September 4, 1996 MEMORANDUM TO: Record FROM: Dennis A. Keyser -- W/NP22 Subject: Changes to OI/SSI-Data Preprocessor (July 12, 1995 Version -- UPDATE # 5) The July 12, 1995 version of the data preprocessor program (PREPDATA) was updated for the fifth time. UPDATE #5 was implemented on 1200 UTC 02 October 1996 on the CRAY. All changes now affect only OBUFR = TRUE switch. Except where noted, the changes below affect both the GLOBAL and RGL/ETA models. ********************* U P D A T E # 5 *********************** I. GENERAL CHANGES 1) Added new namelist switch IEDSTP which will indicate the type of input file from which unpacked NMCEDS formatted TOVS retrievals are returned for RGL/ETA processing (TOVEDS=TRUE). IEDSTP=1 indicates the actual NMCEDS formatted file; IEDSTP=2 indicates the new BUFR dump file. (Default is IEDSTP=1. The NMCEDS file is still used operationally.) 2) Default values for namelist switch IUNIT changed to reflect the addition of the odd GOES-I retrieval satellite (see III.A.1). 3) If either an error or an end-of-file is detected when reading the fixed land-sea-U.S. mask file in IUNIT(3), the program now abnormally stops with a condition code of 49. Prior to this, the program would run normally however it would ignore the land/sea or U.S. in/out checks for certain data types, even if such checks were set-up in the namelist switches. 4) Expanded the sizes of arrays for namelist switches SWINDO, SWNLND, JSWIND, JSMASS to include new (but not-yet- operational) high-density cloud-drift and water-vapor satellite- derived winds (and cloud temperatures) from NESDIS (see III.B.1). 5) Added new namelist switch TOVRAD which, when TRUE, will allow for the processing of TOVS radiances from the new BUFR dump file for GLOBAL processing (TOVBFR=TRUE). (Default is TOVRAD=FALSE. The logic to process TOVS radiances is not yet in place (see III.C.1). 6) Added new subroutines W3FIZX and FIZZ0X which are like existing subroutines W3FIZZ and FIZZ01 except they are set-up to process TOVS radiances in association with a change in the BUFR mnemonic table. These are not actually called anywhere in the program. The will eventually be renamed to the existing subroutine names (replacing them) once the logic to process the TOVS radiances and the new mnemonic table are in place (see III.C.1). 7) Added new namelist switch MSGPSF which, when TRUE, will process surface land reports within the LFM domain with missing station pressure and missing sea-level pressure. In this case, the sea-level pressure is stored as missing and the station pressure is stored as the U.S. Standard Atmosphere value (based on the reported elevation and temperature values). Before, such reports were always excluded from processing (and they still are if outside the LFM domain). A $ is placed in the 6'th character of the station id in order to identify such reports. In addition, the calculated specific humidity for such reports is given a quality marker of suspect since the pressure is used in the calculation. (Default is MSGPSF=FALSE. The default value is currently used in all operational systems. Switch was set to TRUE for special high-resolution ETA model Olympic runs in August 1996.) 8) Logic added to check for wind speed in meters per second times 10 which is stored in Office Note 29, category 8, code figure 924. Currently no operational data types store this, however special ACARS data sets, high-density satellite derived wind data sets (NESDIS), and surface mesonet data sets used in the special high-resolution ETA model Olympic runs (August 1996) did store this. II. SPECIFIC OPERATIONAL CHANGES A. CHANGES TO TOVS SATELLITE RETRIEVAL PROCESSING 1) RGL/ETA: Although not yet switched on (see I.1 above), logic is now in place to return TOVS retrievals in NMCEDS format from the new BUFR dump file for TOVS data. A new subroutine is called, W3TOVEDS, to decode the BUFR messages and return a retrieval in this format. The BUFR file will use the same unit number as the old NMCEDS file did. The unit numbers that currently hold the flag and directory files for the NMCEDS file will no longer be used once the switch is made. B. CHANGES TO WIND PROFILER PROCESSING 1) Logic for processing wind profiler reports will no longer require that the BUFR messages have only edition 0 or 2. This is a result of a future (planned December 1996) change in the edition number of wind profiler BUFR messages from 2 to 3. Those quality markers that were previously identified only for edition 2 are now valid for all editions of 2 or greater. 2) For BUFR edition 2 or greater, levels with quality code indicator of 2 (shear check and/or median check not performed or inconclusive) now get a PREPBUFR table value (q.m.) of 11 (BAD) and they will not be used by any analyses. Before, such levels received a table value (q.m.) of 2 (neutral quality) and were subject to assimilation. III. SPECIFIC NON-OPERATIONAL CHANGES A. CHANGES TO GOES-I SATELLITE RETRIEVAL PROCESSING 1) Updated logic to process retrievals from both the even GOES satellite (currently GOES-08) and the odd GOES satellite (currently GOES-09). Before only the even satellite was set-up for processing. Each satellite read in via a unique unit number. Also, unique unit number assigned for each satellite for re- written BUFR messages within the +/- 3-hour time window. 2) Improved the way the center time and time window are passed into an updated version of the retrieval unpacker, W3FI77. B. CHANGES TO SATELLITE-DERIVED WIND PROCESSING 1) Updated logic to process new high-density cloud-drift and water-vapor winds (and cloud temperatures) produced by NESDIS. New and changed OI/SSI report types are: 241/141 - High-density GOES-ODD satellite visible cloud drift 245/145 - High-density GOES-ODD satellite IR cloud drift 246/146 - High-density GOES-EVEN satellite IR cloud drift 247/147 - High-density GOES-ODD satellite imager water vapor 248/148 - High-density GOES-EVEN satellite imager water vapor 249/149 - High-density GOES-EVEN satellite visible cloud drift 251 - Conventional Indian satellite cloud drift (no mass part processed, high and low levels) 256 - Conventional GOES satellite gradient (no mass part processed, high and low levels) 259 - Conventional GOES satellite water vapor (no mass part processed, high and low levels) 267/167 - High-density GOES-ODD satellite sounder water vapor 268/168 - High-density GOES-EVEN satellite sounder water vapor C. CHANGES TO TOVS RADIANCE PROCESSING 1) Logic is partially in place to process the 27 TOVS radiances.