Observations in the Arctic have shown that sea ice plays a pivotal role in Arctic and global climate change, not only affecting sea-air interactions but also biochemistry. The already observed and further anticipated decline in sea ice will also affect the exchange of climate-active trace gases (CO 2 , CH 4 , O 3 and DMS), which is one of the foci of the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) project. In this thesis we present a modelling study of sea ice processes that are relevant for climate-active trace gas exchange. We perform an evaluation and sensitivity study of the state-of-the-art Los Amalos Sea Ice Model (CICE) with a focus on the essential thermodynamic and dynamic processes driving this exchange. The simulated one-and three-dimensional ice fields are evaluated for the period of the Surface Heat Budget of the Arctic Ocean (SHEBA). The 1D simulations show a simulated ice thickness evolution that agrees well with the SHEBA observations (RMSE 0.34 m), however they also show their sensitivity to the atmospheric forcing data. Using climate-model derived forcing data instead of local observations shows the strong impact of snowfall on ice growth, albedo, onset of the melt season and, consequently, total surface melt. Besides, including ice opening rates in the 1D simulation increased the bottom and lateral melt with 11 and 30 cm/yr respectively. Subsequently, the 3D simulated Arctic ice field agrees on average well with satellite-derived ice concentrations (RMSE of 0.1), but shows larger differences in the marginal ice zone around Svalbard and Greenland. The ice concentration shows a strong sensitivity to the oceanic forcing data where the restoring timescale of sea surface temperature strongly controls the formation of new ice. Both the 1D and 3D simulations show that simulated melt pond areas are significantly different when using different parameterizations, influencing in turn the summer albedo and surface melt. Our results show that important features for climate-active trace gas exchange such a open water fractions and ponds are represented well in CICE, though being very sensitive to the forcing data sets and choice of parameterizations. As such, this study has provided valuable information regarding the required data to optimally constrain CICE simulations for its potential follow-up application studying sea ice biogeochemistry and climate-active trace gas exchange as observed in the MOSAiC field campaign.