POP2 standalone with monthly restoring

[bklinger@gmu_edu]

My student has been trying to get POP2 to run with namelist parameters formulation='restoring' and datatype='monthly' (eg, restoring to annually varying fields).  She tried this with compset G (POP2 + CICE) using lcoupled=.t. and lccsm=.t. and got an error that datatype must be 'none'.  She ran with compset C (POP2 only) with same variable values (error: must have formulation='partially_coupled' or datatype='none'), then changed lcoupled to .f. (error: must have lccsm=.f.), then changed lccsm to .f. (error: not compatible with cpp variables), removed DCCSMCOUPLED from pop2.buildexe.csh but did not help.Is there a way to run POP2 alone (or with just ice) with monthly restoring forcing?  If so, what is it?Thanks, B A KPS Also changed datatype to 'monthly_calendar' and still didn't work (compset C, lcoupled=.t., lccsm=.t., formulation='restoring').

 

[bates]

Barry,Your student needs to set shf_formulation and sfwf_formulation to 'partially-coupled' and shf_data_type and sfwf_data_type should be 'monthly.' If you put 'restoring' for the *_formulation, the model assumes that you are running completely uncoupled, but with the default settings, this isn't even possible right now. Within the CESM framework, one cannot run POP in a stand-alone mode. From the information in your post, it looks like you want to force POP only via restoring to some SST and SSS data. With the present framework and without code modifications, this cannot be done cleanly. Using the settings just mentioned, you can use "large" strong restoring in both T and S, but this will occur in addition to the use of other flux components. However, if the restoring coefficients are large enough, the expectation is that the restoring terms will dominate the surface fluxes. To do this, you would need to set

     sfwf_weak_restore = 0.

     shf_strong_restore = ~100 (or larger) W / m^2 / C

     sfwf_strong_restore = ~ 0.67 (or larger) kg / s / m^2 / msu

     where msu (model salinity unit) = psu (or ppt) / 1000.

   Refer to Large et al. (1997, JPO, v27, 2418) for further details.

   All of this said, POP can be modified to set all other flux components to zero, but this requires
   some detailed knowledge of the code.This response was pieced together from information from both me and Gokhan. Let us know if you need further clarification.Susan

 

[bklinger@gmu_edu]

I think we should do a clean run with ocean-only rather than a kludge.  Does anyone know if the LANL release of POP works on Yellowstone? 

 

[bates]

Barry,I don't know if the LANL version works on yellowstone. Perhaps someone at LANL might be helpful. The LANL version is not necessarily the same as the NCAR version of POP.
Susan

 

[njn01]

Stand-alone POP2 can be downloaded from the LANL code repository by following the links on thehttp://oceans11.lanl.gov/twiki/bin/view/Cosim/CosimModels page.
Susan is correct that this version of POP2 is different from the CESM version, and I agree that you should contact LANL to determine if anyone there has run the stand-alone version on yellowstone.