The good news and the bad news about z

[Vergent]

We have many real time tools to keep track of the x and the y of Arctic ice(JAXA, CT, MASIE, NSIDC, etc.). Here to fore, we have had no real time observation based thickness product. The good news is that this is no longer the case. TOPAZ4 V2 thickness has been forced by observation since June. Every Monday the daily thickness map makes jumps in areas where data has been acquired. By now, the entire ice area is tied to observational data no more than a few weeks old. How do they do it? So what is TOPAZ4?

The operational TOPAZ4 Arctic Ocean system uses the HYCOM model and a 100-member EnKF assimilation scheme. It is run daily to provide 10 days of forecast (one single member) of the 3D physical ocean, including sea ice; data assimilation is performed weekly to provide 7 days of analysis (ensemble average).

EnKF, or Ensemble Kalman filter, is a statistical filter to harvest data and variance from an ensemble of inputs.

http://journals.ametsoc.org/doi/full/10.1175/1520-0493%281998%29126%3C1719%3AASITEK%3E2.0.CO%3B2

How do I know they are measuring thickness. First because they say so. Topaz is tied to an ansemble of five altimeters(five times the data of CRYOSAT2).

http://www.myocean.eu/web/45-sea-ice.php

And second because of the animation weekly jumps in thickness(on Mondays), corresponding to weekly import of data. A GODIVA2 portal where you can do animations is at the following link;

http://thredds.met.no/thredds/godiva2/godiva2.html?server=http://thredds.met.no/thredds/wms/topaz/dataset-topaz4-arc-myoceanv2-be#

Well, how accurate is it? I do not know, they have not published a validation paper yet, but now that August PIOMAS is out, we can compare. PIOMAS average thickness for August was 1.34m, so if we take away all the mid August ice 1.34m or thinner , How much ice disappears?




I tried 1.44m and 1.24m, in the former significantly more than half disappeared, much less than half disappeared in the latter. So, there is probably less than 10cm difference between them in median thickness.

Okay, what is the bad news? Over on the blog no one seemed alarmed by the PIOMAS numbers, everyone seemed to think that 3013 has some aspect of bounce back or recovery.





What is happening to the thickness/volume of ice in the western CAB along Greenland and in the CAA? It has lost about 1.5m of thickness since 2011, mostly this year. It is now too thin to thicken by plastic distortion from lateral pressure, it will form pressure ridges, but the ice cap probably will not thicken past the equilibrium point. The CAB losses should continue next year. There is not much left to lose.

Vergent

[jdallen]

I've had, and aired my suspicions about volume for some time.  The retrograde comparisons look very good, which speaks well to their current estimates.  I think the large data ensemble helps.

I will be watching their publications with interest.

[ChrisReynolds]

Vergent,

Your second link, about altimeters, has a link for the word altimeter - but that link takes you back to the homepage.

This year is clearly not a rebound from 2012, anyone saying so is wrong. It's all weather driven.

[Vergent]

jdallen,

Thanks for your response.

I too am waiting for a validation paper.

Chris,

I share your frustration. They say they are measuring thickness with altimeters, but documentation is sparse. I come to the conclusion that they are actually doing it only from the convergence of their claim that they are doing it with the weekly jumps in thickness that indicate data import, and, of course, the consistency of the data.

I have found no documentation on their methodology. I have found documentation of their intent.

http://150.229.66.66/staff/pxs/wmoda5/Oral/Bertino_Sakov.pdf

http://www.sciencedirect.com/science/article/pii/S1463500313000966

http://www.myocean.eu/automne_modules_files/pscientifpub/public/r149_16_melsom_msd_2010.pdf

http://www.ocean-sci.net/8/633/2012/os-8-633-2012.html

Abstract. We present a detailed description of TOPAZ4, the latest version of TOPAZ – a coupled ocean-sea ice data assimilation system for the North Atlantic Ocean and Arctic. It is the only operational, large-scale ocean data assimilation system that uses the ensemble Kalman filter. This means that TOPAZ features a time-evolving, state-dependent estimate of the state error covariance. Based on results from the pilot MyOcean reanalysis for 2003–2008, we demonstrate that TOPAZ4 produces a realistic estimate of the ocean circulation in the North Atlantic and the sea-ice variability in the Arctic. We find that the ensemble spread for temperature and sea-level remains fairly constant throughout the reanalysis demonstrating that the data assimilation system is robust to ensemble collapse. Moreover, the ensemble spread for ice concentration is well correlated with the actual errors. This indicates that the ensemble statistics provide reliable state-dependent error estimates – a feature that is unique to ensemble-based data assimilation systems. We demonstrate that the quality of the reanalysis changes when different sea surface temperature products are assimilated, or when in-situ profiles below the ice in the Arctic Ocean are assimilated. We find that data assimilation improves the match to independent observations compared to a free model. Improvements are particularly noticeable for ice thickness, salinity in the Arctic, and temperature in the Fram Strait, but not for transport estimates or underwater temperature. At the same time, the pilot reanalysis has revealed several flaws in the system that have degraded its performance. Finally, we show that a simple bias estimation scheme can effectively detect the seasonal or constant bias in temperature and sea-level.

Bold mine.

So, they say they are measuring thickness with altimeters, they say data import improves the accuracy, they say they are importing data on a weekly basis, and the data field jumps on a weekly basis. I am simply connecting the dots and assuming they are not prevaricating.

Vergent

edit; sp

[Shared Humanity]

Vergent,

Your second link, about altimeters, has a link for the word altimeter - but that link takes you back to the homepage.

This year is clearly not a rebound from 2012, anyone saying so is wrong. It's all weather driven.

And doesn't this 9/9/2013 TOPAZ4 ice image show the cumulative effects of the relatively vigorous weather of 2013 on the ice, going back to the late winter cyclones that fractured and moved a large portion of the multi-year ice into the Beaufort? We seem to have a widely dispersed, thin and fragile ice cap. This matches all of the visual images we have been looking at all summer, with rounded flows surrounded by slush throughout the CAB.

The Arctic ice cap seems to have moved into an entirely new state of high mobility. It will continue to be pushed and pulled by normal cyclonic activity. Does this new state create challenges for the existing tools used to measure and assess the state of the ice? Does this new state also introduce an entirely different set of feedbacks not previously present?

I believe this approaching freeze season may be very exciting. And I would like to thank all of you for providing me insights into the Arctic that I did not have a year ago.

[ChrisReynolds]

Vergent,

I suspect they're referring to IceBridge flight transects, as there is no current satellite system aside from CryoSat 2. ICESat ran from 2003 to 2009, and ICESat 2 isn't due for launch for some years.

To explain my annoyance at talk of recovery. Probably the key point is what happened to PIOMAS volume over the spring.

Day        2012    2013   Delta
31-Mar   21.67   21.61  -0.05
30-Apr   21.57   21.27   -0.30
31-May  18.19   19.09   0.90
30-Jun   11.28   13.00   1.72

In 2008 ice volume was above the same period in 2007 throughout the winter and spring - due to a rebound. That was not the case this year, it was only from May that 2013 fell behind 2012 - positive 'deltas' in the table above (delta = 2013 - 2012). There was no rebound, there was weather. I've gone into this in more detail at my blog.
http://dosbat.blogspot.co.uk/2013/09/summer-2013.html

[Vergent]

Vergent,

I suspect they're referring to IceBridge flight transects, as there is no current satellite system aside from CryoSat 2. ICESat ran from 2003 to 2009, and ICESat 2 isn't due for launch for some years.

Chris,

Cryosat2 and ice sat are not the only satellite altimeters, they are the ones whose primary mission was ice.

The altimetry data used for assimilation are the along-track SLA from TOPEX/Poseidon, ERS1, JASON-1, JASON-2, ENVISAT

http://www.ocean-sci.net/8/633/2012/os-8-633-2012.pdf

From 4.2 observations.

So you see they have ensemble of 5 satellite altimeters. They do not use Cryosat2 because they need it for calibration and verification(I assume).

http://i.imgur.com/S8I0yEs.png

http://i.imgur.com/WWLU5Zn.png

http://i.imgur.com/Rhr2YK3.png

Here are Saturday, Sunday, and Monday respectively. You can see that there is a huge data import on Monday. It makes sense that a lot of data will get through the filters when the melt ponds are frozen, yet there is a lot of open water among the ice.

Vergent

[Peter Ellis]

In 2008 ice volume was above the same period in 2007 throughout the winter and spring - due to a rebound. That was not the case this year, it was only from May that 2013 fell behind 2012 - positive 'deltas' in the table above (delta = 2013 - 2012). There was no rebound, there was weather.

I think that's a semantic rather than a scientific debate. One could equally well describe both 2008 and 2013 as rebounds, just occurring for different reasons and at different times of year.

[ChrisReynolds]

Peter,

I don't think it's semantic, it's an issue of mechanism. The 2008 rebound is arguably due to ice growth in autumn/winter. The 2013 state did not recover to above the same time in 2012 over winter, it was only when weather (cold in May) intervened that 2013 fell behind 2012.

Unless there is some weather linkage with the ice involving cold springs, it seems difficult to argue that the 2013 'recovery' is due to the same process that causes the one year lag negative autocorrelation.