Multistatic Observations of Surface Wind and Current Vectors with STEREOID. Paco López-Dekker and others

Transcription

1 Multistatic Observations of Surface Wind and Current Vectors with STEREOID Paco López-Dekker and others 1

2 STEREOID for Earth Explorer 10 Lead Investigator Lopez-Dekker, Paco, Associate Professor LeadInvestigator Proposing Team Name 1 Chapron, Bertrand, Dr. 2 Gommenginger, Christine, Dr. 3 Johnsen, Harald, Dr. 4 Collard, Fabrice, Dr. 5 Hansen, Morten, Dr. Delft University of Technology Department of Geoscience and Remote Sensing Afliation Lopez-Dekker, Paco, Associate Professor Ifremer ProposingTeam 6 Korosov, Anton, Dr Name Rott, Helmut, Dr. Nagler, Thomas, Dr. Zeeuw Chapron, van Dalfsen, Elske, Dr. 1 De Bertrand, Dr. Hooper, Andy, Professor Hanssen, Ramon, Professor National Oceanography Center NORUT Oceandatalab Nansen Environmental and Remote Sensing Center (NERSC) Nansen Environmental and Remote Sensing Center (NERSC) ENVEO IT ENVEO IT KNMI University of Leeds Delft University of Technology US Geological Survey Delft University of Technology University of Twente Politecnico di Milano NORUT DLR Airbus Defence and Space, Spain Cosine Mike, Dr. 2 Poland, Gommenginger, Christine, Dr. Menenti, Massimo, Professor Su, Zu (Bob) Prof. Stefano, Assistant Professor 3 Tebaldini, Johnsen, Harald, Dr. Larsen, Yngvar, Dr. Prats-Iraola, Pau, Dr. Ing 4 Zurita, Albert, Collard, Fabrice, Dr Esposito, Marco Hansen, Morten, Dr. Korosov, Anton, Dr. Rott, Helmut, Dr. Nagler, Thomas, Dr. Country FR GB NO FR NO NO AT AT NL GB NL US NL NL IT NO DE ES NL Delft Universit Department o Affiliation Ifremer National Ocea NORUT Oceandatalab Nansen Enviro Nansen Enviro ENVEO IT ENVEO IT 2

3 Mission architecture 3

4 Sentinel-1 D STEREOID-A STEREOID-B > 250 km 4

5 Mission objectives (1 min overview) Solid Earth 3D surface deformation (volcanic, seismic, landslides) Sudden topographic changes Cryosphere Glacier and ice sheets topography/volume/mas s change High resolution ice fows/ deformation Sea ice drift and topography Marginal Ice Zone variability Oceans High resolution surface currents and wave data for coastal processes Small-scale (100 m to 10 km) ocean dynamics Surface deformation feld () Divergence/strain, vorticity, shear Extreme weather events 5

6 Flight configurations Sentinel-1 D STEREOID-A STEREOID-B > 250 km Stereo formation Maximum line-of-sight diversity Best for surface current vectors and 3-D surface deformation 6

7 Flight configurations Sentinel-1 D STEREOID-A STEREOID-B > 250 km XTI formation Close-formation (TanDEM-X style) Intended for DEM time-series 400 m to 1 km baselines ATI formation 100 m to 200 m along-track separation 7

8 Radial velocities measured by Sentinel-1 8

9 And attempt to contextualize STEREOID (TSCV) Extremes? SKIM I am STEREOID and I take snapshots STEREOID SEASTAR 9

10 Space-segment Platform < 500 kg Dual launch in Vega not an issue Primary payload: radar receiver Split antenna concept On-ground Digital Beamforming or ATI Secondary payload: VNIR + TIR For proposal we assumed Cosine s HyperScout (NL) 10

11 TIR + radar rationale 10 km 11

12 Imaging performance: NESZ (IW mode) sub-swath 1 sub-swath 2 sub-swath 3 NESZ generally adequate to good Probably a bit less gain in exchange of wider elevation beams would be better. Or SCORE + 3dB for ATI mode Position in burst 12

13 Imaging performance: AASR (IW mode) sub-swath 1 sub-swath 2 sub-swath 3 AASR in -17 to 18 db range. Quite good given small total antenna area Sub-swath variability due to Sentine-1 PRFs Position in burst 13

14 Imaging performance: RASR RASR < -20 db Good, but we need to accommodate large dynamic ranges (varying wind conditions) 14

15 Squint angles (top view) ^ ^ ^ 1 ^ Efective squint angle limited by bistatic geometry 15

16 Squint angles 250 km separation 350 km separation 16

17 System Sensitivity (ideal retrieval) Doppler velocity uncertainty 3 km resolution 6 m/s wind 250 km separation TSCV vector uncertainty 17

18 System Sensitivity (ideal retrieval) Doppler velocity uncertainty 3 km resolution 6 m/s wind 350 km separation TSCV vector uncertainty 18

19 System Sensitivity (ideal retrieval) Doppler velocity uncertainty 3 km resolution 6 m/s wind 350 km separation TSCV vector uncertainty If I can have three antennas 19

20 Polarimetry Monostatic Bistatic H Rx Polarization Rx Polarization H At 45 squint geometry V polarized Tx should we equivalent to 45 polarized monostatic Tx V V V Rx Polarization H V Rx Polarization H 21

21 Polarimetry H H Rx Polarization Bistatic Rx Polarization Monostatic V V V Rx Polarization H V Rx Polarization H 22

22 Main sources of systematic (non-geophysical) errors DCA Sentinel-1 pointing Companion mispointing Short-ATI Mispointings weighted by 1/beamwidths Formation knowledge Oscillator frequency ofsets Long-ATI Leads to ATI phase ofset Is a point of concern, but seems technically solved But emphasis on gradients We can mostly live with low-pass systematic errors 23

23 Outlook, status Phase-0 science and system studies and campaigns currently G Call for MA G under preparation to be kicked-off in ow n s r e b m e m No technical show-stoppers up to know. open!!! Exciting and challenging science. A true Earth Explorer, serving multiple communities Exploitation Phase-BCDE Phase-A Phase-0 24

24 Extremely challenging Product resolution (averaging) Instrument and/or measurement calibration Geophysical noise Radiometric resolution Measurement noise Instrument related systematic errors Total error Limiting factor??? Geophysical biases Inversion 25

25 26

26 The value of resolution (for example, vorticity) 12 km grid 4 km grid 2 km grid 28

27 Main open issue: (wind) wave bias Problem: Approaches being studied Measure waves estimate bias Estimate wind from backscatter model waves estimate bias Exploit polarimetric dependency of wave-bias 29

28 Extreme weather [Stereo] cross-pol Stereo confgurati on C-band Surface winds + TSCV 30

29 Main open issue: (wind) wave bias Problem: TanDEM-X experimental dual-beam Doppler feld 31

30 ...but there are possible solutions Measured Doppler feld Wind to Doppler mapping using Asymptotic model Estimated wind 32

31 TSCV retrieval Measured Doppler feld - wind-doppler Estimated current 33

32 Space-segment side challenges Formation flying Formation safety Cross-track baseline knowledge Common Doppler and alongtrack baseline control Synchronization Phase Echo window SAR performance Sensitivity Ambiguities 34

33 Formation flying: cross-track baseline knowledge LOS baseline error translates directly into phase error: Baseline rotation R 800 km 200 m to 400 m Target 10 cm Baseline Baseline knowledge knowledge requirement requirement from from O(0.1 O(0.1 mm)) mm)) 35

34 Common Doppler loss Δ Dop= 0 Should be small compared to processed Doppler bandwidth. 0 TOPS SESAME M MEE A A S S E E S FFoorr S 0 m m)) ( 3 O O( attiioonn ra sseeppaar K O OK 36

35 Space-segment side challenges Formation flying Formation safety Cross-track baseline knowledge Common Doppler and alongtrack baseline control Synchronization Phase Echo window SAR performance Sensitivity Ambiguities 37

36 Phase synchronization Frequency and phase synchronization always a critical issue. From TanDEM-X SyncLink 38

37 Carrier frequency and phase synchronization: lessons we think we have learnt GPS tagging/disciplining Frequency offsets Data driven (AutoSync, etc) Relative phase errors Often good enough Issues for sure Explicit synchronization link Two way synchronization between receivers needed 39

38 Echo window synchronization GPS Instrume nt time position reproducible Radar timing echo delay well known Instrume nt time GPS PRF signal time Radar timing position?? reproducible Situation ok if radar timing derived systematically from GPS-referenced instrument time Position derived timing potential trouble maker. 40

39 Technical challenges Formation flying Formation safety Cross-track baseline knowledge Common Doppler and alongtrack baseline control Synchronization Phase Echo window SAR performance Sensitivity Ambiguities 41

40 Imaging performance: NESZ (Imaging mode) sub-swath 1 sub-swath 2 sub-swath 3 NESZ generally adequate to good Probably a bit less gain in exchange of wider elevation beams would be better. Or SCORE + 3dB for ATI mode Position in burst 42

41 Imaging performance: AASR (imaging mode) sub-swath 1 sub-swath 2 sub-swath 3 AASR in -17 to 18 db range. Quite good given small total antenna area Sub-swath variability due to Sentine-1 PRFs Position in burst 43

42 Sentinel-1 D STEREOID-A SAR ante nna SAR antenna > 250 km Hyper Scout HyperScout 44