Documentation of the Changes in INT2LM
Version 1.22

Besides a consolidation of the grib_api implementation and writing COSMO GRIB2 variables with the new general vertical coordinate, this version of INT2LM contains a modification for the procedure to calculate the COSMO multi-layer soil variables from the IFS and interfaces to couple the COSMO-ART (and POLLEN) part for the interpolation.

Contents:

1. Implementation of GRIB2 (with grib_api)
2. Multi-layer treatment of IFS soil
3. Implementation of ART-part with conditional compilation
4. Technical Changes and Bug Fixes
5. Changes in the Namelists
6. Changes of the Results

1. Implementation of GRIB2 (with grib_api)

(by Uli Schättler)

This version of INT2LM now is able to write GRIB2 data with the grib_api from ECMWF. In addition to Version 1.21, the multi-level atmospheric variables are now written using the new general vertical coordinate (with level type 150).

Note that at least version 1.11.0 from grib_api is necessary to compile INT2LM with grib_api, because of the use of the new general vertical coordinate.

The usage of GRIB2, the new general vertical coordinate (and most important the local use sections of GRIB2) will be documented on the COSMO-Web Page: grib This page is still under construction. We hope to extend it until end of July 2013 significantly.

Implemented new GRIB2 leveltype 'generalVertical' and 'generalVerticalLayer'

The main difference to the other vertical coordinate types 'hybrid' and 'hybridLayer' is, that now there is no computational algorithm how the vertical layers are constructed, but a full 3D file (called a HHL-file for the height of half levels in COSMO) is provided, that has to be processed together with other 3D variables defined on these levels.

With this new vertical level type, also some new keys were introduced for the WMO GRIB2 standard, namely:

• nlev (alias for numberOfVerticalCoordinateValues)
  nlev is ke+1 in COSMO, because the number of half levels are specified here.
• numberOfVGridUsed
  represents the vertical coordinate type ivctype.
• uuidOfVGrid
  is a unique identifier for a special HHL-file.

HHL-files can only be created by the INT2LM, but INT2LM can also process an existing HHL-file. This behaviour can be controlled by the new Namelist variable lnewVGrid (in /LMGRID/): the default is .FALSE., which means that no new file is created

These keys replace the vertical coordinate parameters, which are present in data with level types 'hybrid' or 'hybridLayer'.

The new general vertical coordinate and its usage in the COSMO-Model system will be documented and explained in detail on the COSMO web page (coming soon).

Changes to implement the new vertical coordinate

1. Introduced a new module vgrid_refatm_utils.f90, which contains data structures, variables and routines necessary for dealing with the vertical grid and the reference atmosphere.

   New data structures:

   • refatm_type: which contains variables for the reference atmosphere
   • vcoord_type: which contains variables for the vertical grid
   This module now also contains data and routines, which have been in other modules before:
   • routines to compute the different reference atmospheres (from meteo_utilities)
   • data for the vertical coordinate parameters (from data_grid_lm and data_grid_in, resp.)
   The module vgrid_refatm_utils also contains subroutines to create a UUID (unique universal identifiert) and to compare different UUIDs.

   NOTE:
   In connection with these technical changes we noticed a bug in src_vert_inter_lm: To compute the boundary layer height, the vertical coordinate parameters akh_in, bkh_in from the input data were used. But in case of COSMO_EU for example, which has ivctype=2 (a height based coordinate), akh_in and bkh_in contain height, not pressure. Now, always the pressure coordinates vcoord_in%sigm_coord are used.

   CHANGES OF RESULTS:
   This bug fix changes the results of applications with llm2lm and COSMO input data with a height based vertical coordinate (ivctype_in = 2)!!!

2. Implemented possibility to write HHL in a single file with full precision (but only when writing GRIB2).

   For that, 2 new Namelist switches have been implemented in group /DATA/:

   • ylm_hhl: file name for the HHL-file for the COSMO grid (this file is written in the directory ylm_cat, where the initial and boundary fields for the COSMO-Model are written)
   • yin_hhl: file name for the input HHL-file (when processing data from the COSMO-Model). This file is taken from the directory yin_cat as the other COSMO data.

   In case of writing GRIB2, the subroutine org_lm_output is called during the initialization of INT2LM with a variable list just containing 'HHL' to write the COSMO HHL file.

3. Implemented a new module src_read_hhl.f90, which reads the HHL-files for the COSMO-grid and also for a COSMO-input grid, if necessary.

Other adaptations to GRIB2

• in src_lm_output.f90:
  endStep must not be set for ipds(19)=0 (stepType=instant), because it also resets the startStep to 0
• Bug Fix:
  The upper soil level was not set for T_SO and 'api2', because it was only checked for 'grb1' and 'api1'

Some more organizational data structures:

Introduced arrays to convert some GRIB1 codings to grib_api strings

• GRIB1 level types: ylevltypes_in, ylevltypes_out
• GRIB1 time range indicators: ysteptypes
• GRIB2 scaleFactors / scaledValues: scalefactors

Back to Contents

2. Multi-layer treatment of IFS soil

(by Davide Cesari)

With the current interpolation scheme for the IFS soil model, if you use l_multi_layer_lm=.TRUE. and yinput_model='IFS', the soil is first interpolated to the 2-layer soil model levels, and successively interpolated (actually extrapolated) to the multi-layer levels, so input temperature below the COSMO T_CL level gets lost and deep temperature is unrealistic. With this direct interpolation all the input layers influence the output, so higher precision and less extrapolation is obtained.

The new interpolation is activated when

• yinput_model='IFS' and
• l_multi_layer_in=.TRUE. (and of course l_multi_layer_lm=.TRUE.)

The technique used is linear interpolation of soil temperature, assuming it is valid in the mid point of the layers, and a linear conservative interpolation of volumetric soil moisture. Extrapolation below is done keeping values constant.

To avoid too many if's and branches in src_2d_fields.f90, l_smi=.TRUE. is forced when IFS and l_multi_layer_in=.TRUE., because l_smi code uses a different scaling of soil moisture which allows vertical interpolation, while the default setting does not, for this reason, although src_read_coarse_grid.f90 would allow interpolation with l_smi=.FALSE., this is forbidden in src_namelists.f90 because the results are wrong.

Also the l_use_t_skin=.FALSE. option is not implemented. It could be supported also for l_smi=.FALSE. and l_use_t_skin=.FALSE. at the cost of complicating the code a little bit.

The modifications are:

• src_gribtabs.f90:
  Do not try to add W_SO and T_SO to input list for IFS even if multi-layer
• src_memory.f90:
  Extend the bounds of dt_so_gl
• src_namelists.f90:
  Allow new namelists settings, allocate and compute multi-layer depths for IFS
• src_read_coarse_grid.f90:
  Special treatment for IFS grib fields, which are not really multi-layer in the COSMO sense
• src_coarse_interpol.f90:
  New branch in the interpol_coarse_special_ec routine for interpolating in multi-layer style
• src_2d_fields.f90:
  Allow ke_soil_lm /= ke_soil_in for IFS input

Back to Contents

3. Implementation of ART-part with conditional compilation

(by Daniel Rieger, KIT)

Similar to the COSMO-Model, interfaces have been implemented with ifdef ART to couple an additional component INT2LM_ART. This additional component is provided by KIT together with the COSMO_ART package.

Changes in the Namelist variables:
The namelist variable l_chemistry (in /CONTRL/) has been deleted and has been replaced by two new variables: l_art, l_art_nested

Back to Contents

4. Technical Changes and Bug Fixes

• data_int2lm_parallel.f90:
  Introduced new MPI data type imp_integ_ga for special grib_api integers kindOfSize (which could be a 4 or a 8-byte integer). This new data type is initialized in subroutine init_environment from environment.f90
• data_int2lm_parameters.f90:
  Implemented KIND parameters int_ga for grib_api interface (number of bytes, which could be 4 or 8 byte integers) Implemented global KIND parameter int_dp for 8 byte integers.
• data_int2lm_control.f90:
  Renamed ds_grib to ds_grib_single, ds_gribapi to ds_grib_double as more generic names.
• data_int2lm_control.f90:
  Renamed lprog_qrqs to lprog_qr_qs to be consistent with other names.
• external_data.f90:
  Removed unused variables from the USE sections Introduced calls to new module src_read_hhl, to read hhl files.
• environment.f90:
  Extended interface for init_environment, to pass integer type imp_integ_ga for grib_api integer (for length of message in bytes) Boundary exchange in extend_field only for num_compute > 1.
• io_utilities.f90:
  
  • adapted to latest COSMO-Model version (SR check_input_grid: introduced pv)
  • implemented pv in subroutine check_input_grid as allocatable array
  • adapted interfaces to new integer kind parameter from grib_api.
• parallel_utilities.f90:
  adapted to latest COSMO-Model version (introduced NOMPI).
• utilities.f90:
  adapted to latest COSMO-Model version (just some comments).
• src_2d_fields.f90:
  
  • Subroutine init_multi_layer_gme_ml:
    Corrected interpolation weights when interpolating soil moisture from input soil layers to COSMO soil layers.
  • Eliminated use of aklm, bklm in subroutine ground_fields (which was unnecessary).
    Allow ke_soil_lm /= ke_soil_in for IFS input (Davide).
• src_gme_interpol.f90:
  implemented additional ifdef GRIBAPI, which had been forgotten before.
• src_gribtabs.f90:
  Renamed shortName PRS_MIN to RSMIN, which is the correct name for the product (minimum stomata resistance of plants) (also in: src_namelists.f90, src_read_ext.f90).
• src_namelists.f90:
  compiling lists of initial and boundary variables: add old soil variables only if lmulti_layer_in and lmulti_layer_lm are both .FALSE..
• src_read_coarse_grid.f90:
  
  • implemented additional ifdef GRIBAPI, which had been forgotten before.
  • get ivctype from pv(4+ke1in+1) only, if list of vertical coordinate parameters is long enough (it was not coded for old data).
  • Replaced zhsurfs_lm by hsurfs_in in calls to reference atmospheres subroutines for input grid.
  • moved computation of orography splitting for input orography from src_read_ext.f90 to here, because ivctype_in is not known before..
• src_read_ext.f90:
  
  • implemented additional ifdef GRIBAPI, which had been forgotten before
  • moved computation of orography splitting for input orography to src_read_coarse_grid.
• int2lm_org.f90:
  Implemented additional call to org_lm_output for writing HHL-file Implemented conditional compilation for ART usage (KIT) Allow call to org_2d_fields also for lgsm2lm.

Back to Contents

5. Changes of the Namelists

There were the following changes for the Namelist variables:

Back to Contents

6. Changes of the Results

The bug fix in module src_vert_inter_lm changes the results of applications with llm2lm and COSMO input data with a height based vertical coordinate (ivctype_in = 2)!!!

The results of all other applications are not changed.

Back to Contents