Groth, Clinton et Walia, Rayhan. 2025. « Investigation of observation quality on variational data assimilation methods ». Communication lors de la conférence : CSME-CFDSC-CSR 2025 International Congress (Montreal, QC, Canada, May 25-28, 2025).
Prévisualisation |
PDF
435 - Investigation of observation quali.pdf - Version publiée Licence d'utilisation : Tous les droits réservés aux détenteurs du droit d'auteur. Télécharger (54kB) | Prévisualisation |
Résumé
Data assimilation (DA) is a computational technique that statistically combines observational data with numerical simulations based on an underlying mathematical model to enhance prediction accuracy. A variational DA approach is described herein, with the overarching objective of investigating its predictive performance under various observational conditions. Variations in the observational data frequency, density, and uncertainty and their impact on the resulting DA predictions are all examined. The mathematical model considered here is the simplified one-dimensional partial-differential equation representative of more complex fluid dynamic transport descriptions: the one-dimensional scalar viscous Burgers’ equation. The latter has been selecteddue to both its simplicity and well-posedness. Steady and unsteady solutions of Burgers’ equation are considered with different DA objectives ranging from the optimization of initial conditions for unsteady solutions initial value problems to the optimization of boundary data in the case of solutions toboundary value problems. Beginning with arbitrary initial or boundary data, a perturbation is added to represent the inaccurate (un-assimilated) solution, and the objective of the variational DA approach is then to retrieve the “exact” solution or truth based on observations that are syntheticallyadded at some specified frequency, density, and uncertainty in terms of the desired exact solution. The present investigation provides insight into various observation-related factors affecting the robustness of variational DA algorithms. The latter will assist in future implementation of such DA methods, potentially reducing the level or amount of observational data required to achieve accurate predictions.
| Type de document: | Communication (Communication) |
|---|---|
| Informations complémentaires: | Progress in Canadian Mechanical Engineering, Volume 8. Co-chairs: Lucas A. Hof, Giuseppe Di Labbio, Antoine Tahan, Marlène Sanjosé, Sébastien Lalonde and Nicole R. Demarquette. |
| Date de dépôt: | 18 déc. 2025 14:38 |
| Dernière modification: | 18 déc. 2025 14:38 |
| URI: | https://espace2.etsmtl.ca/id/eprint/32143 |
Actions (Authentification requise)
 |
Dernière vérification avant le dépôt |