Ensemble-based data assimilation at a coastline Conference Paper uri icon

abstract

  • The potential of the Ensemble Kalman filter in the area of model error estimation and reduction in a sea-breeze environment was analyzed. The test was performed after a 12-hour model run using 20 ensemble members and simulated buoyancy observations placed at 20km spacing in the horizontal and 250m in the vertical. The results show that some noise is introduced at the surface, especially around the sea-breeze front where phase difference between the mean forecast field and the truth field is most pronounced. It was suggested that, although this noise does not influence the domain-wide performance of the filter, it does introduce some unbalanced structure.

author list (cited authors)

  • Aksoy, A., Zhang, F., Nielsen-Gammon, J. W., Epifanio, C., & Snyder, C.

publication date

  • June 2004