Opened 4 years ago

Closed 4 years ago

#1895 closed help (fixed)

Problems with new flux diagnostics

Reported by: MikeN Owned by: annette
Component: UKCA Keywords: FCM
Cc: Platform: ARCHER
UM Version: 7.3

Description

Hello,

I have added new diagnostics to asad_flux_dat.F90 in the STASH slots 461 - 478 (Section 34).

/MikeN/FCM/vn7.3_oxbuds_diagnostic/src/atmosphere/UKCA/asad_flux_dat.F90

These relate to the alkyl nitrate chemistry I have added to the model.

I have create a new stash file for these: /MikeN/stash/CheST/s34_CheM_STASH_fluxes_v7.3.3_xmsp

And included them in the STASH panel in the UMUI and initialised them.

However, when running the job (xmspc) I get the message:

UM ERROR (Model aborting) :

Routine generating error: ASAD_CHEMICAL_DIAGNOSTICS
Error code: 34461
Error message:

STASH CODE NOT AVAILABLE

The .leave file is at /home/n02/n02/mnewland/output/xmspc000.xmspc.d16166.t145411.leave

Have I done something wrong in adding them or are these slots not available?

I have also added three new dry deposition reactions (for the three alkyl nitrates) to /MikeN/FCM/vn7.3_oxbuds_diagnostic/src/atmosphere/UKCA/ukca_chem1_strattrop.F90. This has increased the number of dry deposition reactions from 36 to 39. However, when I increase JPDD from 36 to 39 in the hand-edit (/MikeN/hand_edits/xmspa/jp_for_CheST_v1.4_C4C5diag.ed), I get the error:

UM ERROR (Model aborting) :
Routine generating error: UKCA_CHEM1_INIT.check_dims
Error code: 96
Error message:
depvel_defs: size of dimn=3 not equal to jpdd

(e.g. /home/n02/n02/mnewland/output/xmspc000.xmspc.d16166.t113508.leave

If I leave JPDD at 36 the model seems happy!

Many thanks,

Mike

Change History (9)

comment:1 Changed 4 years ago by annette

  • Component changed from UM Model to UKCA
  • Keywords FCM added
  • Owner changed from um_support to annette
  • Platform set to ARCHER
  • Status changed from new to assigned

Hi Mike,

Looking at the code that is extracted on PUMA, your changes are not being picked up:

/home/MikeN/um/um_extracts/xmspc/umbase/src/UM/atmosphere/UKCA/asad_flux_dat.F90

FCM might be getting confused because your branch is listed twice:

fcm:um_br/dev/MikeN/vn7.3_oxbuds_diagnostic/src@21135
/home/MikeN/FCM/vn7.3_oxbuds_diagnostic/src

Can you try switching off the repository version (branch number 16 in the list), and see if that works?

Best regards,
Annette

comment:2 Changed 4 years ago by MikeN

Hi Annette,

The model doesn't even compile if I turn off fcm:um_br/dev/MikeN/vn7.3_oxbuds_diagnostic/src@21135

Output here:

xmspc000.xmspc.d16186.t104406.comp.leave

Mike

comment:3 Changed 4 years ago by annette

Mike,

Can you change the permissions of your files on ARCHER so that I can see them please?

chmod -R g+rX /home/n02/n02/mnewland
chmod -R g+rX /work/n02/n02/mnewland

Annette

comment:4 Changed 4 years ago by MikeN

Done!

Thanks,
Mike

comment:5 Changed 4 years ago by annette

Hi Mike,

I have been looking at the revision log for your branch:
http://puma.nerc.ac.uk/trac/UM/log/UM/branches/dev/MikeN/vn7.3_oxbuds_diagnostic

What was happening before (in your original query) was that only the changes from your branch up to revision 21135, plus the uncommited changes in the working copy were being included. This meant that the changes you added in for the alkyl nitrate chemistry at revision 21168 were not.

Normally including a working copy of a branch adds all the commited changes from the branch as well. I assume this is what you want so that's why I asked you to switch off the other branch listing.

The compilation error you are seeing is a problem with revision 21138 (which was also previously not being included in the code):

depvel_defs_strattrop04=(/                  &
                       ^                      
ftn-253 crayftn: ERROR INIT_STRATTROP, File = ../../../../um/xmspc/ummodel/ppsrc
/UM/atmosphere/UKCA/ukca_chem1_strattrop.f90, Line = 1834, Column = 24 
  The left and right hand sides of this array syntax assignment must be conforma
ble arrays.

The error message is because the dimensions of the arrays don't match up. As you are assigning the array with 90 values it needs to be set as such, i.e:

REAL, DIMENSION(  90) :: depvel_defs_strattrop04

I hope this helps.

Annette

comment:6 Changed 4 years ago by MikeN

Hi Annette,

I have changed the dimensions of depvel_defs_strattrop04 in ukca_chem1_strattrop.f90.

I ran the model with the committed branch on at revision 21135 and including modifications from the user copy.

The model builds fine but then fails just as in my original query (xmspc000.xmspc.d16187.t151654.leave).

%PE1 OUTPUT%

D 15, 218, 60, 5760, 2861573, 3, 2*1
D 30, 453, 1, 96, 2867525, 3, 2*1
STASH CODE NOT AVAILABLE Error code: 34461 PE: 1 K: 28
*
UM ERROR (Model aborting) :
Routine generating error: ASAD_CHEMICAL_DIAGNOSTICS
Error code: 34461
Error message:

STASH CODE NOT AVAILABLE

Mike

comment:7 Changed 4 years ago by annette

Hi Mike,

To be clear what I am suggesting you do is:

  • switch the committed branch (fcm:um_br/dev/MikeN/vn7.3_oxbuds_diagnostic/src@21135) OFF
  • leave the working copy branch (/home/MikeN/FCM/vn7.3_oxbuds_diagnostic/src) ON.

When you use a working copy of a branches you should always switch off the commited version - this is confusing FCM and the changes you made to to asad_flux_dat.F90 are not being picked up which is why you are seeing the error above.

I don't know if the code will work or not but at least you will pick up all the changes you have made.

Annette

comment:8 Changed 4 years ago by MikeN

Hi Annette,

After a couple more changes things seem to be working fine.

Thanks for the help!

Mike

comment:9 Changed 4 years ago by annette

  • Resolution set to fixed
  • Status changed from assigned to closed
Note: See TracTickets for help on using tickets.