GitLab

Changed from grib_api to eccodes. MPI: implemented linversionout=1; fixed calculation of grid_initial fields.

Adding the full GPL license it was apparent that the FSF recommends to
have the statements as file headers

This reverts commit 3481cc18.

bugfix: particles were lost at start of global domain filling run. Error in restarting domain filling run from particle dump

Fixed an issue with nested input fields, and nested output fields that have more than 1000 points in x or y (netcdf output)

Bugfix for double precision dry deposition calculation. Added (hardcoded) option to not use output kernel. Version/date string updated.

Fix for issue #151 (improper use of MPI_IN_PLACE). Also changed default RELEASES file to use SPECIES_001.

SPECIES files converted to namelist format. Fixed-format SPCIES files renamed with .oldformat extension. Added 2 wet depo parameters to old-format SPECIES. Renamed internal variables and parameters used for wet deposition. incloud_ratio increased to 2.2

Code for cloud water should be correct if the total cw is stored in field clwc (old scheme) or in field qc (new scheme). Minor edits in some files.

bugfix: MPI version gave wrong wet deposition when using ECMWF cloud water fields. Cloud water in ECMWF fields now uses parameter qc, or reads clwc+ciwc. Added minmass variable as limit for terminating particles.

Added a modification so thah number of particles in simulation reported to standard output (MPI version) is exact instead of estimated