updates to namelist settings, various additional monthly mean streams echam6 and jsbach
Here are several definitions of mean value streams that have the following purpose:
- they provide critical variables, that are per standard defined as being a snapshot rather than a mean (in particular temp2), robustly as a monthly mean (robustly as the stream definition is explicitly made, there is no more need for interpretation and guess work)
- create streams where general characteristics of the land biosphere, related to productivity and dynamic vegetation, are written out
These streams are intended for production runs, but their use may also be sensible for spinups (to identify the degree of equilibrium of land biosphere for example). The ISM stream has been an idea by @lackerma and is intended for atmosphere / ice sheet coupling in AWI-ESM-2-2 and similar simulations.
There is some overlap with general echam6 streams, but the fact that data is organized thematically should enable to run a cleanup of non-necessary output files in the framework of a post-processing.
This work is based on contributions by @lackerma and @pgierz.
I have tested the settings in the framework of a two year simulation. With the added namelist snippets into a general PALEO ECHAM6 namelist the simulation ran without obvious problems. It has also successfully restarted one time and finished the second year without crashes. Integrity of the additional model output has been tested by conversion of GRIB to NetCDF, followed by a rough inspection of some fields.
Note, that setting variable metadata when convering GRIB output of the new streams to NetCDF will generally not work based on the command
cdo -t echam6 copy input input.nc
Instead it is necessary to explicitly use the code files that are generated in parallel to the output files, i.e.:
cdo -t input.codes copy input input.nc
Consequently, one should consider keeping/archiving also the code files (at least one per simulation) as an integral part of the model output, not only the model output itself.
The code snippets, together with comments outlining output made available via the additional streams, are attached to this issue.
namelist_snippet_echam6_additional_monthly_output_streams.txt
namelist_snippet_echam6_jsbach_vegetation_and_related_output.txt
Attached is also the NetCDF version of the generated output (one month only, standard T63 resolution) for further inspection by others.
test_custom_namelist_final_200001.01_monthly_means_3d_gl.nc
test_custom_namelist_final_200001.01_monthly_means_3d_g3b.nc
test_custom_namelist_final_200001.01_monthly_means_2d.nc
test_custom_namelist_final_200001.01_ism.nc
test_custom_namelist_with_veg_final_200001.01_jsbachmon.nc
test_custom_namelist_with_veg_final_200001.01_vegmon.nc
test_custom_namelist_final_200001.01_monthly_means_3d_sp.nc
File sizes are (per !!month!! raw GRIB output is smaller!):
728K test_custom_namelist_final_200001.01_ism.nc
2,9M test_custom_namelist_final_200001.01_monthly_means_2d.nc
10M test_custom_namelist_final_200001.01_monthly_means_3d_g3b.nc
10M test_custom_namelist_final_200001.01_monthly_means_3d_gl.nc
2,3M test_custom_namelist_final_200001.01_monthly_means_3d_sp.nc
7,4M test_custom_namelist_with_veg_final_200001.01_jsbachmon.nc
12M test_custom_namelist_with_veg_final_200001.01_vegmon.nc