@article{oai:nipr.repo.nii.ac.jp:00015935,
 author = {De Silva, Liyanarachchi Waruna Arampath and Yamaguchi, Hajime},
 journal = {Polar Science},
 month = {Sep},
 note = {Recently, cyclones are quite active and dramatically influence the sea ice distribution in the Arctic region. Therefore, precise prediction of sea ice distribution during the cyclones is crucial for safe and efficient navigation in the Arctic Ocean. A high-resolution (about 2.5 km) ice–ocean coupled model is developed for forecasting the short-term sea ice distribution along the Arctic sea routes. Since a commercial ship navigates to avoid the sea ice as much as possible, a factor to score the forecast skill is considered to be ice edge error which is an averaged distance between forecasted and observed ice edges. First, we discuss the influence of horizontal grid resolution on short-term sea ice predictions during May to November 2015. The model grid resolution is varied from 2.5 km, 5 km, 10 km and 15 km. This grid dependency study suggests that the 2.5 km resolution model predicts the ice edge accurately compared to the satellite observations. Second, the primary experiment was run from 2 August 2016 to 7 August 2016 under the extreme cyclone in the Laptev Sea using ensemble predictions. The ensembles are constructed by using forecasted atmospheric forcing data sets from THORPEX Interactive Grand Global Ensemble (TIGGE) project and European Center for Medium-Range Weather Forecast Interim (ERA-Interim) reanalysis data. The ensemble ice edge errors are computed from 2 August 2016 to 7 August 2016. The maximum forecast skill of ensemble average ice edge error in the ice–ocean coupled model is 11.74 ± 0.54 km with the threshold of 15% ice concentration for the AMSR2 and model predicted ice edges. It can be said that the present model of 2.5 km grids satisfies the ship crew requirement of around 10 km ice edge error for 5-day forecast.},
 pages = {204--211},
 title = {Grid size dependency of short-term sea ice forecast and its evaluation during extreme Arctic cyclone in August 2016},
 volume = {21},
 year = {2019}
}