Details of the Abstract
| Title of paper | Application of Ensemble Kalman Inversion and Model Reduction Using Discrete Cosine Transform in Geo-Electromagnetics |
| List of authors | Börner, R.-U. |
| Affiliation(s) | TU Bergakademie Freiberg, Institute of Geophysics and Geoinformatics, Germany |
| Summary |
Ensemble Kalman Inversion is a powerful iterative method that assimilates observational data with model predictions to update the subsurface property estimates. By leveraging the statistical properties of an ensemble of models, EKI effectively mitigates the effects of noise and provides robust inversion results. However, the high dimensionality of the model parameterization can make the inversion process computationally expensive. To address this issue, we integrate model reduction techniques using the Discrete Cosine Transform. DCT efficiently compresses the data by transforming it into a frequency domain, where the significant components are preserved while discarding the redundant information. This reduction not only decreases the computational load but also enhances the stability and convergence of the inversion process. |
| Session Keyword | 2.0 EM theory, modelling and Inversion |
| File upload |
2.0_application_of_ensemble_k_borner.pdf
|