Ensemble‐based simultaneous state and parameter estimation with MM5

Abstract

The performance of the ensemble Kalman filter (EnKF) under imperfect model conditions is investigated through simultaneous state and parameter estimation for a numerical weather prediction model of operational complexity (MM5). The source of model error is assumed to be the uncertainty in the vertical eddy mixing coefficient. Assimilations are performed with a 12‐hour interval with simulated sounding and surface observations of horizontal winds and temperature. The mean estimated parameter value nicely converges to the true value within a satisfactory level of variability due to sufficient model sensitivity to parameter uncertainty and detectable (relative to ensemble sampling noise) correlation signal between the parameter and observed variables.