Basics Principles Of Rs

REMOTE SENSING Dr. GEORGE JOSEPH DIRECTOR CENTER FOR SPACE SCIENCE AND TECHNOLOGY EDUCATION IN ASIA AND THE PACIFIC (Affiliated to The United Nations)

WHAT IS REMOTE SENSING? `REMOTE SENSING IS THE SCIENCE OF MAKING INFERENCES ABOUT OBJECTS FROM MEASUREMENTS, MADE AT A DISTANCE, WITHOUT COMING INTO PHYSICAL CONTACT WITH THE OBJECTS UNDER STUDY.’ `REMOTE SENSING MEANS SENSING OF THE EARTH’S SURFACE FROM SPACE BY MAKING USE OF THE PROPERTIES OF ELECTROMAGNETIC WAVE EMITTED, REFLECTED OR DIFFRACTED BY THE SENSED OBJECTS, FOR THE PURPOSE OF IMPROVING NATURAL RESOURCE MANAGEMENT, LAND USE AND THE PROTECTION OF THE ENVIRONMENT.’

REMOTE SENSING IN EVERY DAY LIFE

Visual Perception of Remote Sensing Platform

Sensor

Green : Raw

Interpreter

Yellow : Ripe

Remote sensing is recognizing yellow papaya is ripe

PRESENT YIELD LEVELS

LAND USE PATTERN

PRESENT AREA UNDER CROPS

INCREASE LAND UNDER CROPS

IMPROVE YIELDS

ARABLE WASTELAND

INCREASE FOOD GRAIN SUPPLY

DEVELOPMENT/ CONSERVATION

BUFFER STOCK

IMPORT OF

FOODGRAIN

FOOD GRAIN REQUIREMENT

WORLD PRICES

USE OF HIGH YIELD VARIETY

AGRICULTURAL PRACTICES

SOURCES OF IRRIGATION

ADDITIONAL SOURCES

IRRIGATION AREA INCREASE

IMPROVE AGRICULTURE INPUTS AGRICULTURAL PRACTICES

FERTILISERS FOR YIELDS

INFORMATION REQUIREMENT SCENARIO FOR INCREASING FOOD GRAIN SUPPLY

WHY IMAGING FROM SPACE? SYNOPTIC COVERAGE 800 km

10 km 100 m 10 m 1.6 m 4.5 km 11.3 km

35.7 km

357 km 1956 km

Polar Orbit

WHY IMAGING FROM SPACE? GLOBAL COVERAGE & REPEAT OBSERVATIONS 750 4

600 3

450 300 150 2

1

14

13

12

11

10

9

8

7

6

Latitude

00 150 15

Orbit Number

300 5

450 600 750

2820 km

Sensor

Platform (satellite)

Data transmission

Ground Truth/ Accuracy Check Base Map

Decision criteria/ Visual

Photo products

Earth Surface

Thermal emission

Fluorescence

Cloud

Tables Thematic map

Reports

Funds availability/ Other Database Digital

Reflection

Atmosphere

(absorption/scattering/emission)

Sun

Data Analysis

GIS

Digital products

Decision Making

Data products

Data reception & recording

Periodic Monitoring using RS

Monitoring

Discussions with beneficiaries

Implementation at field level

SCHEMATICS SHOWING REMOTE SENSING SYSTEM FOR RESOURCE MANAGEMENT.

ELECTROMAGNETIC RADIATION IF A CHARGED PARTICLE ACCELERATES (MOVES FASTER, SLOWER OR CHANGES DIRECTION), IT PRODUCES BOTH AN ELECTRIC FIELD (BECAUSE THE PARTICLE IS CHARGED) AND A MAGNETIC FIELD (BECAUSE THE PARTICLE IS MOVING). BECAUSE THE MOTION OF THE PARTICLE IS CHANGING, THE ELECTRIC FIELD IS CHANGING AND THE MAGNETIC FIELD IS CHANGING. THE CHANGING ELECTRIC FIELD CREATES A NEW MAGNETIC FIELD AND THE CHANGING MAGNETIC FIELD PRODUCES A NEW ELECTRIC FIELD. THE COLLAPSING AND REGENERATION OF THE ELECTRIC AND MAGNETIC FIELDS IS WHAT ALLOWS EM RADIATION TO PROPAGATE.

Propagation of EM Waves y z

• Changing B field creates E field • Changing E field creates B field

x

C = nλ

E = hn = hc / λ

POLARISATION A CONDITION IN WHICH ELECTROMAGNETIC WAVES ARE CONSTRAINED TO VIBRATE IN A CERTAIN PLANE OR PLANES THE WAVE IS SAID TO BE POLARIZED. POLARIZATION IS GIVEN BY THE ELECTRIC FIELD VECTOR LINEAR POLARISATION

ELECTROMAGNETIC SPECTRUM

UNITS: WAVELENGTH UNITS:

LENGTH

-10

Angstrom (A) : 1 A = 1x10 m; -9 Nanometer (nm): 1 nm=1x10 m; Micrometer (µm): 1 µm = 1x10-6m; Wave number units: inverse length (often in cm-1)

EM RADIATION DESIGNATIONS Optical Infrared (OIR) region Visible

0.4 – 0. 7 µm

Near Infrared (NIR)

0.7 – 1.5 µm

Short Wave Infrared (SWIR)

1.5 – 3 µm

Mid Wave Infrared (MWIR)

3-8 µm

Long Wave Infrared (LWIR) (Thermal Infrared (TIR)

8-15 µm

Far Infrared (FIR)

Beyond 15 µm

MICROWAVES P band

0.3–1 GHz (30 - 100 cm)

L band

1-2 GHz (15 - 30 cm)

S band

2-4 GHz (7.5 - 15 cm)

C band

4–8 GHz (3.8 - 7.5 cm)

X band

8–12.5 GHz (2.4 – 3.8 cm)

Ku band

12.5–18 GHz (1.7- 2.4 cm)

K band

18–26.5 GHz (1.1 – 1.7 cm)

Ka band

26.5-40 GHz (0.75 - 1.1 cm)

Note : 1 GHz = 109 Hz

REMOTE SENSING PASSIVE

IMAGING CAMERA TV MSMR

ACTIVE

SOUNDING

VTPR

IMAGING

SOUNDING

SLAR

LIDAR

SAR

ENERGY SOURCE FOR PASSIVE SENSING

Spectral distribution at the top of the atmosphere for solar irradiance and earth’s emission. Sun T ~6000°K, Earth T ~300°K 104 1000

The radiant exitance from sun and earth follows Planck’s equation

Mλ =

Power radiated at each wavelength

Radiant exitance (W m-2 µm-1)

REFLECTION OF SOLAR RADIATION& EMISSION

Sun

100 10 1

0.1

0.5

1

Earth

10

Wavelength (µm)

10.5 – 12.5 µ is used for THERMAL IMAGING to avoid the ozone absorption at ~9.5 µm

λmax =

[

C1

λ5 eC

2

λT

2898 T

]

−1

µm

For earth, λmax ~9.5 µm

At MICROWAVE frequency C2/ λT <<1 Mλ ∝ εT → BRIGHTNESS TEMPERATURE

ATMOSPHERIC WINDOWS 0.4-1.3, Sun

Absorption by

1.5-1.8,

Data transmission

UV

VISIBLE

µm 3.0-3.6,

MW

Far IR-MW

Vibrational Transitions (H2O,CO2)

Rotational

4.2-5.0,

Forbidden Transitions (O2)

Transitions (H2O)

1cm-30cm INFRARED

MICROWAVE X

1.0

4

5

C

S

5

10

H2O

O2

O2

CO2

O3

3

H2O

1.5 2

CO2

CO2 H2O

H2O H2O 0.5µm

H2O

0

THERMAL IR

REFLECTED IR

RADIO

H2O

Earth Surface 100

7--- 15,

emission

Electronic Transitions (O3,O2) Thermal

Ionization Dissociatio n (O,N2,O2,O3)

Fluorescence

Cloud

2.2-2.6,

IR

UV-VIS-NIR

Reflection

UV

O3

(absorption/scattering/emission) ATMOSPHERIC TRANSMISSION (%)

Atmosphere

molecules

10 15 20µ 0.1 cm WAVELENGTH

0.5

1.0

L

INFLUENCE OF ATMOSPHERE IN MEASUREMENTS

• Molecular Absorption Only window regions of EM Spectrum available • Molecular/Aerosol Scattering Modifies the Spatial/Spectral distribution of incoming and Outgoing Radiation • Atmospheric Turbulence Limits resolution

• REFLECTANCE SPECTRA 80

R E F L E C T A N C E (%)

SILTY

60

CLAY SOIL

VEGETATION 40

MUCK SOIL

20

WATER (Shallow/Deep) 0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.6

Wave length (µm)

4.8

FRESH SNOW GREEN VEGETATION DARK TONED SOIL LIGHT TONED SOIL CLEAR WATER TURBID WATER

GREEN BAND

RED BAND

NEAR IR

SHORTWAVE IR

(0.52-0.59 µm)

(0.62-0.67 µm)

(0.77-0.86 µm)

(1.55-1.75 µm)

COLOUR FORMATION YELLOW (MINUS BLUE)

RED

GREEN

CYAN (MINUS RED)

MAGENTA (MINUS GREEN)

BLUE

COLOUR COMPOSITES

Natural Colour BLUE+GREEN+RED

False Colour Composite GREEN

BLUE

RED

GREEN

NIR

RED

BASIC ELEMENTS OF VISUAL INTERPRETATION TONE (COLOR) SIZE AND SHAPE TEXTURE AND PATTERN RELATIVE& ABSOLUTE LOCATION SHADOWS

SIGNATURE Key to feature identification from space imagery depends on the characteristic changes in the properties of the EM spectrum reflected/emitted from the target surface – referred ‘signature’ Signatures could be inferred through: • SPECTRAL VARIATION • POLARISATION CHANGE • THERMAL INERTIA • TEMPORAL VARIATION

80 R E F 60 L E C 40 T A N 20 C E (%) 0

SILTY CLAY SOIL VEGETATION

MUCK SOIL WATER (Shallow/Deep) 0.4

0.8

1.2

1.6

2.0

2.4

Wave length (µm)

7000

Signatures are not completely deterministic; they are statistical in nature with a mean and dispersion

Frequency

6000 5000 4000 3000 2000 1000 0 2.79

3.08

3.37

Radiance

3.66

3.95

(mw/cm2/

4.24

4.53

str/µm)

4.82

Band 3 Histogram (.52-.59 micron)

Band 4 Histogram (.77-.86 micron)

Barren

Crop

100 50 0

40 45 50 55 60 65 70 75 80 85 90 95 100105110

Pixel count

Water

150

200 150 100 50 50

50 55 60 65 70 75 80 85 90 95 100 105110 115

Grey Level Values

Scatter plot Band 3 versus Band 4 120

Crop, Barren

Water

Grey Level Values

Scatter plot Band 3 versus Band 4 120

Crop

105

Crop

Barren

105

90

Barren

75 60

Band 4

Band 4

Pixel count

250 200

90 75

Urban

60

Water

45 40

50

60

70

80

Band 3

Water

45 90

100 110

40

50

60

70

80

Band 3

90

100 110

Clusters

IMAGING MODES

SENSOR PERFORMANCE PARAMETERS SPATIAL RESOLUTION

A MEASURE OF DISERNABLE PHYSICAL DIMENSION OF THE SURFACE FROM THE IMAGE

SPECTRAL RESOLUTION

A MEASURE OF THE WIDTH OF THEWAVELENGTH (BANDWIDTH) WHICH IS USED TO GENERATE THEIMAGE. NARROWER THE BANDWIDTH HIGHER THE SPECTRAL RESOLUTION

RADIOMETRIC RESOLUTION

• A MEASURE OF WHAT IS THE MINIMUM CHANGE IN RADIANCE THAT CAN BE MEASURED •NE Delta Refl/Radiance/Temp

TEMPORAL RESOLUTION

•FREQUENCY OF OBSERVATION: NUMBER OF DAYS BETWEEN TWO CONSECUTIVE OBSERVATION FOR A PARTICULAR GROUND TARGET UNDER SIMILAR VIEWING GEOMETRY

OTHER SENSOR PARAMETERS OF INTEREST ARE THE NUMBER OF SPECTRAL BANDS, THE POSITION OF THE CENTRAL WAVELENGTH ON THE EM SPECTRUM FOR EACH BAND.

RADIOMETRIC RESOLUTION -

A MEASURE OF THE INSTRUMENT’S CAPABILITY TO DIFFERENTIATE SMALLEST CHANGE IN REFLECTANCE/ EMITTANCE

NOISE EQUIVALENT DIFFERENTIAL RADIANCE

NEΛL

NOISE EQUIVALENT DIFFERENTIAL REFLECTANCE NEΛP NOISE EQUIVALENT DIFFERENTIAL TEMPERATURE NEΛT DEFINED AS THE CHANGE IN RADIANCE/REFLECTANCE/ TEMPERATURE WHICH GIVES A SIGNAL OUTPUT OF THE SENSOR EQUAL TO THE NOISE AT THAT SIGNAL LEVEL

IDEAL FILTER

Intensity

SPECTRAL RESOLUTION 1

RESPONSE = 1 for λc + ∆λ/2 ≤ λ ≥ λc - ∆λ/2 0

= 0 for λc - ∆λ/2 λ ≥ λc + ∆λ/2

λ1

λc

λ2

Wavelength

Bandwidth ∆λ = Full Width at Half Maximum (FWHM)

Intensity

GAUSSIAN FILTER 1

∆λ

0.5

Desirable to have Sharp Roll Off and Roll On to improve out of band contributions.

%Transmissi on

PRACTICAL FILTER

λ0 Wavelength

Wavelength Spectral response of nma practical filter (LISS-B3).

INFORMATION CONTENT VS RESOLUTION

A) OCM (360m)

E) 36m (LISS-II)

B) 360m (OCM)

C) 188m (WiFS)

F) 23m (LISS-III)

D) 72m (LISS-I)

G) 5.8m (IRS 1C PAN)

. `A’ is from a scene from IRS Ocean Colour Monitor (OCM). The area in the small square marked (≈ 4km x 4km) is shown in various resolutions from B to G..

APPLICATIONS OF DIFFERENT RESOLUTION

• 1M+ SCALES

• 1:500K SCALES

• 1:250K SCALES

1m

• 1:50K SCALES

• 1:12500 SCALES • 1:2000/4000/1:8000 SCALES

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

97

103

104

105

106

107

10

109

110

111

82

113

114

115

116

117

11

119

120

121

82

123

124

125

126

127

19

129

130

131

92

133

134

135

136

137

17

139

140

141

91

143

144

145

146

147

99

149

150

151

15

153

154

155

156

157

78

159

DN Value

Enlarged 10 Times

DATA PRODUCTS Input is a digital data

IMAGE RECTIFICATION AND RESTORATION

RADIOMETRIC CORRECTION GEOMETRIC CORRECTION NOISE REMOVAL GEOREFRENCING

IMAGE CLASSIFICATION Assign each pixel to a class based on signature of surface materials belonging to that class The image classification process involves the subdivision of feature space into homogenous regions Scatter plot separated by decision boundaries. Band 3 versus Band 4 1

120

Fallow Forest Water

Wheat

Crop

105

Barren Band 4

2 3 4

90

Barren

75 60

Water

45

Multispectral image

Classified map

40

50

60

70

80

90

100 110

Band 3

THAT IS THE RADIOMETRIC INFORMATION CONTAINED IN THE IMAGE IS CONVERTED TO THEMATIC INFORMATION, SUCH AS VEGETATION TYPE,FOREST,BUILTUP AREA, WATER BODIES Etc

Image Classification

Input data (Digital)

classification process

Output data (Thematic)

‘’A classification is not complete until the accuracy is assessed ‘’

Sub-satellite track

Range direction

SAR Swath

Azimuth direction

Schematics showing the projection of the side-looking radar beam on ground.

STEREO IMAGING dP2

dP1

Negatives

f

h dP = H−h B

(H-h)

h ~ H

dP B/H

dP - parallax difference between the points B - length of airbase H - flight altitude h

dP2

dP1

True base (Datum)

dP

B

Geometry of a stereoscopic pair of aerial photographs. B is the air base (absolute parallax). dP is the difference in parallax from top and bottom of the object.

SPOT HRS

Latitude

70 km

754003 60 450 30002 15 00 0 15 15 300 450 6000 75

1

14 13 12 11 10 9

8 7

Orbit Number

OFF-NADIR VIEWING CAPABILITY

6

5 2820 km

Sampling the Spectrum NIR

BGR 400 nm

700

SWIR 1500

MWIR

3000

LWIR 5000

LOW Panchromatic: one very wide band

MED

Multispectral: several to tens of bands

HIGH Hyperspectral: hundreds of narrow bands

14000 nm

MULTISPECTRAL - HYPERSPECTRAL SIGNATURE COMPARISON Multispectral

Resampled to Landsat TM7 Bands

APPLICATIONS OF REMOTE SENSING FOR EARTH RESOURCES MANAGEMENT TO IDENTIFY THE CATEGORY TO WHICH THE EARTH SURFACE EXPRESSION (MANIFESTED AS DATA) BELONGS, BASED ON SIGNATURE DIFFERENCES. THE SURFACIAL EXPRESSIONS ARE INDICATORS OF CERTAIN RESOURCES, WHICH ARE NOT DIRECTLY OBSERVABLE BY REMOTE SENSING. TO INFER A PARTICULAR PARAMETER OR PHENOMENON (WHICH IS ONLY PARTLY REPRESENTED IN THE DATA) USING SUITABLE MODELLING (YIELD OF A CROP, VOLUME OF TIMBER FROM FOREST, OCEAN CURRENTS, ETC.)

i

Changes in Quilon district (Kerala)

KAKKI RESERVOIR

Landsat 4 MSS (29th Jan 1983)

5

N 0

(1: 250, 000) SCALE (Approx)

KAKKI RESERVOIR

IRS-1D LISS III (28th Feb 2002)

5 Km

LAND

AGRICULTURE & SOIL ¾ Crop Acreage & Production Estimation ¾ Soil & Land Degradation Mapping ¾ Watershed Development ¾ Horticulture FOREST, ENVIRONMENT, BIO ¾ Forest Cover & Type Mapping ¾ Forest Fire and Risk Mapping ¾ Biodiversity Characterisation ¾ Environmental Impact Studies

¾ Landuse/Land Cover ¾ Wasteland Mapping ¾ Urban Sprawl ¾ Large Scale Mapping WEATHER & CLIMATE ¾Extended Range Monsoon Forecasting ¾Ocean State Forecasting

DISASTER SUPPORT

WATER

¾ Flood Damage Assessment ¾ Drought Monitoring ¾ Land Slide Hazard Zonation

¾ Potential Ground Water Zones ¾ Command Area Management ¾ Reservoir Sedimentation

OCEAN ¾ Potential Fishing Zone (PFZ) ¾ Coastal Zone Mapping

EARTH OBSERVATION – APPLICATIONS

GEOGRAPHIC INFORMATION SYSTEM (GIS) ™

TO ARRIVE AT A DECISION WE NEED TO INTEGRATE DATA FROM VARIOUS SOURCES

™

THE DATA SOURCES INCLUDE SPATIAL AND ATTRIBUTE INFORMATION

™

GIS IS A COMPUTER BASED TOOL FOR END TO END PROCESSING FOR A DECISION SUPPORT SYSTEM CAPTURE, STORAGE, ANALYSIS (QUERIES), RETRIEVAL, DISPLAY

Land use slope soil

Various IRS Payloads Specifications Satellite Payload

IRS-P6

AWiFS

Band B2, B3,B4 (VNIR) B5 (SWIR)

Swath

IGFOV

740

56-70

(km)

(m)

Resourse sat-1 Oct 03

B2, B3,B4 (VNIR) ,B5(SWIR) LISS 3* LISS 4

Cartosat-1 FORE May- 05 AFT Cartosat-2 Carto-2 Jan-07

B2, B3, B4 (VNIR)

PAN

141 23.5 23.5 Mx 70 Mono

30

5.6 2.5

PAN

27

2.5

PAN

12

<1

UNIVERSITIES PRESS ISBN 81 7371 535 8

Rs. 425

Image Classification Methods

Supervised Classification

Unsupervised Classification

Distribution Free

Statistical Techniques

Euclidean classifier K-nearest neighbour Minimum distance Decision Tree

based on probability distribution models,

Nearest Mean Classifier (Minimum Distance Classifier)

Advantages: – mathematically simple – computationally efficient

Disadvantages: – insensitive to different degrees of variance in the data (point 2)

Enlarged 10 Times

OPTOMECHANICAL SCANNERS

Ad = a2

a a

Ground target

х

х

SPACE PLATFORM R O T

D R S

Studio HUB

1.

1. S8 Im W

S2 Im TW L L

T L L

APPLICATIONS OF SATELLITES

color violet

(Å)

f (*1014 Hz) Energy (*10

19 J)

4000

4600

7.5

6.5

5.0

4.3

indigo 4600

4750

6.5

6.3

4.3

4.2

blue

4750

4900

6.3

6.1

4.2

4.1

green

4900

5650

6.1

5.3

4.1

3.5

yellow 5650

5750

5.3

5.2

3.5

3.45

orang e

5750

6000

5.2

5.0

3.45

3.3

red

6000

8000

5.0

3.7

3.3

2.5

V I B G Y O R

TYPICALGROUND TRACK OF A RS SATELLITE

Kumbh Mela Site

Kumbh Mela Site

As seen through IRS PAN on 15th April, 2000

As seen through IRS PAN on 15th January, 2001

THE DUAL WAVE-PARTICLE NATURE OF EMR

EMR has both particle and wave properties It is made up of photons which are packets (quanta) of energy The energy E of a photon is directly proportional to its frequency u and inversely proportional to its wavelength λ E = hu = hc/ λ where h = Planck’s constant = 6.6 x 10-34 Js c = speed of light = 3 x 108 m s-1 Why UV is more dangerous compared to visible A UV photon has a shorter wavelength and hence higher energy than a yellow photon. UV photons cause more skin damage than optical photons

UNITS: WAVELENGTH UNITS:

LENGTH

Angstrom (A) : 1 A = 1x10-10 m; Nanometer (nm): 1 nm=1x10-9 m; Micrometer (µm): 1 µm = 1x10-6 m; Wavenumber units: inverse length (often in cm-1)

SUPERVISED SUPERVISED CLASSIFICATION CLASSIFICATION FEATURE SELECTION. ONCE THE TRAINING STATISTICS HAVE BEEN SYSTEMATICALLY COLLECTED FROM EACH BAND FOR EACH CLASS OF INTEREST, A JUDGMENT MUST BE MADE TO DETERMINE THE BANDS THAT ARE MOST EFFECTIVE IN DISCRIMINATING EACH CLASS FROM ALL OTHERS.

CLASIFICATION MULTIVARIATE STATISTICAL PARAMETERS (MEANS, STANDARD DEVIATIONS, COVARIANCE MATRICES, CORRELATION MATRICES, ETC.) ARE CALCULATED FOR EACH TRAINING SITE. EVERY PIXEL BOTH WITHIN AND OUTSIDE THE TRAINING SITES IS THEN EVALUATED AND ASSIGNED TO THE CLASS OF WHICH IT HAS THE HIGHEST LIKELIHOOD OF BEING A MEMBER.

ATMOSPHERIC WINDOWS

Sun

0.4-1.3, 1.5-1.8,

Absorption by

….;’.;.’,: ;:

molecules

Electronic Transitions (O3,O2)

µm 3.0-3.6, 4.2-5.0,

MW

Far IR-MW

Vibrational Transitions (H2O,CO2)

VISIBLE

Rotational Transitions (H2O)

7--- 15,

Forbidden Transitions (O2)

INFRARED

1cm-30cm MICROWAVE

1.0

4

5

C

S

5

10

H2O

O2

O2

CO2

O3

3

H2O

1.5 2

CO2

CO2 H2O

0.5µm

H2O

0

H2O H2O

100

THERMAL IR

REFLECTED IR

RADIO

H2O

X O3

ATMOSPHERIC TRANSMISSION (%)

Ionization Dissociation (O,N2,O2,O3)

UV

IR

UV-VIS-NIR

UV

2.2-2.6,

10 15 20µ 0.1 cm WAVELENGTH

0.5

1.0

L