search

geoschem / geos-chem

GEOS-Chem "Science Codebase" repository. Contains GEOS-Chem science routines, run directory generation scripts, and interface code. This repository is used as a submodule within the GCClassic and GCHP wrappers, as well as in other modeling contexts (external ESMs).
http://geos-chem.org
Other
164 stars 156 forks source link

GCHP fullchem simulation fails when switch on CMIP6_SFC_CH4 #2175

Closed hazel008 closed 3 months ago

hazel008 commented 6 months ago

Name and Institution (Required)

Name: Hansen Cao Institution: University of York

Confirm you have reviewed the following documentation

Description of your issue or question

Please provide as much detail as possible. Always include the GEOS-Chem version number and any relevant configuration and log files. Hi, I am running a fullchem simulation using GCHP14.2.2. It runs fine when I have GMD_SFC_CH4=true and CMIP6_SFC_CH4 =false. But it fails at the beginning, when I have GMD_SFC_CH4=false and CMIP6_SFC_CH4 =true. I would like to run a simulation representing the pre-industrial condition, that's why I want to use CMIP6_SFC_CH4, which has pre-industrial CH4 concentration. I have tried setting the CRE flat in HEMCO_Config.rc to RY, RF, and CYS as shown below, none of them works. CMIP6_Sfc_CH4 $ROOT/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surfaceVMR$YYYY.2x25.nc CH4 1750-1978/1-12/1/0 CYS xy ppbv * - 1 1 Can you please help me? Thank you! STD_anthOFF_3780440.log

I am also attaching my GCHP log file here.

lizziel commented 6 months ago

Hi @hazel008, the error message in your log file is this: pe=00076 FAIL at line=00803 ExtDataGridCompMod.F90 <Found 1 unfulfilled imports in extdata>

This indicates an error in the MAPL ExtData component which handles I/O. The ExtData.rc config file contains the list of imports to the model. In GCHP the imports needed by HEMCO are listed in HEMCO_Config.rc but they are defined in ExtData.rc, including filepath, read frequency, etc. The error message means that one of the expected imports configured in HEMCO_Config.rc was not found in ExtData.rc.

The error log associated with ExtData is called allPEs.log. Does that log file give any additional information? If not you can enable maximum ExtData prints within config file logging.yml in this section:

   CAP.EXTDATA:
       handlers: [mpi_shared]
       level: WARNING
       root_level: INFO   <=== Change this to DEBUG

Also see the following sections of GCHP ReadTheDocs:

hazel008 commented 6 months ago

Hi @lizziel , Thanks for the quick response! I tried you suggestions and enabled the maximum ExtData prints within the file logging.yml and found out why the run failed. It was because the following line in ExtData.rc was commented out: CMIP6_Sfc_CH4 ppbv N Y F%y4-%m2-01T00:00:00 none none CH4 ./HcoDir/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surfaceVMR%y4.2x25.nc

Now I have uncommented it and the simulation runs fine. I have another question regarding the flag in the HEMCO_Config.rc: CMIP6_Sfc_CH4 $ROOT/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surfaceVMR$YYYY.2x25.nc CH4 1750-1978/1-12/1/0 EY xy ppbv * - 1 1

The flag is EY by default. My simulation year is 2022. If I want to drive the simulation using CMIP6_SRC_CH4 from pre-industrial years, which are the years before 1979, what flag should I use? As I understand, EY means the simulation year should be exactly the same year of NetCDF files, do I need to generate a NetCDF file specifically for the year of 2022 using the CH4 concentrations from the year 1978? Thanks!

hazel008 commented 6 months ago

@lizziel ,Sorry, typo error, I want to use CH4 concentration from the year of 1750 to drive the simulation for the year of 2022. Just wondering what flag should I use for the following line in HEMCO_Config.rc: CMIP6_Sfc_CH4 $ROOT/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surfaceVMR$YYYY.2x25.nc CH4 1750-1978/1-12/1/0 EY xy ppbv * - 1 1 Thanks!

lizziel commented 6 months ago

Hi @hazel008, those flags in HEMCO_Config.rc are ignored in GCHP. See the HEMCO_Config.rc page of ReadTheDocs to learn what is used and what is not. Everything not used in HEMCO_Config.rc is in ExtData.rc for GCHP.

In the past we have run into an ESMF issue reading ExtData.rc because the file was too long. Commented out lines reduces the effective length. For this reason we comment out emissions that are not used by default in GCHP. As a courtesy they are still (mostly) included so you do not need to add optional inventories from scratch.

To use only 1750 CMIP6 data for later years you can configure it to use that year as climatology. Change this in ExtData.rc: CMIP6_Sfc_CH4 ppbv N Y F%y4-%m2-01T00:00:00 none none CH4 ./HcoDir/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surface_VMR_%y4.2x25.nc to this: CMIP6_Sfc_CH4 ppbv Y Y F1750-%m2-01T00:00:00 none none CH4 ./HcoDir/CMIP6/v2020-03/2x2.5/CMIP6_GHG_surface_VMR_1750.2x25.nc

lizziel commented 5 months ago

Hi @hazel008, were you able to resolve your issue?

yantosca commented 3 months ago

Closing out this issue