A first meeting of various presentations around the tuning of models has been organised from 5 to 7 September 2017 within HIGH-TUNE at LMD (Paris).
The first day consists of presentations around the importance of tuning for climate models, the presentation of statistical tools that can be used for this tuning, the previous work on the comparison between single-column models and large-eddy simulations.
Afterwards, a coding sprint was organised in order to better understand the statistical tools developped by Daniel Williamson at the University of Exeter and start to apply them to the comparison of Single-Column models and Large-Eddy Simulations. For that, two case studies, one for the clear boundary layer and one for a fair-weather cumulus day.
reports :
- HIGH-resolution simulations to improve and TUNE boundary-layer cloud parameterizations
F. Couvreux
- LES and Single-Column Models Positionning of the HIGH-TUNE project & expectation from the tuning tools
F. Couvreux, F Hourdin, C Rio, R Roehrig, A Jam, V Masson, J Pergaud
- Reduction of systematic errors in climate models : model improvement versus
tuning of free parameters.
F. Hourdin, C. Rio, I. Musat, F. Couvreux, D. Williamson and the LMDZ team.
LMD/IPSL, CNRM, Exeter Univ
- Statistical calibration of the CNRM atmospheric climate model
Aurélien Ribes, Olivier Audouin, Romain Roehrig
CNRM - Météo-France and CNRS
- Ensembles, emulators and history matching; a tutorial
Danny Williamson and Victoria Volodina
University of Exeter
With contributions from Adam Blaker, James Salter, John Scinocca and Slava Kharin