Analysis of urban growth and sprawl from remote sensing data: case of fez, morocco

Analysis of Urban Growth and Sprawl from Remote Sensing Data Case of Fez, Morocco Accepted Manuscript Original Article/Research Analysis of Urban Growth and Sprawl from Remote Sensing Data Case of Fez[.]

Accepted Manuscript

Original Article/Research

Analysis of Urban Growth and Sprawl from Remote Sensing Data: Case of Fez,

Morocco

Abdelkader El Garouani, David J Mulla, Said El Garouani, Joseph Knight

DOI: http://dx.doi.org/10.1016/j.ijsbe.2017.02.003

To appear in: International Journal of Sustainable Built

Environ-ment

Received Date: 21 April 2016

Revised Date: 17 January 2017

Accepted Date: 10 February 2017

Please cite this article as: A El Garouani, D.J Mulla, S El Garouani, J Knight, Analysis of Urban Growth and

Sprawl from Remote Sensing Data: Case of Fez, Morocco, International Journal of Sustainable Built

Environment (2017), doi: http://dx.doi.org/10.1016/j.ijsbe.2017.02.003

This is a PDF file of an unedited manuscript that has been accepted for publication As a service to our customers

we are providing this early version of the manuscript The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Analysis of Urban Growth and Sprawl from Remote Sensing Data: Case

of Fez, Morocco

Abdelkader EL GAROUANI a , David J MULLA b , Said EL GAROUANI c & Joseph KNIGHT d

a

Faculty of Sciences and Techniques, Sidi Mohamed Ben Abdellah University, Route d’Imouzzer, BP 2202, Fez

30060, Morocco, e-mail: el_garouani@yahoo.fr

b Department of Soil, Water, and Climate, University of Minnesota, 1991 Upper Buford Circle, 439 Borlaug Hall,

Minneapolis, Minnesota, USA, e-mail : mulla003@umn.edu

c

Informatic Department, Faculty of Sciences, Abdelmalek Essaadi University, B.P 2121, Tetouan, 93002, Morocco,

e-mail: saidelgarouani@yahoo.fr

d Remote Sensing and Geospatial Analysis Laboratory, University of Minnesota, 1991 Upper Buford Circle, 439

Borlaug Hall, Minneapolis, Minnesota, USA, e-mail : jknight@umn.edu

Abstract:

Fez is the most ancient of the imperial cities of Morocco In Fez the rate of population growth has been spectacular in recent times (484 300 inhabitants in 1982 and 1 129 768 in 2014) The accelerated rate of population growth has generated a large urban sprawl in all its forms and serious environmental problems In this research, we have analyzed the relationship between urbanization and land use changes and their impact on cityscape in Fez and the importance of the increase in impervious surface areas Satellite imageries and census data have been used to identify different patterns of land use change and growth of the city for the period 1984-2013 Classification and analysis of the satellite imageries were performed using Erdas imagine and ArcGIS Software Urban sprawl in Fez was assessed over 29 years (1984-2013) The overall accuracy of land cover change maps, generated from post-classification change detection methods and evaluated using several approaches, ranged from 78% to 87% The maps showed that between 1984 and 2013 the amount of urban or developed land increased by about 121%,

while rural cover by agriculture and forest decreased respectively by 11% and 3%

Keywords: Urban expansion, classification, GIS, Remote Sensing, Fez, Morocco

I- Introduction

Urbanization that is considered as a positive

process linked to modernization,

industrialization and global integration has

economically benefitted only a minority of

the urban population (Bhatta, 2010; Sharma,

1985) During the last century, Moroccan

society was increasingly urban (Fig 1) The

accelerated rate of urbanization in all forms

and the population growth in Morocco has

been generating serious environmental

problems and concern for both the

government and interested stakeholders

(Lehzam, 2012)

The amount of impervious surface in a

landscape is an important indicator of

environmental quality Impervious surfaces

are defined as any surface which water

cannot infiltrate and are primarily associated with transportation and building rooftops (Bauer et al., 2007) Imperviousness increases water runoff, and hence, is a primary determinant of runoff volumes in urbanized areas The impervious surface area provides a measure of land use that is closely correlated with these impacts (Arnold and Gibbons, 1996) It therefore follows that impervious cover information is fundamental to assess flooding risks and flood management in the city

Economic development demands sustainable land management Spatial information on land use/land cover types and their change detection in time series are important means for city planning and new development activities (Ewing et al 2002) The present research is undertaken in that

spirit It will analyze the relationship

between urban growth and land use changes

and their impact on the Fez cityscape This

information is an essential tool in

decision-making and management policy of the city

by the local authority and for ensuring

sustainable urban growth and development

in the study area The period of focus is from

1984 to 2013 Topographical maps,

high-resolution satellite imageries and other

necessary data have been used to detect

land use/land cover changes in the study

area

Figure 1: Population growth in urban and rural areas

in Morocco (HCP, 2015)

Numerous researchers, including Arvind et

al (2006), Lunetta and Balogh (1999), Yuan

et al 2005, Zubair (2006) and others have

demonstrated the value of multitemporal

satellite imagery for classification of land

cover The strong development of remote

sensing and GIS technology has helped us to

study the urban space development

II- Study Area

Fez is located on the northern part of

Morocco (Fig 2) The urban community of

Fez accounted for 1 129 768 inhabitants in

2014 and the city has about 30 km² (El

Garouani et al., 2011) Founded in 789 by

Moulay Idriss 1st and home to the oldest

university in the world (Quaraouiyne

University built in 857) Fez reached its

height in the 13th–14th centuries under the

Merinids, when it replaced Marrakesh as the capital of the kingdom Although the political capital of Morocco was transferred

to Rabat in 1912, Fez has retained its status

as the country's cultural and spiritual center (Aouni et al 1992) Fez is a religious, touristic and academic center Due to its importance, the historic Medina of Fez was added to the UNESCO World Heritage List in

1981 Over the past 20 years, there were many problems and challenges posed by the rapid growth of Fez just like every other city

in Morocco

In fact, the demand for infrastructure, basic services and housing in expanding urban area in Fez are on the increase Moreover, provision of education, health, transportation, water and sanitation services should be accelerated in urban centers

III- Methodology

Present study is based on spatial remote sensing data as well as non-spatial data available from various sources for different periods Urban development has led to expansion of the cityscape of Fez, leading to changes in land use The study specifically focuses on interpreting the city’s land use change patterns and growth based on satellite and demographic data

Our approach combined spring and summer images In the summer image, the Fez urban area appears unvegetated and is

Figure 2: Location of the study area

distinguishable from forests and orchards

However, the spring image is needed to

separate vegetated areas from urban areas

with significant amounts of asphalt and

concrete and other impervious surfaces that

are spectrally similar to bare soil in a

summer image The importance of

multi-temporal imagery was confirmed by

determining the transformed divergences

for the dataset (Two images in 1984 and two

images in 2013) Compared to the single

dates, both the average and the minimum

separability of classes were increased by the

combination of spring and summer images.

This research uses Landsat images to

calculate VSW (Vegetation-Soil-Water) index

images, that distinguish clearly between

Vegetation, Soil and Water elements on the

image This VSW index image is the basis for

detecting clearly urban areas (My Vo Chi et

al., 2009) The VSW index is calculated by

the scatter plot of red band versus

near-infrared band, which are common in almost

of all types of satellite imagery (Fig 3) That

consist to transform or analyze a feature

space mathematically to isolate groups of

pixels that may be related Each axis

represents DNs from one satellite band and

plot each pixel in the feature space from an

image using its DNs

Image processing software has been used for geometric correction of satellite data, supervised classification, accuracy assessment of classification, land use maps (1984 & 2013), change detection, final output maps etc

Figure 3: Scatterplot of TM4 (Y-axis) and TM3 (X-axis)

GIS software has been used for the digitization, integration, overlay and presentation of the spatial and non-spatial data of land use change in the city Field surveys were performed throughout the study area using Global Positioning System

Fez Morocco

Mali

Algeria

Mauritania

Atlantic Ocean

N

Water

Soil Vegetation

Red reflectance

(GPS) to obtain accurate location point data

for each land use class included in the

classification scheme Figure 4 presents the

methodology used to produce maps of land

use change and urban expansion Table 1

presents the classification units of land use

identified in the study area

Figure 4: Flowchart of spatial and temporal changes

in urban land cover

Table 1: Land cover classification scheme

Agriculture Land Crop Land, Olive trees, Orchard

3.1 Satellite Imagery

To accomplish the objectives of the present

study, four available satellite images were

obtained from the United States Geological

Survey (USGS) databases online resources:

It was important to utilize images covering

the summer season to ensure that

agricultural land surrounding Fez are fully

assessed

Landsat Thematic Mapper (TM) and

Operational Land Imager-Thermal Infrared Sensor (OLI_TIRS) data have several

advantages for this application: synoptic view, digital, GIS compatible data, availability of data since 1984, and economical costs This paper extends the methods and results of our previous works (El Garouani et al., 2012, 2014 and 2015) and of the present contribution

Data used in the research include the multi-temporal dataset and topographic maps:

 Landsat TM and OLI-TIRS images (For 18-1984, 04-15-1985, 06-15-2013 and 08-18-2013) - Google earth images (09-12-2013)

 Aerial photographs - Topographic maps

 Demographics and census data (1982,

1994, 2004 and 2014)

 Urban development plan of Fez - Field observations, etc

Landsat images are described below:

The Landsat Thematic Mapper (TM) sensor was carried on Landsat 4 and Landsat 5, and images consist of seven spectral bands with

a spatial resolution of 30 meters for Bands 1

to 5 and 7 (Table 2)

Table 2: Spectral bands description of Landsat TM

Landsat 8 Operational Land Imager (OLI)

images consist of nine spectral bands with a spatial resolution of 30 meters for Bands 1

to 7 and 9 The ultra-blue Band 1 is useful for coastal and aerosol studies Band 9 is useful for cirrus cloud detection The resolution for Band 8 (panchromatic) is 15 meters (Table 3)

Table 3: Spectral bands description of Landsat OLI

Band 1 - Ultra Blue (coastal/aerosol) 0.43 - 0.45

Before analysis, the images were

geometrically corrected During geometric

correction, control points are detected on

the topographic maps and the satellite

images with RMS errors that are estimated

below 0.5 pixel After that, the images for

1984 and 2013 were registered on Lambert

Conform Conic Projection, datum Merchich,

zone I (North Morocco)

A subset image was created from each

Landsat image for subsequent treatment

and classification (Fig 5)

3.2 Image classification

Image classification is a conventional change

detection method The advantage of image

classification is the ability to create a series

of land cover maps We applied the

maximum likelihood supervised classification

(MLC) method for time series of Landsat

bands and VSW index

The maximum likelihood algorithm is one of

the most widely used in the classification of

satellite imagery The method is based on

the likelihood that each pixel belongs to a

particular class The basic theory assumes

that these likelihoods are equal for all

classes and that input bands are uniformly

distributed The method requires a

significant calculation time and is based on a

normal distribution of the data in each band

in the classification It tends to over-classify

signatures with relatively large values in the

covariance matrix (Vorovencii and Muntean,

2013)

MLC is performed according to the following

steps (Richards and Jia, 1993) The method

consisted in choosing the training samples

for each of the desired classes from the color composite image During the training phase, 40 training sites were selected by on-screen digitization of specific polygons (5 training samples by thematic class) The files obtained were saved and used for the images classification Each training field was assigned a number from 1 to 8 representing land cover classes including: water, urban, industrial area, rangeland, olive trees, orchard, forest and agriculture

By using GIS technique (convert vector, overlay and calculate), the urban expansion information (areas, the replacement of land covers to urban area) was assessed over the study periods The areas identified as urban

in 2013, but not developed in 1984 had a high greenness value (due to vegetative cover) in the 1984 imagery and thus had low

to no impervious surface in the 1984 time period

IV- Results and Discussion

4.1 Land use analysis

Urbanization is a major cause of land use changes and land conversions It makes unpredictable and long lasting changes on the landscape An important aspect of change detection is to determine what is actually changing to what i.e which land use class is changing to the other Analyzing the spatial and temporal changes in land use and land cover is one of the effective ways to understand the current environmental status of an area and ongoing changes (Arvind, 2006, Yuan et al 2005 and Zubair, 2006) The land use maps of two points in time, that is, 1984 and 2013 based on automatic classification and visual interpretation respectively depict land use categories changes such as residential, agriculture, industrial, water body, forest, etc (Fig 6).The growth of urban area and accompanying increases in amount of impervious surface area are readily apparent

Figure 5: False color composites of the Landsat and Google Earth images showing Fez in 1984 and 2013.

04-15-1985

06-15-2013

08-18-1984

08-18-2013

09-12-2013

An independent sample of an average of 30

polygons, with about 250 pixels for each

selected polygon, was randomly selected

from each classification to assess

classification accuracies Error matrices as

cross-tabulations of the mapped class vs

the reference class were used to assess

classification accuracy (Congalton & Green, 1999) Overall accuracy, user’s and producer’s accuracies, and the Kappa statistic were then derived from the error matrices (Table 4)

Figure 6: The land cover maps of Fez in 1984 and 2013

El Gaada dam

Dhar Mehrez dam

Table 4: Error Matrix Analysis of field sites (columns) against Landsat classification (rows)

a-1984

Accuracy 84.97 91.63 61.11 97.72 97.46 73.99 54.69 96.07

b-2013

Accuray 77.38 90.67 74.47 76.38 77.41 59.12 91.01 94.90

ErrorO = Errors of Omission (expressed as proportions) - ErrorC = Errors of Commission (expressed as proportions)

4.2 Change detection

Following the classification of imagery from

the individual years, a multi-date

post-classification comparison change detection

algorithm was used to determine changes in

land cover in the interval; 1984–2013

According to the results from 1984 to 2013,

the urban change was large, urban areas

increased from 2041 ha (1984) to 4503 ha

(2013) (Table 5)

Error matrices were used to assess

classification accuracy and are summarized

in Table 4 The overall accuracies for 1984

and 2013 were, respectively, 87% and

78.5% User’s (Field) and producer’s

(Landsat classification) accuracies of

individual classes were consistently high,

ranging from 55% to 98%

Classification maps were generated for two

years (Fig 6) and the individual class area

and change statistics are summarized in

Table 5 From 1984 to 2013, urban area

increased approximately 2462 ha (121%) and olive trees increased 981 ha (89%) while non-orchard agriculture decreased 4124 ha (11%) and forest decreased 19 ha (3%) For the water body surfaces there is an increase

of 247% This large increase can be explained by the construction of two dams (El Gaada in the East and Dhar Mehrez in the South) and, on the other hand, by the heavy rainfall in 2013, which allowed the emergence of wetlands in the NW (Table5) The relationship between population growth and growth in urban land area as determined from the Landsat-derived change maps was also examined Development patterns of Fez reflect the distribution of population and households because residential land uses take over all the land that is developed (HCP, 2015) The average annual growth in urban area determined from the Landsat change detection was 4.2 % from 1984 to 2013 This

rate of approximately 3.9 % from 1982 to

2014 So the growth in urban area is

relatively higher than the population growth

rate Population and urban expansion data

were also tabulated (Table 6) An urban

sprawl index (USI) was calculated as the

ratio of urban expansion to population

increase (OECD, 2013) The urban sprawl

index measures the growth in urban area

over time adjusted for the growth in

population When the population changes,

the index measures the increase in the

urban area over time relative to a

benchmark where the built-up area would

have increased in line with population The

index is equal to zero when both population

and urban area are stable over time It is

larger than zero when the growth of the

urban area is greater than the growth of

population, i.e the density of the

metropolitan area has decreased

In Fez, the USI = 4.2 / 3.9 = 1.08 It is slightly

higher than zero, so the urban area is

slightly higher than the population growth

This index provides a way to assess the

degree of sprawl for each region

To further evaluate the results of land cover

conversions, a matrix of land cover changes

from 1984 to 2013 was created (Table 7) In

the table, unchanged pixels are located

along the major diagonal of the matrix

These results indicate that increases in

urban areas mainly came from conversion of

agricultural, rangeland and orchard land to

urban uses during the period, 1984–2013

Of the 2 462 ha of total growth in urban

land use, 22 65 ha was converted from

agricultural land, 82 ha from orchard and 70

ha from rangeland We note that 70 ha of

rangeland was converted to urban between

1984 and 2013, while at the same time, 5 ha

of urban was converted to rangeland These

changes may be classification errors

Classification errors may also cause other

unusual changes For example, 288 ha of

agricultural land changed to rangeland and

42 ha of rangeland changed to agricultural

land These changes are most likely

associated with omission and commission errors in the Landsat classifications change map Registration errors and edge effects can also cause apparent errors in the determination of change vs no-change This comparative approach has demonstrated how landscape changes can

be derived from satellite imagery in the urban spatial structure Interpretation of Fez’s growth over a period of 29 years allows a deeper understanding of growth mechanisms, underlying drivers of urban expansion, and their effects on local livelihoods

According to our observations, urban sprawl has a negative impact on infrastructure and the sustainability of Fez The information on land use change reveals both the desirable and undesirable changes and classes that are “relatively” stable overtime This information serves as a vital tool in decisions making and policy formulation by the local authority For example, due to urban expansion, Fez lacks vegetation cover Needless to say that vegetation and open green spaces (parks) are the most important parameters of quality of urban environment assessment Hence, a vigorous focus needs

to be given to grow more trees and also develop green belts that can reduce a city’s ecological footprint and carbon emissions significantly A suitable strategy to reclaim industrial wastelands is also required

On the other hand, as the city grows in size and population, harmony among the spatial, social and environmental aspects of a city and between its inhabitants becomes of paramount importance Urban development should be guided by a sustainable planning and management vision that promotes interconnected green space, a multi-modal transportation system, and mixed-use development Diverse public and private partnerships should be used to create sustainable and livable communities that protect historic, cultural, and environmental resources In addition, policymakers, regulators and developers should support