1). A 16 km spatial resolution wave model of the Arctic Ocean was developed to investigate the predictability of surface ocean waves in ice free waters. The model has two forcing: surface wind and sea ice concentration, for which we use the NCEP CFSvs2 hourly and daily products, respectively. In ice-covered waters, the model utilises primitive fractional blocking of wave energy according to the ice concentration. 2). The TodaiWW3-ArCS was validated with a driting wave buoy for 2 months during September and October, 2016 in the Beaufort Sea. It achieved good agreement, however, a model underestimation bias was apparent for a storm in October. Analysis led us to a conjecture that the reduced number of surface level pressure observations, in October compared September, that are assimilated in numerical weather prediction models caused increased uncertainty in surface wind forcing, which led to increased uncertainty in the modelled wave estimates.

Nose et al. 2018 Ocean Dyn; DOI 10.1007/s10236-018-1194-0