FAQ: Data ocean slab mode (DOCN-SOM)

[11438023@zju_edu_cn]

Dear Dave,Sorry to bother you again and agin.Can you say something about how the model tools calcualte Q-flux over ice-coverd regions (with thermo-only and thermodynamic sea ice) for SOM?    For thermo_only sea ice, if I want to stabilize sea ice fraction (ignore seasonal change for simpler understanding here), is qdp = surface_SW - surface_LW - surface_SH - surface_LH ?    I found this only works for open ocean.For thermodynamic sea ice, I checked the code and still trying to understand.  Best wishes

 

[mmg62@psu_edu]

Hi There,I have a question about the q-flux forcing file pop_frc.b.e11.B1850C5CN.f09_g16.005.082914.nc I am looking to use this to run an E-compset case and also create an F-compset case whose boundary conditions (SSTs and SICs) are identical to those used to produce this forcing file.  I was wondering what years of the b.e11.B1850C5CN.f09_g16.005 control run where used to produce the q-flux forcing file?  Also, are the SST outputs from the b.e11.B1850C5CN.f09_g16.005 control run anywhere on Cheyenne that I could access?  Otherwise I can get the history files from HPSS no problem.  Thanks!  

 

[dbailey]

I realize that I forgot to add the years to that SOM file. I believe it was years 1450 to 1500. However, I would recommend using the one dated 150217. This is from years 402 to 1510! Yes, over a thousand years. The SST from the CESM-LE control run is in single variable format under:/glade/p/cesmLE/CESM-CAM5-BGC-LE/ocn/proc/tseries/monthly/TEMPUse level 1 of the ocean TEMP field.Dave

 

[mmg62@psu_edu]

Hi Dave,This is great.  We will use the 150217 file then.    Thank you so much!Cheers,Melissa

 

[chenyh1991@126_com]

Hi, DaveI want to use ETEST（2000_CAM60_CLM50%SP_CICE_DOCN%SOM_MOSART_SGLC_SWAV_TEST）to explore the effects of sea surface temperature on sea ice-atmosphere-land surface for the time span from 1979-01-01 to present, but the SOM data, as you mentioned ， does not properly apply to scientific research in 2000 . What datas in the DOCN-SOM or in cmip5 is more suitable for my research questions ? Could you give me some pertinent advice? 

Thanks a lot
Yuhang Chen

 

[dbailey]

We don't do 2000 controls anymore, so I would recommend using one of the 1850 control SOM datasets. This is a colder climate of course, but at least it is well equliibrated.Dave

 

[rchemke]

Hi all,I would like to run a large enamble of SOM with historical forcing, but to use the same Qflux as in the PI control run. I thought to take the  E1850LENS, and change the forcing to historical. This way, the run will be consistent (same QFLUX and MXLD) with the PI E1850 run, and by that I will be using the same atm/lnd and ice model versions as in the full LENS runs, which allows a proper comparison between the two. The E1850 LENS run uses the pop_frc.1x1d.090130.nc which is in /glade/p/cesmdata/cseg/inputdata/ocn/docn7/SOM. In this directory there are also many files related to the B1850 LENS runs (e.g., pop_frc.b.e11.B1850C5CN.f09_g16.005.150217.nc). Which file I should use for the QFLUX and MXLD? Thanks Rei

 

[dbailey]

I would definitely recommend using the same slab ocean forcing in all your runs. The E1850LENS is set up out of the box to use that file, but I would not recommend that as I say in the FAQ at the top of this forum. There are only two from the LENS control run (b.e11.B1850C5CN.f09_g16.005). The first, dated 082914 was 50 years around 1050-1100 in the control run. The second, dated 150217 is from years 402-1510, so 1100 years. I would recommend the latter. Do an ncdump -h on these files and there is documentation in them.Dave

 

[Rei]

Hi,
I am running the CESM SOM version under historical and RCP85 forcing. When I start a new run, using restrat files from the previous year (not continue a run, but a new run), there is a discontinuity in surface temperature and other fields. I notice that when I moved from historical to rcp8.5, but this also happens under the same frocing, if I start a new run. Any ideas why? Does the SOM does not use the restart files? Does it use the temperature from the forcing file?

Thanks!!

 

[dbailey]

This is an interesting question. I believe since the RCP run is started as a hybrid, it goes back to the initial SST in the forcing file. The SOM only uses the restarts for continue or branch runs. I guess the SOM was never really intended for this purpose. What is the SOM forcing file you are using? The 1850 piControl file? You could fool it by modifying the SST in the SOM forcing file offline and replacing it with the SST from the restart at the end of the 20th century. I kind of think this is not a great idea.

 

[Rei]

Thanks for the reply, David! I have three more questions:
1. If I follow your suggestion, how do I regird the somtp variable in the docn restart file (which is now a long vector of 122880) to the grid of the forcing file (pop_frc.b.e11.B1850C5CN.f09_g16.005.150217.nc) gx1v5? Also, the domain file (domain.ocn.gx1v6.090206.nc) is on gx1v6. Does the SOM remap the forcing file?
2. Is it a bad idea to start the RCP run using branch instead of hybrid? If its not a bad idea, would then the SOM use the restart file?
3. Lastly, when I run an ensmble and change the pertlim variable, in order to use the restrat SOM tempearture, can I also run the different memebrs using brnach and not hybrid?

Thank you so much for your help!!

 

[dbailey]

We have not supported this method before, so I am not sure how best to do it. Maybe you could force a restart read on the hybrid. CAM will not let you do a branch with the change in forcing to RCP8.5. So, these all have to be hybrid regardless.

 

[Rei]

Thanks! I forced the SOM to read the restart file even under hybrid, and it works well: no discontinuity in SST.

 

[11438023@zju_edu_cn]

Hi,

Is that possible to diagnose the Q-flux pattern using CMIP6 archive data, which could then be used in the slab ocean mode?

For example, I would like to diagnose the Q-flux pattern from the last 20-year of the SSP585 case. Can I use the code provided in the repository to do so?

https://svn-ccsm-release.cgd.ucar.edu/model_versions/cesm1_1_1/models/ocn/docn/tools/pop_som_frc/

Best,
Lei

 

[dbailey]

The code there should still work with CESM2 runs as the ocean component is still POP. However, there is always the issue of using a Q-flux from a climate that is not balanced. That is why we recommend using only an 1850 control run.

 

[11438023@zju_edu_cn]

The code there should still work with CESM2 runs as the ocean component is still POP. However, there is always the issue of using a Q-flux from a climate that is not balanced. That is why we recommend using only an 1850 control run.

Thanks for your responses. I am wondering what do you mean by "issue of using a Q-flux from a climate that is not balanced"? I understand that the global and annual average Q-flux from transient simulation would not necessarily be zero. But are there any other potential issues?

 

[dbailey]

That is the primary issue. Have a look at the Bitz et al. paper for more on this.

 

[cpatrizio]

Hello,

Where can I find the surface heat fluxes associated with the slab-ocean model? For example, there are surface latent heat fluxes provided in the history files in both cam and ice directories, so I'm not sure which ones to use...

Thanks!
Casey

 

[dbailey]

The DOCN-SOM does not write history files, so you do not get fluxes over the "ocean". The CICE ones are over the sea ice only. The CAM ones are merged from land/ocean/sea ice. So, where there is no sea ice or land, this is the same as just over the ocean.

 

[cpatrizio]

Hi again!

Just another quick question: I am trying to specify my own SOM forcing file using the following instructions provided in the FAQ:

3. What if my SOM forcing file is in a different directory?

Here you will need to modify the docn.streams.txt file to manually point to the directory you wish. As of CESM 1.1, you can copy $CASEROOT/CaseDocs/docn.streams.txt to $CASEROOT/user_docn.streams.txt and edit it.

However, it seems that in CESM2.1 this does not work. I've edited the user_docn.streams.txt to point to a new directory (my own directory), but after I build the directories shown in /Buildconf/docn.input_data_list still point to /glade/p/cesmdata/cseg/inputdata/ocn/docn7/SOM/.

Thanks for the help!
Casey