Remote Sensing
Definition:
“Remote sensing is the science of acquiring
information about the Earth’s surface without
actually being in contact with it, by sensing and
recording reflected or emitted energy and
processing, analyzing and applying that
information”
Simplest Form of Remote Sensing: Aerial Photography
Cartographers take detailed measurements from aerial
photos in the preparation of maps
Geographers interpret aerial photos to determine land-use,
and environmental conditions
Aerial photographs are not maps
Maps: Directionally and geometrically accurate
Aerial photograps: Radial distortion
GIS’s can account for radial distortion
The Electromagnetic Spectrum
Electromagnetic radiation is energy propagated through
space between electric and magnetic fields. The
electromagnetic spectrum is the extent of that energy
from cosmic rays, gamma-rays, X-rays, ultraviolet,
visible, infrared and microwave energy
Electromagnetic waves can be classified by:
FREQUENCY or WAVELENGTH
Velocity = Speed of light
Electromagnetic Radiation
Consists of electrical field
(E) and magnetic field (M)
Travels at speed of light (C)
The shorter the wavelength,
the higher the frequency
This is important for
understanding information
obtained in remote sensing
Microwaves are longest
wavelengths used in
remote sensing
We are blind to
everything except
this narrow band
UV are shortest
wavelengths practical
for remote sensing
Remote Sensing requires the following:
1. Electromagnetic Energy Source
2. Interaction with a Target
3. Sensor to Record Energy
4. Transmission, Reception and Processing
5. Interpretation
6. Application
Electromagnetic Energy Source
Illuminates or provides electromagnetic radiation to the
target of interest
A-
Source of
Electromagnetic
Radiation
B-
Radiation comes
into Contact
with Atmosphere
C-
Radiation
Interacts with
Target
D-
Sensor Collects and
Records Electromagnetic
Radiation
E-
Recorded Energy Transmitted G to Processing Station (for copy)
F-
Processed Image
Interpreted to
Extract Target Info
Application
Radiation Interacts with Atmosphere
Radiation interacts with the atmosphere on the way to the
target and as the energy travels from the target to the sensor
Interaction with a Target
Radiation interacts with target. The nature of this interaction
is dependent on the wavelength of the radiation and the nature of
the target
Sensor
A sensor (mounted on satellite/plane/helicopter) collects/records
the electromagnetic radiation scattered or emitted by the target
Transmission and Processing
Recorded energy is transmitted to a processing station to produce
an image saved in digital format (or hardcopy)
Interpretation
Visual interpretation or digital (GIS) interpretation to extract
further information about the target
Application
Information applied to solve a problem
* Remote sensing is especially important for
extracting information from harsh environments
or difficult terrain
Transmission through the Atmosphere
Some wavelengths of
E-M energy are
absorbed and scattered
more efficiently than
others
H2O, CO2, and ozone
have the strongest
absorption spectra
Transmission
Light moves through a
surface
Wavelength dependent
(eg. leaves)
Radiation emitted from Earth is of
a much longer wavelength and is of
much lesser energy
Atmospheric Windows
In the diagram below, peaks are windows, while troughs
identify wavelengths that are heavily absorbed
WINDOWS
ABSORPTION
Scattering
Passive Sensors
Measure naturally-available energy
(eg. thermal infrared radiation
emitted from the Earth 24 hours
per day, but solar reflected
radiation only during solar day)
Active Sensors
Sensor emits radiation toward target
Reflected radiation in emitted bands
are detected and measured
(eg. microwaves emitted)
When electromagnetic energy strikes molecules or other
tiny objects, one of three things happens:
1.
2.
3.
REFLECTION
ABSORPTION
TRANSMISSION
Remote sensing is most concerned with REFLECTION
of radiation (or emission from the Earth).
Reflectance
The ratio of the amount of electromagnetic radiation
reflected from a surface to the amount originally striking
a surface
• Specular Reflection
Surface is smooth relative to incident wavelength,
resulting in mirror-like reflection (reflected in single
direction)
May help or hinder remote sensing depending on
where the sensor is situated
• Diffuse Reflection
Surface is rough relative to incident wavelength
Energy is scattered more or less evenly in all directions
Many natural surfaces exhibit a great deal of diffuse
reflection
Surface types yield distinct spectral responses
Examples
Water:
Longer visible wavelengths absorbed more than
shorter visible wavelengths (blue-green)
Leaves:
Chlorophyll strongly absorbs in blue and red,
but reflects green (green colour results)
Healthy leaves efficiently reflect near IR
Absorption vs. Reflection differ for different
wavelengths of electromagnetic radiation
Leaves
Water
Spectral signatures
Degree to which an object reflects incident electromagnetic
energy in different regions of the electromagnetic spectrum
Characteristic signatures can be obtained for specific land
surface classes
Multispectral sensors detect reflectance in more than one
band
Characteristic spectral responses of different surface types. Bands are those
of the SPOT remote sensing satellite.
Visual Interpretation in Remote Sensing and
Aerial Photo Interpretation
The following visual elements are considered in identifying objects:
Tone (Hue or Colour)
Brightness or colour of elements on an image.
Nothing could be discerned without changes in brightness or colour.
Shape
Form, structure or general outline of the objects.
Regular shapes usually indicate human presence and land-use.
Size
Absolute size and size relative to background objects (in context of
scale of the image)
Texture
Smoothness or roughness
Arrangement and frequency of tonal variation (eg. A forest is
rough while an asphalt or cement surface is smooth)
Pattern
Spatial arrangement of objects gives a clue to object character
(eg. random pattern in forest area vs.
Shadow
May help determine relative height but may also hinder
interpretation obscuring objects within
Association and Site
Relationship between target and other objects: context of an
object may lead to its identification
Understanding of stratigraphy alters geomorphological
interpretation of landscape features
Images and Photographs
Representation in digital format
by subdividing image into equallyshaped areas called pixels
The ‘brightness’ of
each area can be
attributed a numeric
value or digital
number
Information from
narrow wavelength
ranges can be stored
in channels, also
called bands
Often, data from multiple channels can be
represented as one of three primary colours
which combine according to brightness.
We are, thus, no longer blind to these ’s.
Orbits and Swaths
Geostationary orbits:
Very high altitude satellites (approximately 36 000 km)
Focus on the same area of the Earth at all times
Continual data collection over a specific area
Eg. Weather and communications satellites
Near-polar orbits
Satellite travels northward on one side of the Earth and then
southwards during the second half of its orbit
In sun-synchronous orbits, ascending path can be on a shadowed
side with the descending path on the sunlit side. Passive sensors
would only record data during the descent.
Swath
The area imaged on the surface.
Swaths vary from very small areas (helicopters and planes) to
hundreds of kilometres (spaceborne satellites)
Earth rotates: Satellite swath may cover new area with each pass
Complete coverage of Earth after one cycle of orbits
Areas at high latitude generally covered more frequently
Spatial Resolution
Size of the smallest possible
feature that can be detected
Instantaneous Field of View (IFOV)
is the angular cone of visibility of
the sensor (See A at right)
This, along with altitude (C), determines
the area visible on the ground (B)
Examples of Remote Sensing Satellites
Each has multiple channels for specific purposes
1.
Weather
GOES (Geostationary Operational Environmental Satellite)
NOAA AVHRR (Advanced Very High Resolution Radiometer)
2.
Land Surface Observation
Landsat (NOAA)
SPOT (Système Pour l’Observation de la Terre)
IRS (Indian Remote Sensing)
MEIS-II and CASI (Airborne Sensors)
3.
Marine Observation
CZCS (Coastal Zone Colour Scanner)
MOS (Marine Observation Satellite)
SeaWiFS (Sea-viewing Wide Field of View Sensor)
Applications of Remote Sensing
There are many applications of remote sensing, most of which are
related to Geography as a discipline
Agriculture:
Forestry:
Hydrology:
Land Use:
Oceans:
Mapping:
Crop type, condition and yield, soil characteristics
Type, health, biomass, burning, species, deforestation
Sea ice, navigation, oil spills, sea surface temperature
Resource management, habitat protection, urban
sprawl, damage assessment, legal boundaries
Currents, winds, waves, phytoplankton concentration,
temperature monitoring, navigation routing, traffic
density, bathymetry, land-water interface delineation,
coastal vegetation
Digital Elevation Models (DEM’s), thematic mapping
AVHRR
Visible, NIR, Thermal
1.1 km Resolution - local area
coverage (LAC)
4 km Resolution - global area
coverage (GAC)
Used for meteorological studies
Vegetation pattern analysis
Global modeling
Broad spectral bands
LANDSAT Thematic Mapper
Sun-synchronous, near-polar orbit, imaging the same
185 km x 0.474 km ground swath every 16 days
Global coverage between 81 degrees north latitude and
81 degrees south latitude
Particularly useful in determining land use classes
Blue/Green, Green, Red, NIR, MIR, Thermal
30 meter resolution 256 brightness values
7 spectral bands
Normalized Difference
Vegetative Index (NDVI)
NDVI = (NIR - red) / (NIR + red)
RADAR - Radio Detection and Ranging
Passive Microwave Sensors:
Applications include meteorology (atmosphere profiles, water and
ozone content), hydrology (soil moisture) and oceanography (sea
ice, currents, oil slicks)
Active Microwave Sensors:
RADAR - Sensor transmits a microwave (or radio) signal toward a
target and detects the backscattered portion of the signal
Strength of backscattered signal discriminates between targets
Time delay between transmitted and reflected signals determines
the distance to the target
Non-Imaging (eg. altimeters) or Imaging Sensors
Imaging Microwave Sensors include RADARSAT (Canada, 1995)
RADARSAT, developed by the Canadian Space Agency, is the
world’s first, operationally-oriented radar satellite system capable of
rapid delivery of large quantities of data
Image Processing
1.
Preprocessing
Radiometric and geometric corrections
2.
Image Enhancement
Improving contrast, and spatial filtering to enhance specific
spatial patterns of interest
3.
Image Transformations
Combined processing of multiple spectral bands for image
enhancement
4.
Image Classification and Analysis
Digital identification and classification of pixels.
Classification: Assigns each pixel to a particular class or
theme based on desired statistical characteristics
(supervised or unsupervised)
Before GIS:
Popularity of
stack maps
Limitation:
Restricted to
consistent scale,
projection and
coverage area
Advantage of Digital Overlay:
1.
Faster
2.
Scale, projection and coverage area
less problematic (Most applications consist of
sources collected by different methods and at
different scales)
3.
Time and error associated with manual
integration and redrafting eliminated
Raster or Vector Implementation
Raster
Implementation
of
Overlay
Overview of Overlay Analysis
1.
Three maps represented with a common
grid
2.
Binary maps converted with Boolean
operators such as AND and OR
Eg. Suitability Analysis
AND =
more than one condition must
occur simultaneously
OR =
identifies areas with either
condition met
Boolean logic: truth tables
Exclusionary approaches rely on boolean logic, where the
value of the statement, "A is true AND B is true" is
determined in a truth table indicated the individual
permutations of A and B:
AND
B is true
B is false
A is true
T
F
A is false
F
F
The value of the statement, "A is true OR B is true" under
the same circumstances:
OR
B is true
B is false
A is true
T
T
A is false
T
F
Boolean Logic
in Raster Overlay
It is often useful
to RECLASS
your data before
performing
OVERLAY
RECLASS
OVERLAY
Task: Given vegetation map and elevation map,
isolate a vegetation type within a particular
altitude range
Map 1: Vegetation Map
(VEGMAP)
Map 2: Digital Elevation Model
(DEM1)
This requires the use of the AND operation
STEP 1:
Our vegetation of interest is class 6.
Reclass to assign a value of 1 for all values from 6 to just less than 7.
All other values are assigned a value of zero.
CLASS6
STEP 2:
Use the Digital Elevation Model to isolate the elevations
between 1700 and 1800 m.a.s.l. In the same way, assign a value of
1 to all values from 1700 to just less than 1801 by using a reclass
function.
DEM1718
Where do the two isolated characteristics coincide ?
Use OVERLAY
Multiply file CLASS6 by DEM1718 to produce the output map
RESULT.
Only pixels with a value of 1 survive to be represented in the output file.
RESULT
OVERLAY is often used in combination with other
operations such as near-neighbour operations
Eg.
Produce a map of riparian vegetation cover within
100 metres of rivers and streams
Locate buffer zone
100m from rivers
Overlay with
vegetation map
Produce resultant
map of riparian cover
Vector Implementation of Overlay
Produces many new polygons due to overlapping
Each new polygon has a unique, new identifier
The identifier is linked to an attribute table
Result is a single layer coverage linked to all attributes