Sea Ice

The National Snow and Ice Data Centre (NSIDC) is a great resource for monitoring Arctic sea ice extent. The Multisensor Analyzed Sea Extent (MASIE) covers the period 2006-2019. They offer a CSV file of data that can be downloaded.
They also offer the sea ice tools which also has a downloadable CSV file for loading into a spreadsheet that covers the period 1978-2019.


A plot of daily readings of northern sea ice extent from 2006 to present.

NSIDC interactive sea ice graph

NSIDC offers a cool interactive chart for displaying the annual variation in ice extent.

However, I find that chart to be too noisy which makes it difficult to follow trends so I downloaded the CSV file to chart the max and min of the ice extent. By noisy I mean the constant random weather changes that makes deciphering trends very difficult without some form of data smoothing. You see that in all the temperature records where year-on-year variation can be quite extreme yet the longer term trends are actually quite small. For example, as explained in the Temperature section, year-on-year temperature swings can regularly be 5C yet the 100 year trend for temperature is .oo5C per year which is quite a bit different. Weather is just an instance of climate and is prone to wild random swings. You have to accumulate the weather for a few years to settle on a climate.


I did find a resource for previous years, from the department of energy which provides a graph of mean sea ice extent going back to 1920.

I calculated the Mean from the NSIDC data set and added the DOE Mean data to derive this chart. There is a data gap between the two but have seen other graphics that show a recovery during that period. Eventually I will find a more complete dataset and provide it here. The newer dataset is from satellite measurement and I imagine the older would be from airplanes and ships monitoring the ice shelf. The satellite era has a curious reference to sea ice extent including areas of the ocean with “at least 15% of ice”. That’s just crazy to include ice cubes that break off from the ice shelf. We really want to know where the edge of the ice shelf is, not where ocean currents could be transporting pieces that break off. Having said that, the reason for measuring sea ice was to be able to determine navigability so pieces of ice affect navigation so there seems to be a reasonable historical reason for it.

I did find this graphic that suggests the missing period may include a rapid recovery of ice extent.

  • Satellite measurement of ice extent is relatively new and we haven’t been measuring change long enough to show the behaviour of even one complete cycle. i.e. Atlantic multi-decadal Oscillation (AMO) cycle has approximately 30 years of warming followed by 30 years of cooling. It is entirely possible that we have only observed the 30 years of warming during the satellite era. And it is highly likely that if we watch it for another 30 years, we will see a recovery like we saw before.

So lets look at the surface temperature for a city in the arctic circle in Iceland, Reykjavik. As you can see, once again we see the all too familiar natural cycles of warming from 1920-1940, cooling from 1940 to 1979, and the warming again from 1979 to 2000.

If we superimpose an ocean circulation pattern, Atlantic Multidecadal Oscillation, (AMO), we see there is very strong correlation that is clearly demonstrating that ice and temperature are strongly controlled by ocean circulation patterns that are cyclical, rather than a linear increase caused by carbon dioxide, which is not cyclical but linear.





  • While Arctic ice extent was declining during this period, Antarctic ice has been growing, which is kind of surprising for a warming earth. I would like to explore more into the possibility of oscillations between the northern and the southern hemispheres.





Ice, Arctic, Antarctic