PIOMAS November 2014

PIOMAS data is out, so here's the usual analysis, with a bit of NSIDC Extent tagged on at the end...



My map-plot of PIOMAS grid box effective thickness shows continued thickening, which is rather trite to note in the autumn.

I noted in my November Status post a thin tongue of ice over 2m thick (blue) levering away from the northern coast of the Canadian Arctic Archipelago (CAA), this feature has thinned out. But there is ongoing thickening into Beaufort and through to the East Siberian Sea (ESS). PIOMAS does have vector data, but I've not as yet had the time to grapple with that (it's rather daunting too), however HYCOM Global does indicate export into that region on top of existing thicker ice there.

By this November the true scale of what I described as a massive volume increase in my September PIOMAS status post is revealed. Using the PIOMAS main series (source) the increase in volume is astonishing. having been at post 2010 volume levels last winter, current volume now rivals 2008!

I have read people expressing their disbelief of these PIOMAS model results, I see no reason to doubt them. Schweiger 2011 gives an uncertainty for October sea ice volume from PIOMAS as +/- 1.35k km^3, this is conservative (a larger range than one might reasonably assume). Taking the PIOMAS main series volume of 13.314k km^3 on 30 November 2014 puts 2014 at roughly the same level as 2008 (13.332k km^3). If one applies a -1.35k km^3 to that 13.314k km^3 then on 30 November the volume would be between 2009 and 2010, but if you do that you are bound to accept that the converse is as likely: that volume may be as high as between 2004 and 2005. When working with an expression of '+/-' uncertainty the door swings both ways.

As usual I will continue to work of the main series figures as given, or the figures given in the gridded data. Note that by 'main series' I mean the data that the Polar Science Centre (mainly Dr Zhang as far as I know) provide for daily and monthly volume. However as regular readers of this blog know, I also do my own calculations from gridded PIOMAS data, some results of which will follow.

Using gridded PIOMAS data I have (as usual) broken down volume into regions, the regions used are the Cryosphere Today regions because these are well know in the amateur sea ice community.

The pattern of regional volume remains that seen throughout the summer, that the volume increase is largely confined to the Central Arctic.

For the whole PIOMAS Domain (north of 45degN) thickness distribution is shown below. Ice volume from grid boxes below 0.9m effective thickness continues to increase, 1 to 1.9m thick volume is up from 2012 but arguably remains part of a recent downward trend. It is from grid boxes between 2m and 3.9m thick that the bulk of the increased volume arises. Thickening beyond such thicknesses is almost totally by ridging and other mechanical deformation due to the insulating properties of sea ice.

This year shows a substantial increase in volume, however the pentad (5 year) averages of thickness distribution demonstrate how much further there is to what could reasonably be considered a recovery of the sea ice (pre 2000s levels).

However in terms of the 2000s the thickness distribution shows a clear high-side bias, due to the higher volume of thicker ice. comparison with individual years post 2010 underlines this.

2010 was a major volume loss event equalling the 2007 loss event, more here. From the above it would seem that the impact of 2010 has turned out to be rather limited, with an increase in volume repairing the damage of that year. However another perspective reveals that this might not be the case.

Now that I have November data it is possible to add one more year onto the series of September to October volume increase, this is an important period because it shows the early season rapid period of volume increase.

But before I go onto that data, Thorndike 1975 provides a graphic of rate of ice thickening as a function of ice thickness.

From the above, it is clear that the period of most rapid gain in thickness (hence volume) is when the ice is below around 1m thick. That is the period from September to December for much of the Arctic Ocean ice pack. This is due to heat conduction through ice, a subject I will go into in more detail at some stage in the next few weeks.

For this post I just wanted to show the gain of volume from September to November calculated using gridded PIOMAS data. That could be done with the main series, but the gridded data allows regional decomposition.

From this it is clear that 2010 ushered in a new period of September to November volume gains in excess of 4k km^3, and it is clear that 2014 is still part of the post 2010 group. Prior to this grouping the only year in excess of 4k km^3 volume gain was 2008.

I have used PIOMAS gridded volume data to decompose this volume gain and show percentage contributions to the volume gain from September to November.

After 2010 the Central Arctic (red) shows a large increase in percentage contribution to overall volume gain. 2014, with thicker ice, shows a small drop off. But 2014 still has a large percentage contribution.

Looking at this in absolute terms here are the volume increases from September to November for all regions of the PIOMAS domain, the increase in volume for the 2010 to 2014 average in the Central Arctic is massive.

Now that it is clear that the area of interest is the Central Arctic we can look at the behaviour of volume gain in that region.

In line with my expectations the increased volume has led (through thickness) to reduced volume growth, down to about 3/4 of the highs of 2010, 2011, and 2012. However I would not have guessed that volume growth would be quite so high based on conditions back in September, I had expected 2010's effects to be totally wiped out, yet volume gain remains historically high.

Thickness plays a role in this due to the relationship between thickness and rate of thickening shown from the Thorndike 1975 graphic displayed above. Here is regional thickness for the Arctic Ocean.

The Central Arctic reveals a strong uptick in thickness, in line with expectations based on volume. Has this wiped out the impacts of 2010, on balance I don't think so. But if 2015 is as poor a melt season as 2013 and 2014 I expect that the effect of 2010 will be totally removed.

Before closing I want to look at NSIDC Extent for the most recent data available.

What strikes me most about this is a period of level extent in the early part of the series, then a drop, after which we see another levelling. That levelling at the end of the series seems to coincide with the hiatus in global warming, which started around 2003 (if you think it happened from 1998, you are wrong).

I know others more knowledgeable than I in statistics have made strong arguments against the supposed hiatus in global warming, I still see a hiatus. I'm not a statistician, I'm just an engineer. But when I see readings levelling I presume that regardless of statistical confidence there probably is a levelling, and see things ramp up sharply has made me reach for the off switch, that has saved money (in damaged kit) without reference to statistical significance. Arguing that such a hiatus represents the end of a long term trend is a statistical question, and as Tamino and others have shown repeatedly the hiatus is not a break in the pre-established warming trend.

So if the pattern of decline in the above graph is an AGW driven decline bound by two periods of quiescence, where has the decline come from? Using Wipneus's calculations of regional extent it is possible to answer that.

Two averaging periods are used, the 1980s and the most recent ten year period 2005 to 2014. Central, Beaufort, ESS, Laptev and the CAA all show little or no decline. The decline comes from regions at lower latitudes and in the adjacent northern Pacific and Atlantic Oceans. 54% of the total decline is in the Arctic Ocean, 44% of which comes from Kara, Barents and the Greenland Sea (leaving 10%). Hudson bay, within the North American continent, contributes 25%. In the Pacific Okhostk & Barents contribute only 8% of the total decline, however those regions have seen a 25% drop in extent from 1980s levels by the 2005 to 2014 period. All of which seems to me to be consistent with an AGW driven loss of extent, in tandem with the overall volume decline from 1995 to 2012 within the Arctic Ocean.