Opened 5 years ago

Closed 4 years ago

#1417 closed task (answered)

UKCA future climate (RCP8.5) set-up

Reported by: fcentoni Owned by: luke
Component: UKCA Keywords:
Cc: Platform: ARCHER
UM Version: 7.3

Description

Hi,

I am trying to set-up my job (xkjtd) in a future climate configuration (RCP8.5) but it does not run.
I have been using as a reference a job of Tara's Bonerjce (xjnzh) who had already set-up her job with RCP8.5.
The problem is that she is using a different model configuration (CheST) FASTJ photolysis scheme,different chemistry scheme and couplings (O3 in radiation scheme).
So I tried to adapt her set-up to my job keeping only her future climate ancillary files and startdump but it did not work.

The error I got is:

lib-4324 : UNRECOVERABLE library error 
  The variable name 'FASTJX_NUMWL' is unrecognized in namelist input.

lib-4324 : UNRECOVERABLE library error 
  The variable name 'FASTJX_NUMWL' is unrecognized in namelist input.

lib-4324 : UNRECOVERABLE library error 
  The variable name 'FASTJX_NUMWL' is unrecognized in namelist input.

lib-4324 : UNRECOVERABLE library error 
  The variable name 'FASTJX_NUMWL' is unrecognized in namelist input.

Encountered during a namelist READ from
lib-4324 : UNRECOVERABLE library error 
  The variable name 'FASTJX_NUMWL' is unrecognized in namelist input.

Encountered during a namelist READ from
Encountered during a namelist READ from unit 5

Encountered during a namelist READ from unit 5

I do not know whether it may be related to the fact we are running with different codes.
There were also two Luke's branches which I did not know what they are used for.

Many thanks for your help.
Regards,
Federico

Change History (8)

comment:1 Changed 5 years ago by luke

Hi Federico,

The errors above are because you have some FastJX related items in your namelist, but the code isn't there to handle them. However, I can't find the hand-edit which inserts them - have you been changing the xkjtd job since you raised this ticket?

Have you checked with Tara which is the best job of hers to reference against? It's polite and always best to do this in these cases, mainly because the job you are considering taking might be a testing job which might not be the best one to use. Also, if you intend to use someone's ancillary files it highlights that they shouldn't be deleted etc.

I don't think you will need to make any code changes - it should just be UMUI settings and ancillary files.

You will need to take the equivalent surface and 3D emissions - there shouldn't be anything you need to do to these, as the emissions are the same for CheT and CheST. You'll also need SSTs and sea-ice, and any CLASSIC emissions that might have changed that you want to consider. As a first guess, I would think that the ones that may need changing are the ones listed here:

http://www.ukca.ac.uk/wiki/index.php/Using_the_UMUI#Running_a_Perpetual_Year

In CheST, CH4 is a tracer with a surface boundary condition. In CheT, it's a constant number. These can be set using the namelist logical L_ukca_prescribe_trace_gases. See the following code in ukca_main1-ukca_main1.F90:

533	      IF (L_ukca_prescribe_trace_gases) THEN
534	         fco2 = um_co2_for_ukca/c_co2
535	         fh2  = um_h2_for_ukca/c_h2
536	         fn2  = um_n2_for_ukca/c_n2
537	         fo2  = um_o2_for_ukca/c_o2
538	         fch4 = um_ch4_for_ukca/c_ch4
539	      ELSE ! present day values
540	         fco2 = 350.0e-6
541	         fh2  = 5.0e-7
542	         fn2  = 0.78084
543	         fo2  = 0.20945
544	         fch4 = 1.76e-6
545	      END IF

These are set in the hand-edit

~ukca/hand_edits/VN7.3/r1.0/CheM_trgas_presentday_full.ed

You can use the scenario routine on PUMA to get the required values for you:

$ ~ukca/bin/scenario --help
  UKCA SCENARIO GENERATOR 
  ======================= 
  
  USAGE: scenario YYYY/MM/DD <scenario>
  
  WHERE <scenario> IS ONE OF: WMOA1
                              RCP8.5
                              RCP6.0
                              RCP4.5
                              RCP2.6
                              USER
  
  THE WMOA1 SCENARIO IS VALID FOR 1950-2100
  THE RCP SCENARIOS ARE VALID FROM 1765-2500
  
  written by N. Luke Abraham 2011/07/13

You might want to consider the upper-boundary conditions for O3 and NOy. Currently you will be using the UM ozone climatology and the Cambridge 2D model for each respectively. Do you have a future ozone climatology you can use? You could use that defined for the CMIP5 specification, or you could always talk to Tara about getting a climatology from her Base run.

It may be overkill to take tracer initialisations from Tara's job, as you are using CheT and the spin-up time for the chemistry is very short. Spinning up the atmosphere might take a few years though.

Thanks,
Luke

comment:2 Changed 5 years ago by luke

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

comment:3 Changed 5 years ago by fcentoni

Hi Luke,

thank you very much for your reply.

The job xkjtd is now my base job which includes a different dry deposition scheme.
My future climate job is xkjte.

I and David Stevenson had already contacted Tara and she actually pointed us at her job xjnzm (CC only RCP8.5).

About CH4 mass mixing ratio, I have updated it along with the other GHGs in the UMUI settings. Are you saying that I am supposed to that manually through code changes?

I actually am not sure whether I am supposed to change future O3 climatology for the experiment I am planning to do. I will try to better figure it out.

Many thanks.
Federico.

comment:4 Changed 5 years ago by luke

Hi Federico,

You shouldn't need to make any code changes so long as the logical is TRUE, and you have updated the UMUI accordingly.

Thanks,
Luke

comment:5 Changed 5 years ago by fcentoni

Dear Luke,

I am still trying to get my future climate model configuration working.

It seems that there may be something wrong with emission or other ancillary files as you can read in the file xkjtg000.xkjtg.d15012.t164111.leave:

*********************************************************************************
 UM ERROR (Model aborting) :
 Routine generating error: UP_ANCIL
 Error code:  321
 Error message:
REPLANCA: PP HEADERS ON ANCILLARY FILE DO NOT MATCH
 *********************************************************************************
Rank 0 [Mon Jan 12 17:32:48 2015] [c1-3c2s12n3] application called MPI_Abort(MPI_COMM_WORLD, 9) - process 0
Rank 5 [Mon Jan 12 17:32:48 2015] [c1-3c2s12n3] application called MPI_Abort(MPI_COMM_WORLD, 9) - process 5

According to what recently discussed by Tara and David Stevenson, it seems that the model is running out of data 1/1/2001 and they think this regards emission ancillary files. David S. suggested me to remake them independent of year (even though it is not that clear to me what that means).
So basically, I made a copy of the emission file ar5_tropisop_n48_2000.anc with aim to make a new emission file with 2090s date headers.

Is this approach right?
If so, am I supposed to do the same for the other emission ancillary files which the model reads in?

However, when I try to create a new ancillary file (I followed your tutorial#5) I get the following error message:

Error: couldn't open "fs2/n02/n02/hum/bin/xancil.namelist" 

You had previously suggested that it may be related to the fact that Xancil might think that its working directory is /fs2/n02/n02/hum/bin/ rather than one of my directories.
However I managed to properly save my job on my archer home so I do not understand why it is not working.

Your feedback would be much appreciated.

Many thanks.
Best wishes,
Federico.

comment:6 Changed 5 years ago by luke

Hi Federico,

You can set the ancillary to be periodic in time. See

http://www.ukca.ac.uk/wiki/index.php/Using_Xancil#Atmosphere_Ancillary_Files

Can you try running Xancil from a directory on /work rather than /home and let me know what happens?

Thanks,
Luke

comment:7 Changed 5 years ago by fcentoni

Hi Luke,

I do not any problem with the ancillary files anymore but thank you.

The only doubt I have left is about to run in a future climate holding CH4 concentrations at present day (2000).

Am I supposed to hold only CH4 mmr (through UMUI setting) at present day but updating the other gas mmrs (according to the values output by scenario tool for RCP8.5)?

I have also created a new hand_edit for the GHGs (CheM_trgas_rcp8.5_CConly.ed) as output by scenario tool.

Many thanks,
Federico.

comment:8 Changed 4 years ago by annette

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

As this ticket has been inactive for some time, I'm going to close it, but you can re-open or post a new ticket if you have further issues.

Best regards,
Annette

Note: See TracTickets for help on using tickets.