Lagrangian tracking of sea ice in Community Ice CodE (CICE; version 5)

Abstract. Sea ice models are essential tools for simulating the thermodynamic and dynamic processes of the sea ice and the coupling with the polar atmosphere and ocean. Popular models such as the Community Ice CodE (CICE) are usually based on non-moving, locally orthogonal Eulerian grids. However, the various in-situ observations such as ice tethered buoys and drift stations, are subjected to sea ice drift and hence by nature Lagrangian. Furthermore, the statistical analysis of sea ice kinematics requires the Lagrangian perspective. As a result, the offline sea ice tracking with model output is usually carried out for many scientific and validational practices. Certain limitations exist, such as the need of high frequency model outputs, as well unaccountable tracking errors. In order to facilitate Lagrangian diagnostics in current sea ice models, we design and implement an online Lagrangian tracking module in CICE under the coupled model system of CESM (Community Earth System Model). In this work, we introduce its design and implementation in detail, as well as the numerical experiments for the validation and the analysis of sea ice deformations. In particular, the sea ice model is forced with historical atmospheric reanalysis data and the Lagrangian tracking results are compared with the observed buoys' tracks for the years from 1979 to 2001. Moreover, high-resolution simulations are carried out with the Lagrangian tracking to study the multi-scale sea ice deformations modeled by CICE. Through scaling analysis, we show that CICE simulates multi-fractal sea ice deformations in both the spatial and the temporal domain, as well as the spatial-temporal coupling characteristics. The analysis with model output on the Eulerian grid shows systematic difference with the Lagrangian tracking-based results, highlighting the importance of the Lagrangian perspective for scaling analysis. Related topics, including the subdaily sea ice kinematics and the potential application of the Lagrangian tracking module, are also discussed.