Landfast ice in the Arctic: parameterization in numerical modeling and effects on the hydrography

Abstract

The Arctic amplification of climate change is strongly connected to the sea ice in the Arctic because of the sea ice albedo and insulation effects. This thesis presents numerical modeling studies on landfast ice and the kinetic energy dissipation within sea ice in the Arctic. Landfast ice is sea ice that is nearly immobile and attached to the coast. As an extension to the land, landfast ice provides a platform for tourism, transportation, and hunting for the local communities, as well as for oil and gas drilling and scientific observation. Although landfast ice plays an important role in the Arctic coastal regions, it is challenging to represent landfast ice in numerical sea ice models with a standard viscous-plastic (VP) rheology. In this thesis, I implement a lateral drag parameterization to improve the representation of the landfast ice in the sea ice model and study the kinetic energy dissipation by sea ice internal stress. This thesis has three scientific questions: (1) How can we improve the representation of landfast ice in the sea ice numerical model? (2) What is the influence on the Arctic Ocean hydrography with the presence of landfast ice cover in the sea ice model? (3) Does the kinetic energy dissipation depend on the details of the VP yield curves? With higher resolution, the heterogeneity of the sea ice cover increases and deformation localizes at distinct linear kinematic features. Does this change have an effect on the mechanical energy cycle?

Firstly, coastal drag parameterization is implemented in the coarse sea ice model, as a function of coastline length, sea ice thickness, and sea ice drift velocity. The results show that the new parameterization improves the simulation of landfast ice in deep marginal seas of the Arctic. Furthermore, the combination of lateral and basal drag parameterization simulates realistic landfast ice distributions in the Arctic.

Secondly, the landfast ice cover is switched on and off in the sea ice model to study the influence of the landfast ice on the Arctic hydrography. The results show that the presence of the landfast ice decreases the salinity in the upper ocean due to the reduced new ice formation releasing less salt in the water. Furthermore, the fresher upper ocean signal formed in the deep marginal sea is transported from the coast to the central Arctic, leading to a fresher upper ocean in the Makarov Basin. The salt budget analysis demonstrates that the salt is advected of the Arctic through the boundaries.

Lastly, the kinetic energy dissipation with different yield curves and three model resolutions is evaluated. The standard ellipse, Mohr--Coulomb yield curve with elliptical plastic potential, Truncated Ellipse Method and Teardrop yield curves are employed in the 4.5 km sea ice model to compare the sea ice thickness, drift velocity, and kinetic energy dissipation differences. The results indicate that the simulations with yield curves with more shear strength lead to smaller sea ice drift and hence to smaller wind energy input and energy loss due to ocean drag. Moreover, the impact of the different yield curves on the net energy dissipation is small, but simulations with similar yield curves have similar kinetic energy dissipation within the ice. Finally, experiments with different model resolutions suggest that the higher the resolution of the simulation, the more deformation and dissipation are localized along shear lines. More localization leads to smaller mean drift, less kinetic energy input, and loss by ocean drag.