Afdust does not finish

I want to run afdust for the whole January month of 2019, but my afdust run stopped at day 13. The log file is attached here(afdust.log.txt (81.5 KB)). In the /intermed/afdust/logs i have the following files:
grdmat_afdust_2019ge_cb6_19k_12US1.log smkmerge_afdust_2019ge_cb6_19k_20190108_12US1_cmaq_cb6ae7.log smkreport_afdust_2019ge_cb6_19k_invgrid.log
helper_scripts_list_afdust_2019ge_cb6_19k_12US1_all.txt smkmerge_afdust_2019ge_cb6_19k_20190109_12US1_cmaq_cb6ae7.log smkreport_afdust_2019ge_cb6_19k_inv.log
smkinven_afdust_2019ge_cb6_19k.log smkmerge_afdust_2019ge_cb6_19k_20190110_12US1_cmaq_cb6ae7.log smkreport_afdust_2019ge_cb6_19k_temporal.log
smkmerge_afdust_2019ge_cb6_19k_20190101_12US1_cmaq_cb6ae7.log smkmerge_afdust_2019ge_cb6_19k_20190111_12US1_cmaq_cb6ae7.log spcmat_afdust_2019ge_cb6_19k_cmaq_cb6ae7.log
smkmerge_afdust_2019ge_cb6_19k_20190102_12US1_cmaq_cb6ae7.log smkmerge_afdust_2019ge_cb6_19k_20190112_12US1_cmaq_cb6ae7.log temporal_afdust_2019ge_cb6_19k_20190101.log
smkmerge_afdust_2019ge_cb6_19k_20190107_12US1_cmaq_cb6ae7.log smkmerge_afdust_2019ge_cb6_19k_20190113_12US1_cmaq_cb6ae7.log timelog_afdust_2019ge_cb6_19k_12US1_all.txt

Everyone of them ends with “Normal Completion”. In temporal_afdust_2019ge_cb6_19k_20190101.log, the file ends on day 13 (attached here
temporal_afdust_2019ge_cb6_19k_20190101.log.txt (73.3 KB))

I have no idea what caused this stop and how to fix it. Can you help with it?

Thanks,
Lin

It looks like SMOKE ran successfully.

The afdust sector uses a “week” temporal approach, meaning that the temporal profile varies by day of the week. For the month, emissions on all Mondays will be the same, all Tuesdays will be the same, etc. Because the emissions are the same only one file per day of the week is output by SMOKE for this temporal approach. The first week of January includes two days classified as holidays, 1/1 and 1/2, which are written as separate files because this is a holiday inclusive approach (“week_Y”). The annual representative date assignments for each temporal approach can be found in the merge dates files.

Afdust is a special case because it is usually post-processed with transport and meteorological adjustments (“afdust_adj”) that bring down the total primary PM. If the adjustments are applied the resulting emissions files will be daily because the meteorology, and thus the emissions, will vary across all days of the month. As a next step you should run the adjustment scripts .

1 Like

Thanks so much! I ran afdust_adj and met the error below:


I followed Afdust_adj NCVGT Error (NetCDF dimension) - #2 by bbaek but my problem still exists. Could you help with this?
Thanks,
Lin

Unfortunately it looks like the pre-compiled executables are not compatible with your system.
(See #2 under Troubleshooting at Running SMOKE using EPA's Emissions Modeling Platforms · CEMPD/SMOKE Wiki · GitHub)

If you are using the 2019 modeling platform you will need to recompile smoke4.8.1/scripts/afdust_adj/apply_precip_adj.f and smoke4.8.1/scripts/afdust_adj/mult.x on your system.

I recompiled it and now it works. Thanks!
Lin

Since there is no error, we believe that this is due to the use of representative days…

The initial afdust job (pre-adjustment) only creates emissions for representative dates.

So there should be 9 days’ worth of emissions in January (one per day of week, plus 1/1 and 1/2 holidays), which is looks like there is.

The second afdust job which applies the xportfrac and met adjustments should create emissions for every day.