A problem of handing PMs in FF10 files using SMOKE

Dear all,

I try to process the 2019 emission files using SMOKE 4.8.1. The raw emission files are from the EPA 2019 Emissions Modeling Platform (https://gaftp.epa.gov/Air/emismod/2019/). Those files are in FF10 format, in which PMs are denoted as “PM25-PRI” and “PM10-PRI”.

After running smkinven, the PM2_5.ncf and PM10.ncf files are not generated and PMC (PM10-PM2_5) is not calculated, either.

The test case in SMOKE reads the ORL format files and does not have this problem.

Could you please advise what are the additional configurations / approaches needed to handle this issue? Thank you so much.

The test case scripts provided with SMOKE 4.8.1 were configured many years ago and are not compatible with how we run SMOKE today.

There is a set of portable scripts for the 2017 platform available here, and these are more compatible with how SMOKE is run now:


See instructions here:


I’ve also just copied up a set of unedited, not yet portable scripts for 2019 platform to this folder:


Eventually we will post portable scripts for 2019 but they are not yet available.

You should be able to use the 2017-based scripts and make some updates to incorporate 2019 inventories, or edit the provided 2019 scripts to be more like the 2017 scripts.

I suggest trying for one sector to determine which seems easier.

Thanks so much, Alison, for sharing those scripts. My runs work now.

The correct INVTABLE files should be used and they can be found from the folder “ancillary_data/ge_dat/speciation/invtable_hapcap/” of the 2019 platform.
This file for the 2005 case in SMOKE 4.8.1 doesn’t have “PM25-PRI” and “PM10-PRI”.

I solved this by using "grep PM25-PRI” of the scripts you shared. Thank again.

Great news – I’m glad that it worked…