Wir bieten regelmäßig MSc-Arbeiten im Bereich der geologischen Modellierung mit quantifizierten Unsicherheiten, der geophysikalischen Inversion, Nutzung von AR/VR in den Geowissenschaften, sowie der effizienten Simulation mit "Reduced Basis"-Methoden an.
Bitte wenden Sie sich direkt an Prof. Wellmann für weitere Informationen.
Process Simulations and Reduced Order Modeling
We offer a wide range of topics in the field of process simulations and reduced order modeling. For the reduced order modeling, we focus on the reduced basis method. We use the method to considerably speed-up our process simulation, allowing in turn, probabilistic model approaches.In these application fields, we cover three main categories:
- Process Simulations: Here, we mainly focus on geothermal and groundwater applications. However, it is also possible to define a thesis in other process simulation applications.
- Inverse Methods: Here, we mainly cover the fields of optimization, data assimilation, uncertainty quantification, and optimal experimental design.
- Combining structural and process uncertainties: This field includes the meshing algorithms and isogeometric analysis.
A thesis in this field contain usually at least two of these three main directions. For a thesis in this field, you should be motivated to learn to program. The level of theory and programming required to perform the thesis is adjustable in accordance with the background of the student.
For more information, please contact Dr.Denise Degen
The estimation of the spatial distribution of geophysical properties in the subsurface has been a major field of research in geosciences for more than half a century. At CGRE we strive to combine well-known geostatistical methods, like varying forms of Kriging, with state-of-the-art implicit structural geomodelling to improve predictions of these spatial distributions. Possible theses topics include the adaptation and application of these methods to specific real-world scenarios as well as the development of innovative workflows to improve prediction quality. Interest in computational geosciences and motivation to learn and use Python is required. Expertise in other fields like structural or economic geology is helpful for the investigation of specific modelling scenarios and interdisciplinary approaches.
For more information, please contact Jan von Harten.
Generally, there are two types of surface representation; parametric and implicit representation. Parametric surfaces are kind of smooth and controllable surfaces based on mathematical rules which are common methods in CAD, animation, gaming and other industries. We try to use this method in geological modelling.
For more information, please contact Mohammad Moulaeifard
Probabilistic geomodelling is an active research area at CGRE institute. We apply Bayesian inference to reduce the model uncertainty by combining our geological knowledge and geophysical measurements, and quantify the uncertainties by methods such as Markov chain Monte Carlo (MCMC). We intend to introduce some recent-developed advanced probabilistic methods into the application of geomodelling. We offer a range of thesis topics in this direction. For a thesis in this field, you have opportunities to learn some cutting-edge techniques in computer science and applied mathematic, such as the popular differentiable programming package TensorFlow. One should expect to learn probabilistic theory and programming in python.
For more information, please contact Zhouji Liang