  lbl2od
  computation of line-by-line optical depth due to molecular absorption

  usage:
  lbl2od  [options]  atm_file  line_parameter_file(s)

  -h             help
  -c   char      comment character(s) used in input,output file (default '#')
  -o   string    output file for saving of cross sections (if not given: write to StdOut)
                 (in case of a list of input files this options specifies just the extension)

  -m   char      mode: 'c' ---> cumulative optical depth
                       'd' ---> difference (delta) optical depth (default)
                       'r' ---> reverse cumulative optical depth
		       anything else: total optical depth

  -i   char      interpolation method   [default: '3' three-point Lagrange,  choices are one of "234lqcsbk"]
  -L   char      Lineshape: V(oigt), L(orentz), G(auss)     [default: Voigt]
  -q   char      quadrature method: default 'T'=trapez, 'S'=simpson
  -s   float     sampling rate used for x-grid (default: 5.0 grid points per (mean) half width)
  -w   float     wing extension (cutoff wavenumber, default 10.0cm-1)

  -n   int       number of grids --- selects 'multigrid mode' for nGrids=2 or 3  (default: nGrids=3 three grid)
  -g   int       gridRatio = ratio of coarse to fine grid spacing  (only 2, 4, or 8, default 8)
  -W   float     transition from fine to coarse grid  (in units of half widths, default 25.0)

 --ToA float     top-of-atmosphere altitude (compute opt.depth only for levels below) 
  -x   Interval  lower,upper wavenumbers (comma separated pair of floats [no blanks!],
                                          default set according to range of lines in datafiles)
 --nm            on output write optical depth versus wavelength [nm] (default: wavenumber 1/cm)
  -r             on output reverse layer optical depth order:  top <--> bottom of atmosphere
  -y             on output write only optical depth without x-values (wavenumbers etc)


  line_parameter_file(s)
  The line parameter file(s) should contain a list of (preselected) lines
  that can be generated from HITRAN or GEISA database with extract.py 
  (See the documentation header of lbl2xs.py for more details)
  
  atm_file 
  The file containing an user's atmospheric profile data has to be in xy  format
  with columns for altitude, pressure, temperature and molecuar concentrations
  (see the documentation for details and the data directory for some examples)

  NOTES:
  interpolation required for multigrid approach and to regrid cross sections to the final dense grid (typically in the top levels)
  * Lagrange linear, quadratic, or cubic interpolation
  * B spline interpolation (uses scipy.interpolate modules splrep, splev)
  * Krogh (1970) interpolation (uses scipy.interpolate.krogh_interpolate)
  Lagrange two-point interpolation (selected by "2", "l", or "L") is clearly the least elaborate and most robust,
  even three-point interpolation can sometimes lead to oscillations associated with some negative xs values 

