Action on the evaluation tools used/developed for ORCHIDEE simulations

Meeting 23/06/2017

This action aims at offering a single tool that may gather the different diagnostics which are currently produced by several people. It will be limited to the production of ‘maps’, a similar action on production of time-series is also ongoing with P. Brockman.

Vladislav Bastrikov, Jérôme Servonnat and Nicolas Vuichard are the key persons involved in the present action (coordinated by Nicolas). Philippe Peylin and Fabienne Maignan are also concerned, (for ORCHIDEE and Evaluation Datasets, respectively).

The features of the tool that we want to produce are

• to have a tool that many people (more than 1 !) may process
• to process simulations stored at different places
• to have flexible tool in terms of selection of season, variables, ...

Overview of the different tools available

• Climaf: developed by IPSL and CNRM (Jérôme Servonnat and Stéphane Senesi, respectively). ​http://climaf.readthedocs.io/en/latest/

Here is an example of what is done with climaf for the pre-CMIP6 simulations: ​https://vesg.ipsl.upmc.fr/thredds/fileServer/IPSLFS/jservon/C-ESM-EP/CM6010_family_14062017_jservon/C-ESM-EP_CM6010_family_14062017.html

• Climaf @ LMD: They use Climaf tool but adding a menu for the selection of season + a brief summary of the simulations used. Here an example: ​https://vesg.ipsl.upmc.fr/thredds/fileServer/IPSLFS/fabric/lmdz/MultiSimu/CONTTROP/BIASGLOBDJF.html
• MAPPER: developed by Vladislav, it generates maps and time-series for selected regions. One may select set of variables (Energy/Water/Carbon?), temporal settings (Season, Annual mean, ...) and bias map option (diff vs OBS, vs reference simulation). Here an example: ​https://orchidas.lsce.ipsl.fr/mapper/FG3.4438.php There is also a summary of the simulations here: ​https://orchidas.lsce.ipsl.fr/mapper/index.html
• Inter-atlas tool: developed by Nicolas Vuichard. It provides global maps of three metrics (annual mean, Standard deviation of monthly variations, correlation with observation) for two simulations. It also generates plots of the distribution of the bias with observation (or of the correlation) overall the grid cells. Here is an example: ​http://www-lscedods.cea.fr/orchidods/EVAL/GLOBAL/r4438/CL5/CL5.4438.L6010.alt1_VS_CL5.4438.L6010.ref/ One may select only one variable and/or one diagnostic with buttons.
• Global Carbon Atlas: Developed by P. Brockmann (​http://www.globalcarbonatlas.org/en/flux-maps). This is a tool that generates “on-the-fly” maps and time-series for a set of preloaded data

Recommendations

• To dissociate the work needed for generating the data that will be later visualized and the work needed for displaying them.
• To be able to produce ‘atlas’ (in static mode) and to think about the capacity of producing maps “on-the-fly” as in the carbon atlas.
• To encourage the use and further development of Climaf, which is the tool with the highest visibility (national).
• To have a look at what is doing at ICOS atm center in terms of data flow and data visualization.

Further actions/meetings planned

• To communicate to the ORCHIDEE group on the ‘Evaluation’ effort. Soon
• Meeting on the Climaf tool and how to “customize” it => Jérôme, Vlad, (Nicolas) Planned for End of June / Early July
• Meeting on the inclusion of additional diagnostics into Climaf => Jérôme, Nicolas, Vlad. Planned for Early July
• To report on the Evaluation effort to the ORCHIDEE group. Next Fall