EGU2016, my impressions

Just like last year, I had the opportunity and time to visit
the European Geosciences Union General Assembly 2016. There were several sessions on subjects related to the cryosphere and remote sensing, but most of the interesting Arctic sea ice-related stuff was happening on Thursday.
So, that's when I went to Vienna.

The day started quite strongly with a press conference called Sea ice decline in the Arctic. Dr. Alexandra Jahn from the University of Colorado talked about models and internal climate variability and Dr. Dobrynin from Uni Hamburg talked about how - as the Arctic loses more sea ice - there will be more, bigger waves eroding coasts and shallow seabeds, etc. This was all very interesting and how I expected a scientific press conference to be: about research looking at the past or the future, based on a painstaking and time-consuming process that involves lots of hypothesizing and data crunching.

However, to my surprise, yet another speaker was talking about the things we like to talk about here and on the Arctic Sea Ice Forum, things that are happening now, things that affect the melting season. Dr. Marcel Nicolaus from the Alfred Wegener Institute explained how a very warm winter resulted in less ice growth (check this previous blog post called CryoSat-2 confirms: sea ice volume is low), and as a consequence this melting season could be seeing new record lows. Depending on the weather, of course.

Robert McSweeney from Carbon Brief asked Dr. Nicolaus and several other scientists about their views on this:

So, that was pretty cool. After the press conference I had time to upload a few images to the ASIF and then went on my way to visit the oral session called Rapid changes in sea ice: processes and implications.

The session consisted of various interesting presentations.
Here's an overview:

Christopher Horvat from Harvard University talked about the Effects of the sea ice floe size distribution on ocean eddies and sea ice melting. Intuitively we all know that smaller ice floes will melt faster than a big one. Sea ice models have this cut-off size of 30 m, below which size is assumed to affect the rate of sea ice volume loss.

Horvat's PowerPoint presentation contained some really cool animations (the animated GIF above shows a part of one of them) of simulations that show how this works in theory. His conclusion is that ocean eddies and floe-edge circulation probably influence sea ice volume loss for floes that measure up to 100 km (much larger than the assumed 30 m). Read more about it on Horvat's website and his paper (under review at Geophysical Research Letters).

Up next was Anja Rösel from the Norwegian Polar Institute who talked about the State of Arctic sea ice north of Svalbard during N-ICE2015. Ice and snow thickness results from the N-ICE2015 cruise produced some interesting data, although Rösel told me later during the poster session that the winter observations were a bit boring because that series of Atlantic cyclones during the 2015/2016 winter dumped so much insulating snow on the ice that the latter hardly grew (just 15 cm in 2 months).

Unfortunately there isn't a lot of data to compare these results to, but the main goal of the expedition, of course, was to compare these ground measurements with airborne and satellite observations. Marcel Nicolaus also had a presentation on snow depth on Arctic (and Antarctic) sea ice, based on buoy data. Both Rösel and Nicolaus told me that practically the only historical snow depth data there is, is the Warren data set based on data from Soviet drifting stations on multiyear Arctic sea ice from 1954-1991 (here's the 1998 paper).

Rösel's name was familiar to me as she had done some very interesting PhD work on melt ponds using MODIS satellite images a couple of years ago (which I wrote about here). To my astonishment this important work wasn't continued due to a lack of funding, even though it shouldn't be too expensive to do, as the satellite images are already there for everyone to see.

My astonishment became even greater when I realized how important an assessment of the distribution of melt ponds during Spring can be to forecast the minimum. All we have right now, is the excellent model work done by David Schröder (see below), but it would be great to have observation-based data as well.

I managed to have a quick chat with Rösel during the break, and practically begged her to continue this important work. She told me that she was trying to get funding, as she had several ideas to expand the research and work together with other scientists who are doing similar research.

During the break, I was also lucky to be able to talk for almost half an hour with Dr. Florence Fetterer from the NSIDC. Recently I had been in contact with her about some other matters and when I saw she would attend the EGU general Assembly this year (unfortunately her presentation was scheduled for the next day), I asked her if she would tell me more about some recent work she's been doing on historical reconstructions of Arctic sea ice cover.

These reconstructions are a fascinating line of cryospheric research, in which various sources of information - satellite data, early monitoring, anecdotal evidence such as logbooks from whaling ships, etc - are pieced together to give an idea of what the Arctic sea ice pack looked like in the past. Back in January I wrote a blog post on reconstruction work that was done by fellow blogger Diablobanquisa (be sure to read the fascinating exchange in the comments with Rob Dekker), and I hope to do one on the work Dr. Fetterer has done with Walsh and Chapman, as soon as it is published later this year.

After the break the session continued with some tipping point stuff from Till J. W. Wagner, a Post-Doc at the Scripps Institution of Oceanography.
A paper he wrote last year with the same co-author, Ian Eisenman, about how Arctic sea ice loss doesn't lead to a tipping point and is thus reversible, received quite some traction in the media (I wrote about it here).

The stuff in this presentation called False alarms: How early warning signals falsely predict abrupt sea ice loss was way over my head (not that it takes much). I though about asking whether this means there is no reliable warning signal or that there is no tipping point, but I'm glad I didn't. You don't want to remove all doubt.

This was then followed by a presentation I was looking forward to: Spring melt ponds drive Arctic September ice at past, present and future climates in coupled climate simulation. The work that David Schröder from the University of Reading has done on modelling melt ponds, has featured quite heavily on this blog in the last two years, as Dr. Schröder has been kind enough to share melt pond distribution maps from his model which gave us an excellent insight into the build-up of melting momentum (see here for an explanation).

I think it's quite exceptional to receive that kind of info from a publishing scientist, and I was hoping to thank David again in person during the poster session. Unfortunately he was ill and left the congress centre after his presentation, but I was able to have short chat with Daniel Feltham, CPOM Professor of Climate Process Physics in Reading and leader of the group Dr.Schöder is a part of. He told me that their work on melt pond modelling was even more promising than they had initially thought and it might prove very useful for GCMs to forecast future Arctic sea ice loss.

One other person I had hoped to have a chat with, was Dr. Wieslaw Maslowski from the Naval Postgraduate School, quite well-known for his modelled projections of an ice-free Arctic in 2016 ± 3 years (see here and here). I wanted to ask how he viewed his model projection (especially this year, given the warm winter and subsequent events), but as he was busy talking to someone else after the oral presentation session ended, I figured I'd get an opportunity during the poster session. But alas, he didn't show up there. I guess I'd make a really bad paparazzi.

This was fully compensated, though, by the opportunity to meet Dr. Julienne Stroeve from the NSIDC, as I had been in contact with her a couple of times through the years. She immediately started telling me about the problems with the SSMIS sensor on the DMSP F17 satellite that have caused quite a few graphs and maps to go haywire since mid-April. I hope to write an in-depth article about that later this month.

The last chat I had at the poster session was perhaps one of the most interesting. Dr. Monica Ionita-Scholz from the Alfred Wegener Institute had already rolled up her poster and was about to leave, but was still kind enough to quickly explain her work to me (and even unrolled the poster again!) on the effects of Arctic sea ice loss in the Barentsz-Kara region on European winter weather:

The cold winters over Europe (low sea ice years) are associated with anomalous anticyclone and the downstream development of a mid-latitude trough, which in turn favours the advection of cold air from the north, providing favourable conditions for severe winters over Europe. We suggest that these results can help to improve the seasonal predictions of winter extreme events over Europe.

Of course, extremely low sea ice levels in the Barentsz and Kara Seas has become a regular feature now, and we did see some extreme cold weather in Europe during recent winters (2013 and 2014 come to mind). The past winter didn't see that much extreme cold in Europe,  and Dr. Ionita-Scholz agreed the effect may have been trumped by the extremely high global temperatures, due to AGW fuelled by a strong El Niño. It will be 'interesting' to see what the effect of low sea ice conditions will have in the next few winters, and if increased scientific scrutiny will lead to incontrovertible evidence of a direct AGW fingerprint.

Just like last year it was a very exciting experience to get so close to science and scientists, and learn things that help me write this blog and make some of the science available to a wider audience. I'd like to thank the European Geosciences Union for providing me with a press pass.

 

Comments

[Aaron Lewis]

Regarding Ian Eisenman,
Non linear feedback systems almost always have tipping points. See the early work of Ed Demming on industrial statistics. This has been very well validated by computer models of industrial systems.

The problem with Arctic ice is estimating what "abrupt" means. On the other hand, once the ice has tipped (gone out of control in Demming's terminology) it is not going back.

A PDF of Eisenman's paper is at http://eisenman.ucsd.edu/papers/Wagner-Eisenman-2015b.pdf

I think his problem is that his noise function needs to simulate noise from El Nino and PDO cycles.

[wayne]

The most amazing now event is of course the Beaufort Gyre persistent Anticyclone, perhaps going on 10 weeks, I lost count, picked up lots of data though, just like 2008 its been sunny every day except for one or two near huge cyclone encroachment. I deal with this now event and its future course :

http://eh2r.blogspot.ca/2016/04/2016-annual-spring-projection-made-by.html

[Yvan Dutil]

Dr. Rösel should go either to croudfunding (Experiment like ) or croudsourcing (Zooniverse like). Her research project is well suites for this type of approach.

[Neven]

Definitely, Yvan. The last word about this hasn't been said yet.

By the way, tomorrow a great workshop starts with a huge amount of interesting presentations: Polar Prediction Workshop.

The presentations are going to be streamed and the stream will be available here.

Things start tomorrow at 9:00am EDT, which is 15:00 o'clock CET, I believe.

I hope to catch some of those presentations. Nice follow-up to the stuff I saw at EGU.

[Avocadocore]

Love your work Neven. I've been enjoying it, and the comments, for a while now. Be sure to let me know when Dr. Rösel has the crowdfunding going - seriously, I created an account just to say this...

[D]

Well, Eisenman's model is getting it's big ENSO kick. I would not be surprised if this El Niño is a tipping point for the northern hemisphere's climate but I don't think we have a good enough systems model to make a good prediction. To model what's happening you need to go from the deep ocean where deepwater forms to the 10mb level of the stratosphere where upward propagating waves trigger sudden stratospheric warmings. It's all linked together in non-linear ways I have trouble visualizing. And I'm just talking about what has been happening this year.

Wayne, according to the GFS the big high is going to slam the Canadian Arctic in the next 2 weeks. The big melt is going to get worse as above freezing temperatures and offshore winds hit some of the thickest ice in the Arctic.

This is going to be a year to remember.

[wayne]

Hi D,

Is like living in the desert except the sand is frozen water, never ending blue skies, worse than 2008. Surface temperatures are more than 7 C above average, soon to be 12 C above. This is not good for Greenland. Sea ice is colder than this progression in surface temperatures, it lags behind, being a heat sink during the day and a big mirage maker during the midnight sun. There is some fog, but it is fleeting, the air is too dry by a few percentage points, just enough to make the fog vanish.

[Kris]

Undoubtly due to global warming and an over heated winter in Canada & Alaska:

The 80.000 inhabitants of Fort McMurray [province of Alberta - Canada] had to be evacuated .

[Rob Dekker]

Please remember that Fort McMurray is the capital of the environmentally so destructive Canadian Tar Sand Industry.

Somehow I find it ironic, that the capital of Tar Sands is being destroyed by wild fires spurred by global warming...

Sort of like Nature strikes back, sort of thing....

[Colorado Bob]

Rob Dekker

And Houston, the "Energy Capital of the World" with a foot of rain just the week before.

Fort McMurray, Alberta
Almanac for Yesterday
Actual Average * Record
Temperature
High 89 °F 52 °F 77 °F (2006)
Low 45 °F 28 °F 17 °F (2002)

https://www.wunderground.com/q/zmw:00000.1.71585

[Artful Dodger]

Regarding Fort McMurray wildfires:

88,000 people have been displaced. Many have no homes to return to. 1,600 homes and other structures have been destroyed as of Noon Thursday (local time).

BMO (one of the largest Charter Banks in Canada) initial estimate of Insurance claims is $9 Billion.

This represents over 35 years worth of civic development up in smoke.

Ironically, oil production facilities North of Fort McMurray were unaffected by the fires. It's just the people that lost everything.

No doubt Workers will be bussed in from refugee centres in the South when oil production resumes.

Gaze upon the stark face of global corporatism: privitize profit and publicize (climate) risk.

Okay, now we should "Feel the Bern". But Investment Bankers and the 1% howl incessently about their "stranded assests".

SMDH.

Regards,
Lodger

[iceman]

Thanks for advocating for observation-based melt pond research, Neven. We really need that early-warning system. I like Yvan Dutil's suggestion of crowdfunding, but am not sure there's sufficient interest outside this and other like-minded communities. Maybe the wealthy homeowners of Miami Beach and Mumbai will kick in?

[Rob Dekker]

In my past projections of Arctic Sea Ice for September, I used snow cover data from Rutgers, and most importantly, NSIDC data for ice-extent and ice-area.

Now with the F17 trouble, NSIDC is not issuing any area and extent numbers, which make accurate projections, much harder to do.
In fact, without area and extent numbers that have a long term calibration history, predictions are baseless.

So I have a request :
Does anyone know where I can get my hands on any (AMSR2 or other) data of ice extent and area (or concentration) that has an as long as possible history ?

Something that could replace NSIDC's area and extent data ?

[Rob Dekker]

And since Dr. Slater has been working on similar projections using ice concentration as a predictor from Sept. minima, does anyone know if Dr. Slater is still doing his projection work this year, given the F17 satellite issues ?

[Artful Dodger]

Hi Rob,

The AMSR/2

[Artful Dodger]

[pardon the interruption] The AMSR/2 raw data is multi-GB per day, and is not easily accessible w/o complex processing. I suggest that you discuss your research objectives with Wipneus, who already has a plastic straw sipping from the fire hose of data. ;^)

Regarding F17 data, NSIDC put out this Press Release 2 days ago:

Due to the compromised data integrity with the DMSP F17 vertically polarized 37 GHz channel (37V) of the Special Sensor Microwave Imager and Sounder (SSMIS), which is a primary channel used for sea ice development, NSIDC is halting production of the Near-Real-Time DMSP SSMIS Daily Polar Gridded Sea Ice Concentrations F17 data and switching over to the F18 satellite, starting with 1 April 2016 data. However, NSIDC is still working on the calibration of the F18 data so temporarily the F18 data will be uncalibrated and using F17 tie points.

Cheers,
Lodger

[Rob Dekker]

Neven, on your graph page, you list this wonderful (daily) northern hemisphere snow cover graph from star.nesdis.noaa.gov.
However, do you know if they present that NH snowcover data in a tabular form somewhere ?
Somehow, I only seem to be able to get some graphs, but no numerical data on snow cover.

[Rob Dekker]

Thanks Lodger,
Yes, I'd love to connect with Wipneus, regarding AMSR2 concentration data.

Regarding NSIDC and the F17 issues, thanks for the update. I hope they can re-calibrate to F18 well before the 2016 melting season gets on its way. It's going to be an epic melting season, and NSIDC's data is desperately needed to provide some calibrated historical perspective, to say the least....

[Neven]

Sorry, Rob, don't know!

I'm going to be off for a week, but will probably/hopefully be able to connect to the Internet (and post the PIOMAS update).

[Rob Dekker]

Some good news : NSIDC resumed their ice-extent reporting graph using F18 data:
http://nsidc.org/data/seaice_index/images/daily_images/N_stddev_timeseries.png

with the following notes :

NSIDC has obtained data from the DMSP F-18 satellite and is in the process of intercalibrating the F-18 data with F-17 data. Intercalibration addresses differences between the series of sensors, in order to provide a long-term, consistent sea ice record. While this work continues, we are displaying the uncalibrated F-18 data in the daily extent image. The daily time series graph shows F-17 data through March 31, and F-18 data from April 1 forward. Initial evaluation of the uncalibrated F-18 data indicates reasonable agreement with F-17, but the data should be considered provisional and quantitative comparisons with other data should not be done at this time.

Because these are provisional data, the Sea Ice Index has not been updated and continues to display only F-17 data through March 31. We expect to make the F-18 data available in Charctic soon.

Looks like they are getting back on track !

[Rob Dekker]

Regarding Dr. Slater's projections, his 2015 projection page here :
http://cires1.colorado.edu/~aslater/SEAICE/
states

The 2015 Melt Season is over; come back in 2016!

It's 2016 now.

And since Dr. Slater was spot-on with his 2015 projection (4.55+/-0.35 million square kilometers)
https://nsidc.org/about/monthlyhighlights/2015/10/evaluating-arctic-sea-ice-predictions
we are REALLY interest in his 2016 projections...

[Rob Dekker]

Dr. Slater updated his site with the following note :

Due to SSMI sensor issues this forecast is suspened until data streams resume

Which confirms what we already suspected : That Dr. Slater's method (just like many other ice projector methods, including my own) relies on the NSIDC data which is currently hampered by the F-17 trouble.

NSIDC is working hard to resolve that issue (by switching to F-18) and we are looking forward to a resumption of Dr. Slater's insightful projection method once that transition is complete.

[Rob Dekker]

Andrew Slater's site is up and running (hat tip to EgalSust and MikeAinOz at the forum).

http://cires1.colorado.edu/~aslater/SEAICE/

The projections are impressive and scary :
For starters, Dr. Slater projects 7.42 M km^2 by July 15. This is while 2012 at that date was at about 8 M km^2. So he projects that 2016 NOT show any significant 'stall' and will maintain its lead over 2012 (by some 600 k km^2) at least until halfway July.