Main menu

Navigation

aerosol and ozone input data error

17 posts / 0 new
Last post
mburt@...
aerosol and ozone input data error
eaton

ar5_mam3_so2_elev_1850-2005_c090805.nc is not the correct filename.  Looks like it should be ar5_mam3_so2_elev_1850-2005_c090804.nc

 

mburt@...

Thanks.

Looking back at the log file,  it looks like the correct set of files were opened further up in the log file:

   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/oxid/oxid_1.9x2.5_L
   1: 26_1850-2005_c091123.nc      196608
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/aerocom_mam3_d
   1: ms_surf_1849-2006_c090804.nc      262144
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/aerocom_mam3_d
   1: ms_surf_1849-2006_c090804.nc      262144
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so2_s
   1: urf_1850-2005_c090804.nc      327680
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so2_s
   1: urf_1850-2005_c090804.nc      327680
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_soag_
   1: 1.5_surf_1850-2005_c100429.nc      393216
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_soag_
   1: 1.5_surf_1850-2005_c100429.nc      393216
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_bc_su
   1: rf_1850-2005_c090804.nc      458752
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_bc_su
   1: rf_1850-2005_c090804.nc      458752
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 1_surf_1850-2005_c090804.nc      524288
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 1_surf_1850-2005_c090804.nc      524288
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 2_surf_1850-2005_c090804.nc      589824
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 2_surf_1850-2005_c090804.nc      589824
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_oc_su
   1: rf_1850-2005_c090804.nc      655360
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_oc_su
   1: rf_1850-2005_c090804.nc      655360
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 1_surf_1850-2005_c090804.nc      720896
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 1_surf_1850-2005_c090804.nc      720896
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 2_surf_1850-2005_c090804.nc      786432
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 2_surf_1850-2005_c090804.nc      786432
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so2_e
   1: lev_1850-2005_c090804.nc      851968
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so2_e
   1: lev_1850-2005_c090804.nc      851968
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 1_elev_1850-2005_c090804.nc      917504
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 1_elev_1850-2005_c090804.nc      917504
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_oc_el
   1: ev_1850-2005_c090804.nc      983040
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_oc_el
   1: ev_1850-2005_c090804.nc      983040
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_bc_el
   1: ev_1850-2005_c090804.nc     1048576
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_bc_el
   1: ev_1850-2005_c090804.nc     1048576
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 1_elev_1850-2005_c090804.nc     1114112
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 1_elev_1850-2005_c090804.nc     1114112
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 2_elev_1850-2005_c090804.nc     1179648
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a
   1: 2_elev_1850-2005_c090804.nc     1179648
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 2_elev_1850-2005_c090804.nc     1245184
   1: Opened existing file
   1: /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a
   1: 2_elev_1850-2005_c090804.nc     1245184

 

santos

I wonder if this is a problem with the dates in that file not being compatible with a 2005 cyclical run. If you run off the end of a file's date range, one of the things that CAM will try to do is to just increment the number at the end of the file and see if it can find more data in a file with the new name. So ar5_mam3_so2_surf_1850-2005_c090804.nc becomes ar5_mam3_so2_surf_1850-2005_c090805.nc.

mburt@...

I think that is exactly the problem Sean.

 

Does anyone have any suggestions on how to get pass this end of file date issue?

 

Thanks. - Melissa

santos

I'm not sure; this file actually does have a date range that covers up to 2006/12/15. Can you attach the CAM namelist (atm_in in CaseDocs, or from your run directory)?

mburt@...

Hi Sean. Thanks for your reply.


I've attached my atm_in file.

 

Thanks.

Melissa

 

Attachment: 
santos

Ah, I see. It looks like you have not set ext_frc_type and ext_frc_cycle_year. The file that's causing a problem is governed by these, rather than the srf_emis settings. So I'm thinking that you can just set the ext_frc settings in the same way, and hopefully that will work?

santos

I think that there may be other problems here as well.

There's a question here of what kind of run you want to do. You are setting up a sort of climatological run by cycling, but the files and settings are from a historical run. I don't think that cycling over year 2005 will do the right thing (and neither does Brian Eaton), since that year is not in the file. I'm not 100% sure that using year 2006 is right either, but it makes more sense.

More likely, what you really want to do is set up a present-day run (e.g. with the FC5 compset) as a hybrid run based on your previous run, rather than simply "branching" by switching these files to loop cyclically.

mburt@...

Okay, thank you. I will setup a present-day run with a different compset as you suggested. I appreciate all of the help! -Melissa

mburt@...

For a fully-coupled transient CO2 experiment simulation, where I would like to increase CO2 at the rate of 1%/yr until it reaches 4x PI concentration and then hold it constant until the climate reaches an equilbiibrium state (mostly), how would I get beyond this 2005 end of file aerosol issue? I have a spinup simulation using the B1850-2000_CAM5_CN compset, could I branch off of this spinup wiht the B1850_rampCO2? and would this run into the same end of file 2005 aerosol issue?

 

-Melissa

eaton

You plan to start from a spunup 2005 state and continue ramping CO2.  What do want to do with the other gases/aerosols?

 

 

mburt@...

Thanks for responding.

I would like to cycle the other gases and aerosols if that's possible.

eaton

Looks like you need a new compset: B_2005_CAM5_RAMPCO2_CN.  This can be constructed by looking at the pieces of existing compsets and the corresponding cam use cases.  To get most of the gases/aerosols fixed look at the B_1850-2000_CAM5 transient case and set the cycle year to the last year of data in the transient datasets.  To get the CO2 ramping you'll need to set scenario_ghg='RAMP_CO2_ONLY' and then use the co2vmr, ..., f12vmr variables to set values consistent with the spunup conditions you plan to start from.  All this info about the gases/aerosols goes into a new cam use case file that the new compset uses.  I think you can start the run as a hybrid startup to make use of the spunup state from a previous run.  This will take a bit of work.

 

mburt@...

Okay, I will look into putting together a new compset.

In response to the gases/aerosols issue. I’ve tried setting:

ext_frc_type                    ='CYCLICAL'
ext_frc_cycle_yr                =2006

Does this swtich (above) apply to the following:

'SO2       -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so2_surf_1850-2005_c090804.nc',
  'SOAG      -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_soag_1.5_surf_1850-2005_c100429.nc',
  'bc_a1     -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_bc_surf_1850-2005_c090804.nc',
  'num_a1    -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a1_surf_1850-2005_c090804.nc',
  'num_a2    -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_num_a2_surf_1850-2005_c090804.nc',
  'pom_a1    -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_oc_surf_1850-2005_c090804.nc',
  'so4_a1    -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a1_surf_1850-2005_c090804.nc',
  'so4_a2    -> /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_aero/emis/ar5_mam3_so4_a2_surf_1850-2005_c090804.nc'


All of these files have the year 2006 in them, but setting it to cycle on this last year of the data doesn’t appear to work.

here's what i get in the log file:

  60: ENDRUN:
  60: GETFIL: FAILED to get /glade/p/cesm/cseg//inputdata/atm/cam/chem/trop_mozart_ae
  60: ro/emis/ar5_mam3_so2_surf_1850-2005_c090805.nc

however, the atm_in file correctly has ar5_mam3_so2_surf_1850-2005_c090804.nc, so it seems that the model is trying to iterate to the "next" file by changing c090804 to c090805 and of course it can't find that file. 

 

raeder

I've struggled with CYCLICAL too, even with a startup run from the default initial file.
Here's some of what I learned.

To answer the question directly, ext_frc_* do not control the surface variables
SO2, SOAG, bc_a1, ...  Those are controlled by surf_emis_*.
To run a forecast beyond the end of the data in the files,
which are specified in surf_emis_specifier, set
 srf_emis_type         = 'CYCLICAL'
 srf_emis_cycle_yr   = a year which actually appears in the file,
                                    not just within the span of years in the file.
NOTE; CAM expects to find data that is more frequent than yearly.
            Data that is yearly will generate the error:
            " flbc_inti: cyclical lb conds require at least two time points"
            It identifies the specified year as both the first needed timeslot index
            and the last, which it refuses to interpolate.

The SO2 which appears in ext_frc_specifier is a 3D variable, not surface
("elev" in the filename, instead of "surf").  This is controlled by
ext_frc_type and ext_frc_cycle_yr.

 

But there's more.


More generally, there are at least 4 (5 in CESM2) separate external forcing
variable groups:
   srf_emis_*
   ext_frc_*
   tracer_cnst_*
   prescribed_ozone_*
   prescribed_strataero_*
it is necessary to make the *_type values consistent BOTH with
the files specified in the corresponding *_{specifier,datapath} variables
and possibly with other forcing variable groups.
That is because 2 groups may access the same file.
For example, the variable 'O3' is read from the same file by both prescribed_ozone
and prescribed_strataero.  Say the forecast is for 2016 and the user_nl_cam has only

 prescribed_ozone_type        = 'CYCLICAL'
 prescribed_ozone_cycle_yr = 2014
 prescribed_ozone_file         = 'ozone_strataero_WACCM_L70_zm5day_18500101-20150103_CMIP6ensAvg_c180923.nc'
 prescribed_strataero_file    = 'ozone_strataero_WACCM_L70_zm5day_18500101-20150103_CMIP6ensAvg_c180923.nc'

That file has data only through 2014.
The first (_ozone_) initialization of the file access uses mode CYCLICAL,
but the second (_strataero_) changes that to the default SERIAL,
because prescribed_strataero_type was not specified.
Then the run will fail when it tries to read O3 through either
prescribed_ozone or prescribed_strataero because it looks for the forecast year (2016)
on the data file (SERIAL mode), doesn't find it, increments the file name to
...c180924.nc in an attempt to find it on the "next" file, and fails to find that file.
This is very confusing, because the prescribed_ozone_* are correct,
but the program seems to ignore those values and fails
because of variables that ARE NOT IN THE user_nl_cam NAMELIST.

eaton

I'm guessing that you're trying to do a branch run which makes use of restart files and ignores most of what's in the namelist files.  To change the dataset interpolation from a time series to a cyclical mode you'll need to do either an initial or a hybrid run.

 

Log in or register to post comments

Who's new

  • 408251463@...
  • 1161561668@...
  • 645505358@...
  • lbqnj@...
  • q1034908455@...