1

PALSAR SCANSAR SCANSAR Interferometry

Masanobu Shimada

Japan Aerospace Exploration AgencyEarth Observation Research Center

ALOS PI symposium, GreeceNov. 6 2008

Introduction

•L-band PALSAR strip mode with 70km swath can effectively achieve the deformation mapping even for the vegetation covered area. (Chuetsu-oki, Noto, Solomon, Iwo, Hawaii, etc.)

•PALSAR SCANSAR with 350km swath is more effective than the repetitive observations of the STRIP mode for global land observation. SCANSAR SCANSAR interferometry (SSI) is also important tool for the earth observation.

•Azimuth signal spectrum is the dispersive and transmission beam synchronization is mandatory for the SSI.

•PLASAR activation timing is controlled by 1 sec as the normal operation. This has been limiting the SSI for long time.

•Here, we conducted the experiomental activation of the PALSAR for SSI performance evaluation. The target area was Africa (Sahara desert and Tanzania) and Amazon.

•We introduce the results of the first PALSAR SSI.

3

PALSARL-band (23.6 cm)Synthetic Aperture RadarPolarimetryDual PolarizationSCANSAR

The Advanced Land Observation Satellite - ALOS

Launch:24 Jan. 2006

4

PALSAR SCANSAR

WB1 : Bw 14 MHz, 5 SCANs, 3looks ~ 7 looks, 120 Mbps

WB2 : Bw28MHz, 5 SCANs, 1~ 3 looks, 240 MHz

6

Timing block diagram (1)

timing

area

Final image

prffDD

⋅ vg

nn:pulses

prffDD ⋅ nn

⋅ vg

spacing

Ts

Timing allocation block diagram

timingnn:pulsesTs

S/C

Frequency (master)

BW = − fDD ⋅rg ⋅ βvg

BSw = − fDD ⋅Ts ⋅Nlook

ρ = BW

BSW⋅L2≅ 5 ⋅ L

2

Azimuth Bandwidth

AB for swath I

resolution

Power spectrum of the SCANSAR in azimuth (spectrum diversity).

SSI condition

2D spectrum (range and azimuth) should be the same.

Assumption10 % allowance for azimuth spectrum

ΔTs < 0.1Ts ~ 0.014 s. ~ 90mWhere, 1st scan, Ts~0.2 sCorresponding position allowance is 90m

9

Expression for the ground target point (rp)

fd =fPRF

nni (− nn

2≤ i < nn

2)

fd =2λ

vs − vp( )⋅ rp − rs

rp − rs

z = F rp( )xp

2

Ra2 +

yp2

Ra2 +

zp2

Rb2 =1 (if z = 0)

fd: Doppler frequency vs:satellite velocityfPRF:Pulse repetition frequency vp:target point velocitynn:burst numbers i: azimuth addressrp:position vector of the pixelrs:position vector of the satellitez:height of the targetF(): function giving the height

InSAR processing Routine(SPECAN)

Master Slave

Telemetry: Doppler Model

Telemetry: Doppler Model

Range compression

SPECAN (Azimuth)

Co-registration

summing(box car)n-look processing Bp correction

Terrain correction

common

Curvature

Range compression

SPECAN (Azimuth)

Curvature

output

fD = a0 + a1 ⋅ r + a2 ⋅ r2Doppler model :

γe jϕ =a ⋅b*

a ⋅ a* ⋅ b ⋅b*

rg' = a + b ⋅ rg + c ⋅az

az' = d + e ⋅ rg + f ⋅az

master slave coherence fringe

Sample result of the InSAR products output from the Browse InSAR processor (SPECAN) : Functionality test

12

Results of Browse-InSAR（2006/8/2-2006/5/2）(Specan processing@SIGMA-SAR)

Coherence DEM Deformation Amplitude

13

Results from FBS DinSAR: 2006/8/2-2006/5/2:Bp=80m(Correlation processing@SIGMA-SAR)

Coherence DTM Deformation Amplitude

14

prf11

prf12

prf21

prf22

prf31

prf32

prf41 prf51

prf42prf52

1 2 3 4 5

time

T1 1) Normal operation•T1 is 1s unit. (Even same latitude is set for the first transmission, 7km dislocates at maximum)•prf is tabulated every 100 m height.

-> SSI could be done by chance( But, never realized the SSI)

2) Special operation•Specify the followings

Latitude of argumentFive prfs

[Note:Latitude of augment from GPS is biased(0.993km).]-> SSI can be done.

Operational constraints and achievement for PALSAR SSI

15

SCANSAR SCANSAR InSAR experimentsDuration April 2 - April 17 2008Sites: Sahara, Tanzania, AmazonMode WB1 and WB2

50%

80%

BS

168m

228m

Bp

Differs in prfNGWB22/234/9

Large BpNGWB22/284/14OKWB13/24/17

Differs in prfNGWB12/294/15NGWB12/284/14NGWB12/274/13

Different locationDifferent location

Area

NGWB12/194/5

OK

NG

Result

WB12/244/10

WB12/164/2

ModeSlaveMaster

16

SCANSAR parameters : Beam synchronization

500m0.5168m20080302200804172800m0.8228m20080224200804101LoΔpBpSlaveMasterCase

5.380.791Total1.160.1711915.732751.150.1642159.835541.090.1601715.227431.020.1502369.735620.9930.1461692.02471Length (km)Duration(s)PRFPulsesNo

Δp j =Lo j

L j

LO

L

17

Location of the test paths

18

350km

Amplitude

April 10 - Feb. 24Bp=~200mBeamS~80%Tanzania

19Coherence

20Flat earth corrected fringe

21Deformation (cancelled for the low frequency)

22

RSP325Sahara

2008/4/17-2008/3/2

Bp=~100m

BeamS~50%Coherence Amplitude

DTM Deformation

23

WB1

WB2

May 16 2006-Oct. 22 2007

SC

AN

SA

R data archives

Conclusions

PALSAR SCANSAR SCANSAR Interferometry has been examined and succeeded by conducting the beam synchronization with the previous images.

The parameters are the PRFs and augment of latitude for the first scan.

Two examples out of 9 examples show that 80% and even 50% of the beam synchronization can produce the SSI over the dense forest and desert of Africa.

It promises that PALSAR can provide the SCANSAR SCANSAR Interferometry for wide range of the target, i.e., earthquake, Antarctica, Greenland monitoring.

Further researchCorrection of the SCAN-SCAN phase offsetCorrection of the atmospheric error and orbital errors.Improvement of the processing speedTrial for Full aperture type processing