event
Tuesday 25 Mar 2014: CliMathNet e-seminar: Accounting for model error due to unresolved scales within ensemble Kalman filtering
Lewis Mitchell - University of Vermont
Harrison 209 15:00-16:00
We propose a method for accounting for model error due to unresolved scales within the context of ensemble Kalman filtering. This method estimates a model error correction to the forecast step by using historical reanalysis increments to build a model error covariance matrix. We compare two different versions of the method; a time-constant model error treatment where the same model error bias correction is added after each forecast, and a time- varying treatment where the bias correction randomly varies with each forecast. We compare both methods with the standard method of dealing with model error through inflation and localization, and illustrate our results with numerical simulations on a low order nonlinear system showing chaotic dynamics. The results show that the filter skill is significantly improved through the proposed model error treatments, and that both methods require far less parameter tuning than the standard approach. This is joint work with Alberto Carrassi.