A modified POD procedure with patterns in time for parameter estimation

Proper orthogonal decomposition (POD) is a well established model order reduction technique, however its efficiency is limited for advection dominated model. In this report, we explore a modified proper orthogonal decomposition procedure with patterns in time instead of the usual patterns in space. This modified procedure is more suitable for this kind of model. Its efficiency is studied in the context of parameter estimation.

Subject

master Thesis
proper orthogonal decomposition
data assimilation
Kalman smoother
ensemble Kalman smoother

To reference this document use:

http://resolver.tudelft.nl/uuid:0d54e5e6-133b-4c7b-a2b7-6be6e9640ed0

Part of collection

Student theses

Document type

master thesis

Rights

Files

PDF

report_TPOD.pdf

13.61 MB

Close viewer