1 IntroductionGIS is weaving relationships between fields and disciplines for the benefit of meeting anumber of objectives, as has been seen in the work of a number of researchers for thepurposes of understanding morphological change (Koster, 1996) and persistence andchange in urban morphology (Scheer and Ferdelman, 1997), promoting participatoryplanning (Geertman, 2002) (Alshuwaikhat and Nkwenti, 2002), and modeling urbanresidential dynamics (Benenson et al, 2002).

GIS is therefore opening avenues for research into urban change (Batty, 1996),facilitating the study of the urban morphology of the built landscape in terms oftime, form, and scale (Moudon, 1994), and, one hopes, furthering our understandingof cities as spatial events in terms of duration, intensity, volatility, and location (Batty,2002), which, notwithstanding the fact that any system of representation is ``a spa-tialization of sorts which freezes the flow of experience and in so doing distorts whatit strives to represent'' (Harvey, 1990, page 206), bears the promise of abating theshortcomings of more traditional methods of representation of space and time.

However, in some cases GIS has the potential of occasionally producing a dataoverload (Scheer and Ferdelman, 1997), a condition that might be particularly acute inwhat may be described as data-rich environments where data are more routinelycollected, where physical conditions are more consistently mapped or documented,and where such data are more easily accessible. In contrast, where research has tobe conducted in what may be described as data-poor environments, GIS has thepotential of maximizing the value of whatever scarce data may be available. Thismay sufficiently characterize the case presented here. In this paper we outline aprocedure for the maximization of research benefits by using data that are generallymore accessible, such as national census data, in combination with a smaller sampleof a rather more restricted-access type of data, concerning landownership registry. Ouraim is to understand physical change and its relation, if any, to change in property

GIS for understanding physical and social change in urbansettings: a case from Amman, Jordan

Nabil Abu-DayyehDepartment of Architecture, Faculty of Engineering and Technology, University of Jordan,Amman, Jordan; e-mail: abudayye@cyberia.jo

Firas ZiadatFaculty of Agriculture, University of Jordan, Amman, JordanReceived 8 August 2003; in revised form 10 June 2004

Environment and Planning B: Planning and Design 2005, volume 32, pages 127 ^ 140

Abstract. Data-poor environments, or where data are not routinely collected and/or poorly archived,or where public records are not easily accessible understandably create a major obstacle to researchand the reaping of the benefits of research for urban policy and urban management. However, in suchrather unwelcoming research environments GIS seems to provide enhanced opportunities for max-imizing the benefits of whatever data are available, as scarce as they may be. This research conditionmay not be unique to the case study presented here but presumably may apply to most countries inthe developing world. The authors' use of GIS for the analysis of the case of an old neighborhood ofAmman outlines a procedure for zooming in on areas of construction activity and understanding therelation between physical and social change, social change being here represented by change inlandownership.

DOI:10.1068/b3185

ownership at the neighborhood level, for Jabal Lweibdeh, an older neighborhood ofAmman, capital city of the Kingdom of Jordan.

Contrary to the typically perceived image of this neighborhood as a more historicalneighborhood (Jarrar, 2002) dating mostly from the 1950s and 1960s, the numbersshow that there was considerable building activity at the heart of the neighborhoodin the 1990s that nonetheless seems to have had little effect on the perceived image ofthe neighborhood as a historical neighborhood with long-standing physical fabric anda concomitant little-changing social fabric. Although the overall assessment of thesocial fabric is beyond the scope of this study, one aspect or indicator of its persistenceor change is the persistence or change in landownership considering the fact that inJordan as a whole, as well as in Amman, more than half the dwellings are owner-occupied (DoS, 1997). Therefore, an analysis was made to investigate the resonance ordissonance between building activity at the heart of the neighborhood in the 1990s andpatterns of landownership change throughout the period in question, 1946 ^ 2002,being the whole period covered by land registry records, which date from the inde-pendence of the Kingdom in 1946 up to the time the data were collected by us in 2002.Our analysis shows that, as a whole, the pattern of landownership change over the wholeperiod resonates well with the construction activity pattern, but it fails to resonate withthe sharp rise in construction activity in the 1990s in one of the main areas at the heartof the neighborhood, an area that witnessed in the 1990s one of the highest rates ofconstruction activity within the whole neighborhood. This means that the recent surgeof construction activity in that central area has been quite divorced from any majorchange in landownership. Recent urban development therefore seems to have beenundertaken by the same landowners, not new owners, thereby contributing to thepersistence of the image of a tight-knit little-changing social fabric.

2 The study areaSince independence in 1946 the Hashemite Kingdom of Jordan and its capital Ammanhave been sent into spirals of urban growth, with an increase in population as wellas considerable spatial, social, and economic transformations. In fifty years from1952 to 2002 the population of Jordan increased from little over half a million inhab-itants to over five million inhabitants (DoS, 2001). The capital city, Amman, witnessedsimilar cataclysmic growth in population, from 110 000 inhabitants in 1952 withinan area of around 25 km2, to over two million inhabitants today living in an area of526 km2 (MGA, 1988).

The physical urban morphology of the city of Amman took form from a mixture ofelements, most importantly from its unique topography. The area lying within a 3 kmradius of the main commercial and historical center of the city has a very distincttopographic formation creating clearly defined `morphological regions'. A series ofrelatively high plateaus surround and define between them a main gorge where theold rivulet, Seil Amman, used to run, now roofed over by a major traffic artery thatruns along the historic core of the city, and other minor gorges which today act asmain radial roads emanating from the city center. These plateaus (in Arabic, Jabals)surrounding the city center are, in general, mature communities mixing an attractivedomestic architecture, planting, and a limited range of materials to create a balancedsetting for family life. Some of these Jabals have developed as low-rise, two-to-fourstorey residential areas initially conceived and planned along the same lines as Britishpostwar neighborhood units (King, 1955; Hacker, 1960). Such is the case with our studyarea, Jabal Lweibdeh, the unique topographic formation of which and the effort of theplanners to develop it as a low-rise residential neighborhood succeeded in making it apleasant setting for family life.

128 N Abu-Dayyeh, F Ziadat

Settlement on this Jabal dates from the early days of the Emirate of Transjordanunder the British Mandate in the 1920s. It had its peak of construction activity in the1950s and, to a slightly lesser extent, in the 1960s. Today, it seems to have reachedmaturity both socially and physically, thereby providing a reasonably long history ofurban habitation, a long range of building activity, and a reasonably long recordof landownership covering a period of over half a century, that is, since 1946. All thesefactors make Jabal Lweibdeh a good place to start inquiring about modalities ofchange in one of the older and better established neighborhoods within a fast-growingmetropolitan area.

According to current administrative boundaries adopted by the Municipality ofGreater Amman (MGA; figure 1), Jabal Lweibdeh covers an area of 221ha of landinhabited, according to data for 2000, by around 25 000 inhabitants, with around 83%of all plots occupied by one or more buildings.(1) Data concerning construction activityin the whole neighborhood are available from the General Census of Population andHousing of Jordan, carried out in 1994 (DoS, 1997). Census data are normally collectedfor each building within statistical blocks, in this case thirty-seven blocks covering

(1) There is some variance between data provided by MGA, by the Department of Statistics (DoS),and by the Department of Lands and Survey (DLS). The MGA lists the total number of plots inJabal Lweibdeh as 1750, whereas the DLS plot count performed by the researchers reached 2206.Moreover, DoS census data lists 1571 buildings, whereas MGA data for 2000 list only 1447;however, in this case the variance may be attributed to the fact that the MGA lists the numberof plots occupied by buildings and not individual buildings; understandably, one plot may havemore than one building on it.

0 2 4 8km

Municipal areas and subdivisions

Jabal Lweibdeh

Figure 1. Greater Amman boundaries and subdivisions, showing the study area, Jabal Lweibdeh.

GIS for physical and social change 129

the whole neighborhood. Disclosure of the exact spatial location of each individualbuilding within the statistical block is prohibited by the 1950 Law of Statistics Num-ber 24, and, at the same time, statistical blocks are so varied in size that data had to beused in an amalgamated fashion, both in order to minimize the differences that existbetween statistical blocks in terms of areas and number of buildings contained withinthem and in order to form discrete and more meaningful morphological areas. Thus,five block groups (BGs) were formed around main foci of open space or landmarks:the Hawooz Circle (BG1), Terra Sancta College (BG2), Al-Muntazah public gardens(BG3), Al-Abdali Camp and Army Headquarters (BG6), and Jordan Hospital (BG5);another block group was formed around the area known as West Lweibdeh, circum-scribed by three main traffic arteries (BG4), thereby making a total of six block groups(figure 2). However, the number of buildings contained within each block group varies,but such variations are inevitable if we want to safeguard the integrity of well-definedand meaningful morphological areas.

Initial investigations revealed that:1. A relatively high percentage of all buildings constructed in the study area are withina morphological area comprising thirteen statistical blocks, hereby named BG1, whichaccounted for 37% of all buildings constructed in the whole neighborhood (figure 3).This area constitutes the older part of the neighborhood, although older buildings havebeen found in small pockets; within BG4, for example.2. The peak of construction activity in BG1 occurred in the same period as the peak ofconstruction activity for the majority of block groups, that is, the 1950s and 1960s, andthe building activity chart for BG1 reflects the general pattern of building activity in allblock groups (figure 4).3. However, there was a surge in building activity in BG1 in the 1990s clearly distinguishablein the particular area of block 91028 (figure 5, over).

Block group 1

Block group 2

Block group 3

Block group 4

Block group 5

Block group 6

6

5

43

2

1

0 250 500 1000m

Figure 2. Map of Jabal Lweibdeh, showing all 37 Department of Statistics blocks, and blockgroups 1 ^ 6.

130 N Abu-Dayyeh, F Ziadat

In general, there was a rise in construction activity in Jordan and in Amman as awhole in the 1990s, particularly relative to the sluggish 1980s (figure 6, over). However,in the case of Jabal Lweibdeh the rise was particularly acute, especially in BG1,thereby warranting closer examination, as has been undertaken. It was hoped that acloser and more detailed study of this particular area of the city might provide someclues about the nature of this construction activity at the core of one of Amman'shistoric neighborhoods. In the research presented here, however, a single factor wasconsidered, patterns of landownership change within this study area and its resonanceboth spatially and temporally, if any, with the acute rise in building activity in this area.

BG19%

BG56%

BG416%

BG314%

BG218%

BG137%

Figure 3. Percentages of constructed buildings in each block group (BG) relative to the totalnumber of buildings in all block groups. Note: for the location of the block groups, see figure 2.

14

12

10

8

6

4

2

0

Percentage

1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s

0.0 0.1

1.5 1.5

3.8

6.4 5.5

3.0

3.1

13.7

Decade

BG1 BG2 BG3 BG4 BG5 BG6

Figure 4. Number of constructed buildings in each block group (BG) as a percentage of the totalin all block groups. Note for the location of the block groups, see figure 2.

GIS for physical and social change 131

3 The basic datasetsCollected data comprised map data and attribute data from a variety of sourcesavailable locally.3.1 Map data3.1.1 Cadastral mapsCadastral maps provide the most basic set of spatial data. In Jordan, these areprovided by the Department of Lands and Survey (DLS), and are available for all ofAmman and most of the urban areas in Jordan. These maps are organized by sheetnumber according to the available hard-copy drawing sheets that form the backbone ofthe DLS's cadastral survey; the digital version of these sheets is organized in the same

80

70

60

50

40

30

20

10

0

Percentage

1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s

0 2 3

19 29

2 30

43

SBs 91022, 91024, and91025

SBs 91026, 91027, and91029

SB 91028

SBs 91030, 91031, and91032 SBs 91033, 91034, and 91037

Decade

Figure 5. Percentage construction activity in statistical blocks (SBs) in block group 1 (BG1).Note: for the location of the block groups, see figure 2.

25

20

15

10

5

0

Con

structionactivity

(%)

Before 1950 1950 ± 54 1955 ± 59 1960 ± 64 1965 ± 69 1970 ± 74 1975 ± 79 1980 ± 84 1985 ± 89 1990 ± 94

Jordan Amman Lweibdeh

Period

3

3 3 2

4 6

89

9

67 7

42

12 11

121414

16

18

20181716

11

15

10

16

Figure 6. Construction activity in Jabal Lweibdeh (JL), in Amman, and in Jordan: constructionactivity as a percentage of the total number of buildings.

132 N Abu-Dayyeh, F Ziadat

manner according to sheet number. These spatial subdivisions do not reflect any clearsense of `natural areas'. For Jabal Lweibdeh, for example, one sheet contains dataabout part of Jabal Lweibdeh and parts of two neighboring Jabals. Therefore, severalsheets had to be edited and joined to create a single map for the whole study area.

These digital data are normally georeferenced by the DLS using the Cassinicoordinate system (Palestine Grid). For the purposes of this study, in order to under-take spatial analysis, these maps need to be registered to the same projection system asthe orthophotos and the land-use maps, as will be explained in more detail below.However, the exact parameters of this system that would enable transformation toother coordinate systems is considered classified' information. Therefore, in view ofthe absence of knowledge of these parameters they were registered to the sameprojection as the orthophoto by an approximate method whereby several outstandingpoints were matched on both maps until the resultant adjustment seemed satisfac-torily accurate, using ERDAS software (Leica Geosystems). Thenceforth, the digitalcadastral maps could be used as base maps for the purposes of further analysis.

3.1.2 Aerial photographsThe Royal Jordanian Geographic Center (RJGC) is the official government organiza-tion that normally undertakes the provision of aerial photographs, satellite images,topographic maps, and other spatial data, particularly georeferenced data maps, forlocal governmental, academic, and research purposes. These are produced or collectedand processed by the RJGC itself, then distributed or sold to other governmentorganizations or departments, or to research institutions, or to individuals associatedor affiliated with such institutions. Aerial photographs and maps are usually providedin hard-copy format for a range of years spanning the period 1918 ^ 2000. These arenormally useful for tracing the physical urban growth of the study area as successivelayers are overlapped. However, this approach was not followed here in view of itstime-consuming and labor-intensive character and the lack of consistency in the scaleand periodicity of the photography itself. Instead, tabulated data about constructiondates of all buildings in the study area were taken from the National Census statisticsand used in our GIS analysis.

Digital orthophoto versions of these aerial photographs are also available fromRJGC. The most recent orthophoto was used as a base map to compile other dataand to trace the physical formation of the building fabric within the study area. RJGCusually provides these orthophotos in any projection system; however, the registration(georeferencing) in the Cassini coordinate system is not totally accurate unless somefieldwork is undertaken to fix a number of control points, which is not always afford-able within smaller project budgets. Therefore, in search for higher accuracy, a JordanTransverse Mercator (JTM) projection was chosen to georeference the required ortho-photo for the study area. One of the advantages of using JTM georeferencing is that allits parameters are known accurately, a matter that facilitates its registration into anyother projection system, and vice versa. This is a necessary criterion for the selection ofthe projection system of a base map.

Subsequently, a new layer was extracted through on-screen digitizing of the outlineof the existing buildings for the whole of the study area, using the base orthophoto inJTM projection as background. The resultant layer was overlaid with the transformeddigital cadastral maps (in JTM projection) and the resultant match between the outlineof buildings and the parcel boundaries seemed reasonably accurate.

3.1.3 Land-use mapsUpdated land-use maps are normally available from the MGA which has an ongoingproject to digitize all available planning-related maps such as land-use maps and to

GIS for physical and social change 133

overlay these maps with the constructed buildings and with all relevant data, includingowners, building license numbers, land-use zoning area codes, number of floors, andso on. However, so far, only a few limited areas of the city of Amman have been sodigitized, and it will take many more years to finish the work for the whole of the cityof Amman, including our study area.

However, hard copies of land-use maps are available and such a map was obtainedfor the study area. This map was useful for updating the cadastral maps, whichnormally do not reflect the exact current boundaries of land parcels because of thelag in the updating of the most recent boundaries of land parcels according to the mostrecent planning acts, which may have affected their shape and area. The currentpractice of the DLS department is to update those parcels on a gradual basis andonly whenever there are any changes in the ownership of the property, placing theburden of updating the data on the individual owner.

Therefore, the hard-copy land-use map was scanned and the image was georefer-enced using an image-to-image correction method using the orthophoto as base toobtain coordinates in the JTM projection system. Subsequently, the land-use imagewas overlaid by the cadastral map and this map was adjusted according to the mostrecent parcel boundaries indicated in the land-use image.

3.2 Attribute data3.2.1 Department of Statistics data on the construction of buildingsJordan's Department of Statistics (DoS) conducts national surveys and national cen-suses: so far, four such censuses have been carried out, in 1952, 1961, 1979, and 1994.The census count is done by a particular method used by the DoS whereby eachgovernorate (muhafatha) is divided into a number of administrative subdivisions,Liwa ', Qada ', Nahiya, Tajammu ', and finally into statistical blocks whereby each blockcontains a number of buildings that can be very varied: in the case of the blocks usedin our study, the numbers varied from a minimum of eleven to a maximum of seventyfour. The spatial boundaries for all divisions and subdivisions are clearly defined butthe location of individual buildings within each block is, as mentioned earlier, prohib-ited by law, which is a major drawback for the purposes of our research. Therefore,statistical blocks were merged to form discrete and more meaningful and comparablemorphological areas, in this case into six block groups (see section 2). This allowed foran overall reading of the urban physical growth within the study area and theimmediate context, but without any close detailed inspection or analysis at the levelof individual plots. In the end, the overall patterns derived from these data werecross-checked by comparison with whatever aerial photos were available from RJGC.

Nevertheless, the level of accuracy of this kind of data may not always be extremelyreliable, on at least three counts. First, there is a good percentage of buildings forwhich the construction date is listed as `unknown'; in the case of our research, theproportion of buildings for which the construction date was listed in the census data as`unknown' was approximately 15% for the whole neighborhood. These are buildingsfor which the field counters were unable to judge or ascertain the actual date of con-struction. Second, data on dates of construction are normally based on oral statementsgiven to field counters by interviewees, who may not always be too knowledgeableabout the correct dates. Third, there is the element of human error in the collection andprocessing of data, which, as always, is an integral part of any data collection. Thus,the data permit only general conclusions to be drawn.

3.2.2 Department of Land and Survey landownership registryRaw data about landownership was obtainable, with special permission, from theDLS. The registries are classified according to drawing sheet number, this number

134 N Abu-Dayyeh, F Ziadat

being the number of the actual hard-copy drawing sheet that contains the cadastralmap of a certain defined area as it exists in the DLS archives. Each register entrynormally contains the name of the owner or owners related to the plot number, theold plot number, and the type and date of any ownership transaction in nine maincategories: initial registry (Taswiyeh), inheritance (Intiqal ), exchange (Tabadul ), gift(Heba), sale (Bai '), settlement between partners (Takharoj ), subdivision (Ifraz), delega-tion (Tafweed '), or correction (tas'hee'h). Combinations of these categories also exist.In total, this gave us 666 register entries covering an area of around 43.3 ha reasonablyclosely matching the area covered by BG1. There were, however, some missing recordsfrom the Land Registry Archive files, but they were very few in number and did notinfluence our research.

The raw data from the registry were reorganized in a different format in order tofacilitate the extraction of more meaningful sets of information. The data were reor-ganized in a horizontal fashion, showing the sequence of ownership changes for eachland parcel, and, for each change, the type of ownership change was noted in additionto the consequent number of owners and the date when the change took place. Thetype of change was divided into twelve categories encompassing the range of changesto landownership: (1) change within the same family, (2) change to a different family,(3) a new partner moving in, (4) a partner moving out, (5) a partner moving out and a newpartner moving in, (6) change to a nongovernmental organization (NGO), (7) change tothe MGA, (8) change of or to a company, (9) change to a religious Waqf, (10) changeto a bank, (11) change to delegated ownership, and (12) change to an embassy. Even-tually, the most significant for our analysis were the first two categories, which describethe change of property either within or outside the same family.

In addition, for the first and last records in the sequence of ownership change ofeach land parcel, the identity of the property owner was noted, whether an individualor group of individuals, an institution, a governmental organization, or an NGO. Also,for individual owners, their religious identity was noted. Furthermore, a code numberwas given to identify the type of change that described the overall landownershipchange between the very first and very last owner or owners, significant mainly fordescribing changes of land property (1) within the same family, (2) outside, to anotherfamily, (3) to an NGO, (4) to the MGA, (5) to a company, (6) to aWaqf, (7) to a bank,or (8) to a governmental organization.

The level of detail of these data was not always crucial to the final analysis but itwas presumed that the compiled database could in the future form the basis for furtherresearch in this area.

4 Queries and results4.1 Spatial development of the neighborhoodThe more usual method of tracing the spatial growth of any urban area is done byoverlaying aerial photographs taken at regular intervals, or whatever available aerialphotography exists, in order to deduce the pattern of physical growth. Compared withthe use of a GIS it has several drawbacks:(a) in some cases, as in the case of Amman, aerial photography may not be available atregular time intervals or at a consistent scale;(b) it is usually a time-consuming and labor-intensive operation; and(c) it does not easily allow the mapping of specific spatial distributions in terms ofpercentages or specific time intervals.

In contrast, GIS can provide not only a less time-consuming and less labor-intensive method of representing physical growth but also, and more importantly,a much more specific representation of data in terms of space and time. This can be

GIS for physical and social change 135

done by linking the tabulated data regarding the date of construction of buildingscollected by the DoS as part of its periodic national census counts (approximatelyevery ten years) to a spatial layer containing polygons representing all buildings asmapped out (traced) from the most recent aerial photograph. However, because ofthe inability to identify every single building individually, these data were linked on thebasis of whole statistical blocks that were in turn merged together to form discretemorphological units or areas (figure 7). This became particularly useful when comparingchange in ownership patterns across the study period with the pattern of constructionactivity in the smaller study area for which landownership data were collected.

4.2 Patterns of landownership changeTabulated data summarized from official land registry archives gave the initial directionfor further inquiries. Initial statistical analysis showed that, although some plots did notundergo any ownership change over the whole study period, other plots changed handsup to a maximum of thirteen times. Moreover, the pattern of change exhibited twinpeaks, in the 1950s and in the 1990s, which in general resonates with the peaks ofconstruction activity for the whole period. A very high percentage of change occurredin the 1990s, which is reflected in the rise in construction activity in all Jabal Lweibdehand in BG1 but not in the core area of block 91028, where landownership witnessed asteady percentage change from the 1980s into the 1990s, at a rate of 12.9% (figure 8).

Mapping the spatial distribution of this pattern we find that there is considerablechange in landownership of all plots in the study area, contrary to the situation in thecore of the neighborhood, formed in significant part by block 91028, where only veryfew plots have undergone change in ownership since 1970 (figure 9).

New construction (%)0 ^ 7

8 ^ 13

14 ^ 20

21 ^ 31

32 ^ 54

0 250 500 1000m

Figure 7. Spatial distribution of construction activity in the 1990s for all Department of Statistics(DoS) blocks in Jabal Lweibdeh.

136 N Abu-Dayyeh, F Ziadat

50

45

40

35

30

25

20

15

10

5

0

Cha

nge(%

)

1940s 1950s 1960s 1970s 1980s 1990s

2.60.0

6.0

11.1

20.0

16.1

22.4

25.130.0

40.8

11.2

19.1

25.1

38.7

3.38.9

9.914.4

19.4

0.01.0

6.412.9

14.517.216.212.9

45.0

39.5

1.7

Landownership change in block 91028

Construction activity in block group 1

Construction in block 91028

All landownership change

Construction in all block groups

Decade

Figure 8. Comparison of land-ownership change in Jabal Lweibdah, 1940s ^ 1990s.

Block 91028

Period1946 ^ 591960 ^ 691970 ^ 891990 ^ 2002

0 50 100 200m

Figure 9. Spatial distribution of plot-ownership change in the study area and in block 91028, forselected periods.

GIS for physical and social change 137

It was therefore concluded that the surge in construction in the 1990s in this blockoccurred almost totally without landownership change. In other words, the land wasdeveloped by its owners. Apparently, no gentrification mechanisms were at work here,unlike the case in Jabal Amman, a similar old neighborhood in the city wheregentrification has for some time had a direct and highly visible impact on the neigh-borhood. This perhaps accounts for the variance between the perceived image of block91028 as a main part of a historical neighborhood with a deep-rooted social fabricdespite the recent considerable change to its physical fabric. It seems that peoplebecome much more cognizant of physical change when it is coupled to landownershipchange.

On the basis of the available data, and without extensive fieldwork, it is notpossible to ascertain with absolute certainty the exact causes for the rise in buildingactivity, but it is well known that the Gulf War of 1991 resulted in what has beentermed the `Third Wave of Immigration' to Jordan (Abu-Ghazalah, 1996), resultingin the return of an estimated 300 000 expatriates to Jordan, of whom only 28.4% ownedtheir own residence and only 22.6% owned their own land (DoS, 1993). Furthermore,official sources have estimated the increase in housing demand due to the third wave ofimmigration at around 30 000 housing units (MoP, 1993), a large percentage of which,judging by previous trends, would be needed in residential zones C and D (MGA,1988), such as in the zoning seen in Jabal Lweibdeh.

Furthermore, by studying landownership records, it has been found that 34% ofcurrent ownership of land in the study area dates from the 1990s. Therefore, it isevident that, although landownership in block 91028 did not significantly change handsin the 1990s, this is clearly not the case for the rest of the study area, which witnessed asignificant percentage of landownership change.

The resistance of the core area to gentrification or ownership change suggests apattern of physical regeneration and/or development in the older and once moreaffluent neighborhoods whereby the core area is influenced by the general rise inhousing demand in the neighborhood and the city at large but remains tightly heldby the original owners. The same does not apply, however, even within the immediatesurroundings, where the demand for development, among other factors, seems tohave resulted in physical development significantly related to landownership change.It seems that in the more historical and once affluent neighborhoods the center orcore area displays a high margin of resistance to ownership change even in periodscharacterized by higher demand for development.

5 Concluding remarksRecently, there have been several projects utilizing GIS for a variety of uses in thefields of physical planning and management in Jordan and in neighboring countries.The development has been fragmented and uneven. For example, a recent reviewof the developing world showed that Qatar has been at the forefront of geographicinformation technology in the developing world, with GIS initiatives beginning in 1988.In contrast, development of GIS in India has not been proportionate to the morefavorable circumstances in this country, including the Great Survey of India, which datesto the 18th century and which covered the entire country with rigorous cartographicsurveys, and Indian Remote Sensing (IRS), which provides high-resolution remotesensing data to global markets through a series of satellites (Montgomery et al, 2002).

In Egypt, GIS has been used by the Supreme Council of Antiquities for thecreation of the Egyptian Antiquities Information System. A pilot study of that projectconcerned Ottoman Alexandria. Meanwhile, the National Center for the Documenta-tion of Egyptian Cultural and Natural Heritage has created its own GIS project for the

138 N Abu-Dayyeh, F Ziadat

documentation and protection of architectural heritage in the center of the city ofCairo. Furthermore, in 1997 the General Organization for Physical Planning solicitedthe URBAMA laboratory at the University Francois-Rabelais in Tours, France, forhelp in defining a strategy for the protection of urban heritage in Greater Cairo, aninitial study project has been launched for the study, with detailed documentation ofarchitectural heritage in Heliopolis, a town situated 10 km from Cairo, founded in 1906ex nihilo by Belgian and French capital. The study is concerned with a 280 ha areacontaining some 2000 buildings, which constitute the historical heart of the old town(Volait and Piaton, 2003).

In Syria GIS has been extensively used for the conservation and management ofthe city of Aleppo and the conservation of the old city of Damascus (the walled city);in Lebanon it has been used by research centers such as the Observatoire de Recherchesur Beyrouth et la Reconstruction and by a German research group affiliated with theGerman Institute for Oriental Studies for the study of Zuqaq al-Blat, an old neighbor-hood of Beirut. In Jordan, the Water Authority has employed GIS technology torestructure the water supply network in Amman, and another standalone system wasdeveloped by the Department of Antiquities at the Ministry of Tourism and Antiquitiesin collaboration with the American Center for Oriental Research for documentingand safeguarding Jordan's antiquities, namely, the Jordan Antiquities Database andInformation System. Added to that is the ongoing project by the Municipality ofGreater Amman for the mapping of all data related to urban management andplanning. There is also the Jerusalem Database Project of the Royal Jordanian Scien-tific Society and the Harvard Graduate School of Design, aimed at documenting theurban history of the old city mainly through Ottoman court records, and the JordanArid Zone Productivity Project, aimed at land-use planning of arid and semi-arid areasin Jordan.

In all these projects, GIS has proven a valuable tool for archiving, documenting,and managing cultural and other resources, but its potential for understanding relation-ships between physical and social change has yet to be fully exploited. The researchpresented here utilizes available data that are relatively easily accessible; however, in thefuture, to make the best use of the available data and to utilize the full potential of GIS,such data should be linked to individual buildings or plots and not simply to statisticalblock areas. This would allow the addition of more specific data to be collected fromthe field about particular owners, building typologies, and so on, in order to obtain abetter understanding of processes of urban change. This should become possible ina few years, that is, once the MGA and the DoS finish their ongoing projects fordigitizing and computerizing spatial and temporal data on land and buildings and ontheir most relevant attributesöfor Amman in the case of the MGA, and for the wholeof Jordan in the case of the DoS. However, it is likely that the linking of data toindividual plots would still be restricted; therefore, the method proposed here may stillbe of benefit for the undertaking of further, more detailed, research.

The particular value of this method is the relative ease with which it can beapplied to other areas of the city, which would then begin to give us a betterview and understanding of the physical transformations in different parts of thecity and subsequently allow us to define varying sets of issues that would in turninvite further investigation, which could eventually give us a better overall under-standing of the spatial and temporal dimensions of change at the more general levelof the city.

GIS for physical and social change 139

Acknowledgements.We would like to acknowledge the financial support received from the Universityof Jordan's Deanship of Scientific Research. Eric Verdeil, Director of the Observatoire de Recherchesur Beyrouth et la Reconstruction, for encouragement and general discussions on various aspectsof physical urban development; Aziz Hallaj, for information about the use of GIS in the con-servation and urban management of Aleppo; the Royal Jordanian Scientific Society, for informa-tion about the Jerusalem Project; the German Institute for Oriental Studies and the Municipalityof Tripoli, for providing a forum for discussing issues related to GIS and its applications tothe documentation and preservation of architecture during the workshop ``Re-evaluating UrbanHeritage'', held at Tripoli, Lebanon, 11 ^ 14 December 2002; and the anonymous reviewers of thispaper.

ReferencesAbu-Ghazalah S, 1996, ` The impact of the Gulf War 1991 upon the housing sector and strategies in

Jordan''Open House International 21(2) 71 ^ 80Alshuwaikhat H M, Nkwenti D, 2002, ` Visualizing decisionmaking: perspectives on collaborative

and participative approach to sustainable urban planning and management'' Environment andPlanning B: Planning and Design 29 513 ^ 531

Batty M, 1996, ` Urban change'' Environment and Planning B: Planning and Design 23 513 ^ 514Batty M, 2002, ` Thinking about cities as spatial events''Environment and Planning B: Planning and

Design 29 1 ^ 2Benenson I, Omer I, Hatna E, 2002, ` Entity-based modeling or urban residential dynamics: the

case of Yaffo, Tel Aviv'' Environment and Planning B: Planning and Design 29 491 ^ 512DoS, Department of Statistics, Public Relations, PO Box 2015, Amman, Hashemite Kingdom of

Jordan1993, ``Study of Jordanian Returnees between August 10, 1991 and December 31, 1992''1997, ` Results of the General Census of Population and Housing of Jordan 1994: Volume 1,

Buildings and Housing Characteristics''2001, ` Jordan in Figures 2000''

Geertman S, 2002, ` Participatory planning and GIS: a PSS to bridge the gap'' Environment andPlanning B: Planning and Design 29 21 ^ 35

Hacker J, 1960, `Amman: a social study'', RP-3, Department of Geography, University of Durham,Durham

Harvey D, 1990 The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change(Blackwell, Oxford)

Jarrar S, 2002, ` Jabal Lweibdeh and cultural identity''Al-Rai 6 September, page 27King G, 1955, ` The Final Report of the United Nations Field Town Planner, Jordan'', edited

and approved by Max Lock, copy may be viewed at the Max Lock Archive, University ofWestminster, London

Koster E,1996, ` GIS applications in town planning history research'', http://odin.let.rug.nl/�ekoster/morfo/nijm eng.html

MGA, 1988, ``Greater Amman Comprehensive Development Plan'', 7 volumes, Municipality ofGreater Amman, and Dar al-Handasah Consultants (Shair & Partners); copy may be viewedat the Technical Library of the Municipality of Greater Amman

Montgomery M, Stren R, Cohen B, Reed H E, Montgomery M R, (Eds), 2004 Cities Transformed:Demographic Change and Its Implications in the DevelopingWorld (National Academy ofSciences; reprinted by Earthscan Publications, London)

MoP, 1993, ` Plan for Economic and Social Development, 1993 ^ 1997'', Ministry of Planning,Amman, Jordan, http://www.mop.gov.jo

Moudon A, 1994, ` getting to know the built landscape: typomorphology'', in Ordering Space:Typesin Architecture and Design Eds K Frank, L Schneekloth (Van Nostrand, NewYork) page 308

Scheer B, Ferdelman D, 1997, ` Morphology as destiny: the story of Over-the-Rhine'', paperpresented at the International Seminar on Urban Form, Birmingham,http://www.gees.bham.ac.uk/research/umrg

VolaitM, PiatonC, 2003,` L'identification d'un ensemble urbain duXXe© me sie© cle en Eè gypte: Heliopolis,Le Caire'', http://www.culture.gouv.fr/culture/inventai/extranet/revue/insitu3/d2/text/d2.html

ß 2005 a Pion publication printed in Great Britain

140 N Abu-Dayyeh, F Ziadat