MiKlip first phase: INTEGRATION

INTEGRAted data and evaluation system for decadal scale predicTION.

The INTEGRATION project provides a standardised data and evaluation system for application within the MiKlip system. The application system enables the assessment of the MiKlip system during each of its development phases. Research progress from different modules of MiKlip directly flows into the system enabling efficient development and continuous improvement of the MiKlip system.
The application system provides efficient and comprehensive access to the model data base as well as to evaluation data sets. The system includes tools for comprehensive analysis and visualisation of results. A module for meta data is developed including detailed information on model output, data used for driving the simulations, datasets used for the calibration of the models and their components as well as basic information on validation and quality control status including information on the data used for validation of the MiKlip forecast system. The scientific development from other MiKlip modules regarding the evaluation and interpretation of decadal scale predictions supports the development of the evaluation system. Additionally, standardised methods like metrics which are suitable for decadal scale forecasts and their evaluation are developed. This enables comprehensive assessment and monitoring of the forecast skill of the system forming the basis for quality check and further development of the MiKlip system.
The application system is developed as an easy to use low-end application minimising technical requirements for users within MiKlip and potential end users outside of MiKlip. It is designed for interactive usage via a website based internet portal. It provides comprehensive information on the forecasts and access to the different data products. Moreover, the installation of the web based portal will finally support dissemination of decadal scale forecasts to the public.

Aims

  • Establishment and application of standardised methods for evaluation of the MiKlip system and different model components (cooperation with FLEXFORDEC, EnsDiVal).Definition of standards for integrated evaluation and the elaboration of strategies for system design and the preparation of according methods (cooperation with FLEXFORDEC, EnsDiVal).
  • Preparation of benchmarks in order to assess the skill of different configurations of the forecast system according to the coordinated MiKlip strategy (cooperation with FLEXFORDEC, EnsDiVal, VECAP).
  • Selection of variables and definition of appropriate metrics focusing on the inclusion of parameters with a potential forecast skill on the seasonal to decadal time scale (coordinated with FLEXFORDEC, EnsDiVal).
  • Optimisation of the evaluation system by exploration and documentation of uncertainties related to the choice of different observation based evaluation datasets and the assessment of the robustness of the employed metrics used for standard evaluation.
  • Development of a meta data system which will provide detailed information on output data from individual model components, steering parameters, data used for tuning the model, data used for validation, general information on validation and on the quality control status as well as information on system interface data (cooperation with FLEXFORDEC).
  • Installation of a module for extensive analysis and visualisation of model output (hindcast/forecast), validation data and results from evaluation with a quicklook and verification tool kit for monitoring, including a web based interface for interactive usage.
  • Preparation of a user manual for comprehensive documentation of the application system.

More on INTEGRATIONNews-Icon

This description regards the project during the first phase of MiKlip. Integration continues in MiKlip II, in Module D WP1 - FLEXFORDEC/INTEGRATION.

Contact

Institut für Meteorologie, Freie Universität Berlin
Christopher Kadow
Prof. Dr. Ulrich Cubasch
Prof. Dr. Uwe Ulbrich