Significantly fewer large-fire IDs in 2024 than in 2023, 2023 has some IDS with only the npf-list file

[mccabete]

I am still diagnosing if this is an issue or not. As part of passing on data to collaborators, I have been assembling dataframe composite of all largefire files in 2022 and 2023. In 2024 we began making composite files of all larege fires automatically, which are much easier to work with, but do have some discrepancies.

The composite laregefire file is being made from reading in the paths of the unique fireID from the largefire files. The 2024 file is made from the allfires object (I think). The number of unique IDs in the 2023 files is 4518, but for 2024 there were only 1741 unique IDs.

Lots of stuff could be happening here, 2023 and 2024 could just be different fire years, but I think this could point to something systematic.
It's also possible that by generating the composite file from the allfires object that we either intetionally or unintentionally absorb any filtering that is happening in the allfire object creation.

• Does the allfire object contain largefires from the start of the season? (If not, then my 2023 FireIDs could be early-season fires)
• Does the allfire object filter out fires that have merged together? (Which would mean many of my 2023 FireIDs are redundant with later versions of the same fire)

[mccabete]

Several 2023 fires only have nfplist files and nothing else. I am not sure why this might happen.

Current IDs: 94118, 164769, 98252, 154778