Much of the work at the institute thus revolves around the use and development of high-resolution mesoscale models. In contrast to mesoscale circulations, boundary layer turbulence is not resolvable in models, nor is it deterministic in its details. The statistics of turbulent flows is, however, to some degree deterministic. Much of the work in this field is directed at determining the effects of turbulence on the mean flow, to improve larger-scale numerical models. Conversely, this means that one must also understand the effect on the turbulence of the mean conditions. Much of the work done at the institute involves participating in, and using data from, field campaigns.

Work on turbulence dynamics has been directed at three areas: the Arctic boundary layer, interaction between turbulence and clouds and boundary layers in high static-stability conditions. The work is carried out within programs such as GABLS, and also involves field experiments in the Arctic, for example ASCOS and SHEBA.