Urban heat islands and the transformation of Singapore

Abstract

Chinese

An urban heat island is defined as an urban area that experiences warmer temperatures than its surroundings. This study examines how Singapore’s planning efforts established after the mid-20th century have affected the thermal environment of the city in association with land transformation, using historical temperature data available from the Meteorological Service of Singapore and some historical studies. Singapore’s planners have carefully regulated the growth of its downtown while promoting expansion in other parts of the city-state. These effects of planning have also unconsciously shaped the location and outline of Singapore’s urban heat island. As a result, new urban heat peaks were found around the centres of newly constructed large-scale new towns compared to industrial areas. This study provides lessons for land planning in mitigating a city’s urban heat island effects.

Keywords

land use planning Singapore transformation urban heat island

Introduction

The urban heat island (UHI) effect has become one of the key problems associated with contemporary urbanisation, and has attracted wide-ranging research in metropolitan areas around the world (De Ridder et al., 2016; Oke, 1973; Wilby, 2003; Wong and Yu, 2005). UHIs, defined as urbanised areas that experience warmer temperatures than their surroundings (Arnfield, 2003; Oke, 1973; Peng et al., 2012; Shepherd, 2005; Voogt and Oke, 2003), have numerous potential effects on water and air quality, vegetation growth (Grimm et al., 2008; Shepherd, 2005) and the increase of heat stress on urban residents, leading to increased mortality (Constantinescu et al., 2016; Gong et al., 2012; Patz et al., 2005). The increased use of air conditioning in UHIs leads to increased energy consumption and greenhouse gas emissions (Zhou et al., 2012), and those living in such areas suffer from a higher risk of heat stroke (Yang et al., 2017; Zhou et al., 2014). While studies are increasingly focusing on the potential hazards that could be caused by UHIs (McDonnell and MacGregor-Fors, 2016), fewer studies have tracked UHIs in relation to urban master plans, which is the focus of this article.

Luke Howard might have been the first to suggest that air temperatures in cities are higher than in rural areas (Howard, 1818), although the term ‘urban heat island’ was not coined until 1929 (Stewart, 2019), and was not subjected to rigorous quantitative analysis until 1969 (Masson et al., 2020). While it may seem straightforward to measure the temperatures of UHIs, there are different metrics that can be applied to these temporally and spatially dynamic phenomena. Oke (1995) has classified UHI measurement into four types: (1) subsurface (difference between temperature patterns in the ground under a city and those in the surrounding rural area); (2) surface (remote sensing satellite images); (3) canopy layer (measurement of the air volume between buildings); and (4) boundary layer (measurement of the air above the buildings).

Research on UHIs also requires a consistent and reliable dataset to derive the UHI intensity and thus understand the influence of spatial factors. Remote sensing images provide consistent wall-to-wall coverage of a desired area, and hence these datasets have become valuable resources for studying large urban areas (Mackey et al., 2012; Schwarz et al., 2011; Tayyebi et al., 2018). Even though remote sensing provides information on the spatial distribution of ground temperatures, however, it is hard to identify long-term trends using this data as there are various factors (e.g. inconsistent resolution, atmospheric conditions, sensor issues) affecting precise temperature measurement at ground level. On the other hand, an assessment of canopy-layer UHI (generally based on weather station readings) is most suitable for a microscale study, allowing researchers to present and model long-term data (Kato and Yamaguchi, 2007; Santamouris, 2014; Steeneveld et al., 2011). Urban surfaces absorb a tremendous amount of heat during the daytime and then release it slowly after dark. As a result, nighttime urban temperatures drop more gradually in cities than in rural areas. This increases nighttime temperature and prolongs the daily heat burden on the urban population, who feel the effect most strongly at night (Azevedo et al., 2016; Lac et al., 2013). Thus, this study uses nighttime canopy-layer temperatures to characterise UHI formations.

Little urban climate research has been done in the subtropical context. However, future urban growth is likely to occur most strongly in cities in these latitudes, where there is a strong demonstrable connection between urbanisation and urban heat island effects. Roth (2007) urges the scientific research community to focus more than they have on (sub)tropical urban climate studies for this reason. Even though tropical maximum UHI intensities are known to be lower than those observed in some temperate cities of similar population size (Oke et al., 1990), the UHI in hot climates likely increases thermal discomfort and associated heat-related problems (e.g. demand for air conditioning) more markedly. It is therefore critical that tropical cities incorporate climatological considerations in their planning and designs for a sustainable environment.

Previous studies have examined a wide range of factors to identify causes of the UHI effects through empirical approaches. Table 1 summarises factors affecting UHIs in terms of type, scale, and effects. The most common factor for analysis has been land cover. Urban surfaces and buildings commonly consist of impermeable covers with a large capacity for absorbing and projecting or reflecting solar radiation. Many studies point out the greater densities of buildings and paved roads as major contributors to UHI intensification (Bouyer et al., 2011; Das Majumdar and Biswas, 2016; Oke, 1982; Zheng et al., 2014). The average sizes of buildings and widths of roads are also important factors (Huang et al., 2008; Oke, 1988, 1995). Urban sprawl patterns likewise influence thermal environments in cities, although their comparative importance is controversial. Some studies in North America have found that low-density sprawl contributes to UHIs (Lemonsu et al., 2015; Middel et al., 2014; Stone et al., 2010; Stone and Norman, 2006; Stone and Rodgers, 2001), while others argue that high-density urban environments have the more significant impact on the UHI effect (Priyadarsini et al., 2008; Svensson and Eliasson, 2002).

Table 1.

Summary of factors affecting urban heat islands.

Type	Factor	Scale	Effects on UHIs
Physical	Paved roads and buildings	Micro	Intensifying
Physical	Built environment density	Local	Mixed
Social	Population	Local	Mixed
Environmental	Water bodies	Micro	Cooling
	Vegetation	Micro	Cooling
	Green roof/green wall	Micro	Cooling

Researchers have also found a positive relationship between population and UHI effects (Chen et al., 2006; Cui et al., 2016). Even though cities with large populations tend to have a higher density of high-rise buildings, some scholars have found a contradictory result: less populous cities with a higher UHI effect than more populous ones (Tran et al., 2006). Oke (1973) and Torok et al. (2001) also pointed out that even towns with populations of 1000 people could have a higher temperature (over 2°C) compared to nearby rural areas. Water bodies act to relieve urban thermal environments, and play an important role in reducing surrounding air temperature by cooling through evaporation (Ishii et al., 1991; Oke, 1987). The cooling effect of vegetation is also well-studied. Vegetation absorbs solar radiation, lowering the sensible heat (Wong and Yu, 2005) and cooling the thermal environment with shading and evapotranspiration (Dimoudi and Nikolopoulou, 2003). Many studies, therefore, emphasise the cooling effects of vegetation and green components for reducing the UHI in urban areas (Lu et al., 2012; Wu and Chen, 2017; Zhou et al., 2014) and suggest the incorporation of more green roofs (Li et al., 2011; Middel et al., 2015) and green walls (Alexandri and Jones, 2008). Others find that urban parks can lower surrounding air surface temperature (Jauregui, 1990; Spronken-Smith and Oke, 1998).

Master plans, documents guiding decisions on urban development (Bell, 2005; Madanipour et al., 2018), can significantly contribute to shaping components such as land use, infrastructure, buildings, natural patrimony, and socio-economic characteristics, all of which are closely associated with UHIs. While many studies examine these specific components’ effects on temperatures as they relate to UHIs, such as land use (Hart and Sailor, 2009; Jusuf et al., 2007) and vegetation (Susca et al., 2011; Weng et al., 2004), relatively fewer analyse master plans to identify how planning directions impact UHI or urban temperatures. One existing study assesses UHI effects before and after implementing the Hanoi Master Plan 2030 in association with the cooling effects of proposed green strategies (Kubota et al., 2017), while another examines ways to integrate urban climate information into urban planning using spatially distributed information (Eum et al., 2013). Kim et al. (2019) analysed master plans’ impact on temperatures around three centres in Seoul in association with that city’s UHI. Although there are some existing studies examining the relationships between master plans and UHIs, they are mainly about the immediate effects of a single plan on temperature changes over time, and only on a fixed urban centre. Broader perspectives are required in planning, however, as climate change-related matters are becoming more pressing and complex. There is little focus in the UHI literature on the considerations of long-term changes in the thermal environment and their spatial distribution as a result of implementing a series of plans with different but cumulative policy directions over time.

This paper offers a three-part analysis of Singapore as a basis for a long-term examination of how planning efforts in their aggregate have affected the spatial transformation of the UHI in a fast-growing, high-density tropical city. The first part reviews four distinct concept plans for Singapore and analyses historical changes in their directions. The second part analyses effects of the concept plans by observing land use changes in association with UHI effects. The final part analyses how the spatial distribution of Singapore’s UHI has changed as the city was transformed as a result of master planning since the 1980s.

Climate and background of Singapore

Singapore is situated near the southern end of Peninsular Malaysia, with an area of 712 km² in 2020 (Department of Statistics [DOS], 2020a). It has a typically tropical climate, with abundant rainfall (annual total of 2809.6 mm at the Changi climate station), high and uniform temperature variation (annual mean of 27.9°C), and high humidity all year round (Meteorological Services Singapore[MSS], 2021). Many of its climate variables, such as temperature and relative humidity, do not show significant month-to-month variation. However, many variables exhibit prominent diurnal variations from hour to hour, indicating the strong influence that solar heating has on the local climate. Singapore’s climate is characterised by two monsoon seasons separated by inter-monsoonal periods (late March to May and October to November). The Northeast Monsoon occurs from December to early March (winter months), and the Southwest Monsoon from June to September (summer months). During the inter-monsoon months, wind direction is highly variable, with generally lower mean wind speeds than in monsoonal periods.

The mean surface air temperature in Singapore has risen by more than 2°C over the last half-century. Furthermore, as a consequence of changes in the heat balance, air temperatures in densely built areas are higher than those in other areas of the city. This UHI phenomenon is a reflection of the totality of microclimatic changes brought about by man-made alterations of the urban surface (Landsberg, 1981), which have their origin in urban planning. In high-latitude cities with cooler weather, heat islands can be an asset in reducing heating loads. However, in low-latitude cities like Singapore, heat islands contribute to the urban dweller’s summer discomfort, ill-health, and significantly higher air conditioning loads. The compiled thermal comfort indices suggest that the optimal comfortable period in Singapore occurs during the winter monsoon, while the most uncomfortable period occurs during the onset of the summer monsoon in April–May (Roth and Chow, 2012).

Singapore’s rapid urbanisation since its independence in 1965 has been notable. As shown in Table 2, the population increased from almost 1.8 million in 1965 to approximately 5.7 million in 2020, translating to a very high mean population density of 7810 people per km² (DOS, 2020b). The percentage of the population living in government-supplied high-rise housing (known as HDB housing after the Housing and Development Board) has dramatically increased from 23% in 1965 to 83% in 2005. In the 1950s and 1960s, most people lived either in in low-rise masonry terraced houses or informal ‘kampong’ settlements of wood and thatch (Eng and Savage, 1985). By the late 1970s, the HDB had removed (actually bulldozed) the informal ‘kampong’ communities and moved their populations (and those in the crowded quarters of the city centre) into high-rise residential clusters known as ‘new towns’ (Table 2). The government’s decision to build high-density high-rise housing in new towns to accommodate an increasingly growing population was critical in changing the urban morphology of Singapore. In 2019, 79% of residential dwellings were HDB public housing, while only 16% of residential dwellings were privately owned high-rise condominiums (DOS, 2020b). Only 5% of residential dwellings were privately owned low-rise houses.

Table 2.

Scale and dimension: Land and population.

	Year	Unit	1965	1985	2005	2020
Land	Land area	km²	581.5	620.5	696.9	728
Land	Urbanised area	km²	177.4	298.8	NA	518 (2015)
Population	Population	Persons	1,886,900	2,735,957	4,265,762	5,685,807
	Density	Persons/km²	3245	4409	6121	7810
	% of Resident Population living in HDB housing	%	23	81	83	79 (2019)

Source: Department of Statistics, Singapore (DOS, 2020b).

Because of the dispersion which accompanied this great re-landscaping and re-housing, the country has evolved as a garden city-state without a distinct boundary between urban and rural areas. This is in great contrast to Hong Kong, also a former British colony, in which significant original green-space was preserved on the periphery, forcing extreme densities in the centre. There are only two predominant green areas in Singapore of significant size: the primary forest in the middle of the island (Central Catchment), which holds a reservoir, and another in the northwest corner (Lim Chu Kang). Singapore’s urbanised area has exponentially increased from 177.4 km² in 1965 to 518 km² in 2015 (Figure 1), mainly at the expense of farmland and forest. A large portion of natural land has been converted into industrial estates, mainly in the Jurong area – at the western end of the island, more than 15 km from the old city. As of 2015, 13% of the island’s total land mass is allocated for industrial use (Tang, 2017: 173). Most built-up regions, like industrial areas, the Central Business District (CBD), and the airport are located in the southern half of the island. Therefore, the northern part of the country can be considered to have ‘rural’ pockets (though still land-banked for possible future development) while the southern part is totally and densely urbanised.

Figure 1.

1943 map and 2010 map of Singapore.

Concept Plans of Singapore, 1958–2001

The basic concepts underlying Singapore’s urban development and transport plans have remained constant since the 1960s, even though some adjustments and refinements in the application of plans have been made over time. Table 3 is a compact summary of planning directions adopted in the four concept plans of Singapore.

Table 3.

Planning directions of Singapore in the concept plans.

Year	Plan directions
1958	The delimitation of a green belt around the city to prevent urban sprawl
	The decongestion of the Central City Area
	The construction of three new towns at Jurong, Woodlands, and Lim Chu Kang
	Open space targets raising the existing ratio of 0.34 to 1.1 hectares per 1000 population
1971	A ring structure of satellite new towns situated around the Central Water Catchment Area
	An island wide system of expressways and a rail-based Mass Rapid Transit (MRT) system to link all major new towns to the city centre and to Jurong Industrial Estate
	A new international airport plan at Changi
1991	A vision of ‘A Tropical City of Excellence’ An extension of the CBD to Marina South
	Regional centres development at Jurong East, Woodlands, Yishun, and Tampines
	Sub-regional centres at Buona Vista, Bishan, Serangoon, Paya Lebar, and Marine Parade
	‘Fringe centres’ introduction (the Singapore River, Outram, Newton, Novena, Lavender and Orchard)
2001	A vision of ‘towards a thriving world class city in the 21st century’ Reaffirmation of the CBD (global business centre)
	Further development at the three regional centres (Tampines, Woodlands and Jurong East)
	New housing construction in the western part of the island and in the new downtown at Marina South
	Green spaces increase in Singapore from 2500 ha to 4500 ha

The 1958 Master Plan

The first 1958 Concept Plan (Figure 2), developed when Singapore was still a British colony, divided the island into three zones: an inner city, a surrounding ‘town’ area, and a rural outer ring. It was in that sense a reflection of existing realities, but also designed such that the city could expand in all directions and be serviced by a network of roads mainly radiating from the centre. Some principal features included: the delimitation of a greenbelt around the city to prevent urban sprawl; the decongestion of the Central City Area by moving one-sixth of the resident population (in order to overcome then-current densities of more than 2500 persons per hectare); the construction of three ‘new towns’, at Jurong, Woodlands, and Lim Chu Kang, to effectuate the planned decentralisation; and open space targets raising the then-current ratio of 0.34 to 1.1 hectares per 1000 population (Dale, 1999: 77).

Figure 2.

Planning maps for Singapore.

The 1971 Concept Plan

The 1971 Concept Plan (Figure 2), the first to be drawn after independence, brought greater coherence to planning and development in Singapore, allowing the government to build public housing, industrial estates, and infrastructures over the island as a whole. Not only was the distinction between core and periphery largely collapsed, but the ‘city’ of Singapore, previously distinct from the rest of the island, became simply the largest hub area within a new ‘national’ space, where other hub areas would develop over time. A key direction in the 1971 Concept Plan was the idea to develop Singapore according to a Ring Concept by organising land use into high-density satellite towns surrounding the central catchment area and with a broad band of development along the east and west coasts, all linked by a planned transport system (Singh, 2017: 130). The Mass Rapid Transit (MRT) network was strategically planned along the development corridors and consisted of a North–South Line from the city centre to the north and an intersecting East–West Line along the south coast. Altogether, an expanded road network and an MRT network formed the key components of the transport layer plan (Figure 2), allowing a systematic expansion of transport infrastructure in line with land development and traffic growth.

The 1991 Concept Plan

The 1971 Concept Plan was updated in 1991 in response to Singapore’s rapid growth in population, which had reached 2.9 million by 1982. One of the key goals of the 1991 Concept Plan was to alleviate congestion in the CBD through accelerated decentralisation, which required the creation of a hierarchy of urban centres outside the central area. The plan divided the island city into five regions – Central, West, North, North-East and East (Figure 2). Each region was to accommodate around a million or more inhabitants and was further subdivided into highly self-sufficient new towns featuring a mix of high, medium, and low-density housing forms. Besides the Central Business District (CBD), Jurong East, Woodlands, Yishun, and Tampines became regional centres (Urban Redevelopment Authority [URA], 1991). The plan also identified several sub-regional centres to further aid decentralisation. In addition, several fringe centres were introduced to complement this function (URA, 1991).

The 2001 Concept Plan

The Concept Plan 2001 (Figure 2) was based on a population scenario of 5.5 million, focusing on housing, recreation, business, infrastructure, and identity. The plan identified new housing to be built in mature estates, the western part of the island, and in the centre, mostly in a new downtown at Marina South (URA, 2001: 13–19). The plan also proposed to increase green spaces in Singapore from 2500 ha to 4500 ha (URA, 2001: 21–27), including opening up more areas in the Central Catchment Reserve to public use. In the city centre, a new white zone was planned to allow for multiple uses, such as housing, offices, and research and development facilities. This plan included a key proposal for a more extensive rail network and a more vigorous reaffirmation of the CBD, to some degree downplaying the role of regional centres compared to the 1991 plan (Table 4).

Table 4.

Urban governance sector.

1. The Housing and Development Board (HDB) is the sole agency responsible for the planning, design, land assembly and construction of public housing.

2. The Urban Redevelopment Authority (URA) is Singapore’s national land use planning authority since 1989. URA prepares long term strategic plans, known as Concept Plans, as well as detailed local area plans for physical development and conservation.

3. The Land Transport Authority (LTA) established in 1995 is responsible for planning, policy, and regulation of all urban transport modes. It also constructs road, MRT, and other public transport infrastructure, in accordance with the provisions of the Concept Plans.

Concept plan effects on land use changes

The directions adopted in the four concept plans (Table 3) have indeed been reflected in Singapore’s land use and urbanisation patterns. The visions presented in the concept plans were supported by transport plans prepared by the Land Transport Authority, and more detailed plans prepared by the HDB (mainly for housing development) and the Jurong Town Corporation (for industrial areas).

Under the 1958 Concept plan, the Jurong Town Corporation (JTC) began constructing a new town with deep-water port facilities in 1968 at Jurong (in the island’s far west), providing industrial land and infrastructure to investors and decongesting the CBD area. The development of the Jurong Industrial Estate was critical to easing the problem of mass unemployment in the 1960s and became a key contributor to the city-state’s manufacturing profile. After 1965, the HDB began removing traditional informal low-rise houses and building high-rise houses in new towns to rehouse the majority of the population. Queenstown, located at the edge of the city centre, was the first new town, and was built during the late period of British control. The first new town planning was not integrated with rail transit development plans.

The 1971 Concept Plan was the first in which land use and transport were closely integrated. It was structured around mass transit spines, with new towns being located around mass transit stations. Arising from the 1971 Concept Plan, the road infrastructure developed extensively from about 800 km to nearly 3000 km of roads between the 1970s and 1990s (Singh, 2017: 132). Singapore’s Urban Redevelopment Authority started to intentionally integrate rail transit development into new town planning through a systematic collaboration with the LTA and HDB. Toa Payoh, first announced in 1965, was the first new town that strategically integrated the public transit infrastructure (then buses) with residential buildings and commercial facilities. It would be the largest of Singapore’s new towns and globally one of the largest such developments of the 1960s–70s. The construction of the first Mass Rail Transport (MRT) lines, East–West and North–South, started in 1983, and a pattern of connecting new towns with MRT stations subsequently became a key planning strategy. In 1981, the international airport was relocated from Paya Lebar to Changi on the far eastern end of the island to make more land available in the centre city for other developments.

The 1991 Concept Plan resulted in key refinements to land use planning in Singapore: the decentralisation of growth in economic activities from the CBD through the development of fringe centres around MRT stations; and the location of employment areas (e.g. business parks and commercial centres) near residential areas. The commercial centres were classified into town, fringe, sub-regional, and regional centres. They were distributed along rail transit to create balance between employment centres and residences (URA, 1991). Tampines was a successful example of new town development within the 1991 plan, becoming a regional centre for shopping and living, and exhibiting the typical characteristics of a transit-oriented development (TOD) pattern, where a compact mixed-use neighbourhood is arranged around a transit station. The Buona Vista Science Hub around the new Buona Vista MRT station (later renamed One-North) was announced in 1991. The development of One-North was Singapore’s first experiment at developing knowledge-based industry clusters in an urbanised mixed-use environment (Tang, 2017: 166). In addition, the formation of the Jurong Islands, comprising several offshore islands in the southern part of the Jurong Industrial area, made the southwestern end of Singapore a petrochemical hub in the 1990s.

The key vision of the 2001 Concept Plan was establishing a global business centre to ensure the sustainability of employment in a globally competitive business environment. The expansion of the CBD, where most corporate headquarters were located, was thus reaffirmed more strongly than before. In the 2010s, the formation of Marina Bay Financial Centre along the south bank of the Singapore River and the addition of iconic buildings completed the current skyline of downtown in Singapore. Further HDB housing constructions continued in regional centres across the city. As a result of green area protection efforts in this plan, 9% of Singapore’s land is committed to parks and green spaces, including four nature reserves.

The transformation of Singapore’s urban heat island as a result of planning

Roth and Chow (2012) reviewed 20 studies since the 1960s examining the influence of urban development on the thermal environment in Singapore. This study examines the land use changes’ relation to the UHI in the 1980s and 1990s, using summer nighttime canopy-layer temperature maps (Roth and Chow, 2012) based on observations from May to August, 1979–1981 (Singapore Meteorological Service [SMS], 1986) and 24–30 June 1996 (Goh and Chang, 1998). The Singapore Meteorological Service (SMS, 1986) mapped 20:00 – 21:00 hour air temperatures obtained from the four meteorological stations located at Changi (CH), Tengah (TE), Paya Lebar (PL) and Seletar (SL) airports, based on observations from nine days total between 1979 and 1981 (Figure 3).

Figure 3.

Summer canopy-layer temperature map based on observations during May–August, 1979–1981.

Between 1979 and 1981, the most prominent UHI peaks were found around the C1/2 (southern commercial and CBD) areas of the island. Even though the focus of the 1958 and 1971 Concept plans was the decongestion of these same areas, they still had the most compact urban development pattern, and were filled with high-rise buildings. The second highest UHI peaks were formed around the Queenstown (Q) new town, Toa Payoh (TP) new town, Ang Mo Kio (A) new town, and the residential areas of Katong/Marina parade (K). These UHI formations around the three new towns are directly associated with the HDB’s new town plans started in the 1960s (Queenstown), 1965 (Toa Payoh), and 1973 (Ang Mo Kio), arising from the 1958/1971 Concept plans (Figure 4). An additional warm belt in western Singapore covered the Jurong Industry (JI) area, which was developed strategically as a result of the 1958 plan for Singapore’s manufacturing output and employment needs. The coolest areas were within and around the Central Catchment (CC) area and the southern part of Kranji Reservoir (KR). This phenomenon manifests the cooling effects of vegetation and water bodies on UHIs. Other cool temperature islands were found around the undeveloped Lim Chu Kang (LC), the northern and eastern parts of the state.

Figure 4.

Locations and boundaries of HDB towns.

Goh and Chang (1998) provided the spatial pattern of the 21:00 hour air temperature of Singapore. The data was collected using sling psychrometers at 15 sites across Singapore and was supplemented by secondary data from six airport weather stations. The measurement was based on four 7-day periods sampled between March 1996 and January 1997. Figure 5 confirms the largest UHI existed over the commercial and CBD areas (C1/2) and the coolest temperate islands were in the forested areas in the Central Catchment (CC) and undeveloped areas around Lim Chu Kang (LC). However, distinct changes were found in Tampines (T) and Pasir Ris (PR) in the east, where extensive high-density residential buildings were constructed as a result of the HDB new town plans guided by the 1991 Concept plan. This area has shown the greatest temperature increase compared to the map of 1979–1981, demonstrating the effect of compact residential buildings on UHI formation. As Goh and Chang (1998) noted, the construction of the Jurong West estate (J) did not contribute to intensifying a UHI in the area, nor did the construction of the Buona Vista (B) Science Hub seem to result in a distinct heat island. Based on the comparison between the maps of 1979 and 1981 and of 1996, the construction of new towns composed of high-rise residential buildings had a higher impact on UHI intensification than urban development in industrial areas.

Figure 5.

Summer canopy-layer temperature map based on observations 24–30 June, 1996.

The Meteorological Service Singapore (MSS) measures the actual air temperature using a thermometer enclosed in a white louvred radiation shield that allows air to circulate within and protects the thermometer from rain and direct exposure to solar radiation. Following the World Meteorological Organization’s guidelines on the siting of radiation shields, thermometers are housed in radiation shields. They are placed between 1.25 and 2 m above ground for measuring temperature by MSS. Minimum daily temperatures during May 2010 collected from 21 weather stations across Singapore (MSS, n.d.) were used to create the isotherm map (Figure 6). Among available average, maximum, and minimum daily temperatures from the MSS, the minimum daily temperature represents the nighttime temperature. Figure 7 provides the locations of the weather stations where the temperature data used for the heat map 2010 were collected.

Figure 6.

Summer canopy-layer temperature map based on observations 1–31 May, 2010.

Figure 7.

Locations of the weather stations that are used for the isotherm map of 2010.

The summer canopy-layer temperature map of 2010 (Figure 6) again shows a similar pattern, with the largest UHI in the southern commercial area (C) and the smallest UHI within the undeveloped Lim Chu Kang (LC) area. Another cooling island was found in the upper part of the Lower Seletar Reservoir (SL) and the Central Catchment (CC). The spatial extent of the cool area around the Central Catchment was significantly reduced by residential development encroaching into the green area. On the continuation of the strong UHI in Tampines (T) and Pasir Ris (PR) in the east from the map of 1998 (Figure 5), intensified UHIs were formed around the Bedok (B) HDB cluster and the Hougang (H) new town. As of 2015, Hougang is the largest HDB new town based on land area. This change again confirms a strong relationship between the construction of the new towns and the UHI formation in Singapore. While there is no notable UHI in the Jurong Industry (JI) estate in the west, the temperature has increased in the Changi (CH) area, where consistent mixed-land use developments occurred after the relocation of the airport from the Paya Lebar to Changi, along with the extension of public transit connecting the CBD to the area. Generally, the south and the east, covered by commercial buildings and HDB housings, have relatively stronger UHIs than the less developed west and north.

Conclusion

This paper analysed Singapore to demonstrate how planning efforts have affected the spatial transformation of the UHI over time. In particular, four concept plans of Singapore established between 1958 and 2001 were reviewed for their impact on the urbanisation patterns in Singapore and hence the footprint of its UHI. It is apparent that the UHIs of Singapore have been spatially transformed since the 1980s. The earlier stages see the hottest part of the city situated mostly around the commercial and CBD areas, but the later stages have shown new UHI peaks, mostly in the upper part of the downtown areas and in the east, that were closely associated with the constructions of the large-scale HDB new towns. Between the 1980s and 2010, cool island formations around forestry areas, including the Central Catchment (CC) and the Lim Chu Kang (LC), and reservoirs have been consistent, reaffirming the cooling effects of vegetation and waterway on urban heat.

The relationships between the locations and boundaries of HDB towns (Figure 4) and the later heat island formations from the UHI maps in the 1980s, 1990s and 2000s were distinctive. These notable changes were possible as a high degree of land use/transport integration was being achieved in planning, and implementing plans on the ground. The concept of TOD has been practised in many Asian cities, giving expression to choices for high-density cities to cope with urbanisation before the rise of TOD theory in the west. This practice was even more efficient in city-state Singapore, given the integrated spatial planning of land uses and transport networks and implementation of new towns and rail infrastructure programmes following concept plans. Toa Payoh, Tampines, and Pasir Ris new towns are good examples of TOD practice in Singapore, featuring a compact and mixed-use neighbourhood arranged around a transit station site. Many advocate the concept and the practice of TOD with the notion of reducing automobile dependence and increasing the accessibility of areas (Cervero and Sullivan, 2010; Still, 2002) in association with smart growth and sustainability principles. However, in Singapore this also brought with it new centres of urban heat peaks. While the planning effort in Singapore to decentralise population into HDB new towns away from the city centre was successful, some have pointed out that the plans aiming at self-containment failed, as up to 80% of employed residents still travel outside their new town to work in the CBD or the major industrial areas (Niu et al., 2019). A failure to connect homes to jobs intensifies urban heat by increasing distance-travelled at a larger scale.

Another notable observation was that industrial areas have a lower UHI intensity compared to residential areas. Goh and Chang (1998) also noted that the construction of the Jurong West estate did not result in a distinct heat island in the west. Newly developed industrial estates around Changi airport did not result in a significant UHI in the east either. Oke (1982) demonstrated that the maximum heat island intensities were related to urban canyon geometry through the canyon height-to-width ratio. Goh and Chang (1999) investigated the statistical relationship between urban canyon height-to-width ratio and nocturnal heat island intensities in Singapore. They confirmed a positive relationship between them in the context of tropical climate. This implies that the construction of new towns filled with high-rise residential buildings, which is commonly practised in many Asian Cities, has a stronger impact on urban heat intensification than the development of industrial estates, which are usually low-rise.

Concept plans and Master plans are key drivers that may shape the UHI of a city. Many cities like Singapore have a long tradition of prioritising accommodation of the burgeoning population and supply of relevant infrastructure as key challenges. It was inevitable that temperature would increase in Singapore because of worldwide climate change and the city’s rapid urbanisation. While the Singapore government’s steady efforts at decongesting the central city area, combined with the evenly distributed developments across the city, had a positive effect on preventing the worsening of UHI in the urban core, they had the effect of creating new UHIs in the rest of Singapore (especially around the centres of newly developed new towns) as the city-state become an urbanised region as a whole. Therefore, this expanded UHI has become a challenge across the city for the government and citizens to tackle.

It is evident that the concept plans were effectively implemented under strong government leadership in Singapore, shaping the city’s urban morphology. The thermal environment of the city has been transformed accordingly. This research provides a good example of planning’s impact on the formation of UHIs and the thermal environment of a fast-growing Asian city. While it is difficult to solely rely on concept/master plans controlling UHI because they cover a wide range of topics, including the built environment and the economic and social issues of cities, establishing a comprehensive plan with a focus on how heat is closely inter-connected with growth could be an effective way of shaping and controlling UHIs. Lately, Singapore’s state authorities have publicly dedicated themselves to pursuing a ‘green’ and sustainable city with an interest in climate-sensitive planning at local scales. For example, a plan for Punggol eco-town was first announced in 1996. This new model will be an integrated laboratory for HDB to test-bed sustainable and innovative solutions in Singapore (Ming et al., 2010). This plan adopts design concepts, such as high-density waterfront housing, the introduction of common green areas, and the use of energy-efficient technologies. Given that Singapore needs more housing to prepare for its population growth and the construction of new towns is one of the major contributors to UHI, lessons from the success or failure of this HDB’s ambitious plan will be a valuable guideline for the future planning fighting against warming in Singapore.

Today, it is crucial to incorporate climatological concerns in urban planning to achieve a better living and working environment as UHIs are expected to become more intense in the future. There is an even more pressing need to prevent the worsening of UHIs in low-latitude tropical cities like Singapore. On top of considering the expected long-term changes in a city’s thermal environment for master planning, developing cities’ heat action plans and adaptation/resiliency plans based on city-specific circumstances are crucial.

Footnotes

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Ministry of Education, Singapore, Academic Research Fund, Tier 2 Grant, MOE2018-T2-2-120. Heat in Urban Asia: Past, Present and Future.

ORCID iD

Yoonhee Jung

References

Alexandri

Jones

(2008) Temperature decreases in an urban canyon due to green walls and green roofs in diverse climates. Building and Environment 43(4): 480–493.

Arnfield

(2003) Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the urban heat island. International Journal of Climatology 23(1): 1–26.

Azevedo

Chapman

Muller

(2016) Quantifying the daytime and night-time urban heat island in Birmingham, UK: A comparison of satellite derived land surface temperature and high resolution air temperature observations. Remote Sensing 8(2): 153.

Bell

(2005) The emergence of contemporary masterplans: Property markets and the value of urban design. Journal of Urban Design 10(1): 81–110.

Bouyer

Musy

Huang

, et al. (2011) Mitigating urban heat island effect by urban design: Forms and materials. In: Hoornweg

Freire

Lee

al.

(eds) Cities and Climate Change: Responding to an Urgent Agenda. Washington, DC: The World Bank, pp.164–181.

Cervero

Sullivan

(2010) Toward green TODs. Working Paper UCB-ITS-VWP-2010-7. Berkeley, CA: UC Berkeley Center for Future Urban Transportation.

Chen

Zhao

, et al. (2006) Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes. Remote Sensing of Environment 104(2): 133–146.

Constantinescu

Cheval

Caracaş

, et al. (2016) Effective monitoring and warning of urban heat island effect on the indoor thermal risk in Bucharest (Romania). Energy and Buildings 127: 452–468.

Cui

Dong

, et al. (2016) Influence of urbanization factors on surface urban heat island intensity: A comparison of countries at different developmental phases. Sustainability 8(8): 706.

10.

Dale

(1999) Urban Planning in Singapore: The Transformation of a City. Shah Alam: Oxford University Press.

11.

Das Majumdar

Biswas

(2016) Quantifying land surface temperature change from LISA clusters: An alternative approach to identifying urban land use transformation. Landscape and Urban Planning 153: 51–65.

12.

Department of Statistics (2020a) Yearbook of Statistics Singapore 2020. Singapore: Department of Statistics, Ministry of Trade and Industry.

13.

Department of Statistics (2020b) Population Trends, 2020. Singapore: Department of Statistics, Ministry of Trade and Industry. Available at: https://www.singstat.gov.sg/-/media/files/publications/population/population2020.pdf (accessed 1 January 2023).

14.

De Ridder

Maiheu

Lauwaet

, et al. (2016) Urban heat island intensification during hot spells—The case of Paris during the summer of 2003. Urban Science 1(1): 3.

15.

Dimoudi

Nikolopoulou

(2003) Vegetation in the urban environment: Microclimatic analysis and benefits. Energy and Buildings 35(1): 69–76.

16.

Eng

Savage

(1985) Singapore landscape: A historical overview of housing change. Singapore Journal of Tropical Geography 6(1): 48–63.

17.

Eum

Scherer

Fehrenbach

, et al. (2013) Integrating urban climate into urban master plans using spatially distributed information—The Seoul example. Land Use Policy 34: 223–232.

18.

Goh

Chang

(1998) The nocturnal heat island phenomenon of Singapore revisited. In: 5th Southeast Asian Geography Association Conference, 30 November–4 December 1998, Singapore.

19.

Goh

Chang

(1999) The relationship between height to width ratios and the heat island intensity at 22:00 h for Singapore. International Journal of Climatology 19(9): 1011–1023.

20.

Gong

Liang

Carlton

, et al. (2012) Urbanisation and health in China. The Lancet 379(9818): 843–852.

21.

Grimm

Faeth

Golubiewski

, et al. (2008) Global change and the ecology of cities. Science 319(5864): 756–760.

22.

Hart

Sailor

(2009) Quantifying the influence of land-use and surface characteristics on spatial variability in the urban heat island. Theoretical and Applied Climatology 95(3–4): 397–406.

23.

Housing & Development Board (2020) Key Statistics: HDB Annual Report 2019/2020. Singapore: Housing and Development Board.

24.

Howard

(1818) The climate of London: Deduced from meteorological observations, made at different places in the neighbourhood of the metropolis (Vol. 1). London: W. Phillips.

25.

Huang

Musy

Hégron

, et al. (2008) 663: Towards urban design guidelines from urban morphology description and climate adaptability. In: Proceedings of PLEA 2008–25th conference on passive and low energy architecture, Dublin, Ireland, 2–24 October 2008.

26.

Ishii

Iwamoto

Katayama

, et al. (1991) A comparison of field surveys on the thermal environment in urban areas surroundings a large pond: When filled and when drained. Energy and Buildings 16(3–4): 965–971.

27.

Jauregui

(1990) Influence of a large urban park on temperature and convective precipitation in a tropical city. Energy and Buildings 15(33–44): 457–463.

28.

Jusuf

Wong

Hagen

, et al. (2007) The influence of land use on the urban heat island in Singapore. Habitat International 31(2): 232–242.

29.

Kato

Yamaguchi

(2007) Estimation of storage heat flux in an urban area using ASTER data. Remote Sensing of Environment 110(1): 1–17.

30.

Kim

Jung

(2019) Transformation of urban heat island in the three-center city of Seoul, South Korea: The role of master plans. Land Use Policy 86: 328–338.

31.

Kubota

Lee

Trihamdani

, et al. (2017) Impacts of land use changes from the Hanoi Master Plan 2030 on urban heat islands: Part 1. Cooling effects of proposed green strategies. Sustainable Cities and Society 32: 295–317.

32.

Lac

Donnelly

Masson

, et al. (2013) CO₂ dispersion modelling over Paris region within the CO₂-MEGAPARIS project. Atmospheric Chemistry and Physics 13(9): 4941–4961.

33.

Landsberg

(1981) The Urban Climate. New York, NY and London: Academic Press.

34.

Lemonsu

Viguié

Daniel

, et al. (2015) Vulnerability to heat waves: Impact of urban expansion scenarios on urban heat island and heat stress in Paris (France). Urban Climate 14: 586–605.

35.

Song

Cao

, et al. (2011) Impacts of landscape structure on surface urban heat islands: A case study of Shanghai, China. Remote Sensing of Environment 115(12): 3249–3263.

36.

, et al. (2012) Regression analysis of the relationship between urban heat island effect and urban canopy characteristics in a mountainous city, Chongqing. Indoor and Built Environment 21(6): 821–836.

37.

Mackey

Lee

Smith

(2012) Remotely sensing the cooling effects of city scale efforts to reduce urban heat island. Building and Environment 49: 348–358.

38.

Madanipour

Miciukiewicz

Vigar

(2018) Master plans and urban change: The case of Sheffield city centre. Journal of Urban Design 23(4): 465–481.

39.

Masson

Lemonsu

Hidalgo

, et al. (2020) Urban climates and climate change. Annual Review of Environment and Resources 45(1): 411–444.

40.

McDonnell

MacGregor-Fors

(2016) The ecological future of cities. Science 352(6288): 936–938.

41.

Meteorological Services Singapore (2021) Annual Climate Assessment Singapore 2021. Available at http://www.weather.gov.sg/wp-content/uploads/2022/08/ACAR_2021.pdf (accessed 10 January 2023).

42.

Meteorological Services Singapore (n.d.) Historical daily records. Available at: http://www.weather.gov.sg/climate-historical-daily/ (accessed 10 January 2023).

43.

Middel

Chhetri

Quay

(2015) Urban forestry and cool roofs: Assessment of heat mitigation strategies in Phoenix residential neighborhoods. Urban Forestry & Urban Greening 14(1): 178–186.

44.

Middel

Häb

Brazel

, et al. (2014) Impact of urban form and design on mid-afternoon microclimate in Phoenix Local Climate Zones. Landscape and Urban Planning 122: 16–28.

45.

Ming

Suan

Toh

(2010) HDB’s next generation of eco-districts at punggol and eco-modernisation of existing towns. The IES Journal Part A: Civil & Structural Engineering 3(3): 203–209.

46.

National University of Singapore (n.d.) Historical maps of Singapore. Available at: https://libmaps.nus.edu.sg/ (accessed 1 July 2022).

47.

Niu

Shen

, et al. (2019) Study on land use characteristics of rail transit TOD sites in new towns—taking Singapore as an example. Journal of Asian Architecture and Building Engineering 18(1): 16–27.

48.

Oke

(1973) City size and the urban heat island. Atmospheric Environment (1967) 7(8): 769–779.

49.

Oke

(1982) The energetic basis of the urban heat island. Quarterly Journal of the Royal Meteorological Society 108(455): 1–24.

50.

Oke

(1987) Boundary Layer Climates. London: Methuen & Co. Ltd.

51.

Oke

(1988) Street design and urban canopy layer climate. Energy and Buildings 11(1–3): 103–113.

52.

Oke

(1995) The heat island of the urban boundary layer: Characteristics, causes and effects. In: Cermak

Davenport

Plate

, et al. (eds) Wind Climate in Cities. Dordrecht: Springer, pp.81–107.

53.

Oke

Taesler

Olsson

(1990) The tropical urban climate experiment (TRUCE). Energy and Buildings 15(1–2): 67–73.

54.

Patz

Campbell-Lendrum

Holloway

, et al. (2005) Impact of regional climate change on human health. Nature 438(7066): 310–317.

55.

Peng

Piao

Ciais

, et al. (2012) Surface urban heat island across 419 global big cities. Environmental Science & Technology 46(2): 696–703.

56.

Priyadarsini

Hien

Wai David

(2008) Microclimatic modeling of the urban thermal environment of Singapore to mitigate urban heat island. Solar Energy 82(8): 727–745.

57.

Roth

(2007) Review of urban climate research in (sub)tropical regions. International Journal of Climatology 27(14): 1859–1873.

58.

Roth

Chow

WTL

(2012) A historical review and assessment of urban heat island research in Singapore. Singapore Journal of Tropical Geography 33(3): 381–397.

59.

Santamouris

(2014) On the energy impact of urban heat island and global warming on buildings. Energy and Buildings 82: 100–113.

60.

Schwarz

Lautenbach

Seppelt

(2011) Exploring indicators for quantifying surface urban heat islands of European cities with MODIS land surface temperatures. Remote Sensing of Environment 115(12): 3175–3186.

61.

Shepherd

(2005) A review of current investigations of urban-induced rainfall and recommendations for the future. Earth Interactions 9(12): 1–27.

62.

Singapore Meteorological Service (1986) A study of the urban climate of Singapore. In: Proceedings of the Conference on The Biophysical Environment of Singapore and its Neighbouring Countries (eds Chia

Lee

Rahman

, et al.), 3–5 May 1985, pp.50–76. Singapore: Geography Teacher’s Association of Singapore.

63.

Singh

(2017) Transportation: Mobility, accessibility, and connectivity. In: Kiang

(ed.) 50 Years of Urban Planning in Singapore. Singapore: World Scientific Publishing, pp.127–151.

64.

Spronken-Smith

Oke

(1998) The thermal regime of urban parks in two cities with different summer climates. International Journal of Remote Sensing 19(11): 2085–2104.

65.

Steeneveld

Koopmans

Heusinkveld

, et al. (2011) Quantifying urban heat island effects and human comfort for cities of variable size and urban morphology in the Netherlands. Journal of Geophysical Research Atmospheres 116: D20129.

66.

Stewart

(2019) Why should urban heat island researchers study history? Urban Climate 30: 100484.

67.

Still

(2002) Transit-oriented development: Reshaping America’s metropolitan landscape. On Common Ground: 44–47.

68.

Stone

Hess

Frumkin

(2010) Urban form and extreme heat events: Are sprawling cities more vulnerable to climate change than compact cities? Environmental Health Perspectives 118(10): 1425–1428.

69.

Stone

Norman

(2006) Land use planning and surface heat island formation: A parcel-based radiation flux approach. Atmospheric Environment (1967) 40(19): 3561–3573.

70.

Stone

Rodgers

(2001) Urban form and thermal efficiency: How the design of cities influences the urban heat island effect. Journal of the American Planning Association. American Planning Association 67(2): 186–198.

71.

Susca

Gaffin

Dell’osso

(2011) Positive effects of vegetation: Urban heat island and green roofs. Environmental Pollution (1970) 159(8–9): 2119–2126.

72.

Svensson

Eliasson

(2002) Diurnal air temperatures in built-up areas in relation to urban planning. Landscape and Urban Planning 61(1): 37–54.

73.

Tang

(2017) Industrial Planning in Singapore. In: Kiang

(ed.) 50 Years of Urban Planning in Singapore. Singapore: World Scientific Publishing, pp.153–176.

74.

Tayyebi

Shafizadeh-Moghadam

Tayyebi

(2018) Analyzing long-term spatio-temporal patterns of land surface temperature in response to rapid urbanization in the mega-city of Tehran. Land Use Policy 71: 459–469.

75.

Torok

Morris

Skinner

, et al. (2001) Urban heat island features of southeast Australian towns. Australian Meteorological Magazine 50(1): 1–13.

76.

Tran

Uchihama

Ochi

, et al. (2006) Assessment with satellite data of the urban heat island effects in Asian mega cities. International Journal of Applied Earth Observation and Geoinformation 8(1): 34–48.

77.

Urban Redevelopment Authority (1991) Living the Next Lap: Towards a Tropical City of Excellence. Singapore: Urban Redevelopment Authority.

78.

Urban Redevelopment Authority (2001) Singapore Concept Plan 2001. Singapore: Urban Redevelopment Authority.

79.

Voogt

Oke

(2003) Thermal remote sensing of urban climates. Remote Sensing of Environment 86(3): 370–384.

80.

Weng

Schubring

(2004) Estimation of land surface temperature–vegetation abundance relationship for urban heat island studies. Remote Sensing of Environment 89(4): 467–483.

81.

Wilby

(2003) Past and projected trends in London’s urban heat island. Weather 58(7): 251–260.

82.

Wong

(2005) Study of green areas and urban heat island in a tropical city. Habitat International 29(3): 547–558.

83.

Chen

(2017) Optimizing the spatial arrangement of trees in residential neighborhoods for better cooling effects: Integrating modeling with in-situ measurements. Landscape and Urban Planning 167: 463–472.

84.

Yang

Ruby Leung

Zhao

, et al. (2017) Contribution of urbanization to the increase of extreme heat events in an urban agglomeration in east China. Geophysical Research Letters 44(13): 6940–6950.

85.

Zheng

Myint

Fan

(2014) Spatial configuration of anthropogenic land cover impacts on urban warming. Landscape and Urban Planning 130: 104–111.

86.

Zhou

Qian

, et al. (2014) Relationships between land cover and the surface urban heat island: Seasonal variability and effects of spatial and thematic resolution of land cover data on predicting land surface temperatures. Landscape Ecology 29(1): 153–167.

87.

Zhou

Weng

Gurney

, et al. (2012) Estimation of the relationship between remotely sensed anthropogenic heat discharge and building energy use. ISPRS Journal of Photogrammetry and Remote Sensing 67: 65–72.