how can i change cam's taux,tauy in restart file?

[zhouqian09@gmail_com]

Hi!  I am using CESM1.0.3 with compset BGCN2000, I want to change atm’s taux, tauy in restart file. When I check rest file, and find:bgcn2000.cam2.i.0051-01-01-00000.ncbgcn2000.cam2.r.0051-01-01-00000.ncbgcn2000.cam2.rs.0051-01-01-00000.nc1.      in file bgcn2000.cam2.rs.0051-01-01-00000.nc, I find var: x2a_Faxx_taux and x2a_Faxx_tauy, I have changed these two vars, but the results do not change. I want to know how these vars are used in the model, and why this method doesn’t work?2.      I check rpointer.atm, and find that bgcn2000.cam2.r.0051-01-01-00000.nc is the restart file for cam,  ncl_filedump this file, I find tauresx and tauresy, but I am not sure what these two vars stand for, and how can I change taux, tauy?

 

[rneale]

The surface stress variables (taux, tauy) are derived variables in CAM based on the near-surface wind and atmopsheric state and so are not used to initialize the model. Therefore chaging them on either the restart or initial files will not change the evolution of the run. 

 

[zhouqian09@gmail_com]

Thank you!So I have to change U V at surface level to archive that ? 

 

[rneale]

I am not sure what you want to do exactly. Chaging U and V will change the evolution of the model as they are used in the model's initial state. It will also change the evolution of the surface stresses. Are you wanting to somehow fix the surface stresses?  

 

[snowyang1988@gmail_com]

Hi Neale,I want to change the TAUX and TAUY value. But I donnot know how to calculate  them ? Can you give me a guide?  Thank you!Jeff

 

[rneale]

TAUX and TAUY are calculated in the CAM routines in and below vertical_diffusion.F90. They are caluclated via the implicit solver in compute_vdiff.

 

[snowyang1988@gmail_com]

Hi Neale,Thank you for your help. Another question:I need a initial field for a certain day (such as 20130516), but I donnot have  the ncdata. So I want to change the ncdata (initial field) usingreanalyze data. But  I find there is no U/V value in the ncdata.( just like $DIN_LOC_ROOT/amt/cam/inic/fv/cami_0001-01-01_4x5_L26_c060608.nc) How the model work with no initial data for U/V? And if so, just changing the other varibales (T and some chemicals) in the ncdata is OK? Regards,Jeff

 

[eaton]

The FV dycore uses a staggered grid.  The U/V velocity components are in the variables US/VS which are located on the staggered grid.  Use an existing cami_* file to determine those locations.  And be sure to use those names in any initial file you create since the FV initialization is looking for variables with those names.