Main menu


Lightning output

10 posts / 0 new
Last post
Lightning output

Dear CESM exports,

I'm learning how to get the lightning data from CESM for the past 60 years.  At the same time, I want to let the temperature match the observation results. Because I'm also interested in temperature results.

I have several questions, hope that I can get some hlep. Thank you very much!

1) I think the best compset meets my need is B20TRCNCHM. Because I think the temperature sensitive needs a transient and ocean coupled run. I don't know exactly what's the different between B and BCN compset, but lightning calculate exists in chemistry source directory, so I think I should use a compset with CHM, is this a correct choise of compset?

2) When made a try with B20TRCN or even B20TR, I set RUN_STARTDATE to 1930 in env_run.xml. But the result showed a decrease at the beginning and did not increase between 1940-2005. By the way, when I try to set a year number after 1930, I don't know why it failed when built. Does RUN_STARTDATE='1930-01-01' mean I start with all the transient date, such as CO2 concentration, of 1930's ?

3) So now I try all these compsets with default RUN_STARTDATE 1850, which cost a lot of time... Does the default setting for a compset for transient 1850-2005 mean that I can get a quick start from 1850 without spinup? I wonder because there still seems some unbalance in the beginning of my run. If B20TR needs a spin-up, shall I spinup with this setting, or I should spin-up with B1850, and use the result as restart file in B20TR?

4) I want to run my case from 1950 to 2010, but the transient case gives a maxmum of 2005. I read some threads in forum and found that I should use something called F_AMIP to solve this problem. But didn't F compset means priscribed SST and CICE? Can I use F to model temperature? Will it get good temperature result?

5) When try B20TRCNCHM case, I succeeded built the case, but just failed to run with a severe(174) error. The log file is pasted in attachment. Hope there's some clue to solve it. Or is there any other way I can use to get lightning result?

Part of the log:

 ionf_mod.F90         245 CFILE closing :     -1376256
 piolib_mod.f90        2817 close:      1376256
 ionf_mod.F90         245 CFILE closing :      1376256
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source
cesm.exe           000000000067A29D  mo_flbc_mp_flbc_g         825  mo_flbc.F90
cesm.exe           00000000005E52D2  chem_surfvals_mp_         459  chem_surfvals.F90

6) As a new learner for CESM model, I felt very uncertain for every setup I did, even after I read some tutorial files and CESM user's guide. Such as B20TRCN and B20TR, I don't know what's CN. The compset setting said that CN is "with CN (Carbon Nitrogen) in clm". Does this mean carbon cycle and nitrogen cycle will be considered in land part of model? So is there any advices about how to begin the study?

Thanks again,




For the question 3, I found that if I change to a F compset, the surface temperature will not change dramatically in the beginning of run.

But I'm still wonder why couldn't B compset do this like I expected? Both B1850 and B20TR compsets' results are lower than observation for around 2 degree after a 150 years' run.

Hope to get some replies. Thanks a lot!

Best regards,



It is very different to run a F compset and a B compset.
- In a F compset: you run an active atmoshere and land model with prescribed SSTs and prescribed sea-ice coverage
- in a B compset, instead of prescribing the SSTs you use a ocean and ice model. This is what we called teh fully coupled mode (active atmosphere, land, oecan and ice).
If you don't need a fully coupled run, I would recommend that you use a F compset instead of  a B compset.

CN means means that you use Carbon-Nitrogen cycle in the land model. If you have the option CN, CLM (this is the land model) includes a fully prognostic treatment of the terrestrial carbon and nitrogen cycles.

If you want to start your un 1950, you can set the RUN_STARTDATE='1950-01-01'.





Thank you, Hannay!

I'm not sure if I need a fully coupled run. I think F will give me a good result, but if I want to run after 2005, I may need to search for observation results and set the prescribed SSTs and sea-ice coverage myself? How do the future predict compsets work?

Best regards,



Hi Hanny,

I am wondering whether you have the F_AMIP hybrid/branch f09_f09 refcase restart data of 2005-01-01?



Yu Huang


I assume from the comspet name you are looking for cam4 data
If you have access to Cheyenne, there should be data in:


Hi Hanny, thank you for your information! 


May I ask one more question about input data? From your CESM1.0 experiments website, I can see that the initialization file for RCP cases should be from the b40.20th.track1.2deg.001. But in my runs, the initial file macthed by the model for cam is /inputdata/atm/cam/inic/fv/ and I do not know what does this file represent. Can you provide the path of b40.20th.track1.2deg.001 restart files on 2005-01-01 or 2006-01-01?


Thanks a lot!



Yu Huang


Restarts from b40.20th.track1.2deg.001 are on the NCAR HPSS under



 However, the NCAR HPSS is currently down. 

Hi strandwg,

Thank you a lot for letting me know about this info. I am wondering if the hpss is down, then what should I do to get the files that I need?

Thank you very much!


Yu Huang


The NCAR HPSS is now back online and in service.

I recommend subscribing to the relevant CISL-provided notification services.

Log in or register to post comments

Who's new

  • hellosyndy@...
  • soorajkrish90@...
  • joao.bettencourt@...
  • oumzh6@...
  • turuncu