Back in 2014 I suggested that there may be an oscillation of extent loss. According to the supposed oscillation this year should have seen a large loss in August, this didn't happen.
Looking at all regions within the Arctic Ocean it can be seen that an oscillation with a four year periodicity seems to emerge after 2000.
The peaks can be seen better in a plot of post 2000 extent loss over August.
The previous peaks are 2004, 2008, 2012, and now we have 2016. In 2016 August extent loss was similar to that in 2015.
Amalgamating the above regions into three larger regions makes the pattern more clear.
What happened this year was a failure of large losses in Chukchi and the East Siberian Sea (ESS). So what happened? Looking at extent in Chukchi and the East Siberian Sea August started with lower extent than the other peak loss years.
However, I've been away from this subject so haven't followed the weather, but August looked like a rather indifferent ice loss month from the point of view of the weather. So this may be a combination of both factors.
The test of whether there is a real cycle continues, next year should see a reduced extent loss in August, the following year should be rather lower before the next two years are higher. It would help if I could put forward a mechanism to test, but I still haven't figured that out.