Update on the 2017 Arctic Melt Season

I made a posting on March 22 where I made a way-too-early forecast on the Arctic sea ice melt season. In that forecast I calculated the sea ice minimum extent for this year would approach 3 million square kilometers, which would easily be the minimum extent ever measured. The season is well advanced, so I thought this would be a good time to update that forecast and see how things are panning out.

To start with, this is the Charctic graph from NSIDC for July 4:

Source: NSIDC

As you can see, the current extent is competing with the lowest extents ever recorded for this date.The actual extent NSIDC reports for July 15 this year is 7.857 million square kilometers, which is more than 1.7 million square kilometers below the 1981-2010 average extent of 9.564 million square kilometers for that date.

This is the sea extent as shown by the Climate Change Institute (click on the image to view a larger version):

Source: CCI

The light blue line shows the 1979-2000 average sea ice extent. The ice-covered area is designated with a white-to-gray coloration. The more gray it is, the more broken-up the ice is with solid ice shown as pure white. As we can see, none, or very little, of the sea ice is shown as pure white. It is almost all some shade of gray. It also easy to see that areas that were historically ice-bound this time of year are now ice free. Take a look at Hudson Bay. It is almost entirely ice free (and has been for several weeks now), but historically would still have extensive amounts of ice. The Northwest Passage across northern Canada and Alaska is almost open, which would be a historically early occurrence for this rare event.

Here is the map of the sea ice extent from the Danish Polar Portal:

Source: Polar Portal

Again, the shade of the ice indicates the amount of coverage with sold ice being pure white. We can see, once again, that there is very little area that is pure ice. Most of the Arctic sea ice is broken up to various degrees. This is important because breaking the ice up into smaller pieces increases the surface area and makes it easier for it to melt. Broken ice will melt faster than solid ice.

But, extent isn't the only factor we need to consider. We also need to look at the ice thickness. This is the sea ice thickness as provided by the Polar Institute with the scale shown on the bottom of the image:

Source: Polar Portal

The cooler the color, the thinner the ice. This shows the ice is not only breaking up, but is also very thin.

Comparing these graphs makes me believe we will see all of the dark blue, purple and violet areas in the above graph melt by the end of the melt season in September. The red and white areas will survive. At this time of year, I think the green areas will, too, but in a much thinner state. Which leaves the light blue areas and I think about half of those areas will melt. On the map above, I estimate the minimum extent will still cover about half of the circular area around the pole and fill approximately two of the 10-degree blocks south of that circle. Using the same math I used in my March 22 posting, this comes out to be about 2 million square kilometers for the northernmost circle and about 325,000 square kilometers for each of the squares. This comes out to a total of approximately 2.7 million square kilometers.

A major factor is the weather. The Arctic weather started the year much warmer than average, but it has been mild-to-cold this summer and not as conducive to melting sea ice as in previous years. Here is a graph from NSIDC showing the cumulative freezing days starting on July 1 of the previous year and continuing to July 1 of the current year. The shaded areas show the percentile ranges (as listed) over the period 1981 through 2010:

Source: NSIDC

We can see the 2017 season has been significantly warmer than any other year and has been far outside of the percentile ranges. But, we can also see the difference from the base-line has decreased in recent weeks.


The weather factor is discussed at length in this PIOMASS posting. In particular, take a look at the graph of the Daily Average Arctic Sea Ice Thickness From PIOMASS located at the bottom of the posting, right before the comments section. This graph shows the thickness is not only historically low, but it has been all year. But, one of the things documented in this post is that the difference between 2017 and 2012 (the year of record low extent) is closing, where 2017 was much lower than 2012 to begin with. Using this input, let's postulate that the ice melt will not occur quite as severely as I speculated above. Based on that input from PIOMAS and NSIDC, let's bump the minimum extent estimate up to an even 3.0 million square kilometers.

This estimate is consistent with my March estimate and would be dangerously lower than the previous lowest minimum extent of 3.4 million square kilometers that occurred in 2012.