Does hybrid continue run need spin-up?

[sunweihao]

Dear scientists,
I am running an AGCM experiment with compsets F2000Climo. I have already run 10 years to spin-up. Based on the 10th year restart files and initial condition, I tried to run ensemble experiments for 1 year. In previous attempts, I set RUN_TYPE=branch and edit the pertlim in user_nl_cam to add perturbation. But I found that the different experiments' output is the same.
I think that I have to set have Run_TYPE to hybrid and perform ensemble experiments but I have question:

The initial and restart file is the 10th year output, in which the model has experience spin-up and reach stable. When I hybrid run, should I spinup again? In other words, whether the first year output in the hybrid run is available? Thank you.

Regrads

 

[altuntas]

@hannay can correct me but I believe that pertlim works only for initial simulations and not for hybrid runs. You should also add a different pertlim value for each ensemble member.

 

[kezhoulumelody]

Hi there, I have a similar question related to hybrid run.

If I have a control run and I want to do an abrupt4xCO2 experiment. What's the difference between an abrupt4xCO2 branch run and an abrupt4xCO2 hybrid run? Say I set the $RUN_REFDATE in the hybrid run same as the date I start the branch run. Imagine both abrupt4xCO2 branch experiment and abrupt4xCO2 hybrid experiment run for 30 years. Are the 30-year averaged changes more or less similar in both experiments? What about the results averaged in the first month?

Thanks,

Melody