* manuscript now follows the rule from TC (e.g., formatting, sections, etc.)
* automatic figure creation that follows naming convention and merges
 multiple figure into one, if required (using Makefile)
* deleted unused old figures
* added new figures: fig01...10.pdf
* update of references -> literature.bib

* update of framework in line with reviewer's comments and suggestions
* update functionality of optimization to account for two different periods: 1990-1999 and 2090-2100; resolves #10
    * add new script to create random sub-sampling forcing data set
    * update of data retrieval scripts
    * add/update Fortran namelist files
* update of parameter sensitivity tools; resolves #11
* data archiving has now a time stamp; resolves #13
* update of SEMIC to model version 1.2; resolves #9
* update of `Makefile`

* update of `manuscript_semic.tex` according to the reviewer's comments
* update of figures to account for periods 1990-1999 and 2090-2099
* update of script `figures.py` accordingly
* delete old figures
* resolves #12

* add changes for missing values over land and ice
* resolve #8

* revised manuscript according to TC editor (Marco Tedesco)
* comply with TC's technical manuscript guidelines; cleaned up
* changed correspondence email to mariokrapp@gmail.com

* add author affiliation
* change git clone instruction to have a tag
* closes #6

* minor correction to manuscript
* update of cost function calculations
  in parameter sensitivity plot

* move common namelist for model parameters 'surface_physics' into another file
  all compenents can access the *same* mode parameters (less copying errors)
* therefore, other components neede to be updated as well,
  e.g., model code for semic.x and run_particles.x
* closes #7

* change program logging messages to begin with 'semic_...' instead of either
  'smb_...' or 'surface_...'

* use allocatble arrays instead of fixed size (causes segfault)
* update FORMAT in write statements (was 'write(*,*)')

* re-run of production runs: optimal, historical, rcp8.5,
  and sensitivity with newly updated model code (commit #b1574ee)

* add subroutine `smb_output` to perform writing of files in one chunk
* remove unused options (monthly, diag output)
* remove unused model configurations (MAR-SUB, MAR-50)
* clean up code