I just did some big data treatment on extent, volume and thickness [Dicke], divided into 8 packs, each of them including five years, to look for the long-term trends on extent, volume and thickness of Arctic sea ice. To do this I calculated the average of 1979 to present for every month and averaged the deviations for each month ("Abw." in the graph) for each of the five-year-packs of these three measures.
Whereas extent seems to decrease more or less constantly in each five-year period, volume doesn't do that in the same fashion, thus thickness (calculated from division of volume by extent) also follows a more step-wise decrease than a "smoothy" one. For a better visualisation I used a second (right) y-axis for the deviation of volume.
See attached graph.