I wrote a post recently on Antarctic sea ice where I only included sea ice extent at minimum month, mostly because it was a nice figure.
Anyway, I just noticed this nicer figure from a new paper in the International Journal of Climatology (Tareghian and Rasmussen, 2013):

Quantile regressions (20th, 50th, and 80th percentiles) and standard linear regression of monthly sea ice extent (1979–2010) for (a) the Northern Hemisphere and (b) the Southern Hemisphere from Tareghian and Rasmussen (2013)
This shows the difference in the annual trends as well as the seasonal ones.
Reference
Tareghian, R. and Rasmussen, P. (2013), Analysis of Arctic and Antarctic sea ice extent using quantile regression. Int. J. Climatol., 33: 1079–1086.