Main menu

Navigation

Mismatch of input dimension

11 posts / 0 new
Last post
rosh.me91@...
Mismatch of input dimension

Hi,

Im running CESM1.2.0 with CLM 4.5 -bgc CN with crop ON. Im running my model at resolution is 1.9x2.5

I tried using the initial data file finidat = ' clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c140111.nc' and I get the following error:

check_dim ERROR: mismatch of input dimension         26947
  with expected value         14891  for variable landunit

I checked the finitdata file which shows the  dimension for landunit 26947.

Why is the initial data file landunits different when the file is for resolution 1.9x2.5?

Do I need to use interpinic ?

Do I have to intitalize my model with coldstart ON?

 

Thank in advance

Roshni Mathur

 

 

sacks

Hi Roshni,

At a glance, it looks like you're pointing to a finidat file that is not the same as the one you get out-of-the-box. In general, you cannot expect this to work (uness you use interpinic). Even for the same configuration and resolution, there can be changes in CLM's internal memory allocation from one version to another that leads to incompatibilities like this. Can you describe why you're using this file rather than the one that comes out-of-the-box for the configuration (if there is one)?

Bill Sacks

CESM Software Engineer

rosh.me91@...

Hi Bill,

Im using the particular finit file because in my run, the crop is ON and methan is OFF. As all the original files do not have CROP On and they ran at 0.9x1.25 resolution, I used this finit file which fits my run compset. I was not able to get a original finit file for my configurations.

For my simulations, I have created my own user compset : 2000_CAM5_CLM45%CN-CROP_CICE_DOCN%SOM. To initialize the CLM model, I using the finit file. 

Shall I interpinic this file according to my configuration or there is any other way ?

Also, how are the landunits, columns and pft dimension calculated in the finit file?

Eg, this finit file shows:

    dimensions:
        gridcell = 5663 ;
        landunit = 14891 ;
        column = 64833 ;
        pft = 124669 ;
        levgrnd = 15 ;
        levurb = 5 ;
        levlak = 10 ;
        levsno = 5 ;
        levsno1 = 6 ;
        levtot = 20 ;
        numrad = 2 ;
        levcan = 1 ;
        string_length = 64 ;
        ntapes = 1 ;
        max_chars = 128 ;
How are each of these calculated? As per my understanding Landunits = no. of gridcells x no. of landunits (5= glacier, lake, wetland,,vegetated,urban). But it doesnt seem to be the case. Similarly how are pft dimention calculated. Im not able to understand the basis of these dimensions.

Please let me know for some clarity!

Thank you

sacks

In a similar CLM version to what you're using, but maybe not exactly the same, I see the following in namelist_defaults_clm4.5.xml:

<finidat hgrid="1.9x2.5"    maxpft="25"  mask="gx1v6" bgc="cn"   ic_ymd="00010101" more_vertlayers=".false." vsoilc_centbgc="on"

ic_tod="0" sim_year="2000" glc_nec="0" crop="on"  >lnd/clm2/initdata_map/clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c130515.nc

</finidat>

 

<finidat hgrid="1.9x2.5"    maxpft="25"  mask="gx1v6" bgc="cn"   ic_ymd="00010101" more_vertlayers=".false." vsoilc_centbgc="no-cent"

ic_tod="0" sim_year="2000" glc_nec="0" crop="on"  >lnd/clm2/initdata_map/clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c130515.nc

</finidat>

 

I would have thought that one of those would be picked up by default in your case. As a general rule, though, if there's no finidat file picked up by default, then you need to either (1) run interpinic using an finidat file that is from a configuration close to your configuration, then run a bit longer to get things spun up, or (2) do your own spinup from cold start. It's hard to explain how the number of landunits, columns, etc. are calculated: this is a very complex operation. See subgridMod.F90 for details.

Bill Sacks

CESM Software Engineer

rosh.me91@...

Hi Bill,

I followed your suggestion and I started my own spinup initial condition file from coldstart for about 121 years. I used the restart file from the output as trial run for user compset I created. Now Im getting the following error:

ERROR:: PFT weights are SIGNIFICANTLY different from the input finidat

  file and fsurdat file(s).

 ERROR:: maximum difference is    0.2379239479654077       max allowed =

   5.0000000000000001E-004

ERROR:: Run interpinic on your initial condition file to interpolate to the new surface dataset


My fsurdat file is: surfdata_1.9x2.5_mp24_simyr2000_c130419.nc

My fsurdat file was picked by default according to my case configurations for the year 2000. 

The colstart case also have the same furdat file. 


following is the lnd_in details for case run from coldstart:

&clm_inparm

 albice = 0.60,0.40

 co2_ppmv = 367.0

 co2_type = 'constant'

 create_crop_landunit = .true.

 dtime = 1800

 fatmlndfrc = '/home/cas/faculty/dilipganguly/cesm/inputdata//share/domains/domain.lnd.fv1.9x2.5_gx1v6.090206.nc'

 finidat = ' '

 fpftcon = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/pftdata/pft-physiology.c130503.nc'

 fsnowaging = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/snicardata/snicar_drdt_bst_fit_60_c070416.nc'

 fsnowoptics = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/snicardata/snicar_optics_5bnd_c090915.nc'

 fsurdat = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/surfdata_map/surfdata_1.9x2.5_mp24_simyr2000_c130419.nc'

 hist_nhtfrq = -8760

 irrigate = .false.


Following is details for case with initial condition provided from the above run:

&clm_inparm

 albice = 0.60,0.40

 co2_ppmv = 368.875

 co2_type = 'constant'

 create_crop_landunit = .true.

 dtime = 1800

 fatmlndfrc = '/home/cas/faculty/dilipganguly/cesm/inputdata//share/domains/domain.lnd.fv1.9x2.5_gx1v6.090206.nc'

 finidat = '/scratch/cas/phd/asz122525/CN_spinup_2000/run/CN_spinup_2000.clm2.r.0121-01-01-00000.nc'

 fpftcon = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/pftdata/pft-physiology.c130503.nc'

 fsnowaging = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/snicardata/snicar_drdt_bst_fit_60_c070416.nc'

 fsnowoptics = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/snicardata/snicar_optics_5bnd_c090915.nc'

 fsurdat = '/home/cas/faculty/dilipganguly/cesm/inputdata//lnd/clm2/surfdata_map/surfdata_1.9x2.5_mp24_simyr2000_c130419.nc'

 hist_empty_htapes = .true.

 irrigate = .true.



Why is the model still asking for interpinic when I have created the initial condition file from coldstart with the same fsurdat file?


Thanks!


Roshni Mathur

sacks

Sorry, I'm not sure why you're running into this problem now. I'll see if anyone else is able to help.

Bill Sacks

CESM Software Engineer

oleson

Based on the namelists you provided, it looks like you have irrigation on for your cold start case and off for your finidat run.  Maybe that's why the weights don't match up.

erik

In addition to what Keith Oleson suggests, my main suggestion is to carefully compare your two cases (the spinup case and the new case) and see what's different between the two. Compare the env_*.xml files as well as all the namelists in the CaseDocs directory. Keith is probably right in that changing the new case to irrigated will change the weights for both the rainfed and irrigated crops, as the non-irrigated case will set the weights to zero and add their weight to the rainfed crops. My main point though is that it's important to carefully compare the cases when you have something unexpected like this, it's easy to miss one namelist setting.

Erik Kluzek ...............

CESM Land Model (CLM) Software Liason

CESM Software Engineering Group, NCAR

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

rosh.me91@...

Hi,

Thank you for your suggestions.

I have 2 questions before I start a new run:

1) Will interpinic work on the current initial condition file that I have created or should I start a new trial spin-up run with irrigation ON or interpinic work only for interpolating gridcells?

2) I checked my *.xml amd namelists in CaseDocs and everything is the same except for the following:

 Spin up Case                                                         Case where initial conditiomn is provided

 

co2_ppmv = 367.0                         co2_ppmv = 368.875

urban_hac = 'ON'                         urban_hac = 'ON_WASTEHEAT'

stream_year_first_popdens = 1850         stream_year_first_popdens = 2000                                                           

Do I need to make all these changes in the spin-up case ? 

rosh.me91@...

Hi,

I tried using interpinic to solve the difference in weights.

Just to try running interpinic, the file I used to interpinic had different dimension. The output says that interpnic has run succesfully but when I read the file, the dimensionas did not change. According to the documentation, the output file is overwritten with interpolated values. Am I looking at an older documention? Where does my new file get saved ? 

Following are the dimensions of my input file:

netcdf CESMsom_Irrig_CROPcyc_ATMcyc_SSTprog_DVoff.clm2.r.0103-01-01-00000 {

dimensions:

        gridcell = 5663 ;

        landunit = 14891 ;

        column = 64730 ;

        pft = 124566 ;

        levgrnd = 15 ;

        levurb = 5 ;

        levlak = 10 ;

        levsno = 5 ;

        levsno1 = 6 ;

        levtot = 20 ;

        numrad = 2 ;

        levcan = 1 ;

        string_length = 64 ;

        ntapes = 4 ;

 

        max_chars = 128 ;


Folowwing are the dimensions of my output file:


netcdf clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c130606 {

dimensions:

        gridcell = 5663 ;

        landunit = 14891 ;

        column = 64833 ;

        pft = 124669 ;

        levgrnd = 15 ;

        levurb = 5 ;

        levlak = 10 ;

        levsno = 5 ;

        levsno1 = 6 ;

        levtot = 20 ;

        numrad = 2 ;

        levcan = 1 ;

        string_length = 64 ;

        ntapes = 1 ;

 

        max_chars = 128 ;

Im also attaching the a text file showing how interpinic worked.

Please let me know if Im doing the right thing or there is something missing?

Thanks in advance!

rosh.me91@...

Sorry, Im attaching the interpinic file here

 

Attachment: 
Log in or register to post comments

Who's new

  • geyer@...
  • jzhan166@...
  • carlos.enciso.o@...
  • pramithamaniyat...
  • lipx@...