Abstract
Over the past 25 years, the minimum sea ice extent has reduced from 6.5 million km2 to 4.5 million km2 in the Arctic Ocean. Waves generated in the emerging open water area are increasing in height, and the Marginal Ice Zone where waves and sea ice interact extensively is enlarging. A coupled prediction of wave and ice is crucial for the development of the Arctic Shipping Route. Under the Arctic Challenge for Sustainability I and II, we have conducted a wave forecast (from 2019) and sea ice forecast (from 2020) in support of the R/V Mirai cruise. Concurrently, wave buoy measurements were made in the Chukchi Sea and the Beaufort Sea in 2016 to validate the models. IcePOM originally developed by a group led by Yamaguchi at the U. of Tokyo has been updated to incorporate nudging of the ocean interior to the RIOPS temperature and salinity fields. The result shows that the RIOPS nudging run outperforms the run without ocean interior nudging. Moreover, the forecast improved by using a reanalysis wind rather than a forecast wind. The study implied that for a regional model, a large-scale coupled forcing is crucial. Given that, a ROMS-Budgell Arctic model was developed. The model was used and compared against IcePOM with the same initialization and forcing by GIOPS and ERA-5. The comparison shows differences between them leaving a large uncertainty in the ice forecast. The impact of sea ice concentration is notable not just among the models but also in the satellite-retrieved data. Different wave hindcast was made for different ice forcing and the resulting uncertainty superseded the uncertainty related to the wave-ice interaction. Wave hindcast uncertainty is subject to atmospheric forcing uncertainty as well. Overall, the study suggests that the regional wave-ice coupled model should be constrained by a large-scale coupled model.
