PIOMAS vs CryoSat

[DavidR]

Odd how the ice in the Barrents is a metre thick even on the ice edge, you'd expect or at least I expected some kind of tapering off of ice thickness.

I am suspecting that some areas shown as white are too thin to reliably measure as well as possibly also being no measurements or no ice. Even so, there is a lot of green and little blue in Southern Baffin Bay as well.

If you check the closeups or the shorter time frames you can see that the image is built from a series of single passes providing  information covering a very narrow band, so anywhere not coloured can be assumed to be not measured unless it is outside the expected area of ice.

Looking at the distribution of the ice, does make me wonder if the isolated pockets of thick ice have the same resistance to melting as the solid area of thick ice above Ellesmere island. I assume not. 

[Jim Hunt]

An extract from the data manual that may be relevant?

Sea Ice near real time thickness products are currently available in ASCII and netCDF formats in either a whole Arctic 5km resolution or as individual sectors at 1km resolution.

For 5km resolution data, a circular operator of radius 25km is applied when gridding the data and all points receive equal weight.For the 5km grids, only grid points with sea ice thickness data are included in the file (ie it is a sparse grid).

For the 1km grid all grid points north of 60N are present.For 1km resolution data, a circular operator of radius 5km is applied when gridding the data and all points receive equal weight. For the 1km grid all grid points north of 60N are present.

Here's one of the sector maps:

[slow wing]

Thanks to CryoSat for this new offering. Really interesting!

[Wipneus]

Here is PIOMAS, AMSR2-SIT and CRYOSAT compared.

[EDIT: fixed a bug in the AMSR2 volume calculation, the volume is noticeable smaller]

[Jim Hunt]

I've just finished (virtually) attending the Sea Ice Prediction Network webinar on "Observations of Arctic Snow and Sea Ice Thickness from Satellite and Airborne Surveys". It seems a recording of the webinar will be made available shortly, but in the meantime the slides can be downloaded from:

http://www.arcus.org/files/SIPN_5May_Webinar_PreliminarySlides_kurtz_1May15_opt.pdf

Here's a very brief overview. Nathan Kurtz from NASA said that IceBridge and CryoSat-2 quick look data will be available from NSIDC "next week". Snow depth will look a lot like the image below.

Slide 18 reveals that PIOMAS does benefit from assimilation of "observed" thickness information when used for the purposes of forecasting, and it seems "weather" reanalysis from MERRA can help also.

Slide 17 says that a "near real time CryoSat-2 sea ice thickness product [is] expected in Fall 2015", so I asked the obvious question. It will differ from the CPOM version by using "waveform fitting" (see slides 12-13) instead of "threshold tracking", and "dynamic" climatology (particularly for snow depth) rather than "static". It seems that getting a handle on sea ice density is currently a bigger problem that snow density.

Some papers that were mentioned:

"Seasonal forecasts of Arctic sea ice initialized with observations of ice thickness", Lindsay et al. 2012: http://onlinelibrary.wiley.com/doi/10.1029/2012GL053576/full

"Interdecadal changes in snow depth on Arctic sea ice", Webster et al. 2014: http://onlinelibrary.wiley.com/doi/10.1002/2014JC009985/abstract

"Using the interferometric capabilities of the ESA CryoSat-2 mission to improve the accuracy of sea ice freeboard retrievals", Armitage and Davidson 2014: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6479282

"An improved CryoSat-2 sea ice freeboard retrieval algorithm through the use of waveform fitting", Kurtz et al. 2014: http://www.the-cryosphere.net/8/1217/2014/

[Jim Hunt]

Here's the recent SIPN IceBridge/CryoSat webinar:



My "obvious question" is asked at 43:25 or thereabouts.

[Steven]

New paper by Tilling et al. about the 2010-2014 CryoSat measurements:

http://www.nature.com/ngeo/journal/vaop/ncurrent/full/ngeo2489.html


Clicking on this link should directly open the pdf-file of the paper (but this may not work for some browsers).  Supplementary information of the paper is here.


See also this blog post on Carbon Brief:

http://www.carbonbrief.org/blog/2015/07/cool-arctic-summer-brought-brief-recovery-in-2013-sea-ice-loss

[Jim Hunt]

See also this blog post on Carbon Brief

I've already pointed them back in this direction!

[plinius]

one thing to assess uncertainties and biases would be to have ex post corrections for PIOMAS pubished. The model will produce wrong dates for the appearance/disappearance of ice, which can be used to infer the model bias. Published anywhere?

[ChrisReynolds]

Not as far as I know Plinius.

[Jim Hunt]

Rachel Tilling suggested I was confused between CryoSat-2 and IceBridge.

http://www.carbonbrief.org/blog/2015/07/cool-arctic-summer-brought-brief-recovery-in-2013-sea-ice-loss#comment-2148344790

In actual fact I don't think I am. I get the distinct impression that UCL isn't up to speed on what NASA have been up to recently.

[Jim Hunt]

Following a frank exchange of views on Carbon Brief Rachel points out that:

We have now released our spring and autumn thickness data for the whole CryoSat-2 period on our website:

http://www.cpom.ucl.ac.uk/csopr/seaice.html?thk_period=0&select_thk_vol=select_thk&season=Autumn&ts_area_or_point=all&basin_selected=0&show_cell_thickness=0

This is the data that is presented in the paper, and has now been peer reviewed. Thanks all for your interest.

This is the "Final" data, not "Quick look".

See also all the usual BS from the usual suspects about this paper amongst others

http://forum.arctic-sea-ice.net/index.php/topic,578.msg57562.html#msg57562

I got a "like" from Ed Hawkins on Twitter

[shendric]

See my comment in Carbon Brief.
http://www.carbonbrief.org/blog/2015/07/cool-arctic-summer-brought-brief-recovery-in-2013-sea-ice-loss#comment-2148344790

There are other papers on the 2013/2014 winter season as seen by CryoSat and the conclusions are not necessarily the same.

http://rsta.royalsocietypublishing.org/content/373/2045/20140157
http://onlinelibrary.wiley.com/doi/10.1002/2015GL064081/full#grl53020-bib-0003%29

[Jim Hunt]

See my comment in Carbon Brief.

Welcome Stefan!

See also my subsequent comment over at Carbon Brief.

Thanks once again for the heads up.

[Wipneus]

Occasionally I visit the Cryosat site, to see whether new thickness/volume data for the year 2015 has been added. We are still missing volume data for March and April and the new measuring season is due to start this month.

What can be noticed is that the volume data for December (2014) and January seem to have been downgraded from precise to near real time, adding some uncertainty to the latest calculated volume data from Cryosat.

[plinius]

Cryosat measurements for the arctic are back online since 23nd of October.
http://www.cpom.ucl.ac.uk/csopr/seaice.html
Not sure how reliable that is, but if their 28-day averages are true, there would be nearly no thick ice left in front of the Canadian Archipelago, and pretty few in the central basin.

[Steven]

Cryosat tracks Arctic sea ice freeze-up
http://www.bbc.com/news/science-environment-34619291

Arctic sea-ice volume during the first two weeks of October was about 6,200 cubic km.

The number comes from Europe's Cryosat mission, which has just restarted its near-real-time data service.
...

Volume of Arctic autumn sea ice: First two weeks of October (average):

2010: 5,900 cubic km;
2011: 4,500 cu km;
2012: 4,600 cu km;
2013: 7,800 cu km;
2014: 6,800 cu km;
2015: 6,200 cu km

[diablobanquisa]

Icesat+Cryosat-2   vs. PIOMAS

(warning: it's just my own rough estimate)





Kwok and Cunningham 2015
CPOM
PIOMAS

[Steven]

Interesting article today at Carbon Brief:

http://www.carbonbrief.org/analysis-the-highly-unusual-behaviour-of-arctic-sea-ice-in-2016

The latest observations of sea ice thickness from the CryoSat satellite show that the ice was, on average, 1.8m thick in March [2016], says [Andrew] Shepherd. This is about 10cm thinner than the same time last year, but about 10cm thicker than the record winter low in 2013.

Shepherd and his team will use measurements of thickness together with sea ice extent to estimate the volume of ice left at the end of winter. These results aren’t public yet, but Shepherd tells Carbon Brief it’s not looking hopeful:

“If things continue as in previous years, I suspect this year could see a tie with that record low for volume as well, but it’s too early to say for sure.”

[diablobanquisa]

2015, 26 Sep-23 Oct



2016, 18 Sep-15 Oct

[diablobanquisa]

CS2, Oct-Nov 2011-2016 (click for a larger version):



http://www.esa.int/Our_Activities/Observing_the_Earth/Space_for_our_climate/Ice_matters

http://www.cpom.ucl.ac.uk/csopr/seaice.html

[slow wing]

Thanks diablobanquisa, that's very interesting.

Presumably they are all 28-day Thickness maps collecting measurements over 24/10 - 20/11?

So this is precise and well-calibrated for year-to-year comparisons?

It shows almost no ice thicker than 2 m at that time of year in 2011, then 2013 a big recovery year before falling away again.

This year at least has some thicker ice off the Canadian Arctic coastline but the overall extent is of course the lowest.

[seaicesailor]

Yes and at this time of the year Hycom ACNFS does not seem really off the mark wrt Cryosat, or not as much as it looked in previous years.
The various compaction periods since end of August have increased this ice accumulation north of the CAA and Greenland coasts and not much Fram export until recently. This is good for that core of ice, but its extent is rather limited. Most of the CAB and the peripheral ice is dangerously thin and young.

[Jim Hunt]

It shows almost no ice thicker than 2 m at that time of year in 2011, then 2013 a big recovery year before falling away again.

You may wish to read the paper mentioned by Stefan Hendricks above:

http://onlinelibrary.wiley.com/doi/10.1002/2015GL064081/full#grl53020-bib-0003%29

We find a positive correlation between buoy snow freeboard and CryoSat-2 freeboard estimates, revealing that early snow accumulation might have caused a bias in CryoSat-2 sea ice thickness in autumn 2013.

[diablobanquisa]

Presumably they are all 28-day Thickness maps collecting measurements over 24/10 - 20/11?

So this is precise and well-calibrated for year-to-year comparisons?

No, but until the final data for 2015 and 2016 are released by CPOM,  this is the best we have. I think it is reliable enough for the overall picture.

It shows almost no ice thicker than 2 m at that time of year in 2011, then 2013 a big recovery year before falling away again.

You may wish to read the paper mentioned by Stefan Hendricks above:

http://onlinelibrary.wiley.com/doi/10.1002/2015GL064081/full#grl53020-bib-0003%29

We find a positive correlation between buoy snow freeboard and CryoSat-2 freeboard estimates, revealing that early snow accumulation might have caused a bias in CryoSat-2 sea ice thickness in autumn 2013.

However, Ron Kwok thinks the 2013 summer recovery is real, and mainly due to ice dynamics (convergence, compaction): http://onlinelibrary.wiley.com/doi/10.1002/2015GL065462/full

"In a different analysis, Ricker et al. [2015] considered the higher MYI freeboard in November 2013 compared to March 2013 to be unlikely after the melt season and surmised that higher ice freeboards may be due to retrieval issues. Even though the retrieval issues merit attention, they neglected to consider the potential of ice convergence and deformation (discussed above) in creating such inconsistencies. Broadly, instead of attributing the ice thickness of this region solely to thermodynamics, the results here highlight the role of dynamics as a source of variability in Arctic Ocean ice thickness that should not be discounted when interpreting retrieval results."

(see also Kwok and Cunningham 2015. And Tilling et al. 2015, of course)

Cheers

[ktonine]

Comparing the Cryosat thickness map to the extent map at minima (circa first week of September), and then taking into account the temperatures that we've seen since then, it's difficult to believe any new ice has grown to 1m thickness this fall/winter.  The number of Freezing Degree Days for N80 is still only half of what's needed to grow 1m thick ice and temperatures from N70 to N80 (where most of the new ice resides) have been even warmer.

If temperatures were a uniform -10C it would take  4 months to grow 1m thick ice. For the area between N70 and N80 we've barely seen a month with temperatures below -10C. Thermodynamically it simply isn't possible for much new ice to be greater than 0.6m thick right now.

[magnamentis]

Comparing the Cryosat thickness map to the extent map at minima (circa first week of September), and then taking into account the temperatures that we've seen since then, it's difficult to believe any new ice has grown to 1m thickness this fall/winter.  The number of Freezing Degree Days for N80 is still only half of what's needed to grow 1m thick ice and temperatures from N70 to N80 (where most of the new ice resides) have been even warmer.

If temperatures were a uniform -10C it would take  4 months to grow 1m thick ice. For the area between N70 and N80 we've barely seen a month with temperatures below -10C. Thermodynamically it simply isn't possible for much new ice to be greater than 0.6m thick right now.

exactly my thoughts all the time, thickness maps are almost useless because most of the time obviously incorrect, showing 1m ice where is zero and the likes.

i boldly lean out of the window and say that the 2016 data are non-sense and close to impossible. as you say, neither temps nor conditions were the likes to grow that much several meters thick ice or piling it up that much. perhaps they measure the highest tip sticking out and apply it to a certain area.

[Tigertown]

Could be the darkness or fear of going for a cold swim accidentally, but maybe everyone is scared to walk out on the ice enough to get adequate cores so as to re-calibrate.
I mean, you can understand the surviving MYI being thick, but not new ice.

[SteveMDFP]

Could be the darkness or fear of going for a cold swim accidentally, but maybe everyone is scared to walk out on the ice enough to get adequate cores so as to re-calibrate.
I mean, you can understand the surviving MYI being thick, but not new ice.

Well, validating ice thickness is difficult, but you don't have to be on top of the ice to do so.  You can determine ice thickness quite accurately from below, with submarine instruments.  That data IS being collected as we speak, but most of it is secured in the data banks of the US and Russian navies.  Interestingly, the military of both nations have been quite pro-active in efforts to plan for such dramatic changes in their operating environments.  I suspect they may know quite a lot that we're only able to guess at.

[Jim Hunt]

Maybe everyone is scared to walk out on the ice enough to get adequate cores so as to re-calibrate.

With good reason! How would you go about manually sampling sea ice thickness Arctic wide on a regular basis? The nearest you'll get to "re-calibration" is comparison with the data acquired by the NASA Operation IceBridge flights. More from Ron Kwok et al.:

http://dx.doi.org/10.3390/rs8090713

[ghoti]

Obviously the Arctic is much too vast to sample manually. Plus
don't forget the tragic deaths of Marc Cornelissen and Philip de Roo in 2015. They were on an expedition in the northern Canadian archipelago measuring ice and observing wildlife when they fell through the ice and died.

[diablobanquisa]

Incidentally, SMOS (click for a larger version):



http://www.iup.uni-bremen.de:8084/smos/

[Tigertown]

No longer ago than early 2015 sea ice was being cored. I remember one sample off the coast of Alaska that turned out to be only one meter thick. It was not really something that had to be done Arctic wide. If you know what the exact measurement is in one place and get readings that it is 32% thicker in another place, you can calculate the depth. They measure many things with satellites using this method, not just ice.

That being said, I personally would not want to go out on the ice, especially nowadays. I do not blame anyone else for trying to use safer methods, whether by airplane, submarine, helicopter, or whatever.

[ktonine]

Not sure if there's a good/easy source to view current temps for latitudes other than N80, but Nick Stokes' at moyhu has an applet that let's you select latitude and longitude (scroll about halfway down the page for the applet).  The data and X-axis are not aligned properly --- I'll send a complaint to Nick

Here's the Pacific sector from N70 to N80:



Climatology would indicate about 890 accumulated FDDs to date
2016 appears to be around 400 accumulated FDDs

Using Lebedev new ice thickness reduces from 0.84m to 0.52m
Using Berillo new ice thickness reduces from 0.68m to 0.43m

Neither of these takes into account the actual date of first ice formation, this means they are likely biased high.

It should also be kept in mind that Cryosat has thickness uncertainties > 0.6m. 

[Archimid]

Interesting tool. Played with it for a little bit and attempted to get regional temps. It is not perfect but I like it. Attached are my best fit for the Chukchi, Barents and Kara.

Using Lebedev new ice thickness reduces from 0.84m to 0.52m
Using Berillo new ice thickness reduces from 0.68m to 0.43m

Neither of these takes into account the actual date of first ice formation, this means they are likely biased high.

It would be interesting to calculate FFD's on a regional basis and somehow apply them to the area of the region.

[Ice Shieldz]

Here is a site that provides 1 week and 1 month GFS hindcast 2 meter temp anomaly means.  As well as archives for previous months going back to 2011
http://www.karstenhaustein.com/climate.php

[diablobanquisa]

http://www.esa.int/Our_Activities/Observing_the_Earth/CryoSat/Arctic_freeze_slows_down

CryoSat volume update. Looks like tied with previous low for November...

Low, but is it better than expected with the current extent numbers and temperature anomalies?

PIOMAS was tying for the lowest last month, will be interesting to see their numbers in a few days.

[seaice.de]

CryoSat2 and SMOS give similar trends. SMOS sees the reduced new ice growth this year.

Click image below to start animation!

[diablobanquisa]

[Neven]

Thanks, Diablo. Very enlightening.

[epiphyte]

Is it even cold enough for cryosat to work right? (Serious question)

[crandles]

Is it even cold enough for cryosat to work right? (Serious question)

This measurement technique works in autumn, winter and spring. In summer, melt ponds prevent us from estimating sea ice thickness

The problem seems to be surface liquid water related rather than temperature related? So storms more of a problem than the temp being as warm as -10C ?

[epiphyte]

Is it even cold enough for cryosat to work right? (Serious question)

This measurement technique works in autumn, winter and spring. In summer, melt ponds prevent us from estimating sea ice thickness

The problem seems to be surface liquid water related rather than temperature related? So storms more of a problem than the temp being as warm as -10C ?

Right... What I should really have asked was, "Is it dry enough [] ?"

[Michael]

The Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI) has released it's CryoSat-2 Sea Ice Thickness data for October 2016 to January 2017, so it is now possible to do a proper comparison with the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) and CryoSat-2 reported sea ice thickness.

The Centre for Polar Observation and Modelling (CPOM) also produce sea ice thickness data derived from CryoSat-2 and their NRT 2 day 1km product is used here to create a monthly average.

In order to properly compare PIOMAS with CryoSat-2, each valid data point in the CPOM dataset is mapped onto the PIOMAS grid, and the corresponding value from the PIOMAS daily gridded dataset is used to create a monthly average. The intention being to sample the PIOMAS grid at the same temporal and spatial frequency as the CryoSat-2 satellite samples the Arctic sea ice and create a CryoSat-2 equivalent monthly average from the PIOMAS data.

The PIOMAS and CPOM data are regridded onto the NSIDC EASE-2 25 km grid, as used by for the AWI sea ice thickness product. The AWI colour bar is used for all three sea ice thickness products.

In the process of generating the monthly averages the opportunity is taken to create a monthly average difference between PIOMAS and CPOM sea ice thickness. To create this monthly average each valid CPOM data point that lies in an ocean grid cell on the PIOMAS gird is subtracted from the corresponding PIOMAS data point

In comparing PIOMAS with CryoSat-2 derived Sea Ice Thickness it is worth noting that PIOMAS reports Effective Sea Ice Thickness, while the CryoSat-2 derived products report Average Sea Ice Thickness. To complicate matters even further, the PIOMAS monthly average presented here is in fact Average Effective Sea Ice Thickness.

The CryoSat-2 satellite does not provided complete coverage of the Arctic in a one month period, and so there are blank areas where no data is avalaible.

Data:

http://data.seaiceportal.de/gallery/index_new.php?lang=en_US]
[url]http://data.seaiceportal.de/gallery/index_new.php?lang=en_US[/url]
http://www.cpom.ucl.ac.uk/csopr/seaice.html
http://psc.apl.uw.edu/research/projects/arctic-sea-ice-volume-anomaly/data/model_grid

Comparison of PIOMAS vS. CryoSat-2 for October 2016

[Michael]

Comparison of PIOMAS vS. CryoSat-2 for November 2016

[Michael]

Comparison of PIOMAS vS. CryoSat-2 for December 2016

[Michael]

Comparison of PIOMAS vS. CryoSat-2 for January 2017

[Neven]

Wow, Michael, this is pure genius. Thanks a lot for making these graphs, I mean comparison maps!

I could sure use such a thing for the next PIOMAS update, but I'm sure you depend on data release by either AWI or CPOM. I've asked Dr Marcel Nicolaus several times when the maps on meereisportal.de were going to be updated, but no response so far.

Edit: meereisportal.de has a http://Jan 2017 CryoSat map as well now.

[DrTskoul]

Yeah, thanks Mike for the comparison!  So that high volume PIOMAS feature has persisted for some time and is different than what Cryostat shows. I wonder what was the genesis of that feature.

[Neven]

Here's a CryoSat animation showing Jan 2016 vs Jan 2017:



Maybe it's my eyeball, but I would say that CryoSat suggests that the ice is slightly thicker right now than it was last year around this time. Which is weird, given the FDD anomaly, the cloudy weather, the storms, and PIOMAS saying there's a difference of 2374 km3 in volume and 17 cm in average thickness.

Weird...