meeting date
date de la réunion / 08-08 2002 / ref/réf / PO-MN-ESA-GS-1349 / page/page / 1
4
meeting date
date de la réunion / 08-08 2002 / ref./réf. / PO-MN-ESA-GS-1349 / page/page / 1
4
meeting place
lieu de la réunion / IFE/IUP Bremen / chairman
président / Johannes Frerick
EEM-PPP
minute’s date
dates de minute / 12 August 2002 / participants
participants / SCCT (L2) group
subject/objet / SCIA-SCCT(L2) progress # 1 / copy/copie / G. Levrini, P. Snoeij
meeting date
date de la réunion / 08-08 2002 / ref/réf / PO-MN-ESA-GS-1349 / page/page / 1
4
description/description /

action/action

/

due date/date limite

The first post-launch level 2 verification meeting was helt at IFE/IUP Bremen.
The meeting addressed the following topics:
Status of level 1b verification (summary)
Verification of selected orbits – definition of verification targets
Additional tasks related to initialisation file parameter verification.
General improvements on task descriptions – how to increase the level of detail.
Place and time for next meeting
List of attachments:
A: General Level 2 verification task aspects (A. Piters)
B: Swath visualisation orbit “type” 2209
C: Swath visualisation orbit “type” 2230
D: Swath visualisation orbit “type” 2222
E: Swath visualisation orbit “type” 2258
Status of level 1b verification (summary)
Presentation by J. Frerick about level 1b status. Details can be looked up in the minutes of meeting of the last level 1b related verification team meeting. They can be found on the verification web page, accessible to all verification team members.
From a level 2 processing point of view, the situation of the level 1b product can be summarized as follows:
  1. A first in-flight sun mean reference spectrum will be available as from beginning of cw 33. It is based on an ESM diffuser state. Whether or not the SMR can be systematically updated (day by day) depends on availability of level 1b data at IECF. More systematic analysis on the quality of the different types of SMRs is still outstanding. I.e. from a level 1b point of view, no clear recommendation can be given yet, which SMR to use for which L2 fitting window.
  2. PPG/Etalon: absolute values of both, the etalon and (if necessary) the pixel to pixel gain correction, still need to be verified. Presuming stable etalon conditions (which is fairly safe due to the overall thermal stability of the instrument) this should not have much influence on level 2 products, because both corrections should cancel out in ratios of atmospheric radiance over solar reference.
  3. Spectral calibration: A first spectral calibration auxiliary file, based on in-flight spectral calibration measurements, will be available as from cw 33. Due to problems with the keydata file as well as insufficient dark current correction in channel 7 and 8), only channels 1, 3,4,5,6 will have in-flight calibration. Channel 2 as well and 7 and 8 will still use on-ground values. There will be no orbit dependent spectral calibration necessary, as deviation from an average spectral calibration is generally below 1/100 of a detector pixel.
  4. Leakage current: First in-flight calibration auxiliary fiel will be available as from cw 33. This should, for the future also improve spectral calibration results for channel 7 and 8.
  5. All atmospheric “radiometric” corrections (straylight, polarisation) still need to be verified. I.e. if they are applied on level 1b products, we cannot judge on their quality/correctness.
Verification of selected orbits – definition of verification targets
From a systematic point of view, it is highly recommendable to start rification of level 2 products for selected, scientifically usefull orbits. Especially when regarding the impact of level 1b/c quality on the level 2 product, it is essential that orbits, which have been analysed for level 1b, verification of atmospheric corrections, are analysed for level 2 accordingly. Otherwise, the interaction between the two processing levels can never be systematically addressed.
For the verification orbits a number of general criteria was worked out by the meeting. The orbits shall
  • Contain both, nadir and limb measurements
  • Contain completely cloud free pixels
  • Have a typical land/sea coverage
  • Contain good radiance targets like deserts (for the best level 1b input possible)
  • Contain a certain number of groundbased stations (or other validation sites) for independent checks.
  • Contain (if possible) data from “geo-physical” events like volcano eruptions.
Based on this criteria the following “typical” orbits have been agreed on (see annex B to E, showing the nadir ground track. The corresponding limb may be taken from the SOST homepage)
2209: Central Europe/northern Africa
2230: North America, Pacific, New Zeeland (ozone sonde)
2222: Central Africa/Congo (Volcano)
2258: Hawai (satellite overpass)
The orbits generally specify the type of orbits, we agreed to analyse. It does not have to be exactly this orbit, i.e. if e.g. the cloud coverage was not optimal, we might take orbit x+43, which has very similar ground coverage. This option certainly holds for type 2209, 2230 and 2222, whereas 2258 needs more careful considerations, as it requires overpass to Mauna Loa.
The products shall be made centrally available, i.e. put on the verification data base server in all processing levels, i.e. level 0, 1b, 2NRT and 2OL (as soon as available).
Additional tasks related to initialisation file parameter verification
Three additional verification tasks have been identified, which are related to some extent to initialisation file settings for the OL level 2 processor.
  1. To investigate correlation matrix along the orbit, aiming an improvement of initial guess parameters.
  2. Profile comparison cross and long track
  3. With some more long term character: to investigate differences in limb profiles, when using solar or atmospheric reference spectra.
For additional task 1, Kai-Uwe Eichmann has been assigned, for task no. 2 Chritain v. Savigny. Both are “new” team members. Therefore, a warm welcome to the team and may the (verification) force be with you….
General improvements on task descriptions – how to increase the level of detail
A. Piters presented a more detailed list of activities, which have to be performed in order to verify the “physical correctness” of all level 2 type data products. Details can be found in annex A.
These activities can be seen as part of the more detailed verification procedures and may be adapted, verification by verfication task.
With respect to the task entry mask still a certain uncertainty abaout what to fill into fields like pass/fail criteria and contingencies exists. The meeting concluded that for most of the level 2 related verification tasks, these criteria cannot be generalised. Consequently, we agreed to leave these fields un-filled.
Place and time for next meeting
CW 37 in parallel to the calibration work shop. / A.I.L2_01
(J. Frerick)
To upload the verification data base with level 0, 1b and 2 products of specified orbits
A.I.L2_02
(A. Piters)
To check cloud coverage and recommend target orbits for type 2209, 2230, 2222 and 2258)
A.I.L2_03
(J. Frerick)
To make L2OL product IODD (input/output data definition) available in its actual version.
A.I.L2_04
(all L2 verificator’s)
To specify in more detail the individual verification task procedures, based on annex A. / 16.08.02
13.08.02
16.08.02
16.08.02