clm-community logo


Working Group: Evaluation


Emmanuele Russo (UNIBE, Chair),  Hendrik Feldmann (IMK-KIT, Co-Chair)



Evaluation of model results is everybody's task applying or developing the model. Often, the biggest challenge is to find on one hand suitable observations (stations as well as gridded data) and, on the other, to define the measures fitting the needs of the evaluation in the best way.

The WG-EVAL brings together all the people who are interested in discussing datasets, evaluation methods and results. 

Tasks and Challenges

The need of more exchange and systematic comparison between different evaluation experiments was revealed as one crucial aspect to be discussed and organized within this COSMO-CLM WG.

This is based on the common interests:

- to better understand the model deficiencies

- to find out ways to overcome them

- to enhance communication of evaluation results (collecting and exchanging information about evaluation from various groups, for various regions)

- to define and introduce a standard evaluation tool (simple rudimentary tool based on CDO or NCO utilities for a small set of common evaluation parameters)

Coordinated Evaluation Strategy

One main task of the WG is to develop a coordinated evaluation strategy for testing new model versions and newly introduced namelist parameters and to assess their influence on the climate change signal.

The process aims to provide a recommended and evaluated standard model version of COSMO CLM.

The evaluation of cosmo5.0 is completed. The process was divided into 3 phases and is documented at 3 appropriate web entries :

Phase 1: testing of new parameters

Phase 2: testing of promising parameter combinations (started in Jan/2015)

Phase 3: tuning on the basis of CON607 following the Omar Bellbrat routine (2012), started Jun/2015 at ETHZ .

The namelist settings are included under the entry 'CCLM-EU165_5.0_clm6' in the Namelist-Tool (list and start of the tool).

COPAT2 for COSMO6.0 and ICON-CLM is in preparation phase 0

Phase 0: set up of the models including model domain, time period, output variables and their resolution.

Phase 1c: testing of new parameters of COSMO6.0

Phase1i: testing of parameters of ICON-CLM

WG Meetings

WG meeting take place regularly at the COSMO User Seminar in Offenbach, Germany, in spring and at the CCLM General Assembly in Autumn. All communitiy members are kindly invited to come to these meetings. 

Minutes from last meetings:

- 2020-09-15  WG EVAL meeting CLM Assembly 2020, WEB session.

- 2019-09-19 Joined PG ICON - WG EVAL meeting at CLM Assembly 2019, Paestum (I)

- 2019-03-21 Joined PG ICON - WG EVAL meeting at ICCARUS, Offenbach (D)

- 2018-03-01 at ICCARUS, Offenbach (D)

- 2016-03-10 at COSMO User Seminar 2016, Offenbach (D)

- 2015-10-01 at CLM General Assembly 2015, Belvaux (L)

- 2015-06-05 webconference Coordinated Evaluation Project - COSMO5.0-CLM 

- 2015-05-27 workshop on the Coordinated Evaluation Project - COSMO5.0-CLM, Offenbach (D)

- 2015-03-05 at COSMO User Seminar 2015, Offenbach (D)

- 2015-01-20 Coordinated Evaluation Project - COSMO5.0-CLM, Offenbach (D)

- 2014-09-03 at CLM General Assembly 2014, Frankfurt (D)

- 2014-03-20 at COSMO User Seminar 2014, Offenbach (D)

- 2013-09-29 at CLM General Assembly 2013, Zurich (CH)   


Summary of evaluation results for COSMO-CLM version 4.8_clm13 (clm17): Comparison of three different configurations over Europe driven by ECMWF reanalysis data ERA40 for the period 1979-2000

Evaluation report for 4.8_clm17: evaluation_report_cclm_4.8_v8.pdf


Evaluation Utilities

 - Observation data -

Everybody needs reliable observation data to evaluate the model results. Carrying out simulations in different areas often opens up questions on available observational data sets for the specific region. We started to collect information on data sets. A list will be published soon but needs your input.

 - Evaluation Tool - 

ETOOL Version 2.0: Peter Berg developed a basic evaluation tool at KIT and provided it to the CLM-Community. Susanne Brienen and Andrew Ferrone extended it by more functions. The tool script is available here.



Bodo Ahrens, Ivonne Anders, Nico Becker, Omar Bellprat, Inge Bischoff-Gauss, Marcus Breil, Susanne Brienen, Erwan Brisson, Edoardo Bucchignani, Ernesto Caetano, Fabien Chatterjee, Evgenii Churiulin, Edouard Davin, Guy de Morsier, Matthias Demuzere,  Andreas Dobler, Alessandro Dosio, Fallah Bijan, Hendrik Feldmann, Barbara Früh, Imoleayo Gbode, Beate Geyer, Matthias Gröger, Maike Hacker, Klaus Haslinger, Samuel Helsen, Anja Hermans, Klaus Keuler, Moritz Kirschner, Sven Kotlarski, Simon Krichak,  Gregor Leckebusch, Hilke Lentink, Shuping Li, Patrick Ludwig, Xiao Luo, Mariano Mertens, Aditya Mishra, Qing Mu, Annika Obermann, Ouedra Ousmane Aly Yabyoure, Hans-Jürgen Panitz, Dirk Pavlik,   Vladimir Platonov, Stefan Poll, Andreas Prein, Kai Radtke, Mohammad Rahimi, Burkhardt Rockel, Emmanuele Russo, Sajjad Saeed, Lucy Samvura Motsi ,Abdoulaye Sarr,  Jan-Peter Schulz, Sonia Seneviratne, Christian Steger, Mauro Sulis, Wim Thiery,  Heimo Truhetz, Nicole van Lipzig, Jesus Vergara Temprado, Fabian Wachsmann, Andreas Walter, Stefan Weiher, Andreas Will, Huan Zhang, Johann Züger

= CLM-Community login required