Set leo_id correctly when reading into ROPP

[Ian Culverwell]

Hans Gleisner (DMI) found that some incorrect values of leo_id were preventing the correct satellite identifier being encoded in the BUFR files. The solution is the wholesale inclusion of

  IF ( data%leo_id(1:2) == 'CH' ) THEN
    data%leo_id = 'CHMP'
  ELSE IF ( data%leo_id(1:4) == 'GRC1' ) THEN
    data%leo_id = 'GRAA'
  ELSE IF ( data%leo_id(1:4) == 'GRC2' ) THEN
    data%leo_id = 'GRAB'
  ELSE IF ( data%leo_id(1:2) == 'CO' ) THEN
    data%leo_id(1:2) = 'C0'
  ELSE IF ( data%leo_id(1:3) == 'MTP' ) THEN
    data%leo_id(1:3) = 'MET'
  END IF

in every read routine (ropp_1dm ropp_2d, atmPrf, atmPhs) of ropp_io/ropp/ropp_io_read_ncdf_get.f90.

[Ian Culverwell]

This has been done at r4915. (This also prevents ropp_io/tools/test2ropp.f90 from generating 'CO??', rather than 'C0??', as a valid leo_id.)

[Ian Culverwell]

And r4916.

[Ian Culverwell]

Checked that ropp_io/tools/gfz2ropp.f90 is OK in this respect - unless and until GFZ start processing COSMIC or GRAS occs, which is presumably fairly unlikely. gfz2ropp.f90 still assumes GRACE = GRACE-A, however, as the DSC file fails to differentiate GRACE-A and GRACE-B.

[Ian Culverwell]

Closing ticket as 'fixed'.