High radiative imbalance in pre-industrial CESM2-CAM5

[NiCosentino]

Hi,
I've run 100 years of a pre-industrial F compset, on CESM2.1.3 with CAM5 and 1-degree resolution (f09_g17), and monthly output files. The radiative balance (FSNT - FLNT) is giving me an annual/global average of approximately +2.1 W/m2, with oscillations in the range 1.6-2.4 W/m2 during the 100 years, with no trends. I've read that this may be too high for pre-industrials. Is it so? If so, may it be related to the fact that I am running CAM5 in CESM2.1.3? What do you suggest me to do if it is indeed too high?

 

[NiCosentino]

UPDATE: the surface energy balance (FSNS-FLNS-LHFLX-SHFLX) gives me an annual/global average of +3 W/m2, with oscillations in the range 2.4-3.5 W/m2 during the 100 years (no trends). Thanks in advance for any comments.

 

[hannay]

As a rule of the thumb a RESTOM of about 2 W/m2 is a F compset is ok (where RESTOM=FSNT-FLNT). This is because the SSTs are prescribed and as a result, the model cannot drift much. In a B compset, this is another story. For a pre-industrial B compset, you would need to keep the RESTOM < 0.2 W/m2 to avoid drift of your simulation.