Exploring Arctic Ice Minima

Every year 2007-2013 had a lower Arctic sea ice extent than every year before 2007.  2014 seems likely to continue this record.  I'll also suggest below that maybe the Arctic has entered a 'new normal', with September ice extents bouncing around 4.7 million km^2. 

For some data to work with further, I pulled the NSIDC September figures.  It's a small, simple text file, so you can check yourself what follows.  First up, let's draw a figure of what we're looking at -- but don't connect the observation dots.  Our eyes tend to be led to conclusions by the superposed lines.

You can check some of the sources for ice before 1979 and see that figures below 5.5 million km^2 are unprecedented in the longer records as well.  To have data precision and consistency, though, I'll stay with the 1979-present.

What else can we say from eyeballing the data?  Since the 1979 starting point:

• There have been 2 record highs (1980 and 1996)
• There have been 8 record lows (1984, 1985, 1990, 1995, 2002, 2005, 2007, 2012)
• There have been more record lows in the last 10 years (3) than record highs in the full 35 year record
• 1996 is about the last year one could say there was no trend in the data
• Versus eyeball curve fitting, 1996 is the most exceptionally high year (not just an absolute record, but even higher above smooth curves we'd try to fit to the data than any other year).
• More recent years look like they have more scatter than the earlier years
• It looks like we might want to divide the period in to 3 intervals -- 1979-1996 (the longest arguably trendless span), 1997-2006 (an intermediate with at least some overlap on the earlier figures) and 2007-present (entirely outside the range of the previous years)

But maybe your eyeballs disagree with mine, and perhaps the appearances are deceiving.  So on to working with numbers, which will also lead us to some additional ideas.

Read more »

Arctic Ice Guesses 2014

Have to bite the bullet here and discuss my guesses for the September 2014 Arctic sea ice extent average.  The thing which has made them so difficult is that they're so different from each other.  Now, one method I've retired.  It was simply so bad last year that there's no point in continuing it.  That is the one I did based on a population growth (of ice-free area) curve.

That leaves, however, two different model-based guessers.  The first one, which appears at the Sea ice prediction network as 'Wang', is based on doing a statistical regression between what the CFSv2 (climate forecast system, version 2) predicts for September ice area and what is observed.  The second also uses CFSv2, but in a different way.  Namely, we know that the model is biased towards ice being too extensive (which the Wang method addresses statistically) and to being too thick.  The Wu method is based on thinning the ice and seeing what the extent is thicker than a critical limit (60 cm it turns out).  (Both Wang and Wu work with me, or vice versa, and we discuss how to work on these guesses.)

The guesses are:
June -- Wang -- 6.3 million km^2 0.47 stdev
July -- Wang -- 5.9 million km^2 0.47 stdev
July -- Wu -- 5.1 million km^2 0.56 stdev
June -- Wu -- 4.8 million km^2 0.65 stdev

One of the things to notice is that the two estimates moved towards each other from June to July.  Wu rose, the higher Wang declined.  The second is, the Wu method has a standard deviation (variability of its estimate) that is double what it was last year.  Whatever is going on in the model, it is much less self-consistent in previous years.  Much more uncertain.  This is one of the reasons for ensemble modeling (part of the Wu approach).

You can also see that the Wang estimate is the highest of all -- even higher than the Watts up with that group estimate.  This is true in both June and July.

So, what's up?  Well, I'm not sure.  Some of it is certainly related to sea ice thickness estimates.  Xingren (Wu) did a different approach based on thickness for June, which we didn't submit, but which landed in between the official June estimates from Wang and Wu.  With the step towards convergence from June to July between Wu and Wang methods, I'm inclined to guess (a meta-guess) 5.5 million km^2 for September.  If this were to occur in reality, it probably suggests something important.  What, exactly, I'm still pondering.

Data are ugly

Current news about whether there really is an increase in Antarctic sea ice cover is reinforcing my belief, shared by most people who deal with data, that data are ugly.  This work argues that the trend that some have seen in some trend analyses has more to do with the data processing than with nature.  I encourage you to read the article in full itself.  It is freely available.

From the abstract:

Although our analysis does not definitively identify whether this change introduced an error or removed one, the resulting difference in the trends suggests that a substantial error exists in either the current data set or the version that was used prior to the mid- 2000s, and numerous studies that have relied on these observations should be reexamined to determine the sensitivity of their results to this change in the data set.

One of the obnoxious things about data sources is that they don't remain the same forever.  This is not so much a problem for my concerns about weather prediction, since the atmosphere forgets what you said you observed in a few days.  But for a climate trend, the entire record is important.  For the data set being discussed, the Bootstrap Algorithm (Comiso) applied to passive microwave, we immediately run in to data obnoxing.  Since 1978, there have been several passive microwave instruments -- SMMR, SSMI F-8, SSMI-F11, SSMI-F13, 14, 15, AMSR-E, SSMI-S F16, 17, 18, and AMSR-2.  They didn't all fly at the same time, and they don't have exactly the same methods of observation.  And none of them exactly observe 'sea ice', which leads to a universal problem which we (people who want to use these instruments to say something about sea ice) all have to deal with.

So a few considerations of what all is behind the scenes of this paper and the earlier Screen, 2011.  The latter paper involved some of my work (read deep in to the acknowledgements).  This one doesn't, but the fundamental issues are the same ...

Read more »

Ice Science Cafe

This Thursday (June 26th) I'll be talking about ice, and, better, yet, answering questions about ice at the Annapolis Cafe Scientifique.  The time will 6:30 PM.  Same location as usual -- Cafe 49 West.  Local folks are invited, and non-local are welcome to pose questions here. 

 

I'll also invite folks to suggest topics for me to prepare for.  My sources of information on sea ice are pretty different than my audience's, so it's hard for me to tell what people have been hearing about.

Our Sea Ice Outlook guesses this year are wildly different -- 4.8 million km^2 and 6.3 million km^2.  The former is above the median of contributions.  The latter is the highest of all, even higher than from Watts and company.  We have a third, unpublished, guess in between the two.  I need to check out a couple of things before writing it up.  Given the spread we're encountering ourselves, I think we're going to learn a lot this year.

AABW in the news!

It only took 25 years, but my thesis topic is now becoming newsworthy!  Gluttons for punishment can see at least the abstract at A model of the formation of high-salinity shelf water on polar continental shelves.  Which is aimed at one of the important ingredients for AABW (Antarctic Bottom Water).

I've been reluctant to blog about the topic because it is, after all, my baby and I'm sorely tempted to post at excruciating length and detail.  (Not that there aren't other people who have studied the topic before or since, but I'm one of the people who has.)

I'll take this note as opportunity to get in to some detail about the weirdness that is sea water, and come to the climate change, carbon dioxide burial, and heat burial, aspects later.  The story of AABW turns on some odd facts about how sea water behaves in Antarctic conditions.  Not least, it can go below freezing.

Read more »

Harry Bulkeley: A few questions about global warming -- Answered

An opinion writer (a retired judge) asked a few questions in his Galesburg, IL local paper, and I'll provide some answers here.  As always, I encourage you to read the original.

The good judge, like the usually informative Mr. Krauthammer, starts off on a very wrong foot, with bad philosophy of science.  There are many facts in science -- the earth is round, the sun is hot, there is a greenhouse effect, and CO2 is a greenhouse gas.  All can be questioned -- but not in the trivial way that Bulkeley and Krauthammer seem to think.  'I question it' is trivial, and pointless.  If you have a _scientific_ question about these things, or any other, it is because, and only because, you have scientific evidence that the 'fact' is false.

Climate change, as even commenters in agreement with Bulkeley note, is indeed a fact.  Climate changes, that's a fact.  One of the tasks of science is to try to understand the hows and whys of that fact.

Let's see about the questions:
1) Average temperature has indeed gone up the past 15 years.  This is a question, apparently, because the author didn't bother to look at the data. One can experiment with time periods and trends at NOAA/NCDC.
http://www.ncdc.noaa.gov/cag/time-series/global/globe/land_ocean/1/2/1880-2014  
It's worth paying attention to the fact that climate trends are defined on 30 year periods, not 15.  Some discussion of why this is the case is at
http://moregrumbinescience.blogspot.com/2009/01/results-on-deciding-trends.html

Read more »

PIOMAS ice volume anomaly q

Q: I've got a question that you may be able to answer with your sea ice hat on.

Why is it that the minimum volume anomaly of ice appears in mid summer and not in September when the minimum volume happens? See piomas.
(From Alastair)

A: I'm hard-pressed to tell what season the anomaly extremes are occurring from this graph, but can say that your reading isn't surprising to me.  At the seasonal minimum of ice volume, you're at the minimum -- so it is relatively difficult to get even lower.  Late winter, towards the maximum doesn't have that problem -- since it's maximum, there's lots of room to go lower.  But the Arctic is cold in winter, so it's going to pile up a lot of additional mass (from the prior minimum) pretty consistently.  The place/time that there's the most room for decreasing the ice volume is a time of year when the ice volume hasn't already declined a lot, but it is melting rapidly -- early- to mid- summer.

As you also see from the figure, the main trend is the year-round decline.  For your question, we're looking at the timing of largest excursions below the line. 

The converse question, for largest excursions above the trend, is that I'd expect those in late fall - early winter, during the freeze-up season.  Then, there isn't much volume, and it's freezing fast.  A couple weeks earlier start to freeze-up makes a comparatively large difference.

Evaluating 2013 sea ice outlook estimates

Very late wrap on my 2013 sea ice season guesses, but, as I was thinking in early August, the straight statistical one busted pretty badly.  The observed (NSIDC) September Average sea ice extent was 5.35 million km^2.  The guesses were:

The original outlooks (end of May) were:

3.9 million km^2 -- Grumbine, Wu, Wang statistical

4.1 million km^2 -- Wu, Grumbine, Wang model-based

4.4 million km^2 -- Wang, Grumbine, Wu model-based statistical

The end-June Wu et al. estimate was 4.7 million km^2.

The end-July Wu et al. estimate is 4.57 million km^2.

As I suggested then, I'm going to get back in to the statistical approach to try to fix it up.  While we don't expect excellent estimates from it, if it's using a sound basis, it shouldn't be wrong by this much.  I have some ideas on how to do it. 

Still, better estimates than Stephen Goddard's end of August (the 26th) prediction -- Doubling of Arctic Ice in 2013.  Doubling September 2012 would have meant 7.26 million km^2.  (I was just touring web sites, and saw this article, which reminded me that I hadn't posted the evaluation of my forecasts.)

Sea ice prediction updates

I'm a little behind on our updated estimates for the September average sea ice extent for the ARCUS outlooks.

The original outlooks (end of May) were:

3.9 million km^2 -- Grumbine, Wu, Wang statistical

4.1 million km^2 -- Wu, Grumbine, Wang model-based

4.4 million km^2 -- Wang, Grumbine, Wu model-based statistical

The end-June Wu et al. estimate was 4.7 million km^2.

The end-July Wu et al. estimate is 4.57 million km^2.

In other words, it looks like the model has picked up on the relatively (compared to last year) slow decline in ice cover and is keeping more ice around.

What will actually happen, well, keep an eye on reality.  But, in the likely event (it seems to me now) that the purely statistical model is very wrong, time for a new statistical prediction method.  I'm starting to get ideas on how to go about the update.

2013 Sea Ice Estimates

Our June guesses for 2013 are 3.9, 4.1, and 4.4 million km^2 for September monthly average sea ice extent as computed by NSIDC.  See the ARCUS SEARCH Outlook for the full current collection of estimates.

Again, I'll put our estimates in context of some other estimation methods. 

• Climatology 1979-2000: 7.03 million km^2
• Climatology 1979-2008: 6.67 million km^2
• Linear Trend Climatology 1979-2008: 5.15 million km^2
•  
• Wang, Wu, Grumbine model: 4.4 million km^2
• Wu, Grumbine, Wang model: 4.1 million km^2
• Grumbine, Wu, Wang statistical ensemble: 3.9 million km^2

All three of our estimates are well below what they were last year -- 4.9, 4.8, and 4.4 million km^2, respectively.  For the two model-based estimates, this is because the model has continued to thin its ice cover, and that ice cover is less extensive.  Then the model (which is coupled air-sea-ice) is not seeing any reason for the ice to be getting markedly thicker or more extensive.

The statistical ensemble reflects two items.  One is, last year being a new record, well below the estimate (by 0.8 million km^2), pulled down the best curves that can be fit through the data.  The second is, the curve itself shows a zone of steep decline.  This is where the feedback of more open water -> warmer ocean -> thinner ice -> easier to melt ice -> more open water in summer is starting to take off rapidly.  (If the basis for my using this curve is reasonably correct, that is.)

The linear 'climatology' estimate last year was 5.23 million km^2, versus the observed 3.61.  Too high that year, and every year 2007-2012, and very likely many more to come.  Merely points to a straight line not being a good representation of how the ice pack is changing, which is no surprise.  But it is also a reminder that talk of sea ice 'recovery', which will no doubt occur in some corners when (if) 2013's extent is greater than 2012's, is nonsense. 

2012 Sea Ice Prediction Evaluations

Before launching in to my estimates for this year, I have to evaluate last year's estimates.  Not much point in making estimates if you don't re-examine how well the previous ones were.

Last year's observed (NSIDC) September monthly average sea ice extent was 3.6 million km^2.  That blew away the old record set in 2007.

Our June 2012 Estimates were:
4.9 Wang, Grumbine, Wu -- statistical correction of CFS
4.8 Wu, Grumbine, Wang -- semi-physical correction of CFS
4.4 Grumbine, Wu, Wang -- Purely statistical from prior Septembers

July:
4.9 Wang, Grumbine, Wu -- statistical correction of CFS
4.7 Wu, Grumbine, Wang -- semi-physical correction of CFS

August:
3.9 Wang, Grumbine, Wu -- statistical correction of CFS

No updates in August for the semi-physical CFS correction as it hadn't looked like it was coming out any different than the prior months.

The Wang et al. was the best of all August estimates, and was still too conservative by 300,000 km^2 (the approximate land area of Poland, the Philippines, Norway, or Ecuador).  The observed mean extent, of 3.6 million km^2 was about 20% larger than India.  The 1979-2000 average was 7.0 million km^2 -- almost double what we just saw, and about the 10% smaller than Australia.

Assessing forecasts

This is actually part of pursuing whether ESMR was screwy, but I decided that to show that nothing was up my sleeve, it was time to talk some about assessing forecasts.  That, and it's something I've been meaning to talk about for a while.  The thing is, forecast assessment is not nearly as simple as we sometimes think.  Having judged many a science fair project that is comparing weather forecasts, I've seen many of the same issues come up there, too.

For precipitation forecasts, people (science fairs included) often think about either 'probability of detection' -- i.e., what fraction of the time that there's rain did the weather forecast call for rain, and 'false alarm rate' -- what fraction of the time did you get no rain even though the forecast called for rain.  Both are potentially meaningful, and both have serious problems if used alone.

Read more »

Was ESMR screwy?

A reader here also asked about the pre-1979 satellite data over at my question place. The thing is, we do have pre-1979 satellite sea ice data -- the ESMR (Electrically Scanning Microwave Radiometer) 1973-1976.  It was a much simpler instrument than the SMMR, SSMI, SSMI-S, and AMSR which started flying in 1978 and since.  The more recent ones have two very important improvements over the ESMR -- they use multiple channels (think of it as colors) and they use both horizontal and vertical polarizations rather than just total power.

Ok, translation to English.  Our eyes look in three channels -- red, green, and blue.  Different creatures use different numbers of channels.  Dogs use only one, black and white (as we do if the lighting is very low).  Mantis shrimp use 10 channels.  Bees use ultraviolet.  And so forth.  The key is that the eyes respond to some number of colors.  Numbers vary, and what color band also varies. 

The other thing about electromagnetic radiation is that it can be polarized -- vibrating in one way versus another.  The two linear polarizations are horizontal and vertical.  ESMR just lumped them together and measured total power.  SMMR and the rest measure horizontal and vertical polarization separately at most of the channels.  Basically, SMMR and all the more recent instruments have high quality color vision versus ESMR being a rather fuzzy black and white.


But ... black and white is still better than not being able to see at all.  The question then arises in retrospect whether we can use the black and white instrument, ESMR, like our more recent color vision instruments.  Now, in part, I know the answer already -- you can't.  More precisely, you can't do it well enough to satisfy my colleagues at NASA-Goddard.  Perhaps, though, it can be done accurately enough to answer some questions of interest, even if not accurately enough to be entirely comparable to the modern instruments.

I have some ideas, naturally, and have been been pursuing them a bit -- enough to know that there's a fair chance of getting useful answers.  Whether it's useful enough to answer questions of interest ... well, I'll also invite questions you all find of interest.

Quick Notes

 Sea Ice:

• Alastair, in case I didn't confirm it earlier, you're on for the 50 quatloos for September NSIDC sea ice extent over/under 4.15 million km^2.  I'll stand with that even with the recent drop in ice extent.
• All are welcome to enter in the poll at the bottom.  I see few are taking the higher side of estimates.  I hope that's not because you're intimidated by my guesses!  When I'm confident about what I'm doing, I'll call them predictions.
• The June Arctic Sea Ice Outlook is out (more in a post to come).  The guesses I'm involved with are 4.4, 4.8, 4.9, with the 4.8 likely to come down when we make a run with more recent initial conditions.  (Again, this first one was from December)
• GCOM-W1 did indeed enter orbit safely and instruments are powering up and being checked out.  The bad news is that AMSR2 is looking at 12-18 months from launch before its data distribution (to me at work) becomes operational.
•  The recent rapid drop-off (and then plateauing) is more or less to be expected.  I think.  If I'm wrong, Alastair stands to win his 50 quatloos. 

Japan:

• My meeting went well.  My thanks to the JMA and JAXA (Japanese Meteorological Agency and Japanese Space Agency). And especially Misako Kachi who took lead in coordinating it.
• Tokyo, Kyoto, Osaka, Nara, Kamakura, and Mt. Fuji all are excellent places to visit as a tourist.  Selected photos to come.
• Someone with approximately zero Japanese language skill, but ok English, can indeed tourist around Japan ok.  (Exhibits 1 and 2 being my wife and me.)
• If you're not Japanese, be sure to stay at a Ryokan if you have a chance.  We stayed at Fujitomori.
• 13 time zones followed by several 12 hour working days are difficult.

 continued ...

Read more »

Sea Ice Predictions 2012 Open Poll

The poll for your guesses on September 2012 sea ice extent (NSIDC) is open.  All the way to the bottom of the page.  If you'd like to explain how you arrived at your guess, please do post a comment explaining it.  If it's a substantial explanation (i.e., one I could steal research for my professional use), I'll also bring it up here to the main post.

Poll is open to the end of June, EDT.

2011 Sea Ice Outlook Verification

Sometimes time passes faster than you'd think.  I'm sure that's where Einstein got the principle of relativity from.  In any case, I realized that we're about to be submitting our guesses for the 2012 sea ice outlook and I never did get around to examining, on the blog, how our guesses for 2011 did.  In brief, pretty well.

• 7.03 million km^2 -- Climatology 1979-2000
• 6.67 million km^2 -- Climatology 1979-2008
• at least 5.5 million km^2 -- Joe Bastardi at WUWT
• 5.31 million km^2 -- Linear Trend Climatology 1979-2008
• 5.0 million km^2 -- Wang, Wu, Grumbine model, June
• 4.8 million km^2 -- Wu, Grumbine, Wang model, June
• 4.61 million km^2 -- NSIDC observed September Average sea ice extent
• 4.6 million km^2 -- Wu, Grumbine, Wang model, August
• 4.4 million km^2 -- Grumbine, Wu, Wang statistical ensemble, June
• 4.4 million km^2 -- Crandles, L. Hamilton (at Neven's, but also cited in comments here)
• 3.9 million km^2 -- Alastair (in comment here)

As I'd expected,  the statistical ensemble was low, and the Wang et al. model was high.  The June Wu et al model was a touch high, and the August version was right on (given its reporting precision).  Also, and reassuring, is that all our guesses were within 0.5 million km^2 of the observation.  That's what we've estimated as the standard variation in sea ice extent.

It's a good sign that we were within that range.  Also, 0.5 million km^2 is the variability estimated by almost all groups that provide an estimate for it.  Even though that number is suspiciously round, we actually arrived at it by data analysis.  It's a statement of how much the ice pack varies, rather than the quality of the methods.  The quality of the methods is how their error compares to the 0.5.

For the coming year, we've got some modifications in mind (experiments) for the Wu et al. approach, and I've already made some for the statistical ensemble. 

Plus, I'm going to be taking a look, perhaps getting Wu to play too, at whether this past winter's heavy sea ice cover in the Bering Sea was something we could have (or did) estimate in advance.  The model runs are already done.  It's 'just' a matter of analyzing them.

Heartland on Ice

What's old is new once again.  I'd actually written most of this in February, when Heartland was in the news.  But between one thing and another, hadn't posted it.  Now that  they're in the news again, it seems once again to be apt.  As usual, my interest is on the science.  Since a number of Heartland supporters are saying things in the vein of Heartland made a misstep in acting as they did, because it takes away from their message on the science.  The supporters think Heartland is doing well on the science.

Now that I've looked in to what Heartland Institute has had to say on sea ice, I can say with confidence that they are not doing well on the science.  They don't know (or lie about) the difference between ice area and ice extent, don't know how much area of ice there is, don't know where it forms, make up numbers even if you ignore the difference between area and extent, lie about what authors say in their scientific papers, treat 2 years as plenty for establishing a climate trend if it is in one direction, but ignore the 30 year trends when it isn't, don't know the difference between sea ice and ice shelves, don't understand how sea level changes, and others I'll let you classify yourself.

The gory details, examples being from their web site, are below the fold.  A different point is, I don't expect everyone to be expert on sea ice, or even pay attention to it.  If Heartland had ignored sea ice, that's fine (at least it is if they're not saying things which require understanding sea ice).  But they chose to write about it.  And the people whose articles I'm quoting are Jay Lehr, their science director, and James M. Taylor, a senior fellow for the Heartland Institute focusing on environmental issues.  In other words, major players in deciding what Heartland says about science, not someone who might once have said something stupid about sea ice while passing through the office.

Having done my homework, I'm comfortable in saying that Jay Lehr and James M. Taylor are unreliable sources on the science, and Heartland Institute is as well.  See below for my homework example.

Read more »

Tempest in an ice pot

There's a bit of a tempest at WUWT and Steven Goddard's regarding satellite data on sea ice.  And there are some anxious comments about FOIA filings to be made regarding data that's been hidden.  I have to concede that it takes a bit of a historian to know what's really being said in the IPCC first report regarding sea ice.  (see page 224, figure 7.20 for the source of the tempest.)  Fortunately, I am such a historian :-)  To give away the ending: There's nothing terribly dramatic going on in the sea ice or the history. But I'll take the chance to talk history of sea ice analysis.

The tempest arises because the figure shown there has sea ice cover in the early 1970s in the Arctic being notably lower than it was through the 1980s.  (Curiously, no interest is displayed at either site regarding the Antarctic being far higher in the 1970s than the 1980s.)  If the figures were commensurate with the satellite observations such as those used in the Cryosphere Today anomaly figures, it would suggest that the Arctic coverage of the last few years wasn't as remarkable as we've generally thought.  (On the other hand, it would also say that the 'record high' Antarctic coverage of recent years was not particularly high at all, a point being neglected at both sites.)

Read more »

AMSR-E failure and fallout

Update 17 October: The meeting last Tuesday gives little hope.  There will definitely be no data for weeks.  I don't know what prevents a conclusion of never.
original:
AMSR-E has failed and is probably permanently out of commission.  For most of you, that's merely news.  Perhaps a source of amusement and interest is now gone.  For me, since I use(d) it in my day job, AMSR-E failing means some real work.  Most of that work was already planned, but now it needs to be done more speedily.

As I've often said here, and even more often in 3d, data are messy and ugly.  One sort of ugliness is that instruments do not last forever.  When (not if) they fail, you have to turn to a different instrument.  Ideally, you already have the replacement in hand and have been running it regularly and intercomparing its results with your current main system and ensured that there are no differences other than those you wanted -- like better resolution on the new instrument.  The present situation is not ideal, so, as we usually do in science, I'm making the best of it that I can.  And making notes for what to do when I have a chance to rework the immediate fixes.

Read more »

Too-early consideration of sea ice estimates

I won't let it stop me, but considering that it is still August as I write, it's premature to evaluate estimates of September's sea ice extent.  On the other hand, it's fun and easy, and some parts make sense even at this early date.

Last May I made, with colleagues, some estimates for September 2011's extent.  As usual, I also included some other forecasts for comparison.  The NSIDC figure for August 29th is:

Where I've added lines in red for the 30 year mean climatology of September extents (6.67 million km^2), the mean monthly September extents if we take a 30 year linear trend (5.31), and lines at 5.0 and 4.5 million km^2.

Read more »