At today's co-chairs meeting, @klindsay28 raised the issue that MOM's history diagnostics are currently wrong if you try to restart in the middle of a history averaging period, because FMS doesn't provide a capability for restarting these history fields. @mvertens feels that this is not a strict CESM requirement, but I'm concerned that this could cause problems for users – including possibly-hard-to-detect incorrect results – if this makes it into a CESM release.
Therefore, I'd like to propose that, if this history restart functionality can't be added in time for the CESM3 release, that a check be added to MOM to abort if it is detected that a user is trying to write a restart file in the middle of a history-averaging interval. (We may want to allow users to override this abort by setting an xml variable or a namelist flag that says they know what they're doing.)
billsacks
commented
2 years ago
At today's co-chairs meeting, @klindsay28 raised the issue that MOM's history diagnostics are currently wrong if you try to restart in the middle of a history averaging period, because FMS doesn't provide a capability for restarting these history fields. @mvertens feels that this is not a strict CESM requirement, but I'm concerned that this could cause problems for users – including possibly-hard-to-detect incorrect results – if this makes it into a CESM release.
Therefore, I'd like to propose that, if this history restart functionality can't be added in time for the CESM3 release, that a check be added to MOM to abort if it is detected that a user is trying to write a restart file in the middle of a history-averaging interval. (We may want to allow users to override this abort by setting an xml variable or a namelist flag that says they know what they're doing.)