Repeat-year (RYF) boundary forcing for regional models - how do we loop back over this forcing?

COSIMA Ocean
1

Hi!

I’ve forgotten how we handle the boundary forcing for mom6 regional models in an RYF simulation. Specifically, can we use the same 365 day forcing segments taken from an OM2-01 RYF run, and somehow loop over this (as we do with the JRA55 atmospheric forcing), or do we need to stitch together multiple repeat years such that our forcing times continue on into an artificial 1992, 1993 etc.

This question arose after I ran the last month of year 1 for an ryf 1991, and hit an error whereby the MOM6 clock was looking for forcing at a time 1991-12-31 12:07:30 however the forcing ends at 1991-12-31 12:00:00 - hence it had no future forcing to interpolate between.

I see the options as:

  1. Find a (potentially non-existent) missing flag that MOM6 has for looping over a 365-day RYF forcing file?
  2. Adding a single day (1st Jan 1991) to the end of the forcing files so that we can complete a single year. Then manually change dates from 1991 to 1992 to appease the MOM6 clock.
  3. Take multiple years from an OM2-01-RYF simulation - essentially what we do for IAF runs.

@ashjbarnes - how did you get around this for your runs?

Thanks!
John

2

Hi John,

I’ve never run any ryf models myself before. I do know you need to flick a switch on rmom6 package boundary forcing function to set ryf=true and i think this changes the metadata encoding for the time field to be modulo

Never tested this myself though

1 Like
3

Yep that sounds exactly like what I’m missing. Have just hit run on the last month so hoping it will magically loop back around now that the modulo attribute is present in the time dimension of all the forcing files. Thanks Ashley!