No GRAS negative longitudes
Sean Healy (ECMWF) has reported that the test GRAS BUFR data contains no occultations in the Western hemisphere (ie no longitude values -180 to 0deg). The BUFR definition for longitude as used for RO is for the range +/-180, so ~50% of GRAS occultations ought to be <0. (A check with COSMIC shows the distributon is as expected).
Martin Sorenesen (DMI) confirms that longitude values for GRAS are in the range 0-360deg and
values >180 are present in the netCDF files. The ropp2bufr tool should cope with such values, as code is in place to detect >180 and convert it to <0. In a specific case, netCDF longitude is 249.2deg, but is (+)110.8deg in the BUFR.
The ROPP I/O code needs to be investigated as to why the longitude value is not being converted to a negative value (the reported sample case would indicate that some conversion is being attempted, but the sign is being lost).
Change history
(4)
| Resolution: |
→ fixed
|
| Status: |
new → closed
|
Found that ropp2bufr effectively had the code:
which should be:
to correctly generate negative values. Fix now in place & tested with sample GRAS netCDF file containing lon=249.2, correctly encoding lon=-110.8 (was +110.8)
Also added a line to print the lat/lon value to stdout as part of -d (diagnostic) switch output.