Resident Perceptions and Expectations of Rooftop Gardens in Singapore

Landscape and Urban Planning 73 (2005) 263–276

Resident perceptions and expectations of rooftopgardens in Singapore

Belinda Yuena,∗, Wong Nyuk Hienb

a Department of Real Estate, National University of Singapore, 4 Architecture Drive, Singaporeb Department of Building, National University of Singapore, 4 Architecture Drive, Singapore

Received 7 April 2004; received in revised form 11 August 2004; accepted 12 August 2004Available online 14 October 2004

Abstract

Using data from the high-rise, high-density city of Singapore, this paper examines residents’ perceptions and expectations ofrooftop gardens in Singapore. In particular, it discusses: To what extent is the roof gardens being used? Why do local residentsvisit roof gardens? What benefits do local residents perceive? The discussion has implications for the potential of rooftopgreening, a new but increasingly common addition of open space in today’s growing urban areas. With the continuing trendtowards urbanization and taller buildings, rooftop spaces stand to offer residents doorstep green oasis in the skyline of high-risebuildings.© 2004 Elsevier B.V. All rights reserved.

ordenaceson

haset-

s-pendofs-

Keywords: Urban greening; Roof gardens; Residents’ perceptions; Singapore

1. Introduction

With international focus on sustainable develop-ment, interest in the role of open space in the urbanenvironment has increased in recent decades (Gehl andGemzoe, 2001; Hollis and Fulton, 2002). Policies forcreating and preserving open space are increasinglyimplemented at a variety of urban densities and spatialscales. Many cities across America have created greennetworks, green belts and programs to protect openspace and restore connections to nature (Gustanski and

∗ Corresponding author. Tel.: +65 6874 3418; fax: +65 6451 3989.E-mail address: [email protected] (B. Yuen).

Squires, 2000; Pendall et al., 2002). In Britain, theUrban Task Force on urban renaissance under LRogers has highlighted the need to provide new opspace systems and better access to existing open spin the face of the continuing trend towards urbanizati(Rogers and Urban Task Force, 1999). The search fornew open spaces in today’s growing urban areascaused some cities to look skywards and create a nwork of ‘skyrise’ or rooftop greenery.

While the phenomenon may be relatively less etablished in North America, many cities across Euro(especially in Germany, France, Austria, Norway aSwitzerland) and Asia have started to green the roof buildings. By 1996, one in ten flat roofs in Ger

0169-2046/$20.00 © 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.landurbplan.2004.08.001

264 B. Yuen, W. Nyuk Hien / Landscape and Urban Planning 73 (2005) 263–276

man cities was greened while 70% of flat roofed in-ner city buildings in Swiss cities are reported to haveroof gardens (Kuhn, 1996; Peck et al., 1999). In Asia,roof gardens are assiduously promoted in Japan, HongKong and Singapore. Tokyo government, for exam-ple, has passed a decree in April 2000 requiring allnew buildings with a floor area of 1000 m2 or larger toplant trees (namely, conifers and cedars) on rooftops(Environment Preservation Bureau Tokyo Metropoli-tan City, 1999). In Singapore, rooftop greening hasgained the support of the government. As announced bythe Prime Minister in June 2001, ‘To demonstrate theGovernment’s commitment to high-rise greenery, thepublic sector will take the lead in implementing skydeck greenery in our public housing estates and pub-lic buildings’ (Singapore Government Press Release,Speech by Prime Minister Goh Chok Tong, 26 June2001).

These gardens are usually provided with paving,seating and deep layers of substrates for garden land-scaping. Incorporated into the building and its archi-tecture, roof gardens are designed to be accessible tothe people. Are they? Some such asOng (2003, p. 201)have begun to question their role in the urban environ-ment, ‘Is a roof garden equivalent to a similar sized gar-den on the ground? Is a high-rise building enwrappedby greenery on its roof as well as facades as effective asa mature tree of similar size?’ Their presence adds to theconcern to provide evidence of the use and role of roofgardens in the pattern of urban living. AsRogers andU c-i orkb eent d it.’(

lessa reast pa-p gar-d dy ofi on-t is-c d res-i in-d partt for-w pens first

glance, this may seem a parochial issue but increas-ingly under the pressure of urbanization, cities haveto explore forms of urban greening beyond groundlevel in the new urban concentrations of high-risebuildings.

2. Greening the sky

In many cities in the western world, there existsa park tradition which in the words of Olmsted is toprovide ‘natural, verdant and sylvan scenery for therefreshment of town-strained men, women and chil-dren’ (cited inOlmsted and Kimball, 1970, p. 523). Therepertory of forms is largely ground or street level gar-dens, greenways, parks and open space systems (Cranz,1982; Floyd, 1991). However, with the continuing trendtowards urbanization internationally (nearly half ofthe world’s population is living in urban areas) andthe rapid construction of ever taller residential build-ings in cities (London, Singapore and many othercities are building 50-storey housing to meet hous-ing needs), urban greenery beyond the ground levelcannot be overlooked.Thompson (2002)in review-ing what should be demanded from urban open spacein the 21st century has shown the need for new pat-terns of open space networks within our changing urbanfabric.

One aspect of the changing urban form as seen inmany American cities is the growth of urban decentral-i rbanc su-a low-d form theo g re-v edu nc-tu pens dU ket( me-w n in-d wns ontp d in-

rban Task Force (1999)remind in the challenge fang British open space planning, ‘Public spaces west when they establish a direct relationship betw

he space and the people who live and work arounp. 57).

While research on urban parks exists, muchttention is being paid to the small-scale green a

hat are close to where people live and work. Thiser takes a closer look at one such area, the roofens, its usage and potential, through a case stu

ts development in Singapore. Our intention is to cribute to the evolving theoretical and empirical dussions on this subject as we seek to understan

dents‘perception of roof gardens, uncovering theividual’s motive to visit and experience, teasing a

he roof gardens experience in practice and puttingard recommendations for future planning of this opace form in the specific case study context. At

zation and development of relatively dense subulusters of commercial and residential activities (ully townhouses and apartment blocks rather thanensity single-family housing that have prevaileduch of the mid-20th century). In most situations,ld central business districts are themselves beinitalized and redeveloped with tall buildings for mixse: office, entertainment, retail and residential fu

ions (Attoe and Logan, 1989; Lang, 1994). With therban form changes, the character and form of opace, asPrice (1987)and more recently,Rogers anrban Task Force (1999)argue, is changing also. Ta

he case of Portland, Oregon, for example. AsPrice1987)traces, the change has resulted in a larger fraork of open spaces that encompasses more thaividual parks and includes old and new downtoquares, in-city and city fringe greenery of riverfrromenades and nature reserves, street level an

B. Yuen, W. Nyuk Hien / Landscape and Urban Planning 73 (2005) 263–276 265

the-sky spaces of pedestrian malls, atriums and roofgardens.

In Singapore, the pressure of urbanization has pro-vided impetus for intense urban concentrations and theskyscraper to be the most ubiquitous development type(Wong and Yeh, 1985; Perry et al., 1997). Singapore isan epitome of a fast growing city or asKoolhaas andMau (1995)describe, ‘a built answer to the shift fromcountry to city which was thought, 30 years ago, toforce Asia to construct in 20 years the same amount ofurban substance as the whole of Western Europe’ (p.1017). In the city center, the tallest commercial build-ings rise to more than 60 stories. In the residential arena,the majority (90%) of the 3.4 million resident popu-lation lives in high-rise apartments: 84% in publiclybuilt housing and 6% in private apartments and condo-miniums. Under the country’s latest long term ConceptPlan 2001, more are expected to live on higher floorsas the population grows to a projected 5.5 million in40–50 years’ time. The private and public housing sec-tors have respectively announced plans to build higherthan the present tallest 30-storey residential blocks (to40- and 50-storey). Despite the trend towards moreurban-style development, the focus is not just on ur-ban development but also a quality urban environment.Consideration of quality of life has become salient asglobalization gathers momentum in Singapore (UrbanRedevelopment Authority, 1991).

Much has been said elsewhere about open spacecontributions to urban quality of life (for example,R Liv-a 3R thatg andm reesa uc-ieKd lar,t gt en-eo -t memp bana iritu-

ally through periodic contact with nature.’ InKnopf’s(1987)analysis, there are four broad dimensions to thiscontact: nature as restorer, as competence builder, assymbol and as diversion. The challenge to planners ishow to increase access to greenery in the urban area asbuildings go taller and pressure on urban land increases.Research on the use of urban parks verifies that the ma-jority of users are visitors on foot (Kaplan and Kaplan,1989; Godbey et al., 1992). These findings broadly sup-port the earlier postulation ofAlexander et al. (1977)that people will visit urban greenery on a regular ba-sis if it is within 3–5 min walk of their home/workplace.

In Singapore, accessibility is a prime considerationand its planners are putting forth an even greener Sin-gapore with a greater variety of parks and open spacesand target of 0.75 ha of parkland per 1000 population(currently 0.6 ha) in their vision to create a ‘dynamic,distinctive and delightful’ world class city in the 21stcentury (Urban Redevelopment Authority, 2001). Themain strategies of urban greening include:

• having more parks and gardens including ‘gardensin the sky’;

• safeguarding the natural heritage; and• making the parks and gardens more accessible.

The approach is to provide ‘seamless greenery’through a network of park connectors and opportunitiesfor greenery to occur beyond ground level. Essentiallyn lingt andn ma-r

O de-v ark-l o thep sivea rget tors,b tiblel odso ob-j eda turalr h forf

ogers and Urban Task Force, 1999; Partners forble Communities, 2000; Van Kamp et al., 200).esearch across different disciplines has shownreenery directly benefits the urban environmentakes the city livable. The arguments are familiar: tnd vegetation may improve urban climate by red

ng carbon dioxide and cooling the heat island (Dwyert al., 1992). Other researchers, includingKaplan andaplan (1989)and Ulrich and Simons (1991)haverawn attention to the human benefits. In particu

he work ofUlrich (1984)is widely cited as illustratinhe restorative health benefits that window view grery can have on patient recovery. ToMiller (1986)andthers such asThompson (2002), the desire for con

act with nature will only increase as people becoore urban in their way of living. AsMiller (1986,. 18)elaborates, ‘Most of their lives are spent in urreas at urban tasks, but they renew themselves sp

ature trails with resting places, jogging and cycracks, the park connectors will connect all parksature areas to form a continuous corridor. As sumized byOi (1998, p. 31),

ne of the main objectives of the connectors is toelop a network of connecting parks so that the pands in Singapore are made more accessible tublic. As land in Singapore becomes more expennd scarce it will be increasingly difficult to obtain la

racts of land for park developments. Park Connecy optimising drainage reserves and other compa

and uses, will be one of the most cost effective methf providing recreational venues. Another major

ective is to provide natural corridors within urbanisreas through which birds can move from one naeserve or refuge to another thus aiding their searcood and breeding sites.


Fig. 1. Pedestrian overhead bridge is provided with planter boxes toadd color to roadside greenery.

Since the early 1990s, 40 km of park connectors hasbeen built. This will be increased to 120 km and whenfully completed, Singapore will be converted into ‘a to-tal playground’ (Lee and Chua, 1992; Urban Redevel-opment Authority, 2002). Opportunity for additionalgreenery is not limited to the spatial level but increas-ingly extended skywards with the buildings to includethe vertical dimension. A key aspect of Singapore’sskyrise greenery is the greening of building surfacesand roofs. Flyover, overhead pedestrian bridges, retain-ing walls and other concrete urban structures are in-creasingly covered with creepers such asFicus pumilaand climbers (Fig. 1). The guiding principle, as theSingapore National Parks Board, the state authorityresponsible for Singapore’s garden city development,shares in its Annual Report 2000, is not only to greenthe land (and provide more new parks, enlarged and im-proved favorite parks, extended network of park con-nectors to link park to park and roadside greenery)but also to move the greenery vertically skywards torooftops and closer to individual homes in the tall apart-ment blocks.

Since the late 1990s, Singapore’s national planningauthority (Urban Redevelopment Authority) has ac-tively facilitated the usage of unused flat roofs as roofgardens. It has reviewed development control guide-lines to exclude the space used for skybridges, sky ter-races and communal open pavilions in roof gardensfrom overall gross floor area computations and relaxeddevelopment parameters to promote greater flexibility

in the design of sky gardens (Urban RedevelopmentAuthority, 2003). In the words of theUrban Redevel-opment Authority (2003, p. 2), ‘In the next phase ofmaking Singapore aCity in the Garden, moving green-ery upwards into the sky and closer to homes is theway to go.’ The motivation for skyrise greenery maybe traced to the growing interest in the environmen-tal benefits of roof gardens including rainwater har-vesting and savings in energy consumption. These areimportant considerations in Singapore’s resource man-agement especially when it has to import much of itsenergy and fresh water (The Straits Times, 4 July 2001).

According toLook Japan (2000), the temperaturein Tokyo could be lowered by 0.11–0.84◦C if 50% ofall available rooftop space were planted with green-ery. This would lead to a savings of approximately 100million yen per day in the city’s electricity bill. In addi-tion, other city studies have contended that roof gardensoffer usable spaces (Willmert, 2000), screen rooftopservices and provide visual relief from dense urban-ity (Dwyer et al., 1994; Johnston and Newton, 1996).Faced with the high premium of open space at streetlevel, roof gardens present the advantage of convenientalternative green spaces in high-rise living. However,persuasions of significance and provision would notnecessarily lead to usage. AsJacobs (1961, p. 100)per-ceptively observed, ‘people do not use open space justbecause it is there and because city planners or design-ers wish they would.’ Currently there is no legislationin Singapore requiring roof gardens to be constructedo

ingp cts.T icalk cia-t ofs.P tantt maye nusedb r re-c thatt ar-d sersw am nts’p un-a otr rban

n new developments.There is seeming reluctance among local build

rofessionals to include green roof in their projehe main reasons include not just the lack of technnow-how but also concern over a lack of appreion of the benefits and performance of green rout simply, many building professionals are reluc

o include roof gardens for fear that these gardensnd up as redundant spaces unappreciated and uy the residents. The general response from ouent survey of architects and developers indicateshe future development and incorporation of roof gens would depend largely on whether the end-uould appreciate it (Wong, 2002). Thus, it is not justatter of academic interest if we can clarify resideerceptions and usage of roof gardens. It would bevailing for any legislation if the general public is neceptive to roof gardens. Even though who uses u


open spaces and why is a topic that has received wideresearch attention in the recreation literature, little isknown about roof gardens usage (National ResearchCouncil of Canada, 2000). Being a new developmentin Singapore no prior analysis has been made on roofgardens.

This paper draws from a larger first study intorooftop gardens in Singapore to understand their ef-fects and usefulness (Wong, 2002). In the study, tech-nical (thermal measurement, energy simulation andbuilding life cycle cost) and people (public attitudeand usage rate) issues were explored. The main in-terest driving the people analysis was to documentthe usage rate of rooftop gardens and residents’ per-ception of the facilities and design desired in theroof gardens to develop design guidelines relatingto the effective design and implementation of roofgardens. This paper shares the findings from sam-pled residents living in Choa Chu Kang new town, anew generation public housing town, to the follow-ing issues: To what extent is the roof gardens beingused? Why do local residents visit roof gardens? Whatbenefits do local residents perceive? The discussionhas implications for the potential of rooftop greeningand planning in tune with public needs and expecta-tions.

3. Understanding community perception

oofg hata d ande wellg en-v up-p tod ng,1 ora elfc ofm nf( re ofo e mosb tiont tion’i seen’

(Rapoport, 1977). It reflects motives, preferences andattitude to space.

Survey work was carried out duringSeptember–October 2001 in the new town ofChoa Chu Kang. Choa Chu Kang new town is a newgeneration 1990s public housing town in Singapore.Located in the northwest of Singapore, Choa Chu Kangnew town has a land area of 583 ha, of which 307 ha isfor residential development. Typical of public housingtowns, the gross plot ratio for residential developmentis 3.0–4.5. At the time of study, the resident populationwas about 128,400 in 38,126 dwelling units (Housingand Development Board Annual Report, 2000/2001).When completely developed, Choa Chu Kang townwill have a total of 62,000 dwelling units. Publichousing towns form the main residential areas inSingapore where 84% of the 3.4 million residentpopulation lives.

With a population of 200,000–300,000 each, Sin-gapore’s public housing towns are typically organizedwithin the framework of neighborhoods and precinctsand provided with a spatial hierarchy of cradle-to-gravefacilities including parks. There is no fee to usage of theparks. There are also no gates, security guards or othermechanisms to prevent outsiders from entering thetown, neighborhoods and parks. Open space and sportsfacilities generally take up about 4% (approximately36 ha) of new town land. The new town open space hier-archy typically includes one town garden (about 5 ha),several neighborhood parks (one per neighborhood,e llingu 000d ro-v undl aceso

lvedta 03).I newt , ther eseb pedi

py-i mT ow-e gs

Understanding the resident’s perception of rardens is critical to arriving at design criteria tre based on how these spaces are really usexperienced by the people. It is an approachrounded in the line of research of geography,ironmental psychology and urban design that sorts the building of cities where people desirewell (Lynch, 1960; Relph, 1976; Ittelson, 1978; La994). Relph (1976), for example, has called fn approach to urban planning ‘that is wholly sonscious. . . that is responsive to local structureseaning and experience. . . that takes its inspiratio

rom the existential significance of place. . .’ As Whyte1988)contends, people are the most potent featupen spaces. They choose spaces that promise thenefits to them. Implicit in this approach is the no

hat perception precedes action. The term ‘perceps used here in the general sense of how things are ‘

t

ach of 1–1.5 ha serving about 5000–6000 dwenits), precinct gardens (0.2 ha, one per 1000–2welling units) and green connector (generally pided along drainage reserve), all of which are groevel public spaces where the largest of these spffer the most facilities.

Over the years, public housing towns have evoo emphasize quality living (Perry et al., 1997; Housingnd Development Board Annual Report, 2002/20

n more recent developments like Choa Chu Kangown, in addition to the street level open spacesooftop spaces of multi-storey car park buildings (thuildings have flat roofs) are increasingly landsca

nto roof gardens (Fig. 2).Typically these are intensive roof gardens occu

ng the roof space (an area of approximately 5502).he car parks (4-storey high) are generally the lst block in the precinct. They are public buildin


Fig. 2. Multi-storey car park buildings in public housing towns areoften built with a flat roof to offer opportunities for greening.

open 24 h, well-lit at night and accessible to the pub-lic and residents of the surrounding taller housingblocks by staircases. Covered walkways provide all-weather shelter from the car park blocks to the apart-ment buildings (Fig. 3). As with the other publicopen spaces in the new town, the roof gardens con-tain several publicly provided amenities: footpaths,benches, shelter/pergola, paving and largely manicuredgarden landscaping, for group or individual activities(Fig. 4).

To enhance the internal validity of research findingsthe mixed method qualitative/quantitative sequentialdesign was used. First, focus group discussion and pilot

Fig. 3. Covered walkway provides linkage from the roof garden tot

Fig. 4. Seats and shrubbery are some standard provision of roof gar-dens on multi-storey car park buildings.

depth interviews with residents were used to develop,pre-test and refine parts of the survey instrument. Thegroup discussion was convened with the help of the lo-cal residents’ committee chairman. It primarily focusedon having the small group of seven resident respondentswho have all visited roof gardens to describe in detailtheir usage, perceived benefits and attitude towards roofgardens provision including views on open space pro-vision in the neighborhood. The discussion conductedin English was preceded by a walk through the nearbyroof gardens and after the discussion the ground planeopen spaces in the neighborhood. It was taped with thegroup’s consent. The tape and notes taken during thediscussion were later transcribed for analysis. Then, ahousehold survey was conducted on a random stratifiedsample of 333 residents comprising households livingnear (<5 min walk) and away (≥20 min walk) from theroof gardens and closely reflecting Singapore’s demo-graphic profile (Table 1).

Residents were interviewed in their homes. This sur-vey method allowed the inclusion of infrequent, off-siteand non-users who may otherwise have been missed inan on-site visitor survey. A structured interview for-mat (included multiple choice, closed and open-endedquestions) was used to guide the interview, each lastingabout 30–45 min. Even though perception is an exceed-ingly complex phenomenon,Shafer (1969)has demon-strated that the direct approach of asking a sample ofpeople how they feel about the natural environmentis one way to measure perception. Full details of the
he public housing apartment block.


Table 1Demographic characteristics of survey respondents

Characteristics Choa Chu Kang new town (%)a Singapore (%)b

Households living nearroof garden (n = 186)

Households living away fromroof garden (n = 147)

EthnicityChinese 79.5 83 76.8Malay 15.7 11.6 13.9Indian 4.3 4.1 7.9Others 0.5 1.3 1.4

GenderMale 42.4 49.6 49Female 57.6 50.4 51

Age (years)16–24 9.1 21.125–34 31.7 29.935–54 55.4 44.955–65 3.8 3.4

a Survey data.b Census of Population 2000.

present study background and methodology are givenin Wong (2002). In this paper, the data on the usagelevel and perceived benefits will be presented.

4. Results and discussion

Three perspectives have emerged from the analysisand interpretation of the data obtained: first, a smallproportion of respondents had visited the roof gardensat the present time; second, most would like to see moreroof gardens provided; and third, they see several bene-fits of roof gardens provision in their neighborhood andthe city that have some implications for this landscapetypology in open space planning.

4.1. Level of usage and awareness

The survey revealed a generally low usage rate andhigh awareness level and willingness to use roof gar-dens. Only 10–20% of respondents had visited roofgardens with higher visitation among those living nearto a roof gardens (Table 2).

Among the households living near the roof gardens,the older age group of 35–54 years old (47.7%) hadtended to visit the roof gardens. More than 70% ofthose respondents had family monthly incomes of S$

1000–4999 (median monthly household income fromwork was S$ 3607 in 2000) and 47.7% had gone tothe roof gardens alone (the rest were with children). Itwas also found that more men (52.3%) had visited theroof gardens. When we examined the data from house-holds living away from the roof gardens, we foundthat a slightly larger proportion (43.8%) of the visitwas made by those in the younger age group, 25–35years old, 62.5% of respondents had family incomesof US$ 1000–4999. Male visitors again predominated(62.5%). While the proportion of single person visitremained high (43.8%), more (56.2%) had visited withtheir children. On the awareness and willingness to useroof gardens in their neighborhood, over 90% of thoseinterviewed claimed to know of the existence of roof

Table 2Usage of roof garden

Respondents (n) Visited roofgarden (%)

Not visitedroof garden(%)

Do not knowwhere/what is a roofgarden (%)

All households(333)

17.7 76.9 5.4

Household livingnear roof garden(186)

23.6 73.7 2.7

Household livingaway from roofgarden (147)

10.5 81 8.5


gardens and over 84% responded ‘yes’ to the question:would you use rooftop gardens if they are providedin your neighbourhood? (households living near roofgardens registered a percentile of 87 to ‘yes’).

On closer analysis, we found that even though manymay not have visited a roof garden, most sampledresidents, 40–65% of respondents, had regularly vis-ited nearby parks other than roof gardens with anotherone-third going to other public parks. The park userswere from all walks of profession, education, income,gender and age. The indication is that some parks at-tract much greater use than others. Not everyone is at-tracted to roof gardens, not that everyone should be.In a more recent study,Tan (2003/2004)reports thepopularity of visiting shopping centers and cinemas asrecreation choices in Singapore over parks and gar-dens. Other international studies relating to the use ofpublic parks have stated the influence of distance fromhome and between competing facilities (Mercer, 1977;Van Herzele, 2003), attraction of familiar places andamenity value (Burgess et al., 1988) and positive parkimage (Hayward and Weitzer, 1984; Grahn, 1991) inpeople’s use of parks.

Part of the reason for the low utilization reported inthe Choa Chu Kang case may be access and visibilitywhich were mentioned as reasons why more use is notmade of the roof gardens. Unlike ground level parks,access to roof gardens in the study site is at presentthrough staircases only, which demands a certain levelof physical fitness and local knowledge of the locationo idualp com-m sup-p re’.‘ lo-c s ofh retr isitt

dt ple’sca gei ivid-u d tog ple’sb eive

in an environment, even though the number involvedis quite small, the implication derived from our findingwould seem to suggest that roof gardens at the presenttime may not occupy a particularly vital part of mostrespondents’ perceived recreation space.

4.2. A sense of motive and purpose

Usage amount by itself is often limited in informa-tion about the qualitative aspects of why people visitthe roof gardens and the benefits of provision. Our cen-tral assertion is that the value and benefits of roof gar-dens must grow out of an understanding of why peoplego to such spaces and what they mean to their usersover time. Information on people’s motives to visit re-flects the demands and needs they expect to be fulfilled(Burgess et al., 1988; Carr et al., 1992). From users’reasons for visiting the particular roof gardens, it ap-pears that respondents were attracted to roof gardensfor a variety of purposes. As summarized inTable 3,‘to take the children out’ was a frequently mentionedmotive to visit the roof gardens. For those living nearthe roof gardens, ‘to get some exercise’ presented an-other principal reason of visit while the reason ‘to findpeace and quiet’ ranks high among those living awayfrom a roof gardens.

‘To find peace and quiet’ signifies personal bene-fits and asBurgess et al. (1988), Kaplan and Kaplan(1989)and many others have argued express the de-sire for escape or to ‘take a break’ from the noise ands f themo pon-d owo Thev allyb gar-d 3).T por-t iesp rs ap It iss l andp osel oxi-m e ofr vities( g or

f those staircases. Part of the reason point to indivreferences as summed up by those respondents’ents on why more roof gardens should not beorted: they are not practical, ‘no one would go the

Roof gardens is too hot for comfort’ (Singaporeated 1◦N of the equator records day temperatureigh 20–30◦C) and ‘nothing there fancies me’ we

wo major reasons found byTan (2003/2004)on whyesidents in another public housing town did not vhe roof gardens.

Burgess et al. (1988)andGrahn (1991)have arguehat landscapes in the minds can influence peohoice of recreation places. Others such asHaywardnd Weitzer (1984)have demonstrated that park ima

s a substantial and consistent predictor of an indal’s decision to use a park. A positive image can leareater usage. If we assume for a moment that peoehavior patterns may be related to what they perc

tress of urban living. The search for peace is one oore striking conclusions inWalpole’s (2000)findingn British park users where many of the survey resents reported visiting the public park for privacy nften unavailable in the media-dominated homes.alue of ‘park as retreat’ is one that has traditioneen advanced as an argument in support of publicen provision in the city (Olmsted and Kimball, 197he other popular attraction is the affordance of op

unities for physical activity and usage of the facilitrovided. In this perspective, the roof gardens offelace to go to rather than a place to get away from.een as a place for recreation, a setting for sociahysical activity, specifically for exercise among th

iving near a roof gardens. The convenience of pral facility seems to be a strong primer for active us

oof gardens. Some users may seek specific actisuch as to exercise or to sit in the garden) hopin


Table 3Reasons for visiting roof garden

Reason for visit Households with visit toroof garden (n = 58) (%)

Household living nearroof garden (n = 43)a (%)

Household living away fromroof garden (n = 15)a (%)

To take the children out 29.3 32.6 20To get some exercise 20.7 27.9 0Because friends were visiting 13.8 18.5 0To find peace and quietness 12.1 7 26.7To enjoy the company of oth-ers

8.6 7 13.3

To get out of the house 6.9 2.3 20Other reasons e.g. to smoke,happen to be there

6.9 4.7 13.3

Do not know 1.7 0 6.7

a n includes only households who had visited roof gardens.

certain that particular provision will be available in theplace. Evidence of the functionality element appearsyet again in respondents’ articulations of the benefitsthey think the provision of roof gardens will bring tothem and their family/colleagues and the several sug-gestions respondents offered to improve the usage ofroof spaces.

Emphasis is on place usage: 65% of respondentsmentioned usage of the roof gardens as a setting foractivities as a benefit of its provision. They suggestedusing the rooftop space for exercise routines and land-scaping and would like to see more amenity facilitiessuch as barbeque pit, sale of drinks and snack areaprovided in roof gardens (these are presently not pro-vided). Currently, outdoor exercise equipment is gen-erally only provided in the larger neighborhood andtown parks. These data seem to illustrate that activityresources and facility development may attract a cer-tain amount of usage. It is tempting to suggest thatplanners seeking to increase the usage of roof gardensshould consider adding facilities. However, it is impor-tant to note that these data represent associations only,not demand and it must be tempered with a review ofthe entire open space system including the intendedpurpose of roof gardens.

The findings have indicated that the roof garden isbut one in a range of open spaces visited. Just what form(and how useful) this open space should take present thechallenge. The various reasons for visits point to the va-riety of functions that roof gardens can serve.Burgesse gso hatp ty of

park facilities and qualities that in turn encouragemany activities. Interestingly, even as they would liketo see more facilities provided in the roof gardens,several respondents especially those from the smallgroup discussion were quick to caution that the roofgardens should be made unique and different fromthe other ground plane parks and gardens. Roof gar-dens are not needed in every block for fear of los-ing the sense of uniqueness. In addition, they sug-gested such enhancement as including garden statues,water features and special plants as possible pointsof distinction. These suggestions act to create focalpoints and a greater definition of community connect-edness to the space. A sense of place identity seemsimportant.

Acknowledging the suggestions highlights adilemma in roof gardens planning: in the face of everlimited resources and varying suggestions, what shouldactually be provided and whose wants should be sat-isfied? The expressed desire for place identity anduniqueness underscores yet the dynamism and chal-lenge in roof gardens planning that recognizes the dif-ferent needs and aspirations of multiple users and inter-est groups. It reflects a timeless constant in open spaceprovision that has preoccupied many writers (for ex-ample,Jacobs, 1961; Cranz, 1982; Cooper-Marcus andFrancis, 1998). Theorists such asCarr et al. (1992)haveurged planners to think about open space in terms ofcreating benefits and cultivating meaning for their usersand responsiveness to the users’ needs, their lifestylea sid-e : theg rden

t al. (1988)in their study of the popular meaninf urban green in the UK have suggested that weople seek in urban green spaces is a diversi

nd socio-cultural conditions. Central to the conration is the fundamental question of relevancearden has to be relevant to the people. A roof ga


that is deemed irrelevant is unlikely to be well used andcherished by the community.

4.3. A glimpse of relevance

Even though roof gardens may not feature stronglyin the usage arena, there appears to be a general sense ofsupport for roof gardens provision. When asked, “Doyou think we should provide more rooftop gardens inSingapore?” about 80% of the respondents voted formore roof gardens provision. It is possible that somerespondents may have been inclined to indicate ‘yes’ asthey were not asked to bear the provision costs. How-ever, review of their responses and articulations to otherquestions, in particular, their usage and reason for pro-vision, would seem to purport that bias was negligi-ble. Respondents were asked to share their personalthoughts in a follow-up open-ended question of whythey support or do not support more roof gardens provi-sion in Singapore. The replies were content analyzed,similar thoughts and words were sorted and groupedunder same representative themes and then indepen-dently verified to unpack the focus of individual atten-tion. As summarized inTable 4, the main reasons pointto considerations of aesthetics: to beautify the environ-ment and personal usage for leisure and relaxation.

The responses present a glimpse of the relevance ofroof gardens in the participant residents’ life. The roofgardens experience of offering relief to the urban pat-terns and ‘near to home’ space for leisure and recreationr roofg where

TR

R ousehen (n =

f

LBGBS 5BE 3S 2ARO 23

ftop ga

they can bring their children to play, entertain visitingfriends, hold block parties or just be alone. The senseof ‘a place to go’ has been found among the most im-portant benefits of open spaces in cities (Kaplan andKaplan, 1989; Walpole, 2000). A lady from the smallgroup discussion shared her joy of being able to spendsome quiet solitary moments in the nearby roof gar-dens (it is perhaps relevant to mention that the averagepublic flat size in Singapore is 90 m2 for a four-roomflat (three bedrooms) and 110 m2 for a five-room flat(four bedrooms); it is not uncommon for household tolive under the extended family structure). She joinedother respondents who were mothers with young chil-dren to suggest that roof gardens should be providedwith more plants that would become the ‘honeypot’ ofinsects (such as butterfly and caterpillar) where theirchildren could observe nature at close range and de-velop an interest in nature that may otherwise be lostin the congestion of urban life. They were keen to seemore nature and less concrete in the garden.

Besides personal benefits, several respondents alsocited wider city benefits to support the provision of roofgardens. These were largely related to the general ur-ban function and aesthetics of the town: beautify theenvironment, better air quality (this quality seemed es-pecially appreciated among those living near roof gar-dens), greenery and nature views, land optimizationand better utilization of otherwise unutilized space.They form a part of the perceived relevance of roofgardens that goes beyond those derived from the ear-l askedi ro-

esurfaced. Many, especially those living near theardens, appreciated the roof gardens as a space

able 4eason for more roof garden provision

eason for more roof garden provision Hgard

eisure and relaxation 23.4eautify the environment 16.5reenery and nature 12.0etter air quality 7.6aving and better utilization of space .1etter living environment 4.4ntertainment and exercise .2ocialization and neighborly relations .5ccessibility and Convenience 1.3educe temperature 0.6ther reasons e.g. no particular reason, safety .4

a n includes only households who gave reasons for more roo

old living near roof158)a (%)

Household living away from roogarden (n = 114)a (%)

16.721.910.50.9

22.85.31.802.60.9

16.7

rden provision.

ier reported usage rate. Respondents were alsof they see any problem(s) with rooftop gardens p


vision in Singapore. The majority of the respondents(61–70%, the higher figure was from those living awayfrom a roof gardens) did not have a problem withroof gardens provision in Singapore. A small minor-ity (8–12%) highlighted concern over issues of main-tenance (whose responsibility) and insects especiallymosquitoes (4–5%).

Safety consideration was another worry highlightedby some 7–9% of respondents. It was cited by about athird of those households living near a roof gardens whodid not support this provision as a reason for not wel-coming more roof gardens in Singapore (interestingly,all those families had not visited the roof gardens in thevicinity of their flats). This very much reflects people’sperception of the level of risk rather than any true mea-sure of actual risk. Public perception of urban safety isan important pre-requisite for people’s use of the placeand may not be neglected (Lang, 1994). Relating topark settings, several researchers includingBurgess etal. (1988), Nasar and Jones (1997)andMadge (1997)have reminded that fears of danger especially to theperson can restrict park usage.

5. Conclusion

As with many other cities keen on maintaining theirconnections to nature, Singapore has in recent yearsexpanded its urban greening in the vertical dimensionto include rooftop gardens as the city’s buildings be-c orkh tries,t in-g ce oi pledr in at ingp senta incet iderc ntialsY

malld gar-d areu cep-t ence

within their particular high-rise, high density urban sit-uation. However, reflecting on the emerging patterns,it does seem possible to suggest some conclusive com-ments that might point to the usage and potential ofroof gardens in Singapore and the urban setting. Thefirst relates to usage level. Our findings indicate that themajority of respondents had not visited a roof gardeneven though one might be provided in the vicinity oftheir neighborhood. A gap seems to exist between re-spondents’ awareness and usage of gardens presentingan alert for closer analysis of residents’ needs, interestsand knowledge of roof gardens to avoid an importantbut often overlooked phenomenon in urban park pro-vision: non-use.Jacobs (1961)in her benchmark studyof American cities has cautioned that proliferation ofopen spaces in the neighborhood do not necessarilyguarantee usage. It reminds an important first principleof park planning: open spaces that do not meet people’sneeds or that serve no important functions for peopleare destined to be underutilized and by the use criterion,unsuccessful (Burgess et al., 1988).

The second relates to perceptions that may offerglimpses of facilitation to greater usage. The presentstudy clarifies some of the perceptions that people haveof roof gardens which may directly affect how they in-teract with those spaces. Our findings indicate that re-spondents are generally supportive of the provision ofroof gardens. The reasons given broadly support otherstudies that argue for ‘home range location’ of greeneryin the city (for example,Alexander et al., 1977; Kaplana ofr ce ofo reatf nity.T itingr re asa willh s forl ongt roupa isitt

to bea rawt fety)m sulti con-c nce

ome taller. While a great deal more empirical was been done on roof gardens in western coun

here is limited study of this landscape typology in Sapore. This paper attempts to address the absen

nformation by presenting a perspective of the samesidents’ perceptions and usage of roof gardensypical new generation public housing town. Knoweople’s perceptions would be useful for the prend future planning of these gardens particularly s

here are plans to proliferate such spaces. On a wontext, the search for the public’s voice is an essetrategy of sustainable city development (Wakely andou, 2001).

It is hard based upon a single case study and sata set, to make universal conclusions about roofens in the city. It seems probable that the findingsnique to Singapore, reflecting local residents’ per

ions of urban greenery formed over years of experi

f

nd Kaplan, 1989). High on the perceived benefitsoof gardens is the aesthetic pleasure and affordanpportunities for children’s play, recreation and ret

or peace and quiet moments amidst dense urbahese are also the most important motives for vis

oof gardens among those who visited. In Singapolso the case in many other cities, not all residentsave the means and time to travel to faraway place

eisure and recreation. Our data indicated that amhose living near the roof gardens the older age gnd those with lower family income had tended to v

he garden.Just as peoples’ needs and activities (such as

lone, to exercise, to take the children out) may dhem to roof gardens, their concerns (e.g. over saay act as inhibitions which in the extreme may re

n diminished garden use. If not addressed, theerns may work to reinforce a person’s ambivale


about the gardens that may seriously limit their roleand contribution to the quality of life. As Hayward andWeitzer (1984, p. 261) explain, ‘with no current parkuse, there is no opportunity to reinforce the positivebenefits of the experience, and the inertia of non-usecontinues.’ There are many ways to encourage usage.In addition to physical improvement of access and lay-out (e.g. introduce more pedestrian-friendly and visibleaccess), there may be greater community involvementsuch as allocating a part of the garden for neighborsand elderly to plant trees and vegetables (Kaplan andKaplan, 1989). In some of Singapore’s public housingtowns, parts of the ground level parks have in recentyears been ‘adopted’ and planted by the residents un-der a community ‘adopt-a-park’ program. Similar com-munity involvement could be extended to the rooftopspaces. Informing and involving the residents is theroad to connecting people to the roof gardens, whichin Carr et al. (1992)hypothesized model of public spacemeanings would be equivalent to encouraging the de-velopment of its relevance. Relevance is an importantpre-requisite of usage.

The third relates to the role of roof gardens. Unlikeground plane parks, roof gardens admittedly take peo-ple away from the public realm of the street into a moresecluded community setting. But for those living in in-creasingly tall buildings (30- and more storey) our argu-ment is that roof gardens may yet offer doorstep recre-ation spaces and contacts with nature in the verticality.Roof gardens present possibilities for carrying the no-t ingd vel,t pens morec Sin-g esti-m d forh 5.5m hec vatea cleart pens

ciala en-t s tot ucha hers

who are less mobile) will require provision nearer tohome. Reports such asRogers and Urban Task Force(1999)have drawn attention to the ageing demographictrend in the western world and the increasing demandsfor the old and those with mobility problems to be pro-vided with better access to open spaces. Singapore toois faced with an ageing demographic trend that raises is-sues of provision and accessibility to appropriate, localopen spaces in the high-rise living milieu. AsYeang(1999, pp. 106–107)forebodes, ‘Ultimately the newvertical city must seek to re-create up in the sky asmuch as possible those conditions that we find exis-tent in the ground place. For instance, these include there-creation of the continuity of spaces and circulationfound at the ground plane, the provision of immedi-ate access to the natural environment and open spaces,etc.’ With cities growing outwards and upwards in theface of increasing urban growth, growing populationand competing demand on urban land, the roof gar-den is poised to take on a new premium in the urbanlandscape.

Acknowledgements

The authors gratefully acknowledge the assistanceand contributions of colleagues in the Singapore Hous-ing and Development Board, Singapore National ParksBoard, National University of Singapore, our studentsa dy ofR dedb

R

A an-

A ystsley.

B the

C ce.

C trand

C setts

ions of nature and open space further in tall buildevelopment. Firstly, at the macro urban pattern le

he competition and demand for urban land for opace has been such that it is likely to become evenritical as urban populations increase. To illustrate,apore planners have in their long-range planningated that another 14,000 ha of land will be needeousing and industry when its population reachesillion (it only has 12,000 ha of developable land).T

onsequential impact is (re)newed pressure to innond (re)develop existing open spaces. There is a

hread in the urban literature of the pressure on opace (Dower, 1963; Nesbit, 1980).

Secondly, at the local neighborhood level, the sond spatial implications of dense urbanity is as m

ioned earlier that while some may have the meanravel to places away from the city, many more (ss the elderly, the young, the lower income and ot

nd research assistants to the research project: Stuooftop Gardens in Singapore. The project was funy the National Parks Board, Singapore.

eferences

lexander, C., Ishikawa, S., Silverstein, M., 1977. A Pattern Lguage. Oxford University Press, New York.

ttoe, W., Logan, D., 1989. American Urban Architecture: Catalin the Design of Cities. University of California Press, Berke

urgess, J., Harrison, C.M., Limb, M., 1988. People, parks andurban green. Urban Stud. 25, 455–473.

arr, S., Francis, M., Rivlin, L.G., Stone, A.M., 1992. Public SpaCambridge University Press, Cambridge.

ooper-Marcus, C., Francis, C., 1998. People Places. Van NosReinhold, New York.

ranz, G., 1982. The Politics of Park Design. The MassachuInstitute of Technology Press, Cambridge, MA.


Dower, M., 1963. Fourth wave. Architect. J. 141 (3), 122–190.Dwyer, J.F., McPherson, E.G., Shroeder, H.W., Rowntree, R.A.,

1992. Assessing the benefits and costs of the urban forest. J.Arboricult. 18, 227–234.

Dwyer, J.F., Shroeder, H.W., Gobster, P.H., 1994. The deep sig-nificance of urban trees and forests. In: Platt, R.H., Rowntree,R.A., Muick, P.C. (Eds.), The Ecological City. University of Mas-sachusetts Press, Amherst.

Environment Preservation Bureau Tokyo Metropolitan City, 1999.Garden-Planning in Buildings, Amendments to the Regulationsregarding Garden-Planning in Buildings, Tokyo MetropolitanCity.

Floyd, J., 2000. Open space. In: Paper Presented at the Royal Aus-tralian Institute of Parks and Recreation Melbourne 1991 An-nual National Conference, Melbourne, Melbourne Water, pp.5/1–5/13.

Gehl, J., Gemzoe, L., 2001. New City Spaces. The Danish Architec-tural Press, Copenhagen.

Godbey, G., Graefe, A., James, S.W., 1992. The Benefits of LocalRecreation and Park Services: A Nationwide Study of the Per-ception of the American Public, Leisure Studies Program, Penn-sylvania State University for the National Recreation and ParkAssociation.

Grahn, P., 1991. Landscapes in our minds: people’s choice of recre-ation places in towns. Landscape Res. 16, 11–19.

Gustanski, J.A., Squires, R.H. (Eds.), 2000. Protecting the Land:Conservation Easements Past, Present and Future. Island Press,Washington, DC.

Hayward, G., Weitzer, W.H., 1984. The public’s image of urbanparks. Urban Ecol. 8, 243–268.

Housing and Development Board (HDB). Annual Report 2000/2001;2002/2003, Housing and Development Board, Singapore.

Hollis, L.E., Fulton, W., 2002. Open space protection: con-servation meets growth management. Discussion Paper.Center on Urban and Metropolitan Policy, The Brook-

e:llis-

I peri-

J tage

J logy

K ridge

K theook

K rge.

KL iley,

L s andSin-

Look Japan, 2000. Greening the Cities: Japan Pioneers in EnergyConservation, vol. 46, No. 535.

Lynch, K., 1960. The Image of the City. The Massachusetts Instituteof Technology Press, Cambridge, MA.

Madge, C., 1997. Public parks and the geography of fear. Tijdschr.Econ. Soc. Geogr. 88 (3), 237–250.

Mercer, D. (Ed.), 1977. Leisure and Recreation in Australia. SorrettPublishing, Melbourne.

Miller, R.W., 1986. Urban Forestry: Planning and Managing UrbanGreenspaces. Prentice-Hall, Englewood Cliffs, NJ.

Nasar, J.L., Jones, K., 1997. Landscapes of fear and stress. Environ.Behav. 29, 291–323.

National Research Council of Canada, Institute for Research in Con-struction, 2000. Greening our roofs: IRC seeks partners to studybenefits and durability of rooftop gardens. Construct. Innovat. 5(1).

Nesbit, R., 1980. History of the Idea of Progress. Basic Books, NewYork.

Oi, K.H., 1998. Park connectors. In: Yuen, B. (Ed.), Planning Singa-pore. Singapore Institute of Planners, Singapore.

Olmsted Jr., F.L., Kimball, T.H. (Eds.), 1970. Frederick Law Olm-sted, Landscape Architect. Benjamin Blon, New York, pp.1822–1903.

Ong, B.L., 2003. Green plot ratio: an ecological measure for archi-tecture and urban planning. Landscape Urban Plan. 63, 197–211.

Partners for Livable Communities, 2000. The Livable City: Revital-izing Urban Communities. McGraw-Hill, Inc., New York.

Peck, S., Callaghan, C., Peck and Associates, Bass, B., Kuhn, M.,1999. Greenbacks from Green Roofs: Forging a New Industry inCanada, Prepared for Canada Mortgage and Housing Corpora-tion, Ottawa.

Pendall, R., Martin, J. Fulton, W., 2002. Holding the Line:Urban Containment in the United States. DiscussionPaper. Center on Urban and Metropolitan Policy, TheBrookings Institution, Washington, DC. Available online:

nda-

P ntal

P tmentn of

R ress,

RR sance.

rs ofthe

S . Be-

T HDB. BSctional

T Land-

ings Institution, Washington, DC. Available onlinhttp://www.brook.edu/dybdocroot/es/urban/publications/hofultonopenspace.htm.

ttelson, W.H., 1978. Environmental perception and urban exence. Environ. Behav. 10, 193–214.

acobs, J., 1961. The Life and Death of Great American City. VinBooks, New York.

ohnston, J., Newton, J., 1996. Building Green. The London EcoUnit, London.

aplan, R., Kaplan, S., 1989. The Experience of Nature. CambUniversity Press, Cambridge.

nopf’s, R.C., 1987. Human behavior, cognition and affect innatural environment. In: Stokols, D., Altman, I. (Eds.), Handbof Environmental Psychology. Wiley, New York.

oolhaas, R., Mau, B., 1995. Small, Medium, Large Extra-LaThe Monacelli Press, New York.

uhn, M., 1996. Roof greening, Eco Architecture 2, OAA.ang, J., 1994. Urban Design: The American Experience. W

New York.ee, S.K., Chua, S.E., 1992. More than a Garden City, Park

Recreation Department. Ministry of National Development,gapore.

http://www.brook.edu/dybdocroot/es/urban/publications/pellfultoncontainment.pdf.

erry, M., Kong, L., Yeoh, B., 1997. Singapore: A DevelopmeCity State. Wiley, Chichester.

rice, L.W. (Ed.), 1987. Portland’s Changing Landscape. Deparof Geography, Portland State University and the AssociatioAmerican Geographers, Portland.

apoport, A., 1977. Human Aspects of Urban Form. Pergamon POxford.

elph, E., 1976. Place and Placelessness. Pion, London.ogers, R., Urban Task Force, 1999. Towards an Urban Renais

Final Report of the Urban Task Force chaired by Lord RogeRiverside. Department of the Environment, Transport andRegions, London.

hafer, E.L., 1969. Perception of natural environments. Environhav.

an, L.T., 2003/2004. Receptivity towards roof gardens amongresidents: a case study of a roof gardens at punggol fields(Real Estate) Dissertation. Department of Real Estate, NaUniversity of Singapore.

hompson, C.W., 2002. Urban open space in the 21st century.scape Urban Plan. 900, 1–14.

http://www.brook.edu/dybdocroot/es/urban/publications/hollisfultonopenspace.htm

http://www.brook.edu/dybdocroot/es/urban/publications/hollisfultonopenspace.htm

http://www.brook.edu/dybdocroot/es/urban/publications/pendallfultoncontainment.pdf

http://www.brook.edu/dybdocroot/es/urban/publications/pendallfultoncontainment.pdf


Ulrich, R.S., 1984. View through a window may influence recoveryfrom surgery. Science 224, 420–421.

Ulrich, R.S., Simons, R.F., 1991. Stress recovery during exposureto natural and urban environments. J. Environ. Psychol. 11,201–230.

Urban Redevelopment Authority Singapore, 1991. Living The NextLap. Urban Redevelopment Authority, Singapore.

Urban Redevelopment Authority Singapore, 2001. Concept Plan. Ur-ban Redevelopment Authority, Singapore.

Urban Redevelopment Authority Singapore, 2002. Skyline. UrbanRedevelopment Authority, Singapore.

Urban Redevelopment Authority Singapore, 2003. Skyline. UrbanRedevelopment Authority, Singapore.

Van Herzele, A., 2003. A monitoring tool for the provision of acces-sible and attractive urban green spaces. Landscape Urban Plan.63, 109–126.

Van Kamp, I., Leidelmeijer, K., Marsman, G., de Hollander, A., 2003.Urban environmental quality and human well-being: towards a

conceptual framework and demarcation of concepts; a literaturestudy. Landscape Urban Plan. 65, 5–18.

Wakely, P., You, N., 2001. Implementing the Habitat Agenda: InSearch of Urban Sustainability. University College London, Lon-don.

Walpole, K., 2000. Regaining an interior world. Landscape Des. 289,20–22.

Whyte, E., 1988. The City: Rediscovering The Centre. Doubleday,New York.

Willmert, T., 2000. The grass is greener on the topside with theseinnovative roofing systems. Architect. Rec. 188 (10), 182.

Wong, A., Yeh, S.H.K., 1985. Housing a Nation. Maruzen Asia,Singapore.

Wong, N.H., 2002. Study of Rooftop Gardens in Singapore. NationalUniversity of Singapore, Singapore.

Yeang, K., 1999. Planning the sustainable city as the vertical-city-in-the-sky. In: Foo, A.F., Yuen, B. (Eds.), Sustainable Cities inthe 21st Century. National University of Singapore, Singapore.

Resident Perceptions and Expectations of Rooftop Gardens in Singapore

Documents

choa chu kang

urban redevelopment authority

urban task force

local residents perceive

open space systems

public housing towns

open space

open spaces