[QUESTION]Questions about CO simulation

[yfsong0709]

Hi，I'm a new user to use the Geos-Chem v [13.1.12]for the tagCO simulation. When I was trying to run this model, I have some questions about it.

1. The CO simulation is a simplified simulation for the CO species only. Should I turn on the chemistry menu in input.geos file below?
   
2. If turn on the chemistry menu, what the _GMD_SFCCH4 and _CMIP6_SFCCH4 used for? I tried to turn off both of them, but the model reported the following error:
3. I attach my input files below and hope to seek some guidance. Thank you! input_file.zip

[msulprizio]

Hi @yfsong0709. No, you do not need to turn on any of the settings in the Chemistry Menu, aside from keeping Turn on Chemistry : T. If you follow the instructions for creating a GEOS-Chem run directory and enter the number for simulation type "Tagged CO" when prompted, then the configuration files (e.g. input.geos, HEMCO_Config.rc) included in your run directory should automatically be selected and set up for that simulation type.

[yfsong0709]

Thank you for your reply. I have fellowed the instructions to set the config-files, and simulated for one month. Then, I try to verify the simulation results by using the flask data from the NOAA ESRL Carbon Cycle Cooperative Global Air Sampling Network. As shown in figures below, I found the simulation results for species ‘CO’ (Fig1), ‘COnmvoc’ (Fig2), ‘COch4’ (Fig3) are exceedingly higher than the normal value (usually up to 1e+8 mol/mol dry), while other species such as ‘CObbxx’ (Fig4) and ‘COxx’ (Fig5) are only about 1e-7 mol/mol dry. Why I got these unreasonable results? How can I fix this problem?

Fig 1. species CO

Fig 2. species COnmvoc

Fig 3. species COch4

Fig 4. species CObbeu

Fig 5. species COeur

[yantosca]

Thanks for replying @yfsong0709. I wonder if your issue might be an initialization problem. Because CO is a long-lived species, you will probably have to spin up your simulation for several years (~10 years). (If you don't have 10 years of met data on disk you could use the same year over and over.) Then start your simulation using the restart file from the end of the 10-year spinup. That could very well be the problem.

Comparing model output to data observations is a scientific problem (rather than a technical or coding problem). If you find that re-running after a 10 year spinup does not make your results look better, please contact the Carbon Cycle Working Group for more information.

[yfsong0709]

Hi, I have tried to spin up my simulation with the default settings for several years. Then, I find the concentrations of species ‘CO’, ‘COnmvoc’ and ‘COch4’ are still exceedingly higher than the normal value. (See Fig.1. The value up to 1e+8 mol/mol dry. Shouldn't this value be less than 1 mol/mol dry?) After extensive experiments, I found that turning off the two switches shown in the Fig.2 below will fix above problem, and these two values become reasonable (See Fig.3). I guess is there is a unit conversion issue going on as those two species (COnmvoc and COch4) come from inputs read in from file (and then get added to total CO)? I checked their original files as Fig.4, and found that the unit given in the file is kg / s, while in HEMCO_Config.rc file, these units are set to 1 as Fig.5. How can I fix this problem?

Fig.1 species COnmvoc

Fig.2 BEGIN SECTION EXTENSION SWITCHES Fig.3 Species COnmvoc after turn off PROD_CO_NMVOC and PROD_CO_CH4

Fig.4 GEOSChem.ProdLoss.YYYYMMDD_0000z.nc4

Fig.5 Chemistry Input