icecover.uk.php
Close up below :
The modified NSIDC map below shows ice loss since the first of the month in red. Where’s Waldo?
The decline in the slope of the area graph is even more dramatic. Area is a better indicator of melt, because it is not affected as much by wind shifting the ice around.
http://www.ijis.iarc.uaf.edu/seaice/extent/AMSRE_Sea_Ice_Area.png
The graph below of average ice thickness shows that 2010 is starting to follow a path more like 2006 than 2007. In 2006 the ice thinned vertically, but did not compress much horizontally. In 2007 the winds compressed the ice horizontally, which led to lots of incorrect news coverage (to this day) that the Arctic was having a meltdown. The average ice thickness actually increased through the summer of 2007.
The video below shows how the winds have changed since earlier in the summer. The ice is no longer being compressed inwards, it is now being stretched outwards by cold northerly winds.
The NCEP forecast for the next two weeks is shown below. Temperatures generally running well below normal in the Arctic. Note – I color shifted the images and scale to make them easier to differentiate. This does not change their data in any way, it just makes it easier to understand.
Temperatures in the high Arctic have been below normal this summer, and in a few days will start their decline towards winter. In 50 days, they will drop below freezing, and remain there.
http://ocean.dmi.dk/arctic/meant80n.uk.php
Meltwater ponds at the North Pole are starting to freeze over.
http://psc.apl.washington.edu/northpole/webphotos/noaa1.jpg
Julienne from NSIDC wrote this yesterday :
I am not surprised that the rate of decline has slowed a bit during the last week. In the Beaufort and Chukchi seas, the ice has essentially melted back to the lobe of old ice that was transported there over the winter (under the negative AO phase). Thus, this is slowing the ice loss in this region. In the Kara Sea, temperatures have been colder than normal and the winds are causing ice divergence (also helping to slow loss of ice extent in that region).
This is the same point I raised questioning PIOMAS back in April.
The computer model is predicting that 3+ year old ice (which is probably in excess of 10 feet thick) is going to melt by early August. That seems rather far fetched. Below is an overlay of the NSIDC map and the U of W simulation for August 18. Note all the multi-year ice that needs to melt.
Like myself, Julienne has forecast 5.5 million km² for the summer minimum.
http://www.arcus.org/search/seaiceoutlook/2010/june
Not everyone at NSIDC agrees. Mark Serreze has been quoted as predicting a possible record low, like he does almost every year.
Mark Serreze of the center forecast the ice decline this year would even break 2007’s record.
Conclusion : There is nothing going on in the Arctic which hints of a repeat of 2007, other than the highly suspect PIOMAS graph.
http://psc.apl.washington.edu/ArcticSeaiceVolume/images/BPIOMASIceVolumeAnomalyCurrent.png
Melt ponds are freezing over at the other North Pole webcam as well.
http://psc.apl.washington.edu/northpole/webphotos/noaa2.jpg
Jarmo
You are confusing area with extent.
Hi stevengoddard,
you keep saying that 2010 will be close to 2006, but at the same time “predict” a september extent of 5.5 km2. I suppose you’re not mentioning the minimum, here, are you? And to what database shall we compare this number? If we look at the JAXA extent chart, 5.5 looks much more like 2009 or 2008 than 2006.
BTW, how do you react to the NSIDC claim that PIPS is a very bad indicator of sea ice thickness, and that it should not be used for year-to-year comparisons nor predictions?
Another good post Steven, thank you.
kadaka
Reading Romm’s site, it is easy to believe that Octopus are more intelligent than humans.
Steven,
If I consider the ice area maximum, it has been shifted approximately one month later. Is it possible that the same phenomena will be seen for ice area minimum?
I think David W is on to the impact of winds. One reason early loss of ice might lead to a larger minimum extent is as follows:
-The wind compresses the ice during the winter, leading to increased concentration and thickness of the core ice pack.
-The open water at the edges refreezes but this ice is of marginal thickness and concentration.
-The ice at the edges melts more quickly in the spring leading to a higher melt rate.
-The melt rate slows significantly when the thicker more concentrated core becomes the edge of the ice pack.
Seems to make sense and, of course, summer weather can have its own influence that may increase or decrease the importance of this scenario.
Rob Vermeulen says:
July 8, 2010 at 5:43 am
Hi stevengoddard,
you keep saying that 2010 will be close to 2006, but at the same time “predict” a september extent of 5.5 km2. I suppose you’re not mentioning the minimum, here, are you? And to what database shall we compare this number? If we look at the JAXA extent chart, 5.5 looks much more like 2009 or 2008 than 2006.
++++
I think that discussion was much more about the slope of the melt in the key July-August period, and is muchly an observation of the (predicted) power of more multi-year ice (more than 2008-2009) to have that slope be more flat in that period.
I suppose re extent you can make an arguement re 2009 vs 2006. If Steve’s number hits, so far as I can see the result would be almost exactly equidistant between 2006 and 2009.
But 2008, as you suggest? No, not at all like it (if Steve turns out right, of course). Neither as to slope in July/August, nor as to extent. 2008 bottomed at 4.7M, so it would be 800K lower than Steve’s 5.5M, while 2006 would be but 300K higher. And the JAXA plots clearly show serious sag in the 2008 July/August slope as a result of the loss of multi-year ice in 2007. 2009 shows much less of that sag, but it is still there. Look at JAXA and compare 2006/2009 at ~Aug 5th vs 2006/2009 at ~Sep 10th. That’s what the relative difference in multi-year ice can do during approach to minimum.
2010, should the wind Gods cooperate, should show even less sag than 2009 in that period, as there is more multi-year ice avaialable.
@Rob Vermeulen says:
July 8, 2010 at 5:43 am
Hmm, I wrote a longer post and it disappeared. Perhaps it will return later.
Take a look at JAXA for 2006 vs 2009, at August 5th and September 10th. What do you see? What caused it? What are its implications for 2010?
[Reply: for some reason it was in the spam folder. It’s posted now. ~db stealey, mod.]
Jack Simmons: re Lewis Pugh
Here’s his website:
——————————————————————————————————
Showman, indeed, with big brass ones, which I’d bet still haven’t dropped back since that swim in May.
Good catch, however other charts are less spectacular:
http://www.ijis.iarc.uaf.edu/seaice/extent/AMSRE_Sea_Ice_Extent_L.png
From Shakespeare:
Caesar: “Soothsayer, the Ides of March have come!”
Soothsayer: “The Ides of March have come, Caesar, but they have not yet gone.”
2010 has come, but not yet gone. It may still have some surprises for all of us.
Casper
The minimum is controlled directly and indirectly by elevation of the sun, and there isn’t much opportunity for it to shift earlier or later.
geo
The shape and behavior is tracking 2006, but 2010 has less ice than 2006. i.e. it is offset downwards.
Thanks for another good update, Steven. I wonder how much the reduction in solar high energy UV will affect the summer minimum?
It’s been fun following this slow motion horse race for the past couple of months and I appreciate the comments that help us to understand the various issues involved.
I’ve got a two questions and an observation. The first question: Why does the AMSRE Sea Ice Extent graph start in 2002? (I’m assuming a new satellite began taking pix, but am not sure.) The second question: How certain is it that there is not a discontinuity between 1979-2001 data and 2002-2010 AMSRE data? That is, are those two data sets reliably comparable?
The observation: Each year since 2002 the rapid melt season typically ends with an inflection point where the melt rate drops to a lower level that is generally sustained without reversion to the earlier higher melt rate. That inflection point occurred in 2006 starting about July 10th. This year, assuming the rather sudden drop off of the melt rate turns out to be the inflection point (I know, big if, but it’s gone on for 8 days already) then it will have occurred about 10 days earlier than 2006. This leads me to wonder if the minimum extent might exceed the 2006 figure and possibly even the 2004 figure and come in in the low 6mm’s. Any thoughts on whether this is at all likely?
Rod Everson
When the Hudson Bay and other areas outside of the Arctic Basin melt out, the slope drops off. This year they melted earlier than in previous years. I don’t think it provides much, if any, useful information about the summer minimum.
Rob Vermeulen
I don’t care if the PIPS thickness numbers are absolutely correct. I am only using PIPS for relative comparisons, so as long as their methodology is fixed from year to year PIPS should do the job just fine. So far, it seems to be spot on.
stevengoddard said on July 8, 2010 at 6:34 am:
For your own good, you really shouldn’t do that. I don’t care how much “excess” the doctors will say you have right now, you may well need those brain cells later on.
🙂
An observation that I’m surprised no one else seems to have made (I could, of course, have missed it through not thoroughly reading all the comments thus far) – in the AMSR-E Sea Ice Area Chart Steven reproduced, there was a flattening in the 2007 curve in late June before it began declining precipitously. In fact, there even appears to have been a brief uptick. I don’t pretend to know how fast the melt is going to go through the rest of July and June but it clearly wouldn’t be unprecedented for the 2010 decline, having slowed down this last week, to speed up again.
So.. Big thick plate of unmeltable ice in the artic, generally cold expansion wind-front and decreased global temps at high latitudes.
Rejoicing about vindication?
Djon
So are you forecasting a rapid drop in ice area?
As much is being made of this (what I think is a short-term) turn in Arctic Sea ice, I think a few charts of interest here. First, one chart really worth watching for the next few weeks is the Arctic Basin Extent:
http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/recent365.anom.region.1.html
Why? Because if forms the largest single area of remaining ice for the summer melt season. Now Steve would have you believe that this is not going to melt much, etc. in the coming weeks, but we see it already is in a negative anomaly condition, but more than that, let’s take a look at some (for better or worse) thickness model data, such as it is, beginning first with TOPAZ:
http://topaz.nersc.no/topazVisual/matlab_static_image.php?action=NA_ARC_NWA_Function&file_prefix=ARC&match_date=20100708&depth=0005&variable_name=hice
Now, this shows exactly what Julienne was talking about a few weeks back. The bulk of the thicker ice is not in the high latitudes of the central Arctic Basin, but is further south, bunched up near the E. Siberian Sea. Notice also the very little is near the N. Greenland coast. This is not dissimilar to what (god forgive me) PIP 2.0 is showing, ableit far less accurately or in detail:
http://www7320.nrlssc.navy.mil/pips2/archive/retrievepic.html?filetype=Thickness&year=2010&month=7&day=8
In the case of the TOPAZ image (which really doesn’t vary too far from what PIOMAS was forecasting for this period) we see that the Atlantic side of the Arctic Basin is very thin, and certainly we’ll be seeing a lot of melting here in the coming days and weeks. Look for this region in particular to keep the Arctic Basin in a negative anomaly condition for the rest of the melt season. Also, if you study the PIOMAS forecast, you see that the melt sort of splits the whole Arctic in two by the end of the melt season, leaving the Artic with two lobes, the thicker region of ice over near the E. Siberian Sea, and the other lobe N. of Greenland into the central arctic, but sort of melting from both east and west toward the center:
http://psc.apl.washington.edu/zhang/IDAO/seasonal_outlook.html
I certainly appreciate Steve’s skill in calling this short-term “turn”, but like the much bally-hooed March “bump” up, or his mentioning of the melt-ponds icing over (short term weather) I don’t see this turn lasting. I think the 2010 trend will make a final turn back down, paralleling 2010 down to the last days of August at least.
http://psc.apl.washington.edu/zhang/IDAO/seasonal_outlook.html
Steven,
No, I’m not forecasting a rapid drop in ice area. At the risk of being alleged to have an abrasive communications style, which part of “I don’t pretend to know how fast the melt is going to go through the rest of July and June…” (“June” being a case of insufficient editing on my part – I meant ‘August”) wasn’t clear enough to tell you I wasn’t making a prediction, only observing that people might be reading too much into a very short term trend?
R. Gates
The Arctic Basin area (not extent) map you linked is probably just reflecting problems with the sensors differentiating between surface meltwater and seawater.
Djon
At the risk of repeating everything I have written in the last few months, I will point out that my analysis has very little to do with the extent/area graphs.