The Arctic sea ice cover is crisscrossed by leads and pressure ridges, often referred to as linear kinematic features (LKFs). Leads are areas of open ocean in the sea ice cover. As such they play a crucial role in the Arctic climate system because they accommodate large amounts of the Arctic heat loss and thus enable sea ice formation. Hence a proper representation of LKFs in a climate model is expected to improve its overall performance. So called viscous-plastic (VP) sea ice models are used in most climate models and are thought to represent the sea ice accurately on large scales. Here, the spatial resolution is pushed to scales of 1 km to show leads emerging in a VP sea ice model at those very high resolutions. In order to avoid unnecessary computational limits, first steps are taken with an idealized environment set up implemented with the MIT General Circulation Model (MITgcm). The modeled sea ice deformation is compared to observational data by a statistical analysis and using the fractal characteristics of sea ice deformation. The probability distributions of the deformation rates have power-law tails in attraction of the Lévy law at high resolutions. This implies that, the localization of strain rate events increases with refining spatial resolution. In addition, it is found that the ice strength depends on the spatial resolution. So that, the ice strength P* needs to be reduced with increasing grid size to obtain comparable dynamic behavior of the modeled ice. Moreover, spatial and temporal scaling laws are explored for sea ice deformation. By increasing the resolution of the wind forcing more small scale strain rate events are induced in the ice leading to a better agreement with fractal characteristics of spatial scaling laws. So that VP becomes comparable to the elasto brittle (EB) rheology. Areas of low sea ice concentration dominate temporal scaling laws. In dense pack ice, deformation events persist over a period of ten days at small spatial scales. In the second part the predictability of the modeled leads is investigated. The position of modeled LKFs is strongly sensitive to uncertainties in the wind forcing and initial conditions. Uncertainties in satellite sea ice products and weather forecast lead to a predictive skill of 1 to 2 days.