Module E - Evaluation of the MiKlip decadal prediction system

Module E is evaluating hindcasts from the MiKlip decadal prediction system focusing the main pillars: i) generation of observational data sets and their use for an improved validation of hindcasts, ii) hindcast verification, i.e. the development and implementation of procedures for a quantitative estimation of forecast quality, and iii) process-oriented validation to enhance the understanding and thus the credibility of the prediction system and its products.

Working towards an operational system in MiKlip II, an additional focus comes up: the transfer of predictions from the MiKlip system into probabilistic forecast products for users. This implies a) bias correction of predictions taking a model drift and a climate trend into account, b) calibration of probabilistic forecasts to increase reliability, and c) the construction of forecasts for user-relevant quantities and events, such as heat-waves, droughts, storm surges or other kinds of large-scale climate anomalies.

These pillars define five Module E objectives paving the way towards a useroriented operational system:
1. Bias and Drift correction, Calibration
2. User-oriented post-processing
3. Process-oriented validation
4. Generation of data sets
5. Hindcast verification

Parametric decadal climate forecast recalibration (DeFoReSt 1.0)

2018 - Geosci. Model Dev., 11, 351-368, 2018

Alexander Pasternack | Jonas Bhend, Mark A. Liniger, Henning W. Rust, Wolfgang A. Müller, Uwe Ulbrich

Rainfall over the African continent from the 19th through the 21st century

2017 - Global and Planetary Change, Volume 165, June 2018, Pages 114-127

Nicholson, S. E., | C. Funk, and A. H. Fink

Contact

Freie Universität Berlin, Institute for Meteorology
Prof. Dr. Uwe Ulbrich

Freie Universität Berlin, Institute for Meteorology
Dr. Jens Grieger