Land Use and Land Cover Changes in Arid Region: The Case New Urbanized Zone, Northeast Cairo, Egypt

doi:10.4236/jgis.2011.33015

Paper Menu >>

Journal Menu >>

Journal of Geographic Information System, 2011, 3, 173-194

doi:10.4236/jgis.2011.33015 Published Online July 2011 (http://www.SciRP.org/journal/jgis)

Land Use and Land Cover Changes in Arid Region: The

Case New Urbanized Zone, Northeast Cairo, Egypt

Rafat Zaki1, Abotalib Z aki 2, Saad Ahmed3

1Department of Ge o logy, Faculty of Scie nce, Minia University, Minia, Egypt

2Geological A ppl i c at ion s an d Mi neral Reso u rces Divisi o n , National Authority for Remote Sensing and

Space Scienc e , Cairo, Egypt

3Department of Ge o logy, Faculty of Scie nce, Al-Azhar University, Cairo, Egypt

E-mail: [email protected]

Received January 21, 2011; revised March 15, 2011; accepted April 10, 2011

Abstract

The spatial characteristics of land cover are useful for understanding the various impacts of human activity

on the overall ecological conditions of the urban environment. The multi-temporal Landsat images (TM)

between the years of 1990 and 2003 were used together with the Geographic Information System (GIS)

techniques to evaluate the environmental changes in the area around Gabal El Hamza and the surrounding

urban expansion in the new urban cities at the northeast side of the Greater Cairo by using the post classifi-

cation change detection technique and field investigation. Five major units were determined including: urban,

cultivated land, Holocene sand dunes, Oligocene basalt and Miocene–Pleistocene sediments. The cultivated

cover changed from 89.6 to 150.4 km2 for the years of 1990 and 2003 respectively. The urban area increased

from 49.5 to 120.9 km2 with a great value of change reached 71.3 km2. The basaltic exposures changed from

3 to 3.75 km2. The sandy cover decreased from 68.9 to 60.1 km2 and the exposures of the rock units changed

from 904.8 to 780.8 km2 with removing 124 km2 in 13 years. The total accuracy of the Landsat-derived land

cover data was 95 and 92% for the years 1990 and 2003 respectively. Landsat TM thermal infrared data in-

dicated that the surface temperature was strongly affected by the land cover changes.

Keywords: Land Cover Changes, Accuracy Assessment, TM Images, Land Surface Temperature, Egypt

1. Introduction

In 1984, the first spark of the urban invasion to the desert

in Egypt was flared-up with the establishment of the

Tenth of Ramadan as a new urban center. Since this time,

new settlements were erected in different parts of the

country from the extreme north (as New Borg El-Arab

near Alexandria) to the south (as El-Minia El-Gedida).

The zone surrounding the Greater Cairo area had the

highest share in urban encroachment in the Egyptian

deserts. This is due to the high population density in

Cairo and is particularly relevant as a result of the ex-

pectation of an increase of population to almost the dou-

ble of its present status by 2050 [1]. Cairo also provides

the highest chances of increase of job opportu- nities in

Egypt as a result of the installation of new factories and

farms around the city, but away from its center. This

necessitates the establishment of dwelling places for the

qualified workin g forces which are generally to be found

near and within the city. The above mentioned reasons,

together with the suitable topographic and geologic set-

ting of the area, make the surroundings of Cairo suitable

for urban e xp ansion.

The northeastern part of Cairo, in which several new

urban settlements were erected, such as the Tenth of

Ramadan, Badr, El-Obour, El-Shorouq and other cities

has suffered from noticeable changes, especially in the

last few decades. These changes were essentially caused

by man-made interference. Therefore, the present study

is particularly important as it will throw light on the im-

pact of human activities on the overall ecological condi-

tions of the urban environment [2]. Land cover change

due to human activities is currently pro- ceeding more

quickly in developing countries than in the developed

world, and this situation will be projected into the year

2020, most of the world’s mega cities will appear in de-

veloping countries [3]. Increasing population in devel-

oping cities has caused rapid changes in land cover and

R. ZAKI ET AL.

174

increased environmental degradation [4]. So Geographic

Information System (GIS) and remote sensing (RS) tech-

niques are powerful and cost-effective tools for assessing

the spatial and temporal dynamics and changes in land

cover and land use [5-7]. Remote sensing data provide

valuable multi- te mpor al data o n the p rocesses and patterns

of land cover and land use change, and GIS is a useful

technique for mapping and anal yzing these patterns [8].

In this work, two Landsat Thematic Mapper (TM)

images (1990 and 2003) were used to detect changes in

the study area. The topographic maps prepared by the

Egyptian Military Surv ey at scale 1: 50,000 with its 2007

update are also used as well as field investigations. Nu-

merous change detection methods have been developed

to assess variations in the land cover using satellite data

and post-classification comparison technique. The accu-

racy of this technique to detect the dynamic changes de-

pends mainly on the accuracy of the individual classifi-

cation of each land cover unit.

Kim, Nichol, Kevin and Timothy, Chen et al., Weng et

al., Wang, Weng and Lu and others ([9-15]), mentioned

the relation between surface temperature and land cover

changes using Landsat TM or ETM thermal infrared

data with a spatial resolution of 120 or 60 meters re-

spectively. Land surface temperature (LST) appears to

be strongly related to the changes of the land cover pat-

terns especially changes concomitant with the urban

expansion. Urbanization is one of the most important

factors affecting the global warming, so studying the

relation between urban surface changes and the land

surface temperature is critical. Ground-based tempera-

ture observations reflect only thermal condition of local

area around the stations. Hence studying thermal distri-

bution in big areas requires a large number of ground

stations and long time of analysis and correlation.

Thermal remote sensing is able to assess the instantly

and accurately temperature distribution over vast areas

economizing time and efforts.

The present study deals with changes of the land

cover between 1990 and 2003 of an area lying in the

northeastern vicinity of Cairo (Gabal El-Hamza area)

and the consequent changes in surface temperature.

Changes include urban and agricultural (cultivated) sur-

faces together with Holocene sand dunes, exposures of

Oligocene basalt and Miocene-Pleistocene sediments. It

is thought that the study can help in understanding the

dynamics of these changes in order to predict and/or

plan for future development.

2. Geologic Setting

The study area lies between 30˚05′N - 30˚24′N and

31˚28′N - 31˚46′E. This study examines in detail 36 km

long and 31 km wide (1116 km2). The area is located at

about 25 km from the Greater Cairo along the Cairo-

Ismailia district and is bounded by the Greater Cairo

from the west, the Eastern Delta from the north, the Is-

mailia Governorate from the east and north Eastern De-

sert from the south (Figure 1).

The study area has a low relief with ground elevation

ranging from 0 m in the northwestern part to 240 m

above sea level in the central and southern parts. The

area encounters many new urban cities as Tenth of

Ramadan, El-Obour, El-Shorouq, Badr, El-Nahda and

El-Hikestep (Figure 2). These cities are surrounded by

series of ridges and scarps related to the Oligocene basalt

with gravelly sands; Miocene gravels, sands, clays, sand-

stones, dolomite with limestone; sandy gravels with

limestone of Pliocene age and Holocene sand dunes

(Figure 3).

3. Methodology

The methodology used in this work is summarized in the

flow chart of Figure 4. The used images were extracted

from two Landsat Thematic Mapper (TM) scenes taken

over the eastern portion of the Nile Delta up to the Suez

Canal in August 1990 and June 2003 (Figures 5(a) and

(b)). A subsene covering the study area was digitally cut

for analysis. Ground resolution for these images is 30 m.

Landsat TM image records data in seven different band-

widths. The latter are broken down into portions of the

visible, reflected infrared and thermal infrared regions of

the electromagnetic spectrum. From these various band-

widths, a great deal of information about the land cover

can be displayed and analyzed. These images were geo-

metrically corrected using topographic maps prepared by

the Egyptian Military Survey at scale 1:50,000 as a ref-

erence data. Atmospheric corrections were applied to

remove the effects of the passage of radiation through the

atmosphere using dark pixel subtraction technique, fol-

lowed by haze and noise reduction in order to diminish

the weather effect. Maximum likelihood supervised and

unsupervised classifications were used in the production

of land cover maps for the study area. The post- classifi-

cation change detection technique was applied and the

temporal change maps were produced using a GIS model.

The ground truth points used for assessing the accuracy

of the classifications were selected using high resolution

SPOT images and field investigation with a Global Posi-

tioning System (GPS) unit. Among the software; Arc

GIS Version 9.2, Erdas Imagine Version 9.1, Envi Ver-

sion 4.5 and Microsoft Excel programs were used.

According to Chen, et al. (2002), the land surface

temperature derived from the Landsat TM thermal infra-

red (band 6) image in two steps:

R. ZAKI ET AL.

175

Figure 1. Location map of the study area.

Figure 2. Base map of the study area showing the distribution of the new urban cities and the main roads.

R. ZAKI ET AL.

176

Figure 3. Geologic map of the study area (modified after Tawfik and Swedan, [16]).

Figure 4. Flow chart showing the procedures of the work.

R. ZAKI ET AL.177

(a)

(b)

Figure 5. (a) and (b) Landsat TM images of the study area in 1990 and 2003.

R. ZAKI ET AL.

178

5. Unsupervised Classification

The first, the digital numbers (DNs) of band 6 was

converted to radiation luminance , mWcm−2sr−1)

using the following equation: 6TM

In the unsupervised classification, the computer separates

the pixels into classes with no direction s from the analyst

[17]. This means that unsupervised classification tech-

niques do not require the user to specify any information

about the features contained in the images. The un-

supervised classification technique (isodata) was applied

for the study area using ENVI Version 4.5. Using initial

cluster options, which were calculated from the file

statistics of 6 classes; with (6) Maximum iteration and

(0.950) Converg ence Thresh old as processing option s. In

the study area the unsupervised classification applied to

the six visible and reflected IR bands of the subscene

resulted in 20 classes.



6255 maxminmin

RV RRR

Where: V represents the DN of ban d 6,

max1.896(mW cmsr)R



and

min0.1534(mW cmsr)R



Secondly, the radiation luminance is converted to abso-

lute temperature in Kelvin , by the following

equation:



At this point, the image is difficult to interpret.

Decisions need to be made concerning which land cover

types that each category falls within. To make these

decisions, other materials and knowledge of the area

were useful. Checking in the field, the various areas

discriminated in the digital image should be examined in

the field to obtain more accurate results. If this

knowledge is not available, scientific reasoning may be

used to group the various categories together in to related

land cover units. In the case of the present study the

second probability was more suitable. So, the twenty

classes obtained from the two sets of images were

merged into five major classes. These classes are

referring to the five land use/cover units including

sedimentary rocks outcrops, urban areas and agricultural

land, Holocene sand dun e s and basaltic exposures.







1ln 21

TKK R b

where, 1 1260.56

K

60.766 and

(mWcm−2sr−1µm−1), b represents the ef-

fective spectral range = 1.239(μm).

2K

Finally, a third step was applied to convert tempera-

ture into Celsius scale using the equation:

273FK

Among the software; Arc GIS version 9.2, Erdas

Imagine version 9.1, Envi version 4.5 and Microsoft Ex-

cel programs were used.

4. Supervised Classification for Obtaining

Land Cover Map However, s everal class es were incorr ectly in terpreted .

Some urban settlements have been misclassified as rock

units or basaltic flows due to the fact that they dis-

pl ayed similar spectral characteristics. Post-classification

change refinements were used to improve the accuracy

of the classification as it is simple and effective method

[18], because urban surfaces are heterogeneous and

composed of a complex combination of features as

buildings, roads, grass, trees, soil and water [19]. After

the performance of the po st classification refinement, the

misclassification has been mostly corrected (Figures 8(a)

and (b)). The post-classification transformation of the

classified raster into shape file vector has been done

using ENVI version 4.5 and using ARC GIS tools. The

obtained shape file was converted into a geo-database in

order to introduce the change detection analysis. Before

this step, it is necessary to make sure that the classi-

fication used for change detection procedu re is matching

fact in the field.

Supervised classification has been developed for satellite

image-processing where it has been applied to the

classification of spectral layers. It depends mainly on th e

experience and accuracy of the user in detecting the

signature differences between various units in the

satellite image using his naked eyes. Each difference in

the signature or pixel value in the processed image is

revealed as differences in the unit tone in the satellite

images. (Figures 6(a) and (b)) show selected training

points on the 1990 and th e 2003. Landsat- TM i mage and

the corresponding spectral profile showing the relations

between the pixel value and the image bands for each

different classes.

Using Erdas Imagine version 9.1, new maximum

likelihood supervised classification analyses were carried

out on the TM (1990) and TM (2003) images (Figures

7(a) and (b)) to identify the land use/ cover changes in

the study area. Five different land cover classes were

identified including: agriculture, sand dunes, urban, basalt

flows and sedimentary rocks units of Miocene to Pleis-

tocene age.

6. Accuracy Assessment

The classified thematic maps are produced for a wide

R. ZAKI ET AL.179

(a)

(b)

Figure 6. (a) and (b) The different classes can be selected by eye in 1990 and 2003 Landsat TM images and the corresponding

spectral profile of these classes.

R. ZAKI ET AL.

180

(a)

(b)

Figure 7. The different land cover units in the 1990 and 2003 Landsat TM images using maximum likehood supervised classi-

fication.

variety of resources; soil types or properties, land cover,

land use, forest inventory, and others. These maps are

not very useful without quantitative statements about

their accuracy. Lillesand and Kiefer [20.] remark “a clas-

sification is not complete until its accuracy is assessed’.

Congalton [21] defined the accuracy asse- ssment as the

accordant between the remote sensing data and the ref-

erence information. The accuracy of digital land cover

classifications can be expressed quantitatively by build-

ing and interpreting a classification error matrix. An er-

ror matrix compares information from a classified image

or land cover map to known reference (truth) sites for a

number of sample points. The accuracy assessment of the

land cover maps extracted from Landsat data include the

generation of 100 random references (truth points) for

each land cover map. Accuracy assessment of the land

cover maps after the post-classification refinements and

merging of the 20 classes into 5 classes covering the

major land cover units was then performed using field

data, high resolution SPOT images, topographic maps

and the results were recorded in a confusion matrix.

By applying the accuracy assessment process on the

R. ZAKI ET AL.181

(a)

(b)

Figure 8. (a) and (b) The unsupervised classifications of the years 1990 and 2003 Landsat TM images after the enhancements

and post-classification refinements.

R. ZAKI ET AL.

182

land cover maps before and after the refinement processes.

It is found to be that the rule-based post-classification

refinements were effective and improved accuracy by

8% - 13%. The total accuracy of the Landsat-derived

land cover data was 95 and 92% for the years 1990 and

2003 respectively.

According to Anderson et al. [22], the standard ac-

curacy for (land use/land cover) mapping studies ranging

from 85% - 90%. From the accuracy reports, it is easily

noticeable that the accuracy of the land cover map de-

rived from the 1990 Landsat TM image is higher than the

accuracy of the land cover map derived from the 2003

Landsat -TM image due to the vast urban expansion at the

2003 image which result in the increasing of the spectral

confusion between pixels contain the urban signature and

pixels contain the argillaceous or basaltic signature.

7. Change Detection Techniques

Many remote sensing change detection techniques have

been developed, and the advantages and disadvantages

of each have been reviewed by a number of authors

([23]). However, the new digital change detection tech-

niques are continuing to be developed primarily

in-response to the range of social and environmental

chal- lenges posed by human transformation of the

Earth’s surface ([24,25]) and the potential of remote

sensing in monitoring related processes ([26-28]). All

change detection techniques rely on the basic idea that

changes in the spectral and/or textural characteristics of

geometrically, atmospherically and topographically cor-

rected remote ly sensed imagery r epresen t change s of the

Earth’s surface.

In this study, the post-classifi cation comparison change

detection technique is the most preferable. The mag-

nitude and location of change (from to change detection

matrix, [29]) were determined for the entire studied years

(1990-2003). The major land cover changes were colour

coded. These changes were from rocky land to urban,

from rocky land to agricultural areas, from sand to agri-

cultural areas and from sand to urban areas. Also the

exposed basaltic extrusions have been found to cover

different spaces in the two dates. Table 1 shows the ob-

served major land cover changes and the area of each

land cover class has been given in m2 with the yearly

average change in each type (each class). Using Arc

Toolbox in ARC GIS Version 9.2 a new model has been

developed to map the change through the thirteen years

between the 1990 and 2003 images.

The data obtained are also shown in a chart form to be

easier to understand for the normal observer and the de-

cision maker, so a diagrammatic view displaying a com-

parison between the same land cover classes in both

dates (Figure 9), and the average ratio of each class that

was occupying the surface area between 1990 and 2003

(Figure 10) were d es igned.

8. Interpretation of the Detection of Change

Analysis

The detection of change analysis is concerned with the

environmental changes and the human impact. These

changes have been detected and identified as the fol-

lowing:

a) Changes concerning agricultural (cultivated)

lands

The surface of cultivated lands shows big changes

through the urban development of the new cities (as

Tenth of Ramadan, El-Shorouq, Badr and El-Obour cit-

ies), or through new cultivation around El-Obour and

El-Nahda cities and on the edges of the Eastern Delta at

Belbis city. Figures 11(a)-(c) show the steps of changes

of the green cover at various dates. The agricultural

cover increased from 89.6 to150.4 km² between the years

1990 and 2003. This means that the agricultural cover

increased during th e last 13 years by 60.7 km² (Table 1).

b) The urban development

Urban development in the study area plays a very im-

portant role through the environmental changes either

through the en largement of the Tenth of Ramadan city or

through construction of the new urban cities such as

El-Shorouq, El-Obour and Badr between the years of

1990 and 2003 (Figures 12(a)-(c)). The large deve-

lopment of the new urban cities, where the urban area in

1990 was 49.5 km2 to reach 120.9 km2 in 2003 with a

large value of chan g e reached 71.3 km2 (Table 1).

c) The bed rock exposures

Bedrock exposures in the year of 1990 were covering

the bulk of the area of study, amounting to about 904.8

km². But this area was largely covered by new elements

related to the urban development and cultivation of new

surfaces. It reached 780.8 km2 in 2003 (Table 1). This

amounts to a decrease in the exposed bed rock surface of

124 km² in 13 years. The remarkable dynamic change is

shown in Figures 13(a)-(c).

Table 1. The land cover changes in m2 in 1990 and 2003 and

the average of change der year.

Classes 1990 Area

(m2) 2003 Area

(m2) Total

change (m2) Difference

per year

Agriculture89696700 150453900 60757200 4673630.769

Urban 49591800 120913200 71321400 5486261.538

Rock 904880700780877800 –124002900 –9538684.615

Sand 68918400 60105600 –8812800 –677907.6923

Basalt 3020400 3757500 737100 56700

Sum. 11161080001116108000

R. ZAKI ET AL.183

Figure 9. Each class area in m² at the time of analysis.

Figsure 10. The proportions of each class to the whole studied area at 1990 and 2003.

R. ZAKI ET AL.

184

(a)

(b)

R. ZAKI ET AL.185

(c)

Figure 11. (a), (b) The distribution of the agricultural covers in the years of 1990 and 2003, (c) The changed area between the

two dates.

(a)

R. ZAKI ET AL.

186

(b)

(c)

Figure 12. (a) and (b) The distribution of the Urban covers in the years of 1990 and 2003 respectively, (c) The changed area

between the two date s.

R. ZAKI ET AL.187

(a)

(b)

R. ZAKI ET AL.

188

(c)

Figure 13. (a)(b) The distribution of the sediments surfaces in the years of 1990 and 2003; (c) The changed area between the

two dates.

d) The sand dunes

Urban and agriculture developments to gether with san d

movement by the wind caused changes in the surface of

the sandy cover from 68.9 to 60.1 km2 between the years

of 1990 and 2003 respectively (Table 1). This change is

shown in (Figu res 14( a)- (c)).

e) The basaltic exposures

The surface of basaltic exposures increased in the con-

sidered period (1990-2003), because in spite of the loss

of a part of these exposu res throug h qu arrying, ad d itional

areas have been exposed by the removal of the Miocene

sediments and Quaternary sands covering the basaltic

flows in order to prepare new areas for quarrying. The

exposures of basalt increased from 3 to 3.75 km2 be-

tween the years 1990 and 2003 (Figures 15( a) -(c)).

9. Land Surface Temperature

Land surface temperature maps (LST) were derived from

the TM images that have been acquired nearly in the

same time of year (4 August 1990 and 8 August 2003) in

order to neglect the seasonal mutations of temperature

and nearly at the same time of day (8.00 and 7.45 AM) to

avoid the effect of the diurnal temperature cycle. In

1990, the land surface temperatures range between 24.05

and 39.45˚C (Figure 16) and from 21.8 and 42.11˚C at

year 2003 (Figure 17).

Superimposing the land surface temperature maps

over the land cover maps indicate that the spatial distri-

bution of LST depends mainly on the land co ver type. In

1990, the rock types were the most common factor af-

fecting the LST distribution. The basaltic exposures and

the surrounding highly ferruginous sandstones are indi-

cated by the pink colour referring to the highest tem-

peratures (36.6˚C - 39.45˚C). The most abundant yel-

lowish colour, which represents the set of temperatures

from 33.6 to 36.5˚C, refers to the non-clastic formations

and sand dunes. The greenish, light green and bright

green colors referring to cultivated lands of different

densities. The urban settlements in 1990 affected the

R. ZAKI ET AL.189

(a)

(b)

R. ZAKI ET AL.

190

(c)

Figure 14. (a)(b) The distribution of the sand cover in the years of 1990 and 2003; (c) The changed area between the two

dates.

(a)

R. ZAKI ET AL.191

(b)

(c)

Figure 15. (a)(b) The distribution of the basaltic flow in the years of 1990 and 2003; (c) The changed area between the two

dates.

R. ZAKI ET AL.

192

Figure 16. The distribution of the land surfaces temperature of the year 1990.

Figure 17. The distribution of the land surfaces temperature of the year 2003.

R. ZAKI ET AL.

193

LST to some extent; along the Cairo-Ismailia, Cairo-

Suez high roads and through the urban built up spaces of

the Tenth of Ramadan city. These urban settlement were

characterized by the highest temperature range (36.6˚C -

39.45˚C) as discordant with the surroundings.

In 2003, the spatial distribution of the land cover units

has suffered much change with the urban invasion. Con-

sequently LST patterns were changed and their distribu-

tions were mostly controlled by the urban distributions.

There is no change in land cover area (the rural culti-

vated lands) at the northwestern part of the study area

corresponds to no change LST values with exiguous de-

crease in temperatures in some parts (from 24.5˚C -

27.8˚C to 21.8˚C - 24.4˚C and from 27.9˚C - 30.7˚C to

24.5˚C - 27.8˚C) give indication that the colder weather

conditions of this day corresponding to the day of 1990.

On the other hand, the central and southern parts that

involved the bulk urban expansion have been covered by

the higher temperature set (36.5 - 39.4). The wind

movement has the north- east direction, so the hot spots

(with the red color) at the southern part may be carried

out the wind from the industrial settlements at the desert

back of Nasr city, which located at the south- west of the

study area.

10. Conclusions

Using remote sensing data in conjunction with Geogra-

phic Information System analytical tools, the present

study enables to monitor sp atial and temporal land cover

changes in the region around Gabal El-Hamza and to

evaluate the dynamics of urban expansion in the new

urban cities at the northeastern side of the Greater Cairo

area. This study was conducted using two Landsat The-

matic Mapper (TM) images acquired in the years 1990

and 2003. The acquisition time of th e two images was in

the same season to keep the weather factor as constant as

possible. The images were taken in August 1990 and

June 2003 respectively. Different procedures of image

enhancement have been performed on the two images to

obtain the land cover maps at the two dates using likely-

hood supervised and unsupervised classifications. The

maps obtained revealed that the unsupervised classi- fi-

cation are more accurate in distinguishing the land cover

units because of human error in digitizing, lack of

knowledge of study area, and other factors generally

contribute to give inaccurate results in the supervised

classification method. These drawbacks are not affecting

the unsupervised classification method. After refinement

of the unsupervised classification, accuracy assessment

analysis had been performed to determine the accuracy

of the classification. The total accuracy of the Landsat-

derived land cover data was 92 and 95% for the years

1990 and 2003 respectively. The obtained results indi-

cate that the accuracy decreased with the increasing of

the urban expansion because the expansion results in the

increasing of spectral confusion between some pixels of

the urban signature and pixels including of argillaceous,

rocky or basaltic exposures. In order to determine the

change in the land cover units through the thirteen years

between the two acquisitio n dates, the post-classification

comparison technique were used. The agricultural cover

changed from 89.6 to 150.4 km2 for the years of 1990

and 2003 respectively. The urban areas show a large in-

creased from 49.5 to 120.9 km2. The basaltic flows cover

changed from 3 to 3.75 km2. The sandy cover decreased

from 68.9 to 60.1 km2 and the exposures of the rocks

decreased from 904.8 to 780.8 km² (i.e., a decrease of

124 km2 in 13 years).

This study shows that an integrated use of GIS and

Remote Sensing data can be effectively used to under-

stand spatial and temporal dynamics of land cover

changes. The interpretation and classification of remote

sensing data could also be useful for estimating the rate

and spatial pattern of the urban expansion around old

cities that have a desert back-ground as Greater Cairo,

Alexandria, El Minia and others. The land cover maps

produced in this study could contribute to both the de-

velopment of sustainable urban land use planning deci-

sions and also for forecasting possible future changes in

growth patterns by using the Polychotomous Regression

modeling and predicting the land cover in the future.

Landsat TM thermal infrared data indicated that the sur-

face temperature is strongly affected by the land cover

changes and needs further research to correlate with Me-

teorology.

11. References

[1] United Nation, “Report of the Meeting-Urbanization: A

Global Perspective,” Proceedings of the Expert Group

Meeting on Population Distribution, Urbanization, Inter-

nal Migration and Development, New York, 21-23 Janu-

ary, 2008.

[2] A. Yeh and X. Li, “Economic Development and Agri-

Cultural Land Loss in the Pearl River Delta, China,”

Habitat International, Vol. 23, No. 3, 1999, pp. 373-390.

doi:10.1016/S0197-3975(99)00013-2

[3] World Bank, “World Development Indicators,” Washing-

ton, D.C., 2007.

[4] M. W. Holdgate, “The Sustainable Use of Tourism: A

Key Conservation Issue,” Ambio, Vol. 22, 1993, pp.

481-484.

[5] S., Hathout, “The Use of GIS for Monitoring and Pre-

dicting Urban Growth in East and West St. Paul, Winni-

peg, Manitoba, Canada,” Journal of Environmental

Management, Vol. 66, No. 3, 2002, pp. 229-238.

R. ZAKI ET AL.

194

[6] M. Herold, N. C. Goldstein and K. C. Clarke, “The Spa-

tiotemporal form of Urban Growth: Measurement,

Analysis and Modeling,” Remote Sensing of Environment,

Vol. 86, No. 3, 2003, pp. 286-302.

doi:10.1016/S0034-4257(03)00075-0

[7] P. Serra, X. Pons and D. Sauri, “Land-Cover and

Land-Use Change in a Mediterranean Landscape: A Spa-

tial Analysis of Driving Forces Integrating Biophysical

And Human Factors,” Applied Geography, Vol. 28, No. 3,

2008, pp. 189-209.

doi:10.1016/j.apgeog.2008.02.001

[8] B. P. Zhang, Y. H. Yao, W. M. Cheng, C. H. Zhou, Z. Lu,

and X. D. Chen, “Human-induced changes to biodiversity

and alpine pastureland in the Bayanbulak Region of the

East Tianshan Mountains,” Mountain Research and

Development, Vol. 22, No. 4, 2002, pp. 1-7.

doi:10.1659/0276-4741(2002)022[0383:HICTBA]2.0.CO

[9] H. H., Kim, “Urban Heat Island,” International Journal

of Remote Sensing, Vol.13, No. 12, 1992, pp. 319-336.

[10] J. E. Nichol, “Visualization of Urban Surface Tempera-

tures Derived from Satellite Images,” International

Journal of Remote Sensing, Vol. 19, No. 9, 1998, pp.

1639-1649. doi:10.1080/014311698215153

[11] P. Kevin and W. Timothy, “Satellite-Based Adjustments

for the Urban Heat Island Temperature Bias,” Journal of

Applied Meteorology, Vol. 38, No. 6, 1999, pp. 806-813.

doi:10.1175/1520-0450(1999)038<0806:SBAFTU>2.0.C

O;2

[12] Y. Chen, J. Wang and X. Li, “A Study on Urban Thermal

Field in Summer Based on Satellite Remote Sensing,”

Remote Sensing for Land and Resources, Vol. 4, 2002, pp.

55-59.

[13] Q. Weng, D. Lu and J. Schubring, “Estimation of Land

Surface Temperature-Vegetation Abundance Relationship

for Urban Heat Island Studies,” Remote Sensing of Envi-

ronment, Vol. 89, No. 4, 2004, pp. 467-483.

doi:10.1016/j.rse.2003.11.005

[14] Y. Q. Wang, “Estimation of Land Surface Temperature

using Landsat-7 ETM+ Thermal Infrared and Weather

Data,” Department of Natural Resources Science, Uni-

versity of Rhode Island Kingston, RI 02881, USA, 2005.

[15] Q. Weng and D. Lu, “A Sub-Pixel Analysis of Urbaniza-

tion Effect on Land Surface Temperature and Its Inter-

play with Impervious Surface and Vegetation Coverage

in Indianapolis, United States,” International Journal of

Applied Earth Observation and Geoinformation, Vol. 10

No. 1, 2008, pp. 68-83.

doi:10.1016/j.jag.2007.05.002

[16] E. Tawfik and A. Swedan, “Geology of Cairo-Suez dis-

trict, Western Part of the N. Eastern Desert,” Geological

Survey, Egypt, 1992, report No. 34/92.

[17] F. F. Sabins, “Remote Sensing, Principles and In-

ter-pretation,” Freeman, New York, 1997.

[18] P. M. Harris and S. J. Ventura, “The Integration of

Geogra- Phic Data with Remotely Sensed Imagery to Im-

prove Classification in an Urban Area,” Photogrammetric

Engineering and Remote Sensing, Vol. 61, No. 8, 1995,

pp. 993-998.

[19] R. J. Jensen, “Remote Sensing of the Environment: An

earth Resource Perspective,” 2nd Edition, Prentice-Hall,

Upper Saddle River, 2007.

[20] T. M. Lillesand and R. W. Kiefer, “Remote Sensing and

Image Interpretation,” John Wiley & Sons, Inc., NewYork,

1994.

[21] R. G. Congalton, “A Review of Assessing the Accuracy

of Classifications of Remotely Sensed Data,” Remote

Sensing of Environment, Vol. 37, No. 1, 1991, pp. 35-46.

[22] R. Anderson, E. Hardy, J. Roach and R. Witmer, “A Land

Use and Land Cover Classification System for Use with

Remote Sensor Data,” USGS Professional Paper, Wash-

ington, D.C., 1976.

[23] D. Lu, P. Mausel, E. Brondízio and E. Moran, “Change

Detection Techniques,” International Journal of Remote

Sensing, Vol. 25, No. 12, 2004, pp. 2365-2407.

doi:10.1080/0143116031000139863

[24] A. S. Goudie, “Land Transformation. In The Challenge

for Geography: A Changing World, A Changing

Dis-cipline, In: R. J. Johnston Ed., Blackwell, Cambridge,

1993, pp. 117-137.

[25] R. M. Turner, “Long-Term Vegetation Change at a Fully

Protected Sonoran Desert Site,” Ecology, Vol. 71, No. 2,

1990, pp. 464-477. doi:10.2307/1940301

[26] G. Gutman, “Land Change Science. In: Monitoring and

Understanding Trajectories of Change on the Earth’s

Surface,” Dordrecht Kluwer Academic Publishers,

Netherlands, 2004, pp. 329-350.

[27] S. I. Rasool, “Potential of Remote Sensing for the Study

of Global Change,” COSPAR Report to the International

Council of Scientific Unions, Advances in Space Re-

search, Pergamon Press, Oxford, Vol. 7, No. 1, 1987.

[28] S. Ustin, “Manual of Remote Sensing: Remote Sensing

for Natural Resource Management and Environmental

Monitoring,” John Wiley & Sons, Chichester, 2004.

[29] R. J. Jensen, “Introductory Digital Image Processing,”

Prentice-Hall, Upper Saddle River, 1995, pp. 467-475.