miscalculated hourly emissions with SMOKE 5.1

callen52 commented 4 months ago

cmv_c3_hourly_test_case_for_github_10jul2024.zip

For CMV sectors, we use hourly FF10 inventories. In SMOKE 5.1, I am finding that the hourly emissions are occasionally being miscalculated, often in the first hour of the day. This was not a problem with SMOKE 5.0.

The attachment contains:

sample annual (annualcsv) and hourly (hourlycsv) inventories, intended to be run through SMOKE for the month of March
sample run scripts (onetime and daily)
hourly Smkreports for one example day when using SMOKE 5.0 (repsmk50) versus SMOKE 5.1 (repsmk51). If you do a diff on the two reports, you'll see where the smk51 emissions are higher. I've confirmed that the smk50 emissions match the hourly inventory and are correct.

hnqtran commented 4 months ago

Hi @callen52, I don't think I have all inputs to replicate the runs. I believe there're at least temporal profiles and repconfig files are missing.

callen52 commented 4 months ago

You can use the temporal profiles/xrefs from 2021 platform for this test (none of the changes in 2022 platform impact CMV): https://gaftp.epa.gov/air/emismod/2021/ancillary_data/temporal_ge_dat_2021hb_17jun2024.zip

Here is the REPCONFIG I used to generate the hourly Smkreports that illustrate the issue: REPCONFIG_for_cmv_bug.txt

hnqtran commented 4 months ago

This issue was found to related to issue #91. Fix implemented for ptagfire test case in issue #91 also fixed the issue for cmv_c3 test case in this issue.

@callen52, would you be able to compile the latest SMOKE github version and let me know if all issues #91, #97 and #100 are fixed on your end? Thanks.

CEMPD / SMOKE

miscalculated hourly emissions with SMOKE 5.1 #100