= Summary of user-affecting changes to PALM =
(e.g. palmrun/palmbuild options/variables, .palm.iofiles, Namelist names and parameters, rrtmg changes require new build of rrtmg, etc)
* palmrun/palmbuild options {{{-h}}} and {{{-d}}} will be renamed {{{-c}}} and {{{-r}}}
* change 2
* etc

= Summary of relevant code changes, new features =
* The basic equations in PALM have been changed to an anelastic approximation. This allows for an air-density change with height and enables PALM to simulate deep convection. The new functionality can be steered with the initialization parameter {{{approximation}}}. A Boussinesq approximation of the basic equations can still be used and is in fact the models default configuration. Note that all internal flux quantities are now quantities per unit volume instead of per unit mass (not kinematic anymore). The input and output of flux quantities with regard to their physical interpretation (e.g. units) can be steered by the initialization parameters {{{flux_input_mode}}} and {{{flux_output_mode}}}.
* The bulk cloud microphysics model code has been modularized and its capabilities can now be controlled within a new namelist called {{{bulk_cloud_parameters}}}. 
 - It is now allowed to enable the bulk cloud model and use complex topography at the same time.
* ocean code has been modularized and put into new file {{{ocean_mod.f90}}}. Turbulence generation by wave breaking, as well as the Stokes drift have been implemented in the default code. New namelist {{{ocean_parameters}}}
* introduced changes to {{{date_and_time_mod.f90}}} for calculations of time indices such as hour of the day or month of the year, according to the running time of the simulation
* New feature Multi Agent System (MAS) has been added. It is modularized in the new file {{{multi_agent_system_mod.f90}}} and can be steered using the new namelist {{{agent_parameters}}}.
* Chemistry: 
 - new module for processing chemistry emissions ({{{chem_emissions_mod.f90}}})
 - introduced deposition of gases and particles in chemistry model
 - subroutines chem_check_parameters and chem_header
 - 2d and profile output for chemistry variables added
 - 'de-cycling' of chemistry variables at lateral boundaries
 - additions to {{{UTILS/chemistry/gasphase_preproc}}} and modifications of {{{kpp4palm}}}: automatic preparation of vector version, inclusion of fixed species, more mechanisms added (CBM4, phstat, phstatp, simplep, passive1)
 - about 1000 (at least if feels like this) small fixes
 - adaptation of namelist {{{chemistry_parameters}}} to new features
* Self nesting
 - nesting of chemistry quantities implemented
 - RANS-RANS and RANS-LES nesting implemented
* New feature Offline Nesting has been introduced
 - time-dependent boundary conditions at the model boundaries as well as time-dependent geostrophic forcing can be derived from the  output of the larger-scale COSMO model via the pre-processing tool INIFOR
 - information is stored in dynamic input file 
 - to initiate turbulence, a synthetic turbulence generator is applied at all lateral boundaries, which imposes turbulent fluctuations onto the COSMO-derived boundary values of the three velocity components, while the strength of fluctuations depend on the actual atmospheric conditions
* The vertical grid stretching mechanism has been extended to allow more than one stretched region. In addition, stretching from coarse to fine as well as from fine to coarse is possible now. For more information see parameter [../../../app/initialization_parameters/#dz dz] or have a look into file {{{init_grid.f90}}}. 
* The model now comprises a biometeorology module that can be used to determine the thermal indices Perceived Temperature (PT), Universal Thermal Climate Index (UTCI) and Physiologically Equivalent Temperature (PET). The indices are calculated as soon as the namelist {{{biometeorology_parameters}}} exists and can be disabled by setting e.g. {{{biom_pt = .F.}}}.
* New feature human exposure model has been added, that calculates the vitamin D3-weighted UV exposure of a human. It is modularized in the new file {{{uv_exposure_model_mod.f90}}} and can be steered using the new namelist {{{biometeorology_parameters}}}.
* Revision of NetCDF data input and extended checks for consistent static and dynamic input files
* Reflection of Lagrangian particles at downward-facing surfaces implemented
* Output of unstructured surface data is enabled via a new namelist {{{surface_output_parameters}}}. This allows further processing of surface data into Paraview-readable files in order to display surface data such as surface temperature at building walls in an urban-environment
* New module to perform virtual measurements {{{virtual_meas_mod.f90}}}
 - NetCDF data input of coordinates from 'real-world' measurements for point-, profile-, or trajectory measurements
 - automatic sampling and output of the quantities according to the measurement coordinates
* urban surface mod
 - output variables {{{usm_t_surf_*}}} for bare building walll and roof tiles was renamed to {{{usm_t_surf_wall_*}}} (* = north/ south/ east/ west/ roof)
 - calculation of short wave radiation inside window layers changed
 - increased building database to provide parameters for all wall and window height levels (ground floor, above ground floor, roof level)
 - window surface and layers temperatures are fixed during spinup to make larger time step {{{dt_spinup}}} possible
 - latent heat flux for green wall and roof tiles and output variables {{{usm_qsws_*}}}, usm_{{{qsws_veg_*}}}, {{{usm_qsws_liq_*}}}
 - calculation of latent heat flax for green roof tiles according to soil water content, green wall tiles dont take soil water content into account
 - output parameter {{{usm_swc_*}}} for soil water content of green substrate layers
 - calculation of green roof substrate layers heat capacity and conductivity according to soil water content and soil parameters
 - building database parameter green_type_roof to choose between intensiv and extensiv green roof type
 - new boolean namelist parameter {{{usm_wall_mod}}} (default = .F.) to decrease heat advection during spinup in the first two wall layers by factor 10 to increase the possible time step {{{dt_spinup}}}
* change 3
* etc