Removing Ozone

[ruk15@psu_edu]

Hello,I want to run a case with no Ozone. In 'Buildconf/cam.buildnml.csh' file, I removed 'D_OZONE:O3' from '&rad_cnst_nl', but I got an error:
"rad_cnst_init: list of radiative gasses must include all radiation gasses for the climate specication"So apparently, I need to specify Ozone. Then I thought that I will specify it, but zero out all the Ozone parameters. So, while specifying 'D_ozone:O3', I did this:
&prescribed_ozone_nl
 prescribed_ozone_cycle_yr              = 0
 prescribed_ozone_datapath              = ''
 prescribed_ozone_file          = ''
 prescribed_ozone_name          = ''
 prescribed_ozone_type          = '

and it gave this error:
"ENDRUN: init_lists: data constituent not found in pbuf: ozone"

So either way, it doesn't like to mess with Ozone.Is there anyway I can not include Ozone in my runs?ThanksRavi KopparapuPennsylvania State University

 

[tilmes]

Hi Ravi,if you want to test a model run that does not have any ozone, you can set the prescribed ozone fields to zero and try that. Simone   

 

[ruk15@psu_edu]

Hi Simone,Thanks. I did that (as mentioned in my above comment). And I got the following error:"ENDRUN: init_lists: data constituent not found in pbuf: ozone"I am using cesm_1_0_5, with E_1850_CAM5 model.    Also, Should I remove D_OZONE:O3  *and*  zero out ozone fields ? Is the way I did to zero out the prescribed ozone  fields the correct way (as shown in my original post)?Thanks muchRavi KopparapuPennsylvania State University  

 

[tilmes]

Hi Ravi,you have to modify  the "prescribed_ozone_file" if you want to change ozone. This means, you copy the one from the model, rename it, change the values to your new values, for instance all zero values, and point to the new file and directory (prescribed_ozone_datapath) in your namelist. Leave everything else as it was before. Your run may very quickly break, since ozone is an important climate gas, good luck, Simone

 

[tilmes]

Hi Ravi,you have to modify  the "prescribed_ozone_file" if you want to change ozone. This means, you copy the one from the model, rename it, change the values to your new values, for instance all zero values, and point to the new file and directory (prescribed_ozone_datapath) in your namelist. Leave everything else as it was before. Your run may very quickly break, since ozone is an important climate gas, good luck, Simone

 

[ruk15@psu_edu]

Hi Simone,Yes, I changed it to very low (1e-20) values, rather than zero.  Then the code gave the following error and quit:*Do you really want to run Hack shallow scheme with a non-standard eddy scheme?*Then, I did shallow_scheme='off', eddy_scheme = 'diag_TKE', and got this error:*ERROR: shallow convection scheme off
  is incompatible with eddy scheme diag_TKE* So, I am not sure what options would work?
 Any suggestions?ThanksRaviPennsylvania State University

 

[andrew]

This seems like the right approach. I have no idea why the Hack scheme gives this message. Something else must be going on.

The parameters for ozone should be left the same. The value could be set to something else.  I could see perhaps that causing issues with the stratospheric heating, but the message is an initialization issue, and should not be even something at run time.

So I have no idea why the code would give an error. I would first verify that the 'new' ozone file works with the same values as the 'original' one, then set the values in the new file to something low and rerun.If it is a run time issue, then the model may not be generating physical radiative solutions without ozone. I am not sure if CAM4 still has a 'top level' for tropospheric physics: but that could be set lower and the Hack Scheme will not run in the stratosphere.

 

[hannay]

By default, CAM prevents to turn on the
  shallow_scheme='off'
  eddy_scheme = 'diag_TKE'but it should be safe to run. Comment the line:call endrun( 'convect_shallow_init: shallow_scheme and eddy_scheme are incompatible' )in vertical_diffusion.F90    ! Check compatibility of eddy and shallow scheme
        if( shallow_scheme .ne. 'UW' ) then
            write(iulog,*) 'ERROR: shallow convection scheme ', shallow_scheme,' is incompatible with eddy scheme ', eddy_scheme
            call endrun( 'convect_shallow_init: shallow_scheme and eddy_scheme are incompatible' )
        endif