Model fails in running MOVESMRG in Onroa_RPP

SMOKE
23

Following is the GRIDDESC file I used in this run:
’ ’
‘US_12km’
2 33.000 45.000 -97.000 -97.000 40.000
’ ’
‘US_12km’
‘US_12km’ -2916000.000 -2304000.000 12000.000 12000.000 486 384 1

24

As @eyth.alison asked, it would be benefit for you to share your met input file (METMOVES) used for you RPP movesmrg run. We need to make sure there is no abnormal values in your file.

25

One other thing you might try is to try running RPP for our 12US1 domain using our METMOVES file and see what happens. If that test works, then your problem is probably related to the METMOVES file. If that test has the same issue as their other runs, then we can cross the METMOVES possibility off of the list.

I’m not sure if possibly some downloaded emission factor tables might be corrupted? Just a thought.

26

I am thinking the same if it is due to corrupted eftables but it seems not since it still stops after I re-download the eftables. I will try to run 12US1 and let you know how it goes. Thanks.

27

Thanks for your response. I uploaded the METMOVES file I used in the run (“SMOKE_DAILYxxx file”) that can be downloaded from the following links:Sign in to your account
Hope to receive your help on this case. Thanks.

28

I’m not sure I understand how to access the METMOVES files. Did you upload any to CMAS or to another server?

30

Hi, I uploaded it to google cloud which should be accessed using the link I shared. Let me know if it is not working or any other server you prefer. Thanks.

31

I have one more question. Do I need to generate my own SRGPRO and SRGDESC for my own domain when running this program?

32

Good question. If your domain is an aligned subset (i.e., same map projection and compatible origin) of the EPA 12km or 4km surrogate domain, SMOKE will “jump” to the right spot of the surrogate. However, if your domain is not aligned or is at a different resolution, you would need to generate your own surrogates.

Is your grid aligned with the 12US1 / 12US2 EPA domain?

34

Not exactly. My domain uses the same projections but 12US1 from EPA is using the following setting:
‘12CONUS1’, -2556000.000, -1728000.000, 12.D3, 12.D3, 459, 299, 1
While mine uses:
‘US_12km’ -2916000.000 -2304000.000 12000.000 12000.000 486 384 1
Is it the reason causing the program to stop? But it works well in all other sectors.

35

It’s possible. Which surrogates are you using – 2016 platform or 2017 platform?

37

I am using the 2016 v1 platform

38

The 2017 platform surrogates have a larger grid – you might try that. Is your map projection the same as EPA’s grid?

39

Yes, both are " 2 33.000 45.000 -97.000 -97.000 40.000
" in the grid description file.

40

I tried to generate MOVES using 2017 platform surrogates and run RPP with it but still failed, the program still ended with the same issue (no error message). It seems like not due to the grid setting. Is there any other potential reason that could cause this issue? Really appreciate your patience with my question.

41

OK – thanks. We think your grid configuration is compatible with the surrogates.

We still think the METMOVES is the most likely issue and don’t think we have ruled that out yet. Did you run RPP for EPA’s 12US1 domain and EPA’s METMOVES output?

If you are able to run that METMOVES test, can you post the log from your Met4moves run (the SMOKE log, not the standard output), and also their Grdmat log. There are a couple of quick things we can check from looking at those particular logs.

Thanks…

42

Sorry this has become such a long thread, but one other thing you might try is this:

Try running onroad RPP from one of the EPA platforms as is with no changes. Our domain (12US1), our METMOVES, our everything – using all of the scripts and inputs as they are packaged. If that doesn’t work, then we know there is something strange going on specific to their platform which is causing RPP to crash. If it does work, then the full logs package makes sense.

44

Thanks for your suggestion. I tried but still stopped at the same point. The log file is attached. I am considering it could be due to the node setting but as we previous discussed if RPD runs well, then nodes shouldn’t have any issues. Do you have any suggestions on this issue?
movesmrg_RPP_onroad_aug_2016fh_16j_20160831_US_12km_cmaq_cb6.log.txt (21.2 KB)

45

Is it possible that too many warning and error messages that cause the end of the program?

46

No too many warnings / errors would not be the problem. Can you please provide an RPP Movesmrg log from the test in which your ran EPA scripts as is (12US1 domain and EPA METMOVES)? The log file you posted was not using the EPA data…