Cooling trend with BGC-BDRD in CESM1.2

[hshin74]

Hello. I am a researcher in South Korea.
Our research team planned a CO2 ramp experiment with CESM1.2.2.

To get a baseline experiment, we ran the model for 800 years fixed at 2000 with BGC-BDRD turned on.
Using the baseline output, we set an initial condition for the CO2 ramp experiment.
Unfortunately, our experiment with increasing CO2 resulted in cooling at the surface.
The simulated global mean surface temperature (GMST) had decreased almost monotonously till the end of simulation that started in 2001 and ended in 2063.
When I tried the baseline experiment using CAM4 (one of the out-of-box compsets) as below, I got a cooling trend of GMST as well.

Baseline:
B_2000_CAM5_CLM4%CN_CICE_POP2%ECO_RTM_SGLC_SWAV_BGC%BDRD (44yrs)
==> GMST oscillates (I assume that this is normal.)

Experiment:
B_2000_CAM5%RCO2_CLM4%CN_CICE_POP2%ECO_RTM_SGLC_SWAV_BGC%BDRD (63yrs)
==> GMST decreases

Out-of-box compset(B_1850_BGC-BDRD):
B_1850_CAM4_CLM4%CN_CICE_POP2%ECO_RTM_SGLC_SWAV_BGC%BDRD (20yrs)
==> GMST decreases

I have tried to figure out what kind of mistake I made in there, but with no fruit. Can you please give me some advice?

Thank you in advance for your help!