Numerical Methods & Scientific Computing

Weather and climate modeling is an interdisciplinary endeavor involving not only atmospheric science, but also applied mathematics and computer science. Atmospheric models cover a wide range of spatial and temporal scales that require robust multi-scale numerical schemes.

In Climate & Space, we investigate modern numerical techniques that are suitable for weather and climate models, remote sensing and statistical applications. We develop and improve numerical schemes for partial differential equations on the sphere (the so-called dynamical cores) and collaborate closely with US modeling centers like the National Center for Atmospheric Research (NCAR), NASA or NOAA laboratories like the Geophysical Fluid Dynamics Laboratory (GFDL). Examples of some particular research areas are Adaptive Mesh Refinement (AMR grids), Cubed-Sphere grids and dynamical core inter-comparisons. In addition, our research addresses high-performance and parallel computing aspects, as well as software and data management for atmospheric models.

The Center for Space Environment Modeling (CSEM) develops high-performance, first-principles based computational models of the space environment and uses these models to predict “Space Weather”, to understand space mission data and to further our understanding of the solar system. The primary focus of CSEM is developing highly accurate numerical models of the space environment using state-of-the-art numerical techniques. We strive to make these models and results available to the community through our collaborations, toolkits and analyses. Through the many projects that we participate in, CSEM models are used to test and predict the behavior of many real-world space-systems. Our focus areas include the space environments of the Sun, the Earth, solar system planets and their moons, comets and exoplanets.