Sea ice roughness: the key for predicting Arctic summer ice albedo

Although melt ponds on Arctic sea ice evolve in stages, ice with smoother surface topography typically allows the pond water to spread over a wider area, reducing the ice-albedo and accelerating further melt. Building on this theory, we simulated the distribution of meltwater on a range of statistically-derived topographies to develop a quantitative relationship between premelt sea ice surface roughness and summer ice albedo. Our method, previously applied to ICESat observations of the end-of-winter sea ice roughness, could account for 85% of the variance in AVHRR observations of the summer ice-albedo [Landy et al., 2015]. Consequently, an Arctic-wide reduction in sea ice roughness over the ICESat operational period (from 2003 to 2008) explained a drop in ice-albedo that resulted in a 16% increase in solar heat input to the sea ice cover. Here we will review this work and present new research linking pre-melt sea ice surface roughness observations from Cryosat-2 to summer sea ice albedo over the past six years, examining the potential of winter roughness as a significant new source of sea ice predictability. We will further evaluate the possibility for high-resolution (kilometre-scale) forecasts of summer sea ice albedo from waveform-level Cryosat-2 roughness data in the landfast sea ice zone of the Canadian Arctic. Landy, J. C., J. K. Ehn, and D. G. Barber (2015), Albedo feedback enhanced by smoother Arctic sea ice, Geophys. Res. Lett., 42, 10,714-10,720, doi:10.1002/2015GL066712.