The Central European limited area ensemble forecasting system: ALADIN-LAEF

Yong Wang
Central Institute for Meteorology and Geodynamics

The Central European limited area ensemble forecasting system ALADIN-LAEF (Aire Limitée Adaptation dynamique Développement InterNational – Limited Area Ensemble Forecasting) has been developed at ZAMG in frame of the international cooperation ALADIN/LACE. It runs with 16 perturbed members and in 18km horizontal resolution. In the ALADIN-LAEF, two new ideas for dealing with the initial uncertainties have been proposed.  Those are: 1) Blending method is used for perturbations to initial conditions by combining the large scale perturbation generated by global Singular Vector and the small scale perturbation generated by regional Breeding; 2) NCSB (Non-Cycling Surface Breeding) technique is employed for perturbations to initial surface conditions.

    In this talk, the idea of Blending, which is based on the digital filter and spectral analysis for generation of initial perturbation in regional EPS, will be presented in detail. The performance of Blending will be compared with Breeding and dynamical downscaling.

    The lack or inadequate representation of uncertainties in the initial surface conditions is quite known in the ensemble forecasting.  In the last few years, various approaches are employed for dealing with the uncertainties that are related to the initial state of the atmosphere, but hardly any methods for perturbing the initial state of the land surface in the ensemble systems. In this talk, the new strategy for perturbing surface initial conditions (NCSB) will be discussed and evaluated.

    The performance of the ALADIN-LAEF has been evaluated by a comprehensive comparison with ECMWF EPS and a verification against the operational deterministic forecast with higher resolution for answering the questions:

    What is the more added value of regional EPS to its counterpart global EPS?
Comparison between ALADIN-LAEF (16 perturbed members, 18 km resolution) and ECMWF EPS (50 perturbed members, 45 km resolution)

    Is a regional EPS adding value to its existing higher resolution deterministic Limited Area Model (LAM) forecast?
Comparison between ALADIN-LAEF and an ensemble system based on time lagged higher resolution (9.6 km) deterministic LAM forecasts

The verification results will be shown in the talk too.