Ionospheric Structure Imaging with ALOS PALSAR

Size: px

Start display at page:

Download "Ionospheric Structure Imaging with ALOS PALSAR"

Sophia Conley
5 years ago
Views:

1 The Second ALOS PI Symposium Rhodes, Greece November 3 7, 008 Ionospheric Structure Imaging with ALOS PALSAR PI Number: 37 JAXA-RA PI: Jong-Sen Lee, Thomas L. Ainsworth and Kun-Shan Chen CSRSR, National Central University, TN (ROC) Naval Research Laboratory, Washington DC, USA

2 Introduction Ionospheric irregularities and scintillations can cause sufficient phase and amplitude changes in L-Band SAR returns to produce distortions in the SAR image formation. Image based algorithms will be used to study ionospheric effects.

3 Effect of Ionospheric Disturbances Simulated effects of the equatorial ionosphere on a.7 GHz electromagnetic wave scattered from a point on the earth (black dot) and received on a satellite at 750 km altitude (red dots). The simulation represents distortion of SAR backscatter by a single bubble irregularity in the equatorial ionosphere. Of course, the same ionospheric disturbances also distort the transmitted SAR signal further complicating detailed data analysis 3

4 Objective Enhance our knowledge of ionospheric processes that produce field-aligned irregularities based on ALOS PALSAR measurements at the HAARP facility in Alaska and three other sites at low latitudes Determine the degree to which the ionosphere distorts ALOS PALSAR imagery Develop robust ionosphere correction techniques for ALOS PALSAR imagery 4

5 Challenge Coordination of ground-based and spacebased data collection with naturally occurring phenomena: Ground-based ionosphere measurements available at proposed test sites ALOS PALSAR imagery should be collected simultaneously with ground campaigns Solar activity affects the ionosphere Currently near a minimum in the year solar cycle 5

6 Proposed Experiment Sites LOCATION (LAT., LONG.) TIME PERIOD PLANNED PRODUCT LEVEL VOLUME PRIMARY TRUTH SENSORS Kwajalien (ALTAIR) (8.7 N, 67.7 E) Aug.-Sept..and. (4-0 collections) ALTAIR COSMIC Indonesia (EAR) (0.0 S, 00.3 E) Mar-Apr., Sept-Oct..and..and. (4-0 collections) EAR COSMIC Alaska (HAARP) (6.39 N, 45.5 W) Summer.and. (4-0 collections) MUIR COSMIC Puerto Rico (Arecibo IO) (8.33 N, W) Any time.and. (4-0 collections) Arecibo Dish ALTAIR: ARPA-Long Range Tracking and Instrumentation Radar EAR: Equatorial Atmosphere Radar HAARP: High Frequency Active Auroral Research Program Arecibo IO: Arecibo Ionospheric Observatory COSMIC (TBB): Taiwan FORMOSAT- 3 satellites (Tri-Band Beacon) 6

7 The Main Thrust Generate artificial field-aligned scintillations using HAARP s high power HF transmitter Synchronization with ALOS PALSAR observations First attempt in September 007 HAARP HF transmitter did not turn on to heat the ionosphere Second attempt in October 007 Poor ionospheric conditions low free electron density Third attempt in October 008 (last week) Opportunistically employ quad-pol PALSAR imagery distorted by ionospheric irregularities 7

8 HAARP HF Radar Array 8

9 Ionospheric Effects on SAR Total Electron Content, TEC path n electron ( x) dx Satellite Altitude ~700km Ionosphere F Layer Alt. ~350km TEC Varies Diurnally and with Solar Sunspot Cycle Ionosphere TEC Causes Refraction, θ Faraday Rotation θ First-Order Effects on Pol. SAR: Faraday Rotation TEC Value Azimuth Shifts TEC Gradient 9

10 Faraday Rotation Estimation O O hh vh O O hv vv = cosω sin Ω sin Ω S cosω S hh hv S S hv vv cosω sin Ω sin Ω cosω Estimated Faraday rotation, Ω, in the circular basis Well-defined based on second-order statistics Insensitive to orientation angles Insensitive to scattering mechanism of the ground Ω = * LR ( < 4 Arg O O RL > ) O RL = [ Ohv Ovh + j( Ohh + O O LR = [ Ovh Ohv + j( Ohh + O π Ω 4 vv vv )] )] π 4 0

11 Ω Ω Ω + Ω Ω Ω Ω + Ω = VV VH HV HH VV VH HV HH VV VH HV HH VV VH HV HH t t t t S S S S r r r r O O O O cos sin sin cos cos sin sin cos VV VH HV HH VV VH HV HH S S S S k k z u uz v uv u w wz z vw w v O O O O γ γ γ γ γ γ γ γ γ γ γ γ α α α α Full Calibration Model where γ = tan Ω, and Ω is the Faraday rotation angle Faraday rotations can destroy diagonal dominance of the [X][G][F] calibration matrix. γ is not necessarily a small parameter! X-talk, [X] Gain, [G] Faraday, [F] T.L. Ainsworth, L. Ferro-Famil & J.-S. Lee, TGRS, vol. 44, 006, pp

12 Alaska at (6.8N, W) (ALOS PALSAR - RA- data) Google Earth HH-VV, HV + VH, HH+VV Faraday Rotation PALSAR

13 Mean =.67 Azimuth Profile of Faraday Rotation Faraday Rotation 3

14 Near HAARP, Gakona, Alaska (April, 007) Google Earth HH-VV, HV + VH, HH+VV Faraday Rotation PALSAR 4

15 Mean =.987 Irregularity Azimuth Profile of Faraday Rotation Faraday Rotation 5

Geomagnetic Latitude N elec [0 e/m 3 ] 59 6 65 68 7 Geomagnetic Latitude N elec [0

16 HAARP Tomographic Imaging Data High-Frequency Active Auroral Research Program 3 March 007; UTC Quiet Day April 007; UTC Disturbed Day Geomagnetic Latitude N elec [0 e/m 3 ] Geomagnetic Latitude N elec [0 e/m 3 ] HAARP ionosphere measurements show spatial and day-to-day electron density variability 6

17 Imaging Ionospheric Structures Satellite Altitude ~700km Ionosphere F Layer Alt. ~350km Spatial TEC estimation by refocusing at ionosphere altitude instead of ground. TEC affects the phase of C RLLR ; ground scattering mechanism is irrelevant. Refocus at altitude by incorporating an additional quadratic phase term. PALSAR imagery displays low X-talk, -40dB, which simplifies TEC estimation. 7

18 Ionosphere Imaging Estimate quadratic phase by sub-aperture analysis Ionosphere structures shift across sub-apertures Determine shift to find quadratic phase correction Refocus SAR images & generate C RLLR phase No ground scatterer effects Only ionosphere effects present Limitations Requires ionosphere structures / gradients Smooth electron densities do not work Better for smaller vertical electron density distributions 8

19 Full Aperture C RLLR Phase Image Gakona Image chip Phase of C RLLR Sub-apertures reduce azimuth resolution but allow estimation of refocusing phase 9

20 Sub-Aperture C RLLR Phase Images Gakona Image chip Phase of C RLLR Ionosphere feature shifts between sub-apertures Use shift to estimate quadratic phase for reprocessing 0

21 Refocused C RLLR Phase Image Gakona Image chip Phase of C RLLR Sub-aperture estimated quadratic phase for reprocessing image Comparison of refocused and original phases Refocused image blurs ground level structures

22 Ionosphere Imaging Focusing at the ground plane collects ionosphere data across a ~0km azimuth window Focusing at the ionosphere altitude limited by the vertical extent of the electron density distribution Electron density layer ~50km thick implies ~700m azimuth window after refocusing Ionosphere height estimated by sub-aperture analysis Possibly further refined by auto-focusing, or a priori information Spatial ionosphere structures / features required Provide variable Faraday effects across image

23 Difficulties Encountered Low solar activity resulted in a quieter than normal ionosphere But a solar maximum is coming Synchronization of PALSAR observations and HAARP heating of the ionosphere RA does not allow Observation Requests 3

24 Conclusion L-band quad-pol imagery can be polarimetrically calibrated in the presence of Faraday rotations & ionospheric effects Faraday rotation precisely estimated by quad-pol (PLR-mode) ALOS PALSAR imagery Refocusing SAR image at the ionosphere altitude provides a high-resolution snapshot of TEC Imaging time scale: ~ second Azimuth length scale: ~ 700 meters High-resolution PALSAR Faraday rotation data may be used for calibration of ionosphere radars More PLR-mode (quad-pol) PALSAR data collection is desired Possibility of Observation Requests would simplify coordination of ground-based campaigns with ALOS PALSAR collections. 4

25 Acknowledgments JAXA ALOS team for providing data in support of this research. Dr. M. Shimada, EORC, JAXA for helpful assistance and coordination ASF ionosphere research team of Dr. F. Meyer and Dr. J. Nicoll for their help and collaboration 5

26 Background 6

27 Ionospheric Electron Density Effects on SAR Signals from Space 7

28 cos Ω sin Ω Faraday Rotations + sin Ω S cos Ω S cos Ω sin Ω + sin Ω HH HV [ S( Ω) ] = VH S S VV cos Ω [ F( Ω) ] = cos Ω cos Ω sin Ω cos Ω sin Ω sin Ω cos Ω sin Ω cos sin θ Ω cos Ω sin Ω cos Ω sin Ω sin cos Ω Ω cos Ω sin Ω sin Ω cos Ω sinω cos Ω sinω cos Ω γ γ γ γ γ γ γ γ γ γ γ γ where γ = tan Ω, and Ω is the Faraday rotation angle Faraday rotation and X-talk matrices are similar Faraday rotation can affect X-talk calibration Setting z = -u = -v = w and X-talk generates an unwanted Faraday-like rotation [ X ] z u uz v uv u w wz z 8 vw w v

29 Differenced Sub-Aperture Images Gakona Image chip C RLLR phase difference between sub-apertures Bright and dark bands show shift Initial images are unshifted Subsequent ones are shifted 0, 0 & 30 pixels 9

HAARP Tomographic Image 3 October 008; 98 937 UTC Disturbed Day

30 HAARP Tomographic Image 3 October 008; UTC Disturbed Day Last Week Geomagnetic Latitude N elec [0 e/m 3 ] 30

31 Near HAARP, Gakona, Alaska (ALOS PALSAR - RA- data) Google Earth HH-VV, HV + VH, HH+VV Faraday Rotation PALSAR 3

32 Faraday Mean = Azimuth Azimuth Profile Profile of Faraday of Faraday Rotation Rotation Faraday Rotation 3

33 Near Kwajalien (9.458N,67.06E) (ALOS RA data) Google Earth HH-VV, HV + VH, HH+VV Faraday Rotation PALSAR 33

34 Mean = Azimuth Profile of Faraday Rotation Faraday Rotation Note: Lower backscattering from ocean surface 34 doest not affect Faraday rotation evaluation

35 HAARP HF Radar Array 35

ALOS-Indonesia POLinSAR Experiment (AIPEX): First Result*

ALOS-Indonesia POLinSAR Experiment (AIPEX): First Result* Mahmud Raimadoya(1), Ludmila Zakharova(2), Bambang Trisasongko(1), Nurwadjedi(3) (1) Bogor Agricultural University (IPB), P.O. Box 2049, Bogor