Documentation of the Changes in the COSMO-Model
Version 5.4

Version 5.4 is a development version, which will not be officially released.

Contents:

1. Changes in the Assimilation
2. Changes in the Dynamics
3. COSMO-ICON Physics
4. Digital Filter Initialization
5. Technical Changes and GPU Management
6. Further Bug Fixes
7. Changes to the Namelists
8. Changes of Results

1. Changes in the Assimilation

(by C. Schraff)

Removal of AOF interface

• Removal of the interface for reading observations from AOF binary files and subsequent processing of these data for nudging. Removal of the dependencies of the AOF interface and code clean-up.
• For VOF ('YUVERIF') file output: passive bit in station characteristics is now always set when height flag, redundancy flag, or blacklist flag (for aircraft) is set.
• The interface for reading GPS ZTD (or IWV) data from ASCII files with the COST-716 Version 2.0 template is maintained (except for the outdated bias correction) and some diagnostic messages on the standard output added.
• The directory to which the NetCDF 'fof' feedobs (or feedback) file is written can now be selected by namelist.

The following source modules have been removed from the active code:

• data_obs_process.f90
• src_obs_processing.f90
• src_obs_proc_aof.f90
• src_obs_proc_util.f90

One new namelist variable has been introduced to specify the directory, where the NetCDF fof output file shall be written.

Several other namelist variables, dealing with the AOF interface, have been eliminated:

• itype_obfile
• yaofpath
• lpraof
• dinlat, dislat, diwlon, dielon
• noctrq

Also the namelist variable lgpsbias has been eliminated.

Back to Contents

2. Changes in the Dynamics

Targeted diffusion of cold pools (by Oliver Fuhrer, Pascal Spoerri, Xavier Lapillonne)

The targeted diffusion for cold pools should avoid unrealistic undershootings in the temperature field near the bottom, if the value of a grid point is considerable colder than the average value of the surrounding. This diffusion is switched on, if the difference is larger than a pre-determined threshold value. Up to now this threshold has been hard-coded to 10 degree Kelvin. Also, targeted diffusion was always switched on in the model.

Targeted diffusion has now been made configurable by introducing two new namelist variables in the group /DYNCTL/:

If the default settings are used, the results will not be changed

Implement simple clipping for Semi-Lagrange advection scheme (by Oliver Fuhrer)

A simple clipping for the Semi-Lagrange advection scheme has been implemented by introducing the option "SL3_SC" for the namelist variable y_scalar_advect.

Hardcode the number of saturation adjustment iterations to 3 for CPP_DYCORE (by O. Fuhrer)

In the COSMO-Model the number of saturation adjustment iterations depends on the maximal vertical velocity. Usually only one iteration is enough, only for larger velocities, 2 or 3 iterations are done. In order to have reproducible results, independent of the domain decomposition, the maximal vertical velocity is determined over the full domain. This requires a global communication.

To save this global communication in the CPP_DYCORE case, the number of iterations to do is always 3. This has been hardcoded.

Update of serialization statements (by P. Spoerri)

The serialization statements were updated to cosmo 5.04. Several fields were added to serialize data consistently and to provide effective tagging of the individual serialized fields.

Affected files:

• lmorg.f90
  
  • Integration of lwest_lbdz, least_lbdz, lnorth_lbdz, lsouth_lbdz, hhl fields, because they are serialized with the ConstantFields and passed to the C++ Dycore. qvsflx, lhfl_s, shfl_s, umfl_s, vmfl_s: needed for the DycoreUnittest.
• fast_waves_sc.f90
  
  • Integration of the irk value into fast_waves_strong_conserv, this field is needed for tagging the serialization savepoint.
  • Merging of the FastWavesSCUnittest.RHS and FastWavesSCUnittest.WPPTP into FastWavesSCUnittest.RHSWPPTP since the update of RHS and WPPTP is now computed level by level. Removal of several intermediate fields.
• src_advection_rk.f90
  
  • Adapted the serialization of AdvectionPDBottUnittest.Init to reflect the changes in advection pd and added the wgtfac_u and wgtfac_v fields.
• src_slow_tendencies_rk.f90
  
  • Integration of ntstep and irk in complete_tendencies_uvwtpp since both fields are used in the serializer for tagging the current timestep.
  • Replacement of t_s_deprecated with t_g in the serializer fields.
• src_runge_kutta.f90
  
  • Integration of rho, rdcoef, dqvdt, qvsflx, lhfl_s, qrs, t_g for serialization.
  • Adaptations for the sedimentation and vertical diffusion.
  • Adding t_g for the VerticalDiffusionUnittest.DoTracers because of changes in the handling of the vertical diffusion.
• src_cpp_dycore.f90
  
  • Removal of t_s_nnow in the serialization.
  • Uncommented SEDIM and SPPT_PERT integration because this integration was not ready upon release.

Back to Contents

3. COSMO-ICON Physics

(by X. Lapillonne)

The module src_block_field.f90 has been refactored to allow for future optimizations (for example doing a multi-copy).

Note, that the OpenACC statements in the copy function are still deactivated at the moment. They can only be activated, if the full physics runs on the GPU.

In addition, some editorial changes have been done:

• Adapt spacing to COSMO style (subroutine declaration, mathematic operation).
• Removed obsolete copylist in gscp_interface.f90 and src_block_fields_org.f90.

Back to Contents

4. Digital Filter Initialization

(by L. Torrisi)

Better treatment of clouds and precipitation during diabatic DFI initialization

The subroutine dfi_initialization.f90 has been modified to have a better treatment of clouds / precipitation during the diabatic DFI step. Filtering the qx-variables tends to smooth the structure of the initial state clouds. To mitigate this problem, two new namelist parameter have been introduced to change the treatment of the qx-variables but also of the soil variables in the DFI. Note that these new variables are only active for the option ndfi=1.

If the default values are used, the results will not be changed. The recommended settings that were tested (results shown at COSMO-GM 2015) are:

• itype_dfi_qx = 2
• itype_dfi_soil = 1

Activating output for diabatic DFI for step 0

For diabatic digital filtering initialization (ndfi==1) no output was produced for step 0 up to now. As this is needed for many operational applications, this has been implemented now by calling:

• near_surface
• organize_satellites
• organize_data ('results')

Back to Contents

5. Technical Changes and GPU Management

Added tracking of boundary fields and copying to GPU (by O. Fuhrer)

The fields which are read as boundary conditions need to be copied from CPU to GPU. Since this is implemented as a dynamic process and since copying from CPU to GPU is an expensive operation, we track the fields which are actually being read in src_input.f90 (only in case #ifdef _OPENACC) and fill a data structure with the information required for the copy from CPU to GPU (which is implemented in acc_utilities.f90 and called from lmorg.f90).

Running on GPU (by X. Lapillonne)

At the moment, only the radiation scheme can run on GPUs. For all other parts of the model, which already contained some OpenACC statements, these are de-activated for now.
(In modules: gscp_interface, meteo_utilities, src_block_fields)

Some corrections and adaptations have been done in modules:
organize_physics, radiation_interface, radiation_rg, radiation_utilities; does not change results!

Device Management on GPUs (by X. Lapillonne)

A module acc_device_management.f90 has been added. This module also contains the routine printGPUMem, which has been in acc_utilities.f90 before.

Further Technical Changes

• COSMO-ART:
  Adaptations for the blocked data structure in modules radiation_rg, radiation_interface, organize_physics; does not change results!
• Information on code owners has been updated in several modules.
• Removal of iintegers and use of KIND parameter for long integers (i8) from kind_parameters.f90 in modules:
  • utilities.f90
  • src_meanvalues.f90
  • lmorg.f90
  to be able to correctly compute the number of seconds or hours passed even in long term climate simulations.
• src_input.f90: The checking of the date has been switched off for the HHL fields.

Back to Contents

6. Bug Fixes

• organize_data.f90:
  Replaced VORTIC1, VORTIC2, VORTIC3 with VORTIC_U, VORTIC_V, VORTIC_W, which are the real names of these variables.
• radiation_utilities.f90 Subroutine compute_sunshine_condition:
  For variable zcosphi the loop indices have been forgotten (does not change results!)
• netcdf_io.f90:
  After reading delta_t and h_scal these variables have not been distributed to the other processors in SR distribute_values_asynio.
• pp_utilities.f90:
  Modified error treatment in LPI computations, to avoid model crashes due to problematic thermodynamic profiles.
• src_obs_rad.f90:
  Bug fix in SR input_obs_satpp: had to move a part outside an IF-statement

Back to Contents

7. Changes to the Namelists

Back to Contents

8. Changes of Results

With the default settings of the namelist variables, there are no changes to the results.

On the Cray XC there are numerical changes to the results due to compiler optimizations.

Back to Contents