Allow ropp2bufr to output satellite POD data

[Ian Culverwell]

Helpdesk enquiry 337 has revealed that the change brought in at ROPP9.0 to use reference POD can sometimes prevent these being included in the BUFR files. This is because ROdata%Lev1a%Missing is being set to .TRUE., which prevents ropp2bufr from using the first r_gns and r_leo of the level 1a data stream to generate POD, if the 'reference' values are unavailable. One solution is to replace

ROdata%Lev1a%Npoints = 0
ROdata%Lev1a%Missing = .TRUE.

at the top of Sec 7.2 of ropp_io/tools/ropp2bufr_ec/mo.f90 by

ROdata%Lev1a%Npoints = MIN ( 1, ROdata%Lev1a%Npoints )

(as was the case before ROPP9.0).

Check this is OK.

[Ian Culverwell]

Note that there is also the requirement to have set

output_lev1a = .true.

in the ropp_pp_occ_tool configuration file, otherwise there will be nothing for ropp2bufr to work on. It would be better to ensure that the reference values, r_leo_ref etc, are correctly set by ropp_pp_occ_tool so that level 1a data need not be output.

[Ian Culverwell]

Confirm that making the above change to ropp2bufr_ec.f90 changes contents of bufr file from

 IN DIRECTION OF 0 DEGREES LONGITUDE, DISTANCE FROM THEM              - - - - -
 IN DIRECTION 90 DEGREES EAST, DISTANCE FROM THE EARTHSM              - - - - -
 IN DIRECTION OF THE NORTH POLE, DISTANCE FROM THE EARTM              - - - - -
 ABSOLUTE PLATFORM VELOCITY - FIRST COMPONENT (SEE NOTEM S-1          - - - - -
 ABSOLUTE PLATFORM VELOCITY - SECOND COMPONENT (SEE NOTM S-1          - - - - -
 ABSOLUTE PLATFORM VELOCITY - THIRD COMPONENT (SEE NOTEM S-1          - - - - -
 SATELLITE CLASSIFICATION                              Code 002020          GPS
 PLATFORM TRANSMITTER ID NUMBER                        NUMERIC               23
 IN DIRECTION OF 0 DEGREES LONGITUDE, DISTANCE FROM THEM              - - - - -
 IN DIRECTION 90 DEGREES EAST, DISTANCE FROM THE EARTHSM              - - - - -
 IN DIRECTION OF THE NORTH POLE, DISTANCE FROM THE EARTM              - - - - -
 ABSOLUTE PLATFORM VELOCITY - FIRST COMPONENT (SEE NOTEM S-1          - - - - -
 ABSOLUTE PLATFORM VELOCITY - SECOND COMPONENT (SEE NOTM S-1          - - - - -
 ABSOLUTE PLATFORM VELOCITY - THIRD COMPONENT (SEE NOTEM S-1          - - - - -

to

 IN DIRECTION OF 0 DEGREES LONGITUDE, DISTANCE FROM THEM              738565.94
 IN DIRECTION 90 DEGREES EAST, DISTANCE FROM THE EARTHSM            -2569257.00
 IN DIRECTION OF THE NORTH POLE, DISTANCE FROM THE EARTM             6656460.00
 ABSOLUTE PLATFORM VELOCITY - FIRST COMPONENT (SEE NOTEM S-1         -589.44513
 ABSOLUTE PLATFORM VELOCITY - SECOND COMPONENT (SEE NOTM S-1         7271.81006
 ABSOLUTE PLATFORM VELOCITY - THIRD COMPONENT (SEE NOTEM S-1         1509.27869
 SATELLITE CLASSIFICATION                              Code 002020          GPS
 PLATFORM TRANSMITTER ID NUMBER                        NUMERIC               23
 IN DIRECTION OF 0 DEGREES LONGITUDE, DISTANCE FROM THEM            -26660274.0
 IN DIRECTION 90 DEGREES EAST, DISTANCE FROM THE EARTHSM               653824.1
 IN DIRECTION OF THE NORTH POLE, DISTANCE FROM THE EARTM             -2427446.5
 ABSOLUTE PLATFORM VELOCITY - FIRST COMPONENT (SEE NOTEM S-1         1383.36682
 ABSOLUTE PLATFORM VELOCITY - SECOND COMPONENT (SEE NOTM S-1        -1811.30249
 ABSOLUTE PLATFORM VELOCITY - THIRD COMPONENT (SEE NOTEM S-1         3093.33545

[Ian Culverwell]

Similarly, if we build with MetDB bufr, then making the above code change in ropp2bufr_mo.f90 has the same effect.

[Ian Culverwell]

(Note again the different precisions of the MetDB and ECMWF BUFR output. ECMWF has

 IMPACT PARAMETER (SEE NOTE 8)                         M              6398258.0

etc, whereas MetDB has

 IMPACT PARAMETER (SEE NOTE 8)                         M              6398258.5

etc.)

[Ian Culverwell]

Commit these changes to the devt branch at r5603.

[Ian Culverwell]

Postpone consideration of the need to ensure ropp_pp_occ_tool sets these reference values appropriately. This might be covered in the long list of DMI mods for ROPP9.1.