November 6, 1997

MEMORANDUM TO:  Record
FROM:           Dennis A. Keyser -- W/NP22
Subject:        Changes to OI/SSI-Data Preprocessor
                (May 1, 1997 Version -- UPDATE # 6)

The May 1, 1997 version of the data preprocessor program
(PREPDATA) was updated for a sixth time.  UPDATE #6 was
implemented on 1200 UTC 24 September 1997.  All changes now
affect only OBUFR = TRUE switch.  Except where noted, the changes
below affect both the GLOBAL and RGL/ETA models. PREPDATA now
exclusively uses JBUFR file input.

*********************  U P D A T E   # 6  ***********************

I.  GENERAL CHANGES

     1) Removed call to subroutine ERRSET, as this is no longer
needed on the Cray.

     2) Copied W3LIB routine W3FS22 in-line into source code (as
is).  This is in an attempt to correct the occasional problem of
data clobbering that always occurs in this subroutine.

     3) Namelist switch MSGPSF has been redefined (see III.B.1
and III.B.2 below).


II. OPERATIONAL CHANGES

A. CHANGES IN THE PROCESSING OF GOES RETRIEVAL DATA

     1) The namelist switch GOESPW has been set to TRUE in the
EDAS/RDAS, ERL-ETA, and RGL versions of PREPDATA (these all share
a common data card).  With this change, the EDAS/RDAS and ERL-ETA
network PREPBUFR files will now contain GOES precipitable water
retrievals.  Only the EDAS-OI and ERL ETA-OI will actually
assimilate these data, since the RDASFERR program will filter
them out in the RDAS-OI runs.  Also, although the RGL network
PREPDATA cards also set GOESPW=TRUE, the RGL data dumps do not
include GOES sounding data, so no GOES sounding data will appear
in the RGL PREPBUFR files.







III. NON-OPERATIONAL (BUT-SOON-TO-BE) CHANGES

A. CHANGES IN THE PROCESSING OF GOES RADIANCE DATA

     1) Ancillary data are now processed along with the GOES
radiances.  These additional data are necessary for proper
assimilation of the radiance information in the GBL-SSI.  W3LIB
subroutine W3UNPK77 was modified to return skin temperature,
cloud amount, solar zenith angle and satellite zenith angle. 
They are stored into unpacked Category 8 code figures 255, 258,
260 and 261, respectively.  These are in turn are now processsed
by PREPDATA and stored in COMMON until they are eventually
encoded into the PREPBUFR file along with the GOES radiances, if
the radiances have been selected for processing.  The
prep.bufrtable has been updated to add the following mnemonics:
'ELEV' (satellite elevation angle) and 'CLAM' (cloud amount). 
The other types, while new to GOES radiance processing, had
already existed for TOVS radiance processing. 


B. CHANGES IN THE PROCESSING OF SURFACE LAND DATA

     1. When the namelist switch MSGPSF=TRUE, and the report is a
land surface report within the old LFM domain (if thinning turned
on, otherwise no retriction to location), then the reported
altimeter setting is now read from unpacked ON124 Category 8,
code figure 20 and stored in COMMON.  Eventually it is encoded
into the PREPBUFR file via the new prep.bufrtable mnemonic
'ALSE'.

     2. When the namelist switch MSGPSF=TRUE, and the report is a
land surface report within the old LFM domain (if thinning turned
on, otherwise no retriction to location), then if BOTH station
pressure and mean sea-level pressure are missing AND if the
altimeter setting is non-missing, the sea-level pressure (PMO) is
encoded in PREPBUFR with the value for the altimeter setting, and
the station pressure (POB) encoded in PREPBUFR is calculated via
the standard T,Z function based on the altimeter setting and the
sea-level pressure.  In this case, the report type is set to
183/284, the pressure quality marker is set to suspect, and
characters 6-7 of the station id are set to '$A'.  If the
altimeter is missing, then processing is as before except
characters 6-7 of the station id are set to '$S'.