Input .nc file into CLM

[Danny Leung]

Hi--
I’m trying to put a .nc file into CLM for it to read and calculate physics in the dust module DUSTMod.F90. I changed namelist_definition.xml and user_nl_clm for reading the namelist_variable.
For namelist_definition_clm4_0.xml:
<!-- soil diameter -dmleung, 22 Apr 2020 -->
<entry id="soil_diam" type="char*256" input_pathname="abs" category="datasets"
group="clm_inparm" valid_values="" >
Full pathname of time-invariant soil diameter dataset for calculating dust saltation threshold.
</entry>

For user_nl_clm:
&clm_inparm
soil_diam='/glade/u/home/dleung/CESM/edit_namelist/soil_diameter-19x25-04222020.nc'
/

I guess I’m doing right above? I am then less certain about how to use PIO library to read the .nc file in DUSTMod.F90 like this (after doing use ncdio_pio, and defining all the needed parameters):
!-----------------------------------------------------------------------
! read in soil diameter dataset (and interpolate to CLM native grid?)
! by dmleung, 29 Apr 2020
call getfil(soil_diam, soil_diam_file, 0)
call ncd_pio_openfile (ncid, trim(soil_diam_file), PIO_NOWRITE)
! Get input data resolution.
ierr = pio_inq_dimid( ncid, 'lon', did )
ierr = pio_inq_dimlen( ncid, did, nlon )
ierr = pio_inq_dimid( ncid, 'lat', did )
ierr = pio_inq_dimlen( ncid, did, nlat )
allocate(dst_lons(nlon))
allocate(dst_lats(nlat))
allocate(soil_diameter_in(nlon,nlat))
ierr = pio_inq_varid( ncid, 'lon', vid )
ierr = pio_get_var( ncid, vid, dst_lons )
ierr = pio_inq_varid( ncid, 'lat', vid )
ierr = pio_get_var( ncid, vid, dst_lats )
ierr = pio_inq_varid( ncid, 'mbl_bsn_fct_geo', vid )
ierr = pio_get_var( ncid, vid, soil_diameter_in )
Am I doing right above? Moreover, I wonder if my dataset is a 2D (long by lat) static map, do I need to regrid the data to CLM native grid? I saw that CAM has interpolation functions like get_rlat_all_p, lininterp, etc. Are there any similar functions in CLM and what functions do I need to read and regrid the dataset?
I attached the detailed scripts, if needed. Any help is highly appreciated.
Many thanks,
Danny

 

[Danny Leung]

Also, when I try to compile the model it gives out the error message (as follows, log file is attached). Why would this error occur even after I specified the path variable in namelist_definition_clm4_0.xml?

/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(55): error #6562: A data initialization-expr is not valid for this object. [SOIL_DIAM]
character(cl) :: soil_diam = 'soil_diam' ! full pathname for soil diameter dataset
-----------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(735): error #6451: A dummy argument name is required in this context. [LBP]
integer, intent(in) :: lbp, ubp ! pft bounds, -dml 23 Apr 2020
---------------------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(735): error #6451: A dummy argument name is required in this context. [UBP]
integer, intent(in) :: lbp, ubp ! pft bounds, -dml 23 Apr 2020
--------------------------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(821): error #6451: A dummy argument name is required in this context. [SOIL_DIAM]
character(cl), intent(in) :: soil_diam
---------------------------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(55): error #6404: This name does not have a type, and must have an explicit type. [CL]
character(cl) :: soil_diam = 'soil_diam' ! full pathname for soil diameter dataset
----------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(55): error #6841: An automatic object must not appear in the specification part of a module. [SOIL_DIAM]
character(cl) :: soil_diam = 'soil_diam' ! full pathname for soil diameter dataset
-----------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(868): error #6404: This name does not have a type, and must have an explicit type. [SOIL_DIAMETER]
allocate(soil_diameter(nlon,nlat))
-------------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(877): error #6284: There is no matching specific function for this generic function reference. [PIO_GET_VAR]
ierr = pio_get_var( ncid, vid, soil_diameter )
-----------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(881): error #6404: This name does not have a type, and must have an explicit type. [P]
do p = lbp,ubp
-------^
/glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90(881): error #6063: An INTEGER or REAL data type is required in this context. [P]
do p = lbp,ubp
-------^
compilation aborted for /glade/scratch/dleung/CESM/diam_case0.3/SourceMods/src.clm/DUSTMod.F90 (code 1)
/glade/scratch/dleung/CESM/diam_case0.3/Tools/Makefile:643: recipe for target 'DUSTMod.o' failed
gmake: *** [DUSTMod.o] Error 1
gmake: *** Waiting for unfinished jobs....

 

[oleson]

I think I would use the "streams" capability in CLM. The streams method can handle interpolation in space and time. For CLM4, you could follow the nitrogen deposition example, src_clm40/main/ndepStreamMod.F90.

 

[Danny Leung]

OK I will look into it, thanks for the suggestion!