Since there is a small but significant circa 14d pseudo cyclic variation in ice cover around min/max events. This is probably due to changes in magnitude of the dirunal tides that are known to have variations of around 13.5 and 14.5 days. There can also be notable short term changes due to storminess, especially around the summer minimum. It is important to isolate this when attempting to measure the length of freezing and melting periods.

The min/max area dates were estimated by finding the first data after the zero-crossing of the time derivative of Arctic ice area.

Ice cover data was processed using a gaussian-derivative* convolution filter. This provides a more accurate estimation of the derivative function as well a providing the low-pass filtering required to remove the ripple. This was found to be the minimal length that gave stable results.

Longer filters gave very similar results but reduced the year to year variation.

Since melting season only gives one point per year, AO index was filtered with a 365 day filter, then scaled and shifted so that it neutral point was six months (0.5).

* gaussian-derivative with sigma = 13 days

The width of the central peak of the 13d gaussian is similar to that of 26 day running mean “smoother” but with a far superior frequency response.

Data for graph above in first two columns

Third col is annual change in area (unchecked: caveat emptor)

# year melting season (d_area ?)

1979 0.528417 -9.43651

1980 0.501037 -9.05165

1981 0.481872 -9.21597

1982 0.503775 -9.5304

1983 0.509251 -8.93974

1984 0.48461 -9.39725

1985 0.459968 -9.50552

1986 0.501037 -8.87501

1987 0.509251 -9.3121

1988 0.487348 -9.70018

1989 0.564009 -8.73724

1990 0.52294 -9.76096

1991 0.514727 -9.43568

1992 0.511989 -9.31876

1993 0.490086 -10.2413

1994 0.501038 -9.3856

1995 0.539368 -9.34537

1996 0.511989 -8.11779

1997 0.528416 -8.94755

1998 0.525679 -10.2618

1999 0.47092 -9.84008

2000 0.509251 -9.67688

2001 0.487348 -9.54703

2002 0.511989 -9.95161

2003 0.503776 -9.86068

2004 0.509251 -9.13722

2005 0.481872 -9.21381

2006 0.503775 -9.05517

2007 0.522941 -10.2829

2008 0.501037 -10.8469

2009 0.501038 -10.1025

2010 0.479134 -10.5337

2011 0.48461 -9.99179

2012 0.479134 -11.3853