Sunday 4 September 2011

Arctic Dipole: A Positive Feedback.

In a previously discussed paper, Zhang et al open their closing discussion with the following line:
"The atmospheric circulation change identified here
could be a mixed response to rising greenhouse-gas-emissions
forcing or unexpected natural climate variability, or both."

More recently the meteorologist Jeff Masters has stated that the Arctic Dipole may be due to reduced sea-ice here. That's an article I strongly recommend.

Overland and Wang 2010 state:
"Most authors including the present ones considered that the
persistent AO and AD patterns were mainly representative of
natural variability of the chaotic climate system in the northern
latitudes. However, due to recent sea ice loss at the end
of summer there is a direct feedback to shifts in the broader
atmospheric circulation in late autumn and winter. Heat being
stored in the upper Arctic Ocean due to reduction in the area of
Arctic summer sea ice is given back to the atmosphere in the
following autumn, which has a direct impact on the temperature
of the troposphere and thus on geopotential height and thickness

First a note of clarification: Overland and Wang use the convention that the "AD pattern has a dipole field with the positive phase associated with a negative SLP anomaly on the North American side of the Arctic." Yet Wu et al state: "When the dipole anomaly remains in its positive phase, that is, negative SLP anomalies appear between the Kara Sea and the Laptev Sea with concurrent positive SLP over from the Canadian Archipelago extending southeastward to Greenland." So the two papers are at odds using opposing definitions of the AD index. This is despite Overland and Wang citing Wu in their paper! In my earlier posts I have used Wu's convention, which agrees with the convention in all the other papers I have read, and I am minded to stick with it. However anyone reading Overland and Wang should be aware of this conflict so as to avoid confusion.

So that issue clarified...

Figure 4 of Overland & Wang 2010.

The above figure is from NCEP NCAR reanalysis, it covers the atmosphere over East Siberian/Chucki/Beaufort to the pole, at around 60deg is the Bering Strait, the period covered is October to December 2002-2008. This is an anomaly plot, as is the following graphic. So where the plot is white that means no (or least) change from the climatological baseline period, any areas of colour represent a departure from that baseline.

The most prominent warming with the highest impact is away from the pole, this is where the ocean's freeze up is delayed by the heat accumulated in anomalous areas of open water due to the rapid loss of sea-ice over that period. It may seem odd that there is also notable warming over the lattitudes covered by ice during the summer. However Overland and Wang find that the positive phase (following Wu) of the AD has happened more frequently in recent years and is associated with meridional flow over the Bering Strait into the Arctic. Furthermore I suspect that the Beaufort High will play a role in distributing heat poleward.

This lower tropospheric warming causes a marked disturbance in the vertical profile of the atmosphere. As one proceeds upwards from the surface pressure drops. So if you take two pressures, 400 and 500mBar for example, there is a height difference between them, this is the geopotential thickness, the thickness of the atmospheric column between two layers of pressure. Take the red area in the graphic below, this is where the height of pressure layers around 550 to 500 hectoPascals is over 35 metres above the average height for the baseline period, 1968 to 1996.

Figure 7 of Overland & Wang 2010.

The lower tropospheric warming produces the above disturbance in the atmospheric column. Because the atmosphere above the pole is usually highly stratified in autumn and winter the lower tropospheric warming causes the marked disturbance to the profile around the Pole, right side of figure, causing the geopotential heights to be raised. Over the open ocean itself, left side of figure, there is a low level drop in geopotential height with increases (green band) higher in the atmosphere. The region in white is the area of minimum disturbance, however the black isobar passing through the middle of that region shows that above the line there is a small increase in geopotential height, below is a decrease - note the scale below the plot, anything less than +/- 5 is white.

Overland & Wang find that these changes in the atmospheric profile cause a reduction in the zonal wind flow, of which the Arctic Oscillation (AO) is the typical mode, this reduction is 40% of the maximum climatological zonal flow. As I understand it; it is this reduction in zonal flow that gives rise to the AD having attained a dominant role in the period since 2002, as noted by Zhang et al and discussed here previously. The most marked reduction is not actually in the area around the pole but over the Beaufort Sea, where sea-ice reductions and persistence of open water in the Autumn have been most marked.

So not only does the AD (in its positive phase) act as a signficant driver to ice loss, (as discussed in an earlier post), it is in turn created by loss of sea-ice and the transfer of heat thus gained by substantial areas of open water to the atmosphere. This is why opinions are changing and the AD is starting to be seen not as 'weather' but as an active and ongoing positive feedback on loss of sea-ice.

The really observant will have noted that in the quote from Overland and Wang above there was reference to impacts on geopotential height thickness fields. I've not addressed that impact in this post but these probably play a role in the recent cold winter events. As I've gone on enough for one post, that will have to wait.

Overland & Wang, 2010, "Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice."

Xiangdong Zhang et al, 2008, "Recent radical shifts of atmospheric circulations and rapid changes in Arctic climate system."


Kevin O'Neill said...

OK, you're to blame for this, so you explain it.

Why has the air temperature /potential temperature above the tropopause over the Chukchi and East Siberian Seas developed large negative anomalies during the melt season since spring 2007?

Is there an oscillation that changed phase between 2006 and 2007?

Kevin O'Neill said...

Chris, per Neven I asked Wayne Davidson about the stratospheric cooling apparent post 2006:

"stratospheric ozone plays a huge role in stratospheric temperatures. In essence there is very little ozone during the long Arctic night
so cooling predominates, CFC's and other ozone depleting chemicals eradicate ozone which would normally be high in concentration when the sun rises especially after the long night, this has caused huge temperature dips of late especially spring of 2011.
The warmer troposphere makes it higher, and also starts the stratosphere from a colder tropopause, the two; ozone depletion and warmer troposphere have caused these anomalies..."

I don't sea a direct link to sea ice loss, but perhaps increased UV?

Kevin O'Neill said...

I may be chasing ghosts in the data, but here's a comparison of 2006 and 2007 from CT. I've outlined the zone with the negative anomaly per NCEP/NCAR . Sure looks like it covers the area of highest sea ice losses.

The more I think about it, the more I think that increased UV could be responsible (and supply the increased energy you were looking for).

*If* this is correct (big IF), then the question becomes: Is this a self-amplifying feedback maintaining the anomaly over this location?

If it is, then it will grow. And sea ice losses will increase. Pretty much what we've seen. A very tenuous chain of conjecture, though :)

And then you have to take into account that it will (if not already) change circulation patterns. In the Antarctic, ozone depletion is credited with *increasing* sea ice. Yeeesh.

Chris Reynolds said...

Sorry for not replying earlier, I missed your replies and didn't have the time to consider last night.

OK. Potential temperature (PT) is the temperature change in a layer of stratified atmosphere that would result from a parcel of the atmosphere being taken to some reference level. These are anomalies, so we don't need to worry too much about what that reference level is.

The plots you provide show that in the stratosphere the PT is declining, implying that the geopotential heights are going down. In the troposphere the PT is increasing showing that the geopotential heights are increasing. I think I'm correct in this because if you take a parcel down from 11 to 1 it warms 1 more than if you take it from 10 to 1. Conversely if you take a parcel from 9 to 1 it warms 1 less than 10 to 1 (The effect is actually non-linear but I'm ignoring that technicality).

I disagree with Wayne only because these aren't necessarily indicative of temperature changes, i.e. the geopotential heights may be changing for other reasons, weather being one.

I need to think about this!

As for the Antarctic, Jinlun Zhang 2007, "Increasing Antarctic Sea Ice under Warming Atmospheric and Oceanic Conditions." pdf here

"The model shows that an increase in surface air temperature and downward longwave radiation results in an increase in the upper-ocean temperature and a decrease in sea ice growth, leading to a decrease in salt rejection from ice, in the upper-ocean salinity, and in the upper-ocean density. The reduced salt rejection and upper-ocean density and the enhanced thermohaline stratification tend to suppress convective overturning, leading to a decrease in the upward
ocean heat transport and the ocean heat flux available to melt sea ice. The ice melting from ocean heat flux decreases faster than the ice growth does in the weakly stratified Southern Ocean, leading to an increase in the net ice production and hence an increase in ice mass. This mechanism is the main reason why the Antarctic sea ice has increased in spite of warming conditions both above and below during the period 1979–2004 and the extended period 1948–2004."

I'm not fully up to date with the Antarctic. I had thought that stratospheric cooling made the Antarctic Polar Vortex go into a positive mode which means a strengthened vortex keeping the Antarctic cold, hence more sea-ice. But there has been warming...

It's another thing I need to investigate. But right now I'm busy looking into the reasons for the recent cold winters.

Kevin O'Neill said...


Don't get hung up on potential temperature; the air temperature (stratospheric) is the same graph, same anomaly. In this case the potential temperature actually reflects the change in stratospheric air temperature.

I think the other clear difference is the highly stratified post-2006 graphs. The tropopause is a clearly delineated layer from 70N to 90N post 2006. You don't find that in the pre-2007 composites.

If this anomaly is of any relevance it is to dynamic changes. It may well be a result of the same weather patterns that are causing the colder winters.

Colder winters can cause ozone depletion and lead to the creation of polar stratospheric clouds which leads to more ozone depletion. Ozone depletion can account for lower stratospheric temperatures. Assume this is the cause of the anomalies I've noted. Why is it spatially limited to the very same seas where we're seeing the most sea ice loss?

My main interest in this is the belief that 2007 ushered in a change. This is the most concrete example I've seen that points to a distinct difference between the pre and post 2007 crash. While it may not be a direct cause of the post 2007 ice losses, it has to be related (the coincidence is just to hard for me to accept). It may be an effect as much as the ice loss itself is an effect, and should therefor serve as an additional clue as to what the true 'cause' is.

The antarctic has been warming, but sea ice has been increasing. There are plenty of differences between the two polar regions that allow for numerous possible explanations. Ozone depletion was considered a highly plausible cause for INCREASING antarctic sea ice despite warming temperatures. Sigmond & Fyfe set out to show this - but found contrary evidence in the simulations. I.e., sea ice DECREASED (contradicted observations) when they forced observed ozone depletion.

The main similarity between the poles is that each are seeing anomalous (new) weather patterns. The obvious prime mover is rising temperatures. The actual mechanics of how individual pieces of the puzzle act upon each other is difficult to isolate. In the antarctic surface topography seems to play a much larger role.

I've sent out a few emails hoping for some illumination, but as yet have not received any replies.

Chris Reynolds said...

I think that to reply needs a proper post. The reply I've just knocked together is too long here due to the length of links for NCEP/NCAR plots. And I could do with some time to ponder.

Check back in 2 days.

Chris Reynolds said...


I've done a post with my best guess as to what's happening here