MiKlip first phase: FLAGSHIP

Feedback of a limited-area model to the global scale implemented for decadal hind-casts and projections

The quality of decadal climate predictions strongly depends on the feedbacks between different processes and spatial scales in the models. Usually, processes, which cannot be resolved are parameterised. As a parameterisation is only an approximation, the question remains, how much information is lost by parameterising processes and whether the skill of the prediction can be enhanced, if these processes are resolved and thus better represented in the model. Therefore the key questions of the FLAGSHIP-project are: first, how does regional climate change influence global climate change and second, how do process feedbacks cross the different spatio-temporal scales?
The focus of this study is North-West-Africa. The West African Monsoon (WAM) is a multi-scale phenomenon influencing the global circulation, while the WAM itself is affected by regional aspects such as land surface types or orography. Furthermore, the WAM influences European climate.

To address the questions in the framework of FLAGSHIP a comprehensive modelling system from the global to the local scale is built, which allows for feedbacks of regional phenomena  to the global scale. The system is based on two community models, the German Weather Service regional forecast and climate model COSMO and the global atmosphere – ocean general circulation model ECHAM5-MPIOM. Both models have  been extended by the powerful Modular Earth Submodel System (MESSy) infrastructure.

After completion of the development of the two-way coupling, simulations zooming on West-Africa will be performed, enabling an evaluation of the feedback of processes on the regional scale to a global model simulation.
In addition to a mere dynamical coupling, FLAGSHIP includes atmospheric chemistry processes (implying aerosols) allowing to investigate their scale-dependent effects.
One focus of this project is dust emissions. Africa is one of the largest dust source regions and it is well known that dust alters the radiation budget substantially, which feeds back to the dynamics and thus to global circulation. Dust emissions depend on the surface properties and the surface wind speed, hence dust emissions are very differently resolved in models of different resolutions. Therefore changes in the dynamic patterns are expected, if the West-African region is better resolved compared to a pure global simulation. Aim of this study is to provide reliable evidence about the importance of these feedbacks for the skill of global decadal climate predictions.


  • Construction of a two-way dynamical and chemical coupling of ECHAM/MESSy and COSMO/MESSy
  • Construction of a two-way dynamical and chemical coupling of COSMO/MESSy and COSMO/MESSy
  • Investigation of the influence of the regionalisation on the skill of the global model based on simulations with one and two nests zooming in on West-Africa

More on FLAGSHIPNews-Icon

This description regards the project during the first phase of MiKlip. For information on Module C projects in MiKlip II, visit the MiKlip II Module C page


Institut für Physik der Atmosphäre, Universität Mainz
Dr. Astrid Kerkweg
Gregor Gläser

Institut für Physik der Atmosphäre, Deutsches Zentrum für Luft- und Raumfahrt
Dr. Patrick Jöckel