To convert emissions of single species to bulk emissions as provided by inventory, some conversion factors need to be applied.

The preprocessor now accounts for the station height and reads model values from the corresponding layer.

Initial commit. Data of German states is written to a file

Timeseries can now be calculated for any time interval, which is to obe defined by calling the "plot_mean_cntry" or "plot_error_cntry" routines. Also, some Id for the plotted timeinterval should be specified (called month in the plotting routine).
Update of the error plot. Now, the relative error of the analysis and background run compared with the reference run are plotted

Routine processed netcdf files that are created per month from plot_realEMFAC. The routine is no ready to efficiently analysis the full timeseries of the emissions factors.

Same as Compare_preprocessed_data but for analysing the full time series, not only single month. Further, reference run is included in all functions and plots.

.gitignore contains checkpoint files

Standardized to process full months.
Added reference run to be read from the preprocessed netcdf file (reference run in preprocessed data is required). However, reference run is not included in the plotting routines (Meanwhile the compare-preprocessed-data routine is outdated. Newer versions named "full_year..." are recommended to use (see later commits).

Optimized treatment of timeseries in the netcdf file. Observation and model timeseries are no longer wrote into single variables but into one variable with dimension (ntimeseries,ntimes), where ntimeseries is the number of timeseries (=3: obs, BG, ANA) and ntimes is the number of timesteps (i. e. hours in the timeseries)
- Addition of a new function dump_to_nc_full_ts that allows to write the full timeseries per station in a single netcdf file. This is needed to efficiently use the preprocessed data in a further analysis/postprocessing.

Unnecessary cells are removed from the notebook. No changes in the important cells

- Date range for plotting the emissions is now flexible (not restricted to single months)
- flag plot_quiver is added to opt out wind arrows on time series plots
- datsets are assigned the same time flags, thus, datset can be modified directly (without using Dataset.values)
- add option to dump emission data (BG emissions and analysis emissions) to a netcdf file. These preprocessed data can efficiently be used be other routines as full_year_realEMFAC to analyse emission corrections averaged for the full timeseries

Addition of plot routines to plot nested domains onto the map (plot_nested_domains), plot number of O3 sonde profiles onto the map at each location,and compare model results with O3 sonde profiles and cams data as monthly averages (o3_sonde_profiles).

Script that shows the location of the 1km grid (Niedersachsen, NRW, Berlin) within the location of the 5km grid (all with respect to the geogrid output). For comparison and evaluation of the location, the ground stations taken for assimilation and validation are plottet.

Added averaged daily emission factors (i. e. the change of the emission factors from one day to another, spatially averaged for each country)
Added WRF wind barbs (averaged over station locations within country to EMFAC-time series plots)
Plot output changed to "png"

Costs are plotted for arbitrary time window (missing data included)

Change in CAMS_ctm_preprocessor. Now, modeled wind components are added at station locations to the preprocessed data file for further comparison.

Improved model layout, added European map for temporal averaged correction factors, improved colorbar (log-norm centered at 1 with 7 ticks)

It seems like in the first commit the compare_preprocessor_data file was empty.

Added observation preprocessor, plotting of costs and computation and plotting of emissionfactors