Title: Comparison of balance and flow-dependency of large-scale background-error variances in two ensembles

Authors: Nedjeljka Zagar (University of Ljubljana, Ljubljana, Slovenia)

Time-averaged and time-dependent structure of short-term forecast-error variances is investigated in two global ensemble data assimilation systems, the ECMWF 4D-Var ensemble and the ensemble adjustment Kalman filter DART/CAM. The applied methodology of the normal-mode function expansion provides an attractive way to measure the balance by splitting forecast-error variances into parts projecting on the balanced and inertio-gravity (IG) circulations, the approach particularly suitable for the tropics where the IG circulation dominates on all scales. The flow dependency is quantified by the flow-dependency coefficient which measures correlation between the forecast-error variances and the mean energy in two-dimensional modal subspaces. Similarities and major differences between the tyo systems over the same time period are presented and discussed. In particular, it is described how the applied inflation field in the ensemble Kalman filer assimilation has a major impact on the structure of the background variance field and its reduction by the assimilation step. A perfect-model assimilation experiment supports the findings from the real-observation experiment.

GMAO Head: Michele Rienecker Global Modeling and Assimilation Office NASA Goddard Space Flight Center

Curator: Nikki Privé Last Updated: May 27 2011