| 302 | | +++DESCRIBE+++ |
| 303 | | Make_run_def -> script to make consistent orchidee_pft.defs with the option to customize the files for your specific application |
| 304 | | COMP/orchidee_ol.card -> to tell ORCHIDEE wich orchidee_pft.def you want to use. |
| 305 | | COMP/sechiba.card -> make sure the PFT map has the correct number of units '''nvm''' vs. '''nvmap'''. |
| 306 | | orchidee_pft.def with 3 (super powerful test-case), 13 (MTCs), 15 (defaults PFTs), 28 (test all availabe parameter sets) and 37 PFTs (13 MTCs with age classes -> probably best to change into 39 PFTs -> 15 with age classes. |
| 307 | | ++++++++++++++ |
| | 302 | A model run now requires three different parameter files: run.def, orchidee.def and orchidee_pft.def. LibIGCM reads these files in COMP/orchidee_ol.card. |
| | 303 | * The run.def contains all the information for the driver. This is also the file where the default global domain can be reduced (LIMIT_N, LIMIT_S, LIMIT_E and LIMIT_W). When the same experiment is to run in off-line and coupled mode, only the run.def should be changed. |
| | 304 | * The orchidee.def contains the details of the experiment toghether with the settings of the orchidee_ol, sechiba and stomate cards in COMP. Each of the predefined configurations in config/ORCHIDEE_OL will have different settings. These files need to set up manually because each experiment will differ. Copy the best matching predefined experiment from config/ORCHIDEE_OL and use it a template to set-up your own simulation experiment. |
| | 305 | * The orchidee_pft.def is available in different configurations. The configuration depends on the interest of the user in vegetation and its dynamics. The default setting for the moment is the same as the trunk, i.e., 15 pfts with a single age class. Nevertheless other useful orchidee_pft.def files are available. Users can make their own files by making use of the script in the MAKE_RUN_DEF folder. '''Note that different numbers of PFTs will require matching land cover maps. Also the order of the PFTs in the run.def should follow the order of the PFTs in the map. There is no quality control to check whether definition files and input maps following the same PFT order.''' If the PFT maps are available, users can run each of the predefined simulation experiments (config/ORCHIDEE_OL) with all available orchidee_pft.def files. The number of PFTs largely determines the speed of the model. |
| | 306 | |
| | 307 | How to use a different orchidee_pft.def file? |
| | 308 | * Go to COMP/orchidee_ol.card and change the value of DefSuffix. DefSuffix = 15pft.1ac will make use of orchidee_pft.def_15pft.1ac, change DefSuffix = 3pft.1ac to have the model use orchidee_pft.def_3pft.1ac. You can also use your own customized orchidee_pft.def, just make sure the follow the naming guideline. |
| | 309 | * make sure the PFT map has the correct number of units '''nvm''' vs. '''nvmap'''. These numbers are set in orchidee_pft.def but affect the expecation of the model when reading the land cover map. The correct land cover map is set in COMP/sechiba.card |
| | 310 | * orchidee_pft.def with 3 PFTs make for a powerful test-case. Most of the output variables in stomate have their units in something per square meter of vegetation. So two identical PFTs with different vegetation fractions should still result in the same stomate outputs. If this is not the case, something went wrong with the units (and likely the way veget and/or veget_max are used in the model. To use this test case remember to set impose_veg = y (PARAM/orchidee.def) and veget_update = 0Y (COMP/sechiba.card). The test case can be run over large spatial domains which may also help to find bugs. |
