UM 6.1 idealised job: Error duplicate Fortran files found

[bethan.white]

Hi,

I have a UM 6.1 job in idealised mode that I am trying to run. Carol has given me a set of Bob Plant's mods to use in setting up this run. I set the job up and edited the mods to eliminate the multiple deletion / insertion errors that I was getting. I have tried to run the job, but it fails to compile, with the message:

qsmncompile(59): * Duplicate Fortran files found
qsmncompile(59): * /work/n02/n02/eebaw/xfdtb/xfdtb/compile_xfdtb/s_main.F90
qsmncompile(59): * /work/n02/n02/eebaw/xfdtb/xfdtb/compile_xfdtb/s_main.F
qsmncompile(59): * Error duplicate Fortran files found, exiting …

My mods are in the following directory:
/work/n02/n02/eebaw/modsets/idealised/common

There appear to be a .F and a .F90 version of s_main in /work/n02/n02/eebaw/xfdtb/xfdtb/compile_xfdtb, but it looks as if they were created at the same time, so I don't think one is left from a past partial build.

The .comp.leave file is:
xfdtb000.xfdtb.d10278.t152344.comp.leave

Thanks,
Bethan.

[willie]

Hi Bethan,

In your idealised modset directory, under the common folder, you have a modset called arw3n602.mf90. This make edits to the deck S_MAIN.dk, which is an mf77 deck. You should rename this modset arw3n602.mf77. This gets us past the duplicate Fortran files problem. However, there are some compile errors, for example in idl_init_data2a.f.

I hope that helps.

Regards,

Willie

[willie]

Hi again Bethan,

In the modsets common/adq2f602.mf77 and idl601_bubble3.mf77, the comparisons "≤" should be replaced by "⇐". This should get past these errors too.

Regards,

Willie

[bethan.white]

Hi Willie,

Thanks for your help with this. The model will now compile but fails when I try to run it; I am getting the message "The variable name 'IDLSURFFLUXSEAOPTION' is unrecognized in namelist input" in xfdtb000.xfdtb.d10280.t143908.leave. In my namelist (/work/n02/n02/eebaw/idealised/IOP3/ideal_swanage1100) I have IdlSurfFluxSeaOption? = 1; this seems to be compatible with /work/n02/n02/eebaw/modsets/idealised/common/afraf602.mf77.

Thanks,
Bethan.

[willie]

Hi Bethan,

You need to add the afraf602 mod to your list of modsets for the model.

Regards,

Willie

[bethan.white]

Hi Willie,

Previously, Bob Plant and Lois had suggested that I leave out arw3f602.mf77 and afraf602.mf77 since they are for the SCM. I have now added both mods (since afraf602 is dependent on arw3f602.mf77). I now have the following errors in xfdtb000.xfdtb.d10281.t123047.comp.leave:

&, L_spec_z0, z0m_scm, z0h_scm

pathf95-113 pathf90-3.0: ERROR NI_BL_CTL, File = bdl_ctl2.f, Line = 125, Column = 20

IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for data object "Z0M_SCM".

pathf95-113 pathf90-3.0: ERROR NI_BL_CTL, File = bdl_ctl2.f, Line = 125, Column = 29

IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for data object "Z0H_SCM".

I have looked through /work/n02/n02/eebaw/modsets/idealised/common/arw3f602.mf77 and it says "If z0h_scm, z0m_scm are absent or < 0.0, sea-surface roughness lengths behave as normal, using z0mseai from the SCM namelists as the initial sea surface roughness."

The problem I have here is that since I am not running the SCM, I don't have an SCM namelist, therefore I have no values for z0h_scm and z0m_scm.

I assume I will have to hardwire these in somewhere in order to get this working. Do you know where the best place would be to do this?

Thanks,
Bethan.

[willie]

Hi Bethan,

Oh dear! You have two different modsets with the same name. You have used the one in the 'common' subdirectory. I think you need the one in the directory bobplant. It has the z0m_scm variables commented out. Someone has been here before!

I hope that helps.

Regards,

Willie

[bethan.white]

Hi Willie,

Thanks. I received two sets of these mods, one from Carol and one from Bob. I had to move the second set to the 'common' directory as the length of the modset names in the 'bobplant' subdirectory had too many characters. I may have unknowingly written over an old file in doing so. I'll replace the one in 'common' with the one in 'bobplant' and see what happens.

I'll let you know how it goes.
Thanks,
Bethan.

[bethan.white]

Hi Willie,

Oddly enough, the version of arw3f602.mf77 in /work/n02/n02/eebaw/modsets/idealised/common/ had the z0m_scm variables commented out, while the version of arw3f602.mf77 in /work/n02/n02/eebaw/modsets/idealised/bobplant/mods/common/ didn't. I was already using the version with the variables commented out.

I'm now trying to run it without those lines commented out.

Thanks,
Bethan.

[bethan.white]

Hi Willie,

It now seems to compile with the mod that doesn't have the z0m_scm variables commented out.

I am now getting the following error in xfdtb000.xfdtb.d10281.t135813.leave:

aprun: -S must be a positive nonzero integer
aprun: Exiting due to errors. Application aborted
/work/n02/n02/eebaw/tmp/tmp.nid00004.25892/modscr_xfdtb/qsexecute: Error in dump reconfiguration - see OUTPUT

I wasn't sure whether to submit this as a new ticket, since I have got exactly the same error with a totally different job, xfhid which I have also tried to run today (xfhid000.xfhid.d10281.t133133.leave).

Let me know if I should bring this up as a new issue.

Thanks,
Bethan.

[willie]

Hi Bethan,

We're preparing for switch to the XT6 and have made some changes to the UMUI. You need to exit the UMUI and relaunch it. An email will appear later.

Regards,

Willie

[bethan.white]

Hi Willie,

This seems to be getting close to working now. I think it's run without crashing now. There are .pp files in /work/n02/n02/eebaw/xfdtb, but when I try to open them with xconv, no fields get loaded.

I might be missing something obvious here. I haven't included a warm bubble yet, but I would have expected the fields to contain data but be unchanging.

Thanks,
Bethan.

[willie]

Hi Bethan,

You job has run out of time. Try increasing from 3600 sec to 7200 sec. It hasn't executed any time steps though. I don't know if the idealised model is supposed to do this.

regards,

Willie

[willie]

Hi Bethan,

I have discovered an error in one of the modsets, afr9f602.mf77. There is a do loop,

  do i=0, idl_max_num_bubbles
:

but the arrays involved start at 1. I have changed this to do i=1, etc (but it could be that the arrays need to be dimensionsed to start at 0?). This takes me to a subscript out of range error during the call to SETCONA: one or more of the indexes to the D1 array is wrong. But it appears in several places and I have been unable to find which one. So it looks like you need to check the modsets carefully - the do loop error appears in several, but I don't think they're being used.

I hope that helps.

Regards,

Willie

[willie]

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 #475.