GVAR changes

last updated 5 March 2001

These are notes related to changes in the GVAR format due to the changes in channels and detector sizes for GOES-M through GOES-O, as of early 2001.

Notes from NOAA

Date: Mon, 05 Mar 2001 13:40:04 -0500
From: "Thomas Renkevens" 
X-Accept-Language: en
To: Thomas Renkevens 
Subject: GVAR Users: *ADDITIONAL* Information on Changes to GVAR datastream for GOES-M and beyond - Issued 3/5/01

Attention Users of GVAR Data:
Below is a memo containing *ADDITIONAL* information outlining changes to
the GVAR data stream that are
scheduled for GOES-M and beyond. This email contains all pertinent
information that was in the last email, in addition to more details
about the upcoming changes to the GVAR data stream.
This email is an effort to reach as many GVAR users as possible, so that
necessary changes can be made to their systems. If you wish to receive
this document in WordPerfect format (tables, etc...) or have other
questions, please email me.

Thanks

Tom Renkevens
Satellite Analysis Branch
Satellite Services Division
NOAA/NESDIS/OSDPD
thomas.renkevens@noaa.gov
-----------------------------------
IMPORTANT INFORMATION:
CHANGE OF CHANNELS ON IMAGERS BEGINNING WITH GOES-M


For GOES-M and -N, the selection of infrared channels on the GOES
Imagers will be different from that currently on GOES-8 through -11. 
Users will probably need to modify their GVAR ingest software to
accommodate this change.  The current 12-um channel (channel 5), which
has two 4-km detectors, will be replaced by a 13.3-um channel with one
8-km detector.  The current 6.7-um channel (channel 3), which has one
8-km detector, will have two 4-km detectors, and its spectral response
will be shifted and broadened, giving it a nominal center wavelength
near 7 um.

The current GOES-8 through -11 detector configuration is as follows:

channel number:    2       3      4       5
                   3.9um   ----   10.7um  12um
                           6.7 um
                   3.9um   ----   10.7um  12um


The GOES-M and -N detector configuration will be as follows:

channel number:    2       3     4       6
                   3.9um   7um   10.7um  ----
                                         13.3um 
                   3.9um   7um   10.7um  ----


To keep the channel numbering system constant in time, NOAA will always
maintain the same association between channel number and wavelength.
Also, NOAA will number the channels in order of increasing wavelength.
The new 13.3-um channel will be labeled channel 6 to distinguish it from
the old 12-um channel (channel 5).   (But since we assume the old 6.7-um
and the new 7-um channel are effectively the same, both are channel 3.)
Beginning with GOES-O, another change will be made--channel 6 will have
two 4-km detectors instead of one 8-km detector.  The configuration will
be as follows:

channel number:    2       3     4       6
                   3.9um   7um   10.7um  13.3um
                   3.9um   7um   10.7um  13.3um

The observations in the infrared channels of the imagers are sent to
users in blocks 1 and 2 of the GOES VARiable data stream (GVAR).  For
GOES-8 through GOES-11, the GVAR does not order the channels according
to increasing wavelength, but by the arbitrary system, 4, 5, 2, 3. 
(This ordering was based on the old pre-GOES-8 AAA data stream). 
Beginning with GOES-M, this will be changed.  The data in GVAR will be
in order of increasing channel wavelength.  For example, for GOES-M the
order will be 2, 3, 4, 6.

The new ordering will apply only to the GVAR for GOES-M and beyond. The
original GVAR ordering will be maintained for GOES-8 through GOES-11
throughout their lifetimes on orbit. More information on these GVAR
changes may be found in the following excerpt from the GVAR Requirements
Document:


***Excerpt from "Requirements for SPS, GVAR, and PM Modifications  to
Accommodate Imager Channel Changes for GOES-M and Beyond"  (Michael
Weinreb, 3/16/99)***

1.   Channel Numbering:  For all satellites from GOES-8 through GOES-P,
the channel numbering will be fixed as a function of wavelength, as
follows: 

Channel      Nominal Wavelength
1                 Visible
2                 3.9 um
3                 6.7 um
4                 10.7 um
5                 12.0 um
6                 13.3 um

No imager will have both channels 5 and 6.  Those on GOES-8 through
GOES-L have channel 5.  Those on GOES-M through -O will have channel 6.
For GOES-P, it is TBD.

This numbering system will be incorporated in the index LICHA, which is
word 5 of the line documentation segment of each record of GVAR blocks 1
and 2.

2. Order of Data Records in GVAR:  The order of the assignment of
channels to the records in blocks 1 and 2 of GVAR will be changed,
beginning with GOES-M.  For GOES 8 through GOES-L, it is

Block 1 (four IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 5 (12.0 um)   North detector
Ch 5 (12.0 um)   South detector

Block 2 (three IR records)
Ch 2 (3.9 um)   North detector
Ch 2 (3.9 um)   South detector
Ch 3 (6.7 um)

The terms "North" and "South" detectors refer to the Earth frame of
reference.

For GOES-M and beyond, the records will be ordered with increasing
wavelength.  For GOES-M and -N the order will become

Block 1 (four IR records)
Ch 2 (3.9 um)    North detector
Ch 2 (3.9 um)    South detector
Ch 3 (6.7 um)    North detector
Ch 3 (6.7 um)    South detector

Block 2 (three IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 6 (13.3 um)

And for GOES-O and -P it will become

Block 1 (four IR records)
Ch 2 (3.9 um)    North detector
Ch 2 (3.9 um)    South detector
Ch 3 (6.7 um)    North detector
Ch 3 (6.7 um)    South detector

Block 2 (four IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 6 (13.3 um)   North detector
Ch 6 (13.3 um)   South detector

(If the GOES-P Imager returns to the 12-um channel in place of the
13.3-um channel, then the last two records in block 2 will carry data
for channel 5 instead of 6.)

The word LICHA in each record's line documentation segment indicates the
channel number.

3.  Imager Factory Coefficients in GVAR:   Beginning with GOES-O, the
imagers will have an additional infrared detector. There will be eight,
instead of seven, IR detectors.   In block 0, the amount of space for
the drift correction coefficients (currently words 5587-6304) will need
to be increased by approximately 15% to accommodate the data for the
eighth detector.  To make room in block 0, the factory coefficients
(currently words 6305-8040) will be removed and will be sent instead in
a new type of block 11.  [The reason we retain the drift correction data
but not the factory data in block 0 is that the drift correction data
are updated with each scan line, whereas the factory coefficients remain
the same for a given satellite for all time.  (And, of course, blocks 0
are sent with each scan line, whereas blocks 11 are sent less often.)]

***The following was added in March 2001: The new "Imager Factory
Coefficients" block 11 will actually be introduced with the GOES-M
satellite.  This block 11 will be identified by the value "20" in words
5 and 6 ("Product ID") of the GVAR block header.  However, the changes
to block 0, including the removal of the Imager factory data and the
addition of drift correction data for the eighth IR detector, will not
become effective until GOES-O.***

4. Versions of GVAR: The changes to GVAR outlined above will not be
retroactive.  Data from GOES-8 through GOES-11 will always be
transmitted to users with the current version of GVAR.  The revisions to
GVAR described above will become effective with the GOES-M satellite. 

***The following was added in March 2001: Users can identify the version
of GVAR they receive from word 8 ("Version Number") of the GVAR block
header.   When this word has the value "0," it indicates the original
(GOES-8 through GOES-11) version of GVAR.  The value "1" will also
indicate the GOES-8 through -11 version of GVAR, but with certain spare
words now utilized to transmit the status of a new algorithm to mitigate
certain calibration anomalies that occur near satellite midnight.  NOAA
expects to make version 1 operational in mid- to late-2001.  Its receipt
will not require any modifications or actions by current users.   For
the GVAR versions for GOES-M/N and GOES-O/P, word 8 of the GVAR block
header will take the values "2" and "3,"  respectively. ***

Date: Wed, 28 Feb 2001 14:13:32 -0500
From: "Thomas Renkevens" 
X-Accept-Language: en
To: Thomas Renkevens 
Subject: GVAR Data Users:  Changes to GVAR datastream for GOES-M and beyond - Issued 2/28/01

Attention Users of GVAR Data:
Below is a memo outlining changes to the GVAR data stream that are
scheduled for GOES-M and beyond.  This email is an effort to reach as
many GVAR users as possible, so that necessary changes can be made to
their systems.
If you wish to receive this document in WordPerfect format (tables,
etc...) or have other questions, please email me.

Thanks

Tom Renkevens
Satellite Analysis Branch
Satellite Services Division
NOAA/NESDIS/OSDPD
thomas.renkevens@noaa.gov

------memo below------

Requirements for SPS, GVAR, and PM Modifications to Accommodate Imager
Channel Changes for GOES-M and Beyond

1.   Channel Numbering:  For all satellites from GOES-8 through GOES-P,
the channel
numbering will be fixed as a function of wavelength, as follows:

Channel         Nominal Wavelength
1                    Visible
2                    3.9 um
3                    6.7 um
4                  10.7 um
5                  12.0 um
6                  13.3 um

No imager will have both channels 5 and 6.  Those on GOES-8 through
GOES-11 have channel 5.  Those on GOES-M through -O will have channel
6.  For GOES-P, it is TBD.

This numbering system will be incorporated in the index LICHA, which is
word 5 of the line documentation segment of each record of GVAR blocks 1
and 2.

2.          Order of Data Records in GVAR:  The order of the assignment
of channels to the records in blocks 1 and 2 of GVAR will be changed,
beginning with GOES-M.  For GOES 8 through GOES-11 it is:

Block 1 (four IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 5 (12.0 um)   North detector
Ch 5 (12.0 um)   South detector

Block 2 (three IR records)
Ch 2 (3.9 um)   North detector
Ch 2 (3.9 um)   South detector
Ch 3 (6.7 um)

The terms "North" and "South" detectors refer to the Earth frame of
reference.

For GOES-M and beyond, the records will be ordered with increasing
wavelength.  For GOES-M and -N it will become:

Block 1 (four IR records)

Ch 2 (3.9 um)    North detector
Ch 2 (3.9 um)    South detector
Ch 3 (6.7 um)     North detector
Ch 3 (6.7 um)     South detector

Block 2 (three IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 6 (13.3 um)


And for GOES-O and -P it will become:

Block 1 (four IR records)
Ch 2 (3.9 um)    North detector
Ch 2 (3.9 um)    South detector
Ch 3 (6.7 um)     North detector
Ch 3 (6.7 um)     South detector

Block 2 (four IR records)
Ch 4 (10.7 um)   North detector
Ch 4 (10.7 um)   South detector
Ch 6 (13.3 um)   North detector
Ch 6 (13.3 um)   South detector

(If the GOES-P Imager returns to the 12-um channel in place of the
13.3-um
channel, then the last two records in block 2 will carry data for
channel 5 instead of 6.)

The word LICHA in each record's line documentation segment indicates the
channel
number.

3.         Imager Factory Coefficients in GVAR:   Beginning with GOES-O,
the imagers will have an additional infrared detector. There will be
eight, instead of seven, IR detectors.   In block 0, the amount of space
for the drift correction coefficients (currently words 5587-6304) will
need to be increased by approximately 15% to accommodate the data for
the eighth detector.  To make room in block 0, the factory coefficients
(currently words 6305-8040) will be removed and will be sent instead in
a new type of block 11.  [The reason we retain the drift correction data
but not the factory data in block 0 is that the drift correction data
are updated with each scan line, whereas the factory coefficients remain
the same for a given satellite for all time.  (And, of course, blocks 0
are sent with each scan line, whereas blocks 11 are sent less often.)]
4.         Versions of GVAR: The changes to GVAR outlined above will not
be retroactive.  Data from GOES-8 through GOES-11 will always be
transmitted to users with the current version of GVAR.  The revisions to
GVAR described above will become effective with the GOES-M satellite.

            The new "Imager Factory Coefficients" block 11 will be
introduced with the GOES-M satellite.  However, the changes to block 0,
including the removal of the Imager factory data and the addition of
drift correction data for the eighth IR detector, will not become
effective until GOES-O.

5.         SPS and PM Data Displays and Printouts: For GOES-8 through
GOES-11, data are displayed in the following order:

Ch 4, det 1
Ch 4, det 2
Ch 5, det 1
Ch 5, det 2
Ch 2, det 1
Ch 2, det 2
Ch 3

Beginning with GOES-M, displays and printouts of infrared
channel/detector data will order the channels by wavelength and the
detectors in the instrument frame of reference (physical order), as
follows:

For GOES M-N:

Ch 2, det 1
Ch 2, det 2
Ch 3, det 1
Ch 3, det 2
Ch 4, det 1
Ch 4, det 2
Ch 6


For GOES O-P:

Ch 2, det 1
Ch 2, det 2
Ch 3, det 1
Ch 3, det 2
Ch 4, det 1
Ch 4, det 2
Ch 6 or 5, det 1
Ch 6 or 5, det 2

6.         Versions of SPS and PM Data Displays and Printouts:  TBD

7.   SPS Calibration/Alignment Database

Beginning with GOES-M, the elements of the data arrays numbered by
detector and channel in the SPS Calibration/Alignment Database will be
ordered by channel with increasing wavelength and by detector in the
instrument frame of reference (physical order), as is shown in section 5
above.

8.   Documentation

The changes in GVAR, SPS, and PM, as they will be perceived by data
users, should be documented and diagrammed thoroughly.  A draft of the
documentation should be made available for review by users before ISI
begins the coding.