WACCM_SC 4xCO2(preindustry) initial condition

[kangwanying1992@gmail_com]

Hi guys,I was trying to run 4xpreindustry CO2 experiment recently in WACCM_SC. I use a boundary condition made from previous 4x(preindustry)CO2 CESM run to force the model but left the initial condition to be the default one. The fact turned out that this inconsistency will cause problem, and model crashed in first several steps. The error report is "Lagrangian path are crossing" at first, and changed to "the enthalpy is not conserved" when I changed nsplitvrm and nsplitrac in the namelist. I was thinking that since this is a very standard experiment, there should be boundary conditions and initial conditions for that somewhere.. I will appreciate to any suggestion!BTW, what I really want is 4xpresentvalue CO2 instead of preindustry CO2... Cheers,Wanying

 

[santos]

I believe that the error "the enthalpy is not conserved" is most likely caused when the model attempts to calculate a saturation vapor pressure outside of the limits of a table in CAM. This will usually occur if you have temperatures above 375K in the model, at a level where the pressure is greater than 1 hPA. It can also occur when the temperature is inside this range, but there is a very large sudden change in temperature, pressure, or humidity that is difficult for the model's cloud feedbacks to deal with.My suggestion would be to run with a lower value of "dtime" for a few model weeks, maybe 900s (as well as using an increased nspltvrm). This will make the model more stable for a while, and hopefully it will "spin up" enough to deal with the sudden increase in CO2. Then you can write out a new initial condition, and use that to run for your increased CO2 case. (You shouldn't use the "dtime" change in future runs, since the cloud physics is tuned for 1800s. However, you may need to keep the increase in nspltvrm.)

 

[kangwanying1992@gmail_com]

Thank you for your suggestion! I will try it. And do you think extrapolate restartfile from previous CESM experiment (low top model) will help?

 

[santos]

"And do you think extrapolate restartfile from previous CESM experiment (low top model) will help?"I don't know. Depending on how you use the low top model data, it might actually be worse (less stable), so I wouldn't do that right now.

 

[kangwanying1992@gmail_com]

Hi Sean,
I even tried to reduce dtime to 10s, and it did not work either... Do you think I could download initial condition somewhere? Or is that possible to make IC file by myself? Will it help if I spin up the model little by little (use multi-year SST and increase CO2 gradually)? Thank you!
Wanying 

 

[marsh]

You may have better luck starting this run if you also multiply the IC CO2 values by 4 using the nco tools. Note, if you are planning on running for 4xCO2 present day values it will be up to you to ensure the CAM-RT code produces scientifically valid cooling rates under those conditions.