Opened 9 years ago

Closed 8 years ago

#475 closed help (fixed)

Running the UM 6.1 in idealised mode on Hector

Reported by: bethan.white Owned by: willie
Priority: high Component: UM Model
Keywords: Idealised 6.1 Cc:
Platform: UM Version: 6.1

Description

Hi, I've been trying to set up the UM 6.1 in idealised mode on Hector. I have taken one of my working jobs, set it up in idealised mode and included a set of modsets that Carol Halliwell has given me. These are:

idl601_rhprofileq_dk.mf77
idl601_diurnalfluxesopt3.mf77
idl601_bubble3.mf77

When I tried to run this setup, the model would not compile. The .comp.leave file (xfdta000.xfdta.d10229.t092550.comp.leave) had the following errors:

Error : Delete ANG0F503.202 deletes already deleted text in deck ATMPHB2 (Modset AFRAF602)
Processed ATMPHB2 : 1 error(s) , 0 warning(s).

Error : Delete AWA0F505.1430 deletes already deleted text in deck SFEXPL8A (Modset AFRAF602)
Processed SFEXPL8A : 1 error(s) , 0 warning(s).

Error : Delete AWA0F505.1348 deletes already deleted text in deck BDYLYR8A (Modset AFRAF602)
Processed BDYLYR8A : 1 error(s) , 0 warning(s).

I've had a look through the modsets and these error appear to be coming from idl601_diurnalfluxesopt3.mf77. I commented out the relevant lines in the modset, but trying to compile the model then resulted in the following error (xfdta000.xfdta.d10229.t182207.comp.leave):

pathf95-197 pathf90-3.0: ERROR BDY_LAYR, File = bdylyr8a.f, Line = 420, Column = 8

Unexpected syntax: ", or EOS" was expected but found "L".

pathf95-550 pathf90-3.0: ERROR BDY_LAYR, File = bdylyr8a.f, Line = 420, Column = 15

"IDLSURFFLUXSEAOPTION" has the INTEGER(KIND=8) attribute, therefore it must not be declared with the REAL(KIND=8) attribute (identifier first appeared at line 128)

pathf95-197 pathf90-3.0: ERROR BDY_LAYR, File = bdylyr8a.f, Line = 420, Column = 57

Unexpected syntax: "object-name" was expected but found "EOS".

make[1]: * [bdylyr8a.o] Error 2

Do you know whether commenting out the relevant lines in the original modset was the correct approach to take, or should I be doing something else to get it to compile?

My hector username is eebaw and the job is xfdta.

Thanks very much,
Bethan.

Change History (16)

comment:1 Changed 9 years ago by willie

  • Owner changed from um_support to willie
  • Status changed from new to accepted

Hi Bethan,

Your modsets are conflicting. You could try switching off the offending one entirely, but you should really discuss the modsets with Carol.

Regards,

Willie

comment:2 Changed 9 years ago by bethan.white

Hi Willie,

Thanks for your reply. I asked Carol yesterday but she said it's been a long time since she's had to add mods this way. Her suggestion was to locate another mod that has deleted these lines and work out what those lines are. If I do have conflicting mods, will switching one off matter, or is it possible to edit them so they no longer conflict?

Thanks,
Bethan

comment:3 Changed 9 years ago by willie

Hi Bethan,

I had a quick look at your job. You have included each of idealised modsets twice. This will cause the error you are seeing. If you ensure that these only appear once in the "modifications for the model" table and remove the commenting out, it should work.

Regards,

Willie

comment:4 Changed 9 years ago by bethan.white

Hi Willie,

Thanks for your reply. Carol told me I would need to include both the .mh and the .mf77 files. Do I only need to include one of these in the "modifications for the model" table?

Thanks,
Bethan.

comment:5 Changed 9 years ago by willie

Hi Bethan,

Sorry. You're quite right. I didn't read to the end of the lines.

Willie

comment:6 follow-up: Changed 9 years ago by willie

  • UM Version changed from <select version> to 6.1

Hi Bethan,

The idealised mod idl601_diurnalfluxesopt3.mf77 conflicts with $UMODS_HRTM/base/apc0f602.mf77.

Looking at these, it is not just the delete that you have commented out that is the concern, but the line following it is being added and they are different in the two files. This will lead to compile errors.

You should discuss this with Carol.

Regards,

Willie

comment:7 in reply to: ↑ 6 Changed 9 years ago by bethan.white

Hi Willie, apologies for the delay in response, I've been away recently.

I have looked at idl601_diurnalfluxesopt3.mf77, and realised that I had not commented out the lines with the continuation marks. I have now done so and have tried again to compile the model.

I now get the following error message (xfdta000.xfdta.d10251.t190446.comp.leave):

make[1]: Entering directory `/work/n02/n02/eebaw/xfdta/xfdta/compile_xfdta'
/lib/cpp -P -C -traditional-cpp -I ./include -I /work/n02/n02/hum/vn6.1/pathscale/include chkideal.F chkideal.i
if [ -f chkideal.i ]; then mv chkideal.i chkideal.f; fi
file chkideal.f size 78377
ftn -i8 -r8 -fno-second-underscore -I /work/n02/n02/hum/gcom/pathscale/gcom3.0/include -O3 -c chkideal.f
/opt/cray/xt-asyncpe/3.6/bin/ftn: INFO: linux target is being used

INTEGER,PARAMETER
max_num_force_times = 100

pathf95-1259 pathf90-3.0: WARNING CHECK_IDEALISE, File = chkideal.f, Line = 1186, Column = 27

"MAX_NUM_FORCE_TIMES" (which first appeared at line 1185) has been given the INTEGER(KIND=8) attribute more than once.


pathf95-554 pathf90-3.0: ERROR CHECK_IDEALISE, File = chkideal.f, Line = 1186, Column = 27

"MAX_NUM_FORCE_TIMES" has the PARAMETER attribute. It must not be given the PARAMETER attribute again (identifier first appeared at line 1185)

I have looked at /work/n02/n02/eebaw/xfdta/xfdta/compile_xfdta/chkideal.f, and lines 1185 and 1186 are identical:

INTEGER,PARAMETER
max_num_profile_data = 100
INTEGER,PARAMETER
max_num_profile_data = 100

Am I right in thinking that one of these is being created by a modset, and one by the basic model? I'm afraid that from the .comp.leave file it's not obvious to me where this is occurring. Are you able to let me know how to find this out?

Thanks,
Bethan.

Replying to willie:

Hi Bethan,

The idealised mod idl601_diurnalfluxesopt3.mf77 conflicts with $UMODS_HRTM/base/apc0f602.mf77.

Looking at these, it is not just the delete that you have commented out that is the concern, but the line following it is being added and they are different in the two files. This will lead to compile errors.

You should discuss this with Carol.

Regards,

Willie

comment:8 Changed 9 years ago by willie

Hi Bethan,

Sorry about the delay in replaying. I'm just back from holiday.

You modset idl601_bubble.mh is including the definition of max_num_force_times when it is already defined by the model (it even has the same value). If you remove this line, then it should work.

I normally find these out by grep'ing the parameter name in the modset and comparing with the source code browser.

Regards,

Willie

comment:9 Changed 9 years ago by bethan.white

Hi Willie,
Thanks for getting back to me. Carol and I have worked through most of the errors that I was encountering and I am now able to get the model to compile with two of the mods that I was using.

However, it's now crashing when it runs, with the message "Idealised namelist vertical profile datadoes not extend to the top of the model". I thought I had been careful to design my namelist so that it did extend to the top of the model.

Is it possible that I've missed something simple like a surface value?

The .leave file is /home/n02/n02/eebaw/um/umui_out/xfdta000.xfdta.d10258.t162549.leave and my namelist is /work/n02/n02/eebaw/idealised/IOP3/ideal_swanage1100

Thanks,
Bethan.

comment:10 Changed 9 years ago by willie

Hi Bethan,

It looks like the ideal_swanage1100 says that num_profile_data=97. I think this should be 76.

Regards,

Willie

comment:11 Changed 9 years ago by bethan.white

Hi Willie,

I made the change you suggested; I also realised that the final value in the height data in my ideal namelist did not match the height of the top of the model in the vertlevs_2G3 file, so I've changed my namelist to match.

The model now runs for a few minutes and crashes with the following error:

qsfinal: thist file copied to /work/n02/n02/eebaw/xfdta/xfdta.thist.9206
/work/n02/n02/hum/vn6.1/pathscale/scripts/qsfinal: Error in exit processing after model run
Failed in model executable

I'm afraid I've not seen this error before so I'm not sure what's the first thing to check.

The .leave file is:
/home/n02/n02/eebaw/um/umui_out/xfdta000.xfdta.d10260.t183305.leave

Thanks,
Bethan.

comment:12 Changed 9 years ago by willie

Hi Bethan,

Near the end of the .leave file, after the first time step, we get

 Vertical grid unchanged - as in input dump 
  BUT    WARNING 
  grid_flat must be set in your NAMELIST input
  For quadratic flattening set grid_flat=3
 ** Quadratic flattening of grid over orography 
 ** starting from surface.  grid_flat =  3
 Eta_grid generated problem_number  3
 timestep =  15.  seconds
 levels =  76
 Equatorial East-West gridlength =  1011.9102033817668  metres  1.0119102033817668  kilometres
 Equatorial North-South gridlength =  1011.9102033817668  metres 1.0119102033817668  kilometres
  Height_domain (top theta level) =  39253.272700000001  metres
  first_constant_rho_level set to  1
  big_layers =  0
  transit_layers =  0
  Westerly component =  0.E+0  m/s 
  Southerly component =  0.E+0  m/s 
  Courant number for constant flow =  0.E+0
  Surface temperature =  300.  K
  Surface pressure =  101000.  Pa
  Brunt-Vaisala frequency =  1.00000000000000002E-2  per second
  Setting pressure fields 
 ** Pressure negative at model top   ***
 **   Cannot run    ***

This suggests that there is still a problem with the idealised namelist.

I hope that helps.

Regards,

Willie

comment:13 Changed 8 years ago by bethan.white

Hi Willie,

I double-checked my namelist profile and there was a problem with my potential temperature profile. I corrected this and re-ran the model with the new theta profile. It compiles and crashes with the following message (xfdta000.xfdta.d10273.t173058.leave):

"qsfinal: thist file copied to /work/n02/n02/eebaw/xfdta/xfdta.thist.21106
/work/n02/n02/hum/vn6.1/pathscale/scripts/qsfinal: Error in exit processing after model run
Failed in model executable"

I've looked all the way through the .leave file and I can't see any error messages or output like the 'cannot run' statement that was previously there. I was hoping that having corrected the potential temperature profile, this would have sorted the problem out, but I'm afraid I'm not really sure what to look at next.

Thanks,
Bethan.

comment:14 Changed 8 years ago by willie

Hi Bethan,

I think it wants you to add grid_flat =3 to your namelist. You're still getting negative pressures - see the tail end of /work/n02/n02/eebaw/xfdta/xfdta.fort.pe6 for example.

Regards

Willie

comment:15 Changed 8 years ago by bethan.white

Hi Willie,

I tried adding grid_flat = 3 to my namelist (/work/n02/n02/eebaw/idealised/IOP3/ideal_swanage1100) but I don't think it's helped - xfdta000.xfdta.d10277.t151135.leave still gives the same message as before.

Thanks,
Bethan.

comment:16 Changed 8 years ago by willie

  • Resolution set to fixed
  • Status changed from accepted to closed

The attempt at a 6.1 idealized job has been abandoned. There are now two standard UM7.5 idealized jobs in the UMUI (user 'umui'), xfqca and xfqcb. This ticket is therefore closed. See also ticket #515.

Note: See TracTickets for help on using tickets.