wiki:ticket/349/TicketSummary

Version 2 (modified by Jin Lee, 5 months ago) (diff)

--

Build and Install SSMIS UPP Averaging Module from NWP SAF

Source

Fiona downloaded source tarball from,

https://www.nwpsaf.eu/site/software/ssmis-pp

A copy of the tarball is,

raijin4:/home/548/jtl548/da/sat_da/processing_package/ssmis_pp/ssmis_upp.tgz

The source was imported to the following SVN repos location,

https://access-svn.nci.org.au/svn/nwp/da/sat_da/processing_package/SSMIS_PP

Initial build

Development branch

Initially I built the package under my personal directory. Once it is successfully tested I will install it under a more general location.

I branched from the "trunk" and my development branch is,

https://access-svn.nci.org.au/svn/nwp/branches/dev/jtl548/da/sat_da/processing_package/ssmis_pp/SSMIS_PP

I then checked out a working copy,

raijin4:/home/548/jtl548/da/sat_da/processing_package/ssmis_pp/SSMIS_PP

Building Met Office Bufr library

Following module environment is used,

Currently Loaded Modulefiles:
  1) intel-cc/15.0.1.133   2) intel-fc/15.0.1.133   3) subversion/1.9.0

Note 1. I ignored warning message about -w90, -w95 and -cm

libbufr.a is built

Building SSMIS UPP Averaging module

Same compile-time environment as for Bufr library build was used.

Successfully built SSMIS_UPP_PREPROCESSOR.exe.

Running SSMIS UPP pre-processor

Created a subdirectory, scripts/ underneath of which a copy of the script given in Appendix 5 of "SSMIS UPP Averaging Module - Technical Description (Version 1.0, 19th November, 2010)" was placed.

The filename format used by UKMO for the NESDIS data stream prior to averaging is,

NPR_TDUP.SB_D{yy}{jjj}_S{hhmm}_E{hhmm}_B{orbit_no}_NS

  • e.g. NPR_TDUP.SB_D13254_S1202_E1347_B3535657_NS

NPR_TDUP.SC_D{yy}{jjj}_S{hhmm}_E{hhmm}_B{orbit_no}_NS

  • e.g. NPR_TDUP.SC_D13254_S1248_E1433_B2010405_NS

Note 1. SA, SB, SC refer to F16, F17, F18 respectively

Q We might have to split each SSMIS Bufr file into separate Bufr files, each containing data only from one DMSP satellite. This probably is needed as there are coefficients valid for each satellite (?)