SMOKE output file missing for ptnonipm sector of EPA 2016v1

Hello @bbaek

Yes, I don’t need to modify emissions in any other sectors. So, I just need to work with the ‘nonipm’ sector. And I can successfully convert CMAQ formatted SMOKE output to CAMx format.

But as I mentioned, SMOKE output has missing days[comment#1] ( because of temporal allocation in L_TYPE & M_TYPE).

I am trying to get one file for each day for CAMx input. When I ran CMAQ2CAMx tool, it’s still giving me those missing days (screenshot:please see the skipped days)


.

How do I get files for the missing days?

Hi @eyth.alison

Yes, I have those premerged data from the IWDW.
But right now, I just want to get emis_mole & inline files for each day for the nonipm sector as base SMOKE run has missing files for each day (comment #1 ) due to temporal allocation in L_TYPE & M_TYPE.

Would you please help me on that?

OK – I think we now understand what is going on. With the EPA platform, there are scripts for doing the CAMx conversion that do not require files to be there for each day.

The 2016v1 package includes CAMx conversion scripts. See section 8 of this file:

https://gaftp.epa.gov/Air/emismod/2016/v1/2016_v1_platform_package_30apr2021.txt

The file that contains the scripts is here:

https://gaftp.epa.gov/Air/emismod/2016/v1/smoke_2016v1_platform_core_04feb2020.zip

If you are unable to get the scripts to work and you daily emissions for ptnonipm instead of representative days, you could:

  • Change L_TYPE and M_TYPE to “all” when running ptnonipm

  • In the SECTORLIST, for ptnonipm, change “mwdss_Y” to “all”

But, we recommend trying to use our CAMx conversion scripts instead of the other program you are trying to use.

Hello @eyth.alison
Sorry for the late reply. I was trying to solve this. But faced the following error:
So, as your suggestions I modified L_TYPE and M_TYPE to ‘all’ and ‘mwdss_Y’ to ‘all’ in the SECTORLIST file.
T1_Annual_ptnonipm_daily_36US3_2028fh_16j.csh.txt (7 KB)

But run stopped automatically after only two days output

The terminal output is

It say’s setenv: Too many arguments.

Would you please help me?
T1_Annual_ptnonipm_daily_36US3_2028fh_16j.csh.txt (7 KB)
sectorlist_2028fh_02aug2019_v0.txt (2.6 KB)

Hi, we realized to create the files for every day you also need to do this:

Replace the last line in the script:

$RUNSCRIPTS/emf/smk_pt_annual_byday_emf.csh $REGION_ABBREV $REGION_IOAPI_GRIDNAME -m “$RUN_MONTHS” $SPINUP_DURATION byday

with this:

$RUNSCRIPTS/emf/smk_pt_daily_emf.csh $REGION_ABBREV $REGION_IOAPI_GRIDNAME -m “$RUN_MONTHS” $SPINUP_DURATION daily

1 Like

Hello @eyth.alison
So, after modifying the last line as your suggestion I tried to run it again. Now, I got the following error in Terminal output:

When I checked the smkmerge_ptnonipm_jan_2028fh_16j_20160101_36US3_cmaq_cb6.log file I can see following error:
Screenshot from 2021-05-26 08-29-18

Thanks in advance

OK, one more script change. Delete the line from the run script that says:

setenv MRG_BYDAY “P”

2 Likes

Hello @eyth.alison
It’s working. Thank you so much. One more question:

How do I confirm that my output has no error. My SMOKE run has generated one file for each day and each files are same size. Is that enough to conclude that run is success?

Thank you

Good news! I think you could do ncdump on the new files and compare them to the previous files that were available / EPA platform files.

You could also run Smkreport to generate some reports based on the new and older files and compare those.

1 Like

Along with what Alison mentioned, their run scripts use m3stat tool to make sure that those output files are created correctly. If the format of output is not complete, there will be an error message.

1 Like

Hi @eyth.alison
How do I run Smkreport?. I can’t find scripts in the 2016v1 platform.

Would you please help me?

Thank you so much

You can edit the “run_setting.txt” run setting file located under run scripts folder to enable Smkreport runs. By default, there should be already some Smkreport reports generated by Smkreport like county, state, county-scc, and so on summary report. Check your log folder to find out which reports are generate by review the Smkreport log files. If you wish to generate any additional reports, then you can update the run setting input file.

1 Like

Hello @eyth.alison and @bbaek ,
So, I was double checking the outputs by using m3diff. According to https://www.epa.gov/sites/production/files/2020-11/documents/2016v1_emismod_tsd_508.pdf , "The value “mwdss” means hourly emissions for one representative Monday, representative weekday (Tuesday through Friday), representative Saturday, and representative Sunday for each month. This means emissions have variation between Mondays, other weekdays, Saturdays and Sundays within the month, but not week-to-week variation within the month"

So, Tuesday represents Wednesday, Thursday, and Friday. So, for the 1/5/2016 (Tuesday) result should match with 1/7/2016 (Thursday). But when I use DEFAULT (A-B) m3diff analysis for emis_mole files for these specific dates, the top differences are:
[Michael@ADEQPlanning ptnonipm_ARDEQ_allday]$ m3diff infileA infileB DEFAULT | grep -w “A:B” newfile | awk -F’@’ ‘{print $1}’ | sort -rk3 | head -10
A:B 9.96206E-07
A:B 9.94815E-07
A:B 9.92429E-06
A:B 9.91113E-06
A:B 9.90092E-06
A:B 9.89054E-07
A:B 9.87940E-04
A:B 9.82778E-06
A:B 9.79428E-07
A:B 9.78180E-05

Is that an issue?

Hi,

The gist of this question is, when you create ptnonipm emissions for every day, should Thursday emissions be the same as Tuesday emissions. This is a good question to ask…

It turns out there is one SCC in the ptnonipm sector which uses a temporal profile which is not the same for every day Tuesday through Friday. It is an airports SCC. So, emissions generated using Thursday temporal profiles will in fact be just a little bit different from Tuesday, but only for that one SCC, and only slightly because that SCC is in just two counties, and because the weekly profile doesn’t have that much Tuesday-Friday variation. So for this reason, the emissions won’t be exactly the same between the days. You are probably better to create some Smkreports to make sure the gist of the emissions totals are consistent.

1 Like

Thank you so much for the explanation. It really helps

@eyth.alison I am trying to get output files for every day in my episode for the airports sector. I tried the fix described here for ptnonipm. I changed my sectorlist to “all” and the M_TYPE,L_TYPE variables to “all”. I commented out MRG_BYDAY and I replaced the final line in the Annual_airports scripts with

$RUNSCRIPTS/emf/smk_pt_daily_emf.csh $REGION_ABBREV $REGION_IOAPI_GRIDNAME -m “$RUN_MONTHS” $SPINUP_DURATION daily

I am getting an error message: SPINUP_LIST: variable not defined

Will this solution not work for the airports sector?

Regarding replacing the final line of the scripts with $RUNSCRIPTS/emf/smk_pt_daily_emf.csh … this particular change should only be done to the ‘daily’ script, not the ‘onetime’ script.

The onetime script should still have “$RUNSCRIPTS/emf/smk_pt_annual_onetime_emf.csh $REGION_ABBREV $REGION_IOAPI_GRIDNAME onetime” as the last line.

We don’t think changing the onetime script would cause the SPINUP_LIST error message though, so it might be something else. Can you provide your airports run scripts and standard output?

Annual_airports_daily_4UDAQ_2017by_17.csh (6.9 KB)
Annual_airports_onetime_4UDAQ_2017by_17.csh (6.8 KB)
directory_definitions_4UDAQ_2017by_17.csh (4.0 KB)
sectorlist_2017by_17.txt (1.2 KB)

Here is my SLURM output, containing the SPINUP_LIST error message.
smk_airports_pt_np_SLURM.txt (325 Bytes)

Thank you for your help with this!

The last line in daily .csh script has stylized quotes around RUN_MONTHS - “$RUN_MONTHS” instead of “$RUN_MONTHS”. That could be the problem.

A great catch! This is fixed! Thank you!