This guide provides a basic introduction to remote sensing, satellite imagery
and Geographic Information Systems (GIS). Click on the links below to view
the relevant section.
The past two to three decades has seen a revolution in our ability to survey
and map our global environment. Digital sensors mounted on satellites scan
vast areas of the earth's surface every day and night. Constellations of satellites
beam out signals, which enable us to accurately and rapidly position ourselves,
and computers store and process quantities of geographical data, which previously
would have been completely unmanageable.
Landsat 7, the latest addition to the Landsat earth
observation satellite program, orbits at 705km above the earth. With onboard
recorders the satellite can store data until it passes within range of
a ground station. Basic geometric and radiometric corrections are then
applied before distribution of the imagery to users.
Satellite Imagery & Remote Sensing
With their continuous development and improvement, and free from national
access restrictions, satellite sensors are increasingly replacing surface and
airborne data gathering techniques. At any one point in time, day or night,
multiple satellites are rapidly scanning and measuring the earth's surface
and atmosphere, adding to an ever-expanding range of geographic and geophysical
data available to help us manage and solve the problems of our human and physical
environments. Remote Sensing is the science of extracting information from
such images.
Satellite Orbits
Most earth observation satellites, (such as the Landsat,
SPOT and IRS series) are in a near polar, sun-synchronous orbit. At altitudes
of around 700-900km
the satellites revolve around the earth in approximately 100 minutes and
on each orbit cross a particular line of latitude at the same local (solar)
time.
This ensures the satellite can obtain coverage of most of the globe, replicating
the coverage typically within 2-3 weeks. With sensors which can be pointed
sidewards from the orbital path, revisit times with high-resolution frames
can be reduced to just a few days. Due to evaporation, the atmosphere normally
contains less moisture early in the morning, so to get a clear picture whilst
achieving sufficient solar illumination, satellites often are set to overpass
at around 9:30 - 10:30 a.m. local time.
Landsat 7 makes over 14 orbits per day, in its
sun-synchronous orbit. During the full 16 days of a repeat cycle, coverage
of the areas between those shown is achieved.
Exceptions to these sun-synchronous orbits include the geostationary meteorological
satellites (such as the Meteosat and GOES satellites). These have a 36,000km
orbit and rotate around the earth every 24 hours remaining above the same point
on the equator, acquiring frequent images showing cloud and atmospheric moisture
movements for almost a full hemisphere. Also, satellites required to obtain
very high resolution (<2m) images which often orbit at altitudes of around
200-300 km are currently not able to operate in a sun-synchronous orbit.
Digital Sensors
Although still in operation today, early satellite designs involve images
being exposed to photographic film and returned by capsule to earth for processing.
However, even the first commercial satellite imagery, from Landsat-1 launched
in 1972, used digital imaging sensors and digitally transmitted the data back
to ground stations.
A Passive sensor is one that records the radiation reflected or transmitted
from the earth. Usually termed optical sensors, these measure in the visible,
near infrared, middle infrared and thermal infrared wavelengths. An 'Active
sensor' is one that transmits it's own microwave radiation, which is reflected
from the earth's surface back to the satellite and recorded.
Optical scanning techniques take one of two main forms. A 'switch-broom' sensor
consists of an oscillating mirror, which directs the reflected light to a few
sensors, building up an image by a few lines of picture elements (pixels) at
a time. However, more common in modern sensors, is the 'push-broom' sensor,
an array of several thousand CCD-sensors each recording a single path (column),
which combine, building an image as the satellite orbits the earth.
Whichever scanning method is used, each satellite records an image of constant
width but potentially several thousand kilometres in length. Once the data
have been received on earth, the imagery is usually split into approximately
square sections for distribution. These nominal image sizes typically range
from 11x11km (IKONOS, 1m pixels) to 185km x 170km (Landsat-7, 15m and 30m pixels).
Individual picture elements (pixels) are the grid of points where the surface
brightness is recorded by a satellite, and which comprise the continuum of
lines and columns of an image. The ground spacing between pixels defines the
image's 'pixel size'. Because the 'spatial resolution' of an image is defined
as the separation that two point features on the ground must have in order
for them to be determined as spatially separate, the resolution of a digital
image is theoretically double the 'pixel size' of the image. However, the term 'spatial
resolution' is now usually used, particularly for optical imagery, to specify
the pixel size of a digital image.
Spectral Resolution, Wavebands and False Colour Composites
The spectral capabilities of a satellite sensor are determined by the ranges
of wavelengths (wavebands) recorded and the precision that the amount of reflected
energy can be measured (quantised). A sensor recording the energy of a single
waveband produces a greyscale (panchromatic) image whilst multispectral images
are produced from sensors simultaneously recording multiple wavebands. The
Landsat Thematic Mapper sensor, for example, records in 7 spectral bands.
Because we can only display and view images using the three visible primary
colours (red, green and blue) any three of the wavebands available need to
be chosen to highlight the particular features of interest. When any combination
other than the visible bands are used, the resulting image created is a 'false
colour composite'.
Materials covering the surface of the earth reflect or absorb the sun's radiation
in a characteristic manner (spectral signature). From the visible wavelengths
(blue-green-red) this gives a material a specific colour which we see. For
a multispectral sensor, recording many wavebands spread over a far greater
range of wavelengths, discrimination between materials is made much more apparent.
The
selection of wavebands for a satellite to record is governed not only by wavelengths,
which aid the discrimination of spectral signatures, but
also
by the incident solar radiation and the wavelengths absorbed and scattered
by atmospheric moisture and gasses. The energy reaching the earth from
the sun peaks in the visible yellow wavelengths and reduces more rapidly
for shorter
than for longer wavelengths.
This, together with strong scattering by atmospheric molecules, such as ozone
and water, makes very short wavelengths, (ultra-violet and x-rays) of little
use to satellite remote sensing of the earth's surface. Although atmospheric
scattering of blue light is greater than for other visible wavelengths, the
blue waveband can still be relatively free from haze intemperate and arid climates,
or at high altitudes, above the relatively thin, low layer of atmosphere containing
most atmospheric moisture. The longer wavelengths, into the infra-red, suffer
less from scattering and so give a clearer image. However, care in selecting
wavelengths to record still has to be taken to avoid the narrow bands of absorbed
radiation, such as at 1.4mm and 1.9mm, where transmission through
the atmosphere is almost completely blocked by water vapour and carbon dioxide.
Geographic Information Systems (GIS)
In order to obtain a clearer understanding of the potentially vast quantity
of data relating to a particular portion or aspect of the earth and its environment,
a Geographic Information System (GIS) enables any available geo-spatial data
to be compiled, analysed and presented.
Many aspects of our lives now benefit from the use of GIS. From the management
and maintenance of the networks of pipelines and cables that supply our homes,
to the exploitation or protection of the natural resources we use. The emergency
services frequently use GIS to ensure resources are allocated to provide adequate
cover where it is needed and also to help establish patterns or causes for
any problems that might occur.
Commercial companies can use demographic and infrastructure data within a
GIS to plan marketing strategies, identify where their service would be most
needed and decide where to locate their business. Insurance companies can use
GIS to determining premiums based on population distribution, crime figures
and likelihood of natural disasters such as flooding or subsidence.
Whatever the application, all the geographically related information that
is available can be input and prepared in a GIS, such that a user can display
the specific information of interest, or combine data contained within the
system to produce further information which might answer or help resolve a
specific problem. From analysis of data that has been acquired, it is often
possible to use a GIS to generate a 'model' of possible future situations and
to see what impact might result from decisions and actions taken.
With a GIS, complex maps can be created and edited much faster than would
be possible by hand, and because the data is stored digitally, the maps are
produced with the same level of accuracy each time.
A GIS can utilise a satellite image to extract useful information and map
large areas, which would otherwise take many man-years of labour to achieve
on the ground. For industrial applications, including, hydrocarbon and mineral
exploration, forestry, agriculture, monitoring of the environment and urban
development, such dramatic and beneficial increases in efficiency have made
it possible to evaluate and undertake projects and studies in parts of the
world which were previously considered inaccessible.
It is such a wide range of commercial uses which has helped to drive the
development of the GIS software and make it the extremely widespread tool for
processing and managing complex, inter-related geographical data.
This guide provides a basic introduction to remote sensing, satellite imagery
and Geographic Information Systems (GIS). Click on the links below to view
the relevant section.
Because we can only display and view images using the three visible primary
colours (red, green and blue) any three of the wavebands available need to
be chosen to highlight the particular features of interest. When any combination
other than the visible bands are used, the resulting image created is a '
The
selection of wavebands for a satellite to record is governed not only by wavelengths,
which aid the discrimination of spectral signatures, but
also
by the incident solar radiation and the wavelengths absorbed and scattered
by atmospheric moisture and gasses. The energy reaching the earth from
the sun peaks in the visible yellow wavelengths and reduces more rapidly
for shorter
than for longer wavelengths. 