I am running a year-long case using EDGAR as my emission source.
The SMOKE is generating one file per day during temporal program, and this is consuming a lot of HD (around 8Gb per file).
I was wondering if there is a way to run smkmerge for the year using, for example, seven temporal intermediate files, one for each day of the week, instead of 365 files.
For example:
January 4th and January 11th of 2021 are both Monday, and they have the same temporal profile, so the smkmerge could use the same temporal intermediate files (atmp and atsup) to process the emissions.
Yes, you can run SMOKE only for representative days. EPA NEI modeling platform has been using that approach from the beginning for the same reason. You can review how the behind-scene scripts work to run for those representative days. You can find some details from the README file from the NEI emission modeling platform. Check out sections 4B, 4C, 11, and 12.
For EDGAR, for the 2-D sectors, EPA runs representative days - 7 days per month, no holidays. For 3-D sectors, EPA runs Temporal for representative days, and then runs Smkmerge daily to support how SMOKE does the layering. You can run EDGAR this way to cut down on the disk space requirements.
If you are using EPA platform scripts use these script settings:
L_TYPE = week
M_TYPE = all (energy, industry, ships, air) or week (ag, transport, residential)
We haven’t recently posted a hemispheric platform, but some scripts are available from a 2011 platform package in the file that starts with smoke that could be used as a starting point: