During the melting season I'm regularly writing updates on the current sea ice extent (SIE) as reported by IJIS (a joint effort of the International Arctic Research Center and the Japan Aerospace Exploration Agency) and compare it to the sea ice extents in the period 2005-2010. NSIDC has a good explanation of what sea ice extent is in their FAQ. I also look at other things like sea ice area, concentration, volume, temperature and weather forecasts, anything that can be of particular interest. Check out the Arctic sea ice graphs webpage for daily updated graphs, maps and live webcam images.
August 27th 2011
We have an exciting week behind us. And that's putting it mildly.
In the last SIE update I referred several times to the unfulfilled potential the ice pack is showing this melting season, even making it the title of the update. A coupe of days later we saw some of that potential being put to use when a big cyclone ripped through the Beaufort and Chukchi regions, causing an unprecedented (I think) case of flash melting in a huge area of the ice pack. Normally you get a bit of ice melt/compaction everywhere at the edges of the ice pack, but this week we saw a massive amount of ice just disappearing from the radar one day to the next. Stunning stuff.
Of course, it is impossible for so much ice to disappear in just 24 hours, but the cyclone managed to disperse the floes, which probably got washed by water in the process, in such a way that concentration fell below the 15% threshold combined with some other factors (there has been much speculation regarding this matter in the Flash melting post). Some of the ice could show up again in the coming days if the winds push it together.
The cyclone was as short-lived as it was big, which means there still is a lot of untapped potential out there. The satellite images and sea ice concentration maps don't lie and show us how weak the ice is at the edges. We even see the water that has been warmed up to the brim in the Laptev Sea starting to eat its way into the ice pack, towards the North Pole.
As usual, weather patterns in these last 2-4 weeks of the melting season will determine the final outcome. But what makes this year so interesting is that this dominant influence is being superseded by the state of the ice. In other words, the ice determines how much influence weather patterns are allowed to have. This week's cyclone, for instance, would have actually made sea ice extent jump up only ten years ago because of its diverging power. But you can't diverge what you disintegrate.
Barring some really strong weather patterns that keep the ice in place and freeze the water over at the edges, we are in for a final phase of the melting season where everything is possible.
Sea ice extent (SIE)
Here's the current IJIS SIE graph:
IJIS extent is about to break through the 5 million square km barrier, for the fourth time in the satellite record, after 2007, 2008 and 2010. This year has managed to nibble off some more of 2007's lead (with the help of a very late century break due to the 'flash melting'), but a difference of 191K square km is still quite significant at this stage of the melting season. 2007 is having some very slow days at the end of the week, so maybe 2011 can get closer still. In the meantime, it seems to be fending off 2008's massive attack quite well, still leading by almost 300K square km.
The current difference between 2011 and the other years is as follows:
- 2005: -820K (-45,000)
- 2006: -984K (-37,697)
- 2007: +191K (-57,041)
- 2008: -295K (-70,333)
- 2009: -582K (-48,654)
- 2010: -468K (-51,736)
Between brackets is the average daily extent decrease for the month of August. 2011's average daily extent decrease for August is currently -59,633 square kilometers.
If 2011 loses as much sea ice extent as...
- 2005 did after this date it will bottom out at 4.50 million square km
- 2006 did after this date it will bottom out at 4.80 million square km
- 2007 did after this date it will bottom out at 4.45 million square km
- 2008 did after this date it will bottom out at 4.41 million square km
- 2009 did after this date it will bottom out at 4.67 million square km
- 2010 did after this date it will bottom out at 4.35 million square km
Sea ice area (SIA)
As I wrote in yesterday's blog post on sea ice area, 2011 guaranteed itself a second place on all area graphs. It even took over first spot from 2007 in the Cryosphere Today sea ice area data file today and is just 65K shy of the record minimum sea ice area. But as SIA tends to go up and down arbitrarily, and with big jumps too, it's difficult to tell if the record will be broken or not. Nevertheless, I know what I would bet on.
The Cryosphere Today SIA anomaly graph is of course showing the largest anomaly ever recorded for this date:
As we can see on the Regional Graphs page, the Arctic Basin and Canadian Archipelago keep going down some more (and are lowest on the MASIE charts). There still is some slush puppie ice left in the East Siberian Sea that will probably completely melt out, and in the Greenland Sea SIA and SIE have been going down some more.
Cryosphere Today area per IJIS extent (CAPIE)
One of the reasons I say there is still quite a bit of untapped potential in the Arctic ice pack, is the fact that CAPIE is still very low. This means that the ice pack is diverged and under the right circumstances could be compacted a lot, which is what happened in 2007. As we see on the graph CAPIE shot up real fast around this time in 2007:
For 2011 to threaten the 2007 record IJIS minimum SIE, CAPIE will have to go up significantly the coming two weeks.
Sea Level Pressure (SLP)
Let's cut straight away to the animation of SLP images from the DMI Centre for Ocean and Ice of the past week:
We see the big, blue cyclone forming over Alaska, doing some shredding and dispersing over the Beaufort and Chukchi regions and then quickly dissipate. There currently is a distribution of sea level pressure areas that has tentatively kicked off ice transport towards the Atlantic (again), but it takes a couple of days to create a constant rhythm and it's highly unsure if things will remain stable enough for it to remain so.
Because this is what the weather has in store for the Arctic the coming five days, according to the ECMWF weather forecast maps:
I don't see anything there that makes me believe there will be a significant extent and area decrease. At least not due to compaction and transport.
Update conclusion
There still is a lot of weak ice. No extreme weather is needed to push it over the edge, just average weather will do. If we do get something like a strong Arctic Dipole and a clockwise spinning Beaufort Gyre, the potential is such that 2011 could break all records on all graphs and all datasets in the coming two weeks. But as always, a fast onset of freezing temperatures and winds blowing the ice towards the coasts could still nip record tentatives in the bud.
It's not really clear what will happen the coming week. We will have to take it day by day. Another serious cyclone would be fun though.
2-4 weeks left to go...
---
TIPS - Other blog posts and news articles concerning the Arctic and its ice:
NOAA ClimateWatch Magazine: Summer Heat Unravels Arctic’s Icy Blanket
OK, I will kick it off with a comment of where we are today.
The most up to date measurements are the daily Bremen maps, and the associated Bremen ice extent measurement made using the data in the graphs. Unfortunately most people (including me) don't have access to the Bremen numbers, but occasionally a commenter will give us a hint on the Bremen number. The most recent data point is that August 24th was around 4.8 million in ice extent. This is an accurate number for what is measured and shown in the Bremen map for that day. A graph of the Bremen ice extent on the Daily Graphs page of this site show that 2011 ice extent is already below 2007.
The most reliable sea ice extent measurement should be the ice extent released by the National Snow and Ice Data Center (NSIDC) for the United States. According to posters here, NSIDC uses a five day average, to report ice extents. A graph of their ice extent is also on the Daily Graphs page, and shows 2011 just above 2007. From eyeballing the graph, the extent is below 5 million, and appears about 4.8-4.9 million. This is the most likely 5-day average extent reported. (A link to numerical NSIDC data would be appreciated, if available and if anyone has that.)
The JAXA-IJIS curves provide a preliminary and final daily extent number at their site (making this site very popular), However they use different averaging and data processing methods, and their report seems to lag changes in the other extents measurement systems by about a week. For example, on August 24th, they reported 5.085 million sq km.
A lot of us are interested in using regional sea ice extents and maps to forecast changes in SIE, and so we use the MASIE site. This site has regional maps, and reports numerical ice extents daily for each region. Unfortunately, the site seems to use ten day old maps to make the regional extent calculations. MASIE also provides a rough daily total ice extent. MASIE apparently uses the same data that is used to produce the NSIDC graph of sea ice extent (NSIDC is one of the two main organizations in the MASIE effort). The most "recent" MASIE report stamped August 26 shows 5.1 million sq km, and shows each region's ice extent that adds up to this number. Unfortunately, the regional charts of ice extent are ten days old, i.e. from August 16. The ice extents calculated, which add to 5.1 million, are from data collected on August 16.
So the site with the SIE closest (5.1 million) to the JAXA report of 5.01 million, uses ten day old data. Clearly the JAXA report is seriously lagging the other SIE measurement systems.
The most accurate assessment of ice extent today, is the Bremen data displayed in the Bremen map, and that ice extent hit 4.8 million several days ago.
How low can we go?
Just follow the Bremen graph, and first the NSIDC will follow, with the JAXA and MASIE reports trailing along.
Posted by: Paul Klemencic | August 27, 2011 at 21:59
Another great explanation, Paul. And perfect timing, as we'll see eventual lags more clearly now that the melting season is coming to a close.
---
BTW, there is one thing I forgot to mention in this SIE update, but I'll stress it some in the coming weeks. We're extremely low in extent and area, even though ice transport through Fram and Nares Straits is much lower than in 2007 (I'm pretty certain this is so, though I don't have the numbers).
And now that the Canadian Archipelago has cleared of ice again, as it did for the first time last year, there could have been some bonus transport of MYI from the Arctic Basin towards the warm waters of the NWP. But this didn't come about either.
The LANCE-MODIS satellite images showed a clear view of Peary and Sverdrup Channels for the last three days, but it's clear the winds aren't pushing the ice southwards.
I'm not exaggerating when I say that this year would already have broken every record there is if ice transport towards the Atlantic and Baffin Bay had been similar to that of 2007. That'll be for another year then.
Posted by: Neven | August 27, 2011 at 22:24
The recent comments about week old (or more) data don't make sense to me - why would they label data from August 16th as being from the 26th? Not only that, just eyeballing the graphs suggests that, at the most, they can't be more than a day or two behind the current day (or they would stop near mid-August), and IJIS shows the most recent day with data as the 26th in their data file. It is more likely that the differences lie in the different sensors and algorithms used; NSIDC recently made a remark about Bremen data showing more open water due to higher resolution.
Posted by: Michael Stefan | August 28, 2011 at 00:01
Yeah, the whole "JAXA lags other sites by a week" thing is something Paul's recently asserted, but not provided any evidence for. It's also in direct contradiction to what JAXA say about their own data product.
Paul: What makes you think there is a week lag in the JAXA numbers? Eyeballing the images really doesn't cut it, nor does comparison to sites using completely different base data such as MASIE.
Posted by: Peter Ellis | August 28, 2011 at 00:10
Paul Jaxa doesn't lag behind. That is misrepresentation.
Jaxa uses a 12.5KM grid while Bremen uses a 6.25km grid.
Jaxa has a 2 day averaging system. Bremen does day to day.
There is no delay. They just have a different grid system.
There Data and UB's is exactly the same. The Algorithm makes the difference and until the ice compacts if it goes they will report slightly higher extents.
Posted by: Frivolousz21 | August 28, 2011 at 00:11
Using two independent servers at NSIDC and JAXA has allowed to stabilize the flow of input data (AMSR-E L1A and L1B swath data) considerably.
UB gets there data through Jaxa, so there is no lag.
Posted by: Frivolousz21 | August 28, 2011 at 00:14
A 2 day average would lag a day by day report by about a day.
Posted by: Andrew Xnn | August 28, 2011 at 00:15
Masie: 4KM grids of 40%
Jaxa: 12.5km Grids of 15%
UB: 6.25km grids of 15%
Norsex: 25km grids of 15%
Posted by: Frivolousz21 | August 28, 2011 at 00:25
Guys, I wish this wasn't that convoluted, but reality is what is.
Collect the data from the different sites, plot it up, and compare it. You'll independently see what I have observed.
Start by checking the NSIDC data provided on the MASIE site.
Use the file at the MASIE site that has NSIDC data in it.
CAUTION: I am reasonably sure that the data in the spreadsheet you download, is ten days older then the date shown in the first column.
Why is that? How can you prove it yourself? Several ways...
First, you can compare the ice maps shown at the MASIE site with the Bremen maps, and you will see the maps on the MASIE site are ten days old. By comparing each map, I determined the ice extent data was consistent with the mapped extent shown for a map from ten days earlier.
Second, examine the data in the downloaded spreadsheet.
For example, the MASIE data in row 2011238 matches sea ice extent data taken on August 16. Why? As the spreadsheet reads, the SIE was 5.142 million, but today's NSIDC graph shows the five day average already below 5.0 (likely around 4.9). That cannot be; for a declining series of data points, the five day average can't be lower than the latest data point.
Since the MASIE spreadsheet and the NSIDC graph are drawing from the same NSIDC data, clearly the MASIE data is delayed.
Another way to show the MASIE data is delayed ten days: If you take the MASIE spreadsheet data, and calculate a five day moving average, and plot it on a graph, then compare it with the graph released by the NSIDC, it compares very well, EXCEPT it needs a ten day offset. I did this as well.
Once you realize the MASIE data is ten day old, then compare the MASIE/NSIDC data with the reported JAXA extents, and you will find a 6-7 day offset.
Clearly the JAXA is also lagging both the Bremen map and the Bremen graph, and what Bremen data we have.
So both the Bremen data we have access to, and the NSIDC data lead the JAXA reported extent data by a week.
Posted by: Paul Klemencic | August 28, 2011 at 00:43
MASIE (the 4km product) uses a variety of data sources to create the final product.
It is downloaded daily.
UB and JAXA both use AMSR-E. That source puts out three resolution products ( 6.2, 12.5, and 25. ) Note the 12.5 product has some issues underestimating concentrations based on noise issues in the real primary source. The real source is available from the data pool or maybe REVERB or WIST, granuals are about 50MB. I have a bunch of them, they are really cool. Data Pool is the source I like best
MASIE is based on the 4km IMS data product. Which is actually put together by an analyst who looks at all available sources for data, that is he looks at more than AMSR-E.
http://nsidc.org/data/docs/noaa/g02186_masie/index.html#dataproc_nsidc
Posted by: Stevemosher.wordpress.com | August 28, 2011 at 01:45
Got quiet.
Let me summarize what happened to me: There are several sets of extent data. The Bremen graph and Bremen map agree with each other, and agree with the NSIDC graph. The JAXA data seemed to agree with the MASIE data at a higher ice extent level.
But when I tried to examine each region in the MASIE data, I ran into big trouble. I was trying to use the regional extent data to forecast future melt in that region. Right away, I noticed the ice extents shown the regional maps were wrong... they were out of date; ten days out of date. But the extent numbers shown, agreed with the map! And the extents added up to the MASIE extent total (for example today's report of 5.1 million, with the spreadsheet showing 5.14 million). But this was consistent with ten day old maps! How could that be?
Then the MASIE data didn't agree with the NSIDC graph until I adjusted the data by ten days. Then I realized the MASIE extent data is ten days old.
If the MASIE data is ten days old, why is just slightly higher than the JAXA data?
Once I adjusted the MASIE data, the JAXA report sticks out like a sore thumb. JAXA isn't consistent with the Bremen map, the Bremen graph, the NSIDC graph, or the corrected MASIE extent data.
Also, If I couldn't find the ice to support the incorrect MASIE report at 5.14 million, until I corrected it, then where is the ice to support the JAXA number at 5.0 million?
It doesn't exist. We are short at least 200k and perhaps over 250k sq km of ice to support the JAXA number, and that shortage has to be in the Beaufort, Chukchi, or E. Siberian regions.
Its hard to hide that much ice in those regions. You'd be able to see it on the Bremen map, or the Cryosphere Map. JAXA is counting ice that is but a memory; the ice its counting melted out a week ago.
Posted by: Paul Klemencic | August 28, 2011 at 01:51
Paul, it still doesn't make sense that IJIS would label 10 day old data as being from yesterday. Also, the 2007 minimum was 4.25 from IJIS and 4.13 from NSIDC. True, it occurred 8 days later according to IJIS, but it is hard to believe that delayed data being passed off as more recent was the cause. On the other hand, the minimum last year occurred within one day in both datasets (ignoring the 5 vs 2 day averaging), but was 200,000 km3 higher for IJIS, similar to what you observe for the present.
Posted by: Michael Stefan | August 28, 2011 at 02:08
It is very odd that the current MASIE map, labeled 26 August, most closely resembles the CT map for 16 August. And, as Paul points out, the MASIE extent of 5.1 is at least 200k larger than the NSIDC chart sitting at 4.9
I can understand a 200k offset between IJIS and NSIDC - or between MASIE and NSIDC - but the maps shouldn't be offset by 10 days. The differences should be more random. Something more than just a difference in algorithms appears to be taking place.
Posted by: Kevin O'Neill | August 28, 2011 at 02:22
Michael, I'm not saying the IJIS is labeling 10 day old data. The MASIE report is based on ten day old data, and this is the only extent data that seems to agree with JAXA report. Right now the MASIE report is 5.14 million while the JAXA report is 5.01 million.
But the MASIE report is really the extent back on August 16th! The current MASIE/ NSIDC extent must be approaching 4.8 million to get a five day average near 4.9 million. Remove the MASIE report, and the JAXA report at 5.0 million looks very suspicious.
If the JAXA reported number has a significant lag as I believe, that becomes less important as we get to the third week of September. The melt rate slows down so much, that at the minimum, there might only be a small difference. But a couple of weeks ago, the difference was quite large, possibly over 350k, and even today, the difference with the Bremen map and graph is likely more than 200k.
So no, at the minimum I don't think its much different. But if you were using a method like the one I was using, looking at ice conditions and extent in each region, and trying to estimate end of melt season regional extents, then adjusting the current ice extent reading; well, then an error in the current ice extent number can throw the forecast way off.
On the Bremen map, I just can't get the regional ice extents to add up to over 5.0 million, with the biggest problems in the E. Siberian, Chukchi, and Beaufort regions. The current Bremen ice extent is likely midway between 4.7 and 4.8 million, and that appears to be an accurate range for the current ice extent.
Posted by: Paul Klemencic | August 28, 2011 at 02:41
Here are the comments from the previous thread pertaining to this issue:
In this comment I raise the issue of problems with MASIE.
In this comment, I point to an event, a separation in the ice pack over a week earlier, that demonstrated the MASIE maps and data were over a week old.
And in this comment, I propose this seemingly idiotic idea.
And scanning through comments on down the thread, anyone can pick up the gist of the problem, and how it became more defined.
Posted by: Paul Klemencic | August 28, 2011 at 03:02
There is no lag. Bremens graph uses a different Y axis. They are identical most of the year except when the Ice is slushy with water holes all over. Jaxa uses a 2 day running average to keepit smoother.
Posted by: Frivolousz21 | August 28, 2011 at 04:10
SIE under 5 million, but only down by 29k. INMHO, that rules out the possibility of any lag. More likely to be a sensor issue!
Posted by: Phil263 | August 28, 2011 at 05:15
The beafuort had a lot of convergence today. Let's see how UB dies tonight I bet it's a flat day or very small decline. Almost certain. Java has a 2-Day running average. So a -29K prelim from a-46k final Suggests a very tiny melt day which the UB map also backs.
Posted by: Frivolousz21 | August 28, 2011 at 05:27
Damn, Phil scooped me. But the actual preliminary IJIS SIE number for 8/27 is:
(drumroll)
4,980,938 km2
Posted by: Kevin McKinney | August 28, 2011 at 05:33
Stacking up the JAXA and UB time series, I don't see a consistent lag pattern. The cross-correlation function peaks at zero lag, as it should. The most obvious difference is not a time offset, but the higher peak extent (on the order of 700k this year) that UB sees in wintertime.
UB and JAXA converge in spring, until their positions reverse -- around July 1 this year. Since then UB has generally been lower, currently around 300k.
That's just a numerical view, I leave interpretation to smarter folks.
Posted by: L. Hamilton | August 28, 2011 at 05:35
Off Thread, I understand, but I think this important issue is not being addressed.
forwarding...
I'd like to bump this thread a bit because I think the discussion earlier today is important. Scientific understanding is not well represented in the blogosphere and it is valuable, even essential, to have more of the scientifically literate venture into the climate denialosphere.
I agree in particular with the comment from Timothy Chase above. I have been posting comments about the Arctic Sea Ice on a denialist dominated discussion forum for several years now and have learned a few things.
If you should choose to go here, first thing (as Timothy mentioned), do not focus simply on your antagonist, but realize, when venturing into this lion's den, you will encounter an the unending stream of objections. But remember, any number of people click into such discussions to learn (on the forum that I frequent, the ratio of views to posts is 5 or 10 to 1).
Consider; civility can score you points with the audience you really need to reach.
Also, don't presume that logic is king. Denialists have a twisted logic all their own. Evidence and knowledge, however, do accumulate over time. Persistent participation adds up.
It would definitely be useful to increase the knowledge content of any number of internet forums.
I would conclude; this is not a task for the thin skinned, overly emotional, or easily discouraged, but getting informed content out on the net could have an important impact.
Tom Z
Posted by: Tzupancic1 | August 28, 2011 at 05:43
Given grid resolution Bremen likely dies a Better job with a slush puppy ice pack. More grids will come up ice free compared to Jaxa. Like in the 135E to 105E area. UB will do much better there. Same with East Siberian Sea and that hole between the two arms of ice.
Posted by: Frivolousz21 | August 28, 2011 at 05:44
I don't think the IJIS-JAXA data are lagged. Paul seems to have wandered down a rabbit hole and hopefully will be able to extricate himself without too much trouble.
If one compares the absolute values of sea ice extent numbers from different data providers (IJIS, NSIDC, UB, NORSEX), there will be systematic differences among them, due to both differences in the input imagery (as mentioned above) and perhaps differences in how they model sub-pixel ice area for pixels with sea ice concentration << 100%. (I can think of a couple of other potential sources of difference among the data sets, but don't want to speculate about things I haven't had a chance to investigate adequately).
One consequence of this is that, if 2011 ends up very close to 2007, it may not be possible to definitively say which year had a lower minimum ... because 2007 might be lower in, e.g., the IJIS daily minimum, while 2011 might be lower in the NSIDC monthly minimum.
The important thing is to be consistent. I have no doubt that WUWT will focus like a laser beam on whichever metric (area, extent) and whichever provider (IJIS, UB, ...) is least in conflict with their party line that "the ice is recovering." I hope people here can avoid the mirror-image trap of deciding that whichever data set shows the most dramatic loss must be correct, and then coming up with Rube Goldberg explanations for why the other data sets are unreliable.
Posted by: Ned Ward | August 28, 2011 at 06:43
Neven writes: I'm not exaggerating when I say that this year would already have broken every record there is if ice transport towards the Atlantic and Baffin Bay had been similar to that of 2007. That'll be for another year then.
That's OK, this year will have its place in the record books regardless. And in terms of understanding the real long-term trend, we're actually much, much better off not having 2007-style outliers.
Look at how the anomalously high global mean temperature associated with the 1998 El Nino set us up for a decade of trolling from the pseudo-skeptics ("temperatures haven't risen since 1998, doncha know?") For sea ice, the perfect-storm combination of melt, compaction, and advection in 2007 created excessively high expectations for 2008, 2009, and 2010.
It's exciting to smash records, of course. But that tends to lead to a weird psychological whiplash, in which people first panic because the world seems to be ending, then become jaded and complacent when the next year "merely" returns to the long-term downward trend.
Posted by: Ned Ward | August 28, 2011 at 07:01
Ther is a Ton of 30-50% ice on the Bremen map. Quite incredible. We can match the fast decline of the outer ice with Healys webcam.
This is a new ring around the ice pack. The bottom melt has to be going to town. Many of these areas have been cloudy with -4 to -8 or lower 85os the last week and still quickly melting out.
00z gfs is out and has gone back to a late season
Ice melt machine.
1. The temps are warm. The models again throw old the medium
Range cold. Now throw 3-7C into the Kara/laptev around 20-30kt winds for 3 days before modifying. This torches the entire sea. Expect Ssts to respond. And that warm water will be thrushes at the ice already melting out from warm says there already.
2. Beaufort temps torch. Models continue to bring in 3 to upwards of 8c 850s along the shore. The flow is directly into the arctic basin from the bearing sea. The water is already warm there. Will get a boost.
3. Compaction in the Beaufort, western arctic basin, east siberian sea, Kara, Greenland sea and Canadian basin will commence.
Instead of a dipole anomaly. The arctic gets wind flows and temps distributed as good as can be in the arctic late August to keep heat on the ice. The only cold is over areas that are already solid or undercutted by warm surface winds. And Ssts.
Posted by: Frivolousz21 | August 28, 2011 at 07:04
It's exciting to smash records, of course. But that tends to lead to a weird psychological whiplash, in which people first panic because the world seems to be ending, then become jaded and complacent when the next year "merely" returns to the long-term downward trend.
I couldn't agree more Ned, and we are paying the political price for this phenomenom right now. After the high level of AGW awareness in 2007, we now have conservative politicians and think tanks who want us to believe that Business as Usual is a sensible strategy!
Posted by: Phil263 | August 28, 2011 at 08:53
UB looks like it went up slightly. Jaxa and UB both got the same data. Jaxas 2 day running avg. Caused it to drop some while UB dropped none. Same reason yesterday UB dropped more than jaxa. Then there is whatever grid res accounts for. As the ice compacts jaxa will drop faster then ub
Posted by: Frivolousz21 | August 28, 2011 at 09:14
I have updated the Slush Puppie animation.
Posted by: Neven | August 28, 2011 at 11:39
Looks like a few days of slow melt ahead. The AO is moving towards positive territory, so we should start to see some storms that will do some signficant damage to the ice, but I don't see anything significant yet over the arctic ocean.
The DMI north of 80 temps are beginning to drop significantly, and I expect a signicant freeze over of the melt ponds and I'm expecting a jump from Cryosphere Today, as a result.
However after the melt ponds are frozen over, the Cryosphere Today results should be more accurate, with a slow drop to the new minimun area.
NSIDC SIE is showing very little difference between 2007 and 2011, perhaps 50-100K.
Posted by: Lord Soth | August 28, 2011 at 16:44
CT area swung up today by 98K (2007 retook the lead), and so did CAPIE as a consequence.
I also think things are going to be slow. The weather's perfect for that.
Posted by: Neven | August 28, 2011 at 16:46
A Decent SLP is forming over the Western Arctic, moving into the Arctic basin by tonight.
this is going to create a decent inflow from the warm waters of the bearing and the warn waters of the North Atlantic.
http://vortex.plymouth.edu/gifs/110828151811.gif
http://vortex.plymouth.edu/gifs/110828151916.gif
http://vortex.plymouth.edu/gifs/110828151958.gif
http://www.wetterzentrale.de/pics/Rhavn062.gif
If the Beaufort Ice didn't get converged from a heavy E to SE wind the last 4-5 days we would be down to 4.8km2 or so by now. But that will be quickly stopped. That area is still torching for this time of year.
http://ocean.dmi.dk/arctic/satellite/plots/satsst.arc.d-00.png
http://ocean.dmi.dk/arctic/satellite/plots/satanom.arc.d-00.png
The ice will see some bigger drops the next 3-4 days.
Posted by: Chris Biscan | August 28, 2011 at 17:30
http://lance-modis.eosdis.nasa.gov/imagery/subsets/?subset=Arctic_r02c03.2011240.aqua
The Fram Ice is about to drop below 15% as warm water and soon warm air with sun gets pumped into this area for the next 3-5 days. This will aid in the extent dropping again.
Posted by: Chris Biscan | August 28, 2011 at 17:35
Chris, it'd be cool if you were right, because it would confirm what I've said in this SIE update (ie, the state of the ice supersedes dominance of weather conditions).
But I have to admit I'm torn between what I learnt last year and what seems to be new this year. I still think that we need some really compacting weather to last 3 weeks for 2011 to get close to the 2007 IJIS record minimum SIE.
Posted by: Neven | August 28, 2011 at 17:42
Whoever is lagging whom, an excellent observation to note that the JAXA and Bremen Y-axis are identical... good way to zoom in and put the old ruler on top, MASIE took the opportunity to catch up with a -56227,51 KmSq change for 26th > 27th. The Standings:
MASIE 5086117
JAXA 4990156
Difference 95961
Good chunks gone from CA, Chukchi, ES, Kara. Small increases on Arctic CB, GS and Barents to name a few.
CT gaining 98K... well it's ping-pong time.
Posted by: Seke Rob | August 28, 2011 at 17:47
in this case I would think both will be contributing factors.
the weather isn't bad. The Extent has been saved by thin ice converging in the face of warm winds.
The winds have been blustery along the NW Territories into the Beaufort causing masssive Convergence of thin ice. The winds are about to go calm on the West side and do a 180 on the East side as an HP gets stronger and slides south towards Alaska. This brings a 15K wind pushing into the ice pack up there. with warm water and air.should see compacting kick in fast. That ice is weak..
Posted by: Chris Biscan | August 28, 2011 at 17:47
http://vortex.plymouth.edu/gifs/110828155456.gif
12z GFS coming in stronger with SLP.
the models have trended stronger with this piece for 2 days..4 days ago it didn't exist...
thank you Hurricane Emily.
As you can see that map is perfect for compaction this time of year.
http://vortex.plymouth.edu/gifs/110828155633.gif
By tonight the flow is established and ripping. The Fram has a constant inflow that only gets stronger as the week goes on.
The flow down the right side is compacting the ESB, Kara, and Laptev, with the Barrents expanding but not much because the flow diverges and splits back to Greenland. While the other side flows out to warm waters.
http://vortex.plymouth.edu/gifs/110828155834.gif
Warm air:
http://vortex.plymouth.edu/gifs/110828160006.gif
The Kara and Beaufort have 10-12C 850 positive anomalies right now.
this is reflected well on the SST charts.
Winds are 15-20Kts inflowing warm air/SSTS into the Western arctic and compacting the ice.
Posted by: Chris Biscan | August 28, 2011 at 18:03
http://vortex.plymouth.edu/gifs/110828160408.gif
by day 2.5 inflow is still rolling over the Western Arctic.
The Fram still has a strong flow up around Greenland..
these two will compact ice over the Canadian Basin.
by day 3: the inflow starts to weaken.
the inflow over the fram gets stronger and warmer.
http://vortex.plymouth.edu/gifs/110828160541.gif
http://vortex.plymouth.edu/gifs/110828160726.gif
Warm air is relentless coming out of Alaska this will keep SSTs warm and pushing into the Western Arctic.
While Warm Air/Water is pushed into the Greenland Sea.
the central arctic will see melt ponds freeze over but no ice will form over open seas as to much warm water is around and to much wind flows.
Can't see the rest of the run for a bit..but it is already setting up for a powerful SLP to push into the Western Arctic by day 5-7 which would mean massive compaction.
Posted by: Chris Biscan | August 28, 2011 at 18:09
UB Prelim out showing the beggnings of what I described above.
Jaxa might go above 5,000,000km2 for a day though.
Posted by: Chris Biscan | August 28, 2011 at 18:23
Today marks another 'spring tide'
With the northern portion of the ice shelf in front of Independence Fjord already decimated during the last high tide, and the southern tip hanging by a thread, I'm wondering what might happen today or tomorrow
It appears as though a strong southerly flow through Fram could tear the whole thing from the land
I couldn't find evidence that any of this shelf had been disturbed in 2010, and its destruction could be the biggest event of the year
Posted by: Twemoran | August 28, 2011 at 19:17
Humboldt and Petermann glaciers:
Have you all noticed the melt lakes or fjords that are made in between Petermann and Humboldt glaciers, I wonder what that will end up with eventually, something similar what happened to the Ward Hunt Ice shelf with the Disraeli Fjord? Also by my estimates Ward Hunt is smaller in area than even 2010.
Regards Espen
Posted by: Espen | August 28, 2011 at 20:16
Not less than 3 Oceanographic research boats are indeed very close to the North Pole these days.
http://www.sailwx.info/shiptrack/researchships.phtml
Perhaps Polarstern and Healy (currently the most famous vessel in the arctic on this blog, due to the webcam!) are planning to meet in a couple days, a honourable sign that would mark.
Besides, the web cam stills from USCGC Healy, tell of almost absent presence of MYI, will "Game Over" for The North Pole Summer Ice be next year!?
Finally, before entering the bed: This year is a thrill, and the contribution on this blog feel good and nourishing, at least for the soul! So better hang on guys, 'till last round...or 'till White Winter returns in the north...
http://www.youtube.com/watch?v=DrQRS40OKNE
Posted by: Christoffer Ladstein | August 29, 2011 at 00:26
Polarstern recorded a 30 nm area of -0.6 C seawater between 88-89N, near the 140W meridian. SSTs near the pole were as high as -1.0 C
http://www.sailwx.info/shiptrack/shipposition.phtml?call=DBLK
These temps are well above the freezing point of seawater, which is near -1.8 C depending on salinity.
Posted by: Artful Dodger | August 29, 2011 at 02:11
Is there any sea ice measurements coming out of the Healy.
Posted by: Chris Biscan | August 29, 2011 at 02:37
Hi Chris. You just need to know Healy's callsign to use shipwx.info ship tracker:
Ships at sea: Callsign:
Polarstern DBLK
USCGC Healy NEPP
CCGS Amundsen CGDT
CCGS St-Laurent CGBN
CCGS Laurier CGJK
CCGS Terry Fox CGTF
Posted by: Artful Dodger | August 29, 2011 at 05:16
No 'bounce' on IJIS; it's roughly 4.960M on the prelim, which IIRC is closed enough to 30K decline.
Posted by: Kevin McKinney | August 29, 2011 at 05:41
Ack. "close enough."
Posted by: Kevin McKinney | August 29, 2011 at 05:42
Yeah the Beaufort didn't eat as much ice as the prelim had.
since that connection is gonna stop....extent drops will pick up...possibly 50-70K the next 2-3 days.
Posted by: Chris Biscan | August 29, 2011 at 05:57
Chris said : Is there any sea ice measurements coming out of the Healy
As I mentioned in a side-thread, the Healy seems to be steaming ahead at at least 3kts (some 5 km/hour). That is its cruising speed as long as ice is no thicker than 4 ft (some 120cm). If there were any MYI around, it would have to back up and ram, which will slow it down considerable.
Also note that according to ARCc (PIPS follow-up model) the Healy is currently in an ice field 4 meter thick, while it has been breaking through 3 meter ice over the past week or so. Remember this if anyone ever wants to bring up ARCc (or PIPS) as a valid ice-thickness model.
Posted by: Rob Dekker | August 29, 2011 at 10:57
North East Greenland / Joekelbugt:
Heavy cracks in the remaining Shore Ice, I believe about 60% of if is on the loose now according to the prelim images from Modis today.
http://rapidfire.sci.gsfc.nasa.gov/cgi-bin/imagery/single.cgi?image=crefl2_143.A2011241054000-2011241054500.250m.jpg
Regards Espen
Posted by: Espen | August 29, 2011 at 11:26
MASIE shot past JAXA to 4,931,552 dropping -154564 square kilometers.
Loosers in main
Beaufort -56,981
Chikchi -67,726
East Sib. -32,490
Arctic Cntrl - -8,633
Winners in main
Greenland Sea 7,721
Canadian Arch. 4,123
Not much else on that front. Looking forward to press level worthy reports by the 3 research vessels way up north... their ice thickness findings will greatly help the fact checking and 'tuning' of what that new prophetic ESA satellite.
Posted by: Seke Rob | August 29, 2011 at 16:56
Using the football analogy, between 2007 and 2011.
It's second down, after the quarterback got sacked on the first down with a loss of yardage (difference between 2007 and 2011)
We definitely need to gain some yardage, or else the upcoming 2007 5 day stall will still leave us short of a touchdown.
With regard to the three research vessels in the high arctic, the Louis St. Laurant and Healy are on a joint Canada USA science mission, and are doing tag team icebreaking - multibeam bathymetric surveys.
Posted by: Lord Soth | August 29, 2011 at 17:12
FrankD and Neven:
I am not seeing the steep drop in the Central Arctic Basin if 2011 is compared to 2007.
For the 2010 v. 2011 comparison, I do not remember seeing consistently big differences between the beginning point on the Cryosphere Today Arctic Basin chart and the end point on the graph. Admittedly, I have not overlaid the 2010 chart over 2011 to see what differences have occurred over the last 365 days, but I would be shocked if 2011 was consistently 7.5% below 2010 since in the deep freeze months they are identical.
FrankD:
re: your comment:
Indeed. Cryosphere Today's graph for the Arctic Basin is ~190,000 km^2 down on 2010, a 7.5% delta. The data does not support the claim that there is "no significant difference" between the two years.
The data are not showing anything approaching a 7.5% annual decline rate for the Arctic Basin when multiple data points are plotted. Even if it did, this rate of decline is not sufficient to create an ice free Arctic Basin in less than 10 years.
You are comparing 2011 to 2010 for a single day to say there is a trend of a steep decline. That is not a valid methodology.
The Cryosphere today chart for the Arctic Basin showed a sharp uptick after the August 27th measurement date you are using. Neven's post above indicates it was 98K, so about half of the difference you cited disappeared in a single day.
http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/recent365.anom.region.1.html
The multi-year ice area chart for September for the Central Arctic Basin provided by Tivy does not show anything approaching an annual trend of a 7.5% drop even if you add in 2010 and YTD 2011 (September will not be significantly below the late August YTD measurement as the Central Arctic Basin ice starts to expand before the September minimum for the Arctic as a whole is reached is reached).
The Central Arctic Basin as measured in the Tivy graph was 2.1 million km2 in 2007. Are you predicting that the 2011 September data point will be below this?
http://www.arcus.org/files/page/images/639/figure8.png
The MASIE ice extent chart for August 27th has 2011 at 3.03 million km2 and 2007 at 3.06 million km2 and 2010 at 3.12 million km2. No 7.5% drop there.
I would not be surprised if the data for August 28th and 29th for 2011 showed an extent higher than 3.03 million km2 on the MASIE chart due to the uptick on the Cryosphere today chart.
What the MASIE chart does show is there is some variability between the years for any given day. For example, August of 2008 was generally higher than 2007 and 2009 was higher than 2008. Ice extent in 2010 was lower than 2008 and 2009 but higher than 2007. The MASIE chart shows that on some days the years change their ranking position. For the first 9 days of August 2011 was generally above the same day level as 2010, so what does that show?
ftp://sidads.colorado.edu/DATASETS/NOAA/G02186/plots/r11_Central_Arctic_ts.png
The "big drop" from August 9th to August 20th for 2011 was approximately a 150,000 km2 drop and the ice extent remains above 3.0 million km2 per MASIE chart. Since the "big drop", ice extent in the Central Arctic Basin has remained close to the 3.03 million km2 mark through August 27th. If the uptick on the Cryosphere Today chart is real, the MASIE chart should show a similar uptick after August 27th.
Posted by: William Crump | August 29, 2011 at 18:16
William Crump, the data for the 28th is out, which was posted in this thread some 1.5 hours ago: http://neven1.typepad.com/blog/2011/08/sie-2011-update-18-ten-yard-line.html?cid=6a0133f03a1e37970b015434eef3a0970c#comment-6a0133f03a1e37970b015434eef3a0970c
MASIE fell 154Kfrom the 27th to the 28th in summary.
Posted by: Seke Rob | August 29, 2011 at 18:28
William, I know you discussed this before over at Patrick Lockerby's blog, but can't remember the details.
Maybe this is a stupid question, but wouldn't it be logical that the Arctic Basin would be the last region to show a marked decline? If the Arctic was moving to an ice-free state of course.
Obviously this stage hasn't been attained yet, but the system seems to be on the threshold. I guess that when surrounding regions start melting out earlier, so will the Arctic Basin region.
Posted by: Neven | August 29, 2011 at 18:40
I know extend, area and volume are important factors when it comes to tell the situation in the Arctic Sea, but I find it equally important to see what is happening to the shore ice in the area, and to my eyes the situation is really bad and it tells me a lot more than the above mentioned factors we have less than 200 km of fast / shore ice left in the whole Arctic Sea!!!!
Regards Espen
Posted by: Espen | August 29, 2011 at 19:04
Neven, you make a good point about the 2-dimensional coverage of the Central Arctic Basin. But William ignores volume, for which we have several datapoints. Since 2001, Arctic Sea Ice thickness in the Central Arctic has declined from about 2 meters thickness to about 0.9 meters thickness. And that over a period of just 10 years.
To focus on area and extent for the Central Arctic is to focus on its weather, while the thickness (as a surrogate metric for volume) is all about the climate.
After all, the volume's the thing...
Posted by: Daniel Bailey | August 29, 2011 at 19:19
And with the loss of volume, we see loss of shore ice!
Regards espen
Posted by: Espen | August 29, 2011 at 20:02
William,
CT covers Sea Ice Area and concentration.
Masie covers Sea Ice Extent.
Unless there is a Masie Area page I am missing.
The lack of meteorology in this thread is rather disturbing. I have a hard time believing many can grasp the arctic ice and what is going on without at least giving the meteorology of this some modest attention.
I spent 2006-now learning meteorology which helped me understand the arctic and the arctic process quickly and pretty well.
I am currently turning my focus to Oceanography and all things SSTs and sea salt water related to melt.
William again,
you should do some research on Ships and buoy data in regards to ice thickness and Temperature profiles under the ice. You may not understand this yet. But the warm waters do not just vanish when they run into the ice or colder waters..they stream south under the ice and under the layer of fresh water at the surface(right under the ice)The warm water filled with salt moves faster, as in the particles move faster which infiltrate and by default warm the colder fresh water from below.
While 2010 stopped surface melt in mid September and you saw rapid ice freeze on the top. Then snow quickly cover it, which gives a false positive on thickness color. Which makes the ice to appear much thicker on modis.
For instance. This Winter areas of the ESB have 30-80CM of melt to go still, like as of now.
this thin ice will form.
Lets say a 200km2 block forms..in 3 days t goes 10CM thick. during that 3 days most of it will lose 3-6CM from the bottom. But the models that do not account for bottom melt say it's 10 CM.
well over the course of the Winter from that moment that area reaches 2-3 Meters on Pips. But there was 30-80CM loss at the bottom so now it's anywhere from 1.2 meters to 2.7. With an error towards 1.4-2.4 meters.
10-30 years ago this didn't happen near as much or strong.
Right now the BEAUFORT IS TORCHING. Warm SSTs have exploded again along the NW Canadian and Alaskan shore. An SLP has come into play over the Western Arctic Basin.
This SLP will turn the flow with the help of an HP over the beaufort Sea W-E.
http://vortex.plymouth.edu/gifs/110829183307.gif
You can see there is the SLP and HP..the flow is directly around the area's of lowest and highest pressure.
http://vortex.plymouth.edu/gifs/110829183411.gif
850mb temps(5000 feet) are warm in Alaska bleeding over into the Arctic.
These temps are downslooping off the Mtns in Alaska.
This helps mixing to the surface. There is also enough sun around to facilitate windier conditions and mixing down.
http://vortex.plymouth.edu/gifs/110829183700.gif
over 1-2 millionkm2 of open ocean/ice wind speeds are 12-30 mph between now and 72 hours all moving relatively W-E.
This is transporting that warm air and Warm Waters and slamming it into the cold pool of air over the central arctic.
http://vortex.plymouth.edu/gifs/110829183838.gif
you can see there the wind direction. If you notice the Fram/Greenland Sea is taking the same beating. The Last few days the kara/Laptev/barrents took the same beating. And water temps in the Kara shot up.
http://vortex.plymouth.edu/gifs/110829184108.gif
2M temps in Alaska are decieving because they are relative to the mtns. The air that comes off this over the Coast with 5-10C SSTs and 6-10C 850s is warm..this gets modified but creates melt/ocean water warming.
those SSTs get slammed into the ice with those winds or under the surface water....this speeds up water particles and aids in melt.
Overall this means Volume has taken a beating the last few years.
only super thin ice has saved extent and near ideal winds all summer.
Posted by: Chris Biscan | August 29, 2011 at 20:44
The MASIE numbers released and dated today, match the extent losses on the flash melt day of August 22nd.
Of the 154k sq km reported lost; 157k came from the Beaufort (57k), Chukchi (68k), and E. Siberian (32k) seas. There is only one day this month that showed such massive losses in these seas at roughly these levels; and that day is August 22nd.
I predicted a MASIE reported loss of over 120k before August 30 (first attempt), or by September 1st (second attempt). The prediction proved out.
We now have a very strong case to show the MASIE (NSIDC) numbers show the regional extents with a one week delay.
On my spreadsheet, the average 5 day rolling average jumped to 88k today for MASIE, whereas JAXA-IJIS 5 day rolling average stands at 31.5k. Clearly these two data sets are reporting for different time periods.
The IJIS reported daily numbers are also getting processed (not simply a two day average) apparently to avoid reporting big swings, including gains in extent mixed in the losses this time of year. I am finding the IJIS data harder to unravel, but have some good clues that I can discuss in the next comment.
Bottom line: MASIE/NSIDC showed a massive 157k decline in the slush puppy seas on August 22nd, and the sea ice extent on that day was 4.93 million sq km.
Posted by: Paul Klemencic | August 29, 2011 at 20:45
UB prelim map shows decrease in the Beaufort.
The Extent drops will pick up steam now.
Without any divergence we may see 3+ days around 75,000 per day with those inflows on both sides.
stay tuned.
Posted by: Chris Biscan | August 29, 2011 at 20:46
Clarification: the 5 day rolling average loss in extent for MASIE are 81k versus only 31.5k for JAXA-IJIS.
Posted by: Paul Klemencic | August 29, 2011 at 20:48
The Euro sets up a reverse DPA.
This might lift the entire Ice pack off Greenland enough for some unreal images of the arctic come 10-15 days if that happens.
Posted by: Chris Biscan | August 29, 2011 at 20:48
Yes, I noticed this too. It'll be interesting to see what happens if this comes about. Should make for an early end of the melting season though (in theory).
Thanks for the meteorological updates, Chris. I know a bit how to interpret those GFS and ECMWF maps, but not to the extent that you do.
That should take this year past the 2010 minimum.
Posted by: Neven | August 29, 2011 at 21:26
It seems some of that buoy data has been updated: 80cm in the Beaufort.
Posted by: Neven | August 29, 2011 at 21:28
90 cm at the Russian Arctic Station NP-38.
Posted by: Neven | August 29, 2011 at 21:36
thanks neven.
The bottom melt is crazy.
Posted by: Chris Biscan | August 29, 2011 at 21:37
That won't end the melt season.
http://lance-modis.eosdis.nasa.gov/imagery/subsets/Arctic_r03c03/2011241/Arctic_r03c03.2011241.terra.1km.jpg
http://lance-modis.eosdis.nasa.gov/imagery/subsets/?subset=Arctic_r03c03.2011239.terra.1km.jpg
This isn't our Fathers Arctic Sea Ice.
hell this isn't my Teen years(I am 28) Arctic sea ice.
The thing to take from that graph is that thick sea ice can't move like that from mediocre winds in 36 hours.
Posted by: Chris Biscan | August 29, 2011 at 21:49
The ice off NE Greenland is falling apart rapidly.
http://lance-modis.eosdis.nasa.gov/imagery/subsets/?subset=Arctic_r03c03.2011241.aqua
Just brutal.
Posted by: Chris Biscan | August 29, 2011 at 22:05
If I recall correctly, earlier this month, Neven wrote that JAXA-IJIS never reported a daily gain in extent in August.
Doesn't that bother anyone?
August starts with 6.5 to 7.0 million sq km, and so even just a 2-3% error would be 130-210k sq km. Compare this to an average two day ice loss of 120k. And many times we observe very little ice extent loss. In fact, during re-freeze periods, we would expect a gain in extent.
Then why isn't the IJIS report ever showing a gain, when supposedly the only adjustment they make is two day averaging?
Just how accurate are these sea ice sensors, that day to day they couldn't make just a 2-3% measurement error? And consider periods where ice extent could gain 80-100k in two days (MASIE daily data show a two day gain of 91k earlier this August).
Then how does a two day rolling average not ever report a gain in August? It boggles the mind.
Even more interesting: IJIS and Bremen supposedly get their data from the same source, then why the big difference in graphed extents? And the MASIE data doesn't agree with IJIS either, now that we've shown the MASIE reported data shows data from a week ago.
Clearly the IJIS is using some kind of adjustment or longer term averaging to avoid reporting extreme gains/losses of the type we see in the daily NSIDC data, and not a simple two day average. This results in the reported ice extent number (and daily changes in ice exten), that don't match the Bremen map.
Earlier this month, we saw a huge shift in the pack toward Alaska, and expected a big IJIS loss, but the reported numbers didn't seem to match the obsevations from the map. Now looking the MASIE/NSIDC data, we see a 320k loss date stamped August 12, followed by a 145k loss data stamped August 13. Those losses match the Bremen map changes we saw about a week earlier.
Those big losses are quite differently distributed than the flash melt day loss on August 22; very little ice extent was lost in the Beaufort or Chukchi regions, which were benefiting from the shift of ice in that direction. The ice extent losses were greatest in the Laptev, Kara, Canadian Archipelago, and Hudson's Bay. Again, this matches the Bremen map observations quite well.
The IJIS reports, not so much. Although they reported some big losses around this period, they fell far short of the MASIE/NSIDC numbers, and don't match up well. Since the really slow melt days at the end of July, the IJIS report has responded much more slowly to observed ice pack loss.
Posted by: Paul Klemencic | August 29, 2011 at 22:10
@ Paul K
Please check you FB email.
Posted by: Daniel Bailey | August 29, 2011 at 22:14
If you guys line up the UB maps just right you can see the compaction in all of the places I described well under way.
If that area in the Laptev around 85N opens up the next 3 days we could have a century break.
Posted by: Chris Biscan | August 29, 2011 at 22:24
Thanks Neven for the US Army Ice Mass Balance (IMB) buoy update from the Beaufort. Note that this buoy (2010E) was placed in a slap of MYI, so the 80cm should represent sort of the upper limit of ice thickness in the Beaufort.
In regards to bottom-melt please also note the (continuously updated) FluxBuoy nearby :
http://www.oc.nps.edu/~stanton/fluxbuoy/deploy/buoy21_deltaT.html
This FluxBuoy shows bottom heat flux of 50-100 W/m^2 over the past month, which translates to 1-2 cm/day of bottom melt, which is very much consistent with the 2010E IMB buoy (1.8 cm/day over the past month).
I love it when independent data sources confirm findings, even when they report alarming melting rates.
Either way, the FluxBuoy also shows that water in the Beaufort under the ice is still very warm (-0.5C) and only going down very slowly. So bottom melt will continue for a long time in that area.
Also read this page from the US Army IMB buoy scientists. Very interesting read of ice thickness profile over two years :
http://imb.crrel.usace.army.mil/2006Csum.htm
Posted by: Rob Dekker | August 29, 2011 at 22:30
Paul the Y axis is different on the graphs.
For most of the year the numbers are nearly identical.
http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm
http://www.iup.uni-bremen.de:8084/amsr/ice_ext_n.png
Start with 2003 Paul. the only time they diverge is when the ice has water inside the extent's reach.
like this year.
most of the time other then smoothed data the are nearly identical.
How is that so hard for you to see?
This is from there website:
Method for calculating sea-ice extent
* The sea-ice extent is calculated as the areal sum of sea ice covering the ocean where sea-ice concentration (SIC) exceeds 15%. SIC data of JAXA’s AMSR-E standard products are used for this purpose (http://sharaku.eorc.jaxa.jp/AMSR/products/pdf/alg_des.pdf). The algorithm for calculating SIC was developed and provided by Dr. Comiso of NASA GSFC through a cooperative relationship between NASA and JAXA.
* The black dot seen at the North Pole is an area lacking data where AMSR-E cannot observe the Earth’s surface due to the limit of its observational coverage (i.e., orbit inclination of 98deg. and swath width of 1600km). Please note that this area is also counted as sea-ice cover in our estimation of sea-ice extent. We may change the policy (i.e., filling the gap with full coverage of sea ice) in the near future due to the recent drastic reduction of Arctic sea ice. We will announce this if it is implemented.
In principle, SIC data could have errors of 10% at most, particularly for the area of thin sea ice seen around the edge of sea-ice cover and melted sea ice seen in summer. Also, SIC along coastal lines could also have errors due to sub-pixel contamination of land cover in an instantaneous field of view of AMSR-E data
Averaging period and the update timing of daily data
* In general, sea-ice extent is defined as a temporal average of several days (e.g., five days) in order to eliminate calculation errors due to a lack of data (e.g., for traditional microwave sensors such as SMMR and SSM/I). However, we adopt the average of two days to achieve rapid data release. The wider spatial coverage of AMSR-E enables reducing the data-production period.
* Usually the latest value of daily sea-ice extent is fixed and updated at around 1 p.m. (4 a.m.) JST (UT). Before the value is fixed, we also assign a preliminary value of daily sea-ice extent several times (usually three to four times) as an early report, which is determined without the full two-day observation coverage. (The fixed values of sea-ice extent are determined with the full coverage of observation data.
They admit to a 10 percent error. Right now there are tons of areas not counted because of holes in the ice. But the graphs show the data is nearly the same. Except differnt algorithms. Both places update there data all day until there final numbers they extract from the exact same files.
Posted by: Chris Biscan | August 29, 2011 at 22:33
Paul Klemencic
http://www.ijis.iarc.uaf.edu/seaice/extent/plot.csv shows:
08,23,2002,6109688
08,24,2002,6117188
Posted by: Lucia (The Blackboard) | August 30, 2011 at 00:33
Chris: Thanks very much for the information. I will learn more about IJIS procedure.
But Chris, think bout this: The daily average melt loss shown in the MASIE report has averaged 81k per day for the last five days, with huge losses in the Beaufort and Chukchi, especially yesterday.
Do you see these losses on the Bremen map? I can't see them. Up until the preliminary Bremen map today, the last five days seen areas of ice show back up or ice pack spreading out on the Bremen map in the Beaufort region, and to a lesser degree, in the Chukchi. The MASIE reported extents make no sense if the dates are correct.
And although the IJIS data seem more realistic over this period, I think high too. And I simply couldn't find the ice consistent with the extent in the IJIS report on the Bremen map; I keep coming up about 200-250k short. Please note, this in spite that the Bremen map shows ice down to 10% in concentration. According to the legend on the map, the dark blue areas are 0-10%, light blue 10-20%, and teal green 20-30%. In spite of this, the Bremen map seems to clearly show less ice extent than reported by IJIS.
The maps on the MASIE site are ten days old; and the data listed in the spreadsheet today, match Bremen map observations from August 22nd.
We also know that NSIDC puts data on their graph using a five day average, which leaves the last average plotted three days behind today. They supposedly do this to avoid the day to day variation due to measurement errors, so they don't want to release raw data every day. But then why would they allow MASIE to report their NSIDC previous day's daily data every morning? That doesn't make sense.
My hopefully reasonable guess is that MASIE is receiving data that was embargoed by NSIDC for five days to avoid complicating their release of the five day average, and then reporting it the next morning for a six day delay.
If we adjust for that delay, the reported numbers match our observations of the Bremen map beautifully. Regions that should lose ice extent, do, and close to estimates of ice extent made from the map. The total extent matches the observed extent. And the day to day extent change matches up well with the changes in the map. The 'delayed' MASIE data fit the Bremen map like a glove.
I still don't understand the IJIS reported numbers:
they don't make sense to me since they never show an extent gain in August (MASIE occasionally does);
they are incompatible with the total area shown, or the changes observed, in the Bremen map;
the total extent reported is higher than the other reports (in August);
and differences in measuring open water gaps that lead to different ice extent measurements don't seem to explain the differences.
Posted by: Paul Klemencic | August 30, 2011 at 01:07
Lucia, You had to go back nine years to find one data point? You are missing the point. That ice pack with over six million sq km could easily see big extent gains. Its not happening in recent history. But it should. You are way too literal, and not using critical thinking.
The system variation (freezeup, and ice floe consolidations etc) should change the ice extent level by 2-3 percent occasionally, and this alone should lead to reported extent gains. The measurement system also should have variation in measured extent of several percent. When you have a system whose output (daily extent) can vary 2-3 percent; and added to that, additional variation of 2-3% due to measurement variation; then the reports generated on this system by this measurement should vary enough so that even with a 1% average daily declines, the reports should occasionally show a gain in extent for 2-day averages. The MASIE show two such gains just this August; but the IJIS does not.
You can construct such a system on your statistical tool, and you will see reported gains in the output. If you disagree, tell me why.
Posted by: Paul Klemencic | August 30, 2011 at 01:23
Lucia, You had to go back nine years to find one data point?
No, she was just showing you the first such occasion in the data set. If you want a more recent example, look at August 29th of last year. There are others, too.
Posted by: Ned Ward | August 30, 2011 at 01:26
Paul,
The difference is because it's late summer and Jaxa uses a 12.5 grid res and UB uses a 6.25km grid res.
From the NSIDC:
Data from the Advanced Microwave Scanning Radiometer – Earth Observing System (AMSR-E) sensor, processed by the University of Bremen, show ice tracking near 2007 levels. The AMSR-E instrument can detect small but widespread areas of open water within the ice pack in the Beaufort and East Siberian seas, because of its resolution (6.25 kilometers or 3.88 miles). Normally, NSIDC uses data from the Defense Meteorological Satellite Program (DMSP) F17 Special Sensor Microwave Imager/Sounder (SSMIS). F17 provides a longer time series of data, but at a 25-kilometer (15.5-mile) resolution.
UB probably picks up 50,000km2 more of open water it the region between 135E and 105E from 85N to 80N compared to jaxa. The Larger grid boxes will be prone to pick up more ice.
this is why as I said go to 2003 not 2011 now and see the differences back then when the summers had less holes like this the stats were the same.
This is the same reason norsex is way off.
I don't think there is any other way to explain this.
Posted by: Chris Biscan | August 30, 2011 at 01:36
Paul--
Go back? The data are organized from early to late. I opened, scanned the first year and found a data point in the first August for which JAXA reports. I have no idea what I would have found if I'd looked later. Of course I can construct such a thing. I can permit errors to exhibit non-zero autocorrelation which I think may be realistic, instead of assuming that they must be iid which is likely unrealistic.
Because I can construct such a thing, I don't find JAXA's failure to show numerous positive extent changes in August suspicious.
Posted by: Lucia (The Blackboard) | August 30, 2011 at 01:49
OK, Ned and Lucia, I cede the point about IJIS never reporting extent gains. The question still remains about variation, or lack thereof, in the IJIS data compared to the NSIDC data.
I have daily data now for NSIDC. I going to try and de-construct estimated daily data from the 2-day averages reported by IJIS.
Then I will use both daily databases to build 5 day averages to compare with the NSIDC report, and 2 day averages to compare the IJIS report and data. Then I will return ...
Posted by: Paul Klemencic | August 30, 2011 at 02:19
Paul Klemencic
Do you mean daily extent data? Back to what year? Is it online? I'd love to see that.
Posted by: Lucia (The Blackboard) | August 30, 2011 at 02:28
Paul,
I believe you are correct that the MASIE data is time-shifted in comparison to UB or JAXA.
The lack of increases in August, OTOH, is not such a surprise. A very large portion of the ice pack is susceptible to only minor sensor error. Where the error is significant are the ice edges - and even that should be a systemic error, not random. In addition, with two-day smoothing we really can't say with certainty how many days showed an increase; we'd need the individual day totals from JAXA to make that claim.
My quick back-of-the-envelope calculation is a mean circumference of 8000km. Consider the ice-edges to be 25km wide. This gives a high uncertainty area of 200,000 km^2. If a typical August decrease is 40,000km^2 we'd need an error of 25% or more to see increases due to sensor error. Or, since they state in the AMSR-E validation documents that errors can approach 10%, we'd need decreases of less than 20km^2/day before possible sensor errors give false positives.
Posted by: Kevin O'Neill | August 30, 2011 at 02:38
OK, Ned and Lucia, I cede the point about IJIS never reporting extent gains. The question still remains about variation, or lack thereof, in the IJIS data compared to the NSIDC data.
If you detrend both the IJIS and NSIDC data sets, and examine their residuals, you will see that IJIS has a much lower variance than NSIDC.
This is not surprising, suspicious, or indicative of any "tampering" with either data set.
As you have repeatedly been told, the IJIS data are based on imagery from the AMSR-E sensor, while the NSIDC data are based on SSMI. They are two different instruments, with different spatial resolutions, signal-to-noise ratios, etc.
Posted by: Ned Ward | August 30, 2011 at 04:04
Golly Lucia, I asked you to read my comments on the other thread, in fact, practically begged you. You are too smart sometimes... The whole point of this discussion from the very beginning was to review the NSIDC data points in the MASIE site spreadsheet, and see if we could use them. That is why I spent so much time evaluating the MASIE data.
If we can confirm the date issue, then it provides day by day data, region by region, going back some years.
Posted by: Paul Klemencic | August 30, 2011 at 04:09
Golly Paul. I read many of your comments on the other thread.
So what is the answer to the question I actually asked. Is the daily NSIDC NH extent data on line? How far back?
Not withstanding whatever it is you think is "the whole point", the "whole point" of my question is to try to find out if there is NSIDC NH extent data on line, how far back it goes and whether there is a link.
Posted by: Lucia (The Blackboard) | August 30, 2011 at 04:22
really big drop on the UB final.
Compaction all over the ice pack..not sure if it will show up tonight in Jaxa fully. but it will be like this the next few days maybe a bit faster.
Posted by: Chris Biscan | August 30, 2011 at 04:34
Chris,
I have to say, I've been following your posts all summer long, and I have to say your predictions just don't seem to happen like you say.
bro, we're all on the same side here and maybe I'm just a crazy Slovenian guy. But it is looking like things will end up around 4.7 or 4.8.
Posted by: Tzupancic1 | August 30, 2011 at 04:46
4.7 or 4.8?
Could be--but tonight we're sitting right at 4.9, and we're not quite out of August yet.
Posted by: Kevin McKinney | August 30, 2011 at 05:47
I added the most 1 week change in the 7 day average extent to my weighted statistical model. The regression using that zoomed right to the top of the heap taking a weight of 80%. My current best fit says 4.47 million km for the minimum 7-day average of JAXA.
Posted by: Lucia (The Blackboard) | August 30, 2011 at 05:56
TzPancic,
We are at 4.9 million km2 on jaxa and 4.7-4.8 milliom km2 on UB.
We have two more days in August with very favorable conditions. Jaca runs a two day average. So today was close to a century break.
We will be at 4.75 - 4.80 million km2 on September 1st. With 2-3 weeks maybe more to go. But you are preductung that and calling my forecasts bad.
You also posted a link to a H5 vort chart and saud it wasnt lookImg good fir ice loss that week.
Are you trolling?
Posted by: Frivolousz21 | August 30, 2011 at 06:14
Sorry for the typos. On I phone.
Posted by: Frivolousz21 | August 30, 2011 at 06:20
Lucia, funny how your number keeps coming down... by the time we hit the minimum of 4.20, you should be right on target : )
To find the MASIE site: go to Neven's daily graph page by clicking the tab at the top of the page. On his page, he has the links to the most important Arctic ice related sites. On the right side of the links shown, he has the links to the NSIDC/NIC MASIE homepage, and a link to the download site for Daily Data. To really understand the data, you need to go to the home page and it shows the downloadable CSV file, and it shows these choices:
MASIE Home
Browse Regions
Time Series Plots
MASIE FAQs
About MASIE
Documentation
Browse regions gives a map, and by clicking on each region in the map, it shows the area for each region, that they give the ice extent for that region. These regional data are in the CSV file. The other choices are self-explanatory.
Since MASIE gets the data from NSIDC, you will note the file is labeled MASIE NSIDC/NIC Daily Ice Extents by Region. The column market N. Hemisphere is the total for the Arctic.
Enjoy.
Posted by: Paul Klemencic | August 30, 2011 at 06:21
Just a few quick comments on the Bremen map movement today. The interaction between the high pressure over the NP and the low on the Siberian coast, is pushing the ice pack on the boundary right toward Severnaya Zemlya and past it at an incredible rate. If this keeps up for another day or two, this section of the pack will be sheared and moved into the Kara.
Ice moving receded away from Svalbard toward the pole at pretty fast clip too. The edge of the pack is now 200 km above Svalbard.
The compaction everywhere was big... the extent lost over 100k today, which we will see in the MASIE numbers in six days.
Posted by: Paul Klemencic | August 30, 2011 at 07:18
Paul, that is now happening in the Beaufort too and is going on in the Fram and Greenland sea.
these next 3 days this pattern holds.
We could have a century break tomorrow if that ice in the Beaufort/Western Arctic basin gets smashed in and the ice can move off Greenland some more.
If that area between 135E and 105E melts out, look out.
Posted by: Chris Biscan | August 30, 2011 at 07:41
A bit of bickering is fine by me. We have some people pushing the envelope of careful wording, and some people pushing back against that. Nothing wrong with that, keeps things juicy and on our toes. I'm not with or against anyone, as I don't know what to think (of the IJIS issue and the short-term weather forecasts).
But let's not ask questions like the one above, okay? That's just producing negative energy.
Posted by: Neven | August 30, 2011 at 09:59
No Problem.
Posted by: Chris Biscan | August 30, 2011 at 10:16
Cool.
With that prelim drop of 64K in IJIS SIE, it looks like you are on the right track, Chris. I'm looking forward to what the next days will bring. 64K is already pretty big in this phase of the season.
But again, if that negative ADA comes about, I'm not sure if it will cut the melting season short. Unless we get some heavy flash melting because of it. I'm confused this year... :-)
Posted by: Neven | August 30, 2011 at 10:32
I have updated the Fram Strait and Slush puppie animations.
Posted by: Neven | August 30, 2011 at 11:51
I have to take this back. I've been reading up on some posts from last year (when I was less confused) and would say the cut-off date is around September 10th. If after that we have low pressure areas dominating the American side of the Arctic, things will come to a screeching halt.
So this negative ADA might slow things down, or whatever, but not end the melting season. It will if it persists after the 10th of September.
Posted by: Neven | August 30, 2011 at 12:28
@William Crump: You are comparing 2011 to 2010 for a single day to say there is a trend of a steep decline. That is not a valid methodology.
You're right, William, it's not valid methodology. Which is why I would never say something so stupid. Did I use the word trend? No. Did I refer to ice free conditions in less than 10 years? No. I merely reflected on the merits of your contention, which at the time was, IMO, flawed. You said that there "is no significant difference" between 2010 and 2011. I looked at the data as it stood when I posted and noted a 7.5% difference. I did not reflect on the season as a whole, or whether there was a difference at maximum, or whether I think it will be 7.5% at minimum. Only that at the time you said that, the data did not support you. So at the time you said that, there was a 7.5% difference in area. Clearly you don't think that is significant. I think it is sufficiently noteworthy that your handwaving was worth pointing out.
"annual trend of a 7.5% drop"
Really William, this misquoting is beneath you. Did I talk about an annual trend? Why is that you have to misrepresent what I have said every time I disagree with something you say.
You are (of course) quite right that the area ticked upwards the day after I posted. However, unlike you I have never claimed the power to travel through time, so I could not know that would be the case. I merely reported the facts as they stood at the time I posted.
"I have not overlaid the 2010 chart over 2011 to see what differences have occurred over the last 365 days".
I haven't either, so I can't (and wasn't pretending to) comment on the year as a whole. But then, since you regard central basin area as the holy grail, maybe you should do that. Do let us know how you get on.
"Are you predicting that the 2011 September data point will be below this?"
Presuming that this is also addressed to me - I don't deal in predictions, any more than I can travel through time. As I have repeatedly stated. I have made one prediction in two seasons here - In late June last year I predicted an IJIS minimum of 4.8 million, 13,600 km^2 out on the final number. That is the only prediction I have made here, and the only prediction I will make. I don't even vote in Neven's polls. So if you want me to venture a number, you'll be wanting in vain.
I don't really care that you are wrong about this, William, (well, IMO) but I really wish you would stop lying about what I have said in a sad attempt to bolster your own arguments. If you are going to quote me, quote me. If not, please just ignore me.
Posted by: FrankD | August 30, 2011 at 12:51