How to generate smk_merge_dates for 2020

Hi all,
I am trying to run 2020 with 2017 NEI data. Do I need to replace “MRGDATE_FILES” with “smk_merge_dates_202001.txt”? How to generate the “smk_merge_date” for 2020, then?
Thanks so much, and any suggestions are appreciated.

We have posted these here:

1 Like

Alison, Thanks so much. That solved my problem.

Hello Alison, just a follow up question – is there a way to generate files that are not on the FTP server? For example, how would one create the merge dates files for 2021?

To model 2021 you will also need a 2021 holidays file. We have posted one here that includes 2021:

To make new merge dates files for either of those years, download the SMOKE utilities package from 2016v1 platform:

Navigate to the smk_merge_dates/ directory in that package, where there is a script to run and a README. The README says the script uses the holidays file but it actually does not anymore. The script attempts to figure out the holidays on its own although the calculations may not be 100% accurate for every year. If holidays are not correct, you might get errors – not when running the sectors themselves, but when attempting to generate annual post-SMOKE reports and when running the final sector merge.

In that case you might need to make some minor tweaks to the smk_merge_dates files after the fact

Let us know if you have issues with running the script.

1 Like

Thank you! Yes, I could figure out how to generate the holidays file by just looking up the dates of holidays, however, it looks like this script did not generate the correct dates for easter, thanksgiving, and memorial day of 2021. However, for my simulation, I will be using August/February which do not have these holidays … so perhaps this error may be circumvented. However, my spin up period does start in July, so is there a way for me to check to see if the merge dates file doesn’t make sense?

What error are you specifically referring to? The incorrect dates for some holidays?

Sorry for being unclear – I mean how could I double check the representative date file to be correct? Or is that not a known error for this script? Thank you!

Probably the easiest way to check the accuracy of the date files is to run a sector with representative dates and see if the dates that are created by SMOKE match the merge date files. Two good sectors to run would be pt_oilgas (compare with the mwdss_Y column in the date files) and nonpt (compare with week_Y).

But if you are only running Feb and Aug, you won’t really need to worry about the holidays, assuming that your spinup doesn’t go far enough back into July to include the July 4th weekend.

1 Like

Ok perfect, thank you!