Exceptional warmth in the Arctic regions has grabbed a lot of attention in 2019. Some regions have seen long-standing temperature records broken by wide margins. In many cases this has been tied in with sea and land ice loss on an unprecedented scale.
How did the Arctic get into this terrible state? Let me take you on a journey through time and climate…
Late 20th Century: Trimming the Edges
The global average temp rise was already well underway in the 1980s and 1990s. However, even the warmest of those years (1998) was nearly half a degree cooler than the warmest of the 2010s. There was also much less of a difference in warming between the Arctic and elsewhere than we’ve seen this past decade (more on that later).
The observed impact on ice is very interesting when viewing two different measures; area and extent. These have some similarities, but also some important differences, which you can read about in this FAQ.
Below (Figure 1), I’ve plotted the extent and area data for 1980-1999.
Figure 1: Plots of tri-monthly averaged sea ice extent (left) and area (right) for 1980-1999. The letters in the key correspond to the groups of months e.g. JAS = Jul-Aug-Sep. I’ve used tri-monthly averages to allow for quick assessment of sub-annual trends (as the dominant mechanisms for freezing and melting are not constant, meaning seasonal differences are to be expected).
Clearly, extent declined overall across all seasons, but area either held near-steady (JAS) or increased. During these two decades, the Arctic sea ice generally consisted of a vast expanse of sea ice near or at 100% concentration (i.e. few if any gaps per grid square). Only on the periphery was concentration very variable. The combination of reducing extent but increasing area implies that the peripheral belt saw losses while the central expanse grew slightly larger.
I’ve gone into so much detail here because it will serve to emphasise just how much the structure and loss pattern of the Arctic has changed in the 21st Century.
21st Century: Capitulation by All Measures
As this century got underway, the Arctic wasn’t sending a clear message that it was in serious trouble.
That would soon change, however. Area began to decline across all seasons as well. Meanwhile, the extent loss rate became even steeper, especially in the summer and autumn months (see Figure 2).
Figure 2: Plots of tri-monthly averaged sea ice extent (left) and area (right) for 2000-2018. The letters in the key correspond to the groups of months e.g. AMJ = Apr-May-Jun.
Clearly, there’s a lot of variability between individual years, but the overall trend is irrefutable. Bear in mind that we’re looking at millions of square kilometres in these plots. The Jun-Sep averaged sea ice extent, for example, has dropped more than 1,500,000 km in the space of 18 years!
So, what’s behind this change of fortunes for area and pace in extent? Well, the overall rise in global temperatures not only continued, but triggered a multitude of feedback mechanisms. These have regionally enhanced the rate of warming in the Arctic. Technically, these are called ‘positive’ feedbacks, but I’ll call them ‘additive’ here, to escape the ‘good’ connotation.
A Frenzy of Feedbacks
I will now summarise the most significant feedbacks that have been discovered and explained to date. This is one of the quickest routes to understanding how the Arctic climate is changing so rapidly.
Reflection Deficit – More Heat Hoarded
The most direct additive feedback is driven by changes in albedo of the surface within the Arctic. Albedo is a measure of what proportion of shortwave radiation, such as that from the sun, a surface will reflect. It ranges from 0.0 (total absorption) to 1.0 (total reflection).
Bare sea ice typically has an albedo between 0.5 and 0.7, but open ocean only has an albedo of approximately 0.06. So, as sea ice coverage of the ocean reduces, the amount of energy from the sun that’s absorbed increases immensely. The Arctic ocean warms, which in turn adds to the atmospheric temperatures above, further increasing sea ice loss.
Circulation Patterns in Disarray
There’s evidence that this accelerated warming relative to the rest of the planet has disrupted the Northern Hemisphere circulation patterns. This has increased the frequency and extent of large movements of unusually warm air into the Arctic and cool air out of it. There have been some mind-blowing examples of this in recent years (see Figure 3).
Figure 3: Daily mean air temperature anomalies (versus 1981-2010 baseline) for two example days on which exceptional movements of warmth into the Arctic (and cool or cold air out of it) occurred. It’s worth noting that the actual temps are in fact highest in the June example, due to much warmer baseline. Variability is also less in summer; 10°C above normal in June is at least as exceptional as 20°C above normal in November.
Increased Natural Emissions – as Bad as it Sounds
Meanwhile, recent observations tell us that vast expanses of permafrost, beneath the ocean and adjacent Arctic lands, are breaking down. Permafrost is soil which stays frozen for at least two consecutive years at a time. It should only experience brief thaws, if any at all. Within the Arctic region, much of this permafrost resides above large stores of organic carbon, methane, and some other natural gases such as nitrous oxide.
When it’s warm enough, decomposition of the organic carbon by microbes produces those natural gases. The trapped natural gases are a consequence of the Earth’s internal heat. As the permafrost melts, the microbial action accelerates and expands. There comes a point at which the gases are able to seep through breaks in the permafrost and enter the Arctic atmosphere. In some environments, this process can be alarmingly fast.
Like carbon dioxide, many of the natural gases produce a ‘greenhouse effect’ in the atmosphere. In fact, methane and nitrous oxide are a lot more potent. The resulting climate warming is regionally variable, depending on how the atmosphere is moving around relative to the gas emissions. Generally, though, it’s largest in the Arctic – another additive feedback.
The Arctic Ocean is Losing its Cool Head
Finally, we have what could prove to be the most serious additive feedback in the long-run.
A critical component for seasonal refreeze of sea ice in the Arctic Ocean is a layer of relatively fresh water (low salt continent i.e. low salinity) at the surface. This layer is sustained by inflow from rivers fed by a combination of melting snow or ice and rainfall (varying with the seasons).
There is also a large influx of more saline water into the Arctic Ocean, but this sinks down to the ocean depths, flowing beneath the surface layer rather than mixing with it. This is possible because the more saline water is, the denser it is. The fresher surface layer is more buoyant – its literally water floating on top of water.
This structure of fresher atop more saline water is the defining feature of the Arctic halocline (derived from Greek hals/halo – ‘salt’ and klinein – ‘to slope’).
The density difference is so large that the water at about 200-500 m depth can be some 2-3°C warmer than those at the surface, despite the fact that density reduces with increasing temperature. That typically means they’re above freezing, even at the coldest time of the year.
This layering is, however, vulnerable to mechanical mixing, by wind and wave action. Large expanses of thick, continuous ice, such as observed in the 1980s-90s, protect the layering by dampening the effect of those weather features. However, these have become increasingly scarce in the two decades since.
This is evident in the volume numbers. The 2000’s average annual sea ice volume peak was near 26 million km3, but the 2010’s average was near 23 million km3. That’s a loss of 3 million km3 in one decade!
Very strong storm events are now able to damage the Arctic halocline within the region of the Arctic they hit. Two recent examples are the so-called ‘Great Arctic Cyclone’ of early August 2012 and a slightly weaker storm in August 2016. The sea ice doesn’t respond well to the warmer, saltier water that then assaults it from the depths.
The layered structure is able to recover from these individual events, but there is a net loss of freshwater content to the surface layer; it thins and shrinks.
There is a very real risk that sometime within the next few decades, a much-weakened halocline could be completely mixed out during one August or September storm.
In such an event, the Arctic Ocean would become more akin to the North Atlantic. This means only seasonal sea ice formation, restricted to the continental shelves (adjacent overland temps would still fall well below zero, albeit not as far as they used to). There would potentially be no ice at all for several months of the year.
A Ball Rolling Down a Steepening Slope
I’ve spent a lot of this blog piece talking about additive feedbacks. In Figure 4, the power of these feedbacks is evident by the proportional melt change. This measure looks at what proportion of the March sea ice has melted away by September.
Figure 4: Plot showing the percentage of the March-averaged sea ice extent being lost based on the averages of each subsequent month within the melting season.
Sea ice extent loss in winter is naturally focused on the periphery of the Arctic Ocean. This ice is the ‘easiest’ to melt away, as the adjacent landmasses heat up to well above freezing during the summer months. So, the melting seasons are starting with an increasingly large proportion well away from land, and close to or across the North Pole. This ice is much more difficult to melt, due to the sun being lower in the sky (the sunlight ‘glances off’ more) and no land assist in summer.
Based on that in isolation, the proportion of March sea ice extent being lost would be expected to be reducing over the years. Yet instead, we see it increasing – and more than just a little bit!
The Arctic climate is akin to a ball rolling down an increasingly steep mountainside. It’s going to take something increasingly dramatic to stop it tumbling all the way down into the warm, rainy valley below.
James Peacock MSc
Head Meteorologist at MetSwift