H/T to Andrew for alerting me to this new paper published in Quaternary Science Reviews.
From the paper:
To investigate the feedback processes linking solar activity and sea-ice cover, we used the coupled climate model COSMOS, which indicates that a decrease in solar radiation results in increased sea-ice cover (Fig. 7a) and decreased sea-surface temperature (Fig. 7b). A strong negative correlation between sea-ice variability and solar forcing is observed along the eastern and southwestern coast of Greenland and in the Arctic Ocean, indicating that in this model solar variability is critical for simulating changes in local sea-ice production. A small change in incoming shortwave radiation, and associated ice-albedo effects, resulted in a large response of local ice formation, according to ‘bottom-up’ (solar heating of the sea surface) mechanisms (Gray et al., 2010; Hunke et al., 2010).
In addition, the significant 216-year period is similar to the well known de Vries solar cycle of ~210 years, while the periodicity of 26 years may be related to the periodicity of the 22-year Hale cycle (Fig. 4). We note, however, that the latter periodicity (~26 years) lies close to the Nyquist period and therefore may be an artefact of the sampling frequency. Notably, the 56-year cycle in our sea-ice record is not associated with any well-known solar cycles, but falls within the range of the 55e70 year spectrum calculated for the Atlantic Multidecadal Oscillation (AMO) over the last 8000 years (Knudsen et al., 2011), although the sea-ice variability linked to solar forcing may also have impacted the AMO cycle (Knudsen et al., 2014).