Opened 7 years ago

Closed 7 years ago

#1056 closed help (fixed)

UKCA ancillary emissions file

Reported by: fcentoni Owned by: luke
Component: UM Model Keywords:
Cc: Platform: HECToR
UM Version: 7.3

Description

Dear Ros (or Luke),

I was trying to look at the emissions file

/work/n02/n02/ukca/ANCILS/QESM/ar5_tropisop_n48_2000.anc

which I had already explored using Xconv but now I get 'permission
denied'. I do not understand why.
The last time I managed to do that, I saw there are several emission fields but it is not clear which chemical species they correspond to. Is there any way to get that?

Many thanks. Regards,
Federico.

Change History (9)

comment:1 Changed 7 years ago by luke

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

Hi Federico,

I am able to view and open the above file on HECToR:

[luke@hector-xe6-9 ~]$ ls -l /work/n02/n02/ukca/ANCILS/QESM/ar5_tropisop_n48_2000.anc
-rwxr-xr-x 1 ukca n02 7143424 2011-02-04 14:49 /work/n02/n02/ukca/ANCILS/QESM/ar5_tropisop_n48_2000.anc*
[luke@hector-xe6-9 ~]$ xconv -i /work/n02/n02/ukca/ANCILS/QESM/ar5_tropisop_n48_2000.anc
[luke@hector-xe6-9 ~]$ 

Could you please replicate the above commands and, if you are still having issues opening the file in Xconv, copy and paste the output into the ticket?

The order of the emissions is defined in ukca_setd1defs.F90:

            em_chem_spec =                                             &
                 (/'NO        ','CH4       ','CO        ','HCHO      ',&
                 'C2H6      ','C3H8      ','Me2CO     ','MeCHO     ',  &
                 'C5H8      ','NO_aircrft'/)

Where the NO_aircraft emission is in the user multi-level ancillary file. It should be noted that, for historical reasons, the emissions of NO are actually in units of NO2, and this is converted within ukca_emission_ctl.F90. If you are making up new NO emissions you will need to ensure that these are in units of kg NO2/m2/s.

You can also see the user STASHmaster file

/home/ukca/userprestash/VN7.3/r1.0/s0_CheM_STASH_emissions_v7.3

which defines the emissions used in your job (and the assoicated "Initialisation of user prognostics" UMUI panel), although the em_chem_spec array in ukca_setd1defs.F90 is the master list.

I hope this helps.

Thanks,

Luke

comment:2 Changed 7 years ago by fcentoni

Many thanks Luke,

your informations about the emissions were very useful!

Actually, I was trying to lunch another run (job xiepd) with the aim to look at the daily dry deposition of O3 and NOy. For O3 and NOY I set up the time to TDAYMON, domain to DIAG, usage to UPJ and then I added up LAI diagnostic, I turned on some emissions diagnostics(NOx surface emissions, CH4, CO, N2O, NH3, SO2) .
It turned out the run failed.
May you have a look at the file xiepd000.xiepd.d13113.t110001.leave?

Thanks,
Federico

comment:3 Changed 7 years ago by luke

Hi Federico,

Looking in your .leave file, there is the line

aprun: file /work/n02/n02/fcentoni/um/xiepd/bin/qxreconf not found

which is causing the run to stop. This is because you don't have a /bin directory within your /xiepd directory - this /bin directory contains the model and reconfiguration exectutables, as well as other executables and scripts.

Can you turn on compilation of your xiepd model and reconfiguration executables before sending the job off again?

Also, I would suggest making a new time profile for outputting daily UKCA fluxes. The UKCA chemical solver is only called every hour, and so the flux data is not valid at all timesteps (TDAYMON calculates the mean from all timesteps). I would suggest

  1. Copy the TDAYMON profile:
    • In the STASH panel, highlight the TDAYMON profile
    • Go to Profiles → Copy profile → Copy time
    • Enter a new name for this profile. I would suggest something like TDYMUKCA
  2. Open the TDPMUKCA profile (which already exists)
    • Go to Profiles → Edit profile → Edit time
    • Make a note of the sampling frequency. This should be:
      • Frequency (every) 3 timesteps
      • Sampling offset 2 timesteps
    • Close this window
  3. Open the TDYMUKCA profile you have just created in the same way as above
    • Edit the Frequency and Sampling offset fields to match what is in the TDPMUKCA above (these should both be 1 currently)
    • Close this window
  4. Use this new TDYMUKCA time profile for your output instead of TDAYMON.

I hope this helps.

Thanks,

Luke

comment:4 Changed 7 years ago by fcentoni

Hi Ross (cc Luke),

I ran the model following your instructions as above but it seems the run failed.
Looking at the .leave file xiepd000.xiepd.d13120.t134009.leave, it says:

*
UM Executable : /work/n02/n02/fcentoni/um/xiepd/bin/xiepd.exe
*

_pmiu_daemon(SIGCHLD): [NID 00924] [c8-1c0s1n0] [Tue Apr 30 13:16:30 2013] PE RANK 126 exit signal Segmentation fault
[NID 00924] 2013-04-30 13:16:30 Apid 4476442: initiated application termination
diff: /work/n02/n02/fcentoni/tmp/tmp.hector-xe6-13.1410/xiepd.xhist: No such file or directory
qsexecute: Copying /work/n02/n02/fcentoni/um/xiepd/xiepd.thist to backup thist file /work/n02/n02/fcentoni/um/xiepd/xiepd.thist_keep
xiepd: Run failed

Can you help me out to fix it?

Many thanks,
Federico.

comment:5 Changed 7 years ago by luke

Hi Federico,

I think that the error is associated with the extra STASH diagnostics that you have requested (since these are the only differences from my original job). I'm going through options and let you know what I find.

Thanks,

Luke

comment:6 Changed 7 years ago by luke

Hi Federico,

I think that there are a few issues with your STASH settings:

The block of diagnostics

34 180 NOX SURFACE EMISSION (AS NO2)       TDMPMN   DIAG     UPMEAN   
34 181 CH4 SURFACE EMISSION                TDMPMN   DIAG     UPMEAN   
34 182 CO SURFACE EMISSION                 TDMPMN   DIAG     UPMEAN   
34 190 NOX AIRCRAFT EMISSION AS NO2        TDMPMN   DALLTH   UPMEAN   
34 191 N2O SURFACE EMISSION                TDMPMN   DIAG     UPMEAN   
34 202 SO2 SURFACE EMISSION                TDMPMN   DIAG     UPMEAN   
34 207 NH3 SURFACE EMISSION                TDMPMN   DIAG     UPMEAN   

should be removed since these diagnostics don't actually work (even though they are available in the STASH panel. You will need to expand the src/atmosphere/UKCA/asad_flux_dat.F90 file to create these, but you need to put them in a STASH number greater than 34,301 (I would just put them in at the end of the existing diagnostics in that routine). I'm sorry about the confusion here.

The diagnostic

19   7 LEAF AREA INDEX PFTS AFTER PHENOLOG TDYMUKCA DPFTS    UPJ

is causing the model to crash - turning this off (and the above 7 emissions diagnostics) allow the code to run. I'm not certain exactly why here - it may be because you are requesting it on incorrect timesteps (the Help states "Available with both A19_1A and A19_2A. Only available on phenology timesteps."). The Help text does state that it is the same as 0,217, which does not seem to be available - perhaps this diagnostic cannot be outputted at all in this model configuration. Also, it needed to have the DPFTS domain, rather than DIAG (although this is a moot point).

The diagnostics

34 321 Ox BUDGET: O3 DRY DEPOSITION (3D)   TDYMUKCA DIAG     UPJ      
34 322 Ox BUDGET: NOy DRY DEPOSITION (3D)  TDYMUKCA DIAG     UPJ      
34 331 Ox BUDGET: NOy WET DEPOSITION (3D)  TDYMUKCA DIAG     UPJ      

Should actually be outputted on DALLTH, since with the interactive dry deposition scheme deposition can occur throughout the boundary layer. Also, the same is true for the diagnostic

34 353 TROPOSPHERIC O3                     TDYMUKCA DIAG     UPJ

A very useful function within the STASH panel is the

Diagnostics
  -> Verify diagnostics (control-v)

This will bring up a panel listing possible errors. It should be noted here than with the UKCA release configurations this will appear to give errors for several of the UKCA tracers (within the range 34,001 - 34,150). This isn't actually going to cause problems, it is just that the tracers are turned on at the PROCESS stage of job submission, which the STASH panel doesn't know about yet.

I hope this helps. Let me know if you need anything else.

Thanks,

Luke

comment:7 Changed 7 years ago by fcentoni

Hi Ros (cc Luke),

thank your detailed answer.
I was wondering, is it possible to get the LAI diagnostics working anyway?
This because I would need to investigate the sensitivity of O3 deposition on LAI considering that it currently is the only vegetation parameter included in the non-stomatal resistance parameterization.
What do A19_1A and A19_2A stand for?And what about phenology timesteps?

Secondly, I would like to figure out how to break down the dry deposition loss rate in order to quantify what's the contribute of each resistance (Ra,Rb and Rc).
Particularly, I would need to partitionate stomatal and non-stomatal O3 deposition and to understand how much O3 ends deposited on the different 9 type of surfaces.
My doubt is about to get the model comes out with these output fields. Do I have to work on the code or there is another way to print out the informations I need (the most part of which are contained in the subroutine UKCA_SURF_DD_RS)?

Many thanks. Regards,
Federico.

comment:8 Changed 7 years ago by luke

Hi Federico,

I'm not sure how to turn on the LAI diagnostic correctly. I suggest opening a separate ticket to ask that question since this ticket is UKCA specific (and the original question has been solved).

A19_1A and A19_2A are used to tell the code which routines to compile. These will be set depending which options have been turned on. A19 refers to atmosphere section 19, and 1A and 2A are two different routines that essentially do the same thing, and the user can switch between them. If scheme 1A is set then 2A will not be and vice-versa. For instance UKCA is A34_1A.

You can adapt or expand the code within asad_chem_flux_diags.F90 to allow you to output the extra diagnostics you require, add the request of this into asad_flux_dat.F90 and make a new STASHmaster file to tell the UM that you want these diagnostics to be turned on. You would then call this diagnostic from UKCA_SURF_DD_RS or any other routine you need output from, and the structures contained within asad_chem_flux_diags.F90 will send the diagnostic out through STASH.

Since your originally query has been fixed, I will close this ticket. Please open a new ticket to cover the other questions that you have mentioned above, if required. This will allow someone who is more experienced with the land-surface scheme to deal with your first query, while I will be able to help you with your second if you need it.

Many thanks,

Luke

comment:9 Changed 7 years ago by luke

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