Observation and simulation of landfast sea ice using thermodynamic modeling in Prydz Bay, East Antarctica

Landfast sea ice plays a crucial role in the Antarctic coastal ecosystem and iceberg dynamics. This study employs a single-column thermodynamic model to simulate landfst sea ice near Zhongshan Station from 2016 to 2017. Forcing variables, including wind, temperature, specific humidity, and radiation, were obtained from in-situ automatic weather stations (AWS) and reanalysis atmospheric data (ERA5 and JRA55). Although there was alignment with observations, reanalysis data consistently overestimated precipitation, with a correlation coefficient of less than 0.8. Sensitivity experiments were conducted by varying precipitation inputs. In Group 1, the simulated sea ice thickness showed minimal biases of 6.3 cm and 5.3 cm when using AWS and ERA5 data, respectively. Additionally, the sea ice temperature simulations in Group 1 were within 0.3 degree celsius of the observed values. The minimum bias in simulated sea ice thickness using JRA55 across all experiments was 10.9 cm. Overestimated snow depth, particularly with JRA55 data, resulted in thicker snow ice. Excessive precipitation led to overestimated snow depth, which revealed the dual impact of snow. Excessive snow depth led to greater snow ice formation, while reduced snow depth resulted in colder inner sea ice temperatures, promoting more bottom sea ice growth. Bottom melt primarily reduced sea ice thickness, while lateral melt reduced sea ice fraction, with latent heat flux dominating energy loss during melting. This study underscores the significance of snow depth, largely influenced by precipitation, in understanding the ablation process of landfast sea ice. Despite the discrepancies between simulations and observations, which indicate a need for refinement, this study highlights the potential of reanalysis data in sea ice thermodynamic simulations.