Topic: Non-linear Data Assimilation

Thursday, May 08, 2025 from 15-17h UTC

Non-linear Data Assimilation

Organizers and Conveners: Nachiketa Chakraborty (University of Reading, UK)
Hristo Chipilski (Florida State University, US), Javier Amezcua (U. Iberoamericana, Mexico; U. Reading, UK)

Traditional data assimilation methods are optimal under the assumption of linear relationships between the observations and the state variables (observation operator), linear forecast models, as well as Gaussian sources of error. Linearizations are performed when these requirements are not met. However, with increasing resolution and complexity in the models and observations, these linearizations can fail, and one may need to use nonlinear methods. We invite contributions on the areas of particle filters, methods derived from measure transport theory, and methods which leverage tools coming from the AI/ML community.

*** OPEN CALL FOR ABSTRACTS (click here) ***

Presentations:

• t.b.a.

Time Zones:
15 – 17 UTC
Europe:           04 – 06 pm BST (London)     | 05 – 07 pm CEST (Berlin)
Asia/Australia: 11 – 01 am CST (Shanghai)  | 00 – 02 am JST (Tokyo)      | 01 – 03 am AEDT (Sydney)
Americas:        08 – 10 am PDT (San Fran.) | 09 – 11 am MDT (Denver)   | 11 – 01 pm EDT (New York)

How to attend to ISDA-Online events?
Simply subscribe to the ISDA-Online mailing list