  Lecture 16  
SATELLITE IMAGERY
1. Satellite Images for Mapping
This technology initially evolved from the Cold War space race. First TV image from space. We see them now on a daily basis in
news and weather reports. hurricane Katrina flood images animation: world air traffic
a. Photography and multi-spectral scanning
(mss)
Scanning can be used at any wavelength, and does not require the
satellite to return to earth for film processing: the data are
transmitted back to earth from the satellite.
orbits b. Earth monitoring satellites
These have one of two types of orbital paths:
- the first is most important for mapping
Sun-synchronous: the satellite orbits the
earth, 9 degrees off the polar axis, between
9.30-10.30am local time.
(Landsat) Altitude ~ 700-900km - swath path- 'earthnow'
Geo-stationary: the satellite rotates
with the earth, so that it is always in the same position relative to
the earth. This is the case with our weather satellites (such as NIMBUS
and GOES). Altitude ~ 36,000km
Spatial Resolution (pixel size) ~1 metre to 1 kilometre
Low resolution (usually free) ..................1km - 10km (international) - suitable mostly for weather
Medium resolution (usually free) 100m -1km (national) - suitable for < 1:250,000
High resolution (costs less) ..............10 -100 metres (regional) - suitable for 1:50,000 - 1:250,000
Very High resolution (costs more) 1 - 10 metres (Local) - suitable for > 1:50,000
2. The Landsat Program
a. Multi-Spectral Sensor (MSS): 1972 - 1982
The first Landsat satellite was launched in 1972. Satellites 1,2,3 had
a Multi-Spectral Sensor (MSS) scanning in the green, red and two near
infra-red wavelengths. Pixel size was 80 metres in resolution.
This was suitable for mapping and map updating at 1:250,000 and
smaller scales, example : flooded areas in Quebec
b. Thematic Mapper (TM): 1982 and 1984 - present
Landsat satellites 4 and 5 have the MSS sensor but also the higher
resolution (30 metres) 'Thematic Mapper' (TM). This has 7 bands:
blue, green, red, near IR, two bands in the mid-IR and one in the
far (Thermal) IR (120m resolution). The two main advantages over MSS are that the mid-IR is included and the pixels are 30 metre instead of 80
metre.
c.
Enhanced Thematic Mapper (ETM+) 1999 - [2003]
Landsat 7 (1999) is similar to the earlier TM
sensor, but data cost dropped 80%
and copyright was removed, enabling websites where scenes
can be downloaded. [The sensor malfunctionned in April 2003.]
Landsat 7 data are the basis of Google maps and Google Earth, and the mosaics in imap and www.mapplace.ca
3. Satellite Imagery versus Aerial Photography
Table : Satellite images versus
Aerial Photos
|
Advantages of satellite imagery
|
Characteristics of aerial
photographs
|
|
Greater Areal Extent
|
Higher Resolution
|
|
Digital data
|
Analogue photos
|
|
Repetitive Coverage
|
Lower cost of 'launch'
|
|
Regular (predictable) Distortion
|
3D Stereo Effect
|
|
Greater Wavelength Range
|
Higher understanding
|
|
Analysis / GIS
|
Easier interpretation
|
4. Image display and enhancement
a. Digital images
The range of digital numbers (DN) is generally 8 bit data =
256 values: this gives each image dataset a range from 0 to 255.
Therefore with 8 bit data DNs per channel and with three channels
displayed (in red-green-blue), you could have up to 2563 = 16, 777, 216 colour combinations (and this is why it 'looks like a
photo')
b. Contrast stretch
Digital images are initially low in contrast and features and need
enhancement. The poor contrast results from the gray tones or color
assignment rarely matching the Digital Number (DN) range (0-255) not filling the potential display range. In order to enhance
the image the reflectance values can be stretched to match those of the
potential display - a contrast stretch. Nifty example of Shroud of Turin
c. Colour composites
A colour composite is a three band colour image with each band
projected through one of the RGB colour guns; thus only three bands can be
displayed simultaneously.
TM bands 1,2,3 in the BGR
guns respectively will produce a "normal colour" image
Bands 2,3,4 will produce the equivalent of 'false' colour film (healthy vegetation displayed in red)
Bands
3,4,5 will produce the highest contrast image, with one band each in
the visible, near-IR and mid-IR.
see also resmap
5. Other Satellites
a. SPOT (Systeme Probataire
d'Observation de la Terre)
This series of satellites was first launched in 1986 by France and
provides three multispectral bands with 20 metre pixels in the red,
green and near IR wavelengths, creating a colour composite equivalent
to 'false-colour' film. A fourth band is 'panchromatic' across the
visible bands at 10 metre resolution, almost equivalent to high
altitude aerial photos. Generally, the higher resolution of SPOT did
not compensate for the extra bands in Landsat, although the latest
SPOT-5 satellite (May 2002) includes the mid-IR, and even higher
resolution (2.5m or 5m).
SPOT home page www.spot.com In Canada, Iunctus distributes SPOT data
b. ASTER
(Advanced Spaceborne Thermal Emission and Reflection)
This sensor launched in 2000 aboard the
Terra satellite is 'Landsat-like' but with 15 metre
resolution ASTER homepage
The Terra satellite also carries a lower (250m-500m-1km) resolution sensor for regional observation MODIS homepage
c. GOES (Geostationary Orbital Earth Satellite)
These are the weather satellites, also with 1 kilometre pixels. A
series of geo-stationary satellites cover the globe, yielding several
earth images per day.
GOES home page www.goes.noaa.gov
Example of passive microwave (not affected by clouds): Sea temperature
d. RADARSAT
RADARSAT is Canada's first environmental satellite launched in 1995, and dedicated entirely to RADAR wavelengths.
It was aimed mostly at monitoring arctic sea ice, as RADAR is not
affected by cloud cover or darkness. Radarsat 2, with higher spatial
resolution is hoped to be launched in 2007.
RADARSAT: antarctica RADARSAT 2: radarsat 2
Canada Centre for Remote Sensing: CCRS
.
e. Very high resolution: IKONOS and
QUICKBIRD
Ikonos was the world's first corporate imaging
satellite (the others are launched by national government agencies)
It was
launched in 1999, followed by Quickbird in 2001. These images are often
used in news media to show dramatic events such as the 2004
Tsunami and Hurricane Katrina.
Ikonos: 4metre multispectral / 1m Panchromatic UNBC campus 1m / 4m
Quickbird: 2.4m multispectral / 60cm Panchromatic Quickbird sample imagery
These provide most of the high resolution data seen in Google Earth and Google Maps
Google has sponsored its own satellite in 2008, the highest resolution sensor (50cm) Geoeye
General satellite image: gallery .... Earth as Art Gallery
These satellites may be good, but they cannot see
everything and monitor your movements as portrayed in movies such as Enemy of the State or
by now they should have found Osama
bean laden
|
