Technical Points About Measurement of Sea Ice Extent
A few technical items on how sea ice extent is measured:
- Satellite observations of the poles have been available since September 1979.
- Microwave energy behaves differently for sea ice versus open water, and passes through cloud and is “visible” during the long polar night. Therefore, satellite microwave channels are used.
- The ocean domain is broken up into 25 by 25km squares or pixels. Sea ice extent is defined as the areal extent of pixels with 15% or more sea ice.
- Each 25x25km square times the areal sea ice coverage can also be summed up to obtain the sea ice coverage.
- Sea ice extent is used instead of coverage because, while less accurate, it gives a less “noisy”, more consistent year-to-year value.
We will be using sea ice extent throughout this post.
The 2021 Melt Season in Context
With the impacts of the climate crisis seemingly all around us this summer, one might have expected a precipitous decline in sea ice extent in the Arctic Ocean as well. However, that did not ultimately happen, even with 2021 trending pretty much with the record minimum in 2012 until about mid-July.
In the next graphic, we show the Arctic sea ice extent since the beginning of June, compared to 1981-2010 normal and the record low minimum ice year of 2012. The 2012 ice pack was more dominated by first year (seasonal) ice, which is both thinner and more prone to break up and melting from unusual weather events. Such an event, a strong low pressure system that moved over the mid-Arctic Ocean in early August 2012, served to speed up the loss of the thin ice already on the verge of melting completely.
But in 2021, August was dominated by fairly standard near-surface temperatures and a lack of extreme events over the Arctic Ocean. The mean near-surface (2 meter) temperature analysis for August 2021 is shown below, centered on the North Pole.
The Arctic Ocean surface air temperature was near to even slightly cooler than normal in August. In limited Arctic Ocean near-shore areas, surface air temperatures were much warmer than normal because of an anomalous lack of sea ice. With all that, however, Eastern North America. most of central and northern Asia and Greenland were up to 6-7oC (11-13oF) above normal. It’s not that surprising that for the first time in the modern climate record, liquid precipitation and above freezing temperatures occurred at the highest point in Greenland.
We can look a bit more closely to this year’s melt season when we compare it to those over the last decade (2011-2020). That’s shown next.
The heavy dotted line marks the 2021 sea ice extent; 2021 looks to be tied with 2018 and 2017 for third highest extent over the last 11 years, with 2013 and 2014 tied for 1st place. So in the context of the last decade, this has not been a bad year for Arctic sea ice extent. Even so, the 2021 minimum extent will be in the lowest 10% of the 1981-2010 record.
What Is the Trend In Sea Ice Volume?
Sea ice extent is like a thin slice of the amount of ice present; to determine the amount of energy to melt the entirety of that ice, we need to know the volume. That value is harder to obtain directly from observations, though satellite observations are useful. Another way to get at the volume of sea ice is to run a computer model that assimilates weather observations, includes an ocean model that provides ice edge and bottom ice/water interface boundary conditions, and satellite-observed sea ice area coverage and thickness (area coverage x thickness = volume) for the starting point (a.k.a. initial condition) and as an ongoing “sanity check”.
The data below are from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS, Zhang and Rothrock, 2003), developed by the Polar Science Center at the University of Washington. In the first graphic below, we see the sea ice volume anomaly (gray time series, with error bars) and trend (with the dark shading around the trend line representing about 68% of the trend uncertainty, and the light shading representing about 90% of the uncertainty).
Looking at the time series itself, we can see decade-long trends where the volume stays either more steady or more sharply declines. From 2010 to 2021, for example, we seem to fluctuate at about a 6,000 km3 negative anomaly, while from about 1996 to 2010, the anomaly falls steeply from about plus 4,000 km3 to about negative 5,000 km3. Some climate skeptics have interpreted the last decade’s “pause” in declining sea ice volume as the end of the disappearing Arctic sea ice, but this is something to be expected in a warming world because of natural climate variability. One explanation for the pause is that the sea ice has been recovering from the catastrophic 2012 drop in volume, associated with that extreme August 2012 cyclone that moved into the central Arctic Ocean. The recovery is because the average Arctic climate state during the 2012-2021 period is not warm enough overall to sustain the low 2012 value.
An Aside: The Mean Global Temperature Reflects All Earth System Impacts!
Finally, here’s the trend in mean global surface temperature from 1958 to 2020 (apologies for the dark background). It surely doesn’t look like warming has ceased, especially when considering that 2016 was a record El Nino year.
As in the case with Arctic sea ice extent, however, the land/ocean surface temperature is only a slice of the earth system as a whole. The heat held in subsurface soil is negligible in the grand scheme of things, though at times it’s important locally. However, ocean heat content through its depth is another thing altogether. Not only does the ocean surface have a much larger area than the land surface, it has a tremendous heat capacity relative to the soil volume! Over long periods of time (decades to centuries), it is the greatest redistributor of heat in the Earth system.
Below we show the trend in ocean heat content anomalies relative to the 1981-2010 mean in the upper 2000 meters of ocean volume, during the same period as the graphic above (1958-2020). The units are in ZetaJoules (ZJ; a joule is a unit of heat, zeta is a LOT of units … to be exact, one sextillion or 1,000,000,000,000,000,000,000 of them!)
Let’s compare the capacity of the atmosphere versus the ocean to hold heat energy. To heat the depth of the atmosphere 1oC takes 2.2ZJ, while to heat the full mass of the ocean 1oC is 1000ZJ, almost 500 times that of the atmosphere. Year over year changes in ocean heat content would, if immediately released to the atmosphere, result in a 5-10oC change in whole atmosphere temperature (but more likely, mostly in the bottom few kilometers.
The bottom line is that global ocean heat content is by far the biggest component in the Earth System, and more or less is the final arbiter when it comes to a warming earth. It’s why warming the system has taken a deceptively long amount of time to notice for most human beings, and why even when we stop dumping greenhouse gases into the atmosphere, global warming will be far from over. All that heat will need to be radiated from the ocean surface, through the atmosphere, and finally out into space.
Final Thoughts
The Arctic sea ice this melt year continues to get something of a reprieve since its precipitous decline in 2012, as the ice works to get to a new, temporary equilibrium with the full Arctic environment. I would expect the decline to resume in the next couple of years, especially given how warm the surrounding land areas have been each summer. Stay tuned.