GNSS Reflectometry: Innovative Remote Sensing

Transcription

1 GNSS Reflectometry: Innovative Remote Sensing J. Beckheinrich 1, G. Beyerle 1, S. Schön 2, H. Apel 1, M. Semmling 1, J. Wickert 1 1.GFZ, German Research Center for Geosciences, Potsdam, Germany 2.Leibniz University Hannover, Institute of Geodesy, Hannover, Germany Slide 1 WISDOM: GNSS-R flood monitoring, Jamila Beckheinrich, Vietnam 2013

2 WP within the WISDOM project GNSS Reflectometry Principle Experimental Setup First Results Outlook & Conclusion Slide 2

3 WP within the WISDOM project GNSS Reflectometry Principle Experimental Setup Preliminary Results Outlook & Conclusion Slide 3

4 WP within the WISDOM Project: Climate changes have caused severe changes in the Mekong Delta: extreme flood events important to monitor coastal area with dense population Test the possibility of using low elevation GNSS Reflectometry for flood monitoring of the Mekong Delta Develop and implement algorithms to process GPS signals into water levels Slide 4

5 WP within the WISDOM project GNSS Reflectometry Principle Experimental Setup Preliminary Results Outlook & Conclusion Slide 5

6 GNSS Reflectometry Principle: Global Navigation Satellite System GPS, GLONASS, Galileo, Beidou Space-based radar and laser altimeters: high altimetric accuracy insufficient spatial and temporal resolution Ground-based instrumentation: high temporal resolution but for a point location only GNSS Reflectometry could fill this gap. Slide 6

7 GNSS Reflectometry Principle: Main advantages of GNSS: large and increasing number of available GNSS signals signals for civilian are for free High quality signals: dual frequency, long-term availability and stability Inexpensive: passive system dense global coverage Some surfaces like water, ice or wet soil show high reflectivity for the GNSS L-Band signals. Multiple simultaneous measurements with high temporal and spatial resolution. Slide 7

8 GNSS Reflectometry Principle: GNSS signals scatter off the reflecting surface. The reflected signal has to travel a longer path relative time delay Δt phase offset Δf w.r.t. to the direct signal E direct signal 2hsin E h ρ E Reflected signal Reflecting Surface Slide 8

9 WP within the WISDOM project GNSS Reflectometry Principle Experimental Setup Preliminary Results Outlook & Conclusion Slide 9

10 Experimental Setup: Measurements in Can Tho City at the An Bihn Hotel in Vietnam March 2012 Two time series with two different antenna heights above the reflecting surface: Terrace with ~10 m Roof with ~20 m Slide 10

11 Two antennas: Experimental Setup: RHCP oriented to the zenith (direct signal) LHCP tilted to the reflecting surface (reflected signal) Direct signal L1/L2 GPS RHCP Reflected signal RHCP/ LHCP L1/L2 GPS Slide 11

12 Experimental Setup: Measurements in Can Tho City at the An Bihn Hotel in Vietnam March 2012 Two different antenna heights above the reflecting surface: Terrace with 10 m Roof with 20 m Leveling staff Every 30 min. Slide 12

13 Experimental setup: Water level changes Slide 10

14 WP within the WISDOM project GNSS Reflectometry Principle Experimental Setup First Results Outlook & Conclusion Slide 14

15 First results: Phase Observations Phase Model - Cycle slip detection Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 15

16 First results: Phase Observations Phase Model - Cycle slip detection Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 16

17 Coherent Altimetry: Rayleigh criterion: Specular reflection change into diffuse scattering depending on: Signal wavelength λ surface roughness σ Elevation angle E Slide 17

18 Coherent altimetry: 7

19 First results: Phase observations Reflection events with an approximate antenna height of 10 m Vietnam, Can Tho, Receiver River Bank Time serie 1: 10 m antenna height ~5 days Refl. Events: ~254 h Coherent: ~65 % Refl.Events Reflection events with an approximate antenna height of 20 m Vietnam, Can Tho, Reflection point traces Time serie 2: 20 m antenna height ~3 days Refl. Events: ~112 h Coherent: ~57 % Slide 17

20 First results: Phase Observations Phase Model - Cycle slip detection Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 20

21 Input: First results: Phase model Ionospheric and tropospheric delay Satellite and receiver clock error Satellite elevation Relative height between reflecting surface and receiver Observations Model Slide 21

22 First results: Phase Observations Phase Model - Cycle slip detection Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 22

23 First results: Cycle slip detection Cycle slip detection, PRN 13 Vietnam, Can Tho, 25 th February 2012 Slide 23

24 First Results: Cycle slip detection Hypothesis H 0 : Hypothesis H a : Slide 24

25 First Results: Cycle slip detection Hypothesis H 0 : Hypothesis H a : Slide 25

26 First Results: Cycle slip detection Hypothesis H 0 : Hypothesis H a : Slide 26

27 First results: First conceptual Draft Phase Observations - Cycle slip detection Phase Model Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 27

28 First results: unwrapped phase differences Input height [m] Slide 28

29 First results: unwrapped phase differences Input height [m] Slide 29

30 First results: Example Slide 30

31 First results: Accuracy: 0.1 m (1σ) Slide 31

32 Preliminary results: Unwrapped Phase Obs. Unwrapped Phase Model - Cycle slip detection Extract. polynom coefficient (Least-Squares Method) Data Snooping Average height Ambiguity fixing Slide 32

33 Outlook & Conclusion: GNSS-R is a promising technique that could be used as a complement to other measurement methods Improvement of the phase model Multipath mitigation and tropospheric correction Improvement of stochastic model Least-Squares and Data snooping results LHCP and RHCP antenna for the reflected signals Implement the second part of the algorithm Ambiguity fixing by using a Kalman Filter Slide 33

34 Thank you for your attention Slide 34

35 Slide 35 i sat rcv i sat rcv sat rcv sat rcv sat rcv sat rcv i direct i i i i i i N iono trop t t c i sat rcv i sat Po refl rcv Po refl sat Po refl sat rcv sat rcv reflected E h N iono trop trop t t c i i i i i sin 2 int. int. int. Preliminary results: Phase model i multipath sat Po refl rcv i sat Po refl rcv Po refl sat Po refl direct reflected i i i N iono trop trop ds int., int. int. int.

36 GORS Receiver: GORS Channel Up to 10 channels run in parallel (master and slave correlators) 10 reflections can be recorded simultaneously Slide 36

37 Coherent altimetry: Coherency causes the typical rotation of the Phasor

38 Coherent altimetry: Coherency causes the typical rotation of the Phasor

39 First Fresnel zone, 10 m height antenna Slide 39

40 First Fresnel zone, 20 m antenna height Slide 40

41 First results: Least-Squares extraction of polynomial parameter a,b,c: unknown polynomial parameters t : time l : observations A : Jakobi Matrix v : noise Amp slave : Slave Amplitude Amp master : Master Amplitude Slide 41