 |
| Titel |
Some experiments with artificial neural networks in data assimilation |
| VerfasserIn |
Haroldo Fraga de Campos Velho, Fabricio Harter, Rosangela Rosangela Cintra, Helaine Furtado |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250052473
|
|
|
|
|
|
| Zusammenfassung |
| Abstract:
Data assimilation is an essential step for operational forecasting
systems by means of a weighted combination between observational
data from a mathematical model. Artificial neural networks (ANN)
have been proposed as a new technique for data assimilation. The
new method is presented with applications on Lorenz system under
chaotic regime, atmospheric models, and space weather (the latter,
a three-wave model of auroralmradio emissions).
The performance of the ANN isevaluated with different data
assimilation methods: Kalman filter (KF), variational method,
and particle filter. In addition, we explore two ANN implementations:
multilayer perceptrons, and radial base function. |
|
|
|
|
|
|