May 2016 Status

The data is now in, so here's a look at May and prospects for the coming summer.



Using NCEP/NCAR data for May, at the surface temperatures have been amongst the warmest since the data started in 1948, however they were within the range of temperatures for recent years.

Aloft temperatures were even higher, the second warmest on record

The vertical cross section shows the warming is at depth within the atmosphere, to over 500mb.

The warming is associated with mass influx of warmer air, particularly over the North Pacific, this is shown in the following large graphic.

The following graphic has three panes, the upper is air temperature anomaly with height, over which is superimposed meridional flow, both are for the band 60 to 65degN. The meridional flow is shown in black overlay, northwards flow is labelled N, southwards is labelled S, I dislike the terms northerly and southerly as they're confusing, but a northwards flow is a southerly wind (i.e. wind from the south) and vice versa.

The largest flow is a northwards flow over Alaska which is associated with a high to the south of Alaska (panel 2) and a northwards kink in the jetstream (panel 3, approximate average location of the jet is shown in black). There is also a flow northwards over the Baltic Region (Western Russia and Scandinavia).

This pattern of meridional flow is similar to the climatology, which is shown below. This is not a surprise as the fixed land/ocean pattern fixes the pattern

However referring to the three paned image above, the correlation of Nino 3.4 is shown below. From this it seems that the pattern of strong northwards flow over the Baltic and Alaska is associated with the El Nino.

The El Nino is fading, however it will be later this year before a La Nina sets in and the impact on global average temperature and Arctic temperature will likely persist over the coming summer.

So what effect has this had on sea ice during May?

PIOMAS gridded data is now out, I have used the PIOMAS HiDay gridded files of grid box effective thickness to calculate the volumes for all regions on days 121 and 151 (approximately the start and end of May). This allows an initial indication of the first part of the spring volume loss. Once June's data is in I will look at this in more detail, but for now I look at May volume loss.

The sectors are:

• Pacific: Okhotsk & Bering Seas.
• Peripheral: Beaufort, Chukchi, East Siberian, Laptev Seas.
• Atlantic: Kara, Barents, Greenland Seas.
• Central Arctic: Central Arctic & Canadian Arctic Archipelago.
• Canadian: Baffin Bay & Hudson Bay (Gulf of St Lawrence does not feature in PIOMAS). 

The above graph shows the overall increase in May volume loss, but it isn't very clear, so I have recalculated that data as differences from the average for each sector from 1980 to 2016.

What interests me are the large losses in 2010 and 2012. 2010 was a large volume loss year, but remnant ice in the East Siberian Sea kept September extent from falling to very low levels, whilst 2012 was a substantial new record. Comparing 2016 to those years shows that May 2016 volume loss in the Peripheral Seas was far from as severe as in 2010 and 2012, this might prove significant as a factor for the coming summer because over winter thickening in the East Siberian Sea and Laptev has been quite normal for recent years, and in May Laptev saw a marginal net gain in volume. This suggests ice is entering the summer proper in those regions is in rather a healthy state similar to recent years.

Indeed only Chukhi has seen greater May losses than in May 2012. It might seem strange that Beaufort has lost less than in May 2012 but the massive loss of extent over May has come from the the loss of very thin ice in a poor state caused by wind driven divergence over winter (export towards Chukchi).

Over May Beaufort extent has collapsed very early, which has been due to wind driven export.

The slight recent levelling may be the start of a reduced rate of loss due to the presence of multi-year ice, I recently suggested may be about to start. However my attitude to this is simply to watch and see what happens, I don't know if this is just a minor pause, or the start of a change of loss rate.

The recent sensor failure has stopped me from using sea ice data based on that system, so I have had to resort to the shorter AMSR2 dataset. This poses a problem for assessing long term significance. So the following graph for Beaufort combines both datasets and shows extent for a seven day average centred on 1 June.

Clearly Beaufort has more open water than any time in the record at the start of June.

However Beaufort is a relatively small part of the peripheral seas and it is in those seas at the coastal edges of the Arctic Ocean that the serious melt will start to happen through June. Conditions in those seas set the scene for the coming summer.

Overall extent falls in the peripheral seas is not exceptionally unusual.

Extent really tells us about the ice edge, what is the ice state behind the ice edge in the peripheral seas?

Compactness is the ratio of sea ice area and sea ice extent (explained here), it gives an indication of how compact the ice is behind the ice edge. Compactness falls during the summer as floes melt and open water between them increases, current compactness isn't exceptional in the peripheral seas (nor indeed in Beaufort itself).

And that slight drop recently in the 2016 drop probably isn't significant. Again I have calculated the 1 June centred seven day average, this time for Peripheral Seas compactness. Here AMSR2 has had to be scaled to match the overlapping three days. But compactness in the Peripheral Seas doesn't seem to be doing anything really surprising.The situation at the end of June might be different.

Overall extent remains low.

However this is mainly due to low sea ice in the Pacific and Atlantic regions.

In the Arctic Ocean the effect of the Pacific Ocean extent is left out.

Extent tracks below recent years, and comparing AMSR2 and NSIDC Extent is likely to be at record low levels.

Beaufort is one factor here, but the Atlantic ice edge plays a role. The Central Arctic is already showing a very early drop in extent as the ice edge has been moved polewards of Svalbard by what seems to be Atlantic Ocean warmth.

Going back to the difference from 2012 in volume, the overall pattern is that from the Siberian coast to Canada there is a central band of ice that is in a reasonable condition similar to other recent years, however Chukchi and Beaufort provide an avenue of warmth from ocean warming due to open water and the ice/ocean albedo feedback.

I am not convinced the Atlantic ice edge has much bearing on extent within the Arctic Ocean Basin, the trend of ice loss in Barents is far steeper than in the Arctic Ocean Basin's seas, it does not seem to be a particularly 'infectious' trend.

Overall volume is similar to other recent post 2010 years, but is below May 2015. For completeness below is the May volume for all regions.

How much is the Central Arctic keeping overall volume up? Below I have calculated the total May volume, for both Central Arctic and the Total volume minus Central Arctic.

Remove the Central Arctic from the equation and volume is the lowest on record, but not by a large margin.

A new record is possible this year, but from my reading of all the data it is far from certain. With poor melt weather we could yet again see an increase in September volume on last year. The El Nino is still having a substantial impact on the atmosphere, what impact this will have on the creation of an Arctic Dipole?

The pressure patterns that drive the Arctic Dipole are shown above. Winds from the high pressure over Greenland and the Canadian side of the Arctic Ocean act together with low pressure dominance over the Siberian side and drive a flow that draws in warm winds and Pacific water. This pattern has dominated summers since 2007 and has played a role in low extents since that year.

In general a good melt weather pattern in the summer is clear skies over much of the ice through June and August, but storms and cloudy skies in August. The early summer clear skies let the sun shine on the ice, then later in the summer as the sun is getting low and the energy from it declines, clouds trap infra-red emitted from the ice and ocean, radiating it back to the surface. Storms in the late summer can disrupt the ice pack and mix warm water from the surface under the pack causing more melt.

The winds that have so damaged Beaufort have blown along the Alaskan coast towards Siberia. Ekman Transport would cause the movement of the ocean column (where there is open ocean) and the surface ice to be towards the right of the wind direction. In other words northwards. Has this drawn warmer Pacific waters in through the Bering Strait? If it persists, will it drive water warmed by the sun under the ice causing enhanced melt?

There is a lot of uncertainty, but it is possible that we will see September extent approaching that of 2012, possibly beating 2012, either outcome will need the weather to be conducive to melt over the coming summer.