The project is, as it was during the first phase, organised around a synthesis module and four research modules. The synthesis module (Module D) hosts the central prediction and evaluation systems. Furthermore, during MiKlip II, Module D puts an increased focus on making the system operational both from a technical and application point-of-view. The latter aspect is covered by two user-interaction projects: one project focuses on government agencies (lead DWD) and one focuses on the private sector (lead the Climate Service Centre Germany – GERICS).
The research modules continue to pursue questions in the four main research areas that were defined at the beginning of MiKlip: Initialisation (Module A), Processes and Modelling (Module B), Regionalization (Module C) and Evaluation (Module E), as these areas adress the most crucial open questions in the field of decadal climate prediction. Nevertheless, compared to the first phase, increased focus is given to the operational and application aspects of the system; Module C concentrates on the focus region of central Europe, and module E extends its evaluation to also investigate probabilistic analyses for potential users.
The work of MiKlip II builds upon the results of the first phase and the first main step for MiKlip II was the implementation and production of the fourth generation of the decadal prediction system. The production has started in mid-2016. Once available, the output of the simulations will be shared on the project-owned server at the German Climate Computing Centre (DRKZ). Here all project partners have access to the output of all generations of the prediction system, to observational data provided by Module E, as well as access to the computing resources necessary to apply the central evaluation system.
Click on the titles to learn more about the research of the different modules:
Module A improves estimates of initial conditions and initialization procedures.
The main goal of Module B is to improve the understanding of processes leading to decadal climate variability.
Module C establishes a regionalization component to the MiKlip decadal prediction system.
Module D provides the scientific and technical infrastructure for the decadal prediction system.
Module E evaluates hindcasts from the MiKlip decadal prediction system and works towards transferring predictions from the MiKlip system into probabilistic forecast products for users.