SCIAMACHY: The new Level 0-1 Processor - elib.dlr.de · Gunter Lichtenberg¨ 1 Sander Slijkhuis1...

1
SCIAMACHY: The new Level 0-1 Processor unter Lichtenberg 1 Sander Slijkhuis 1 Bernd Aberle 1 Denis Scherbakov 1 Markus Meringer 1 Stefan No¨ el 2 Klaus Bramstedt 2 Patricia Liebing 2 Heinrich Bovensmann 2 Ralph Snel 3 Mathijs Krijger 3 Richard van Hees 3 Pieter van der Meer 3 Christophe Lerot 4 Angelika Dehn 4 Thorsten Fehr 5 1 Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR), Oberpfaffenhofen, Wessling, Germany 2 Institute of Environmental Physics (IUP), University of Bremen, Bremen, Germany 3 SRON, Netherlands Institute for Space Research Utrecht 4 Belgian Institute for Space Aeronomy (IASB-BIRA), Brussels, Belgium 5 ESA-ESRIN, Via Galileo Galilei, Casella Postale 64, Frascati, Rome, Italy 6 ESA-ESTEC, Keplerlaan 1, Noordwijk, The Netherlands Introduction Level 0-1 processing provides calibrated radiances. The calibration of the data is done with a combination of on-ground thermal vacuum and ambient measurements and in-flight measurements. Currently version 9 of the processor is implemented. The Level 1 products with the updates shown here will be released in 2017 after thouroughly testing the changes on the basis of retrieval results. For details of the new Level 2 processor see ATMO-37 - New Developments in the SCIAMACHY L2 Ground Processor Updated Degradation Correction The degradation correction is based on a scan mirror model that fits the thickness of con- taminant layers. The scan mirror model was first imple- mented in Version 8 of the processor. In version 9 sev- eral improvements were added that lead to a better correc- tion of the optical bench degra- dation. Together with the re- calculated polarisation and ra- diometric calibration data from on-ground measurements it is expected that the calibrated ra- diances will be improved over the whole mission life time. Details Poster ATMO-26 - Mir- ror Contamination in Space: On-ground SCIAMACHY mod- elling Figure 1: Top: Measured and mod- eled degradation of SCIAMACHY at 350 nm for 4 different lightpaths: Subsolar, limb, extra mirror and dif- fuser, from left to right over 9 years of data. Bottom: Ratio of measured and modelled m-factors at 350 nm. (SRON) Better Pointing Correction I Limb and occultation retrievals need very accurate pointing information for proper tangent heights. I First update of mispointing parameters was already applied in 2007. I IUP fitted new correction parameters by adjusting calculated solar and lunar positions to measured ones. I For the first time, lunar measurements have been utilized for this purpose. Details at poster ATMO-30, Bramstedt et. al. -10 -5 0 5 10 15 -800 -400 0 400 01-Jan-03 01-Jan-04 01-Jan-05 01-Jan-06 01-Jan-07 01-Jan-08 01-Jan-09 01-Jan-10 01-Jan-11 01-Jan-12 dλ [mdeg] dTH [m] lunar solar subsolar Elev. offset: OLD Elev. offset: OLD -10 -5 0 5 10 15 10000 20000 30000 40000 50000 -800 -400 0 400 dλ [mdeg] dTH [m] #orbit lunar solar subsolar Elev. offset: NEW Elev. offset: NEW 0 40 80 120 -6000 -4000 -2000 0 2000 01-Jan-03 01-Jan-04 01-Jan-05 01-Jan-06 01-Jan-07 01-Jan-08 01-Jan-09 01-Jan-10 01-Jan-11 01-Jan-12 dϕ [mdeg] horiz. offset [m] lunar solar subsolar Azim. offset: OLD Azim. offset: OLD 0 40 80 120 10000 20000 30000 40000 50000 -6000 -4000 -2000 0 2000 dϕ [mdeg] dTH [m] #orbit lunar solar subsolar Azim. offset: NEW Azim. offset: NEW Figure 2: Elevation and azimuth angle offsets. OLD: Mispointing with the current processor. NEW: Remaining mispointing with the improved pointing parameters (IUP). Improved Polarisation Correction The polarisation correction for V. 9 of the processor will be improved by implementing several additions and revisions to the polarisation correction: I Implementation of the CHEOPS correction already used in GOME-1 for a better correction in the UV I Addition of a retarder matrix to the scan mirror model to consistently handle the phase shift that is caused by the OBM of SCIAMACHY I Better handling of anomalous PMD data (spike filtering and electronic delays) I Better handling of cases where U cannot be measured because of insufficient instrument sensitivity I Complete Revision of the Limb polarisation calculation SS q 0.7 - 0.6 - 0.5 - 0.4 - 0.3 - 0.2 - 0.1 - 0 0.1 0.2 0.3 P q 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 0.4 0 5 10 15 20 25 30 35 40 45 SCIAMACHY q (PMD 1, RO 0, TH #10), 2007/08 current algorithm SS q 0.7 - 0.6 - 0.5 - 0.4 - 0.3 - 0.2 - 0.1 - 0 0.1 0.2 0.3 P q 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 0.4 0 10 20 30 40 50 60 SCIAMACHY q (PMD 1, RO 0, TH #10), 2007/08 new algorithm SS u 1 - 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 P u 1 - 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 0 20 40 60 80 100 120 SCIAMACHY u (PMD 1, RO 0, TH #10), 2007/08 SS u 1 - 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 P u 1 - 0.8 - 0.6 - 0.4 - 0.2 - 0 0.2 0 10 20 30 40 50 60 70 SCIAMACHY u (PMD 1, RO 0, TH #10), 2007/08 Figure 3: Plot of determined Limb polarisation values q (top), u (bottom) vs single scattering values for the old (left) and new (right) algorithm (IUP). Improvements to SWIR Channel Calibration I Bad Pixel Mask based on Individual Pixels: The new algorithm analyses each pixel individually over time to decide if it still useful I Better Dark Correction I Spectral Calibration I SRON derived new polynomial coefficients for the calibration of channel 8 I For channel 6 (Methane) a new adaptive scheme for spectral calibration is investigated. Figure 4: Prediction and measurement of the orbital variation of the dark signal in channel 8 (SRON). New Level 1 Product Format I Both products, Level 1 and Level 2 will be transferred from the current ENVISAT format to netCDF V4 format I Guidelines: I Product structure will be as far as possible similar to those developed for Sentinels (SCIAMACHY is more complicated though) I All information in the current product will also be in the new products I Level 1 will additionally contain all relevant operations information for LTDP I The geolocation will be available for all pixels (the ENVISAT format required users to caclulate the coordinates from the provided fine grid) Further Information I Level 0-1c ATBD: http://atmos.caf.dlr.de/sciamachy/documents/level 0 1b/scia01b atbd master.pdf (will soon be updated for Version 9) I Gottwald, Manfred, and Heinrich Bovensmann, eds. SCIAMACHY – Exploring the changing Earth’s Atmosphere. Springer Science & Business Media, 2010.

Transcript of SCIAMACHY: The new Level 0-1 Processor - elib.dlr.de · Gunter Lichtenberg¨ 1 Sander Slijkhuis1...

Page 1: SCIAMACHY: The new Level 0-1 Processor - elib.dlr.de · Gunter Lichtenberg¨ 1 Sander Slijkhuis1 Bernd Aberle1 Denis Scherbakov1 Markus Meringer1 Stefan Noel ... (IUP), University

SCIAMACHY: The new Level 0-1 ProcessorGunter Lichtenberg1 Sander Slijkhuis1 Bernd Aberle1 Denis Scherbakov1 Markus Meringer1 Stefan Noel2 Klaus

Bramstedt2 Patricia Liebing 2 Heinrich Bovensmann2 Ralph Snel3 Mathijs Krijger3 Richard van Hees3 Pieter van derMeer3 Christophe Lerot4 Angelika Dehn4 Thorsten Fehr5

1Remote Sensing Technology Institute (IMF), German Aerospace Center (DLR), Oberpfaffenhofen, Wessling, Germany2Institute of Environmental Physics (IUP), University of Bremen, Bremen, Germany

3SRON, Netherlands Institute for Space Research Utrecht 4Belgian Institute for Space Aeronomy (IASB-BIRA), Brussels, Belgium5ESA-ESRIN, Via Galileo Galilei, Casella Postale 64, Frascati, Rome, Italy

6ESA-ESTEC, Keplerlaan 1, Noordwijk, The Netherlands

Introduction

Level 0-1 processing provides calibrated radiances. The calibration ofthe data is done with a combination of on-ground thermal vacuum andambient measurements and in-flight measurements. Currently version9 of the processor is implemented. The Level 1 products with theupdates shown here will be released in 2017 after thouroughly testingthe changes on the basis of retrieval results. For details of the newLevel 2 processor see ATMO-37 - New Developments in theSCIAMACHY L2 Ground Processor

Updated Degradation Correction

The degradation correction isbased on a scan mirror modelthat fits the thickness of con-taminant layers. The scanmirror model was first imple-mented in Version 8 of theprocessor. In version 9 sev-eral improvements were addedthat lead to a better correc-tion of the optical bench degra-dation. Together with the re-calculated polarisation and ra-diometric calibration data fromon-ground measurements it isexpected that the calibrated ra-diances will be improved overthe whole mission life time.Details Poster ATMO-26 - Mir-ror Contamination in Space:On-ground SCIAMACHY mod-elling

Figure 1: Top: Measured and mod-eled degradation of SCIAMACHY at350 nm for 4 different lightpaths:Subsolar, limb, extra mirror and dif-fuser, from left to right over 9 yearsof data. Bottom: Ratio of measuredand modelled m-factors at 350 nm.(SRON)

Better Pointing Correction

I Limb and occultation retrievals need very accurate pointing informationfor proper tangent heights.

I First update of mispointing parameters was already applied in 2007.I IUP fitted new correction parameters by adjusting calculated solar and

lunar positions to measured ones.I For the first time, lunar measurements have been utilized for this

purpose. Details at poster ATMO-30, Bramstedt et. al.

-10

-5

0

5

10

15 -800

-400

0

400

01-Jan-03 01-Jan-04 01-Jan-05 01-Jan-06 01-Jan-07 01-Jan-08 01-Jan-09 01-Jan-10 01-Jan-11 01-Jan-12

dλ [m

deg]

∼ d

TH

[m

]

lunar solar subsolarElev. offset: OLDElev. offset: OLD

-10

-5

0

5

10

15

10000 20000 30000 40000 50000

-800

-400

0

400

dλ [m

deg]

∼ d

TH

[m

]

#orbit

lunar solar subsolarElev. offset: NEWElev. offset: NEW

0

40

80

120-6000

-4000

-2000

0

2000

01-Jan-03 01-Jan-04 01-Jan-05 01-Jan-06 01-Jan-07 01-Jan-08 01-Jan-09 01-Jan-10 01-Jan-11 01-Jan-12

[m

deg]

∼ h

oriz. offset [m

]

lunar solar subsolarAzim. offset: OLDAzim. offset: OLD

0

40

80

120

10000 20000 30000 40000 50000

-6000

-4000

-2000

0

2000

[m

deg]

∼ d

TH

[m

]

#orbit

lunar solar subsolarAzim. offset: NEWAzim. offset: NEW

Figure 2: Elevation and azimuth angle offsets. OLD: Mispointing withthe current processor. NEW: Remaining mispointing with the improved

pointing parameters (IUP).

Improved Polarisation Correction

The polarisation correction for V. 9of the processor will be improvedby implementing several additionsand revisions to the polarisationcorrection:I Implementation of the CHEOPS

correction already used in GOME-1 fora better correction in the UV

I Addition of a retarder matrix to thescan mirror model to consistentlyhandle the phase shift that is causedby the OBM of SCIAMACHY

I Better handling of anomalous PMDdata (spike filtering and electronicdelays)

I Better handling of cases where Ucannot be measured because ofinsufficient instrument sensitivity

I Complete Revision of the Limbpolarisation calculation

SS q

0.7− 0.6− 0.5− 0.4− 0.3− 0.2− 0.1− 0 0.1 0.2 0.3

P q

0.8−

0.6−

0.4−

0.2−

0

0.2

0.4

0

5

10

15

20

25

30

35

40

45

SCIAMACHY q (PMD 1, RO 0, TH #10), 2007/08

current algorithm

SS q

0.7− 0.6− 0.5− 0.4− 0.3− 0.2− 0.1− 0 0.1 0.2 0.3

P q

0.8−

0.6−

0.4−

0.2−

0

0.2

0.4

0

10

20

30

40

50

60

SCIAMACHY q (PMD 1, RO 0, TH #10), 2007/08

new algorithm

SSu1− 0.8− 0.6− 0.4− 0.2− 0 0.2

P u

1−

0.8−

0.6−

0.4−

0.2−

0

0.2

0

20

40

60

80

100

120

SCIAMACHY u (PMD 1, RO 0, TH #10), 2007/08

SSu1− 0.8− 0.6− 0.4− 0.2− 0 0.2

P u

1−

0.8−

0.6−

0.4−

0.2−

0

0.2

0

10

20

30

40

50

60

70

SCIAMACHY u (PMD 1, RO 0, TH #10), 2007/08

Figure 3: Plot of determinedLimb polarisation values q (top),u (bottom) vs single scatteringvalues for the old (left) and new(right) algorithm (IUP).

Improvements to SWIR Channel Calibration

I Bad Pixel Mask based on Individual Pixels: The new algorithmanalyses each pixel individually over time to decide if it still useful

I Better Dark CorrectionI Spectral Calibration

I SRON derived new polynomial coefficients for the calibration of channel 8I For channel 6 (Methane) a new adaptive scheme for spectral calibration is

investigated.

23627.0 23627.5 23628.0 23628.5 23629.0 23629.5 23630.0Orbit number

5110

5120

5130

5140

5150

Dark current (BU/s)

Monthly orbit 23628 predicted by vardark data of monthly orbit 23213, pixel 621, ch8

Figure 4: Prediction and measurement of the orbital variation of thedark signal in channel 8 (SRON).

New Level 1 Product Format

I Both products, Level 1 and Level 2 will be transferred from the currentENVISAT format to netCDF V4 format

I Guidelines:I Product structure will be as far as possible similar to those developed for Sentinels

(SCIAMACHY is more complicated though)I All information in the current product will also be in the new productsI Level 1 will additionally contain all relevant operations information for LTDPI The geolocation will be available for all pixels (the ENVISAT format required users

to caclulate the coordinates from the provided fine grid)

Further Information

I Level 0-1c ATBD: http://atmos.caf.dlr.de/sciamachy/documents/level 0 1b/scia01b atbd master.pdf (will soon be updated for Version 9)I Gottwald, Manfred, and Heinrich Bovensmann, eds. SCIAMACHY – Exploring the changing Earth’s Atmosphere. Springer Science & Business Media, 2010.