November 6, 1997

MEMORANDUM TO:  Record
FROM:           Dennis A. Keyser -- W/NP22
Subject:        Changes to OI/SSI-Data Preprocessor
                (May 1, 1997 Version -- UPDATE # 2)

The May 1, 1997 version of the data preprocessor program
(PREPDATA) was updated for a second time.  UPDATE #2 was
implemented on 1200 UTC 11 June 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   # 2  ***********************

I.  GENERAL CHANGES

     1) Removed logic testing for SEQNTL=FALSE, OBUFR=FALSE and
NLWRD=4 since these are never satisfied with output now only to
PREPBUFR file.  This saves about 1% in CPU time but streamlines
the source code.  There are more changes that can be made in this
way. 

     2) Removed the function MARARC since it pertained only to
the obsolete 36-day HDS archive processing.

     3) The report d-time stored in NOBS(7) is changed from
hundredths of hours to thousandths of hours.  Related to this,
the PREPBUFR mnemonic DHR was been redefined in the prep
bufrtable read in by PREPDATA.  DHR is now scaled to 10**3 rather
than 10**2, with the number of bits for storage increased from 13
to 16 and the reference value changed from -2400 to -24000.  All
this was done in preparation for the input of the new MDCARS
ACARS data from ARINC (in a week or so) which defines report time
out to seconds.

    4) Added new Namelist switch IVADTP to select the format of
input file for the still non-operational NEXRAD (VAD) wind
reports.  When IVADTP = 1, the format is expected to be the WMO
BUFR file that used to be present on the HDS but is no longer
available.  This format uses the W3LIB routine W3FI77 to read the
data. When IVADTP = 2 (DEFAULT), the format is expected to be the
JBUFR file from either the /dcom BUFR tank or a time-windowed
data dump file.  This is the new and expected format for these
data.  This format uses the W3LIB routine W3UNPK77 to read the
data.  This was made available through a corresponding change to
the W3LIB routine W3UNPK77.

     5) Added new Namelist switch FLDMGS to (when TRUE) flag the
moisture (specific humidity) data on all levels for the new Gulf
Stream dropwinsonde reports.  These moisture data will receive
quality marks of 15 preventing them from being used by any NCEP
analysis.  The DEFAULT is TRUE.


II. OPERATIONAL CHANGES

A. CHANGES IN THE PROCESSING OF AIRCRAFT DATA

     1) For AMDAR/ASDAR aircraft reports, the entire (true)
station id (flight id) is now encoded in the PREPBUFR file.  This
ID can be up to 8-characters in length.  This true ID is read in
through unpacked ON29 category 8 code figures 917-920 (2-
characters each).  This was made available through a
corresponding change to the W3LIB routine IW3GAD.  Also of note,
the downstream program PREPACQC was modified to now expect this
new form of ID from the AMDAR reports.  Prior to this, the old
HDS/unpacked ON29 truncated version of the AMDAR ID was encoded
into PREPBUFR.  In this version, still reported in words 11 and
12 of unpacked ON29 (to keep with unpacked ON29 convention), the
ID is always truncated to 6-characters with a "Z" in the 6'th
position to identify the report as AMDAR/ASDAR.

     2) For ACARS aircraft reports, the 7'th and 8'th characters
of the station id (flight id) are now encoded in the PREPBUFR
file.  These 2-characters are read in through unpacked ON29
category 8 code figure 920.  This was made available through a
corresponding change to the W3LIB routine IW3GAD.  If the ID is
less than 8-characters in length, blank fills will appear here. 
This is in preparation for the input of the new MDCARS ACARS data
from ARINC (in a week or so) which creates "bogus" ID's usually
8-characters in length.  The actual ACARS ID's, used up to now,
were never more than 6-characters in length.  This was fine for
the prior version of PREPDATA which encoded only the first 6
characters of the ID into PREPBUFR.  In keeping with unpacked
ON29 convention, unpacked words 11 and 12 will still contain only
the first 6-characters of the ID.

     3) For ACARS aircraft reports, the report time to the
nearest 1000'th of an hour is now read from unpacked ON29
category 8, code figure 921.  This will allow the increased
precision available in the ARINC MDCARS ACARS report time
(becoming operational in a week or so) to be used and
subsequently encoded into the PREPBUFR file. 




B. CHANGES IN THE PROCESSING OF UPPER-AIR DATA

     1) The new (about to become operational) Gulf Stream
dropwinsonde data will set the quality marker for moisture
(specific humidity) on all levels to 15 so it will not be
assimilated by any NCEP analysis (see I.5 above).  The moisture
from these drops in considered to be of bad quality. 


III. NON-OPERATIONAL CHANGES

A. CHANGES IN THE PROCESSING OF NEXRAD (VAD) DATA

     1) NEXRAD (VAD) wind reports can now be read from the JBUFR
data dump file and encoded into the PREPBUFR file (see I.4
above).  These data will VERY SOON become operational in some or
all NCEP systems.