The m3stat report you posted shows that your file contains emitted species ZO and ZO2 which I am not familiar with (though I’ll admit that I’m not familiar with the full list of species in most chemical mechanisms).
In addition, the log file from your original posts does not list these two species as being unreferenced by your emissions mapping (in contrast to HOZO, YO, YO2, VOC etc. which are in your emission file but which aren’t mapped to a model species - as an aside, I also haven’t seen HOZO, YO, and YO2 before), so in your GC_namelist file (it seems you are using CMAQv5.2.1) you probably have mapped these ZO and ZO2 species to some model species. Is that correct?
Your log file actually does show gas phase species HOZO, YO, YO2, ZO, and ZO2 while also showing that you’re using cb6r3_ae6_aq. These species aren’t standard cb6r3 species. Did you modify the chemistry somehow?
REPORT.txt (334.6 KB) REPORT_success.txt (223.7 KB)
I just attached them here, I don’t think the failure one has issue on the emission rate. I feel that it might due to the significant figures of the numbers in each grid cell?
If there are any modifications you made to either the CMAQv5.2.1 code or the namelist files compared to the released version, please post or summarize these modifications.
for this run, I did not change anything
Thanks. The log file you linked in your original post shows the following:
IC/BC Factors used for transported gas-phase species
1 NO2 1.0000
2 NO 1.0000
78 YO2 1.0000
77 YO 1.0000
80 ZO2 1.0000
79 ZO 1.0000
These aren’t standard model species I am aware of. Can you please provide more context on why these species are listed in your log file?
That’s what I changed in the input file, to add the nonreactive NOx to track the transport and mixing of the chemicals
In which input file did you make these modifications - the GC namelist file? Non-reactive tracer species should be added to the TR* namelist file (i.e. Species_Table_TR_0.nml), not the GC or AE namelist file.
This is outside my area of expertise and I could be giving wrong advice, but if you indeed added these species to the GC namelist file, my suspicion is that it causes problems with the ebi chemical solver. To test if that’s the case, I’d recommend running your simulation again using the original, released, unmodified GC namelist file.
Hello Huan, after checking with some of my colleagues, it seems that adding the tracer species to the GC namelist file shouldn’t have caused the problem you’re experiencing. My apologies for leading you down a dead end.
Have you tried compiling and running the executable in debug mode to see if this provides any additional information on the calculations leading to the model crash?
Hi Christian,
Thank you for exploring that for me
Did not try the debug mode, how to do that?
Huan
In the bldit_cctm.csh file, uncomment the line “#set Debug_CCTM”. This changes the Fortran compiler flags from those defined as “myFSTD” to those defined as “myDBG” in config_cmaq.csh
And be careful to remove all the previous object and module files in the BLD_* directory.