Problem while running WRF-Chem with MOZART-MOSAIC scheme (chem_opt = 202 and 201)

1,448 views
Skip to first unread message

mriganka biswas

unread,
Mar 15, 2021, 12:35:28 PM3/15/21
to wrf-chem

Hello everyone,

I am encountering a problem while running WRF-Chem with MOZART chemical mechanism and MOSAIC aerosol scheme (chem_opt = 202 and 201). The program got stuck without any error message. No output or log entry appears even after 12 hr. I had to kill the program forcibly. The last few lines of the rsl.error file are:

"d01 2013-12-01_00:00:00 microphysics_driver: returning from
d01 2013-12-01_00:00:00  call moist_physics_finish
d01 2013-12-01_00:00:00 zero out small condensates, zero out negative vapor
d01 2013-12-01_00:00:00 zero out small condensates, zero out negative vapor
d01 2013-12-01_00:00:00 zero out small condensates, zero out negative vapor
d01 2013-12-01_00:00:00 zero out small condensates, zero out negative vapor
d01 2013-12-01_00:00:00 calling inc/HALO_EM_THETAM_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_EM_THETAM_inline.inc
d01 2013-12-01_00:00:00  call chem polar filter
d01 2013-12-01_00:00:00 calling inc/HALO_EM_D3_5_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_D3_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_MOIST_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_CHEM_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_TRACER_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_SCALAR_inline.inc
d01 2013-12-01_00:00:00  call set_phys_bc_dry_2
d01 2013-12-01_00:00:00  call spec_bdy_final
d01 2013-12-01_00:00:00 calling inc/HALO_EM_C_inline.inc
d01 2013-12-01_00:00:00 calling inc/PERIOD_BDY_EM_E_inline.inc
d01 2013-12-01_00:00:00  call set_w_surface
d01 2013-12-01_00:00:00 calling inc/HALO_EM_PHYS_W_inline.inc
d01 2013-12-01_00:00:00 calling inc/HALO_EM_PHYS_HCW_inline.inc
d01 2013-12-01_00:00:00 --> TOP OF AFTER ALL RK STEPS
d01 2013-12-01_00:00:00 --> CALLING DIAGNOSTICS DRIVER
d01 2013-12-01_00:00:00 --> TOP OF DIAGNOSTICS PACKAGE
d01 2013-12-01_00:00:00 --> CALL DIAGNOSTICS PACKAGE: ACCUMULATED AND BUCKET DIAGNOSTICS
d01 2013-12-01_00:00:00  call HALO_RK_E
d01 2013-12-01_00:00:00 calling inc/HALO_EM_E_5_inline.inc
d01 2013-12-01_00:00:00  call HALO_RK_MOIST
d01 2013-12-01_00:00:00 calling inc/HALO_EM_MOIST_E_5_inline.inc
d01 2013-12-01_00:00:00  call HALO_RK_CHEM
d01 2013-12-01_00:00:00 calling inc/HALO_EM_CHEM_E_5_inline.inc
d01 2013-12-01_00:00:00  call HALO_RK_SCALAR
d01 2013-12-01_00:00:00 calling inc/HALO_EM_SCALAR_E_5_inline.inc
d01 2013-12-01_00:00:00  call end of solve_em
d01 2013-12-01_00:00:00  DEBUG wrf_timetoa():  returning with str = [2013-12-01_00:00:00]
d01 2013-12-01_00:00:00 calling mozart_mosaic_4bin_aq_kpp from chem_driver
d01 2013-12-01_00:00:00 calling chem_opt=? from chem_driver
d01 2013-12-01_00:00:00  call chem_prep
d01 2013-12-01_00:00:00 calling emissions driver
d01 2013-12-01_00:00:00 No tracer option selected
d01 2013-12-01_00:00:00 MOSAIC or SORGAM sea salt emissions
d01 2013-12-01_00:00:00 AFWA modified Gocart dust emissions
d01 2013-12-01_00:00:00 Gocart dms emissions
d01 2013-12-01_00:00:00 no biomass burning
d01 2013-12-01_00:00:00 biogenic emissions: calling megan v2.04
 calculate MEGAN emissions at ktau, gmtp, tmidh =            1  0.0000000E+00  1.0000001E-02
d01 2013-12-01_00:00:00 emissions_driver calling add_anthropogenics
d01 2013-12-01_00:00:00 add_anthropogenics
d01 2013-12-01_00:00:00 emissions_driver calling add_biogenics
d01 2013-12-01_00:00:00 emissions_driver not calling add_emiss_cptec
d01 2013-12-01_00:00:00 emissions_driver calling mosaic_addemiss
d01 2013-12-01_00:00:00 calling optical driver
d01 2013-12-01_00:00:00 optical driver: process aerosols true
d01 2013-12-01_00:00:00 optical averaging
d01 2013-12-01_00:00:00 optical driver: call optical averaging
d01 2013-12-01_00:00:00 optical driver: after call optical averaging
d01 2013-12-01_00:00:00 calling photolysis driver
 photolysis_driver: called for domain            1
d01 2013-12-01_00:00:00 calling ftuv_driver
d01 2013-12-01_00:00:00 MOZART_MOSAIC_4BIN_KPP, MOZART_MOSAIC_4BIN_AQ_KPP aerosols initialization
d01 2013-12-01_00:00:00 calling conv transport for chemical species
d01 2013-12-01_00:00:00 calling calc_het_n2o5
d01 2013-12-01_00:00:00 chem_driver(01): Calling dchm_tstep_init
d01 2013-12-01_00:00:00 calling kpp_mechanism_driver
d01 2013-12-01_00:00:00 kpp_mechanism_driver: calling mozart_mosaic_4bin_aq_interface"

I am using ERA-intering forecast data for meteorological conditions and mozart-4 global output for chemistry BC and IC. Using same meteorological data and chemistry data WRF-Chem run was successful for RADM2 chemistry and MADE/SORGAM aerosol scheme (chem_opt = 11).

I am attaching the namelist.input, rsl.error, rsl.out and mozbc.inp files. Please help me to solve the problem.

Thanks and regards,
Mriganka

mriganka biswas

unread,
Mar 15, 2021, 12:36:44 PM3/15/21
to wrf-chem
Please find the attachments.

Thanks and regards,
Mriganka

--
To unsubscribe from this topic, visit https://groups.google.com/a/ucar.edu/d/topic/wrf-chem/HCoxDKWd-qs/unsubscribe.
To unsubscribe from this group and all its topics, send an email to wrf-chem+u...@ucar.edu.
mozbc.inp
namelist.input
rsl.out.0000
rsl.error.0000

Mary Barth

unread,
Mar 16, 2021, 12:52:05 PM3/16/21
to mail...@gmail.com, wrf-chem

Mriganka,

Sorry that you are having this issue. One quick check is to be sure you are not running out of disk space. However, since the error is at the start of the simulation, disk space may be okay.

There's nothing obviously wrong in your namelist and mozbc.inp files. So, I suggest looking at the data in the input and emissions files to be sure there's nothing odd looking (i.e. wrfinput, wrfchemi). It's generally the case that wrfbiochemi is read during the real.exe stage and you do not need to include it in the namelist for wrf.exe.

Also, look at the end of all the rsl.out files to see if an error message appears. You may also want to specifically search for "cfl" for CFL errors, but I doubt that since your model run stops at the beginning.

Lastly, when you try different things to test the simulation, I recommend running for a very short time period. At the moment, you can run for just 10 minutes to make sure wrf-chem runs a few time steps okay. If you set history output to 10 minutes, you can also check the output to be sure nothing wrong happens. Then, once you're comfortable with how the model is running, you can conduct the longer simulation.

-- Mary

-- 

Mary Barth

unread,
Mar 17, 2021, 1:37:26 PM3/17/21
to wrf-chem, mail...@gmail.com

Mriganka,

It does sound like compiling with KPP could be the issue. Since I am less of an expert on this topic, I am forwarding your email back to the WRF-Chem discussion group. But one step you could try is clean -a, configure, and compile again.

Mary



-------- Forwarded Message --------
Subject: Re: [WRF-Chem] Problem while running WRF-Chem with MOZART-MOSAIC scheme (chem_opt = 202 and 201)
Date: Wed, 17 Mar 2021 22:59:30 +0530
From: mriganka biswas <mail...@gmail.com>
To: Mary Barth <bar...@ucar.edu>


Hello Dr. Barth,

1. Is it possible that the issue I am facing, is due to compilation problems? Is there any way to check that?
 
2. Today I tried running WRF-Chem with RADM2-MADE/SORGAM mechanism with KPP library (chem_opt = 106) and it got hanged with the similar log message as earlier. Earlier WRF-Chem with RADM2-MADE/SORGAM scheme without KPP (chem_opt = 11) ran successfully. Probably the problem is related to KPP. I am attaching the rsl.error and namelist file. Please let me know your suggestions.

Regards,
Mriganka
rsl.error.0000.chem_opt_106
namelist_chem_opt_106.input

ANIE K LAL

unread,
Oct 21, 2021, 12:23:52 AM10/21/21
to wrf-chem, mail...@gmail.com
Hi Mriganka,

I am also facing the same issue as yours. My run gets stuck after calling Photolysis driver:

...
photolysis_driver: called for domain            1
d01 2019-12-28_00:00:00 calling ftuv_driver
d01 2019-12-28_00:00:00 MOZART_MOSAIC_4BIN_KPP, MOZART_MOSAIC_4BIN_AQ_KPP aerosols initialization
d01 2019-12-28_00:00:00 calling conv transport for chemical species
d01 2019-12-28_00:00:00 calling calc_het_n2o5
d01 2019-12-28_00:00:00 chem_driver(01): Calling dchm_tstep_init
d01 2019-12-28_00:00:00 calling kpp_mechanism_driver
d01 2019-12-28_00:00:00 kpp_mechanism_driver: calling mozart_mosaic_4bin_aq_interface

I am using version 4.3, chem_opt =202, and CAM-Chem for chemical boundaries (namelist attached).
Where you able to solve the problem?
Seeking help regarding this.

Thank you
Anie
namelist.input

cho...@gmail.com

unread,
May 20, 2022, 1:04:06 AM5/20/22
to wrf-chem, aniek...@gmail.com, mail...@gmail.com
Dear Anie,
Hello,

I am facing the same problem as you.
I am using version 4.4, chem_opt =202, and CAM-Chem for chemical boundaries.

It worked fine in version 4.2, but I have the same problem as you in version 4.4.
Did you solve the problem you are experiencing?
If you have solved the problem, please let me know how.

Thank you
Jae-Hee, Cho

2021년 10월 21일 목요일 오후 1시 23분 52초 UTC+9에 aniek...@gmail.com님이 작성:

cho...@gmail.com

unread,
Jan 26, 2023, 8:44:18 PM1/26/23
to wrf-chem, cho...@gmail.com, aniek...@gmail.com, mail...@gmail.com
Hello everyone.

The problem I was facing was resolved in 1 year.
It went through a lot of installs and runs.
The solution was 'mozart_ph_diag = 1'.
I couldn't find it anywhere in the user guide.
Please refer to the bulletin board below.
Thank you.

https://forum.mmm.ucar.edu/threads/possible-bug-with-mozart-mosaic-scheme-while-writing-output.12438/#post-29457

Jaehee Cho

2022년 5월 20일 금요일 오후 2시 4분 6초 UTC+9에 cho...@gmail.com님이 작성:
Reply all
Reply to author
Forward
0 new messages