Also fixed option_verbose=10 in readinput.f90

Flexpart-wrf didn't complain if the species list contained more entries
(numtable) than the maximum number of species (maxspec). This resulted in
terrible out of bound nastiness.

Furthermore a faulty label/address was fixed (see ticket #110).

(Problems after merge JB)

Conflicts:
	read_ncwrfout.f90
	readinput.f90

Conflicts solved by using LOCAL

Some changes in par_mod.f90
(This is mainly for personal use)

If the total number of particles exceed the maximum expected number, an
error message is given before stoping. The old message gave suggestions
for solving the problem which would never work with the new approach for
allocating memory.

It has been replaced to better inform the user.

Mainly a fix in netcdf_output_mod.F90 but it required updated calls in
the concoutput* subroutines.

Removed some debugging comments
Removed old netcdf output related files

Fixed length of variables
Removed redundant definintions

The netcdf3 routines have also been updated but not tested...