independent verification of the sentinel-1 system calibration · 2019. 2. 14. · • shape of the...
TRANSCRIPT
www.DLR.de/HR • VG 1 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt> www.DLR.de/HR • VG 1 <Title> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Independent Verification of the Sentinel-1 System Calibration
Marco Schwerdt, Kersten Schmidt, Gabriel Castellanos, Nuria Tous Ramon, Björn Döring, Pau Prats, Manfred Zink ASAR/CEOS Workshop 2013
www.DLR.de/HR • VG 2 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Challenge to calibrate Sentinel-1
Strategy / Recommendations
• Successful Execution of all Calibration Activities
• Delivery of Calibrated SAR Data Products as soon as possible
DLR’s Calibration Scenario for Sentinel-1 • S1-coverage across the DLR’s calibration field
• Calibration Procedures (Absolute Radiometric Calibration, Antenna Pointing, Geometric Calibration, Antenna Model Verification, Internal Calibration)
In-Orbit Calibration Plan
End-to-End Radiometric Calibration Budget
Conclusion
Overview
www.DLR.de/HR • VG 3 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
• 4 different modes operated at least over 10 Years
• Dual polarization
• Wide range of swath positions (18°-47°)
• Tight performance in all Modes
1.0dB (3σ)
Commissioning Phase 3 Months
Efficient Calibration
Strategy
Chart by Astrium GmbH (Germany)
Challenge to calibrate Sentinel-1
www.DLR.de/HR • VG 4 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Internal Calibration Facility • compensate for drift effects by internal calibration pulses • derive actual settings of the TRMs by pulse coded technique (PCC) for
tuning/optimizing the antenna model
Antenna Model • shift most of the antenna characterization from the CP to pre-launch activities
=> shift the effort from space to ground
• provide all reference patterns for radiometric correction of the SAR data • derive the actual pointing of the antenna
Goal • reduce the calibration effort during the commissioning phase
• ensure stable and reliable operation of a precise SAR system over a period > 10 years
only a few selected beams have to be really measured
• short duration of 3 months CP • reasonable effort for long term system
monitoring
Calibration Strategy I
www.DLR.de/HR • VG 5 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Relative radiometric calibration of all SAR data products has to be already performed by applying:
• internal calibration for drift compensation
• shape of the antenna patterns
• gain offset between different beams
Then, absolute radiometric calibration can be performed by measuring the whole Sentinel-1 system against reference targets independent of both:
• the target position within the swath and
• the beam and mode being operated.
Thus, one absolute calibration factor is valid for all operation modes
Minimum number of measurements required against reference targets is defined by the end-to-end system calibration budget:
• worst case parameters across all modes are applied because measurements are performed independent of beam and mode
Hence, the end-to-end system calibration budget for one specific mode will be better because not all worst cases are combined by one mode.
derived by the antenna model
Calibration Strategy II
www.DLR.de/HR • VG 6 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
High Radiometric Accuracy 1dB (3σ) • Measuring at least one beam of each mode
• Against 3 reference targets deployed within the swath
• Measuring at least one beam with low, one with mid and one with high incidence angle
• Measuring at least one beam in both transmit polarisations
• Absolute radiometric reference knowledge by 2 different types of reference targets (transponder and corner reflector)
• Execute equal number of measurements for asc and des passes
Tight Schedule of 3 months CP • Co-/cross polar receiving channels should be measured simultaneously
• Test site within crossover area of ascending and descending swathes
Recommendations/Rules for Beam & Site Selection
Which beams are suitable?
www.DLR.de/HR • VG 7 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
SM2 asc Day 10
SM1 des Day 5
• 2 x low • 2 x mid • 2 x high
IW1 des Day 12
IW3 asc Day 8
SM5 des Day 7
EW3 asc Day 3
WV1 des Day 5
incidence angle
DLR Oberpfaffenhofen
suitable Test Site for all
External Calibration Measurements
S1A Coverage across DLR Cal-Field
• Selected beams
• Asc and des passes for low, mid and high incidence angles within a repeat cycle
SM1, SM2, IW1, EW3, SM5, IW3 SM1 SM1, SM2 SM1, SM2, IW1 SM1, SM2, IW1, EW3 SM1, SM2, IW1, EW3, SM5 SM1, SM2, IW1, EW3, SM5, IW3
www.DLR.de/HR • VG 8 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
SM2 asc Day 10
SM1 des Day 5
• 2 x low • 2 x mid • 2 x high
IW1 des Day 12
IW3 asc Day 8
SM5 des Day 7
EW3 asc Day 3
WV1 des Day 5
WV1 asc Day 10
incidence angle
• Selected beams
• Ascending and descending passes for low, mid and high incidence angles
• Transponder test site D39, D40 and D41 is driven by modes enabling the co- and cross polar channel
SM1 SM1, SM2 SM1, SM2, IW1 SM1, SM2, IW1, EW3 SM1, SM2, IW1, EW3, SM5 SM1, SM2, IW1, EW3, SM5, IW3
D39 x D40
x D41
x DLR
Oberpfaffenhofen
S1A Coverage across DLR Cal-Field
www.DLR.de/HR • VG 9 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
SM2 asc Day 10
SM1 des Day 5
• 2 x low • 2 x mid • 2 x high
IW1 des Day 12
IW3 asc Day 8
SM5 des Day 7
EW3 asc Day 3
85 km
D38 x
WV1 des Day 5
WV1 asc Day 10
incidence angle
• Selected beams
• Ascending and descending passes for low, mid and high incidence angles
• Transponder test site D39, D40 and D41 is driven by modes enabling the co- and cross polar channel
SM1 SM1, SM2 SM1, SM2, IW1 SM1, SM2, IW1, EW3 SM1, SM2, IW1, EW3, SM5 SM1, SM2, IW1, EW3, SM5, IW3 SM1, SM2, IW1, EW3, SM5, IW3 and WV1
D39 x D40
x D41
x
at least 4 Targets per Pass
• D39, D40, D41 Transponder • D38, D42, D43 Corner Reflector
DLR Oberpfaffenhofen
S1A Coverage across DLR Cal-Field
D43 x D42
x
www.DLR.de/HR • VG 10 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Schwerdt, Folie 10
Beam / Mode
Day within Repeat Cycle Test Site
4.21 day (des) D38 - D41
9.68 day (asc) D38 - D40
SM2 9.68 day (asc) D39 - D43
SM3 11.20 day (des) D39 - D43
SM4 2.68 day (asc) D40 - D43
SM5 6.20 day (des) D39 - D42
6.20. day (des) D38 - D40
7.69 day (asc) D38 - D42SM6
SM1
Beam / Mode
Day within Repeat Cycle Test Site
4.21 day (des) D38 - D41
9.68 day (asc) D38 - D43
EW2 11.20 day (des) D40 - D43
EW3 2.68 day (asc) D39 - D43
EW4 6.20 day (des) D39 - D42
EW5 7.69 day (asc) D40 - D42
4.21 day (des) D38 - D39
9.68 day (asc) D39 - D40
WV2 2.68 day (asc) D41 - D42
EW1
WV1 Black ≥ 4 Targets per Pass Blue = 3 Targets per Pass (except for WV 2 Targets)
• 2.68 day (asc) IW1, IW2, SM4, EW3, WV2
• 4.21 day (des) SM1, EW1, WV1
• 6.20 day (des) IW2, IW3, SM5, SM6, EW4
• 7.69 day (asc) IW3, SM6, EW5
• 9.68 day (asc) SM1, SM2, EW1, WV1
• 11.20 day (des) IW1, SM3, EW2
Beam / Mode
Day within Repeat Cycle Test Site
11.20 day (des) D38 - D42
2.68 day (asc) D38 - D40
6.20. day (des) D41 - D43
2.68 day (asc) D41 - D43
6.20. day (des) D38 - D40
7.69 day (asc) D38 - D41
IW1
IW3
IW2
6 Passes within
1 Repeat Cycle
S1A-Coverage over DLR Cal-Field Ref Orbit S1A_TEST_MPL_ORBRES_20130814T120000_20130827T120000_1004.EOF
www.DLR.de/HR • VG 11 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Pre- Phase
Launch T 0 T 4 T 3
Radiometric Calibration
Antenna Model Verification
Antenna Pointing
Rainforest Cal Targets
GeoCal
T 2 T 1
End of CP
Internal Calibration
T 5 1. 2. 3. 4. 5. 6. 7.
6.5 Repeat Cycles => 2.5 Months
In-Orbit Calibration Plan
www.DLR.de/HR • VG 12 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Absolute Radiometric Calibration
Beam / Mode 4. Cycle 5. Cycle 6. cycle
11.20 day (des) 11.20 day (des) 11.20 day (des)
- -
- -
7.69 day (asc) 7.69 day (asc) 7.69 day (asc)
SM1 4.21 day (des) 4.21 day (des) -
SM2 9.68 day (asc) 9.68 day (asc) -
SM5 6.20. day (des) 6.20. day (des) 6.20. day (des)
EW3 2.68 day (asc) 2.68 day (asc) 2.68 day (asc)
- - 4.21 day (des)
- - 9.68 day (asc)WM1
IW1
IW3
7 beams (IW1, IW3, SM1, SM2, SM5, EW3) within 1 repeat cycle
2 passes per beam
Equal number of passes for asc/des
4 targets for IW3 (asc), SM1 (des) 5 targets for IW1 (des), SM2 (asc), SM5 (des), EW3 (asc) 2 targets for WV1 (asc/des)
At least one beam in both transmit polarisations modes (grey boxes)
=> within 3 repeat cycles 100 measurements
sufficiently to be compliant with the end-to-end system
calibration budget in all modes (see VG 18)
www.DLR.de/HR • VG 13 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Schwerdt, Folie 13
Rainforest
2-way elevation pattern with notch on receive
different orbits (attitude control check)
Ground Receiver Unit
1-way notch-pattern in azimuth
2 acquisitions at low, mid and high incidence angle
1 notch-beam in second polarisation
lowANP1
ANP2H
D38 - D41
D39 - D43
4.21 (des)
9.68 (asc)
mid ANP3 H D39 - D43 11.20 (des)
mid ANP4 V D40 - D43 2.68 (asc)
highANP5
ANP6H
D39 - D42
D38 - D42
6.20 (des)
7.69 (asc)
not critical with respect to the schedule
• 2 passes (asc/des) for low, mid and high inc. angle (3)
• 3 ground receiver units per pass
18 measurements within 1 repeat cycle
Antenna Pointing
www.DLR.de/HR • VG 14 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Schwerdt, Folie 14
Rainforest (Elevation Pattern)
2-way pattern, 4 acquisitions per selected beam
3 StripMap beams (low, mid, high inc. angle)
Gain offset between different beams by IW / EW
not critical with respect to the schedule
• 2 passes (asc/des) for low, mid, high inc. angle
• 2-3 ground receiver units per pass
• also applicable for absolute Rad-Cal
16 measurements per Pol within 2 repeat cycle
Ground Receiver Unit (Azimuth Pattern)
1-way azimuth pattern, 2 acquisitions per Pol at
low, mid, high inc. angle (S1/S2, IW1, IW3)
For mid inc. angle (IW1) both transmit polarisations
IncidenceAngle
Beam Test Sites 2. Cycle (H-Pol) 3. Cycle (V-Pol)
lowSM1
SM2
D38 - D41
D39 - D43
4.21 (des)
9.68 (asc)
4.21 (des)
9.68 (asc)
mid IW1D38 - D42
D38 - D40
11.20 (des) (1. cycle)
2.68 (asc) (2. cycle)
11.20 (des) (1. cycle)
2.68 (asc) (2. cycle)
high IW3D38 - D40
D38 - D41
6.20 (des)
7.69 (asc)
6.20 (des)
7.69 (asc)
Geometric Calibration
can be performed simultaneously
Imaging mode
Antenna Model Verification
www.DLR.de/HR • VG 15 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Next Generation of DLR’s Reference Targets
designed for Sentinel-1
Remote controlled
Corner Reflector
Remote controlled Transponder
www.DLR.de/HR • VG 16 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Complete electronic is covered by the housing including also the antennas
Remote controlled
Coherent chirp detection and flexible backscatter matrix (HH, VV, HV, VH)
Center Frequency 5.405 GHz
Bandwidth 100 MHz
RCS 60 dBm²
Radiometric Stability ≤ 0.1 dB (1σ)
Absolute Radiometric Accuracy 0.2 dB (1σ)
DLR’s Transponder Specification
⇒ Verified by real measurements campaigns using RADARSAT-2
⇒ Thanks to MDA for supporting our activities
www.DLR.de/HR • VG 17 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
First RADARSAT-2 image of new DLR transponder
• 8 successful acquisitions using RadarSAT-2 executed in April 2013
• absolute radiometric accuracy of 0.2 dB (1σ)
www.DLR.de/HR • VG 18 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
DLR
Passes per Repeat Cycle 6
Beams for AbsCal 7
Passes per Beam (H/V) 2 / 1
Targets per Pass 4 - 5
Targets Accuracy 0.6 dB (3σ)
End-to-end System Budget S-1A (3σ)
H (co/cross) 0.94 dB / 0.97 dB
V (co/cross) 0.96 dB / 0.99 dB
Compliant with the strong requirement of 1 dB (3σ) end-to-end system budget in all operation modes
S-1 System Budget by DLR‘s Calibration Scenario
www.DLR.de/HR • VG 19 <CEOS 2013-10, S-1A CAL Scenario> • <Satellite-SAR-Systems> • <Calibration> • <Schwerdt>
Conclusion Efficient calibration strategy has been developed for Sentinel-1 based on:
• PCC method (developed by DLR and successfully applied by TerraSAR-X) providing the actual setting of individual T/R modules
• precise antenna model providing antenna patterns and beam-to-beam gain offsets
• only 1 absolute calibration factor for all operation modes
Based on different rules/recommendations a set of 7 beams (IW1, IW3, SM1, SM2, SM5, EW3, WV1) has been established for measurements against at least 4 reference targets per pass
The in-orbit calibration plan for DLR’s complementary calibration campaign with a test site of 3 transponders and 3 corner reflectors deployed in South Germany is compliant with
• the end-to-end system calibration budget in all modes: < 1.0 dB (3σ) and
• the schedule of the commissioning phase of 3 months
Next generation of the DLR’s reference targets (transponders and corner reflectors) was designed and developed for the Sentinel-1A mission