November 6, 1997

MEMORANDUM TO:  Record
FROM:           Dennis A. Keyser -- W/NP22
Subject:        Changes to OI/SSI-Data Preprocessor
                (May 1, 1997 Version -- UPDATE # 7)

The May 1, 1997 version of the data preprocessor program
(PREPDATA) was updated for a seventh time.  UPDATE #7 was
implemented on 1200 UTC 22 October 1997.  Except where noted, the
changes below affect both the GLOBAL and RGL/ETA models.

*********************  U P D A T E   # 7  ***********************

I.  GENERAL CHANGES

     1) Removed namelist switches LNDHRY and BUOHRY as asynoptic
surface data processing is no longer performed (see III.B.1
below).

     2) Removed namelist switch FILLP, and all logic testing it's
value, since it had always been set to TRUE anyway.

     3) Removed namelist switches SEQNTL, OBUFR, LIMITA, LIMDAY
and LIMON (ans all associated testing logic) since these are
obsolete with the conversion of the data processing to the Cray
and the prior removal of the old 36-day archive processing logic
from this program.

     4) Namelist switch arrays SWINDO, SWNLND, JSWIND and JSMASS
all expanded from (6,5,7) to (6,5,10) and element 3 redefined in
order to handle the processing of the new NESDIS high-density
satellite cloud-vector data (see III.A.1 below).


II. OPERATIONAL CHANGES

A. CHANGES IN THE PROCESSING OF ACARS AND AMDAR/ASDAR AIRCRAFT    
   DATA

     1) RGL/ETA only: The namelist switches IACFTL(2) and
IACFTL(4) have been changed to -200 from 2286 and 0,
respectively.  This drops the floor on the processing of ACARS
aircraft data from 0 to -200 meters and drops the floor on the
processing of AMDAR/ASDAR aircraft data from 2286 to -200 meters.





B. CHANGES IN THE PROCESSING OF SATELLITE CLOUD-VECTOR DATA

     1) GLOBAL only: The namelist switch JSMASS has been set to
REMOVE (rather than just flag) all satellite cloud-vector mass
reports.  (RGL/ETA: Never was processed.)

C. CHANGES IN THE PROCESSING OF GOES SATELLITE SOUNDING/RADIANCE  
   DATA

     1) RGL/ETA only: The field-of-view number is now encoded
into the PREPBUFR file for all GOES sounding reports under the
mnemonic descriptor "ACAV". (NOTE: For GLOBAL: non-operational
GOES radiances reports also now include this.)  This necessitated
a change in the bufrtable as well.

D. CHANGES IN THE PROCESSING OF CLASS SOUNDING DATA

     1) The definition of a CLASS sounding report was made more
restrictive.  Instead of simply being defined as a report with
ON29 report type 13, the additional constraint is that the report
id must begin with "CLAS".  This is necessary because ON29 report
type 13 is also used by mobile radiosondes.  The GLOBAL currently
does not process CLASS mass and wind data (LSMASS and LSWIND are
set to 9999), but it should process these mobile radiosondes, and
this change allows for this.


III. NON-OPERATIONAL CHANGES

A. CHANGES IN THE PROCESSING OF SATELLITE CLOUD-VECTOR DATA

     1) Logic has been added to process the new NESDIS high-
density satellite cloud-vector data. Specifically, high-density
IR cloud drift (as report type 245/145), visible cloud drift (as
report type 251/144), imager water vapor [both deep-layer (as
report type 247/147) and cloud-top (as report type 246/146)],
sounder water vapor [both deep-layer (as report type 249/149) and
cloud-top (as report type 248/148)], and new processed (but not
high-density) picture triplet (as report type 255).  This logic
replaces old logic which handled high-density winds as they
appeared a year and one-half ago.  The namelist switch JSWIND has
been set-up to process the following types, but flag them for
non-use by any analysis (the remaining new types are not yet
available and are set to not process): High-density IR cloud
drift winds, high density imager water vapor winds (both deep-
layer and cloud-top), new-processed picture triplet cloud-drift
winds.



B. CHANGES IN THE PROCESSING OF ASYNOPTIC SURFACE DATA

     1) All logic pertaining to the processing of asynoptic
surface land and marine buoy data has been removed.  Such data
were not processed operationally through this logic, since they
are part of the BUFR data dumps under the conversion to the CRAY
data processing.  This logic had been used by the old HDS 36-day
archive processing which is no longer in PREPDATA of course.