Simulations with increased vertical background diffusivity in the ocean

[adagj]

I have run 3 simulations:

1. n1850.ne16pg3_tn14.noresm3_0_beta20.470.ref461.20260626 #470 BDMC2 = 1.8e-5 (default 1.5e-5) and NUBMIN = 5.e-6 (default 2e-6)
2. n1850.ne16pg3_tn14.noresm3_0_beta20.473.ref461.20260630 #473 BDMC2 = 1.8e-5 (default 1.5e-5)
3. n1850.ne16pg3_tn14.noresm3_0_beta20.474.ref461.20260630 #474 BDMC2 = 1.8e-5 (default 1.5e-5) and ksno = 0.2 (default 0.3)

For all simulations, I increased clubb_c8 back to 4.95, from 4.85 in #461 , as a positive TOA imbalance was expected. All simulations turned out rather balanced

ENSO

---

One main concern is the muted ENSO in #470 (compared to the reference simulation #461). That made us reset NUBMIN which seems to help, as #473 and #474 have improved ENSO

Sea Ice

---

The Sea ice extent seems to decrease in the SH (desirable). However, it increased in the NH - in #470 . Hence, to compensate, we reduced ksno in #474 , since this reduction tends to decrease the NH ice volume (desired) and increase the SH ice volume (not desired).However, the variability is rather large.
Oct - Nov - Dec:

April - May - June:

Diagnostics

---

ADF #470 vs #461
noresm_diag #470 vs #461
noresm_diag #473 vs #461
noresm_diag #474 vs #461

[adagj]

@kjetilaas @matsbn @jmaerz @JorgSchwinger @JensBDebernard @DirkOlivie

[JorgSchwinger]

Thank you @adagj

the ksnow does not seem to do anything systematic to the SH sea-ice according to the time-series. So, the concern with ksnow is the ENSO. But #474 is not so bad, e.g. in month 2-10 it is comparable or better than #473, it just misses the month with the highest variability.

So maybe the idea with a compromise ksnow between #474 and #473 is viable?