The Typhoon Investigation using GNSS-R Interferometric Signals (TIGRIS)

Transcription

1 The Typhoon Investigation using GNSS-R Interferometric Signals (TIGRIS) F. Fabra 1, W. Li 2, M. Martín-Neira 3, S. Oliveras 1, A. Rius 1, W. Yang 2, D. Yang 2 and Estel Cardellach 1 1 Institute of Space Sciences ICE-CSIC/IEEC, Spain 2 School of Electronic and Information Engineering SEI, BeiHang University, China 3 European Space Agency ESA-ESTEC, Netherlands SPACE REFLECTO rd Conf. on passive reflectometry using radiocom space signals

2 Outline

3 Outline Frame Typhoon remote sensing

4 The frame of this work Frame Typhoon remote sensing China-Europe Joint Initiative National Remote Sensing Center of China (NRSCC), Ministry of Science and Technology School of Electronic and Information Engineering (SEI), BeiHang University China Meteorological Administration (CMA) European Space Agency (ESA) Institute of Space Studies from Catalonia (IEEC) + Other institutions TIGRIS experiment Typhoon Investigation using GNSS-R Interferometric Signals Main purpose: Research on GNSS-R towards remote sensing of typhoons But also: Provide recommendations for a future GNSS-R space-based mission Related work: aircraft experiments done by S.J. Katzberg and NASA s CYGNSS mission

5 Frame Typhoon remote sensing Typhoon remote sensing with GNSS-R Basic idea... Typhoons can be characterized by its impact over the sea surface Remote sensing measurements through intense rain fall (lower attenuation at L-band) better understanding on typhoon s air-sea interaction Potential retrievals Wind Speed (WS) waveform s amplitude and area Significant Wave Height (SWH) waveform s leading edge and coherence time Sea Surface Level (SSL) anomalies (up to 1 m) altimetry Interferometric waveforms (all GPS codes) with different wind speeds simulated by W. Li under PARIS IoD nadir geometry

6 Outline Scenario Instrumentation Collected data

7 Scenario Instrumentation Collected data Location: Xichong bay, South-East of China

8 Scenario Instrumentation Collected data Scenario: Antennas installation in Shenzhen site

9 Campaign: Instrumental setup Scenario Instrumentation Collected data IEEC s Equipment PIRA Eth USB Power RHCP U H RF1 LHCP D H RF0 Nav A1) RHCP A2) RHCP Power Supply Jack GOLD RTR RHCP D H Link 3 LHCP D H Link 2 RHCP U L Eth Link 1 Power Nav PS 1 In 2 Antenna Inputs RF Coaxial N Type Female 1 PS 2 IN 1 PS 2 IN Power Input AC 220V/50Hz CEE 7/7 plug 1 PS 2 IN SEI s Equipment 2 channel BeiDou IF signal collector IN1 RHCP U H 1 PS 2 3 IN 4 USB IN2 LHCP D H 4 channel GPS IF signal collector IN1 LHCP D H IN2 RHCP U L USB RHCP D H IN3 IN4 A3) LHCP A4) RHCP INSTITUTIONS IEEC: Institut d Estudis Espacials de Catalunya SEI: School of Electronic and Information Engineering POWER SPLITTER (PS) IN Input DC THRU 1 Output DC THRU 2, 3, 4 Outputs DC BLOCKED ANTENNA SYMBOLS Up looking Low gain Antenna Up looking High gain Antenna Down looking High gain Antenna

10 Scenario Instrumentation Collected data Campaign: IEEC s GNSS-R Instrumentation GOLD-RTR GPS Open Loop Differential Real Time Receiver GNSS-R dedicated hardware receiver (standard approach using GPS L1 C/A code) 10 channels compute cross-correlations (waveforms) of 64 lags every millisecond 50 ns lagspacing 15 meters PIR PARIS Interferometric Receiver GNSS-R dedicated hardware receiver (direct cross-correlation) 1 waveform of 320 lags every millisecond 12.5 ns lagspacing 3.75 meters

11 Scenario Instrumentation Collected data Campaign: SEI s GNSS-R Instrumentation BeiDou IF signal collector Dual-front-end for both direct and reflected signals MHz sampling rate (18.75 meters resolution in delay) Configurable to BeiDou, GPS and GALILEO GPS IF signal collector Quad-front-end channels MHz sampling rate GNSS-R software receiver in post-processing Computes waveforms using standard or interferometric approaches Compatible with both Beidou B1 and GPS L1 band signals

12 Collected data during typhoons Scenario Instrumentation Collected data Relevant aspects Campaign s duration: from end of July until end of September 2013 Two typhoons and one tropical storm were monitored! An additional (but simpler) setup collected data from Yangjiang station Name Classification Min. pressure Dates Distance JEBI Strong Tropical Storm 985 mbar 31-July/3-Aug 470 km UTOR Super Typhoon (cat. 4) 925 mbar 9-Aug/15-Aug 270 km USAGI Super Typhoon (cat. 5) 910 mbar 16-Sept/23-Sept 100 km given next Altimetric retrievals from GOLD-RTR around UTOR and USAGI Scatterometric measurements from BeiDou IF signal collector around JEBI and UTOR

13 Outline

14 Main aspects Results from IEEC s GOLD-RTR High gain antennas should avoid direct signal contamination However, we try to remove waveforms with presence of residual direct signal Estimated coherence time of 100 msec 100 sec of incoherent integration to minimize speckle (120 sec for practical purposes) Two periods analyzed: around UTOR and USAGI N N W W Specular points over the ocean surface from GPS tracks during 14th Aug Elevation [deg]

15 : specular delay 120 Main aspects Position of waveform s peak power (instead of maximum of first derivative) Altimetric delay [m] H sin(elevation) Related to altimetry Elevation [deg] Low elevation angles magnify altimetric errors during delay-height inversion Only values obtained for elevations > 25 are considered Height residual [m] Elevation [deg]

16 : specular delay 2.0 UTOR s landfall 1.5 Estimated SSL [m] August 13 August 14 August 15 August 16 August 17 August USAGI s landfall 1.5 Estimated SSL [m] September 18 September 19 September 20 September 21 September 22 September 23

17 : specular delay (OSU Tide model) 2.0 UTOR s landfall 1.5 Estimated SSL [m] August 13 August 14 August 15 August 16 August 17 August USAGI s landfall 1.5 Estimated SSL [m] September 18 September 19 September 20 September 21 September 22 September 23

18 : specular delay (tide-corrected) 2.0 UTOR s landfall 1.5 Estimated SSL [m] August 13 August 14 August 15 August 16 August 17 August USAGI s landfall 1.5 Estimated SSL [m] September 18 September 19 September 20 September 21 September 22 September 23

19 : Discussion 2 30 First impressions Altimetric retrievals show moderate agreement with tides, but errors are larger than expected Altimetric errors show track-evolution multipath effect due to direct signal contamination! Next step: to retrieve altimetry from more robust phase observables (experience from previous campaigns) Estimated SSL [m] Estimated SSL [m] August August Elevation [deg] Elevation [deg]

20 Main aspects Results from SEI s BeiDou IF signal collector BeiDou GEO #1 and #4 satellites are monitored stable geometry and sharper auto-correlation function (2.046 Mchips) Off-specular reflections spatial filtering of antenna footprint and coastline Two observables are analyzed: Area of the waveform and coherence time Two periods processed: around JEBI and UTOR Specular points from BeiDou GEO #1 and #4 satellites

21 : Area of the power waveform (JEBI) Area of the power waveform Expected to be sensitive to the sea surface wind speeds (proportional)

22 : Area of the power waveform (JEBI) Area of the power waveform Expected to be sensitive to the sea surface wind speeds (proportional)

23 : Area of the power waveform (JEBI) Area of the power waveform Expected to be sensitive to the sea surface wind speeds (proportional)

24 : Area of the power waveform (UTOR) Area of the power waveform Expected to be sensitive to the sea surface wind speeds (proportional)

25 : Coherence time (JEBI) Coherence time Expected to be sensitive to the Significant Wave Height (inversely proportional)

26 : Coherence time (JEBI) Coherence time Expected to be sensitive to the Significant Wave Height (inversely proportional)

27 : Coherence time (UTOR) Coherence time Expected to be sensitive to the Significant Wave Height (inversely proportional)

28 : Coherence time (UTOR) Coherence time Expected to be sensitive to the Significant Wave Height (inversely proportional)

29 Outline

30 Moderate agreement with tides, but errors are larger than expected due to direct signal contamination still have room for improvement! More robust phase observables might overcome these limitations Good correlation with the in-situ wind speed and SWH measurements Off-specular reflections (spatial filtering of antenna footprint and coastline) benefit coastal applications BeiDou GEO satellites provide good opportunities for sea state measurement On-going analysis Phase-based and polarimetric measurements, GPS interferometric signals... Combined analysis is required to enable a proper typhoon characterization

31 Thank you for your attention