Backfill HRRR OPS 6-hourly precip verification

[JeffHamiltonNOAA]

Due to an error in how the valid times were matched up in our sub-24 hourly precip application, all of our 6 hour accumulations of HRRR OPS precip were erroneous from the onset of the application. I've removed all of the bad CTCs from our database prior to the fix being put in place in late 2023. However, the modelers have requested backfilling from 2019 onward, with higher priority data periods as follows:

• March-April 2024 (although the algorithm was corrected by then, an outage occurred, likely due to the rocky Rocky 8 transition)
  • June 1, 2023-August 7, 2023 (two of three month-long RRFS retro evaluations)
  • February 2022 (third month-long RRFS retro evaluation)

In addition, RRFS retros were recently run over these shorter time periods:

• April 19-27, 2023
• August 26-31, 2023 These weeks are a lower priority than the first list but may get some attention.

After these periods are done, fill in the gaps back to 2019

[JeffHamiltonNOAA]

I have the Feb period reprocessing and stats look reasonable compared to later Feb periods thus far. I'll target the Rocky8 outage time period next and further investigate statistical consistency with the before and after periods

[JeffHamiltonNOAA]

I have the Feb period reprocessing and stats look reasonable compared to later Feb periods thus far. I'll target the Rocky8 outage time period next and further investigate statistical consistency with the before and after periods