Stephan Juricke

profile image

Stephan Juricke

Postdoctoral Research Assistant


I am postdoctoral research assistant working with Tim Palmer in the Predictability of Weather and Climate group and Laure Zanna in the Climate and Ocean Physics group. My research focuses on the development of stochastic parametrizations in ocean and sea ice models, to include measures of uncertainty as well as better representations of sub-grid scale variability in current climate models.

Research Goals

Stochastic parametrizations can help to improve the representation of small scale, high frequency, unpredictable processes in a statistical sense, rather then trying to actually resolve these processes or their average impact using classical deterministic sub-grid scale schemes. While they are already widely used in weather forecasting, the application of these schemes in models of other climate system components such as the ocean and sea ice are fairly recent. Stochastic schemes can provide an estimate of uncertainty originating from the diversity of small scale processes given similar large scale background states. Furthermore, they can also help to improve the model in terms of its predictive skill and biases. In a non-linear, chaotic system such as the atmosphere, ocean and sea ice, small scale fluctuations can lead to changes in the large scale mean state. As a consequence, using stochastic parametrizations to better represent the statistical properties of small scale processes can change and potentially improve model biases as well as the skill of probabilistic forecasts.
Especially on longer timescales, from seasons to decades, an adequate simulation of the ocean and sea ice becomes crucial in the development of skilful coupled climate models. This is true in terms of reducing model biases, representing climate variability accurately and providing measures of model and, consequently, forecast uncertainty. This aspects are the major purposes of the stochastic parametrizations I'm working on.