Status of Level 1Bs and Plans for Updates

Below is updated information on the status of the current level 1B data
with anomalies and plans for updates. If you have any question please
contact either myself (301-457-5262) or Arlington Morgan (301-457-5262

STATUS OF LEVEL 1B UPDATES: These are the corrections that we are
currently making for the anomalies found in the NOAA-16 and 15. All
previous notices indicating when every thing changed are on the web site
(/PSB/PPP/PPP.html click on Level 1B notices).

We are no longer using the pitch correction for NOAA-16 instrument
data. Instead, we are using a timing correction of 1 second for all
instruments to fix the along track error along with the normal SOCC
clock drift correction. For Example: On May 16 at 0000Z the clock
error determined was +862 milliseconds (ms) (+1 sec for along track
error and 138 ms SOCC clock drift error).

The across track error seen in the AVHRR data is corrected by using a
scan angle of +/-55.25 degrees (stepping angle of 0.05398143624 degrees
instead of the normal +/-55.37 (stepping angle 0.05409868099 degrees).

The across track error seen in the HIRS data is believed to be a mirror
misalignment. To properly earth locate the data we are correcting for a
roll error +1.8 degrees. Any user wishing to collocate the corrected
HIRS data with other instruments must realign the spots to match the
other instruments unless they are using lat/lons to collocate.

The AVHRR is still having problems and is not providing reliable data.
When the data is good, we are correcting earth locations for clock drift
errors as well as correcting for a yaw error of +0.2 degrees.

1.) Correct the NOAA-16 AVHRR PRT values. (User impact to be
determined) We are currently using NOAA-15 PRT values for NOAA-16 and
will change them to the appropriate values as identified by the
instrument scientist.

2.) Correct calibration for first scan line of HIRS pass. (No impact to
users) Sometime since the NOAA-16 AIP processing became operational, it
was noticed that the intercept in the first scan in the HIRS NOAA-15/16
1b*/1b was incorrect. The problem caused the temperatures in the first
scan line to be wrong and appear as a distinct anomaly in the images
generated from the HIRS data. Complicating the problem was the
erroneous setting of a time sequence error flag. Analysis revealed that
the the time sequence flag started being set when clock corrections were
turned on to correct an along track error.
It was discovered that the flag was being inadvertently set because the
initial expected time used for comparing successive scans was not a
corrected time. This error was corrected and rather than concentrate on
a time sequence problem, personnel began examining the HIRS calibration.
After much examination of the entire calibration process, an anomaly was
discovered in the part of the secondary mirror interpolation scheme
which uses a polynomial algorithm to convert counts to secondary mirror
temperatures. The interpolation algorithm assumes that three
temperatures exist; a previous, a current and a subsequent mirror
temperature. At the
beginning of an orbit there are usually missing scans and very rarely a
complete superswath. Consequently one or two of the counts were missing
and were assigned a value of zero. The algorithm that converts from
counts to temperature was bypassing code based on the count being a
missing value (not zero). The algorithm was executed in some cases
where some of the counts were 0 causing erroneous secondary mirror

3.) NOAA-M updates - a.)Change HIRS calibration method for first 24
after launch so that raw coefficients are produced instead of using a
previous satellite's data - no impact to user.
b.) Update the HIRS encoder position table to correct the slew values
NOAA-16 and make it easy to change the values per satellite. - no
impact to
c.) Put the HIRS 63rd element header in the right position and assign
HIRS scan quality flags after the 63rd element is properly aligned. -
possible user impact

4.) Change NOAA-15/16 preprocessor to correctly select the primary CPU
- no
impact to user

Recent problem seen in AMSU-B data related to moon glint correction:
Under the current algorithm for detecting moon in the space view, we
noticed that during the peak of the moon occurrence that three out of
four space view samples per scan are showing moon glint contamination.
this occurs for more than four continuous scans, no calibration
are computed for that channel. Channel 16 is affected the most, 17 to a
lesser extent, and we haven't noticed a loss of data in 18, 19, or 20.
We've noted that space view (spot) 1 of the 4 appears unaffected as
to scans of surrounding good data during this anomaly. We are open to
suggestions for possible fixes. One possibility could be to use scan
1 when the other 3 are out of tolerance. The anomaly must be
further. One example of the anomaly was seen in the NOAA-16 orbit
scans 1325 (23:13:09Z) through 1334 (23:13:33Z), JDAY 121.

Please click the number of slides to view the report on this issue.
1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21 

Name: vince_rpt.ppt
Type: Microsoft PowerPoint Show
vince_rpt.ppt (application/
Encoding: base64