PIOMAS vs CryoSat

[Neven]

Here's a comparison of January 2012-2017:



CryoSat-2 clearly has Jan 2012 and 2013 thinner than Jan 2017. But PIOMAS has this:



Difference with 2012 and 2013 at the end of January is 2270 and 1571 km3 respectively. Okay, CryoSat is showing an average, so let's the take the average of the differences with 2017 at Jan 1st and 31st: 2118 and 1193 km3 respectively.

[DrTskoul]

Weird indeed.

I know we have discussed about it before but Wlwhat is Cryosat's skill in distinguishing snow?

[epiphyte]

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...

My hand-waving hypothetical explanation of this is as follows:

1. Cryosat 2 is at it's heart a radar altimeter.
2. Space-based SAR (non-military, at least) has a resolution of at best 10s of meters.
3. Thin, soft ice is more frangible than thick, hard ice.
4. Broken-up ice is more mobile than sheet ice.
5. unstable weather causes ice drift to change direction more than the historical high-pressure, temperature-inverted arctic regime.

--> 3,4 and 5 in combination create a low-density ice field with high-granularity variation in surface height above sea-level. (6)
--> 1 and 2 in combination make it impossible for cryosat to distinguish between  (6), above, (which has low density, volume) and solid sheet ice. (which has high density, volume) (7)

... so my SWAG on this is that under this year's conditions (i.e. characterized by 3,4 and 5 above) Cryosat is likely on the high side in many areas. This might be testable if there's a higher-resolution radar than Cryosat (S1A/B?) available for comparison. Areas which show up on the latter as uniformly reflective (i.e. dark or light depending on the angle) should show better Cryosat-PIOMAS correlation than the noisier regions.

Not saying that PIOMAS is right either, mind you     (Have been pointedly not saying that since pretty much my first post on this forum!)

[S.Pansa]

... what is Cryosat's skill in distinguishing snow?

I was wondering about that too. The PIOMAS numbers did make a lot of sense to me ...

For what it is worth, below the snow thickness for late January 2012, 2013 and 2017 - according to TOPAZ4.
On the first glance the shape of the  2017 snow cover could partly explain the band of relatively thick ice (~3 m)  from North Greenland to the ESS in Cryosat. But then again, there was more snow in 2012 (and even in 2013). 
If anything, the snow cover could perhaps explain the odd fat bulge near the pole in piomas' estimates for 2017. To sum it up: I have no clue what's going on 
 

[seaicesailor]

Here's a comparison of January 2012-2017:



CryoSat-2 clearly has Jan 2012 and 2013 thinner than Jan 2017. But PIOMAS has this:



Difference with 2012 and 2013 at the end of January is 2270 and 1571 km3 respectively. Okay, CryoSat is showing an average, so let's the take the average of the differences with 2017 at Jan 1st and 31st: 2118 and 1193 km3 respectively.

There is a problem with the palette used in these maps, in that it is easier to distinguish ice from 2 to 4 m than from 0 to 2 m. Very difficult to compare the main extensions of ice which are under 2m !

Regarding Michael's comparison, to me it looks that for whatever reason Cryosat and PIOMAS best agree in FYI zones with independence of snow.

[seaicesailor]

Why does 2017 map say Cryosat "NRT"? Is it maybe an enhanced version? That might invalidate comparison (just saying)

[Neven]

There is a problem with the palette used in these maps, in that it is easier to distinguish ice from 2 to 4 m than from 0 to 2 m. Very difficult to compare the main extensions of ice which are under 2m !

I agree. I wish I could do more than eyeball. I don't even know how to count pixels.

I'll just wait for someone to come up with a graph of CryoSat-2 data for January, or look for the data myself. But even my eyeball can see that CryoSat considers Jan 2012 and 2013 to have lower volume than Jan 2017.

[shendric]

Why does 2017 map say Cryosat "NRT"? Is it maybe an enhanced version? That might invalidate comparison (just saying)

NRT refers to the near real time data stream from ESA. January thicknesses based on the regular data will only be available at the beginning of March. The difference in sea ice thickness between both Cryosat input products is marginal though (see: http://www.the-cryosphere.net/10/2003/2016/).

[shendric]

Weird indeed.

I know we have discussed about it before but Wlwhat is Cryosat's skill in distinguishing snow?

Ku-Band radar (CryoSat) is sensitive to snow grain size. A larger anomaly in the snow microstructure (depth hoar, ice lenses) may result in too high freeboard/thickness values.

[diablobanquisa]

There is a problem with the palette used in these maps, in that it is easier to distinguish ice from 2 to 4 m than from 0 to 2 m. Very difficult to compare the main extensions of ice which are under 2m !

 

[CognitiveBias]

Now, that's funny!

[Neven]

In another thread crandles remarked the following:

However, I have what is a rather different quibble. We seem to be saying FDD is low this year which will result in thinner ice. However there is a serious problem of cause and effect. At the other extreme, we could say there is less volume ie thin ice so heat transfers from ocean to atmosphere faster and therefore the atmosphere is warmer. So the low FDD could be a consequence of low volume and as there is fast transfer of heat from ocean to atmosphere, the volume of ice could be catching up to where it normally is. It isn't going to be one extreme or the other but some combination. Is it possible that the low volume and increased storminess could this year have altered the balance between the two causes making it more a case of ocean warming atmosphere. Then would using low FDD as a reason for low volume be giving a false indication of particularly low volume? Perhaps Cryosat2 data shouldn't be dismissed in favour of PIOMAS quite so readily?

And maybe there is more heat going through the ice because waters are warmer, either caused by absorption and mixing (during the melting season and fall), like Hyperion proposes in the other thread. Or by a heat pulse brought in by ocean currents.

What speaks against this (heat coming out of the Arctic Ocean through the ice), is that we have seen all those lows bringing in heat (and moisture) from the Atlantic. This would also imply that there has been more snowfall on the ice and maybe that's what causes the difference between PIOMAS and CryoSat.

Because after the 2012 melting season there also was a lot of thin ice, but the Arctic didn't heat up nearly as much as now during the 2012/2013 freezing season. Quite the contrary, if my memory doesn't fail me.

Snow might explain the difference between CryoSat and PIOMAS. I've asked one scientist about snow depth data around the North Pole, compared to previous years, but the sensors failed quite quickly this year.

[diablobanquisa]

I think that the difference could be due to ice dynamics (drift, compaction).

I believe that PIOMAS is good at modelling the thermodynamics that affect the ice thickness , but maybe it is not so good at modelling the changes in thickness caused by drift, compaction and ridging of the ice.

 

[crandles]

I've been thinking about that too - but came to the conclusion that if the ocean is warming the atmosphere more than heretofore, it is symptomatic of something new - and none of the possibilities I can think of seem ultimately conducive to higher than expected volume. viz:

 - thinner ice causing more ocean heat to be lost to the atmosphere. No volume increase there.
 - less snow cover causing greater heat loss and more freezing? Yes, but it's been stormy, so why would there be less snow? Also, the storms would seem to imply greater mixing with the warmer southern latitudes.
 - halocline breakdown. Ouch.

>"- thinner ice causing more ocean heat to be lost to the atmosphere. No volume increase there."

My reaction is depends on circumstances: If it is thinner ice allowing more heat to be lost then rate of heat loss is a good indicator of ice mass being formed. However if it is stormy weather stirring up heat from below then it is much less of a good indicator of ice mass formed.

Cause and effect problems are complicated and I am not at all sure I am getting my head around it all correctly. I have no real reason to think PIOMAS is not doing it all correctly. Nevertheless with the FDD total being so weird, I think we should be cautious and not go jumping too far ahead to too many conclusions.

[epiphyte]

I've been thinking about that too - but came to the conclusion that if the ocean is warming the atmosphere more than heretofore, it is symptomatic of something new - and none of the possibilities I can think of seem ultimately conducive to higher than expected volume. viz:

 - thinner ice causing more ocean heat to be lost to the atmosphere. No volume increase there.
 - less snow cover causing greater heat loss and more freezing? Yes, but it's been stormy, so why would there be less snow? Also, the storms would seem to imply greater mixing with the warmer southern latitudes.
 - halocline breakdown. Ouch.

>"- thinner ice causing more ocean heat to be lost to the atmosphere. No volume increase there."

My reaction is depends on circumstances: If it is thinner ice allowing more heat to be lost then rate of heat loss is a good indicator of ice mass being formed. However if it is stormy weather stirring up heat from below then it is much less of a good indicator of ice mass formed.

Cause and effect problems are complicated and I am not at all sure I am getting my head around it all correctly. I have no real reason to think PIOMAS is not doing it all correctly. Nevertheless with the FDD total being so weird, I think we should be cautious and not go jumping too far ahead to too many conclusions.

Concur on the first - even more so in the absence of snow cover - which is why I think beaufort might end up putting a fight this year by reaching near-normal volume. What I should have said to be clear was "no year-on-year volume increase there"

On the second , things in many areas are so different from normal that I see no reason to think that PIOMAS _is_ doing it correctly!

[seaicesailor]

There is a problem with the palette used in these maps, in that it is easier to distinguish ice from 2 to 4 m than from 0 to 2 m. Very difficult to compare the main extensions of ice which are under 2m !

 

Thank you Diablobanquisa.

[Neven]

I've been able to look some more into this, with the help of Michael, who sent me the data that allowed me to create this graph:



Michael sent me the CryoSat-2 data from AWI (CryoSat-2 average volume on a combined PIOMAS-AWI grid), and I simply added the daily PIOMAS volume data for January and then divided by 31 to get an average as well.

As you can see the trend lines more or less move the same way, except for this year. According to CryoSat-2 there is slightly more volume now than last year and so the trend line goes up a bit. For PIOMAS the trend line crashes.

The same thing happens in December:



As shendric said:

Ku-Band radar (CryoSat) is sensitive to snow grain size. A larger anomaly in the snow microstructure (depth hoar, ice lenses) may result in too high freeboard/thickness values.

This is most probably the reason for this enormous divergence, as there have been so many Atlantic storms hurled into the Arctic this winter.

Here are a couple of quotes from Robert Ricker's PhD thesis paper. Abstract:

Radar altimetry measurements of the current satellite mission CryoSat-2 show an increase of Arctic sea-ice thickness in autumn 2013, compared to previous years but also related to March 2013. Such an increase over the melting season seems unlikely and needs to be investigated. Recent studies show that the influence of the snow cover is not negligible and can highly affect the CryoSat-2 range measurements if it is assumed that the main scattering horizon is given by the snow-ice interface. Our analysis of Arctic ice mass-balance buoy records and coincident CryoSat-2 data between 2012 and 2014 adds observational evidence to these findings. Linear trends of snow and ice freeboard measurements from buoys and nearby CryoSat-2 freeboard retrievals are calculated during accumulation events. 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.

Look at the December graph where you can clearly see a similar jump (relative to PIOMAS) in 2013.

More quotes:

Moreover, backscatter from both interfaces superimpose each other and cause broadened radar returns, which is largest for snow depths >20 cm (Kwok, 2014). As a result, freeboard estimates can be biased high with the presence of thick snow layers.

(...)

For wet snow at the beginning and the end of the melting season, the dielectric properties of the snow layer might even limit the physical penetration of radar waves.

(...)

We hypothesize that the snow cover significantly affects the CS-2 freeboard retrieval by snow backscatter which would affect also sea-ice thickness and volume, independently of the range retrieval method.

(...)

We find high differences of up to 45 cm (30 cm) for the 40 % threshold retrieval and up to 30 cm (20 cm) for the 80 % threshold retrieval from the comparison between November 2013 and 2012 (November 2013 and March 2013) north of Canada.

(...)

It is still difficult to quantify the snow-scatter induced bias without knowledge of the regional distribution and temporal evolution of snow depth and snow stratigraphy. Snow, accumulated early, may undergo a partial melting and subsequent freezing as well as wind compaction. This leads to a very heterogeneous snow density distribution, while for the propagation of the Ku-band signal it is widely assumed that the snow density is homogeneous. In this way formed layers may affect the location of the main reflecting horizon.

(...)

 We conclude that snowfall can have a significant impact on CryoSat-2 range measurements and therefore on ice freeboard, thickness and volume. The assumption that the CryoSat-2 main scattering horizon is given by the snow-ice interface cannot be justified in regions with a thick snow layer.

My preliminary conclusions:

- PIOMAS has it more right than CryoSat-2, although probably underestimating thickness slightly.
- The discrepancy is caused by either a thick snow layer, or short melt events due to heat incursions changing the snow stratigraphy, or a combination of both.
- Bad news for the ice, as 1) snow insulates, causing the ice to thicken less, 2) snow melts more easily than ice, which can set off feedback processes earlier (melt ponding, etc), especially if it has already melted for short periods during the winter-spring transition (Stroeve published a paper about this last year).

Questions:

- Is there any observational data (buoys, atmospheric data) that enables us to quantify this, or...
- Give an idea of which areas are affected most? Unfortunately PIOMAS seems to be experiencing a problem (that bulge of thick ice hovering over Fram Strait), so I don't know how useful a regional breakdown would be.

Either way, this is pretty big, IMO, as it tells us something about snow depth on the sea ice which may have consequences for the state in which the ice pack enters the melting season.

I might write about it on the blog, or mention it in the next PIOMAS update, but I thought I'd share it here first. Maybe together we can squeeze more out of this.

[Jim Williams]

I've been able to look some more into this, with the help of Michael, who sent me the data that allowed me to create this graph:



Michael sent me the CryoSat-2 data from AWI (CryoSat-2 average volume on a combined PIOMAS-AWI grid), and I simply added the daily PIOMAS volume data for January and then divided by 31 to get an average as well.

As you can see the trend lines more or less move the same way, except for this year. According to CryoSat-2 there is slightly more volume now than last year and so the trend line goes up a bit. For PIOMAS the trend line crashes.

The same thing happens in December:



As shendric said:

Ku-Band radar (CryoSat) is sensitive to snow grain size. A larger anomaly in the snow microstructure (depth hoar, ice lenses) may result in too high freeboard/thickness values.

This is most probably the reason for this enormous divergence, as there have been so many Atlantic storms hurled into the Arctic this winter.

Here are a couple of quotes from Robert Ricker's PhD thesis paper. Abstract:

Radar altimetry measurements of the current satellite mission CryoSat-2 show an increase of Arctic sea-ice thickness in autumn 2013, compared to previous years but also related to March 2013. Such an increase over the melting season seems unlikely and needs to be investigated. Recent studies show that the influence of the snow cover is not negligible and can highly affect the CryoSat-2 range measurements if it is assumed that the main scattering horizon is given by the snow-ice interface. Our analysis of Arctic ice mass-balance buoy records and coincident CryoSat-2 data between 2012 and 2014 adds observational evidence to these findings. Linear trends of snow and ice freeboard measurements from buoys and nearby CryoSat-2 freeboard retrievals are calculated during accumulation events. 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.

Look at the December graph where you can clearly see a similar jump (relative to PIOMAS) in 2013.

More quotes:

Moreover, backscatter from both interfaces superimpose each other and cause broadened radar returns, which is largest for snow depths >20 cm (Kwok, 2014). As a result, freeboard estimates can be biased high with the presence of thick snow layers.

(...)

For wet snow at the beginning and the end of the melting season, the dielectric properties of the snow layer might even limit the physical penetration of radar waves.

(...)

We hypothesize that the snow cover significantly affects the CS-2 freeboard retrieval by snow backscatter which would affect also sea-ice thickness and volume, independently of the range retrieval method.

(...)

We find high differences of up to 45 cm (30 cm) for the 40 % threshold retrieval and up to 30 cm (20 cm) for the 80 % threshold retrieval from the comparison between November 2013 and 2012 (November 2013 and March 2013) north of Canada.

(...)

It is still difficult to quantify the snow-scatter induced bias without knowledge of the regional distribution and temporal evolution of snow depth and snow stratigraphy. Snow, accumulated early, may undergo a partial melting and subsequent freezing as well as wind compaction. This leads to a very heterogeneous snow density distribution, while for the propagation of the Ku-band signal it is widely assumed that the snow density is homogeneous. In this way formed layers may affect the location of the main reflecting horizon.

(...)

 We conclude that snowfall can have a significant impact on CryoSat-2 range measurements and therefore on ice freeboard, thickness and volume. The assumption that the CryoSat-2 main scattering horizon is given by the snow-ice interface cannot be justified in regions with a thick snow layer.

My preliminary conclusions:

- PIOMAS has it more right than CryoSat-2, although probably underestimating thickness slightly.
- The discrepancy is caused by either a thick snow layer, or short melt events due to heat incursions changing the snow stratigraphy, or a combination of both.
- Bad news for the ice, as 1) snow insulates, causing the ice to thicken less, 2) snow melts more easily than ice, which can set off feedback processes earlier (melt ponding, etc), especially if it has already melted for short periods during the winter-spring transition (Stoeve published a paper about this last year).

Questions:

- Is there any observational data (buoys, atmospheric data) that enables us to quantify this, or...
- Give an idea of which areas are affected most? Unfortunately PIOMAS seems to be experiencing a problem (that bulge of thick ice hovering over Fram Strait), so I don't know how useful a regional breakdown would be.

Either way, this is pretty big, IMO, as it tells us something about snow depth on the sea ice which may have consequences for the state in which the ice pack enters the melting season.

Neven, this is worthy of a blog post, if you wouldn't mind.

[crandles]

Great post Neven.

Yes, thick snow and melt event causing various layers in the snow sounds much more likely than my don't get too carried away with FDD causing low volume because that might be cause and effect wrong way around.

Might layered snow be a more effective insulator as there is little scope for convection though snow layers?

Also snow layers could perhaps limit the scope for thick snow to remain dry on top by water trickling downwards so perhaps thick layered snow does loose its high albedo quickly?

That would be a bad combination but 2014 melt season was poor. However the Dec 2013 mismatch had corrected by Jan 2014 so perhaps we shouldn't expect a poor 2014 melt season.

2017 it hasn't corrected by Jan. Maybe it might correct itself by Feb or March?

[charles_oil]

Is it possible that the snow is a significantly different grade / quality this year when the weather has been warmer than even last year ?   Thinking fine powder vs coarse "wetter" snow...

[Neven]

I think there's simply more snow to start with. And then, on top of that, it may be that the snow's structure is such that it throws the CryoSat-2 radar waves off.

I want to look at SLP maps for all they months/years involved in the comparison to see if that can tell us something. Maybe some other maps as well.

Might layered snow be a more effective insulator as there is little scope for convection though snow layers?

Yes, perhaps. At the same time, it's the air bubbles in the snow that make it so insulative, I think. And I guess there's less air in layered snow.

That would be a bad combination but 2014 melt season was poor. However the Dec 2013 mismatch had corrected by Jan 2014 so perhaps we shouldn't expect a poor 2014 melt season.

I'm not seeing that. I see the PIOMAS trend line going in a straight line from 2013 to 2015, but CryoSat still has a spike (relatively speaking) in 2014. And it's still there in February, March and April.

It will be corrected once the snow melts and there is not as much snow during the next freeze-up.

[crandles]

However the Dec 2013 mismatch had corrected by Jan 2014 so perhaps we shouldn't expect a poor 2014 melt season.

I'm not seeing that. I see the PIOMAS trend line going in a straight line from 2013 to 2015, but CryoSat still has a spike (relatively speaking) in 2014. And it's still there in February, March and April.

It will be corrected once the snow melts and there is not as much snow during the next freeze-up.

Perhaps yes. However, I don't think you should compare only with 2013 and 2015. The Jan 2014 gap of 2.5k is similar to 2011 and all the gaps are 2.5k to 3.5k with just one exception of 2017 gap of very weird 0.4k.

With Feb, yes 2014 is the smallest gap but this is no more unusual than 2013 being the largest gap.

[seaicesailor]

I've been able to look some more into this, with the help of Michael, who sent me the data that allowed me to create this graph:
...

Of all this, what amazes most is your eyeballing capabilities. Man, that you were able to detect this from the Cryosat maps with the colors they have ...

[Neven]

It was the difference with 2011 and 2012 that made me sure something was off (just by looking at the yellow colours).

Perhaps yes. However, I don't think you should compare only with 2013 and 2015. The Jan 2014 gap of 2.5k is similar to 2011 and all the gaps are 2.5k to 3.5k with just one exception of 2017 gap of very weird 0.4k.

With Feb, yes 2014 is the smallest gap but this is no more unusual than 2013 being the largest gap.

I was planning to look at the gaps today or tomorrow. Maybe I should divide one by the other, and thus create super data mutant CryoMAS! Just like CAPIE telling us about compactness, it will give us an idea of how much snow there is on the sea ice pack, compared to other years.

The problem is that it's difficult to align CryoSat-2 data with PIOMAS data, having to do with land masks and grid sizes, and what not. But Michael is looking into it. As soon as he has the time, he will send me data for both for the Central Arctic (or actually, the Arctic minus peripheral seas). That should be interesting.

[DrTskoul]

That is why Neven's EyeCrometer is so powerful
..

[diablobanquisa]

Hi Neven, have you considered whether the divergence between PIOMAS and CS2 could be due (at least partially) to ice dynamics, in addition to the 'snow-related' retrieval issue? (Kwok, R. 2015. Sea ice convergence along the Arctic coasts of Greenland and the Canadian Arctic Archipelago: Variability and extremes (1992–2014). Geophysical Research Letters, http://onlinelibrary.wiley.com/doi/10.1002/2015GL065462/abstract)

I have noticed that the divergence was already there in November...




... but it had dramatically deepen by the end of December:




What happened in December? Many things, of course   But, regarding sea ice drift, we had this:




I think that this pattern could have caused compaction and ridging from the Atlantic side towards the North Pole, CAA and North Greenland. The drift during December 2016 reminds me of the pattern that Kwok argued to be the cause of the increase in thickness in summer 2013:


https://nsidc.org/arcticseaicenews/2015/09/2015_arctic-minimum/


I think that this feature could be due to compaction and ridging...



... and it is just where one of the most striking differences between PIOMAS and CS2 seems to be located:




Just another theory.

[Neven]

Thanks, Diablo, that could have something to do with it as well. Either way, both factors seem for a large part to be caused by the series of Atlantic storms. If ridging does play a large part in this, the tables would be turned and it might be an indication that there could be a rebound this year. But I'm jumping to conclusions here.

I've notified the PIOMAS folks and they might ask around as well. So, let's see what happens.

[jdallen]

Thanks, Diablo, that could have something to do with it as well. Either way, both factors seem for a large part to be caused by the series of Atlantic storms. If ridging does play a large part in this, the tables would be turned and it might be an indication that there could be a rebound this year. But I'm jumping to conclusions here.

I've notified the PIOMAS folks and they might ask around as well. So, let's see what happens.

Considering the deficit of cold weather, I'm at a loss as how we could possibly have a volume rebound at this juncture.  The thermodynamics don't justify it.

[DrTskoul]

Thanks, Diablo, that could have something to do with it as well. Either way, both factors seem for a large part to be caused by the series of Atlantic storms. If ridging does play a large part in this, the tables would be turned and it might be an indication that there could be a rebound this year. But I'm jumping to conclusions here.

I've notified the PIOMAS folks and they might ask around as well. So, let's see what happens.

Considering the deficit of cold weather, I'm at a loss as how we could possibly have a volume rebound at this juncture.  The thermodynamics don't justify it.

Only mechanics (ridging etc)

[Neven]

And new ice growing where the winds have pushed the ice away.

[jdallen]

And new ice growing where the winds have pushed the ice away.

I'd be more optimistic and more in agreement were it not for the continuous melting along the Atlantic and export to melt from Baffin bay and Bering Sea.

New ice is forming, but I'm not sure it is overcoming that Volume loss.

[Archimid]

Considering the deficit of cold weather, I'm at a loss as how we could possibly have a volume rebound at this juncture.  The thermodynamics don't justify it.

I think it has a lot to do with the jump in extent. If I understand it correctly, once ice forms the first .5m grow really fast, explaining the rapid jump in volume.

And new ice growing where the winds have pushed the ice away.

I would expect to see a corresponding spike in temperatures. If it is cold enough the spike might be too small to even be measured, but there should be a spike.

[Michael]

To complement my earlier post where I compared average daily PIOMAS data with averaged CPOM 2 day CryoSat-2 data, here I have compared PIOMAS monthly data with AWI CryoSat-2 monthly data.

Because of the problems of reconciling the PIOMAS curvilinear orthogonal grid with the AWI azimuthal equal-area grid, the PIOMAS data are sampled using the EASE-2 1 km grid as a 25 x 25 subgrid of the EASE-2 25 km grid used by AWI.

Because the coverage of the Arctic by the CryoSat-2 satellite is incomplete on a monthly time scale, only those grid cell for which AWI report sea ice are included in the comparison.

The difference is calculated by subtracting AWI CryoSat-2 Sea Ice Thickness from POIMAS Sea Ice Thickness.

[Michael]

CryoSat-2 derived Sea Ice Thickness for February 2017 is now available so it is possible to compare them with the PIOMAS model output.

As before, PIOMAS daily gridded data compared to CPOM 2 day 1 km Sea ice Thickness.

Diference is calculated as PIOMAS minus CPOM.

[Michael]

Also as before, PIOMAS monthly gridded data compared to AWI Cryosat-2 Sea Ice Thickness.

Difference is calculated as PIOMAS minus AWI.

[epiphyte]

CryoSat-2 derived Sea Ice Thickness for February 2017 is now available so it is possible to compare them with the PIOMAS model output.

As before, PIOMAS daily gridded data compared to CPOM 2 day 1 km Sea ice Thickness.

Diference is calculated as PIOMAS minus CPOM.

There are radial striations visible in the plots which apparently line up with the orbital plane of the satellite (i.e. 92 degrees - tangential to edge of 'pole hole'). These just have to be artifacts, don't they? Assuming so, and given that the difference in measured freeboard between the peaks and troughs is as much as 2-3m, (most visible between 75-85N on the Russian side) what are we to make of it?

[crandles]

See http://www.cpom.ucl.ac.uk/csopr/seaice.html
 and click on 2 days, 14 days and 28 days.

It takes a month to cover all the area shown so it seem likely a result of comparing thickness at dates up to a month apart.

[oren]

The main difference, and the outstanding feature in PIOMAS this year, is the thick blob above Greenland. It's missing in both Cryosat and Hycom, at least in the magnitude alluded to by PIOMAS. I keep wondering about it, especially as it seems to be headed eventually towards the Fram which might cause a sharp drop in PIOMAS volume (if and when that happens).

[Michael]

There are radial striations visible in the plots which apparently line up with the orbital plane of the satellite (i.e. 92 degrees - tangential to edge of 'pole hole'). These just have to be artifacts, don't they?

The radial striations are due to the CryoSat-2 Satellite having limited coverage. Adjacent grid cells may contain data dated many days apart.

The CryoSat-2 satellite has a nominal orbital periodicity of 100 minutes, which means there are 14 to 15 orbits a day. Combine that with the fact that the SIRAL instrument uses “Synthetic aperture radar altimetry” to reduce the size of the instrument footprint to approximately 0.3 km by 1.5 km along track and across track, respectively (Laxon et al 2013), means that the area covered in one day is relatively small.

To help overcome this paucity of information, AWI use a 25 km grid. While CPOM utilise a gridding procedure that gives each data point a footprint roughly 10 km in diameter on the 1 km grid, and 50 km in diameter on the 5 km grid.

To illustrate this limited coverage I have done a comparison between the PIOMAS daily gridded data and CPOM 2 day 1 km data for 27_28 February 2017, on the EASE-2 12.5 km grid.Even this is potentialy misleading because the satellite track is much narrower than shown and there are two days worth of orbital tracks.

http://onlinelibrary.wiley.com/doi/10.1002/grl.50193/abstract

[epiphyte]

There are radial striations visible in the plots which apparently line up with the orbital plane of the satellite (i.e. 92 degrees - tangential to edge of 'pole hole'). These just have to be artifacts, don't they?

The radial striations are due to the CryoSat-2 Satellite having limited coverage. Adjacent grid cells may contain data dated many days apart.

The CryoSat-2 satellite has a nominal orbital periodicity of 100 minutes, which means there are 14 to 15 orbits a day. Combine that with the fact that the SIRAL instrument uses “Synthetic aperture radar altimetry” to reduce the size of the instrument footprint to approximately 0.3 km by 1.5 km along track and across track, respectively (Laxon et al 2013), means that the area covered in one day is relatively small.

To help overcome this paucity of information, AWI use a 25 km grid. While CPOM utilise a gridding procedure that gives each data point a footprint roughly 10 km in diameter on the 1 km grid, and 50 km in diameter on the 5 km grid.

To illustrate this limited coverage I have done a comparison between the PIOMAS daily gridded data and CPOM 2 day 1 km data for 27_28 February 2017, on the EASE-2 12.5 km grid.Even this is potentialy misleading because the satellite track is much narrower than shown and there are two days worth of orbital tracks.

http://onlinelibrary.wiley.com/doi/10.1002/grl.50193/abstract

[Neven]

Maybe they've been reading here, but it seems people at the NSIDC are aware of the situation:

Data from the European Space Agency’s CryoSat-2 satellite indicate that this winter’s ice cover may be only slightly thinner than that observed at this time of year for the past four years. However, an ice-ocean model at the University of Washington (PIOMAS) that incorporates observed weather conditions suggests the volume of ice in the Arctic is unusually low.

[epiphyte]

Maybe they've been reading here, but it seems people at the NSIDC are aware of the situation:

Data from the European Space Agency’s CryoSat-2 satellite indicate that this winter’s ice cover may be only slightly thinner than that observed at this time of year for the past four years. However, an ice-ocean model at the University of Washington (PIOMAS) that incorporates observed weather conditions suggests the volume of ice in the Arctic is unusually low.

...and am I correct in saying that in Beaufort, at least, the (few) on-the-ground thickness measurements (from the new 2017 IMBs) indicate that PIOMAS may itself be somewhat on the high side?

[Peter Ellis]

...and am I correct in saying that in Beaufort, at least, the (few) on-the-ground thickness measurements (from the new 2017 IMBs) indicate that PIOMAS may itself be somewhat on the high side?

"few" = one.  One is the number of active IMBs measuring thickness, and the number of the buoys shall be... one.

And no.  A point measurement cannot be easily compared to PIOMAS, which actually has an entire distribution baked into every pixel - X% at one thickness, Y% at a different thickness, etc.  The numbers reported and used for the graphs are the "grid box effective thickness" which is a sort of average over the whole pixel.

When a buoy is placed, it's likely to be in a thinner spot than average, because the average is pulled upwards by ridges and hummocks.  Nobody is going to drill down through the middle of an ice ridge to place a buoy.

Remember what arctic ice actually looks like - a rolling landscape with flat fields (floe ice), hummocks (deformed ice) and hedges (ridges between floes).  Over half of all the ice volume is in the ridges, and the buoys don't see those at all.

[Neven]

All we can do is wait for the results of Operation IceBridge to come in, and hope that will give us some more clues.

[Jim Hunt]

...and am I correct in saying that in Beaufort, at least, the (few) on-the-ground thickness measurements (from the new 2017 IMBs) indicate that PIOMAS may itself be somewhat on the high side?

No, as Peter pointed out. If the buoy doesn't get crushed in a ridge it should give a good idea how thermodynamic thickening (and eventually melting) goes from here. 2017A will certainly be on first year ice, but we can't say with any certainty when the floe started to freeze. < 1m seems on the low side to me for ice that's been increasing in thickness all winter long, and the Canadian Ice Service maps show thicker ice in that area. However I haven't checked location/motion/temperature yet to see if that helps explain matters.

I've been too busy winding up Anthony Watts amongst others!

 

[epiphyte]

...and am I correct in saying that in Beaufort, at least, the (few) on-the-ground thickness measurements (from the new 2017 IMBs) indicate that PIOMAS may itself be somewhat on the high side?

No, as Peter pointed out. If the buoy doesn't get crushed in a ridge it should give a good idea how thermodynamic thickening (and eventually melting) goes from here. 2017A will certainly be on first year ice, but we can't say with any certainty when the floe started to freeze. < 1m seems on the low side to me for ice that's been increasing in thickness all winter long, and the Canadian Ice Service maps show thicker ice in that area. However I haven't checked location/motion/temperature yet to see if that helps explain matters.

I've been too busy winding up Anthony Watts amongst others!

I wanted to mention that there's a recent sentinel 1 image of the neighborhood which may inform this discussion. FYI I just posted on it in the "What the Buoys are Telling Us" thread.

[Steven]

New paper by Ricker et al.:

Satellite-observed drop of Arctic sea-ice growth in winter 2015-2016

Abstract:

An anomalous warm winter 2015-16 lead to the lowest winter ice-extent and highlights the sensitivity of the Arctic sea ice. Here, we use the 6-year record of an improved sea-ice thickness product retrieved from data fusion of CryoSat-2 radar altimetry and SMOS radiometry measurements to examine the impact of recent temperature trend on the Arctic ice-mass balance. Between November 2015 and March 2016, we find a consistent drop of cumulative freezing degree days across the Arctic, with a negative peak anomaly of about 1000 degree days in the Barents Sea, coinciding with an Arctic-wide average thinning of 10 cm in March with respect to the 6-year average. In particular, the loss of ice volume is associated with a significant decline of March first-year ice volume by 13%. This reveals that due to the loss of multiyear ice during previous years, the Arctic ice cover becomes more sensitive to climate anomalies.

(click to enlarge image:)

[Feeltheburn]

...and am I correct in saying that in Beaufort, at least, the (few) on-the-ground thickness measurements (from the new 2017 IMBs) indicate that PIOMAS may itself be somewhat on the high side?

You could be on to something. According to DMI modeled ice thickness, the ice in Beaufort, especially near land, are thicker this year compared to last. If that is wrong, then DMI must be overestimating thickness overall. You can see this by clicking + and 1 365 days and comparing the two:

http://ocean.dmi.dk/arctic/icethickness/thk.uk.php

[CognitiveBias]

If you compare this year to last you see that the ice pulled off the Alaskan coast last year vs the Russian coast this year.  Overall the ice still looks worse this year.

[Adam Ash]

As soon as the CAA Garlic Press gets underway, all the remaining 'thick' ice will head south to oblivion.  Then all that will remain is a mush of barely-one-year-old ice filling the basin.  The risk of that melting out in anything like an above average melt season seems acute.

The atmospheric circulation arising from this year's melt season (ever more open water) re heat transport from equator to the Arctic will be very interesting to watch.  Are there any signs of the Ridiculously Resilient Ridge or its relations restoring itself?