After running CMAQ-ISAM, I found that the results of all the tags in the CCTM_SA* file, except for OTH,ICO,BCO, are all 0 values as shown in the figure. However, the result in CCTM_* is normal. I would like to know what causes this phenomenon and how can I fix it? I have attached the relevant file below.
Thank you for your answer!
EmissCtrl_cb6r3_ae7_aq.txt (31.0 KB)
run_cctm.csh (35.4 KB)
isam_control.txt (2.0 KB)
Your target region is labeled as CC in your isam_control.txt file, however, I did not find it in your EmissCtrl_cb6r3_ae7_aq.txt. This is why all concentrations are 0 because of no emissions from CC region that does not actually exist. You should define your region named CC and create data file for CC first if it is a specific self-defined region (nether a nation nor a state…), and then edit the line in EmissCtrl_cb6r3_ae7_aq.txt immediate after
! | Region Label | File_Label | Variable on File
“MYCC” ,‘ISAM_REGIONS’ ,“CC”
in your isam_control.txt file (for example)
TAG CLASSES |OZONE,
TAG NAME |CC
EMIS STREAM(S) |AGL
I hope this helps.
At least two items in your GRIDMASK file seem incorrect:
the VAR-LIST attribute should list the variables as character*16 strings, i.e. “CC HB JL” … etc., while your attribute is "CC HB JL SP SY "I see now that this likely was a markdown issue, not an issue with your file. Apparently markdown removed those extra spaces following the variable names both in my comment and your copy of the file header, so probably you do see those required extra spaces when you ncdump the file on your system.
The variables representing the region masks are expected to be float, not int by the ISAM code.
How did you create your GRIDMASK file? The header mentions M3MERGE
, but M3MERGE shouldn’t have created that type of VAR-LIST attribute. Did you do any editing / file manipulation after running that tool?
The discussion in the following threads suggests that you could try a combination of ncdump and ncgen to convert the variable type to float, but I don’t know if that approach tried by that user was successful. Maybe you could also try to reach out to them to see if they can share their experience:
Yes, the variable in your region mask file should be in real format. You may straightforward use ncflint to convert them from integer to float or refer to Cristian’s suggestion, and see if your CMAQ/ISAM works.
Try to use m3fake instead of t3mask in t3tool, please refer to
BTW, I created my mask files using ArcGIS and ncgen.