 |
| Titel |
3dVar analysis of glider data with a hybrid covariance model |
| VerfasserIn |
Max Yaremchuk, Dmitri Nechaev, Chudong Pan |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250036136
|
|
|
|
|
|
| Zusammenfassung |
| A forecast error covariance model for assimilatiing glider data is proposed. It is based on the
assumption of statistical independence of the errors described by the leading dynamical
modes and by the small-scale components of the errors’ projection on the subspace
orthogonal to that modes. The associated cost function penalizes error projections on the
modes and the magnitude of the high-pass filtered projection of the errors on the orthogonal
subspace.
The covariance model is tested by assimilating glider observations into the NCOM model.
It is shown that the proposed forecast error covariance approximation provides better forecast
accuracy compared to the Gaussian error covariance model widely used in sequential data
assimilation of oceanographic data. |
|
|
|
|
|