Module C - Regionalized Decadal Prediction

The objective of Module C is to establish a regionalization component to the MiKlip decadal prediction system. Downscaling ensembles are created and assessed with respect to the predictive skill and added value in specific target regions. During MiKlip the target regions Europe, Africa and Central America/North Atlantic where addressed using up to three regional climate models (COSMO-CLM, REMO and WRF) und several approaches regarding the downscaling techniques, process descriptions and initialization methods. For MiKlip II Module C focusses on the target region Europe and a single RCM, namely COSMO-CLM, for the regional prediction system. The goal is to improve and consolidate this system and provide robust estimates of its predictive skill.

To achieve this, the research has been grouped into three major objectives divided into a total of eight work-packages with contributions from five institutions:

Objective C1 Advance the COSMO-CLM towards a regional climate-system model (2 work packages)      

The focus is here on two main components of the regional climate system, which showed promising potential to improve process descriptions and predictability during the first phase of MiKlip:

  • Implementation and testing a coupled regional ocean model(CCLM/NEMO) for the marginal seas around Europe (C1-WP1)
  • Inclusion of more sophisticated soil-vegetation exchange models (SVATs) into the predictions system (C1-WP2)

Objective C2: Examine the application potential of regional decadal predictions and the relevant process behind decadal predictability over Europe (3 work packages)

  • Study the skill and added value of the regionalization for user relevant climate parameters (C2-WP1)
  • Examine large regional climate anomalies and their impact and predictability over Europe (C2-WP2)
  • Improved attribution of the long-term climate evolution in Europe to climate trends due to green-house gas emissions and longer-term climate variability (caused for instance by variations of the oceanic processes in the North Atlantic). This will be achieved by enlarging the investigation period over the 20th century and thereby increasing the robustness of the skill estimates. (C2-WP3)

Objective C3: Generation and optimizing regional decadal prediction ensembles (3 work packages)

  • Post-processing of regional decadal predictions, to improve the results of the raw simulation ensembles towards a reliable forecast (C3-WP1)
  • Improve the composition of the ensembles with respect to a well-balanced ensemble and computational efficiency (C3-WP2)
  • Generation of the regional ensemble predictions and assessment of the optimal configuration (C3-WP3)

See also: http://www.clm-community.eu/index.php?menuid=261

The smoother extension of the nonlinear ensemble transform filter

2017 - Tellus, Vol. 69(1)

Kirchgessner, P. | J. Tödter, B. Ahrens, and L. Nerger

The regional climate model and the model components coupled via OASIS3-MCT: description and performance

2017 - Geoscientific Model Development, Vol. 10 (4), pp. 1549–1586

Will, A. | N. Akhtar, J. Brauch, M. Breil, E. Davin, H. Ho-Hagemann, E. Maisonnaive, M. Thürkow, and S. Weiher

Decadal and multi-year predictability of the West African monsoon and the role of dynamical downscaling

2017 - Meteorol. Zeit., doi:10.1127/metz/2017/0811

Paeth H. | A. Paxian, D.V. Sein, D. Jacob, H.-J. Panitz, M. Warscher, A.H. Fink, H. Kunstmann, M. Breil, T. Engel, A. Krause, J. Toedter and B. Ahrens

Is there potential added value in COSMO–CLM forced by ERA reanalysis data?

2017 - Climate Dynamics

Lenz, C.-J. | B. Früh, and F. Davarary Adalatpanah