AGE 400 Remote Sensing and Resource Management

KENYATTA UNIVERSITY- INSTITUTE OF OPEN LEARNING
AGE 400
REMOTE SENSING AND RESOURCE MANAGEMENT MODULE
CHRISTOPHER MISATI ONDIEKI

&
DAVID ABEL MAGIYA OSEBE
DEPARTMENT OF GEOGRAPHY
KENYATTA UNIVERSITY
©JUNE 2005
1
TABLE OF CONTENTS PAGE
1.0 LESSON ONE: INTRODUCTION TO REMOTE SENSING 03
2.0 LESSON TWO: SIGNIFICANCE AND HISTORY OF REMOTE

SENSING 12
3.0 LESSON THREE: THE ELECTROMAGNETIC SPECTRUM 25
4.0 LESSON FOUR: INTERPRETATION AND SOURCES OF ERROR

IN IMAGE DETECTION 37
5.0 APPLICATION OF REMOTE SENSING IN AGRICULTURE 44
6.0 LESSON SIX: APPLICATION OF REMOTE SENSING

IN NATURAL RESOURCES MANAGEMENT 50
7.0 LESSON SEVEN: SATELLITE WEATHER AND CLIMATE

FORECASTING 61
8.0 LESSON EIGHT: INTRODUCTION TO GEOGRAPHIC

INFORMATION SYSTEMS (GIS) 68
9.0 LESSON NINE: GIS COMPONENTS AND FUNCTIONALITY 82
10.0 LESSON TEN: SPATIAL DATA MODELS 93
11.0 LESSON ELEVEN: APPLICATIONS OF GEOGRAPHICAL

INFORMATION SYSTEMS 103
2
1.0 LESSON ONE: INTRODUCTION TO REMOTE SENSING
1.1 Introduction
This lesson deals with fundamental concepts of remote sensing. Here, remote sensing is
defined as the collection of data by a device, which is not in direct contact with the
phenomena to be investigated. Other concepts that are covered by this lesson include
types of remote sensing, the definitions of satellites, resolution and multi-concept. The
lesson is concluded with advantages and disadvantages of camera and satellite systems.
1.2 Learning Objectives

By the end of this lesson, you should be able to:
(a) Define remote sensing.
(b) Describe two types of remote sensing.
(c) Describe the various types of resolution and multi-concept.
(d) Describe satellites.
(e) Describe the advantages and disadvantages of photographic cameras and
satellites.
1.3 FUNDAMENTAL CONCEPTS

1.3.1 Remote Sensing
Remote sensing means feeling, measuring or imaging sensations from near or distant
objects, areas and others through the analysis of data acquired by a device that is not in
direct contact with the phenomena to be investigated. The remotely collected data can
be of many forms including:
(a) Acoustics: A branch of science that deals with sound-reflecting character at an

interval space by recording, reinforcement, hearing, or perceptions of sounds and
hearing aids. (b) Force of gravity: One of the four fundamental forces of nature, the
other three being the electromagnetic force, the weak nuclear force, and the strong
nuclear force. The gravitational force is the weakest of the four forces, but it acts over
great distances. Gravity is force of attraction that arises among objects by virtue of their
masses. On Earth, gravity is the force of attraction between any object in the earth’s
gravitational field and the earth itself. (c) Seismology: The study of earthquakes, the
3
seismic waves they produce, the processes that cause them and the effects they have.
By examining the global pattern of waves produced by an earthquake, seismologists
can deduce the nature of the materials through which they have passed. This leads to an
understanding of the earth’s internal structure. On a smaller scale, artificial earthquake
waves generated by explosions or mechanical vibrators can be used to search for sub-
surface features such as oil or mineral explorations. (d) Radio astronomy: The branch
of astronomy that deals with the origin and nature of emissions from extraterrestrial
sources in the radio waves range of the electromagnetic radiation rather than in the
visible range. It specifically deals with celestial phenomena through measurement of
the characteristics of radio waves emitted by physical processes occurring in space. (e)
Forensic medicine: The use of scientific techniques to solve criminal cases. It is a
multidisciplinary field embracing chemistry, physics, botany, zoology and medicine
including identification of human bodies or traces. Ballistics, the study of projectiles
such as bullets is another traditional forensic field that makes use of such tools as the
comparison microscope and electronic microscope. Traditional methods such as
fingerprinting are still in use. Among other techniques is the electrostatic document
analysis (ESDA), used for revealing the indentations on paper, which helps in
determining if documents have been tampered with. (f) Electromagnetic energy
distribution: energy rays that are either emitted or reflected as radiation by objects.
Satellite sensors collect this energy, but what the detectors capture is only a small
portion of the entire electromagnetic spectrum.
Note that: These are all good examples of remote sensing because the useful data from
these branches of science are through the analysis of data acquired by a device that is
not in direct contact with the phenomena to be studied.
In environmental sciences, remote sensing is understood as the acquisition, recording

processing and classification of data obtained through the use of electromagnetic
radiation sensors. Here, remote sensing is understood as the imaging sensations of the
earth's atmosphere, land and water surfaces from the outer space. Normally, this space
is cloud-free and has limited air, and therefore the personnel sensing the area below it
has to carry oxygen masks.
4
It should be stated here at the outset that useful environmental data obtained from
exploitation of acoustic waves, force fields and others from sensors that are not
designed to produce any form of imagery from their fields of view are excluded from
the scope of this module. However, air photographic interpretation, in which much
of the information is acquired using the electromagnetic energy without actually setting
foot on the ground, part of the general field of remote sensing will form the bulk of the
module.
1.3.2 Types of Remote Sensing

Two types of remote sensing are recognized namely: active and passive remote sensing.
(a) Active Remote Sensing

This type of remote sensing uses imaging sensors that supply their own source of
energy to illuminate features of interest. Part of this illumination is reflected back to the
sensor for recording. All active sensor systems such as radar operate in the microwave
portion of the electromagnetic spectrum (EMS), namely from about 1 millimeter to 1
meter wavelength range (lesson 3).
(b) Passive Remote Sensing

In passive remote sensing, the sensor system relies upon the naturally available energy,
normally from the sun. Many passive sensor systems operate from the visible to the
infrared regions of the electromagnetic spectrum, namely from about 0.25 µ to 2.3 µ
wavelength bands.
A simple and common example of an active system is a camera used with flash bulbs.
When the same camera is used in sunlight without the flash bulbs, it becomes a passive
sensor system.
1.3.3 Satellite
This is a body in space orbiting around a larger body such as the moon that orbits
around the earth. Manmade satellites such as the telecommunication satellite are those
put in orbit around a planet, which relays back to the earth telephone messages or radio
5
and television (TV) signals received from any part of the earth. It should be noted that a
satellite image is not different from a TV signal (picture) for both are built up and
reproduced electronically from dot-like elements called pixels. Since pixels are the
smallest item of information that can be registered by the satellite sensors, it determines
the resolution of the image. In other words, any object size that may be smaller than
the pixel of the real world cannot be detected by the satellite sensors/scanners even if its
radiometric and spectral resolution are greater than the surrounding areas.
Remote sensing satellites use two types of orbits: geostationary and polar.
Geostationary satellites are put into orbit 35,900 km directly above the equator. At this
altitude the speed of the satellite exactly matches the earth's rotation. To all intends and
purposes, therefore, the satellite is geo-stationary-hovering continuously over the same
part of the earth. Consequently, geo-stationary satellites are used either to transmit
telecommunications signals or to get a very broad view of the weather.
Polar orbits, on the other hand, are lower in altitudes, 600-1500 km above the earth
surface. In a polar orbit, a satellite follows a north-south path close to the poles, and
gradually traverses every part of the earth over a period of days. It then starts the same
cycle again, and since in effect it constantly retraces its steps, it is ideally placed to
view and monitor environmental changes. Some of these changes include seasonal
rhythms, for example, or short-terms events (floods, forest fires, pollution and the like)
or changes in land-use patterns in farming and urban areas.
The second generation of American LANDSAT Satellites returns to the same point on
the globe every 16 days. The orbits are sun-synchronous, i.e. the satellites cross the
equator always at about 9.45 AM local time on the ground below.
1.3.4 Resolution
Resolution is an indication of the sharpness of images or it is the ability of a sensor to
show details of a phenomenon. In other words, it is an expression of the size of the
smallest object that can be depicted. The term is most commonly associated with the
raster data model where the resolution of a raster or grid is equal to the size of the cell
6
in the real world. For example, the resolution of a remotely sensed image may be 10m
(meaning each cell is representing 10mx10m on the ground). Increased resolution leads
to larger storage requirements, increased processing and higher costs for a given area.
Many things affect resolution including pixel counts, format, receiver quality, lenses,
and lighting, among others. The quality of image is measured by the quality of
information extractable from it, which is strongly influenced by three types of
resolution namely spatial (ground), radiometric and spectral.
(a) Spatial Resolution

This is the ability of the lens-film-filter combination to image distinctly small details. In
other words, it is the smallest feature that is covered by an instantaneous field of view
(IFOV) and can be seen from any a given altitude. Many IFOVs form the building
blocks of imagery.
(b) Spectral Resolution

It is the capability of a remote sensing system to picture ground objects that reflect light
of different intensities (colors). It encompasses the width of the regions of the
electromagnetic spectrum and the number of channels that are sensed. Sometimes
chromatic aberrations may prevent certain light intensities from reaching the sensor
system, and thus preventing important images from being registered. A chromatic
aberration is a defect in an optical image, and every sensor system has some degree of
aberration.
(c) Radiometric Resolution

This is the sensitivity of a sensor system to differentiate signal strength.
(d) Temporal Resolution

It is defined as the length of time intervals between measurements. It is usually
recommended that the data for temporal resolution be obtained on the same time of the
same date for comparative analysis.
7
1.3.5 Multi-Concept
The remote sensing specialist uses different definitions that are meant to enhance
resolution, thereby incorporating what is known as the “multi-” approach: thus the use
of multi-spectral, multi-scale, and multi-date, among others:
(a) Multi-spectral refers to the use of different parts of the electromagnetic spectrum,
and is synonymous with multi-band.
(b) Multi-date is the acquisition of data at different moments in time, and is

synonymous with multi-temporal.
(c) Multi-scale is the use of remotely sensed data at different scales and hence from
different altitude. The multi-scale approach is imperative when extrapolation of
large-scale photography is used in conjunction with satellite imagery.
(d) Multi-polarization refers to the various ways that comprise a beam or an

illumination of energy usually from the sun that can be considered to be vibrating in
all sensor platforms as they travel through the atmosphere and impinge on the
earth’s surface. However, the light that is reflected back into the atmosphere from
one kind of feature on the earth surface such as a body of water may be strongly
polarized (that is, vibrating primarily in one platform) while that reflected back
from some other kind of feature such as vegetation or fractured rocks may be
polarized only slightly, if at all. Since the polarizing capability of a feature is a clue
to its identity, more information about earth resource features can be obtained from
multi-polarization sensor than from only one polarization.
1.4 Advantages and Disadvantages of Cameras and Satellites

The detection and capture of electromagnetic energy can be performed either by
cameras or satellites. The camera systems offer many advantages that include:
They are relatively simple in operation.

They provide a high degree of spatial resolution and geometric integrity. That is,
they have superior resolution capabilities and little rectification is required.
8
Minimal space and power are required.
The equipment’s cost is low.
The disadvantages are:
The camera systems have a limited range of the spectrum in which they operate,
ranging from UV (0.25 µ) to near infrared (0.9 µ).
They have a problem of operating large amounts of data.
Satellites are all electronic sensors that generate an electrical signal that corresponds to
the energy in the original scene. They have the following advantages:
The information from satellites is in electronic form; therefore, relatively easy to

radio it down to ground receiving stations for reconstruction or reconstitution into
computer generated images.
They operate in a wider spectral range (UV through infrared and microwave).
Unlike cameras that have one great problem of weather, the satellites circumvents the
problem of weather by ignoring the use of visible and infrared as sources of
information. Since satellites are active systems, they are all-weather system and are in
operation for 24 hours.
Other advantages include:
They have a practical output, either digital or analog for computer processing,
Further, they have a greater range in detectors than films in the camera,
They are fast in processing and analysis of data by computers,
It is relatively easier to obtain quantitative radiometric data from calibrated
scanners, and
It is easier to compare data of different scales and wavelengths.
Disadvantages:
Satellites have lower resolution.
9
The equipment costs are higher.
1.5 Summary
Remote sensing is defined as a means of collecting data/information from objects,
areas or conditions by a device that is not in direct contact.
Active and passive are two types of remote sensing.
A satellite is defined as a body in space orbiting around a larger body such as the
moon that orbits around the earth.
Resolution as an ability of a sensor to show details of a phenomenon, determines
the image quality, which in turn is strongly influenced by three types of resolution:
spatial, radiometric and spectral.
Multi-spectral, multi-date, and multi-scale photography, among others enhance
resolution of images.
The camera and satellite systems offer many advantages and disadvantages.
1.6 Review Questions

1. Define the following:
Remote Sensing
Air Photo-Interpretation.
2. Distinguish between the following:

Radiometric and Spectral resolution
Active and Passive Remote Sensing.
Geostationary and Polar-Orbiting Satellites.
3. What are the advantages and disadvantages of photographic cameras and

satellites?
4. Do you think satellite images can replace aerial photographs in earth resource
mapping? Justify your answer.
1.7 Further Reading

B.N.S.C. (1989): The Earth Below: A Look at Remote Sensing; Department of Trade
and Industry, London: Moore and Mathers.
Carter, J.D. (1986): Remote Sensing Source Book: A Guide to Remote Sensing
Products, Services, Facilities, Publications and Other Materials; London:
Macanta ltd.
Cracknell, A. and L. Hayes (1993): Introduction to Remote Sensing; London: Taylor
10
and Francis.
Hota, K.R. ed. (1973): The Surveillant Science: Remote Sensing of the
Environment; Boston: Houghton Mifflin Co.
Janza, J.F; Blue, H.M. and J.E. Johnston (1974): Manual of Remote Sensing;
American Society of Photogrammetry, Virginia: Falls Church.
Lillesand, T.M. and R.W. Kiefer (1994): Remote Sensing and Image Interpretation;
New York: John Wiley & Sons Inc.
Rabenhorst, T.D. and P.D. McDermott (1989): Applied Cartography: Introduction
to Remote Sensing, Ohio: Merrill Publishing Co.
Reich, H. (1966): The World from Above. New York: Hills Wang.
Thomas, L.F. (1987): Classification of Remotely Sensed Images, Bristol: Adam
Hilger.
Sabina, F.T. Jr. (1978): Remote Sensing Principles and Interpretation; San
Francisco: W.H. Freeman & Co.
11
2.0 LESSON TWO: SIGNIFICANCE AND HISTORY OF REMOTE SENSING
2.1 Introduction
In this lesson, significant advantages related to remote sensing that use the
electromagnetic spectrum (EMS) sensor systems over other systems are discussed.
History of remote sensing, which was envisaged in c.400 BC is also described. Further,
in this lesson major historical developments of remote sensing from space that earnestly
begun between 1946 and 1950, when small cameras were carried aloft the rockets at the
White Sands Proving Ground in New Mexico, America are traced.

By the end of studying this lesson, you should be able to:
(a) Describe the significance of remote sensing.
(b) Trace the major historical development of remote sensing.
2.3 Importance of Remote Sensing

Modern remote sensing is directed to the surface of the earth. Such sensing ordinarily is
designed to facilitate the management of the earth’s resources in an efficient manner. The
rationale for Remote Sensing can be expressed in a four-part statement:
Whether viewed on a local, regional, national or global basis the human demand for
most kinds of earth resources is rapidly increasingly at the time when the supply of them
is rapidly dwindling and the quality of others is rapidly deteriorating.
Belated to the situation is a call for the wisest possible management of these
resources.
Wise management is greatly facilitated if timely, and accurate inventories are
periodically made available to the resource managers so that they will know reasonably
well at all times the amount and condition of each kind of resource that is present in each
portion of the earth that they seek to manage.
Almost invariably, such inventories can best be made by means of remote sensing,
i.e., by obtaining aerial photography and satellite imagery, and related data from aircrafts
and spacecrafts during periodic over-flights.
12
Significant advantages related to remote sensing that uses electromagnetic spectrum
sensor systems over other systems are:
(a) Improved Vantage Point

From the air an observer’s field of view is tremendously increased. Depending on the
clarity of the atmosphere as well as other factors such as altitude, an observer may be
able to see and detect critical objects or conditions several kilometers away. Aerial RS
provides a bird’s-eye-view of large areas, which enables the observer to see features on
the earth surface in the context.
(b) Increased Spatial Resolution and Geometric Fidelity

With proper selection of camera, film and other flight parameters, it is possible to record
more spatial detail on an imaging system (photograph or imagery) than can be seen with
the unaided eyes. This is particularly important when viewed under magnification. With
proper ground reference data, accurate measurements of positions, distances, areas,
heights, volumes and slopes can be obtained easily from air.
(c) Broadened Spectral Sensitivity

Human vision is spectrally limited. Other RS imaging systems, however, can “see” and
record over a wavelength range so many times broader than can be seen with unaided
eyes. The human eye operates only in the vision region (0.4-0.7 microns) range, whereas
other imaging systems operate in the visible, infrared and microwave portions of the
EMS. Hence the possibility of some sensors recording phenomena in the form of visible
image that cannot be seen with unaided eye.
(d) Records
Satellite imagery and air photographs are virtually permanent records of the then existing
conditions. These records can be studied at leisure under office environment rather than
field conditions. In this way a single image can be studied by a large number of end
users. One area in Kenya where the records of aerial images have been extensively used
is the study of soil erosion. For example in their book, More People Less Soil Erosion,
13
Gichuki, et al (1994) have used aerial images to demonstrate that many parts of the
former larger Machakos District had been subjected to serious soil erosion before
independence. They have compared these images with the post-independence ones, and
noted that the post-independence Machakos has not had any serious soil erosion. Thus
coming up with the title of the book: “More People Less Soil Erosion” because if one
compares the two periods, the post-independence has had more people than pre-
independence when there was serious soil erosion in Machakos.
(e) Other important uses are capability of “Stop Action,” economy, illustration, and
communication.
2.4 History of Remote Sensing

Socrates envisaged the RS concept in c.400BC, when he observed that: man must rise
above the clouds to the top of the atmosphere in order to understand the
environment in which he lives…” This prediction by Socrates might have formed the
basis of remote sensing:
Nicephore Niepce is credited with the invention of the photographic camera in 1822,
thereby giving humankind the opportunity to record vision. In 1858 a camera was hauled
aloft a balloon in the air and pointed downward to take the first aerial photograph. A
Pevisian photographer, Gaspard later in 1858 used a balloon to ascend to a height of 80m
from where he took a “bird’s-eye view” photograph of Petit Bicetre in France. After these
hilarious events, balloon photography flourished widely.
Wallace (1860) is also remembered for having taken from a balloon, the earliest existing
aerial photograph over Boston. Although kites were initially used for collecting
meteorological data, they later became useful platform from which aerial photography
could be launched. Archibald (1882) is credited for having obtained the first aerial
photography from the kites.
The aeroplane invented in 1903 was not used as a camera platform until 1909 when a
“biosphere” motion picture photographer accompanied Wilbur Wright to take picture
14
over Centocelli in Italy during one of Wright’s training sessions for Italian naval officers.
Thereafter, photography from aircraft received heightened attention in the interest of
military strategists who realized the potential of this medium in the acquisition of military
information. During the two decades between 1st and 2nd Wars, many non-military
applications of aerial photography appeared. Civil engineers were among the first people
to use aerial photography for non-military purposes when they begun using it in the early
19th century as a means for surveying and mapping. In the USA, the Department of
Agriculture began nation-wide application of aerial photography for soil survey during
the 1930s. George Eastman, perhaps more than any other individual pioneered in the
improvements of photographic processes, which made photography widely available to
the average person. This began after the 2nd World War when specialized cameras were
developed for improved photography.
The practice of using a human interpreter to analyze photographs obtained from airborne
cameras was well established by 1940’s. Further progress towards both and color infrared
photography greatly improved the effectiveness of aerial camera as an important sensor
of the earth surface. It is important to point out that most advances in active remote
sensing (air photography) were initiated because of actual or impending wars or conflicts,
which is true even today. False color film, originally called “camouflage detection film”
was developed to enable quick, and accurate discrimination between live vegetation and
recently cut foliage or green-painted canopies used to cover military targets. This
technology is now used to help identify crop types, vegetation under stress, water quality,
and haze-penetration.
Active remote sensing systems that use the microwave portion of the spectrum were
earnestly used for the first time during the World War II. As the acronym radar suggests-
radio detecting and ranging, its primary function was to detect and determine the
distance and speed of approaching ships and aircrafts. Later technological advances led to
recording of radar images directly on the film. Today, radar imagery is collected and
analyzed by oil companies for mapping geological structures, especially in tropical
environments where cloud cover inhibits conventional remote sensing systems.
15
2.5 Remote Sensing from Space
The use of rockets as platforms for remote sensing equipment began in early 1900s, when
rockets equipped with cameras were launched over German landscape. However, it is
estimated that earnest remote sensing must have begun between 1946 and 1950 when
small cameras were carried aloft the rockets (V-2) that were shot at White Sands Proving
Ground in New Mexico. Over the succeeding years numerous flights involving
photography were made by rockets, ballistic missiles, satellites and manned spacecraft.
The photographs produced during the early flights were generally of inferior quality
because early missions were made primarily for purposes other than photography. But
crude as they were by today's standards, the early photographs demonstrated the potential
value of remote sensing from space.
Up to mid-1960's, aerial photography using photographic film to record visible light

and near infra-red energy reflected by objects at the earth's surface constituted the
dominant remotely sensed image products. Thereafter, increasingly more versatile and
sensitive non-photographic systems were designed to record both reflected and emitted
energy such as the LANDSAT to produce imagery, the final product of non-
photographic systems. This development coincided with the design and launch of
automatic satellites that dramatically increased the capability of remote sensing.
Beginning with TIROS-1 in 1960, the early weather satellites took imagery of coarse
views of cloud patterns with unclear images of the earth surface. With the
improvement in the imaging sensors aloft the meteorological satellites, images of both
atmospheric, and terrestrial objects and conditions became clearer. The meteorological
satellites were finally directed to conduct intensive studies on surface areas of the earth
to collect data on water, snow and ice features. And this was the time "when looking
through the atmosphere" had just begun.
The features of remote sensing became more exciting as manned space programmes
(Mercury, Gemini, Apollo, Skylab and Space Shuttle) were developed. In 1961, a sub
orbital Mercury flight was made, on which excellent photographs were taken. These
photographs showed only the sky, clouds and ocean because they mainly aimed at
16
capturing them. In 1962, three historic orbits were made around the earth and Mercury
Mission took colored photographs, MA-6. The photographs showed mostly clouds and
water, and most of the deserts of North-west Africa. On later Mercury Missions,
reversal photographs were taken with 70 mm lens Hasselblad Cameras. A specially
modified Hasselblad camera with an 80 mm lens soon became the workhorse for space
photographic experiments conducted in the Gemini Programme.
Mission GT-4 of Gemini included the first formal photographic experiment from space
specifically directed at geology. This imagery soon led to new and exciting discoveries
in tectonology, volcanology and geomorphology. With the success of the programme
in geology, subsequent missions included a host of similar experiments aimed at
investigating various oceanic and topographic phenomena. By the end of Gemini
programme, nearly 2500 photographs, mostly oblique views had been taken with more
than 1000 of these providing high quality colored imagery for earth resource
applications. Hence, the significance of remote sensing from space had become well
recognized during this programme.
Thereafter, serious thinking began on systematic and repetitive coverage of the globe.
The introduction of Apollo programme extended the knowledge of space photography
among the scientific community. The Apollo flights were mainly dedicated to the
objective of lunar landings, but Apollo 6 and 9 (1967) included experiments in
automatic color photography and multispectral orbital photography for earth resource
studies that proved important in the later development of the LANDSAT satellite
system.
In 1973, the first American Space Workshop, SKYLAB was launched to accommodate
a wide range of space-related experiments. Its astronauts took over 35,000 scenes of
the earth with the earth Resources Experiment Package (EREP) on board. The EREP
incorporated a multi-camera array with long focal lengths, a 13-channel multi-band
scanner and two microwave systems. The multi-camera array provided imagery at
spatial resolutions of between 30 m and 80 m that are still useful to researchers and is a
function of the size of the data collection applications specialists. It should be
17
mentioned that the EREP Programmes were the first to demonstrate the complementary
nature of photography and electronic imaging from space.
In early 1975, US, and USSR conducted a joint venture, Apollo-SOYUZ TEST (ASTP)
on a space station experiment with a remote sensing component. In this venture, earth
resource imaging was not a primary goal and for various reasons, the overall qualities
of most of the images from the ASTP were disappointing. Like SKYLAB, the ASTP
mission demonstrated that trained crewmembers could obtain useful and sometimes
unique earth resource data from visual observation and discretionary imaging. The
result of training crewmembers to look for specific earth resource phenomena and
selectively record important events crystallized the complementary nature of manned
and unmanned observation systems.
The technological development of automatic and unmanned satellites (LANDSAT,

HCMM, SEASAT, METEOSAT, ETC.) to acquire remotely sensed data came into
effective use in 1970s. LANDSATS were originally called Earth Resources
Technology Satellites (ERTS) and their present acronym of LANDSAT programmes
were acquired in 1975. LANDSAT 1 was launched in 1972 and was equipped with a
Return Beam Vidicon (RDV) television camera system and a four-wave band
Multispectral Scanning System (MSS). The latter operates in two channels of the
visible spectrum and two in the near infrared. It records radiant energy from the field
of view by means of on-board detectors that converts this, electronically on to a
magnetic tape and it can also be telemetered directly to ground receiving stations.
LANDSAT 2 and 3 launched in 1975 and 1978 respectively were virtually carbon
copies of the prototype. These first three LANDSATS had an orbit height of 910 km,
and a capacity to image the entire earth's surface (below 81°) every 18 days.
LANDSAT 4 and 5 launched in 1982 and 1984 respectively incorporated important

improvements. They orbit at an altitude of 705 km and have a global repeat imaging
frequency of 16 days, but also carried a seven-wave band sensor known as the
18
Thematic Mapper (TM). From the start of the programme, LANDSAT data were
made available to the worldwide user community free of access or copyright controls.
This, together with the quality and versatility of information made available, has been
of great significance in promoting the civilian use of remote sensing from space in the
last two decades.
The HCMM Satellite was experimentally operational between 1978 and 1980, and was
equipped with scanning radiometer that obtained data in visible, near infrared and
thermal infrared wavebands. The orbital configuration was designed to maximize
opportunity to gain data on surface temperatures. Geological and vegetation mapping,
snow and ice melt predictions and industrial pollution are amongst some of the
applications that have made positive uses of HCMM data.
SEASAT was launched in 1978 on experimental basis to obtain information on the

World's Oceanic surfaces (wave heights, patterns sea ice, etc.) and has provided
original data on land cover. It was the first satellite to use a high-resolution radar-
imaging instrument, synthetic Aperture Radar (SAR). It was also equipped with a radar
scatterometer that measured oceanic wave heights, and a visible and infrared
radiometer for determining sea-surface temperatures and cloud cover.
Other major contributions, outside United States of America are the USSR military
satellites and SPOT. Several series of both manned spacecraft and automatic satellites
have been operated by the USSR since the late 1960s including the METEOR and
COSMOS Programmes. Unfortunately, details of these programmes are not found in
the West and there is no access to the imagery archives for Western Scientists and other
users, and therefore no direct reference can be made to the Russian contribution to
global remote sensing. It is known that the main thrusts of application development
have been confined to the national territories of the USSR and CMECON states with
the exception of global meteorological and oceanographic sensing.
SPOT was the first civilian Earth Resource Satellite developed independently of the
USA and USSR. The first satellite in the series was launched in 1985. Primarily, the
19
French Space Center (CNES) developed it with contributions from Sweden and
Belgium. It supplies the global remote sensing data-user community on an open access
and commercial basis.
SPOT satellite senses in three bands in the visible and near infrared in the multi-band
mode and one band in the panchromatic mode. The main sensor in SPOT is the
"Pushroom" scanner that gives several advantages over conventional Multispectral
Scanning (MSS) devices such as improved radiometric and geometric accuracy and
higher spatial resolution. The two scanners in SPOT-1 record in both panchromatic and
MSS Modes - the former giving the best known optimal spatial resolution of 10 m and
the latter of 20- 25 m. But what makes it so innovative compared with LANDSAT is
its potential capability of "off-nadir" viewing [tilting of sensors which allow it to make
a sideways look at areas which it has already visited or is about to visit]. The sensor
can be directed to +27° either side of field of view defined by the vertical axis between
the ground and the satellite. This means that the frequency of repetitive coverage is
lowered from one to three days depending on the latitude.
The near-polar orbits are sun-synchronous, so, whenever the SPOT crosses the polar it
is always at a given local time on the ground below. It surveys the globe from an
altitude of 832 km in 26 days. With tilting sensors it can view an area much more
frequently: 98 times a year at the Equator and 152 times at 45 degrees North.
The advantages in terms of frequency of viewing are obvious. Since the satellite can
sense the same area from two angles, it has the facility to produce stereoscope pairs,
and hence the capacity to provide a new form of remote sensing information. In
addition the possibilities for monitoring rapidly developing or changing events are
considerable. The main shortcoming of SPOT includes the use of a limited range of the
electromagnetic spectrum in the visible and near infrared to sense the earth resources.
This means that it can only take high quality imagery when there is a cloud-free sky.
20
Until mid-1980s, USA enjoyed the position of operating the only satellite providing
unclassified multi-spectral imagery available on a regular basis. Now several nations are
in the race for controlling of the remote sensing market. Since the French successfully
launched a high-resolution scanner (SPOT) that has continuously for a long time now
pursued a very aggressive sales program, Japan, Canada, China India, Brazil, and
European Space Agency (ESA) have also launched other advanced sensor systems.
The late 1990s saw a worldwide trend of governments privatizing many of their services
and facilities. In the trace of this wave, was the USA government that transferred its
NASA LANDSAT program to NOAA (National Oceanic Atmospheric Administration),
and late altogether transferred to private enterprise (EOSAT). This was meant to foster
competition in remote sensing technology, as well as cutting the price of satellite
products. Some of the private companies licensed for permanent earth observation were
SpaceImaging, EarthWatch, and OrbImage.
The Orbview 1 developed by OrbImage, launched in mid-1990s was for scientific

meteorological research. Orbview 2, launched in 1997 is also used for research purposes,
and in particular for studying the ocean and coastal areas.
The efforts made by Space-Imaging yielded Ikonos-2, was successfully put in orbit on
September 24, 1999. The panchromatic black-and-white Ikonos images have a 1m
resolution, while the multi-spectral images have a 4m resolution. The showpiece of the
EarthWatch is the Quick Bird 2 whose panchromatic black-and-white images have a
ground resolution of 70cm, and multi-spectral images at 3m. The Tables 2.1, 2.2 and 2.3
below give a summary of some of the sensors of much higher spatial resolutions that
have been recently launched:
Table 2.1 Multi-spectral High Resolution Sensor Systems

Sensor Origin Resolution Resolution Swath Available
(Nadir*- (Nadir*-IR) width
visible)
CLARK NASA 15km 3m 30km 1998
Quick Bird 1 Earth-watch
3.3m 4m 22m 1998
21
Orbview Orb-image 4m 4m 8km 1998
Space-Imaging 1 Space 4m 4m 11km 1998
Imaging
SPOT 4 CNES 20m 4 60km 1998
SPOT 5 CNES 10m 43m 60km 2002
Table 2.2 Low Resolution Sensor Systems

Sensors Origin Resolution Resolution Swath Available
(Nadir*- (nadir*-IR) width
visible)
Folder CNES 4km NA 2500km 96-97
SPOT4 CNES 1.1km NA 2200km 1998
EDS/MODIS NASA 275m 0.5km 2300km 1998
IRSP 4/OCM India 350m ? 1500km 1998
MERIS ESA 250m 500m 1500km 1999
MSG Eumetsat 1km 3km Earth 2001
Currently, the Department of Remote Sensing and Resource Surveys (DRSRS) applies
NOAA data at 1km spatial resolution in monitoring range conditions and estimating end
of season biomass production in Kenya. With the launching of NOAA satellite of
improved spatial resolution (250m), DRSRS is expected to improve in precision and open
up new areas of research in fields such as crop harvest forecasting and forest cover
monitoring.
Table2.3 High Resolution Sensor Systems (Panchromatic)
Sensor Origin Resolution Resolution Available
(visible- (IR-nadir*)
nadir*)
CLARK NASA 3km 6km 1998
Orbview Orbimage - 4km 1998
Quick Bird 2 EarthWatch 70cm - 2001
Ikonos-2 Space Imaging 1m 4m 1999
LandSAT 7 CNES 15m 185km 1999
*Nadir- the point vertically beneath the sensor when a plumb-line is drawn meets the
point at 90 degrees.
The high spatial resolution satellite imagery has not been accessible to Kenyan scientists.
A 20 m ground resolution of SPOT imagery is the best so far used in vegetation mapping
in Kenya. The availability of higher resolution to the Kenyan scientists will definitely
provide high quality mapping products for vegetation and other land use and cover
mapping.
22
2.5 Summary and Concluding Remarks
The development of remote sensing from satellites has not relegated conventional
aerial photography to a secondary position. In fact, it remains a primary technique
where it is necessary to resolve the detail of ground conditions as in most forms of
resource and environmental planning. It is a long established technique for surveying
and mapping. Aerial photography has a very high resolution - its' ability to register
small objects from hundreds of kilometres up in space. Its disadvantage is its limited
spectral range. Normally, it records reflected light across the whole of the visible
range, but with special film it can also pick up a certain amount of infrared radiation
(the so-called "near infra-red") at the wavelengths just outside the visible spectrum.
The earth observation from above may follow the same line of development like
computer technology. When this trend is followed, the coming decade will witness the
birth of small dedicated earth observation systems, aimed at thematic and topographic
mapping.
It is not possible to list any satellite program as suitable for earth observation for
resource mapping, because the number of such programs are overwhelming and by the
time you are through with this course many of them will be out of date.
With the advancement in space technology, remote sensing will accord scientists
with indispensable information for resource management.
The modern remote sensing, which is directed to the earth surface is ordinarily
designed to facilitate resource management in an efficient manner.
Remote sensing has significant advantages that include improved vantage point,
increased spatial resolution and geometric fidelity, broadened spectral sensitive, and
permanent records, among others.
Remote sensing concept was envisaged by Socrates in c.400 BC, when he predicted
that man must rise above the clouds to the top of the atmosphere in order to understand
the environment in which he livers. This was followed with the invention of
photographic camera in 1822.
Other major historical developments of remote sensing included the use of balloons,
kites and aircrafts, with the latter revolutionizing the remote sensing industry.
The earnest use of rockets and satellites as platforms for remote sensing data is
estimated to have begun between 1946 and 1950.
23
In the 1990s witnessed a worldwide trend of governments privatizing many of the
remote sensing services, which of course has led to the development of satellites whose
spatial resolution is less than 1 meter. However, many developing countries including
Kenya have not managed to access this high spatial resolution satellite imagery.

1) Trace the major developments in the history of remote sensing.
2) Outline the significance of remotely sensed data.
3) What are the benefits of establishing a regional satellite ground receiving
station in Kenya?
2.7 Further Reading

Macanta ltd.
and Francis.
Rabenhorst, T.D. and P.D. McDermott (1989): Applied Cartography: Introduction to
Remote Sensing, Ohio: Merrill Publishing Co.
Hilger.
24
3.0 LESSON THREE: THE ELECTROMAGNETIC SPECTRUM
3.1 Introduction
In this lesson, electromagnetic spectrum (EMS) and electromagnetic radiation (EMR) are
defined, with properties of EMR related to wavelengths broadly outlined. Emphasis is
given to the visible light, one of the smallest portions of the EMS. Also discussed in the
lesson is the interaction of energy with matter, and uses of remote sensors to collect
radiation. In particular, the interaction of EMR with the atmospheric constituents is
extensively covered. The lesson is concluded with a description of photographic films.

By the end of studying this lesson, you should be able to:
(a) Define Electromagnetic Spectrum
(b) Describe the visible spectrum
(c) Describe the types of interaction between energy and matter
(d) Discuss the effects of atmospheric constituents on radiation.
(e) Describe two main types of photographic film.
3.3 The Electromagnetic Spectrum

3.3.1 Introduction to Electromagnetic Spectrum
The electromagnetic spectrum (EMS) is a wavelength continuum of electromagnetic
radiation as shown in Figure 3.0. Electromagnetic radiation (EMR) is the means by which
electromagnetic energy is propagated in the form of waves. EMR is a force field with
electric and magnetic properties, which can be generated or detected either naturally or
with special equipment. Visible light is EMR and constitutes a very small portion of the
EMS, which also includes cosmic rays, gamma rays, x-rays, ultraviolet light, ___,
infrared light, thermal infrared, microwaves (radar), television waves, radio, and audio
waves, in that order (Figure 3.0). Properties of EMR related to wavelength are:
The shorter the wavelength the greater the energy,

The shorter the wavelength, the greater the scatter and absorption, and
The higher the temperature of an object the greater the total energy emitted.
The interaction of energy with matter is wavelength dependent.
25
Visible light falls in the mid-portion of the spectrum and is detectable by most animals
with their eyes: but it can also be detected with cameras and other instruments. Generally,
either special instruments or film are necessary to detect or record the EMR in other
portions of the spectrum, which most animals cannot naturally see. However, some
animals, notably certain poisonous snakes can sense infrared thermal energy.
Figure 3.0 The Electromagnetic Spectrum
Naturally, the sun is the most obvious source of EMR in all wavelengths with the greatest
amount in the visible range. In nuclear reactions and all bodies with temperatures above
absolute zero (0ºK = -273ºC) do also generate EMR. Further, man has learned to generate
radiation artificially in almost all regions of the spectrum. Therefore, in this way,
terrestrial (ground) objects constitute important source of radiation, although in
considerably different spectral composition and magnitude from that of the sun.
26
3.3.2 The Visible Spectrum
This is the most widely used band/region of the EMS. This region is extremely small
because the spectral sensitivity of the human eye extends only between 0.4µm and 0.7µm
(Figure 3.0). Although it is a tiny portion of the EMS, the visible region does not only
allow our eyes as sensors, but do also let a variety of photographic and imagery devices
to record the radiations.
The energy radiated through the visible light portion of the spectrum is known as white
light. It is a blend and can be physically separated by a prism into discrete components,
each with its own wavelength and color. Color results from the interaction of white light
with an object. More specifically, the color of an object is produced by the absorption and
reflection of different wavelengths in the visible spectrum. A combination of various
amounts of blue, green and red light can form any color in the spectrum. These primary
colors are defined by a continuum, which run from 0.4 to 0.7 microns. For example:
The color blue ranges from 0.4 to 0.5µm.

The color green ranges from 0.5 to 0.6µm.
The color red ranges from 0.6 to 0.7µm.
Thus the key to creating color pigments is to obtain materials that reflect the desired
wavelengths of visible light. The parts of the spectrum that man cannot see are at the
extremes- UV “above” violet and near-IR “below” red. IR means below.
The UV radiation adjoins the blue end of the visible portion of the EMS. Adjoining the
red end of the visible portion of the EMS is the infrared wave. At much longer
wavelengths in the range between 1mm to 1m is the microwave portion of the EMS.
In remote sensing, we deal with visible light, ultraviolet, infrared and microwave
wavelengths of the EMS. (A wavelength is the distance between two successive
waves). Although an individual wavelength can be sensed, the most common practice is
to sense energy from a range wavelengths referred to as a band. The wavelengths
associated with a particular band share similar characteristics: For example, all infrared
27
band/wavelengths travel, are reflected and are absorbed in a similar fashion. Therefore
acquiring information from a band requires:
Developing a sensor that “reads” or gathers information exclusively from the band,
Displaying the information in a form that can be interpreted.
3.3.3 Energy/Matter Interaction

When energy strikes an object, five types of interaction are possible (Figure 3.1). The
energy is:
(a) Transmitted, energy passes through with a change in velocity as determined by the
index of refraction for the two media in question.
(b) Absorbed, energy is given up to the object, generally in the form of heat although
other forms of energy such as light are possible.
(c) Reflected, energy is returned unchanged with the angle of incidence equal to the
angle of reflection. Reflectance is the ratio of reflected energy to the incident energy
on the body. The wavelength reflected determines the color of an object.
(d) Scattered, the energy undergoes random change of direction.
(e) Emitted, the energy is first absorbed and then re-radiated (figure 3.1).
3.3.4 Sensors and The Spectrum

All remote sensors collect raw data in analog form and transform the data into a more
usable format. Figure 3.2 depicts the four basic components of any remote sensor
(system), which are:
The collector, for example, lens or antennae.

Detector, for example, the photographic film.
Signal processor, for example, the amplifier.
28
Output, for example, the contact prints- photographs and strip charts.
Figure 3.1 Interactions of Electromagnetic Radiation and Matter
Incident Energy Reflected

Emitted
(Re-radiated)
Scattered
Absorbed
Transmitted
3.3.5 Interaction of Electromagnetic Radiation with the Atmospheric Constituents

The same laws guide interaction of EMR with the atmosphere as EMR interaction
with a target, that is, reflection, absorption, scattering, or re-emission.
The interaction of EMR and the atmosphere is wavelength dependent.
Complete absorption takes place in the x-ray region because of the interaction of the
high-energy radiation and atmospheric matter.
Ultraviolet (UV) radiation is strongly-often completely absorbed by atmospheric
gases, mainly by nitrogen, oxygen, and ozone. The breakdown of ozone layer (O ) poses
possibly serious consequences.
29
Figure 3.2: Basic Components of a Sensor
Collector
Detector Output
Signal
(Imagery)
Processor
Energy from Source
3.3.6 The Atmospheric Effects on Radiation

The atmospheric effects on radiation are mainly caused through the mechanism of
atmospheric scattering and absorption:
(a) Scattering. Atmospheric scattering is basically an irregular random reflection of

radiation in all directions. Three main types of scattering have been identified namely:
Raleigh, Mie and Non-Selective scatters.
(i) Raleigh scattering arises when radiation interacts with the atmospheric molecules and
other tiny particles of much smaller diameters than the wavelength of the interacting
radiation. In this way, there is a higher tendency for shorter wavelengths to be scattered
by this mechanism than long wavelengths. A “blue” sky is a manifestation of Raleigh
scattering, and its absence the sky will be black. When the sun interacts with the earth’s
atmosphere, it scatters the shorter (blue) wavelengths more rapidly than other visible
forms, and consequently we see a blue sky.
At sunrise and sunset, the sunrays travel through longer atmospheric path lengths, which
make the scatter and absorption of shorter wavelengths more complete so that we see
only the less scattered long wavelengths of orange and red. In RS, Raleigh scattering is a
principal cause of haze, which visually diminishes crispness or contrast.
30
(ii) Mie Scattering exists when atmospheric particle diameter essentially equals the
wavelengths of the energy being sensed. Water vapor and dust particles are major sources
of Mie scattering.
(iii) Non-Selective Scattering occurs when diameter of the scatter causing particles are
much larger than the energy wavelengths being sensed. Water droplets, which commonly
have diameters in the range of 5 to 100 microns, are major causes of non-selective
scattering. They scatter all the visible light and IR radiations about equally. In the visible
wavelengths, equal quantities of blue, green and red light are often scattered
indiscriminately, which results in fog and clouds appearing white in color.
(b) Absorption. Absorption results in the effective loss of energy to atmospheric

constituents. This normally involves absorption of energy at given wavelengths (Figure
3.3) and thus resulting in certain wavelengths of radiation being affected far more by
absorption than by scattering. Radiation absorption is particularly common with regard to
infrared radiation and energy wavelengths, which are shorter than visible energy. The
most efficient absorbers of solar radiation are water vapor, carbon dioxide and ozone:
because these gases tend to absorb EM energy in specific wavelength bands, they
strongly influence “where-we-look” spectrally, when using any given image recording
system. These atmospheric regions through which energy of the desired wavelengths can
travel without significant absorption interference are commonly referred to as
atmospheric windows.
In other words, windows are places within the atmosphere where specific wavelengths
pass through virtually unimpeded. One of the principal windows for transmission of data
is located between 0.4 and 0.7 micrometer, which is the visible part of the spectrum. The
window at 1 to 3microns is used to explore reflected IR energy. Another wavelength used
window is located between 8 and 14 microns. It is used to examine emitted IR energy,
which is proportion to the heat of an object. Other lesser windows, which are not
transparent to energy transmission, are found scattered about the spectrum between the
major windows.
31
Figure 3.3 Some Portion of the Electromagnetic Spectrum with
Wavelength Absorption at the Maximum and Minimum
Although the atmosphere is a barrier to data transmission in the shorter wavelengths, its
blocking effect does benefit man/woman kind. If the shorter wavelengths, less than 0.4
microns were allowed to pass through unhindered, most living organisms would be
adversely affected by strong doses of UV, X-ray and gamma radiations.
3.4 Photographic Films

Films are usually made using several types of emulsions. Emulsions are light sensitive
substances such as silver chloride and bromide. They are normally used for coating
plates, or papers that are sensitive to different wavelengths of the EMS, each with its
own spectral sensitivity.
Two types of films are commonly used in remote sensing:
Panchromatic, and
Infrared film.
3.4.1 Panchromatic Film

This is a film that is sensitive to all color portions of the visible part of the EMS, or
simply all wavelengths of the visible spectrum. However, most air photographs are
acquired by using Pan-minus blue film. Pan-minus blue film refers to a panchromatic
film exposed through green, yellow, orange and red portions of the visible spectrum. In
32
the process, the blue portion of the EMS is left out because it is scattered more than the
rest: thus blurring distinct impressions on the photographic images.
Color aerial photography is primarily panchromatic in that the entire visible spectrum is
represented and the resulting photography has approximately the same visible colors as
it exist in nature.
3.4.2 Infrared Film

This is a film that is sensitive to violet, blue, and red light of the visible spectrum in
addition to infrared. Infrared color film is sensitive to green, red, and near infrared
wavelengths; that is, the film is particularly sensitive within the EMS portion between
0.5 µm and 0.9 µm. A major advantage of infrared photography is that fact that it
penetrates haze better than panchromatic photography.
Apart from ordinary infrared color photography, there is also color infrared (CIR)
sometimes referred to as camouflage detection. Unlike the ordinary color infrared,
which consists of three colored layers of blue, green, and red, the CIR also consists of
three layers, but the blue sensitive layer is replaced by near infrared layer so that during
photo-processing, the true colors are also replaced by others such as:
Green is replaced by blue.

Red is replaced by green.
Infrared is replaced by red.
In such circumstances, vegetation that normally appears green will appear red on the
film. Unhealthy or disturbed vegetation will appear blue-green (cyan). This constitutes
the basis for camouflage detection property in the CIR film on the principle of which it
is commonly referred to as false color photography, because colors in the final image
do not correspond to the natural colors.
False color photography was developed for military purposes to detect and pierce the
natural camouflage. During World War II, when CIR was developed, targets
33
camouflaged to appear green in forests could easily be detected because disturbed
vegetation could poorly reflect the infrared energy. Thus objects that were painted
green to look like vegetation would normally have low infrared reflectance and
appeared cyan on the film and therefore easy to locate. For example, shades of cyan in
an area of red coloration normally meant a sign of human disturbance in the form of
military camps and equipment
3.5 Summary
The electromagnetic spectrum (EMS) is a wavelength continuum of the
electromagnetic radiation (EMR). The EMR is the means by which electromagnetic
energy is propagated in the form of waves.
The EMS constitutes visible light, x-rays, cosmic rays, ultraviolet radiation,
television waves, microwaves (radar), infrared waves, radio waves and audio waves.
While human and other animals’ eyes can detect the visible light, the other portions
of the EMS can only be sensed by use of special sensors.
The sun is the most obvious source of EMR, with a small portion being generated
by nuclear reactions and all other bodies with temperatures above absolute zero.
The visible light is the most widely used band of the EMS. Although it is a tiny
portion of the EMS, the visible region not only permits our eyes to sense but do also let
a variety of photographic and imagery devices to record vision.
The visible light, ultraviolet radiation, infrared radiation and microwaves are the
most important portions of the EMS used for remote sensing.
Transmission, absorption, scattering, reflection and emission are five different types
of interaction when energy strikes an object. The atmospheric effects on radiation are
mainly caused by scattering and absorption.
Three types of scattering: Raleigh, Mie and non-selective are recognized.
Absorption of energy takes place at certain wavelengths and thus resulting in some
wavelengths of radiation affected far more by absorption than scattering.
Atmospheric windows are regions through which energy of the desired wavelengths
travel without significant absorption interference.
Panchromatic and infrared are two types of films commonly used in remote sensing.
34
(a) Draw the Electromagnetic Spectrum (EMS) and discuss the significance
of visible, infrared, microwave (radar) and ultraviolet portions of the
EMS in remote sensing.
(b) Describe the interaction between the atmosphere and electromagnetic
radiation.
(c) Explain why the sky looks blue and why sunset and sunrise look
different.
(d) Why do clouds and fog look white?
(e) Why is it difficult to replicate the natural colors during photography?
(f) Describe two main types of photographic films.
(g) Explain the significance atmospheric windows in the EMS that are
known to affect the image recording systems.
3.6 Further Reading

Macanta ltd.
and Francis.
Rabenhorst, T.D. and P.D. McDermott (1989): Applied Cartography: Introduction to
Remote Sensing, Ohio: Merrill Publishing Co.
Hilger.
35
4.0 LESSON FOUR: INTERPRETATION AND SOURCES OF ERROR IN
IMAGE DETECTION
4.1 Introduction
This lesson describes the act of interpreting satellite imagery. It explains how temperature
differences, season and time of the day, and soil chemistry, among other factors influence
the detection and identification of target features on the imagery. It is concluded with
sources of error in image detection.

(a) Describe the factors that affect the detection and identification of
subject/background contrast of satellite imagery.
(b) Determine the influence of attenuation, sensor lag, and foreshortening, among
other sources of error on image detection.
4.3 Subject Background Contrast

Successful interpretation of satellite imagery has been made in many fields of
application. These include such diverse fields as determining rock type and structure,
locating geological faults, mapping soil type and soil moisture, locating irrigation canal
leaks, determining the thermal characteristics of volcanoes, locating hot springs and
geysers, studying evapo-transpiration from vegetation, determining the extent of active
forest fires, and locating subsurface fires, among many others.
Human ability to detect and identify objects or conditions in air photographs and satellite
imagery depends on reflection differences between subjects and their backgrounds. Many
factors combine either to create or to reduce contrast differences among subjects and their
backgrounds. These include:
(a) Temperature Differences

Most scanning operations such as geological and soil mapping are qualitative in nature.
In these cases, it is not usually necessary to know absolute ground temperatures and
emissivities, but simply to study relative differences in the radiant temperatures within a
36
scene. However, some scanning operations require quantitative data analysis in order to
determine absolute values of the phenomenon.
Many factors influence the selection of an optimum time or time for acquiring the radiant
data. Mission planning and imagery interpretation take into consideration the atmospheric
effects and temperature variations. For instance, the effects of diurnal effects is shown in
Figure 4.0, which illustrates the relative radiant temperatures of soils and rocks versus
water during a typical 24-hour period. It should be noted that:
Just before dawn, the quasi-equilibrium condition is reached where the slopes of
the temperature curves for these materials are very small.
After dawn, the equilibrium is upset and the materials warm up to a peak, that is,
reached some time in the afternoon.
Figure 4.0 Generalized Diurnal Radiant Temperature Variations for

Soils and Rocks Versus Water
Radiant Temperature
DAWN
SUNSET
00 4 8 12 16 20 00
Midnight Noon Midnight
- - - - - -Radiant Temperature for Soils and Rocks

Radiant Temperature for Water
37
Maximum scene contrast normally occurs at about this time and cooling takes
place thereafter.
Temperature extremes, heating and cooling rates can often therefore furnish
significant information about the type and condition of an object.
Terrain features’ temperature is normally higher than for water during the day,
and lower than water during the night.
(b) Season and Time of the Day

Some seasons are favorable for locating features on the ground. In the northern
hemisphere air photographs and satellite imagery for mapping purposes are usually
taken during winter months, after the leaves have fallen. This is also the most suitable
season for the study of stream patterns, for gulley and stream course analysis, for the
study of roads and trails and for studies in which differentiation between coniferous
and deciduous vegetation is necessary. In the autumn, when many trees have brightly
colored leaves, identification of deciduous vegetation is frequently easier.
In springtime, it is a most favorable season for locating springs, seeps- hanging water
table, all of which are marked by luxuriant vegetation. It has also been demonstrated
that spring flowers often help to identify terrain containing certain minerals and
spring blossoms to identify trees. Similarly, summer photography is beneficial in
studies of aquatic vegetation, because the growth of vegetation is usually greatest
during the late summer period.
Minimum shadows are usually specified for mapping photography because shadows,
especially in forested areas may obscure ground information. Mapping photography
is normally taken between 10.00am and 2.00pm in order to minimize shadows. If
shadows are required, the photography is either taken earlier or later in the day.
(c) Soil Chemistry, Soil Porosity and/or Soil Permeability, and the Depth to the Water
table, indicated by vegetation height are of great significance to the forester, the
geologist and the agriculturalist. Differences in plant height are also significant in
archaeological studies, in the study of water supply, and in pollution control
38
reconnaissance, because plant growth is accelerated by high percentages of residual
nitrogen, phosphorous and potassium, all of which may be present in the soil as a
result of the deposits of human wastes, thus marking the vicinity of septic drain fields
or ancient human habitations. Minor differences in the height of vegetation, relevant
to the study of wildlife and range conservation management, include evidence of
excessive grazing on watersheds.
(d) There are several other factors that may restrict or enhance subject/background
contrast. These include vibration and many little-understood atmospheric conditions. It
must be remembered that several thousand meters of atmospheres separate aircraft and
the ground below when sensor images are being obtained. Scale is another
consideration: thus with the larger the scale, the greater the contrast, and vice-versa.
Others are spectral and spatial resolutions.
4.4 Sources of Error in Image Detection

The following can be sources of error that can limit our ability to perform proper
analysis of satellite imagery:
(a) Attenuation
Attenuation is a decrease in the amount of radiation reaching the satellite due to
absorption and scattering by the intervening medium. Attenuation increases with
distance from the satellite position. This error is most pronounced in the infrared bands
where the primary absorbers are water vapor, carbon dioxide, and ozone.
(b) Background Contrast

When clouds or other phenomena have a radiative temperature that is close to that of
the underlying surface, the analyst may find it difficult to distinguish them from the
background. This source of error is widespread in radiometric measurements from
space. The primary technique to account for this error is to sample the same scene in
multiple spectral intervals.
39
(c) Contamination
This is caused by radiation reaching the sensor from beneath a cloud. This problem is
most significant with thin clouds such as cirrus. This can be accounted for and detected
by sensing in multiple channels.
(d) Displacement
It occurs as a result of the geometrical viewing-angle and the projection of image data
on a two-dimensional plane for displaying an image. Figure 4.1 shows how the top of a
cloud is displaced to a point away from the satellite position. This error can be
corrected if the height of the cloud, the satellite position, and the coordinate of the
cloud are known. Data from high latitudes collected by GOES satellites are, however,
often difficulty to use because of this error.
Figure 4.1 Displacement due to Cloud Height
40
(e) Foreshortening
This is due to the effect of the earth’s curvature on the resolution of the image. As the
satellite scans the earth surface, pixels cover larger areas as one moves from the nadir
position. For GOES satellites, foreshortening reduces the usefulness of the data beyond
60 degrees from the satellite nadir point.
(e) Sensor Lag

It produces an error in the radiative phenomenon assigned to each pixel. As the sensor
moves along the scan line, it remembers the radiance it senses for a few microseconds.
A rough estimation is that the output of the sensor is 75% of the radiance from the
current pixel, 20% from the previous pixel, and 5% of the pixel before the previous
one.
(f) Signal Interference

The satellite signal must be transmitted to the ground, and in some cases such as the
GOES satellites, it must be processed, sent back to the satellite and retransmitted to
earth stations. During this process, there is a potential for interference of the transmitted
signal. Missing lines are easily detected and can be corrected by a number of averaging
schemes that assign values to each pixel in the missing line based on the value of
neighboring pixels.
4.5 Summary
Remote sensing techniques can be applied in many fields such as determining rock
type and structure, location of geological faults, mapping soil type and moisture, and
determining the thermal characteristics of volcanoes, among others.
Human ability to detect and identify objects/conditions in photographs and satellite
imagery depend on reflection differences between objects and their backgrounds.
Temperature differences, season and time of the day, and soil chemistry, among
other factors combine to influence contrast differences between subject and
backgrounds.
41
Attenuation, background contrasts, contamination, sensor lag, foreshortening,
displacement, etc are among the sources of error that limit our ability to perform proper
analysis of satellite imagery.

1. Describe the factors that affect the detection and identification of
subject/background contrast of satellite imagery.
2. What is the influence of attenuation, sensor lag, and foreshortening on image
detection?
4.7 Further Reading

B.N.S.C. (1989): The Earth Below: A Look at Remote Sensing; Department of
Trade and Industry, London: Moore and Mathers.
Macanta ltd.
Cracknell, A. and L. Hayes (1993): Introduction to Remote Sensing; London:
Taylor and Francis.
Lillesand, T.M. and R.W. Kiefer (1994): Remote Sensing and Image
Interpretation; New York: John Wiley & Sons Inc.
Hilger.
42
5.0 APPLICATION OF REMOTE SENSING IN AGRICULTURE
5.1 Introduction
This lesson examines the major land use/cover categories that can be differentiated by
the existing satellite sensors. It also describes Air Point Sampling (APS) as an
excellent alternative data acquisition technique in tiny African plots, with wide
incidence of intercropping and highly varied calendar.

By the end of studying this learning, you should be to:
(a) Describe the use of remote sensing in land uses/covers.
(b) Evaluate the potential air point sampling as alternative technique.
5.3 Remote Sensing in Land Uses and Covers

To evaluate fully the difference between general land resource potential and the
current use status of the resource base, timely and accurate information is needed on
the land cover/use at the national level. The resolution of existing satellite sensors
available for civilian use is sufficient to differentiate between major land cover/use
categories in Table 5.1:
Table 5.1 A Typical Land Cover/Use Definitions as Derived from LANDSAT

Agriculture Land use for the production of food and/or fiber.
(a) Sugar: Major agricultural areas of land planted predominantly
to sugar cane interspersed with few other major crops except
improved pasture.
(b) Intensive agriculture: All the other major agricultural areas
predominantly (75% or more of the land) used for field and tree
crops usually associated with prime agricultural areas.
(c) Marginal agriculture: Less intensive agricultural areas of a
predominantly subsistence nature with 25 to 74% of the land used
for field and tree crops. Smaller fields interspersed with
unimproved pasture, range, trees and open land in hilly terrain and
foothills usually characterize it.
(d) Pasture: predominantly improved pasture used for grazing.
Forest Land with a predominant crown closure of 75% or more. The trees
can be either predominantly deciduous or coniferous.
43
Rangeland Areas with a predominant brush and grass vegetation cover.
Limited potential for grazing. Presence of xerophytes common in
the foothills. There are others such as limited rangelands that are
areas with major limitations for grazing caused either by steep
slopes or heavy brush cover.
Urban and Manmade structures for residential, industrial, commercial and
Built-up transport related land uses in contiguous areas of more than 1
square kilometer.
Wetlands Areas with a hydrologic regime accommodating aquatic or
hydrologic vegetation. Excluded are areas under rice production.
Barren/Open Areas with exposed soil and little or no vegetation cover. Surface
mining areas are included in this category.
Water Inland water surfaces.
Cloud Cover Areas where cloud free satellite imagery could not be obtained.
Source: Witter et al 1980
The resolution of many satellites poses severe limitations for some tropical
agricultural areas where small scale subsistence farming, intercropping and relief-
induced spectral variations make it difficult to derive consistently accurate land
use/cover data. However, in many regions in the world especially in Africa where
comprehensive, multi-seasonal and timely imagery coverage is incomplete or non-
existing, air point sampling provides an excellent data acquisition alternative.
5.4 Air Point Sampling

Agricultural geography of African settlement is a complicated one because the
various specialised production units have same tone and texture on the aerial
photography or satellite imagery. In addition, the analysis of land covers is
complicated by use of black and white photography together with the relatively small
scale. Another complication comes from the tiny size of fields or plots, the wide
incidence of intercropping and highly varied calendar that reduce the effectiveness of
more conventional aerial survey methods.
44
Due to this complication, Ecosystems Limited (ESL) developed a technique known
as Aerial Point Sampling [APS], which quantifies land use patterns within the
context of regional planning for the African Smallholder Farming System (Norton-
Griffiths, 1987). The APS was founded when it was realized that conventional large
format photography suffers from two main drawbacks:
[a] Costs prohibit the use of scales, large enough to resolve the intricate patterns of
smallholder land use, and
[b] The climatic "window" during which high altitude photography can be carried
out, especially in East Africa rarely coincides with the stage of maximum crop
differentiation, but usually coincides with the post-harvest period.
APS overcomes these problems in two ways: Firstly, by flying high altitudes
between 120 and 400 m above the ground and under any weather conditions.
Secondly, the APS method uses large scale and sample colour photography. The APS
method is a 3-level system of flight lines of primary units that consists of vertical
photography, geo-referencing the sample photographs and ground truthing.
[i] Vertical photography involves taking vertical sample photographs on high-speed

colour film (diapositive). Flying height and focal length of lens are selected to give a
nominal image scale of 1:7 000. However, the exact height of the aircraft is recorded
each time a photograph is taken so that the exact scale can be calculated later. At this
nominal scale, each photograph covers about 4.2 hectares (pixel) and on average there
is one sample photograph for every 2.5 km².
[ii] Geo-referencing photography is a stage at which each photograph is treated as a

point sample and their exact locations are identified. Every time a photograph is
taken, the distance along the flight line from the start point, and the distance on the
left or right of the track is recorded from the navigation system (OMEGA/VLF).
From these two parameters, it is possible to plot the location of each sample
photograph on a map overlay of 200-300 m².
45
Typically, a 1:250 000 scale map overlay is made showing the position of each of the
sample photographs. This overlay is then used to assign each individual point sample
to any layer (stratum) such as district, division, location, catchments or sub-
catchments, soil type, climatic zone or UTM grid cell. This is done by placing the
overlay onto appropriate 1:250 000 scale base map of the stratum boundaries. Once a
sample point has been assigned to a stratum such as a location or soil type it can be
associated with any characteristic of that stratum such as population density or
inherent erodibility. The APS method thus provides a powerful tool for integrating a
wide range of ancillary data into land use survey.
All data associated with the sample points can now be used to create statistical
estimates for the different strata, for example each district, division and location
within the survey area. Estimates can also be made for individual UTM grid cells that
can then be applied in a cell format.
[iii]Since the bulk of the land use data from an APS survey comes from the
interpretation of the sample aerial photographs, ground truthing is of great
significance. The first stage is for the interpreters to fly on the surveys and make diary
of the crops, intercrops, vegetation and other items of interest that they see, and in
particular whenever they see something that they do not understand. Next the aircraft
as soon as possible for a closer look that should be often sufficient to solve the
problem visits these areas.
All "problem areas" are nonetheless noted onto a master map and after the main
survey is completed, the aircraft visits each in turn making low passes so that the
observers can clearly see the crops and intercrops. The problems are again discussed,
and solutions and interpretations proposed. Finally, a helicopter is sent out to visit the
few remaining problem areas about which the interpreters remained uncertain. Low
hovers and landings are made until all conflicts are resolved.
46
5.5 Summary
To fully evaluate the difference between general land resource potential and the
current land use status of the resource base, timely and accurate information is needed
on the land use/cover at the national level.
The typical land use/cover categories that can be derived by the LANDSAT are
agriculture, forest, rangeland, urban and built areas, wetlands, water, barren/open
spaces, and cloud cover.
Air Point Sampling (APS) is a technique that quantifies land use patterns within the
context of regional planning for the African smallholder farming system. It overcomes
drawbacks such as high costs and climatic window, which hardly coincides with the
maximum stage of crop differentiation associated with conventional large format
photography.
The APS method is a three-level system of flight lines that consists of vertical
photography, geo-referencing and ground truthing.

(a) Differentiate between Air Point Sampling and Ground Truthing.
(b) How important is Air Point Sampling in the acquisition of remote sensing data
for the African small-scale agriculture?
5.7 Further Reading

Macanta ltd.
Taylor and Francis.
47
Hilger.
48
6.0 LESSON SIX: APPLICATION OF REMOTE SENSING IN NATURAL
RESOURCES MANAGEMENT
6.1 Introduction
This lesson examines the application of remote sensing in natural resource
management. Remote sensing use in vegetation, wildlife and water resource studies is
highlighted.

By the end of learning this lesson, you should be able to:
(a) Explain the application of remote sensing in natural resource management.
(b) Describe the use of remote sensing technology in vegetation, wildlife and
water resource studies.
6.3 Introduction to Natural Resource Management

Tropical countries face conditions of increasing population and scarce resources that
call for the urgent need to develop comprehensive survey methods for database with
an allowance for regular updating. In many aspects of local and national planning,
aerial photography provides one way of achieving this goal. In East Africa, there are
some interesting examples of aerial photography applications in different fields. In
Kenya, for example, colour infrared air photographs have been used by the Kenya
Soil Survey in the inventory of soil associations in the country's rangelands, while
standard and small camera photography has been widely used in wildlife studies.
Sequential air photography has also been used in studies of soil erosion. Aerial
photography has further been used in other areas such as the estimation of standing
biomass and in phenological mapping.
The significance of aerial photography in resource inventory, analysis and

management is illustrated here by a general description of its application in
vegetation, wildlife and water resource studies. Three case studies are provided to
49
show the application of aerial photography in the analysis of physical features,
vegetation characteristics and rural and urban land use.
6.4 Aerial Photography in Vegetation Studies

Aerial photographs have been used extensively for vegetation mapping purposes
particularly in the preparation of forest inventories. Interpretation of air photographs
is therefore an established technique in vegetation mapping and analysis. Such
interpretation is, however, based on an approach that may be highly subjective being
mainly centered on observed patterns whereby one investigator interprets and
classifies vegetation on the basis of such elements as tone, colour, texture and pattern,
that is observed on each image. To avoid extreme generalization, accurate vegetation
analysis from aerial photography must always be complemented by ground data
without which wide error margins are bound to arise.
A suitable approach in aerial vegetation survey should involve systematic integration

of survey techniques whereby intensive stereoscopic image interpretation is combined
with detailed ground truthing sampling. An accurate survey requires that a number of
procedures be adopted, beginning with the assembly of photographs or photo mosaics
followed by a systematic stereoscopic photo interpretation. In later stages, transfer of
photo information into a base map can be made. By utilizing ground truth data, some
corrections will eventually be made on the original vegetation map due to new
findings, and confirmation from the field sampling exercises.
The application of aerial photography in vegetation analysis is mainly based on

spectral reflectance characteristics. The effects of plants' physiological and life-form
changes on vegetation reflectance characteristics are important with regard to
vegetation photo analysis. An understanding of the spectral response patterns of plant
communities within different electromagnetic wavelengths during different seasons is
essential for effective application of aerial photography in vegetation studies.
Outstanding examples of areas in which aerial photography is applicable to

vegetation analysis include the analysis of general stand characteristics: species and
50
cover type identification, pest impact analysis, phenological studies and estimation of
vegetation depletion rates.
6.4.1 Species Identification

Plant species identification on aerial photographs can easily be undertaken through
the process of elimination, whereby all species that are unlikely to prevail in an area
due to existing environmental conditions are initially eliminated. When one is
analyzing air photographs taken from the Kenyan rangelands, for example, it is very
unlikely that the vegetation observed would include species that exist within the
humid highlands such as bamboo [Arundinaria alpina]. It is however, very likely that
the vegetation may consist of one of the many species of Acacia which are
widespread in the dry lands.
The second step in species identification will then involve establishing possible
groups or formations of species that may occur in an area based on the knowledge of
common species association. The final stage will involve the identification of
specific plant species by making use of basic image spectral characteristics from
photographs of small vegetation stands. Any two vegetation stands consisting of
different plant species will manifest different spectral characteristics as a result of
canopy structure and other life form characteristics. It must be realized that two
stands of similar plant species composition may exhibit different spectral
characteristics due to plant age status. In most cases, however, vegetation spectral
reflectance characteristics will normally depend upon tree height, leaf structure,
shape, number and distribution.
The task of species identification from aerial photographs will depend on a number of
factors: photo scale, quality of photographs, and the season when photography was
done. Seasonal changes in the appearance of trees throughout the year may
particularly enable discrimination of species that are indistinguishable from
photographs taken on single dates. In East Africa, such a distinction can be made
after the long rains when some species that may have shed their leaves during dry
spell turn green again.
51
Air photographs that are most suitable in the task of species identification are
panchromatic paper prints at a scale of between 1:15 000 and 1:24 000. Black and
white infrared films can also be used for vegetation stands that are characterized by
high species diversity. Both colour and colour infra-red [CIR, False colour]
photographs are also being used with increasing frequency particularly in small photo
scales.
In Sudan, an inventory of forests and woodlands has been carried out for the southern
parts of the country using aerial survey. At the first stage, a vertical monochrome
photograph of the whole area was taken. Later, smaller units of the area were
photographed again using 35 m Ektachrome film and the photographs examined at a
six times enlargement. By using these techniques, it was possible to identify up to
80% of tree species forming the top canopy.
6.4.2 Other Vegetation Characteristics

Apart from species identification, other stand characteristics can also be studied from
aerial photographs. Common examples include stand coverage, tree height and
density, biomass estimation, stand phonology, and plant disease infestation. These
analyses are particularly vital in the quantitative analysis of woodlands and forests.
Stand coverage can easily be estimated using photographs of known scales through
ordinary geographic techniques. Similarly, stand density can be measured by
estimating the percentage of crown cover in a given stand. In forestry, a number of
visual density grids have been developed to assist in this process. These typically
consist of squared or circular grids with black dots to represent trees of varying
percentages of crown density.
6.4.3 Density Estimators of Trees

Using ordinary statistical techniques such as regression analysis, the volume of
individual trees can normally be determined as a function of species and other
characteristics such as crown diameter and tree height.
52
Aerial photography has successfully been used in phenological studies. Phonology is
defined simply as a branch of science that studies periodic phenomena in plants as a
response to local oscillations in environmental conditions. The detection of
phonological events for specific vegetation types is achieved by interpreting related
spectral responses. Using such responses, different phenological sequences may be
established. A green wave for example, may be recorded in colour photography
during the time of maximum foliage development in plants. On the other hand, a
brown wave can be identified with vegetation stages such as plant maturation and
fruition or leaf coloration and abscission. Such green and brown waves can be useful
in rangeland studies as a way of monitoring forage availability and development for
sustainable range management.
Aerial photography has been extensively used to study tree damage as a result of
diseases that emanate from pest infestation as well as other sources. Biological tree
damage is attributed to a variety of organisms such as bacteria, fungi, viruses, and
insects. The damage caused by diseases from these agents could be detected using
aerial photography. The technique has been widely used around the world to detect
forest damages emanating from such insects as budworm, beetle, moth, butterfly,
weevil, and so forth. For instance, exploration work was conducted to assess the
damage by the recent attack on Kenyan exotic Cypress forest plantations by aphids.
Aerial photography can also provide valuable information about shallow deposits of
oil hydrocarbons. Generally, two approaches can be adopted. One of these involves
the analysis of spectral data to detect parameters such as the following; absence or
reduced density of vegetation; location, identification and distribution of specific
indicator species; identification of morphological and spectral characteristics and
quantifying differences in vegetation zones. The other involves a more complex
process of developing a vegetation potential model and comparison of the actual and
potential vegetation to detect the vegetation anomalies. This area of aerial
photography application is still young and largely developmental.
53
6.5 Wildlife Studies
Vegetation and landscape patterns affect the quality of habitat for wildlife. Within
the tropics there is an urgent need to develop quantitative methods to assess wildlife
habitats that must be considered in land use management planning. Aerial
photography can provide the means for mapping vegetation cover over large areas for
wildlife habitat management. In most studies where aerial photography has been
applied, it has manifested great advantages in terms of speed, convenience, economy
and replicability.
The application of aerial photography in wildlife studies is probably most common in

the making animal population censuses. Inventories of wildlife populations are
necessary for the planning of management programmes, and formulation of
exploitation regulations. In such censuses, a major problem may arise from the fact
that some animals may be concealed from view by vegetation and/or other animals in
a group. Animals that have been photographed for censuses taking can, however, be
counted under a low power microscope. Recounting under high power magnification
can make checks on the accuracy of the counting techniques.
One area that has widely benefited from aerial photography in Kenya in recent years
is wildlife and livestock management. The Department of Remote Sensing and
Resource Surveys (DRSRS) has been monitoring spatio-temporal wildlife game and
livestock population changes through aerial surveys. As a state agency under the
Ministry of Planning and National Development, DRSRS was established in the
1970s and has operated with financial and technical assistance from Canada through
the Canadian International Development Agency [CIDA]. The major objectives of
DRSRS aerial survey section has been to establish a monitoring programme to
determine the seasonal abundance and distribution of livestock and wildlife
herbivores throughout the Kenya range lands and training Kenyan personnel in the
techniques of aerial and ground surveys. During 1977, DRSRS conducted systematic
aerial surveys of livestock and wild herbivores throughout the rangelands that cover
about 500,000 km2 of the Kenyan land. The surveys were done from Cessna 185
aircraft equipped with global navigation systems and radar altimeters at a flight height
54
and speed of 300 ft and 150 km/hr, respectively. Censuring was achieved from strips
of 112 meters wide along east-west transects spaced 10 km apart. The exercise was
repeated in 1978 beginning in south-east Kenya in January, and thereafter moved to
the north eastern regions and ending in the northern Rift Valley in November. The
survey took approximately eight months to complete.
During the exercise, observers in each of the aircrafts made actual counts and took
photographs of both wild game and livestock from which population estimates were
corrected by viewing colored slides under a microscope. In 1979, DRSRS also
conducted aerial surveys over Meru National park and Bisanodi National Reserve
which together covered some 1 600 km2. The aim of this aerial survey was to stratify
these areas into zones of different herbivore density and to distinguish major habitat
types. The entire exercise was conducted from Cessna 185 and Partenavia aircrafts
with flight height of 300 ft and mean survey speed of 150 km/hr.
Aerial photography has also been used in Kenya to carry out flamingo and pelican
sample censuses on Lakes Nakuru and Bogoria. This was done using a piper PJA 12
Cruiser aircraft. In this exercise, vertical aerial photographs were taken from a height
of 457 m above sea level with a Trikon F camera fitted with 55 mm Nikkon lens
mounted in the rear fuselage. The camera was fitted with a F250 motor-back taking
250 exposure length 35 mm film operated by a home made electrical timer, fitted
with a light that enabled the pilot to get the aircraft wings level for photography. The
photography of flamingo flocks was done with about 30% forward overlap.
Photographs were later assembled into a photo mosaic and the area of overlap
delimited. The flamingoes were then counted under a binocular microscope. The
greatest problem in this exercise was to distinguish between the lesser and Greater
flamingos [Phoeniconaias minor and Phoenicopterus ruber, respectively]. The aerial
photographs also did not allow a reliable distinction to be made between white and
pink-backed pelicans Pelecanus onocrotalus and Peleccanus refescens,
respectively].
55
6.6 Water Resource Studies
Although tropical countries have permanent water bodies, they continuously face
serious water supply and management problems. Management of water resources
must begin with the establishment of a hydrologic inventory and surveillance
systems. Aerial photography is capable of offering multi-temporal data and
substantial area coverage constitutes a well-suited tool for providing input into a
geographic information database designed for natural resource management but more
specifically for water management.
Apart from providing spatial information about the location and extent of water
bodies, aerial photography also indicates the structure of drainage networks, and is
capable also of providing details on geological lineament in terms of location, extent,
orientation, configuration and intensity. Such details are vital in understanding the
distribution of water resources in an area.
Aerial survey can be used in a number of ways in the inventory of water resources
and related resource planning. Although the technique is particularly suitable for the
study of surface waters, it can also be used in ground water investigations whereby
inference is made from prevailing vegetation characteristics. A good example of this
is the close association between the existence of the yellow fever tree [Acacia
xanthophloea] in high water table areas. In this and other ways, aerial photography
assumes an important role in the planning of water development schemes and in
monitoring water use in irrigation schemes. Seasonal variations in surface water flow
at both short and long terms can also be studied effectively through multi-temporal or
sequential analysis of air photographs.
Apart from studies of rivers and streams, aerial survey can provide an accurate base
for monitoring many of the dynamic aspects of lakes and large man-made reservoirs.
In Kenya, the constant changes of sizes and water chemistry in Rift Valley Lakes
such as Naivasha, Nakuru and Elemeinteita can be observed from periodic air photos
in the area. Unlike lake water level fluctuations that are much more straightforward to
discern from photographs, application in water quality studies is based on spectral
56
characteristic changes in water as a result of variations of electrical conductivity and
water salinity. This variation causes the dielectric constant or permittivity of the
water, and hence the reflectivity and emissivity of water surfaces to change. Such
spectral characteristics can then be observed from colour infrared photography.
The inference of water chemical conditions as a result of differences in spectral

characteristics is a very difficult task because it can easily be confused with certain
other characteristics of water bodies. Deep-water bodies for instance might appear
darker than shallow water basins while sedimentation may also affect the colour or
tone on air photographs. Water bodies that are infested by aquatic weeds can also
exhibit unique photographic characteristics that may require some form of ground
truthing. In Kenyan lakes, the distribution and abundance of notorious weeds such as
Salvinia molesta and Eichorrnia crassipes provide yet another dimension in which
aerial photography can be used for the sustainable management of aquatic
ecosystems.
Air photography can forms a versatile tool in monitoring cover changes and species
structure within wetland ecosystems such as swamps, marshes, bogs, flood plain and
estuaries. In this way, colour infrared photography has been effectively used as an
aid in the mapping of wetland plant communities. This kind of mapping is vital as a
source of baseline information for long-term monitoring of subtle changes in these
very vital ecological systems. More rapid changes such as the depletion of Cyprus
papyrus swamps in and around the Kenyan lakes by livestock and agricultural
industries can also be monitored. In this kind of studies, infrared colors and textures
together with lifeform information and ground reconnaissance data is used to identify
wetland vegetation types, sub types and species. Normally, an assignment of several
sets of colour and texture combinations is made for plant species existing under
certain substrate conditions. A general increase in vegetation complexity is, for
example, expected from saline to freshwater wetlands.
57
6.7 Summary
The remote sensing technology is important in inventorying, monitoring, analysis
and management of vegetation, wildlife and water resources.
The application of remote sensing in vegetation analysis is mainly based on
spectral reflectance characteristics. Outstanding examples of areas in which remote
sensing is applicable to vegetation analysis include species cover type identification,
pest impact analysis, phonological studies and estimation of vegetation depletion
rates.
The application of remote sensing in wildlife studies is mostly in the making of
animal population census.
In water resource management, aerial photography is not only used in locating the
extent of water bodies, but also the structure of drainage networks. Remote sensing
can also be used to monitor cover changes and species structure within wetland
ecosystems such as swamps, marshes, bogs, floodplains and estuaries. Thus forming a
vital source of baseline information for long-term management of subtle changes in
these vital ecosystems.

(a) Citing specific examples, evaluate the potential and challenges of remote
sensing for the study of weather forecasting in Kenya.
6.9 Further Reading

Macanta ltd.
Taylor and Francis.
58
Hilger.
59
7.0 LESSON SEVEN: SATELLITE WEATHER AND CLIMATE
FORECASTING
7.1 Introduction
This lesson deals with satellite weather and climate forecasting. It specifically
examines the role of NOAA, GOES and DMSP, and concludes the lesson with
weather forecasting in Kenya.

(a) Describe NOAA, GOES, and DMSP.
(b) Explain how weather forecasting is conducted in Kenya.
7.3 Weather and Climate Forecasting

Since 1960s, when the first meteorological satellites were launched, mankind was
provided with a vintage space opportunity to observe global perspective of the
environment. These satellites made it necessary to monitor and predict weather and
climate, and make accurate observations of the current state of the atmosphere and the
surface.
The satellites, mainly NOAA and GOES carry a wide range of meteorological sensors
that enable them to have an advantage of global coverage at very high temporal
resolution. In addition, they are useful in natural resource applications, where
frequent, large area mapping is required and fine details are not necessary. Further,
the advantage of large area coverage reduces the volume of data to be processed for a
particular application.
i) NOAA
This series of satellites is named after the National Oceanic and Atmospheric
Administration (NOAA). The satellites' orbit crosses nearly the north and south
60
poles, and hence normally referred to as near-polar satellites. They orbit the earth
with an inclination relative to the Equator of nearly 90o. Their orbit height varies
between about 270Kms and 1600Kms, and they are usually sun-synchronous -
meaning that they cross the Equator at the same sun-time everyday. By having this
type of orbit, the satellite visits any particular point above the earth at the same time,
which is useful for the comparative analysis of multi-temporal weather and climatic
data.
ii) GOES
These are satellite-sensing systems that are placed in a high orbit at approximately
35,900Kms above the Equator. At this attitude, the speed of the satellite can exactly
match the speed of the earth's rotation. That is, the orbital plane of the satellite
precesses 360o in 365 days. The precession of the orbital plane at this rate keeps it
always in the same orientation with respect to the sun, day-by-day, and revolution-by-
revolution. The fact of remaining stationary means that they can achieve a high
temporal resolution, but the great height means that the spatial resolution is somewhat
poor, and therefore having a limited number of uses such as getting a broad view of
the weather, and transmission of telecommunication signals.
GOES normally image the earth's surface between latitudes 80oN and 80oS. They are
able to transmit the data on their whole viewable area every 30 minutes. NOAA
shares this characteristic, and both have an average working life expectancy of four
years. Also note that all satellites have been designed to have similar basic
characteristics such as co-coordinated design approach which permits the reports
from the Data Collection Platforms (DCPS) moving from one satellite coverage
area to another to be received and transmitted by any one of the satellites in the
network. GOES' greatest contribution is, therefore, assisting in the retransmission of
conventional analogue pictures of weather and information that can be picked from
DCPS for retransmission to Meteorological Data Distribution Centers (MDDCs)
Worldwide.
61
iii) DMSP
According to Lillesand et al (1987) some countries administer their own
meteorological satellite programmes called the Defence Meteorological Satellite
Programmes (DMSP). Some of the data produced from the operational satellites in
this program are available to civilian users on an unclassified basis. For instance, the
U.S. Air Force DMSP satellites carry a range of meteorological sensors whose
scanners onboard produce images in the 0.4 to 1.1µm (visible and near-Infra-red)
band, and the 8 to 13µm (thermal infra-red) band.
The U.S. Air Force DMSP satellite has a unique capability of scanning nighttime
visible band imaging. This comes about through the ability to "tune" the amplifiers
of the system to obtain images under low illumination conditions. Both the thermal
and daytime visible images of the DMSP have been used for such civilian
applications as urban extent, auroral displays, oil and gas fields and forest fires that
have been detected with the low light sensors.
7.4 Satellite Weather Forecasting in Kenya

Over the past few years, the Kenya Broadcasting Corporation (KBC) Television
Station has been displaying satellite images of Kenya, Africa and other continents
during weather forecasting programmes in real space and time on a daily basis for
public consumption. The images are in convenient formats that have latitude and
longitude grids with clearly marked coastal lines. The displays usually involve
remotely sensed meteorological information (e.g. cloud distribution) from the visible
and Infrared channels of the European Space Agency (ESA) METEOSAT.
The French space and meteorological authorities that performed the feasibility studies
and initiated the pre-development of the satellite system started the METEOSAT
programme. Later, there were consultations with the Western European governments
which saw the tasks of operations, processing, archiving and extraction of
meteorological parameters being taken over by the ESA after 1978. By mid-1983,
plenipotentiaries of several European meteorological services had signed an
62
agreement, the EUMETSAT Convention which created a METEOSAT Operational
system; and came into effect in November 23, 1983.
The METEOSAT system comprises the satellite itself and the ground system of
which the main components are the Data Acquisition, Telemetry and Tracking
Stations (DATTs) and the METEOSAT Ground Computer System (MGCS) all
located in Darmstadt, Germany.
The raw data (pre-processed data) acquired by the satellite is geometrically corrected
at the receiving station, Darmstadt to appear as if they had been taken by a perfect
radiometer on-board of an ideally stable spacecraft. This is accomplished by satellite
operator or users equipped with a computer-linked receiving station. The data
detected by the sensors is then digitized and processed on the ground in order to
extract qualitative information.
Some of the important meteorological parameters collected by the METEOSAT

system include cloud coverage, cloud motion vectors, cloud top heights, sea surface
temperatures, upper tropospheric humidity and vertical soundings. In most cases,
these products are automatically processed, and when ready for submission to human
quality control, the information is either produced as computer print-outs or telexed
or telefaxed, or transmitted back to the satellite or Data Collection Platforms
(DCPs), from where it can be retransmitted to the Meteorological Data Distribution
Centres (MDDCs) that are accessible to the users community through ordinary
telecommunication satellites.
The Dagoretti Meteorological station in Kenya is one of the MDDCs in Africa. The
station receives the METEOSAT geometrically corrected information from
Darmstadt through the International Telecommunication satellite (INTELSAT) off
the Indian Ocean via another INTELSAT station on Mt. Longonot as shown in the
Figure 7.1
In conclusion, the main role of the METEOSAT can be summarized as follows:
63
(a) Generation of satellite images showing distribution of clouds and water vapor
in the globe in every 30 minutes;
(b) Extraction of meteorological parameters from the original data;
(c) Dissemination programmes which include:
(i) Processed METEOSAT images,
(ii) Relay of conventional meteorological charts as conventional analogue
transmission pictures,
(iii) Maps of cloud top heights, extracted from image data, also as pictures,
(iv) Administrative messages, sent in image format as conventional analogue
transmission picture, and
(v) Retransmission of conventional analogue pictures with image broadcasts via
INTELSAT, or relaying of information data from DATA Collection Platforms
(DCPs) at remote sites to the actual users.
7.5 Summary
NOAA is an acronym for National Oceanic and Atmospheric Administration.
NOAA ‘s orbit crosses nearly the north and south poles, and hence normally
referred to as near polar-orbiting satellites.
GOES are remote sensing systems placed in high orbits at approximately
35900km above the Equator.
The remotely sensed meteorological data use by the Kenyan government is
obtained from European Space Agency (ESA).
The METEOSAT comprises the satellite itself and the ground systems of which
the main components are the data acquisition, telemetry, and tracking stations.
The role of METEOSAT includes generation of satellite images, extraction of
meteorological parameters, and dissemination of meteorological information.

(a) Citing specific examples, evaluate the potential of Satellite Remote Sensing
for the study of weather forecasting.
(b) Provide a detailed account of weather forecasting that may be shown on the
Kenya televisions during weather forecasting.
(c) Show the extent to which the following concepts can be differentiated:
64
Aerial Photography and Remote Sensing
Air Point Sampling and Reference data.
Polar-Orbiting satellites and Geostationary satellites.
(d) Give a detailed account of the LANDSAT system as an Earth Resource

Satellite.
(e) Examine the relevance of remotely sensed data for environmental and natural
hazard monitoring.
(f) Either, Show how you will employ the principle characteristics of Air photo
Interpretation in identifying different types and stages of crop development in
African traditional rural areas.
Or, Discuss how you will employ air photo Interpretation principle characteristics
to identify and classify major land uses in the neighborhood of an urban slum
settlement.
(g) Discuss, how you would use Aerial Photography and Remote Sensing for the
planning of a new settlement area in the Sahelian region of Africa.
7.7 Further Reading

Macanta ltd.
Taylor and Francis.
Kidder, S.Q. and T.H.V. Haar (1995): Satellite Meteorology. London: Academic
Press.
Hilger.
65
Figure 7.1 Transmission of Meteorological Information to Dagoretti
Meteorological Data Distribution Center in Kenya from European space Agency
(ESA) METEOSAT
66
8.0 LESSON EIGHT: INTRODUCTION TO GEOGRAPHIC INFORMATION
SYSTEMS (GIS)
8.1 Introduction
In this lesson, the definition of GIS is provided and the ways in which GIS
technology’s impact on the society are presented. Also discussed is the way in which
GIS has developed, tracing the relevant influences in other information systems that are
concerned with the representation of geographic data.

y the end of this lesson you should be able to:
Define Geographic Information Systems
) Describe the significance
Describe the questions GIS can answer
) Trace the major developments in the history Geographic Information Systems
8.3 Defining Geographic Information Systems (GIS)

One of the most general definitions developed by consensus among 30 specialists is:
“GIS is a system of hardware, software, data, people, organizations

and institutional arrangements for collecting, storing, analyzing and
disseminating information about areas of the earth” (Ducker and
Kjerne 1989, pp.7-8).
Accordingly, GIS is an organized activity by which people or any information

management system which can:
“Measure aspects of geographic phenomena and processes,

Represent these measurements, usually in the form of a computer database to
emphasize spatial themes, entities and relationships,
Operate upon these representations to produce more measurements and to
discover new relationships by integrating disparate sources, and
67
Transform these representations to conform to other frameworks of entities and
relationships.
Retrieve information based on its spatial location,
Identify locations within a targeted environment which meet specific criteria,
Analyze the related data spatially as an aid to making decisions about that
environment,
Facilitate selecting and passing data to application-specific analytical models
capable of assessing the impact of alternatives on the chosen environment,
Display the selected environment both graphically and numerically either before
or after analysis” (Hrisman 1996; Francis Hanigan 1988).
However, the most common understanding of GIS is tools (Hrisman 1996, pp.5) that
have been designed to be effective and efficient for a certain range of purposes within
a social and a historical context to serve changing needs. These tools allow for the
processing of spatial data into information tied explicitly to and used to make
decisions about some portion of the earth (Demers 1997). The term GIS is now and
again becoming hybridized and modified to conform to intellectual, cultural,
economic, and even political objectives as shown in Table 8.0 below:
Table 8.0 Examples of synonymous terms for geographic information

system and the source or motivation behind their derivation*
Terminology Source
Geographic Information System United States Terminology
Geographical Information System European Terminology
Geomatique Canadian Terminology
Georelational Information System Technology-based Terminology
Natural Resources Information Disciplined-based Terminology
System
Geoscience or geological Disciplined-based Terminology
Information System
Spatial Information System Non-geographic derivatives
Spatial data analysis System Terminology based on what the
system does.
*The Table confirms that GIS is becoming extremely elastic resulting in an
increasingly confusing jargon due to new definitions that constantly creep into
both the scientific and popular literature.
68
8.4 Significance of GIS
One of the strongest and most successful application areas for GIS has been in
addressing problems of the environment. Virtually all resource management agencies
now have some form of GIS program and associated policies. However, this pattern
gives no impression of the diversity of GIS applications within agencies. The following
provides a categorization of different types of applications within the broad area of
problem solving:
1. GIS is a powerful tool for handling spatial data. Large quantities are maintained and
retrieved at greater speeds, lower costs per unit when computer-based systems are used.
The ability to manipulate the spatial data and corresponding attributes, and integrate
different types of data in a single analysis and at high speed are not matched by any
manual methods.
2. The analysis of complex, multiple spatial and non-spatial datasets in an integrated

manner forms major part of a GIS capabilities. These spatial analysis capabilities of
GIS, enable geo-referenced information to be created and used in a completely different
context which had not been seen anywhere before.
3. GIS frequently allows geographic entities to have multiple attributes, usually

including capabilities for storing and handling relationships between entities that also
include the capabilities of digital cartography in its input and output subsystems. Of
particular importance is the manner in which digital cartography and GIS handle
notions of continuous variation. Cartographically, such continuous variation are
represented in the form of map features, typically color contour lines and irregularly
spaced spot heights, and digital cartography that provides capabilities for processing
these as line and point objects.
In practice, however, many applications of GIS technology turn out to be little more
than digital cartography. The map is a very persuasive form of data display, and a
computer-drawn map carries the authority of a powerful technology.
69
4. There are so many ways of representing geographic variation and such diversity in
data structures between GIS products and databases include substantial facilities for
input and output of data in different formats. Other useful preprocessing functions
include the ability to extract information in a user-defined window, scale and projection
change including knowledge of such coordinate system as UTM (Universal Transverse
Mercator) and SPC (State Plane Coordinates); and capabilities for re-sampling. All
these give GIS as a manager of data, particularly if data must come from diverse
sources in mutually inconsistent formats.
5. Some of the greatest interest in the use of GIS for problem solving has come from
those who would apply the technology to translate the results of modeling into policy.
Post-processing is essential if the results of a spatially distributed model are to be used
for policy development. Results are often aggregated by administrative units, or
brought into consistency with social and economic data for comparison and correlation.
Finally, increasing use is being made of the paradigm of spatial decision support, in
which technology is made available directly to decision makers for scenario
development, rather than being confined to use by analysts.
Although the GIS as a system for handling geographical information come in many
different forms:
It may not necessarily be a solution to all problems. It thus requires considerable

thought for successful implementation.
There is no clearly agreed definition of when a computer system is or is not a GIS:
it is possible to identify the key characteristics, which distinguish these systems from
others.
Many GIS contain user-friendly map editing and formatting subsystems, and allow
the user with little cartographic training to make a convincing product quickly and
efficiently. But cartographers have accumulated centuries of knowledge and experience
about the effective visual communication of geographic information, and GIS may be
doing its users a disservice if it encourages them to ignore this (Buttenfield and
Mackaness 1991).
70
8.5 Questions GIS Can Answer
GIS has been described in two ways:
Through formal definitions, and

Through its ability to carry out spatial operations by linking data sets together.
However GIS is distinguished by the type of questions it is able to answer that
include:
(a) What is at a specified Location? This is the first of these questions that seeks to
find out what exists at a particular location. The location can be described in many
ways for instance, using name place, postcode, or geographically referenced
coordinates such as (longitudes, latitudes), or (x, y) coordinates.
(b) Are certain Conditions Satisfied? This is the second question, which is an
inverse of the first question, and requires spatial data to its answer. Here the questions
is not asking what exists at a given location, but rather wants to find the locations
where certain conditions are satisfied such as all the un-vegetated areas of at least
2000 square meters within 100 meters of a road and with soils suitable for supporting
buildings.
(c) What are the Trends? This questions combines the above first two questions and
seeks to find the differences within an area over time such as land use or elevation on
agricultural productivity.
(d) What Spatial Patterns Exists? This question is asked to determine the anomalies
that may exist or the phenomena that have a most recurrence in the area under study.
An important example is a question that may ask whether landslides are occurring
mostly near streams, or is there any land use that does not fit the pattern at the present
location?
(e) What if there is change? The what if questions are asked in order to determine
what happens is a new phenomenon such as a road is added to a network, or a toxic
71
substance seeps into the local groundwater supply. Answering this type of questions
requires both geographic and other types of information, as well as specific models.
8.6 The Development of Geographic Information Systems

The use of the term Geographic (or Geographical) Information System (GIS) dates
back to mid-1960s (Goodchild 1993), where it seems to have originated in two quite
different contexts. In Canada, it was devised to refer to the use of a mainframe
computer and associated peripherals (notably a scanner) to manage the mapped
information being collected for the Canadian Land Inventory, and to process it to
compute estimates of the area of land available for certain types of uses. A rigorous
analysis was used to show that a computer was the only cost-effective means of
producing the fast numbers of measurements of area required by the project, even with
the primitive and expensive nature of digital technology at the time, because manual
measurement of area remains an inaccurate and labour-intensive task. Much of the
proposed analysis was concerned with measuring areas simultaneously on two maps, to
answer questions like:
“How much area is class 1 agricultural land and not currently used for agriculture?”
The ability to overlay two or more maps for analysis (in this case a map of soil capability
for agriculture with a map of land use) provided the impetus for GIS development,
because it had been so cumbersome by hand.
Almost the same time, researchers in the United States were struggling with the problems
of accessing many different types of data required by large-scale transportation models
then in vogue, and conceived of a GIS as a system capable of extracting appropriate data
from large stores, making them available for analysis, and presenting the results in map
form (Coppock and Rhind 1991). Such models combined information on population
distributions with other spatially distributed information on places of employment and
transportation routes, and required access to data in a variety of formats.
72
Almost 40 years later, these same arguments are still among the most frequently heard
justifications for the use of GIS, particularly in geographical modelling and policy
development. It is now seen as a general-purpose technology for handling geographic
data in digital form, and satisfying the following specific needs, among others:
The ability to pre-process data from large stores into a form suitable for analysis
including such operations as reformatting, change of projection, re-sampling, and
generalization.
Direct support for analysis and modelling such that forms of analysis, calibration
of models, forecasting, and prediction are all handled through instructions to the
GIS.
Post-processing of results including such operations as reformatting, tabulation,
report generation, and mapping.
In all of these operations, the typical GIS user now expects to be able to define
requirements and interact with the system through a “user-friendly” intuitive interface
that makes use of such contemporary concepts as graphic icons and desk metaphors
(Mark and Gould 1991).
GIS has evolved dramatically since the early days of mainframe computing,
particularly in the past 22 to 25 years. Its first commercial success came in the early
1980s, primarily in resource management, but now there is a large software market in
local government, utility companies and a host of activities that use geographic data or
manage geographically distributed facilities.
The development of GIS can be viewed in relation to two other important areas of
geographic information handling by computer, namely:
(a) Computer-Assisted Cartography (CAC), and

(b) Remote Sensing and Image Processing.
73
8.6.1 Computer Systems
It is assumed that the students taking this Unit, AGE 400 have basic idea of the way in
which computers work, and it thus safe move to the next section. However, for
beginners in computer systems, it is strongly urged to seek a fuller introduction to the
development and role of computers in geography and planning in many textbooks such
as Exploring Geographic Information Systems by Chrisman 1997, pp.3 or Geographic
Information Systems by Martin 1996, pp.9-13)
8.6.2 Computer-Assisted Cartography (CAC)

Computer-Assisted Cartography (CAC) is an umbrella term used to cover a variety of
specialised systems for map and plan creation using a computer. In other words, it is the
use of many complex operations of computer in which conventional cartography is
performed, using a coded data representation to rapidly reproduce the conventional
product in a more flexible way. One such specialised system is Computer-Aided
Design (CAD), which is widely used in architectural and engineering design
environments, and which when applied to the production of maps, is known as
Computer-Aided Mapping (CAM). Within CAM lies the related aspects of automated
cartography, a term generally used for the production of conventional topographic maps
by automated means, and the more specialised research activity of statistical and
thematic map creation. According to Rhind (1977) CAC covers all aspects of map
making using computer assistance. The CAC has led to the development of vector GIS,
in which spatial coordinate storage and high resolution graphics have played an
important role.
Monmonier (1982) regards automation as an important step in the evolution of

cartography, which he portrays as a line of development stretching back to the earliest
recorded maps, and which includes the invention of printing in the 15th century. It was
noted that the way in which ‘softcopy’ map images have replaced paper maps in many
applications, outlines potential developments that points towards the integrated
geographic information system. Moellering (1980) classified the new computer
representations of maps as ‘virtual maps’ that offered more flexibility than the
traditional paper map, which he termed ‘real maps.’ The new virtual maps had only an
74
ephemeral existence on the computer screen and would avoid large numbers of paper
copies.
Rhind (1977) observes that few geographers or professional cartographers were

involved in the earliest attempts at drawing maps with computers. Initial developments
came from applications in geology, geophysics and environmental sciences. Although
there were suggestions for the use of computers for such cartographic tasks as map
sheet layout, name placement and the reading of tabular data such as population
registers, the early developments generally represented very low quality cartography.
They were designed to produce output on standard line printers, using text characters to
provide variations in shade density as shown in the Figure 8.0 below. The most famous
of these early programs was SYMAP (Synagraphic MAPping system), developed at the
Harvard Computer Laboratory. SYMAP and LINMAP (LINE printer MAPping), a
program written to display the results of the 1966 UK Census of population was seen as
representing the first generation of widely used computer mapping systems with the
apparent influence of hardware capabilities. SYMAP could produce maps for areas
with constant values or interpolate surfaces between data points to produce shaded
colour maps for spatially continuous variables, but the line-printer orientation meant
maps were limited in resolution and individual data cells could not be represented by
square symbols.
Figure 8.0 An Example of Line-Printer Map Output

********************Ooooooooooooooooooooooo1111111111111111
****************Ooooooooooooooooooooooooo111111111111111111
***************Ooooooooooooooooooooooooooooooooo11111111111
**********************Oooooooooooooooooooooooooo11111111111
******************************Oooooooooooooooooo111111111111
********************Ooooooooooooooooooooooooooooo11111111111
*****************************Ooooooooooooooooooooooooo111111
**********************************Ooooooooo1111111111111111
75
An advance was the use of coordinate-based information, and the use of pen plotters
such as in the NORMAP package developed in Sweden. This produced maps on either
plotter or printer, and much care was given to the methods of map construction, in this
case with data based mainly on points. For Nordbeck and Rystedt (1972), the technique
used for the presentation of the finished map formed an important part of this process
but even more important are the methods used for processing the original information
and for the actual construction of the map.
The use of coordinate data and plotter output was the starting point for the use of
automated cartography. As the capacity of hardware to store and retrieve digital map
data began to increase the advantages of storing data in this form became more
apparent. For example, a geographic database that was revised continuously would be
available on demand for the production of new editions without recompilation.
Generalisation and feature-coding algorithms allowed for the creation of hardcopy
maps at a variety of scales, and the complex issues of projection and symbology that
separated from the actual task of data storage: thus the real map ceasing to be the main
data store and becoming a merely a way of presenting selected information for the
customer’s requirements from the comprehensive cartographic database held in the
computer. It became possible to preview the appearance and design of a new product on
graphics screen before committing it to paper and production itself was much faster
once data were entered into the computer.
Morrison (1980) outlines a three-stage model for the adoption of automation,

suggesting that technological change has been an almost constant feature of
cartography. The three stages identified were as follows:
The early 1960s: rapid technical development, but a reluctance to use the new methods
and the fear of the unknown new technology;
Late 1960s and 1970s: acceptance of CAC, replication of existing cartography with
computer assistance;
1980 onwards: new cartographic products, expanded potentials and full
implementation.
76
According to Martin (1996) various reasons contributed to the slow take-up of CAC, which
included:
(a) High initial costs of entering the field with initial installations disappointing and thus
discouraging other map-producing agencies from taking the first step;
(b) The need to make heavy investments in hardware and software in an environment
where both were evolving rapidly;
(c) Transformation of enormous quantities of existing paper maps into digital format, with
the latest developments requiring surveying instruments that record all measurements
digitally for direct input to digital mapping systems.
(d) Related to (c) above is lack of implementation of efficient mass digitizing procedures,
which although conceptually simple has proven difficult to achieve.
8.6.3 Computer Aided Design (CAD) and Computer Aided Mapping (CAM)
Contemporary CAD and CAM systems provide the capacity to store a number of thematic
overlays of information (e.g., roads, water features, contours, etc.), each of which may be
associated with a particular symbology. Some specialized CAD software contains
sophisticated editing and three-dimensional modeling functions, while modern CAM systems
are more concerned with the ability to vary projection and symbolization of output maps.
However, neither have strongly developed attribute data handling capabilities and generally
store information as series of feature-coded points and lines without any associated
topological information or area-building ability. The ‘attribute data’ is here referred to as the
non-location characteristics of a particular geographic object, such as the population of a
zone or the name of a street.
8.6.4 Image Processing

RS may perhaps be traced back to the earliest aerial photography. However, massive growth
has only been experienced in the last forty years with the arrival of satellite sensing systems
that are returning enormous quantities of continuously sensed digital data. The next
technology of importance to GIS development has been that of remote sensing (RS) and the
activity of image processing (IP) of remotely sensed data. The use of data from satellite and
77
airborne sensors for the monitoring and management of the environment is also increasingly
referred to as earth observation. Satellite RS is the largest single source of digital spatial
data, and has therefore been a significant factor in the development of mechanisms for fast
and efficient processing of these data. The orientation here has been towards raster (image-
based) systems and the entire field has only been made possible by the development of
automated systems, which are able to cope with the volumes of data involved. In raster
systems, the map is in the form of a grid, each cell of value relating to that region.
IP systems have been developed to handle satellite-sensing systems that are returning
enormous quantities of continuously sensed digital data. There is an increasing range of
‘integrated’ IP/GIS systems such as ERDAS and IDRIS that contain a range of functions for
image interpretation and manipulation together with more general-purpose raster GIS
abilities. The raw data received from the satellite sensors require constantly radiometric and
geometric correction before any attempt can be made at image interpretation, and a number
of strategies exist for the interpretation operation itself. Image data consist of a matrix of
digital numbers (DN) for each band, each cell (pixel) of which represents the value recorded
for the corresponding area on the ground.
Usually, there is need to remove known distortions resulting from the curvature of the earth
and suppress the effects of atmospheric scattering and satellite wobble. The values recorded
may be affected by skylight and haze and, over large areas differential degrees of shade and
sunlight. Radiometric restoration and correction consist of the adjustment of the output of
each detector to a linear response to radiance. Geometric correction, on the other hand
involves changing the location of lines of pixels in the image and complete re-sampling with
reference to ground control points of known locations in order to georeference the image
data. Stretching of the distribution of values in each band to utilize the whole DN (8-bit)
density range, usually in a linear fashion enhances images.
Once the image data has been ‘cleaned’ the major task remaining is one of image
classification. If RS data are to be interpreted with existing GIS or digital map data, it is
necessary to interpret the DN values and to combine and classify the image bands according
to some recognizable classification scheme.
78
8.6.5 Contemporary GIS
These definitions all emphasize the historical development of GIS as a combination of
Computer Aided Design (CAD) and Digital Cartography handling capabilities integrated
with database software. Although these definitions may help:
(a) GIS as a system for handling geographical information come in many different forms
and, may not necessarily be a solution to all problems. It will thus require considerable
thought for successful implementation.
(b) It must also be stated that there is no clearly agreed definition of when a computer
system is or is not a GIS: it is possible to identify the key characteristics, which distinguish
these systems from others.
8.7 Summary
GIS is a system of hardware, software, data, people, organizations, and institutional
arrangements for collecting, analyzing, storing, and disseminating information about
areas of the earth.
The significance of GIS is seen in the categorization of five different types of
applications within the broad areas of problem solving: (i) a powerful tool for handling
large quantities of spatial data, (ii) analysis of complex, multi-spatial and non-spatial
datasets in an integrated manner, (iii) ability to allow geographic entities to have multiple
attributes such as storing and handling relationships between entities, (iv) ability to input
and output of data in different formats, and (v) modeling.
The term geographic information system (GIS) dates back to mid 1960s.
The development of GIS can be viewed in relation to computer aided cartography
(CAC), remote sensing (RS), and image processing (IP) of geographic information using
a computer.
CAC is an umbrella term used to cover a variety of specialized systems for map and
plan creation using a computer.
Computer aided design (CAD) and Computer aided mapping (CAM) systems
provide the capacity to store a number of thematic overlays of information (e.g. roads,
water features, contours, etc) each of which may be associated with a particular
symbology.
79
Remote sensing is the largest single source of digital spatial data.

(a) Define GIS.
(b) Trace the major historical developments of Geographical Information system (GIS).
(c) Differentiate between GIS and CAC.
(d) Describe the significance of geographic information system.
(e) Describe the questions that must be answered by a GIS.
8.6 Further Reading

Bracken, I. and C. Webster (1990): Information Technology in Geography and
Planning Including Principles of GIS, London: Routledge.
Burrough, P.A. (1986): Principles of Geographical Information Systems for Land
Resource Assessment, Monographs on Soil Surveys 12, Oxford: Oxford
University Press.
Huxhold, W.E. and G.A. Levinson (1995): Managing Geographic Information System
Projects, New York: Oxford University Press.
Maguire, D.J. and M.F. Goodchild (1994): GIS: Principles and Applications, London:
Longman.
Martin, D. (1996): Geographic Information Systems: Socio-Economic Applications,
London: Routledge.
Rhind, O. and J. Raper (2001): “GIS: Time for a Re-think.” In GeoEurope, No.10, Issue
5: pp 47-48.
80
9.0 LESSON NINE: GIS COMPONENTS AND FUNCTIONALITY
9.1 Introduction
Lesson nine examines the basic components of GIS. It also deals with GIS software
programs used for the analysis of geographic information, with some explanations on
differences between GIS software programs. The lesson is concluded with GIS
functionality.

By the end of studying and reading this lesson, you should be able to:
(a) Describe the basic components of GIS.
(b) Describe GIS software programmes.
(c) Explain GIS functionality.
9.3 GIS Components

Geographic Information Systems consist of four basic components (Figure 9.0). These
components are described here below as follows:
9.3.1 Collection, Input and Correction

(a) Collection
These data can be collected through image processing and interpretation; field mapping
and monitoring; laboratory analysis; and collection of existing data.
(b) Input and Correction

The data input component converts data from their existing format into one that can be
used by the GIS in a process called digitization. Geo-referenced data are commonly
provided as paper maps, tables of attributes, and electronic files of maps, air
photographs and satellite imagery.
The data to input in a GIS are mainly of two types: spatial data and associated attributes
of non-spatial data. The spatial data represent the geographic location of the represented
81
‘objects”: points, lines and areas that are used to represent geographical elements such
as roads, lakes, rivers, etc. The non-spatial attribute data provide descriptive
information like the name of a street, the hydrologic composition of Water Lake, etc.
Figure 9.0 A Typical Model of GIS as Components
INPUT
COLLECTION, INPUT AND

CORRECTION
STORAGE AND RETRIEVAL
DATABASE
MANAGEME
MANIPULATION AND
NT SYSTEM
ANALYSIS
OUTPUT AND REPORTING
OUTPUT
82
There are 4 types of data entry systems commonly used in Geographic Information
Systems:
i) Keyboard
ii) Manual Digitizing
iii) Scanning
iv) Existing Digital Files.
9.3.2 Storage and Retrieval (Management) mechanisms include the control of

physical storage of data in memory, disk or tape, and mechanisms for retrieval to
serve the needs of the other three components. Storage and retrieval entails a
database, which is a collection of information about things and their relationships to
each other. The objective of a D-base is to relate facts and situations that were
previously separated. The creation of a D-base in which different kinds of data are
combined by:
Ensuring that all data layers have the same dimensions and coordinate
systems,
Designing a useful D-base structure for the tabular data, and
Establishing links between map and tabular data.
Note that the data management part of GIS is comparable to other D-bases and
therefore an exchange of data files mainly of non-spatial data is usually possible.
9.3.3 Manipulation and Analysis represents the whole spectrum of techniques

available for the transformation of the digital model by mathematical means. These
are the core elements of a GIS, and are the features that distinguish GIS from
Computer Assisted Cartography (CAC). A library of data-processing algorithms is
available for the transformation of spatial data, and the results of this manipulations
ma be added to the digital database and incorporated in new visual maps. Using
these techniques it is possible to deliberately change the characteristics of the data
83
representation in order to meet theoretical requirements. It is equally possible to
mishandle or unintentionally distort the digital map at this stage.
Manipulation and analysis allow the efficient use of large volumes of data in
planning and decision-making by providing a rapid method for evaluation of past
and potential developments in relation to management measures and decisions. For
example, the use of the following models does assist in the management of data in a
Geographic Information Systems:
Aggregation: permits combination of detailed map categories in order to

create new detailed categories;
Buffering: creates a zone of specified width around one or more features, for
example, an area within 100m of a lake;
Network analysis: permits such determinations as finding the shortest path
through a street network, determining stream flow in a basin, or finding the
optimum location of a fire station;
Inter-visibility operations: use elevation data to permit viewshed mapping of
what terrain features can be seen from a specified location.
9.3.4 Output and Reporting. This is a procedure by which information from the
GIS is presented in a form suitable to the users. It involves the export of data from
the system in computer- or human-readable form. Data are output in one of the 3
formats: hardcopy, softcopy or electronic file.
Hardcopy outputs are permanent means of display. The information is provided on

paper, photographic films or similar materials. Softcopy output is the format as
viewed on the computer monitor. They are used to allow operator have interactions
and preview the data before the output.
9.4 GIS Software

Currently, there are several software programs in the market for analyzing
geographic information. The programs available for PC/Microcomputers and other
84
types of computers include EASI/PACE®, ERDAS (v), IDRISI®, ILWIS (rv),
MIPS (rv), PAMAP (rv), ARC/INFO (rv), PMAP ®, SPANS (rv), STRINGS (v),
TERRITORY (v), and TOPOLOGIC (rv), among others.
The differences between the above GIS programs can be summarized as follows:
Hardware type and requirements

PC-DOS is the most common operating system with a few GIS using OS2 or the
MacIntosh operating systems. All the above programs have extensive digitizing
capabilities: with some programs working on a hardware configuration with two
monitors, while others use one for text and menus and another one for maps.
Currently, many GIS experts are working towards the development of a GIS system
with a single screen of very high-resolution monitors in the Windows.
Data structure
Although many analytical systems can handle and process vector in a limited way
by making a provision for vector-to-raster conversions, few systems can process
quadtree data structure because they do no have extensive image processing
capabilities. All the systems mentioned above are two-dimensional, and the three-
dimension systems such as LYNX, and Intergraph are, however, not available for
PC’s.
Internal database
Only a few programs possess an internal database management system, and those
without, rely on external Database Management Systems (DBMS). These programs
have a provision for table import/export and to or fro some of the most common
systems.
Image processing facilities

Some programs are hybrid GIS and image processing programs such as IDRISI or
ILWIS, whereas others such as ERDAS, EASI/PACE and ARC/INFO are mainly
image processing programs.
85
Command language
The use of command language with macros and batch files is indispensable for those
applications that require complex modeling or many repetitive manipulations.
However, many programs lack this option.
Menu-user interface
Practically, all programs have a menu-user interface with some making use of
Windows for displays. Some programs such as IDRIS and ILWIS are designed with
academic environment such as teaching tools, and thus making them very suitable to
use.
Price
Prices for PC-based GIS range from a few hundred dollars to tens of thousands.
When purchasing a GIS program, it is important for the users to give priority to
design of tasks it can perform but not costs.
9.5 GIS Functions

The heart of GIS is formed by the analytical capabilities of the system. Although data
input is in general the most time consuming part, it is data analysis that forms the
most important element of GIS. The analytical function employs the spatial and non-
spatial attributes in the database to provide answers to the questions in the real world.
The objective of data analysis is to transform data into information in order to satisfy
the requirements of decision making at all levels of details. An important use of the
analysis is the possibility of predicting the occurrence of activities in one location or
point in time.
The range of analytical procedures can be subdivided into four basic categories:
86
9.5.1 Retrieval, Reclassification and Measurement Operations
(a) Retrieval
Data retrieval involves selective search, manipulation and output of spatial and/or
attribute data. Retrieval operations can be subdivided into the following groups:
To find out what exists at a particular location. In GIS, programs exist that permit
reading of the contents of maps and related attribute tables for a specified (x, y)
coordinate in vector or in a pixel (rater data).
To retrieve spatial data such as points, lines, polygons or mapping units in a raster
figure that may answer the question: where is it? For example, such as: where is the
forest?
To retrieve information using conditional, logical and arithmetic operations.
(b) Reclassification
Reclassification involves the re-assignment of thematic values to categories of an
existing map. It is also called renumbering. Typical examples of reclassification are:
A soil map can be changed to include rate of infiltration; or

An elevation map reclassified to show contours intervals of 50m.
It important to recognize that use of any formulae works well when there are few
units for reclassifying maps.
(c) Measurement Operations

This operation involves the measurement of distances between points, lengths of
lines, area, perimeter of polygons and determination of volume and counting. Some
examples of measurement operations are:
The distance between two traffic cycles,

The total acreage of wetlands in a watershed,
Total length of first order streams in a catchment, and
Counting of numbers of first order streams in a river basin.
87
Most of the abovementioned operations such as computation of the area, or perimeter
of the polygon, or length of lines are automatically performed. For other operations,
special programs are required. One such example is ILWIS Program that calculates
distances in meters from pixels in a map with a value of 0 (source).
9.5.2 Overlay Operations

Overlay Operations involve the combination of two or more maps according to
Boolean conditions (e.g. If A is greater than B and A is less than C) and may result in
the delineation of new boundaries. In such cases, it is therefore essential that the
spatial and attribute data are a correct representation of the real-world phenomena. An
example would be the overlay of an enterprise zone on to a base of census wards.
This would be appropriate to determine the ward composition of the zone, but may
not allow an accurate estimate of the population falling within it, as may not be exact
coincidence of the boundaries. Thus the operation is only appropriate if the intended
interpretation of the data is meaningful.
9.5.3 Neighborhood Characterization or Operations

Neighborhood Characterization involves ascribing values to a location according to
the characteristic of the surrounding region. Such operations may involve both
summary and mean measures of a variable, and include smoothing and enhancement
filters. These techniques are directly analogous to contextual image-classification
techniques to be found in image-processing systems.
Using these operations, the characteristics of an area surrounding a specified location

are evaluated. As these operations are generally applied to the data in raster format,
the surrounding (neighborhood) is defined as the eight-raster cells surrounding the
central cell.
The neighborhood operations are subdivided into three important functions:
88
(a) Retrieval functions that use the values of its neighbors to calculate values for a
specific pixel. This is done by moving small calculation window of 3 by 3 cells,
or 5 by 5 cells over the map, and shows the calculated results in the central pixel.
The retrieval operations can be executed both on scalar data (e.g., water depth,
elevation), or on ordinal (e.g., class 1, class 2,), or on nominal (e.g., lithological
type) data. On the scalar level, properties such as average, standard deviation,
total, range, majority, and maximum/minimum can be calculated. For nominal
and ordinal data, all the statistics can be computed except mean, standard
deviation and total, which have no meaning except when the pixels represent
geological units.
(b) Interpolation functions, which estimate unknown values at sampled sites using
known values of existing observations at neighboring locations. In other words, it
is an operation in which values are predicted for locations without observations
based on the known observations of neighboring pixels. There are quite a number
of interpolation methods such as polinomian regression, fourierseries, weighted
means, Thiessen polygons, and kriging, whose applications differ from cases to
case.
(c) Topographic functions, which compute values that describe the topography at
specific locations using local terrain information. Some of the values that can be
obtained using the operation include digital elevation models such as slope angle,
direction, convexity, hillshading, automatic catchment delineation, drainage
pattern, and slope length computation.
Neighborhood operations working on topographic input data can also be used for
the so-called dynamic modeling: calculating groundwater flow, the turnout
distance of debris flow, the area occurred by lavaflows, etc. The models are found
among the existing finite element models that have been implemented in GIS.
89
9.5.4 Connectivity Operations
Connectivity and Distance measurement include both simple measures of inter-point
distance and more complex operations such as the construction of zones increasing
transport cost away from specified locations. Some systems will include sophisticated
networking functions tied to geographic database. Connectivity operations include, e.g.
view-shed analysis that involves the computation of inter-visibility between locations in
the database. In other words, the use of these operations permits the characterization of
spatial units that are connected.
This group of operation can be subdivided into:
Contiguity functions: these are functions that connect areas that share common
characteristics such as badlands as may be affected by gullies. In very general terms,
these types of functions are applied in order to identify or retrieve areas with specific
size and characteristics.
Proximity functions: connect areas with same distance in absolute distance, time
and cost to a given point, line or area.
Network functions: using lines to connect areas that form a set of interconnected
features through which resources move from one location to another. Note that a
network is formed by a series of interconnected areas or lines that form a certain feature
such as road network, or river system via which resources can be transported. The
network functions are important in three major ways: (a) prediction of the capability of
the network, for example, discharge modeling along a stream network; (b) calculation
of optimal route, for example, the shortest way to a fire outbreak from fire-brigade
tracks; and (c) identification of resources, for example, the division of a functional
region such as a city into zones with a defined size for planning of first aid centers for
which the supply of ambulances can take place within a specified period of time.
Spread functions: these are a mixture of proximity and network functions, which are
used to evaluate phenomena that are accumulating, or dispersing with distance from a
90
certain source. For example, the spreading of groundwater pollution from a point
source or the accumulation of water behind a dam.
Seek functions: usually using lines to connect areas that form an optimum pathway
according to specified rules. A good example of this is the determination of water
pathway flow, which known to flow only to lowest neighboring points.
9.5.5 Modeling
Basically, a GIS consists of static models or descriptions of the past or existing
situations. Manipulation of data by various functions facilitates the understanding of
relationships and constraints but does not add any new information. On the basis of the
distribution of spatial elements and characteristics, predictions can be made about
processes and impacts such as soil erosion, flooding, and anthropogenic influences.
Consequently, its capability to evaluate present conditions together with changes
resulting from different interventions and management measures that provides a
convenient tool for decision-making.
Therefore, GIS is a flexible and powerful tool for the storage and analysis of data on
impact monitoring. This range of applications is even wider by combining dynamic
simulation models that have non-spatial attributes.
9.6 Summary
The four basic components of GIS are (a) collection, input and correction, (b)
storage and retrieval, (c) manipulation and analysis, and output and reporting.
There are several software programs for analyzing geographic information, with
differences in these programs including hardware type and requirements, data structure,
internal database, image processing facilities, menu-user interface, and command
language.
The most important functions of GIS are divided into (a) retrieval, reclassification
and measurement, (b) overlay operations, (c) neighborhood characterization, (d)
connectivity operations, and (e) modeling.
91
9.7 Key New Terms
Quadtree. This is when a data structure subdivides any given space into 4 quadrants
and continues to subdivide each quadrant in a similar way until they are uniform or the
basic resolution of the data is reached. It is mostly used to compress raster data.
Topology. These are relationships in spatial terms between connected or adjacent

geographical entities. It is used to apply intelligence to data held in the vector data
model.

(a) What are the components of GIS?
(b) Describe the functions of GIS.
9.9 Further Reading

Bracken, I. and C. Webster (1990): Information Technology in Geography and
Planning Including Principles of GIS, London: Routledge.
Burrough, P.A. (1986): Principles of Geographical Information Systems for Land
Resource Assessment, Oxford.
Huxhold, W.E. and G.A. Levinson (1995): Managing Geographic Information
System Projects, New York: Oxford University Press.
Maguire, D.J. and M.F. Goodchild (1994): GIS: Principles and Applications,
London: Longman.
Martin, D. (1996): Geographic Information Systems: Socio-Economic
Applications, London: Routledge.
Rhind, O. and J. Raper (2001): “GIS: Time for a Re-think.” In GeoEurope, No.10,
Issue 5: pp 47-48.
92
10.0 LESSON TEN: SPATIAL DATA MODELS
10.1 Introduction
This lesson describes how complex reality can be simplified using spatial data models.
It explains the four spatial elements: point, line, area and surface are used for the
simplification of reality. It also describes raster, vector and object-based models that
have considerable influence on the functionality of the GIS. The lesson is concluded
with the description of GIS data structure.

By the end of studying and reading this lesson, you should be able to:
(a) Define spatial data model.
(b) Explain source of environmental data for spatial data models
(c) Describe the raster, vector and object-based models.
(d) Describe GIS data structure.
10.3 Spatial Data

Spatial data is information, which is linked to a specific location, for example, the
population of a town or the occupant of an address. The term geocoding is used to
describe the process of setting up data for a GIS that link information to a location. For
a GIS, within a particular data set there must exist an element that specifies its location.
Ideally this would be a map coordinate, but it can also be a postcode or street address.
The element within the data that identifies the location is known as its geocode.
Data sets can be divided into those about people, socioeconomic and those concerning
the environment:
(a) Socioeconomic data is widely available, often from national and local government
and is usually the product of population surveys and censuses. These data is also used
by a number of commercial vendors who combine census information with other data
sets to produce neighborhood profiles classifying particular areas for marketing
93
purposes. This ability to recognize particular markets based on geographical datasets is
known as Geodemographics and is one of the fastest growing areas within GIS.
(b) Environmental data are related to collection and analysis of information about the
environment, and these data were the driving forces behind the development of GIS and
continues to be an important area of GIS application. Environmental data sets often
tend to be large and require considerable management.
Sources of environmental data include:
Existing topographic maps,

Thematic maps, specifying geological aspects, soils, etc.
Remote sensing. Sources of remote sensing data for GIS applications in various
environmental aspects include; aerial photographs, LANDSAT, SPOT, NOAA,
METEOSAT and ground based weather radar. Databases resulting from the
joint use of several distributed information maps are able to support a series of
applications in different fields of environmental management and social-
economy. Users select pertinent information for the different purposes.
Environmental data often includes boundaries between vegetation types, for example,
which are fuzzy, i.e., a simple line does not define them. Conversely, socioeconomic
data is usually related to administrative boundaries, which are sharp if artificial.
10.4 Spatial Data Models

A data model is a generalized user-defined view of data representing the real world. It
is used because reality is too complex for even the most sophisticated GIS software to
represent. In order to represent reality in a spatial database, a simplification of reality is
created. This simplification is known as a spatial data model. In a spatial data model,
reality is simplified into four spatial entities or elements, which can be used to represent
the real world. These four spatial entities are:
The point
94
The Line
The Area, and
The Surface.
In a telecommunications GIS application, for example, a point may represent the

location of a junction box; a line might represent a section of a cable; an area may
represent a building; and surface may represent the land surface through which cables
are laid (Ed. Parsons). Other commonly encountered examples of each of the spatial
entities are given in Figure 10.1. Distance (or length) is the fundamental geographic
dimension, and spatial entities may be classified according to the number of length
dimensions they possess: zero for a (x, y) point, one for line, two for an area and
three for a surface.
Figure 10.0 Examples of Spatial Entities

Entities Point Line Area Surface
Class
Dimension 0 1 2 3
Example Fence post Road section Land parcel Physical
terrain
+ Post Plot
Attributes are then attached to these spatial entities, for example, the type of cable,
the address of the building and the height of any particular point. The linking of
spatial entities with their attributes is one of the key concepts of GIS.
Spatial entities and their attributes are stored using a number of spatial data models by
specific GIS software, and it is therefore important to understand the characteristics
of each since the data model used has considerable influence of the functionality of
the GIS. The spatial models are:
95
(a) The Raster Model, and
(b) The Vector Model.
10.4.1 The Raster Model

The raster data model is the simpler of the two and is based on the division of reality
into regular grid of identically shaped cells. In other words, the data is presented by
subdividing the space regularly into cells: the location of geographic entities or
phenomena is defined by row and column position of the cells they occupy. The area
that each cell represents defines the spatial resolution available. Given that positions are
defined by the cell row and cell column numbers the position of a geographic feature is
only recorded to the nearest cell.
In the raster approach, the space is populated by a large number of regularly distributed
cells, with each cell assigned a single value, which represents the attribute for the area
of that cell. In a soil data set, for example, a cell may have a value of 001, which may
represent clay soil. The area that each cell represents varies from a few meters to
kilometers and is known as the resolution of the grid. The higher the resolution of the
grid, the more cells are required to represent a given area.
10.4.2 The Vector Model

Cartesian coordinates, commonly used to represent linear features, represent
geographical objects in this model. Each feature is represented by a series of
coordinates which define its shape and which can have linked information. In other
words, the vector data model is similar in its operation to joining the dots in the books
we all used as kids. An object’s shape is represented by dots, which are located where
the shape of the object changes. Straight lines join the dots. In the vector data model the
dots are known as vertices. A straight line known as a segment or arc joins each
vertex. Where segments join, the vertex is called a node. A series of segments that
return to the same node form an area or polygon.
The vector data model appears to be similar to the data structure used by Computer
Aided Design (CAD) systems and simple computer drawing packages. The GIS vector
96
data model is slightly more complex at each vertex, segment, node and polygon is
uniquely identified and the relationships between them are stored in the database.
The relationships between the elements of a vector data model, in terms of relative
location and connections are known as Topology. Topology gives the vector data
model a level of ‘intelligence,’ which means that the GIS can recognize which
segments are joined to each other, and identify those polygons that are adjacent to one
another.
The vector data model is best suited to representing linear features such as cable
networks, and the relationships between areas. The main drawback of the vector model
is when datasets are combined and analyzed, as a much higher level of processing is
required.
In addition to these two main data models, a third, the object-based model exists and
has become increasingly popular. This represents the world in the form of objects the
user would recognize, i.e. on a highway; it would represent the whole highway and not
the individual segments that go to make it up. This has a number of advantages since
the model is less abstract and easier to understand. However, the processing
requirements are high.
10.4.3 GIS Data Structure

Data structure is a logical arrangement of spatial data in a format suitable for the user
and user system to manage. For one to use a raster or vector data model it is important
to make a decision which model may be better suited to the application. However, the
choice of data structure to be used for any application is often an arbitrary decision,
since GIS software will generally support one particular model as fully as another.
Whichever model and structure one may choose, of course, one requires time to convert
the data into a format that can be used by the GIS. Converting data into digital format is
a labor-intensive activity, and on average accounts for up to 80% of the total system
cost.
97
Central to any data capture plan is a thorough internal data audit. This helps to
determine the size, scope and cost of the task ahead. Data capture is also an opportunity
to improve the quality of the data by incorporating new information with the old.
10.4.4 Scanning or Table Digitizing

Scanning is a data capture technique that digitizes information from paper or film hard
copy into digital raster data. The process is rapid, but the resulting raster data set only
has color, grey scale or black and white attributes associated with it. In effect, the result
of scanning is a raster image of the original source material. Scanning offers ease and
speed, but the resulting raster images lack the intelligence needed for vector-based GIS.
A fair degree of operator expertise is also required and compression techniques
(typically run-length encoding) will need to be applied to keep the files to a manageable
size.
On the other hand, table digitizing also known as vectorisation are the two common
methods of converting data. On the other hand, vectorisation can be applied
automatically or interactively to produce intelligent vector files.
Vectorisation has the advantage of employing inexpensive digitizing equipment.

However, operator training is needed to obtain good results. Conversely the procedure
is laborious, time-consuming and hence costly.
10.5 Summary
Spatial data is described as information, which is linked to a specific location, for
example, the population of a town or the occupant of an address.
This ability to recognize particular markets based on geographical datasets is known
as Geodemographics.
Sources of environmental data for GIS models include existing topographic maps,
thematic maps, and remote sensing.
A data model is a generalized user-defined view of data representing the real
world.
98
In a spatial data model, reality is simplified into four spatial entities or elements,
which can be used to represent the real world. These four spatial entities are the point,
the line, the area, and the Surface.
The spatial entities and their attributes are stored using a number of spatial data
models by specific GIS software. The models that are known to have considerable
influence of the functionality of the GIS are raster and vector.
Data structure is a logical arrangement of spatial data in a format suitable for the
user and user system to manage.
10.6 Key Terms

Attribute. This is an item of text, a numeric value or an image that is characteristic of a
particular spatial entity.
Coordinate. Coordinate are numbers representing the position of a point relative to an

origin. Cartesian coordinates express the location in two or three dimensions as the
perpendicular distances from two or three orthogonal axes.
Data Model. It is a generalized user-defined view of data representing the real world.
Digitizing. This is the conversion of existing maps from an analogue form (paper) into
digital information, usually in the form of Cartesian coordinates. This may be via a
digitizing table or tablet with a hand-held cursor, or via a scanner.
Network. It is a model representing the interconnected elements through which some

form of resources can be transmitted or will flow. In GIS this is represented as series of
nodes connected by segments, which has attributes representing flow characteristics,
e.g. a road or pipeline network.
Point. This is a spatial entity that represents the simplest geographical element;
represented in the vector model as a single (x, y) coordinate, and in the raster model as
a single cell.
99
Polygon. Polygon is representation of an enclosed region defined by a segment or a
series of segments that make up its boundary. Polygons may have attributes describing
the region they represent such as the population of a census ward.
Raster. It is a data structure composed of a grid cells. Groups of cells represent

geographical features; the value in each cell represents the attribute of the feature.
Scanning. It is a data capture technique that digitizes information from paper or film
hard copy into digital raster data.
Segment. This is one of the basic spatial entities, and a basis for spatial models. It is
formed from a set of ordered coordinates (vertices) that represents the shape of a
geographical object. A segment begins and ends in a node.

(a) What is a spatial data model?
(b) Demonstrate how the four elements, or spatial entities can be used to
represent the real world.
(c) Describe the three spatial data models that have considerable influence on
the functionality of the GIS.
(d) Differentiate between Scanning and Table Digitizing.
(e) Define a data structure.
10.8 Further Reading

Antenucci J. C., Brown K., Croswell, P.L and Kevany, M.J. (1991): Geographic
Information Systems: A Guide to Technology; New York: Chapman Hall,
301 pp.
Blackmore, M. (1983): Generalization and error in Spatial Databases. Proc. Auto-
Cart, 6 pp 313-322.
Bracken, I. and Webster, C. (1990): Information Technology in Geography and
Planning Including Principles of GIS. London: Routledge.
Burrough P.A (1986): Principles of Geographic Information Systems for Land
Resource assessment, Oxford: Clarendon Press.
Clark, M. J. (1998): “Putting Water in its Place: Perspectives on GIS in Hydrology
and Water Management: Hydrological Processes 12, 823-834.
100
De Roo A. P.J Hazelhoff, I and Burrogh P. A. (1989): Soil Erosion Modeling Using
ANSWERS and Geographic Information System; Earth Surface Processes
Landif, 14, 517-532.
Goodchild, M. F. (1980): The Effects of Generalization in Geographic Data
Encoding. In: Freeman, H. and Pieroni, G. (eds) Map data processing, New
York: Academic press, Pp 191-205.
Goodchild, M.F, Parks, B. and Staeyart, L. (1993): Environmental Modeling with
GIS; New York: Oxford University Press.
Heuvelink, G. B. and Burrough, P. (1993): Error Propagation in Cartographic
Modelling Using Boolean Logic and Continuous Classification. In:
International Journal of Geographic Information Systems. Vol. 7 No. 3
Taylor and Francis Ltd. London. Pp231-247.
Maidement, D. R (1993): GIS and Hydrologic Modeling. In Environmental
modelling with GIS. (Eds. E.M. Goodchild, B. Parks & Staeyart, L.; Oxford:
Oxford University Press.
Marble, D.F. and Amundson S.E (1988): Microcomputer Based Geographic
Information System and theirRole in Urban and Regional Planning.In
Environment and Planning, 15:305-324.
Masser, I and Campebell, H. (1996): The Impact of GIS on Local Government in
Great Britain. In: Geographic Information Handling- Research and
Applications. Ed. Paul M Mather. London: John Wiley and Sons Ltd.
Muller, J. C. (1991): Generalizations of Spatial databases. In: Maguire, D.J,
Goodchild, M. F. and Klind, D. W. (eds), Geographic Information Systems:
Principles and Applications. Harlow: Longman, pp457-475.
Reusing, M, Schneider, T and Ammer, U. (2000): Modeling Soil Loss Rates in the
Ethiopian Highlands by Integration of High Resolution mom-02/d2 Stereo-
Data in a GIS, In: International Journal of Remote Sensing Vol 21 No.9.
Taylor and Francis Ltd. London. Pp 1885-1896.
Thurner, T, Jones, A. B. and Brown D. (2000): A GIS based Coastal Management
System for Climate Change associated flood risk assessment on the East
Coast of England. In: International Journal of Geographic Information
Science, Ed. Fischer P. Miller, H. and Lees. Vol. 14 No. 3. Taylor and Francis
Ltd. Pp 265-282.
Viex, B. E. (1991): Geographical Information Systems and Non-Point Source Water
Quality and Quantity Modeling; Hydrological Processes, 5 101, 101-113.
Xiang, W. (1993): A GIS Method for Riparian Water Quality Buffer Generation. In:
International Journal of Geographic Information Systems VOL. 7 No.1
Ed. J.T. Coppck and K. C. Clarke. Taylor and Francis Publishers, London.
Pp57-70.
101
11.0 LESSON ELEVEN: APPLICATIONS OF GEOGRAPHICAL
INFORMATION SYSRTEMS
11.1 Introduction
This lesson examines the need for GIS applications in resource management: resource
mapping, risk identification, prediction, modelling and forecasting, among others. It
also describes the opportunities and constraints a GIS may have in its use for support
of public policy decisions regarding natural resource use such as land, water, forests,
fish and wildlife.

(a) Evaluate the need for GIS applications in resource management.
(b) Describe the opportunities and constraints a GIS has in its use for support
of public policy decisions regarding natural resource use such as land,
water, forests, fish and wildlife.
11.3 General Applications

A GIS application involves interdisciplinary studies in social-economic and
environmental sciences and its strength lies in its capability to combine independent
datasets. The GIS technique deals with themes that integrate geographical and non-
geographical information. Environmental databases for land registration and
comprehensive topographic data for every country have been emphasized in Europe.
In Canada, for example, emphasis has been put on forests access with timber
harvesting potential. In Asia, especially, China and Japan, monitoring and modeling
of environmental changes have been of great interest.
Nearly in all places, there are large amounts of data in space and governments can use
it to plan, zone, assess property and land records, public safety, management of
forested, agricultural and coastal lands. Further, a GIS can support daily activities by
automated mapping and facilitation of management of electricity supply,
telecommunications and sewer management. In demography, a GIS is used to target
market analysis and facility siting, among other uses in research. Since all these
102
diverse applications are carried out using similar software and techniques, a GIS is a
general purpose tool that may be used to integrate data from various sources to
determine resource availability, plan ecological aspects, transport, urban and
landscape development, among other uses.
Potential GIS end users include resource and environmental planners, land drainage
engineers and hydrologists. They all need informed decisions in their areas of
specialization and need GIS systems that are easy to use, that can perform spatial
analyses and modeling, support common data formats and can be able to produce
quality outputs. The output leads to policy formulation in many areas such as water
resources; conservation plans; and this can influence politicians and stakeholders who
need information on an easier, assimilative and understanding ways. This helps in the
management and visualization of geo-referenced data. Different applications
however call for different requirements on a GIS. Some of the applications are
summarized here in figure 11.0.
11.3.1 Environmental Planning and Management

Environmental decision-making requires a large number of variables from nearly all
disciplines. The environmental information is inherently spatial in nature and all
variables observed are considered to be a function of space: be it physical transport of
contaminants, extent of health or economic risks associated with remedial activities.
A GIS plays an important role in environmental information management by offering
spatial data analysis and management capability in the application of Environmental
Decision Support Systems (EDSS) as management tools.
An EDSS is a computer-based system supporting human decision makers in

environmental management arena. They focus on particular class of problems designed
to support a particular category of decision makers. It often integrates monitoring data,
providing a coherent view of these data and enhancing human comprehension of them.
103
Figure 11.0 Some Applications of Geographical Information Systems
Planning Public safety Public works Highways and

transportation
GIS
Parks and
recreation Telecommunica
tions
Environmental
protection Resource
inventories Electricity
Economic Impact
development assessment Water and
wastewater
Land use
Health and planning
human services
Permits and
licensing
104
Through the use of modelling and simulation, the EDSS support the analysis of
candidate decision alternatives, and help evaluation of the uncertainty of the problem at
hand. In other words, GIS provide a general platform for analyzing spatial data while
EDSS provide analytical procedures targeting specific problems. GIS such as
Geographic Resource Analysis Support System (GRASS), which is a GIS modelling
product of the US Corps of Engineers supports EDSS by facilitating data compatibility
with other common GIS’s.
The GIS that allow for real time presentation of development processes have a way of
dealing with environmental problems in a systematic way, solving complexity through
multi-disciplinary work and can optimise decision choices based on all involved
features. Indeed, environmental issues are among the most important challenges facing
decision makers and the dynamics of the earth systems imply that all environmental
systems are dynamically and spatially inter-related and impact in one location have
effects on others. For most environmental modelling projects, a GIS is seen as a
convenient and well-structured database for handling large quantities of spatial data
needed. Whereas GIS databases contain information on location, spatial distribution
and spatial relationships, environmental models work on mass and energy transfer and
the two can be integrated using dynamics and continuity. The environmental processes
on the earth’s surface need to be modelled for purposes of environmental protection of
air quality and water quality.
Erosion Models have been interfaced with GIS using Universal Soil Loss Equation in
Illinois, USA by adopting vector-grid approach. The drainage basins used illustrated a
full integration of topographic analysis within a GIS. This provided an environment for
effective evaluation of various approaches to erosion and deposition risk assessment for
landscape scale applications. Furthermore, modelling soil loss rates in the Ethiopian
Highlands has integrated the use of Remote Sensing Data in a GIS. It led to the
production of thematic maps that emphasized the spatial impact of different soil erosion
factors and showed the expected soil loss in an area in the central Rift Valley of the
Ethiopian Highlands. Assessment of mean soil loss rates was derived by overlaying all
surveyed spatial erosion factors which included; rainfall erosivity, soil erodibility, land
105
cover, and slope factors in an Arc/Info computer program. Remote Sensing and GIS
have also been used to assess the spatial distribution of potential erosion and erosion
factors in the Kenya’s Tana River Basin. To do this, digital coverage of erosivity,
erodibility and slope classes were overlain and classes of erosion susceptibility derived.
The digital process involved in this work was performed using Arc/Info and IDRISI.
The results gave an indication of erosion hazard within the drainage basin.
A GIS also plays an important role in an Environmental Impact Assessment (EIA) of

various resources. US Federal Water Pollution Control Act have used GIS methods for
riparian water quality buffer generation. Using generally available data sets, the riparian
GIS method determines buffer width in terms of local physical conditions and expected
effectiveness in the generation of riparian environmental buffer in the USA. Further, it
has been used in mapping risks of nitrate leaching in soils and soil erosion in Denmark
by employing Arc/Info. It was also used to pinpoint agricultural areas where erosion
could pose a hazard to a nearby aquatic environment. On the East Coast of England,
GIS was employed to assess climate change with associated flood risk. The study used
the Arc/Info package to assess the implications of sea level rise and hence determine
the coastal vulnerability to flooding. As a first step, a GIS based risk assessment model
was developed. Vulnerability was then modelled as being a function of three main
coastal processes; tidal and storm surge activities, sea level rise and isostatic
adjustment.
11.3.2 Hydrology and Water Resources

The dynamic nature of hydrologic processes and the wide scope of many hydrologic
studies result in some of the most demanding requirements of GIS. The requirements
include continuous presentation of terrain, land and stream-based characteristics and
their relationships, land use, and soils. The information must be stored and accessible
in a format that does not limit choice of models nor distort the underlying model
assumptions as well as be flexible to cater for different area sizes. Applications of GIS
to solutions of hydrological problems, therefore require spatial data presentation
capabilities of linking the GIS systems to process based models offering different levels
of coupling. Applications in hydrology are extremely varied. The hydrological
106
processes are presented in distributed models and spatial resolutions of water resources
infrastructure can be used to handle interrelationships and management. The spatial
resolution helps in the GIS framework and the structure of the database that supports
the GIS will influence the applications. With the increasing availability of high
resolution Digital Elevation Models (DEM), the most widespread application of GIS in
hydrology is the identification of drainage pathways and runoff contributing areas
based on topographic form and their coupling with hydrologic models. The key models
within GIS applications include those investigating the impacts of precipitation patterns
and vegetation presentation and topography on water yield; impacts of forest harvesting
on peak stream flow and the estimation of catchment scale soil erosion and sediment
yield. GIS applications go beyond local and catchment scales to regional scales.
For example, a GIS methodology has been applied for monitoring water level on three
East African Rift Valley lakes (Nakuru, Naivasha and Elmentaita) with the primary
objective of creating a GIS database, which could then be used to implement a
hydrological model to accurately predict the lake level changes. To generate the GIS
database, the study divided the lake catchments into a series of land reference units
(LRU’s) with each unit having a unique set of parameters defining their area, slope and
permeability. These layers of grid data were overlaid with rainfall and actual
evapotranspiration within the LRU’s on monthly basis over the study period. This GIS
overlay of multiple layers of grid data gave the outputs of spatial data integration and
analysis as a series of tables which could be interfaced directly with the hydrological
model. The model receives files from the GIS that are then processed to obtain lake
level for each month with a possibility of output table of each LRU ‘s monthly water
flow.
Some further examples of GIS applications include the modeling of slope stability and
landslide activity at both site and regional scales, assessment of non-point source
ground water contamination and assessment of factors controlling coastal sediments. A
GIS attractively provides a framework for integrating data from disparate sources.
Raster data from remotely sensed and fine DEM grids, vector data derived from
contemporary and historical map and air photograph sources and point data from
107
hydrological and ecological monitoring networks are some of the examples that utilize
the power of the GIS to integrate, manipulate and improve visualization. By using
distributed physically based models, GIS can play an important role to support
management and help in decisions and operations in basin agencies.
A GIS for hydrology must further be able to provide all information required to trace
and predict the movement of a drop of water from time it reaches the earth’s surface
along its course into and through a stream system. A GIS can be used to effectively
study the spatial distribution of basin characteristics and their influence on runoff
generation by utilizing certain procedures that integrate the physically based
hydrological models with a GIS. A number of such models exist that can be used with
a GIS an example of which is the TOP hydrologic model whose operation is based on
analysis of topography of a catchment which also accounts for differences in soils and
vegetation characteristics. The GIS that is useful in integrating with this model is the
GRASS. The integration provides better understanding of hill slope runoff production
within a catchment and the influence of soil variability on runoff generation including
geomorphologic characteristics and their impacts on catchment response.
A GIS handles modeling by incorporating pre-processing of data into a form suitable

for analysis in terms of scale, coordinate systems, data structure, etc; direct modeling
tasks such as calibration and predictions; and post processing of data through
reformatting, tabulation, mapping and report generation. One relevant application of
such linkages is: vulnerability or hazard assessment through summed influences.
Generally, a GIS for hydrological purposes need to have capability of dealing with both
raster and vector data. Such a GIS should also be able to store, manipulate and display
geomorphologic data related to basin landscape domain to afford appropriate set of
operational tools oriented to solve hydrological problems. In flood forecasts for
example, topographic information, digital distributed maps of soil type, land use/cover
and soil moisture are required.
Currently, most detailed information available for catchment modeling is terrain

information in the form of distributed elevation data, requiring that the catchments
108
should be modeled using DEM’s. Using GIS techniques, several products can be
calculated from a DEM that are useful for soil erosion modeling and the most important
are slope gradient, aspect or local drainage direction and contributing upstream area.
Some models can be coupled to GIS using conversion programs or user interfaces. To
couple the models and the GIS, the output maps of a GIS are converted to the specific
input format needed by the model. Examples of erosion models that have been linked to
a GIS are the Areal Non-point Source Watershed Environmental Response Simulation
(ANSWERS) model and Limburg Soil Erosion Model (LISEM). The latter model
simulates runoff and erosion in catchments during and immediately after a single
rainfall event. The model useful for scientific planning and conservation purposes as it
incorporates the main hydrological processes of rainfall, interception, surface storage,
infiltration, percolation, overland flow, channel flow, detachment by rainfall, and
detachment by overland flow and transport capacity. The distributed parameter model
is used to model surface runoff and erosion. The catchment to be modeled is assumed to
be composed of square elements with values of key variables defined for each element.
11.3.3 Settlement, Urban Planning and Urban Growth Modeling

A GIS is also applied in social sciences, such as human geography, economics,
planning and sociology. A good example of this is the research on impact of GIS on
local government in Britain. A combination of methods was used to assess the extent
of the take- up of GIS facilities and explore its impacts on administration practices. The
extent of take-up was evaluated by a comprehensive telephone survey of all local
authorities while the impacts of introduction of GIS were explored in the context of
case studies of selected authorities with at least two years of GIS experience. Results
indicated that the level of take-up varied with local authority, function, size and
location.
North West Water Authority in Britain has employed GIS to connect user community,
schedulers, planners, project managers, and customer service personnel. This allows for
operations, development and maintenance. Concurrent users are connected and large
records and database operations used for successful implementation. A large percentage
of the information is tied to geography including location of customers, infrastructure
109
management and operations. Location must be built into planning, design, operation
and maintenance through the use of GIS.
In insurance services, natural hazard management benefits from high-resolution data on

risky areas supplied by GIS applications. The UK Ordinance Survey, for example,
provides accurate digital data for GIS parameterization in the identification of flood
prone zones. Flooding and erosion are major problems in many areas where they cause
damage to property. The prediction of peak flows, source areas, soil erosion and
deposition amounts are necessary for solving the settlement risks problem.
One aspect of urban planning is urban drainage. Although urban drainage models are
well established and adequate, there is a problem with acquisition, manipulation and
storage of data. The combination of remote sensing and GIS technologies allow for a
solution. General purpose GIS are applicable in urban planning and the application
involves generation of maps and inventories, a link with simulation models and
geographic presentation of results. A digital map and its accompanying database is one
of the final products. The GIS allows environmental data from disparate sources to be
geo-referenced and related missing or discontinuous data values to be interpolated and
many different combinations and resolutions of geographic data to be explored.
GIS tools have been applied to integrate and summarize existing data sets and to
transform multiple data layers into new information by modeling farmland dynamics
and land degradation on the Sahelian landscapes using remotely sensed data and socio-
economic data. Dynamics of landscape patterns need to be characterized and quantified
in meaningful ways before ecological impacts of changes of landscapes can be
understood. Landslides can deliver substantial volumes of sediment to river channels
with adverse effects to water quality and fish habitat. In planning land use, it is
important to consider risks posed by landslides and account for its effects. This is
especially true of groundwater recharge areas where timber harvesting is done, as is the
case in the Pacific Northwest.
110
11.3.4 GIS Applications in Urban Modeling
Information about urban areas is required at a variety of spatial and temporal scales.
This is frequently required in connection with land use change especially in respect
with urban growth in terms of height and volume of individual buildings required by
various authorities. Whereas this information can be derived from existing paper maps,
these become soon outdated due to the rapidly changing urban environment. Detailed
ground surveys provide information, but are labour intensive, time consuming and can
only be carried out less frequently. Remote sensing is a third source of information on
urban areas and when integrated with other data, important information for GIS
modelling can be generated. The GIS strength is the power of being able to combine
independent data sets, multidisciplinary and easily resolving problems that could
otherwise be tedious. An interactive stereo digitizing system allows three-dimensional
structure of urban areas to be determined from stereo pairs of digital remotely sensed
images. Importing data generated using this system together with the original image
data into an existing GIS package provides tools for visualizing urban environments
and also permits more meaningful interpretation of existing two dimensional data sets,
by distinguishing between buildings on the basis of differences in height and
architectural style.
Enhancements in the spatial resolving power of satellite sensors offer great possibilities
for inferring land use and land use change in and around urban areas by examining the
spatial mixture of spectrally distinct land cover types in the resultant multi-spectral
images. Incorporating three-dimensional spatial data into existing GIS packages
improves their utility considerably allowing for data query, analysis and scene
visualization. The potential applications include inter-visibility for telecommunications
planning, scene visualization for landscape architecture and urban design studies and
Environmental Impact Assessment. Natural hazards can be modelled using simple
overlay operations with GIS techniques. Identification of the spatial extent of these
hazards present a necessary step in the process of environmental assessment, definition
of control strategies and design of sanitation measures. The outputs of such modelling
are hazard maps that can be used as a basis for land use planning in urban areas. The
local authorities and other developers of urban areas are potential users of the
111
commercial GIS. Urban land capability mapping caters for the wide range of erosion
and instability problems that may be encountered from developments incorporating a
wide range of urban land uses. Urban land capability investigations is based on
information on physical data associated with climatic data, social-economic information
which can be imported into existing GIS packages so that information such as slope,
potential hazard and land-use suitability can be determined and displayed in response to
a point and click query. At micro-scale, GIS operations can collect data over smaller
urban areas especially extracting three dimensional models of buildings, both for
dynamic visualization of urban scenes and to provide three dimensional data sets for
inputs into existing GIS packages. Availability of accurate detailed information on the
height, shape and volume of individual buildings within urban areas allows additional
analysis including telecommunications, urban drainage design and Environmental
Impact Assessment. The data extracted from plan details on elevation from stereoscopic
remotely sensed images are incorporated into a digital mapping GIS package, which
allows for heights of buildings above the underlying terrain to be integrated. Many
systems, however, are unable to cope with two points that differ in terms of height, but
the same X, Y location.
Urban growth (expansion) encroaches areas that may be under agriculture, forestry,
grassland and so on. Expansion into these areas has significant relationships with other
factors such as slope, terrain, soil, climate, drainage surface hydrology and socio-
economic aspects such as population. These factors determine the susceptibility of
these peripheries to invasion by urban growth, and their potential to remain intact. The
strong dependence of urban expansion on all the factors means that it should be
possible to map potential distribution of urban areas.
An understanding of these factors and mapping of their distribution is important in

urban capability assessment. Hannan and Hicks have developed an urban land
capability classification to cater for the wide range of erosion and instability problems
that may be encountered from developments incorporating a wide range of urban land
uses; a classification that could be applied flexibly to different levels of urban
development.
112
The data processing is to interrelate the various variables on a common base to identify
the spatial distribution of various urban capability classes within the boundaries and
around areas of urban expansion. GIS Vector format makes it possible to directly refer
or retrieve geographic conditions and the possible effects on changes due to urban
expansion. GIS operations for all the data sets considered are expected to come up with
a composite output map showing potential hazards that affect urban land use.
Applications of GIS to urban growth modeling have received great attention in recent
years. A number of researchers have developed computer based technologies especially
graphics-based and mathematical models for cities including applications of fractal
geometry and cellular automata to urban structure as evidenced in the works of Michael
Batty. Most of his works are concerned with urban studies and planning involving the
development of computer models and computer graphics in land use and transport
planning, the spatial analysis of urban form, geographic information systems
technology, impact of information technology on cities, and formal methods of decision
making in policy analysis. Benefit from his works is published works in urban
modeling in which algorithms, calibrations, predictions and microcomputer graphics on
design and creative modeling factual cities by way of geometry of form and function
and other methods of studies and planning and design to the development of models is
quite profound. Models of urban shape and form can be achieved using fractal
geometry, information infrastructure and its impact on cities, urban information systems
in developing countries, and the development of spatial analysis and models within GIS
are some of the recent works that have potential applications. Animation of urban
growth in different regions may be achieved using a variety of data sources accessible
in different media.
Most of the objectives for urban growth research include the study of changing
landscapes or metropolitan areas. A new housing sub-division or a new shopping
center are examples of a city undergoing change. Change is measured when the
geographic extent of urban growth is compared. Urban growth models can be applied
by community in economic development decisions and communications using
appropriate GIS technologies and policy and community service networks as has been
113
attempted by the Urban Growth and Land use (UGLE) modeling research that presents
scenes over different year, 1935, 1951, 1973, 1991 and the possibility of using the
trends to give a scene for 2050. Here the changes have been measured by comparing the
geographic extent of urbanization through time. Human induced land transformations
may also be characterized by temporarily analyzing the conversion of land cover to
land use. The evolution of an urban area can be anticipated when modeling and
predicting future urban growth patterns to understand the impact that the change has on
a region’s natural resources, economy and people. In order to investigate how humans
are altering the landscape, the United States Geological Survey (USGS) are currently
studying the anthropogenic impacts on the Western environment under urban dynamics
and Global Change research programme and results on land use trends, land
transformations, urbanization etc would be presented on quarterly reports. These will
reveal foreseeable scenarios for urban growth for planning purposes based on initial
growth predictions. Growth projections for entire metropolitan areas can then be
calibrated with possible modifications made to allow for expenditure commensurate
with urban growth factors that are input to the models.
The modeling of interactions between urban activities and infrastructure renewal and
other economic activities are based on historic data. Apart from transport modeling,
modeling of ozone in some urban areas such as Chicago are also critical. The urban
land use allocation model (ULAM) makes use of modeling software and related
planning and engineering activities for traffic modeling with urban land use concerns.
Such Urban planning models have been used in Florida for projecting urban
development patterns. They integrate land use probabilistic modeling of variability and
uncertainties in urban air toxic emissions using pollution modeling on the urban,
regional and global scales.
11.4 Constraints and Opportunities for GIS Applications
11.4.1 Constraints
A GIS has many advantages in its use for support of public policy decisions regarding
natural resource use such as land, water, forests, fish and wildlife. A number of
limitations however exist in its applications. These include: limited user awareness,
114
poor co-ordination among users, data quality and propagation of error during analysis,
costs and poor spatial coverage data, limited capacity and generalizations for
heterogeneous areas due to a scale problem resulting from lack of all the information.
Most GIS systems portray sophisticated data processing with limited analytical
capabilities. In particular, they fail to take cognition of the imprecision of many
categories of spatially referenced data handled so mechanically. Most spatial
information systems assume real world phenomena to be modeled by discrete entities
such as points, lines, polygons, etc which are deemed to have exact attributes.
For most processes, remote sensing is used along with data from other sources such as
digitized maps of soils, geology, elevation etc. The technique of merging information
from different sources into one system has to be carried out with care using same
coordinate system. Usually, a GIS does not offer information on equal size pixel, thus
merging of different information gives rise to severe errors. One practical example is in
hydrological modelling where water stored in the soil is relevant and this depends on
soil type and land use. Data from land use may be obtained from land use classification
derived from LANDSAT imagery while soil types may be derived from digitized maps.
Maximum soil storage from each pixel is then determined on the basis of merged
remote sensed data and map information. Layers of diverse origins and with diverse
degrees of accuracy cause problems due to lack of consideration of the reliability of
each source in addition to possibility of unreliable rules for combining the information.
Outputs from such GIS applications depict no indication of how reliable the
classifications are and inference about the real world situation must be treated with
caution since exactly correct GIS output is not possible. There is a certain amount of
generalization involved. Generalizations for heterogeneous areas result from scale
problem for lack of all information. Such generalization effects are compounded when
one looks at more than one feature simultaneously. There are errors associated with
generalization in both raster and vector systems. These are: area and point estimation
and a creation of spurious polygons, geo-coding errors and altered quality by
115
decreasing location and attribute accuracy which affect completeness and consistency
within the data.
In research, generalization of spatial data is considered deliberately induced, such that

if scale of a map is decreased, there is less physical space to represent the geographic
features of a region and as the process continues, such features become exaggerated
in a smaller scale. As the features ‘fight’ for representation in a reduced map space,
some features will be eliminated and those remaining may be further simplified,
smoothed, displaced, aggregated or enhanced. In some of the extreme cases the map
losses its geometric properties. The main reasons for generalization in a GIS are:
i) To create a level of detail appropriate for the scale of display or analysis,

ii) To permit analysis of data at multiple levels of resolution and
iii) To minimize the data storage and input/output operations.
Generalization facilities available within a GIS allows the selection of generalization

algorithms (instructions) that minimizes selected generalization effects and therefore
optimize the process from the point of view of the user making him/her aware of what
is happening with the data. GIS tools such as overlays used along with models depict
significant incompatibilities preventing true integration since most GIS manage static
and discrete data, while environmental models deal with dynamic and continuous
phenomena. In natural world entities, there are no real sharp class boundaries. For
example, uses of logical operations of overlays in the exact measurement of attributes
of continuous classes propagate error in the data generation in soil characteristics,
climatic and vegetation types, among others.
The problem of data quality related to resolution database useful for a range of many
applications such as flood insurance is lacking. The challenge is for technology, to be
able to present distributed data in the right resolution for respective purposes.
Creating a viable GIS application from routinely available spatial data can result in
data error and user misconception.
116
11.5 Opportunities of GIS
There is a tendency towards a GIS software being integrated to mathematical models
and coupling of GIS with efficient databank. There will be a shift from specific
application models to spatial decision support and expert systems to ensure top
management have access to relevant information on a day-to-day basis. Identification
and development of critical management indicators is therefore necessary.
Expert systems need to be developed which will help the user in the decisions on
which software to be used and for what purpose. With the development of new space
systems, we will have at our disposal new types of remotely sensed data that will
require more complex GIS hardware and software. With the continued threat of
environmental decline, more research on computer simulation models describing
processes in each environmental compartment; air, water and soil will be required
precisely. Databases, models, and expert systems can be combined to achieve
practical solutions for environmental problems. This could be achieved through
delegating sub-tasks to simple models or procedures and allowing nested expert
systems (expert systems that contain rules that call for other expert systems).
GIS’s have improved in recent years due to direct research activities, user
involvement and vendor cooperation. This initiative increases user awareness of
spatial thinking and also calls for a multi-media approach. Multi-media is necessary
because of unavailability of critical materials in convenient formats importable into
software packages for data to be assimilated into a GIS to produce maps with sites
having features of interest.
Aerial photographs, satellite imagery and technical reports are examples of multi-
media documents for which systems providing structured, exploratory and
explanatory navigation will be required to assist planning and management through
‘best links’. With such systems, users could have their questions answered
interactively. A GIS could be used to integrate the range of relevant data sets for
producing supporting maps and statistics with no more restrictions to operations
offered by database management systems. GIS will be expected to be more open to
117
users’ demands of better interaction and versatile customization. Integrated
information and decision support systems brings together large volumes of
background data and interactively distill from it decision- relevant information. A
GIS is one of the key technologies, others being workstations, networking, databases,
interactive graphics, modeling, optimization and expert systems. The development of
algorithms for automated acquisition of remotely sensed information holds promising
perspectives for wider applications in future.
Capability of handling time-variant data within a single system which allows spatial
data to vary with time and user interactive flexibility need to be incorporated to model
and GIS in future. Fast on-line data management, inserting, updating, retrieval in real
time, while allowing for visual effects of choice of model structure has potential for
the future. Generally, an urban capability map can be developed from an assessment
of an interaction between landform, soils, land-cover, and hydrological features.
Various classes based on physical limitations to urban land-use will then be defined
using raster format, defined by row and column position of cells occupied.
11.6 Summary
A GIS application involves interdisciplinary studies in social-economic and
environmental sciences and its strength lies in its capability to combine independent
datasets.
The GIS technique deals with themes that integrate geographical and non-
geographical information.
GIS handles large amounts of data in space, and many institutions can use it to
plan, zone, assess property and land records, public safety, management of forested,
agricultural and coastal lands, among others.
Since all these diverse applications are carried out using similar software and
techniques, a GIS is a general purpose tool that may be used to integrate data from
various sources to determine resource availability, plan ecological aspects, transport,
urban and landscape development, among other uses.
Although a GIS can have many opportunities in its use, it has also a number of
limitations in its applications that include limited user awareness, poor co-ordination
118
among users, data quality and propagation of error during analysis, costs and poor
spatial coverage data, limited capacity and generalizations for heterogeneous areas
due to a scale problem resulting from lack of all the information.

(a) Describe any organization that may be using GIS, and how the
organization is benefiting from the technology.
(b) What are the opportunities and constraints of a GIS?
11.8 Further Reading

Antenucci J. C., Brown K., Croswell, P.L and Kevany, M.J. (1991): Geographic
Information Systems: A Guide to Technology; New York: Chapman Hall,
301 pp.
Blackmore, M. (1983): Generalization and error in Spatial Databases. Proc. Auto-
Cart, 6 pp 313-322.
Bracken, I. and Webster, C. (1990): Information Technology in Geography and
Planning Including Principles of GIS. London: Routledge.
Burrough P.A (1986): Principles of Geographic Information Systems for Land
Resource assessment, Oxford: Clarendon Press.
Clark, M. J. (1998): “Putting Water in its Place: Perspectives on GIS in Hydrology
and Water Management: Hydrological Processes 12, 823-834.
De Roo A. P.J Hazelhoff, I and Burrogh P. A. (1989): Soil Erosion Modeling Using
ANSWERS and Geographic Information System; Earth Surface Processes
Landif, 14, 517-532.
Goodchild, M. F. (1980): The Effects of Generalization in Geographic Data
Encoding. In: Freeman, H. and Pieroni, G. (eds) Map data processing, New
York: Academic press, Pp 191-205.
Goodchild, M.F, Parks, B. and Staeyart, L. (1993): Environmental Modeling with
GIS; New York: Oxford University Press.
Heuvelink, G. B. and Burrough, P. (1993): Error Propagation in Cartographic
Modelling Using Boolean Logic and Continuous Classification. In:
International Journal of Geographic Information Systems. Vol. 7 No. 3
Taylor and Francis Ltd. London. Pp231-247.
Maidement, D. R (1993): GIS and Hydrologic Modeling. In Environmental
modelling with GIS. (Eds. E.M. Goodchild, B. Parks & Staeyart, L.; Oxford:
Oxford University Press.
Marble, D.F. and Amundson S.E (1988): Microcomputer Based Geographic
Information System and theirRole in Urban and Regional Planning.In
Environment and Planning, 15:305-324.
Masser, I and Campebell, H. (1996): The Impact of GIS on Local Government in
Great Britain. In: Geographic Information Handling- Research and
Applications. Ed. Paul M Mather. London: John Wiley and Sons Ltd.
119
Muller, J. C. (1991): Generalizations of Spatial databases. In: Maguire, D.J,
Goodchild, M. F. and Klind, D. W. (eds), Geographic Information Systems:
Principles and Applications. Harlow: Longman, pp457-475.
Reusing, M, Schneider, T and Ammer, U. (2000): Modeling Soil Loss Rates in the
Ethiopian Highlands by Integration of High Resolution mom-02/d2 Stereo-
Data in a GIS, In: International Journal of Remote Sensing Vol 21 No.9.
Taylor and Francis Ltd. London. Pp 1885-1896.
Thurner, T, Jones, A. B. and Brown D. (2000): A GIS based Coastal Management
System for Climate Change associated flood risk assessment on the East
Coast of England. In: International Journal of Geographic Information
Science, Ed. Fischer P. Miller, H. and Lees. Vol. 14 No. 3. Taylor and Francis
Ltd. Pp 265-282.
Viex, B. E. (1991): Geographical Information Systems and Non-Point Source Water
Quality and Quantity Modeling; Hydrological Processes, 5 101, 101-113.
Xiang, W. (1993): A GIS Method for Riparian Water Quality Buffer Generation. In:
International Journal of Geographic Information Systems VOL. 7 No.1
Ed. J.T. Coppck and K. C. Clarke. Taylor and Francis Publishers, London.
Pp57-70.
120

AGE 400 Remote Sensing and Resource Management

Uploaded by

Copyright:

Available Formats

AGE 400 Remote Sensing and Resource Management

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AGE 400 Remote Sensing and Resource Management

Uploaded by

Copyright:

Available Formats

KENYATTA UNIVERSITY- INSTITUTE OF OPEN LEARNING

REMOTE SENSING AND RESOURCE MANAGEMENT MODULE

CHRISTOPHER MISATI ONDIEKI

1.0 LESSON ONE: INTRODUCTION TO REMOTE SENSING 03

2.0 LESSON TWO: SIGNIFICANCE AND HISTORY OF REMOTE

3.0 LESSON THREE: THE ELECTROMAGNETIC SPECTRUM 25

4.0 LESSON FOUR: INTERPRETATION AND SOURCES OF ERROR

5.0 APPLICATION OF REMOTE SENSING IN AGRICULTURE 44

6.0 LESSON SIX: APPLICATION OF REMOTE SENSING

7.0 LESSON SEVEN: SATELLITE WEATHER AND CLIMATE

8.0 LESSON EIGHT: INTRODUCTION TO GEOGRAPHIC

9.0 LESSON NINE: GIS COMPONENTS AND FUNCTIONALITY 82

10.0 LESSON TEN: SPATIAL DATA MODELS 93

11.0 LESSON ELEVEN: APPLICATIONS OF GEOGRAPHICAL

1.2 Learning Objectives

1.3 FUNDAMENTAL CONCEPTS

(a) Acoustics: A branch of science that deals with sound-reflecting character at an

In environmental sciences, remote sensing is understood as the acquisition, recording

1.3.2 Types of Remote Sensing

(a) Active Remote Sensing

(b) Passive Remote Sensing

(a) Spatial Resolution

(b) Spectral Resolution

(c) Radiometric Resolution

(d) Temporal Resolution

(b) Multi-date is the acquisition of data at different moments in time, and is

(d) Multi-polarization refers to the various ways that comprise a beam or an

1.4 Advantages and Disadvantages of Cameras and Satellites

They are relatively simple in operation.

The disadvantages are:

The information from satellites is in electronic form; therefore, relatively easy to

Other advantages include:

1.6 Review Questions

2. Distinguish between the following:

3. What are the advantages and disadvantages of photographic cameras and

1.7 Further Reading

2.2 Learning Objectives

2.3 Importance of Remote Sensing

(a) Improved Vantage Point

(b) Increased Spatial Resolution and Geometric Fidelity

(c) Broadened Spectral Sensitivity

2.4 History of Remote Sensing

Up to mid-1960's, aerial photography using photographic film to record visible light

The technological development of automatic and unmanned satellites (LANDSAT,

LANDSAT 4 and 5 launched in 1982 and 1984 respectively incorporated important

SEASAT was launched in 1978 on experimental basis to obtain information on the

The Orbview 1 developed by OrbImage, launched in mid-1990s was for scientific

Table 2.1 Multi-spectral High Resolution Sensor Systems

Table 2.2 Low Resolution Sensor Systems

2.6 Review Questions

2.7 Further Reading

3.2 Learning Objectives

3.3 The Electromagnetic Spectrum

The shorter the wavelength the greater the energy,

Figure 3.0 The Electromagnetic Spectrum

The color blue ranges from 0.4 to 0.5µm.

3.3.3 Energy/Matter Interaction

(d) Scattered, the energy undergoes random change of direction.

3.3.4 Sensors and The Spectrum