Variational Data Assimilation for Numerical Weather Forecasting. |
|
|
|
|
|
|
| For the very large systems that arise in meteorology and oceanography, the available observations are not sufficient to initiate a numerical forecasting model. The technique of data assimilation enables the measured observations to be combined with the model predictions to generate accurate estimates of the expected system states - both current and future. Often there are data sparse areas where good state estimates are difficult to construct - such as in the equatorial Pacific Ocean. In 4D-variational schemes, the dynamical model equations act to spread information into these regions, but the mechanism for this is not well understood. We will describe one approach, involving singular value decompositions, that can be used to demonstrate the critical features in the process, and we will discuss the significance of the results for tracking important features such as weather fronts. |
|
|
algweb@cerfacs.fr Last Update: Apr 4, 2003 |