Improved representation of Arctic sea ice velocity field in ocean–sea ice models based on satellite observations

• Published in: Climate dynamics Vol. 57; no. 9-10; pp. 2863 - 2887
• Main Authors: Toyoda, Takahiro, Kimura, Noriaki, Urakawa, L. Shogo, Tsujino, Hiroyuki, Nakano, Hideyuki, Sakamoto, Kei, Yamanaka, Goro, Komatsu, Kensuke K., Matsumura, Yoshimasa, Kawaguchi, Yusuke
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.11.2021; Springer; Springer Nature B.V
• Subjects: Analysis; Arctic Ocean; Arctic sea ice; climate; Climate models; Climate prediction; Climatology; Coefficients; Constants; Earth and Environmental Science; Earth Sciences; Environmental aspects; Error reduction; Geophysics/Geodesy; Ice models; Ice pressure; latitude; Momentum; Ocean models; Oceanography; Oceans; Parameter sensitivity; Parameters; prediction; reproduction; Satellite observation; Satellites; Sea ice; Sea ice concentrations; Sea ice dynamics; Sea ice models; Velocity; Velocity distribution; Arctic Ocean; Arctic Ocean; Sea ice velocity; Ocean reanalysis; Data assimilation
• ISSN: 0930-7575; 1432-0894
• DOI: 10.1007/s00382-021-05843-4

Accurate description of sea ice velocity is necessary for improving the reproduction and prediction of high-latitude climate in modeling studies. In order to improve the simulated sea ice velocity field based on satellite-derived daily data, a set of sea ice dynamic parameters of the ocean–sea ice model was optimized with a Green’s function method, which uses linear combination of forward sensitivity operators for these parameters. An improved sea ice velocity field with 15–20% error reduction was obtained for both simulations with and without the data constraint of sea ice concentration. Moreover, dependencies of the sea ice dynamic parameters on background conditions were examined and linear dependency coefficients were also optimized by extending the above method. The sea ice velocity field was further improved for both free-drift and high-pressure regions with mean error reduction of about 25%. This resulted in decreases of surface salinities in the marginal ice zones, making them closer to observations. While the parameter adjustments as constants tended mainly to compensate influences of biases in surface forcing, the obtained dependencies were generally consistent with previous observational and theoretical studies and thus improved parameterizations of air–ice–ocean momentum exchange and ice pressure which are widely used in OGCMs. The enhanced ocean–sea ice initialization and parameterizations would be useful for operational seasonal prediction and climate prediction studies.