Synthetic Aperture Radar (SAR) Basic Principles & Concepts

Irena Hajnsek Earth Observation and Remote Sensing Institute of Environmental Engineering ETH Zurich Email: [email protected] German Aerospace Center (DLR) Microwaves and Radar Institute (DLR-HR) Pol-InSAR Research Group VU 1 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

German Aerospace Center (www.dlr.de) DLR is Germany’s Aerospace Research Center and Space Agency 31 research institutes and scientific / technical facilities DLR Oberpfaffenhofen: mostly space activities 5 research institutes 3 technical facilities including: –

German Space Operations Center (GSOC);

–

German Remote Sensing Data Center (DFD).

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 2 > Autor Name Microwaves and Radar Institute > 30.05.2006

1

Electromagnetic Spectrum

Microwave Window

Optical Window

Electromagnetic spectrum and attenuation caused by Earth’s atmosphere VU 3 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

35,3 km

21,3 km

1994: Shuttle Radar Lab SIR-C/X-SAR

Microwaves and Radar Institute > 30.05.2006

X (3cm)

CETH(5cm) Zü Zürich – DBAUG - IFU - Earth Observation Chair L (23cm)

VU 4 > Autor Name

Microwaves and Radar Institute > 30.05.2006

2

VU 5 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 6 > Autor Name

3

Infrared Image (Spaceshuttle Columbia)

Radarbild (Spaceshuttle Endeavour) Endeavour)

VU 7 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 8 > Autor Name

4

Unique Characteristics of Microwave Remote Sensing Independent of Weather Conditions: Penetrate clouds, rain, (smoke); (Lower Frequencies) Penetrate into / through a wide class of natural cover types as: Sand / Ice / Vegetation; Sensitive to objects of dimensions from cm to m: (Complementary to Optical and IR remote sensing); Very accurate (differential) distance measurements (employing interferometric techniques); (Active) Microwave systems are able to operate day and night.

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 9 > Autor Name Microwaves and Radar Institute > 30.05.2006

Range Resolution

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 10 > Autor Name Microwaves and Radar Institute > 30.05.2006

5

Christian Hülsmeyer: Inventor of Radar (1904)

VU 11 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Radar Measurement Principle Range distance ro

co

(Velocity of light)

Tx

co

object

Rx

Received echo signal (back-scattered signal of imaged object): Total time delay =

Total distance = Velocity of light

2 . ro co t (time)

receive

transmit ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 12 > Autor Name Microwaves and Radar Institute > 30.05.2006

6

(Slant) Range Resolution Pulse Duration τ

Range difference Δro

co Tx

co Rx Δro Time difference Δt =

Total range difference Velocity of light

=

2 .Δr roo co

Δro

co Tx

co Rx Δro Two scatterers can be resolved because they have different time delays VU 13 > Autor Name ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair i.e. because their echoes arrive at different times

Microwaves and Radar Institute > 30.05.2006

(Slant) Range Resolution Pulse Duration τ

Range difference Δro

co Tx

co Rx Δro Time difference Δt =

Total range difference Velocity of light

=

2 .Δr roo co

Δro

τ

co Tx

co Rx Δro When the time difference Δt (i.e. the difference of time delays) becomes VU 14 >smaller Autor Name ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair than the pulse duration τ then the objects cannot be separated from each other ! Microwaves and Radar Institute > 30.05.2006

7

(Slant) Range Resolution Range resolution: The min. distance between two objects that can still be separated from each other i.e. the distance that corresponds to a time difference Δt equal than the pulse duration τ

δr

co Tx

co Rx Δro Time difference Δt := Pulse duration τ =

2 Range resolution = Velocity of light

2δ c

δr 

→ rRange resol.:

cτ 2

The range resolution is independent on the distance between the sensor and the scatterer; it depends only on

δr 

the pulse duration τ that is inverse proportional to the pulse bandwidth W:

cτ c  2 2W

Example: A pulse bandwidth of W=100MHz leads to a range resolution δr=1.5m (corresponding to a pulse VU 15 > Autor Name duration of τ=10ns) . ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

The Dilemma The energy of a pulse defines the (maximum) distance that a object can be detected:

Range distance ro

Received signal energy ERX from an object at distance r0: The energy of the transmitted pulse and pulse duration τ: E Tx  PTx τ

Object

ERx ~

ETx r04

EisTxgiven by product of instantaneous peak power P

P is limited by the sensor hardware – especially in the case of spaceborne sensors so that for increasing the pulse energy long pulses are required. But on the other side: For achieving a high spatial resolution short pulses (realised by wide bandwidths) are required. The solution is to use pulse modulation. Commonly used: Linear frequency modulation (FM Chirp) ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 16 > Autor Name Microwaves and Radar Institute > 30.05.2006

8

Amplitude variation

Linear Frequency Modulated Pulse

FM Chirp

The frequency f0 is linearly changed along the pulse time

fr ( t )  k r t

Frequency variation

τ 2

Frequency variation



τ 2

τ:= Pulse duration



τ τ t 2 2

kr maybe positive (Up-Chirp) or negative (Down-Chirp)

The bandwidth

Wr : is Δfindependent on the pulse

length τ. So it becomes possible to generate pulses

Wr  Δf

with large pulse lengths τ and large bandwidths Wr

kr

τ 2

τ 2

Phase Modulation



with

Chirp bandwidth

Wr : k r τ

Pulse compression gain

Cr  10 log10 ( Wr τ r )

time

φr ( t )  πk r t 2

Phase modulation

Note: The phase of the chirp has ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair a

VUon 17 > time. Autor Name quadratic dependency Microwaves and Radar Institute > 30.05.2006

Matched Filter The high range resolution is achieved after reception: The long chirp(ed) pulse so(t) is compressed using a “matched filter”. The output pulse uo(t) will have the same total energy of the input pulse so(t) but will be significantly shorter.

n(t)

Matched filter xo(t)

so(t)

ho(t)

uo(t)





u 0 ( t )  s 0 ( t )  h 0 ( t )  s 0 ( T )  h 0 ( t  T ) dΤ

where  stands for convolution



The matched filter function ho(t) is given by the complex conjugated, time-reverted transmitted shirp h0 ( t )  m  s0 ( t ) ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 18 > Autor Name Microwaves and Radar Institute > 30.05.2006

9

Range Compression by Convolution

τ

Range SAR Signal

t Convolution

Reference Function

t

δr

e

Point Target Response

t ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 19 > Autor Name Microwaves and Radar Institute > 30.05.2006

Range Compression by Convolution Range SAR Signal

t Reference Function

Convolution

t

δr

e

Point Target Response

t ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 20 > Autor Name Microwaves and Radar Institute > 30.05.2006

10

Matched Filter Assuming white Gaussian noise (e.g. thermal noise): ►Pulse compression with matched filtering leads to best possible Signal-to-Noise ratio.

τ

The length of the compressed pulse is limited by the bandwidth of the chirp:

1 Wr

Examples of transmitted pulses:

Pulse with 10 ns width

Chirp with 5 s width and 100 MHz

leads to 1,5 m resolution.

bandwidth leads to 1,5 m resolution

(Bandwidth=100MHz)

(after pulse compression).

Gain in the average power by using chirp signal:

G = 5 s / 10 ns = 500 !!! VU 21 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Radar Types Nadir Looking Implementations:

Side Looking Implementations

Radar Altimeters

Scatterometers

Radar Sounders

Imaging Radars (SLAR & SAR)

Left (A) and right (B) echoes are not separable.

A

The side looking geometry resolves the left-right ambiguit

B

A ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

B VU 22 > Autor Name Microwaves and Radar Institute > 30.05.2006

11

Basic Radar Block Diagram

Transmitter Antenna

Radar Pulse Circulator

Receiver

Transmitter: generates generates aa high high power power pulse pulse •• Transmitter: Data Recording

Circulator (Switch): (Switch): switches switches transmitted transmitted pulse pulse to to antenna, antenna, •• Circulator returned echoes echoes to to receiver receiver returned Antenna directs directs transmitted transmitted pulses pulses towards towards the the target target area area •• Antenna Receiver amplifies amplifies the the received received signal signal and and converts converts to to base base band band •• Receiver

VU 23 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Side-Looking Imaging Geometry

Microwaves and Radar Institute > 30.05.2006

z

• Pulsed radar system • Two-dimensional imaging (azimuth x slant range) Antenna Azimuth Range

• Timing of the Radar: Tx

Tx Rx

Swath



x

y

Rx

Illuminated area T = 1/PRF

PRF = Pulse Repetition Frequency ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 24 > Autor Name Microwaves and Radar Institute > 30.05.2006

12

Slant Range 2 Ground Range Resolution δr  θ d Depression Angle

cτ 2

Slant Range resolution:

Look Angle θ 0 (or Incidence Angle)

δr 

cτ c  2 2Wr

Wr:= Chirp bandwidth τ := Pulse duration δr 

cτ 2

is the “natural” radar range coordinate Ground Range resolution: δGr 

δr cos(θ0 )

Ground range is Slant range δr 

δGr 

cτ 2

projected on the geoid δr 

cτ 1 2 cos(θ0 )

δGr 

cτ 2

cτ 1 2 cos(θ0 ) VU 25 > Autor Name

Slant Range 2 Ground range transformation by resampling

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Radar measures distance !!!

VU 26 > Autor Name

13

Shadowing

Shadows: Areas in the scene that are not illuminated for example when the incidence angle is larger than the negative terrain slope

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Foreshortening

VU 27 > Autor Name Microwaves and Radar Institute > 30.05.2006

Compression of topo features in the scene Becomes stronger when the the (local) terrain slope becomes closer the incidence angle.

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 28 > Autor Name Microwaves and Radar Institute > 30.05.2006

14

Layover

Incidence angle = the (local) terrain slope.

VU 29 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 30 > Autor Name

15

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Layover

VU 31 > Autor Name Microwaves and Radar Institute > 30.05.2006

Extreme case of foreshortening that occurs When the incidence angle becomes smaller than the (local) terrain slope.

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 32 > Autor Name Microwaves and Radar Institute > 30.05.2006

16

VU 33 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 34 > Autor Name

17

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 35 > Autor Name Microwaves and Radar Institute > 30.05.2006

Azimuth Resolution

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 36 > Autor Name Microwaves and Radar Institute > 30.05.2006

18

Real Aperture Radar Angular resolution:

θa 

λ da

Spatial resolution (Az):

δa  θaR0 

λ R0 da

d  δa (R ) A p e r t u r e

θa

da

R0

R0

δa (R )

x

δa (R )

The azimuth resolution decreases with increasing distance to the scatterer. High azimuth resolution requires high frequencies, large antennas & small ranges.

y

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 37 > Autor Name Microwaves and Radar Institute > 30.05.2006

Real Aperture Radar Angular resolution:

θa 

λ da

Spatial Resolution (Az):

λ R0 da

d  δa (R )

R e a l A p e r t u r e

δa  θaR0 

θa

da

R0

R0

δa (R )

δa (R )

Example Example1: 1:X-Band X-Band Airborne Airbornesystem, system,50 50MHz MHzbandwidth, bandwidth,33mmantenna, antenna,3000 3000mmrange range

r ==33m m r

a == 30 30m m a

Example X-BandSatellite Satellitesystem, system,50 50MHz MHzbandwidth, bandwidth,12 12mmantenna, antenna,800 800km kmrange range Example2: 2:X-Band

x y

r ==33m m r

a == 2000 2000m m !! a

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 38 > Autor Name Microwaves and Radar Institute > 30.05.2006

19

Synthetic Aperture Radar

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 39 > Autor Name Microwaves and Radar Institute > 30.05.2006

SAR Basic Principle

swath width

a

x y

• Pulsed radar system • Radar system must be coherent (stable local oscillator). Phase information is preserved

• Two-dimensional imaging (azimuth x slant range) • Azimuth resolution is independent on range distance ! ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 40 > Autor Name Microwaves and Radar Institute > 30.05.2006

20

SAR Azimuth Resolution

• Angular resolution of synthetic antenna:

v P Platform velocity

θSA 

λ d  a 2L SA 2R0

• Length of the synthetic aperture: S y n t h e t I c A p e r t u r e

L SA  δa  θaR0 

L SA

θa

R0

λ R0 da

• Azimuth resolution: a

θa

δSA  θSAR0 

da 2

Azimuth Resolution = Half Antenna Length in Azimuth

x The azimuth resolution is independent on range (i.e. the distance to the scatterer).

y

It is determined only by the size of the real antenna. ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 41 > Autor Name Microwaves and Radar Institute > 30.05.2006

SAR Azimuth Resolution

S y n t h e t I c A p e r t u r e

L SA L SA

θa

R0

R0

x The loss of resolution with increasing range (i.e. distance to the scatterer) is

y

compensated by the increase of the sythetic aperture. ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 42 > Autor Name Microwaves and Radar Institute > 30.05.2006

21

Synthetic Aperture Formation Lsa Point Target Flight Direction

Beamwidth of Real Aperture Antenna

SAR Sensor



Two-dimensional received SAR signal

VU 43 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

SAR Phase History

Microwaves and Radar Institute > 30.05.2006

•Phase history

v P Platform velocity

φ( t )  2

 4π 2π 2π v P2 t 2   r( t )   R0  λ λ R0   λ

Range variation r(x): S y n t h e t I c

r( x )  R02  x( t )2  R0 

x2 v2 t2  R0  P 2R0 2R 0

where: x : v P t

L SA

θa

R0

A p e r t u r e

• Frequency variation:

r(x)

fa ( t ) 

1 1 dφ( t ) 2 v P2 ω(t )   t 2π 2π dt λ R0

• Doppler Rate:

x

κ a : 

due to the quadratic range var.

2 v P2 λ R0

• Azimuth (or Doppler) Bandwidth:

y

2 P a tot ETH Zü ü rich – DBAUG IFU Earth Observation Z 0 Chair

W :

2v t λR



2v P2 R0θa 2v P  da λR 0 v P

VU 44 > Autor Name

Microwaves and Radar Institute > 30.05.2006

22

SAR Azimuth Signal v P Platform velocity Phase history S y n t h e t I c

L SA

Amplitude variation

θa

R0

A p e r t u r e

Frequency variation

r(x)

PRF 2

x



PRF 2

Azimuth (Doppler) Bandwidth

y

VU 45 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

2D-SAR Image Formation Raw data

Range compressed data

range range

Image data

azimuth

azimuth

azimuth Range reference

Azimuth reference

function

function

point target

range

far range

Near range amplitude

azimuth ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 47 > Autor Name Microwaves and Radar Institute > 30.05.2006

23

Sp. Res: 3 x 3 m ISLR Az: -15.1 dB ISLR Rg: -14.1 dB ISLR 2D: -11.5 dB

MGD-SE ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Spatial Resolution: 3>xAutor 3m VU 48 Name Microwaves and Radar Institute > 30.05.2006

SAR Imaging Modes

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 50 > Autor Name Microwaves and Radar Institute > 30.05.2006

24

VU 51 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 52 > Autor Name

25

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 53 > Autor Name Microwaves and Radar Institute > 30.05.2006

Az

Rg

Comparison: ScanSAR vs. Stripmap vs. Spotlight (TerraSAR-X)

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 54 > Autor Name Microwaves and Radar Institute > 30.05.2006

26

Comparison: ScanSAR vs. Stripmap (TerraSAR-X) ScanSAR (HH) 150 MHz 17 m resolution 1 (az) x 6.9 (rg) looks 7.25 m spacing θincid ≈ 19.7° .. 30.4° ascending orbit

Stripmap (HH) 150 MHz 7 m resolution 3 m spacing 2.9 (az) x 3.4 (rg) looks θincid ≈ 24.8° .. 28.1° descending orbit

3 days time separation

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 55 > Autor Name Microwaves and Radar Institute > 30.05.2006

ScanSAR EEC-RE 17 m res.

illumination

~3 km x 4 km

ScanSAR

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 56 > Autor Name Microwaves and Radar Institute > 30.05.2006

27

Stripmap

VU 57 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Squint Angle & Doppler Centroid • Phase history

v P Platform velocity

φ( t )  2 S y n t h e t I c

where:

θa θs

L SA

R0

A p e r t u r e

r(x)

x

 4π 2π 2π v P2 ( t  t 0 )2   r( t )   R0  λ λ λ R0   t 0 :

R0 sin(θ s ) The scatterer appears vP earlier/later due to the squint angle

• Frequency variation: fa ( t )  fD  κ a t 

2v P sin(θ s ) 2v P2  t λ λR 0

due to the linear due to the quadratic range variation range variation 2v P sin(θ s ) λ

• Doppler Centroid:

fD :

• Doppler Rate:

κ a : 

2 v P2 λ R0

y ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 58 > Autor Name Microwaves and Radar Institute > 30.05.2006

28

SAR Azimuth Signal v P Platform velocity Phase history S y n t h e t I c

Frequency variation

θa θs

L SA

R0

A p e r t u r e

r(x)

PRF  fD 2

x



PRF  fD 2

Azimuth (Doppler) Bandwidth

y

VU 59 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Moving Targets

Microwaves and Radar Institute > 30.05.2006

•Phase history

v P Platform velocity

S y n t h e t I c

Amplitude variation

2 2  4π 4π 2π ( v a  v P ) t  φ( x )   R0  vr t   λ  λ λ R0  

Range variation r(t):

 v L SA

 vr

R0

A p e r t u r e

 va

r(t)  (R0  vr t)2  (vat  vPt)2  R0  vr t 

(va  vP )2 t2 2R0

• Frequency variation: 2 2( v a  v P )2 fa ( t )   v r  t  fD  κa t λ λR 0

r(x) • Doppler Centroid (due to vr): (causes an azimuth shift)

x

2v r λ

2( v a  v P )2 λ R0 (azimuthal velocities changes the doppler rate)

• Doppler Rate:

y

fD : 

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

κa : 

VU 60 > Autor Name Microwaves and Radar Institute > 30.05.2006

29

Tehachapi Loop, California, 19.12.2007 VU 61 > Autor Name Microwaves and Radar Institute > 30.05.2006

azimuth

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

V range

Tehachapi Loop, California, 30.12.2007 ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 62 > Autor Name Microwaves and Radar Institute > 30.05.2006

30

VU 63 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

SAR History History of SAR 1951

Doppler beam sharpening concept

Carl Wiley, Goodyear

1958

First focused SAR Image

University of Michigan

1960

First operational airborne system (UPD-1)

Univ. Michigan/Texas Instrum.

Airborne SAR with digital processing

Military industry, USA

1978

First SAR satellite (SEASAT)

NASA/JPL, USA

1979

Real-time airborne SAR processing

MDA, Canada

Digital SAR processing with workstations

NASA/JPL, USA

1991

First European SAR satellite (ERS-1)

ESA

1994

First demonstration of fully polarimetric, multi frequency spaceborne SAR (SIR-C/X-SAR)

NASA, USA,DLR, Germany / ASI, Italy

2000

First single-pass interferometric mission in space (SRTM)

NASA, USA, DLR, Germany

2002

First dual-polarised SAR satellite (ASAR/ENVISAT)

ESA

2007

Launch of TerraSAR-X (first German civilian radar satellite)

2010

Launch of TanDEM-X

early 1970

early 80s

DLR / Astrium

DLR / Astrium ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 64 > Autor Name Microwaves and Radar Institute > 30.05.2006

31

Air- and Spaceborne Sensors

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 65 > Autor Name Microwaves and Radar Institute > 30.05.2006

Commonly Used Frequency Bands Frequency band

Frequency range

Application examples

VHF

300 KHz -

300 MHz

Foliage/Ground penetration, biomass

P-Band

300 MHz -

1 GHz

soil moisture, biomass, penetration

L-Band

1 GHz

-

2 GHz

agriculture, forestry, soil moisture

C-Band

4 GHz

-

8 GHz

ocean, agriculture

X-Band

8 GHz

-

12 GHz

agriculture, ocean, high resolution radar

Ku-Band

14 GHz

-

18 GHz

glaciology (snow cover mapping)

Ka-Band

27 GHz

-

47 GHz

High resolution radar

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 66 > Autor Name Microwaves and Radar Institute > 30.05.2006

32

Polarimetric Airborne SAR Sensors

AES1

AIRSAR

DOSAR

RENE

AeroSensing (D) GulfStream Commander X-Band (HH), P-Band (Quad)

NASA / JPL (USA) DC8 P, L, C-Band (Quad)

EADS / Dornier GmbH (D) DO 228 (1989), C160 (1998), G222 (2000) S, C, X-Band (Quad), Ka-Band (VV)

UVSQ / CETP (F) Écureuil AS350 S, X-Band (Quad)

ESAR

EMISAR

MEMPHIS / AER II-PAMIR

STORM

DLR (D) DO 228 P, L, S-Band (Quad) C, X-Band (Sngl)

DCRS (DK) G3 Aircraft L, C-Band (Quad)

FGAN (D) Transal C160 Ka, W-Band (Quad) / X-Band (Quad)

UVSQ / CETP (F) Merlin IV C-Band (Quad)

PISAR

RAMSES

PHARUS TNO - FEL (NL) CESSNA – Citation II C-Band (Quad)

NASDA / CRL (J) ONERA (F) GulfStream Transal C160 ETH(Quad) Zü - Earth Zürich – DBAUG - IFU L, X-Band P, L, S, C,Observation X, Ku, Ka, Chair W-Band (Quad)

SAR580 CCRS (CA) Convair CV-580 VU 67 > Autor Name C, X-Band (Quad) Microwaves and Radar Institute > 30.05.2006

Spaceborne SAR Missions since 1978

SEASAT

ERS-1/2

J-ERS-1

SIR-C/X-SAR

NASA/JPL (USA) L-Band, 1978

European Space Agency (ESA) C-Band, 1991-2000 & 1995-today

Japanese Space Agency (NASDA) L-Band, 1992-1998

NASA/JPL, L- and C-Band (quad) DLR / ASI, X-band April and October 1994

RADARSAT-1

SRTM

ENVISAT / ASAR

ALOS / PALSAR

Canadian Space Agency (CSA) C-Band, 1995-today

NASA/JPL (C-Band), DLR (X-Band) February 2000

European Space Agency (ESA) C-Band (dual), 2002-today

Japanese Space Agency (JAXA) L-Band (quad), 2005

CosmoSkymed

TerraSAR-X

SAR Lupe BWB Germany X-Band, 2006

ASI / Alenia German Aerospace Center (DLR) / Astrium X-Band (dual), (quad), 2007 ETH Zü rich – DBAUG - IFU - Earth X-Band Observation Chair Zü2007

RADARSAT-2 Canadian Space Agency (CSA) VU 68 > Autor Name C-Band (quad), 2007 Microwaves and Radar Institute > 30.05.2006

33

ENVISAT/ASAR

Launch: 28.02.2002 MIPAS

MERIS

AATSR SCIAMACHY MWR

GOMOS DORIS RA-2 LRR

Size: 26 x 10 x 5 m, Weight: 8.2 t

ASAR

Most ambitious RS satellite for environmental monitoring Unique combination of 10 RS instruments  ESA ETH Zü Chair Zürich – DBAUG - IFU - Earth Observation

ALOS/PALSAR

Wavelength

0.236 m

Chirp Bandwidth

14 MHz

Peak Transmit Power Duty Cycle Noise Figure Antenna Size (Tx , Rx) Quantisation

Launch date Weight Solar Power Orbit Altitude Revolution Yaw steering Inclination Attitude error

ASAR-Antenna (ca. 10 m x 1.33 m VU 69 > Autor Name and 320 T/R-Modules) Microwaves and Radar Institute > 30.05.2006

January 2006 4000kg ~7Kw@EOL Sun Synchronous 691.65 km 14+27/46 ON 98.16 degrees ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair 0.4e-4°

2 kW 3,5 % (7 % / 2) 4 dB 8.9 m x 3.1 m 5 bit (BAQ)

VU 70 > Autor Name Microwaves and Radar Institute > 30.05.2006

34

Launch date: 14. Dec 2007 Frequency 5.405 GHz (C-band) Repeat cycle: 24 days

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 71 > Autor Name Microwaves and Radar Institute > 30.05.2006

Launch date: 14. June 2007 PPP with EADS ASTRIUM GmbH Wet mass: 1209 kg Size: 5 m height  2.4 m diameter

Solar Generator

Thrusters

X-Band Radar Antenna 384 Transmit/Receive Modules S-Band TM/TC Antenna X-Band Downlink Antenna Data Rate: 300 MBit/sec ETH Zü – DBAUGMass - IFU - Earth Observation Chair Zürich State 256 Gbit Solid Memory

VU 72 > Autor Name Microwaves and Radar Institute > 30.05.2006

35

Information Content -Speckle -Radiometric Resolution -Radar Cross Section -CS’s

VU 73 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

 E

Imaginary

Received Electric Field:  | E | A

 : R



4



Microwaves and Radar Institute > 30.05.2006

 E  A exp( i )

Scattering Amplitude R Phase

Real

R Distance Antenna-Scaterrer

Image Intensity

~

 | E |2  A 2

Single Single scatterer scatterer in in the the resolution resolution cell cell

Deterministic Scatterers (Point Scatterets) ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 74 > Autor Name Microwaves and Radar Institute > 30.05.2006

36

 | E | A

2 : 

4



 E R2

Imaginary

Received ReceivedElectric ElectricField: Field:

 | E | A

1 : 

R1

3 : 

R1

4

R3



Real

 R2

 | E | A

4



 E  A exp( i )

R3

The Themovement movementofofthe thescatterer scattererwithin withinthe the resolution resolutioncell cell isisassociated associated totoaachange changeofof the thephase phaseofofthe thereceived receivedsignal signal… …

… …due duetotothe thechange changeofofthe the Antenna-Scatterer Antenna-Scattererdistance. distance. Image Intensity

~

 | E |2  A 2

is not changing ! VU 75 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Received Electric Field:

 E 

 Am exp( im ) m

The The received received signal signal isis given given by by the the coherent (i.e. vectorial) sum of coherent (i.e. vectorial) sum of the the signals signals received received from from each each scatterer scatterer inin the theresolution resolutioncell. cell.

Im

Am  A

m : 

4



Rm

Re

“Random Walk” Process

m mscatterers scatterers within withinthe theresolution resolutioncell cell

Stochastic Scatterers (Distributed Scatterets) ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 76 > Autor Name Microwaves and Radar Institute > 30.05.2006

37

The The movement movement ofof the the scatterers scatterers within within the the res. res.cell cell isisassociated associated toto aa change change ofof phase phase and and amplitude amplitude ofof the the received receivedsignal. signal.

Im

Re

“Random Walk” Process

Image Intensity

changes !

Stochastic Scatterers (Distributed Scatterets)

VU 77 > Autor Name Assumption: ofof–the resolution than ETH Zü DBAUG - IFU - Earthcell Observation Chair Zürich Assumption:Wavelength Wavelengthλλ 30.05.2006

Phase Image

Probability density function (pdf) of the phase of a SAR image of distributed scatterers with fully developed speckle. pdf (  )

Uniform Distribution [- π,π]: ► No information in the phase of a single SAR image. 



ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 80 > Autor Name Microwaves and Radar Institute > 30.05.2006

39

Fully Developed Speckle

 E

Received Electric Field:

NS

A

m

exp( iφ m )

m 1

When:

The number of scatterers is large within the resolution cell; The scatterers are independent from each other; Amplitude Am and phase φm are independent random variables; The phase φm is uniformly distributed between –π and π;

Then:

Fully developed speckle:

  In terms of the central limit theorem: Re{ E } and Im{ E } are independent random Gaussian variables with zero mean and the same standard variation σ.

 pdf (Re{ E }) 

 1 Re{ E } exp(  ) 2σ σ 2π

 pdf (Im{ E }) 

 1 Im{ E } exp(  ) 2σ σ 2π

  The pdf of the Intensity I  (Re{ E }) 2  (Im{ E }) 2 is

pdf (I) 

The pdf of the Amplitude A  I is

pdf ( A ) 

1 I exp(  ) 2σ 2 2σ 2 A A2 exp(  ) 2 σ 2σ 2

Exponential distributed

Rayleigh distributed

VU 82 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Speckle in Multi-Look Images Multi-Look: Incoherent summation (average) of resolution cells. The average can be performed in the frequency domain … by averaging images of individual frequency parts in range and/or azimuth; or in the spatial (time) domain … by averaging individual resolution cells in range and/or azimuth.

Multi-Look Intensity

IML 

1 N aN r

N a Nr

I

k

Ik

where

k 1

Nr Na

… Intensity of the kth resolution cell … Number of res. cells used in Rg / Az

L  Nr N a … Number of Looks The pdf of the Multi-Look Intensity is:

pdf (I) 

LL IL 1 1 L I exp(  ) (L  1)! ( 2σ 2 )L 2σ 2

Gamma distributed

The same mean 2σ 2 but a reduced standard deviation by a factor L compared to the 1 look intensity

The pdf of the Multi-Look Amplitude is: pdf ( A ) 

2LL A ( 2L 1) LA 2 exp(  ) Gen. Gamma distributed 2 L (L  1)! ( 2σ ) 2σ 2

VU 83of > Autor Name Speckle can be reduced at the expense of spatial resolution by increasing the number looks. ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair Microwaves and Radar Institute > 30.05.2006

40

MGD-SE Spatial Resolution: 3 x 3m ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 84 > Autor Name

Microwaves and Radar Institute > 30.05.2006

MGD-RE Spatial Resolution: 7 x 7m VU 85 > Autor Name (~ 9 Looks)Microwaves and Radar Institute > 30.05.2006

41

What does the Radar measure ? Radar reflectivity (backscattered signal) of the target as a function of position. 1 The radar transmits a pulse (pulse travelling velocity is equal to velocity of light)

2. Some of the energy of the incident radar pulse is reflected back towards the radar. … and is measured by the radar. It is known as Radar Backscatter o (Sigma Nought or Sigma Zero).

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 86 > Autor Name Microwaves and Radar Institute > 30.05.2006

What does the Radar measure ? Normalized radar cross-section (backscattering coefficient) is given by:

o (dB) = 10. Log10 (energy ratio) Whereby: received energy by the sensor energy ratio = “energy reflected in an isotropic way”

The backscattered coefficient can be a positive number if there is a focusing of backscattered energy towards the radar

or The backscattered coefficient can be a negative number if there is a focusing of backscattered energy way from the radar (e.g. smooth surface) ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 87 > Autor Name Microwaves and Radar Institute > 30.05.2006

42

Backscattering Coefficient o

Typical scenario

Levels of Radar backscatter Very high backscatter (above -5 dB)

Man-Made objects (urban) Terrain Slopes towards radar very rough surface radar looking very steep

High backscatter (-10 dB to 0 dB)

rough surface dense vegetation (forest)

Moderate backscatter (-20 to -10 dB)

medium level of vegetation agricultural crops moderately rough surfaces

Low backscatter (below -20 dB)

smooth surface calm water road very dry terrain (sand)

VU 88 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 89 > Autor Name

43

Radar image color composit (Shuttle Imaging Radar-C) Two large ocean eddies next to sea ice in the Weddell Sea, Antarctica, on October 5, 1994, The eddies sweep up loose bits of sea ice, mostly pancake ice, in their rotating currents. Very new ice is seen in the darkest areas. First- year ice is green. Open ocean is blue. The image has a size of 240km by 360 km. L-band VV is blue, L-Band HV is green, and C-band VV is red.

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 90 > Autor Name Microwaves and Radar Institute > 30.05.2006

91 > Autor Name Bergen, Norway, March 13,VU2008

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

44

VU 92 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

Surface Scattering

VU 93 > Autor Name

45

VU 94 > Autor Name

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

VU 95 > Autor Name

46

Multi-Parameter SAR Acquisitions -Multi-Temporal -Multi-Angular -Multi-Frequency -Polarimetric

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 96 > Autor Name Microwaves and Radar Institute > 30.05.2006

F-SAR (DLR), Kaufbeuren, X-Band

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 97 > Autor Name Microwaves and Radar Institute > 30.05.2006

47

F-SAR (DLR), Kaufbeuren, X-Band, vollpolarimetrisch

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 98 > Autor Name Microwaves and Radar Institute > 30.05.2006

Polarimetric SAR

X-band PI-SAR / Test Site: Gifu-Japan

VU 99 > Autor Name R:HH-VV G:HV+VH B:HH+VV

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

Microwaves and Radar Institute > 30.05.2006

48

Frequency and Polarisation Diversity

C-band R: HH G: HV B: VV

L-band P-band VU 100 > Autor Name ETHHH Zü Zürich R: G:– DBAUG HV B:- IFU VV- Earth Observation Chair R: HH G: HV B: VV

C-band

L-band

Beaufort Sea, 1998 AIRSAR / NASA-JPL

Microwaves and Radar Institute > 30.05.2006

P-band R: C-TP, G: L-TP, B: P-TP

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 101 > Autor Name Microwaves and Radar Institute > 30.05.2006

49

Multitemporal

Az Rg

High Resolution Spotlight Image (300 MHz): Australia, Sydney, Harbor Bridge

1. Layer: Dec 12, 2007 2. Layer: Jan 1, 2008 3. Layer: Jan 12, 2008

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 102 > Autor Name Microwaves and Radar Institute > 30.05.2006

Multitemporal

Az

High Resolution Spot Light Image (300 MHz): Australia, Sydney Harbor Bridge

1. Layer: Dec 12, 2007 2. Layer: Jan 1, 2008 3. Layer: Jan 12, 2008

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

VU 103 > Autor Name Microwaves and Radar Institute > 30.05.2006

50

Rg

Drygalski Glacier Oct 2007 – Jul 2008

6 m/day

by Remote Sensing Technology Institute

ETH Zü Zürich – DBAUG - IFU - Earth Observation Chair

19/04/06

RGB Pauli

06/06/06

VU 104 > Autor Name Microwaves and Radar Institute > 30.05.2006

05/07/06

VU 105 > Autor Name Multitemporal L-band Quad-Pol - AGRISAR Campaign 2006 ETH Zü - IFU - Earth Observation(E-SAR) Chair Zürich – DBAUG Microwaves and Radar Institute > 30.05.2006

51