Title: Modelling of flow-dependent ensemble-based background error correlations using a wavelet formulation.

Authors: Hubert Varella (Meteo-France)
Loik Berre (Meteo-France)
Gerald Desroziers (Meteo-France)

A wavelet formulation on the sphere is considered for modelling heterogeneous background error correlations for the Meteo-France global Numerical Weather Prediction model. This approach is compared to the operational spectral formulation, which is horizontally homogeneous to a large extent. Moreover, while this wavelet formulation is used operationally at ECMWF to specify heterogeneous but static correlations, a flow-dependent wavelet representation of correlations is investigated in this study.

Firstly, diagnostic studies have been conducted to examine geographical variations of 3D correlations over the whole globe. The results show the ability of wavelets to represent horizontal and vertical heterogeneities of correlations. For instance, the contrast between relatively broad horizontal correlations in the tropics and sharp ones in the mid-latitudes is well represented by the wavelet formulation.

The impact of the flow-dependent formulation on the forecast quality has also been investigated during two different periods, with different levels of cyclonic activity in the North Atlantic. While the impact of the flow-dependent formulation is globally positive during the two periods, the impact over Europe tends to be more spectacular during the cyclonic period. This seems to be related to spatio-temporal variations in correlation length-scales over the North Atlantic, which are well captured by the wavelet formulation.

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

Curator: Nikki Privé Last Updated: May 27 2011