Navigation – Plan du site

AccueilRubriquesCartographie, Imagerie, SIG2022Urban Design Guidelines for Impro...

2022
1032

Urban Design Guidelines for Improvement of Outdoor Thermal Comfort in Tropical Cities

Lignes directrices de conception urbaine pour l’amélioration du confort thermique extérieur dans les villes tropicales
Directrices de diseño urbano para la mejora del confort térmico exterior en Ciudades Tropicales
Lea A. Ruefenacht, Ayu S. Adelia, Juan A. Acero et Ido Nevat

Résumés

Le changement climatique et la hausse des températures sont l'un des plus grands défis pour les villes tropicales. Ces villes doivent être conçues dans une optique d'adaptation à une forte croissance démographique et d'offre d'un environnement thermique confortable. À cet égard, nous proposons un cadre pour l'élaboration de politiques de conception urbaine axées sur l'amélioration du confort thermique dans les environnements extérieurs, un concept connu sous le nom de Confort Thermique Extérieur (OTC). La proposition est fondée sur le calcul de variables climatiques telles que la température, la vitesse du vent, l'humidité, etc., pour diverses stratégies d'urbanisme à l'échelle du district. De plus, nous utilisons l'occupation spatiale pendant la journée sur une telle échelle pour estimer une note globale de performance OTC. Enfin, les directives de conception sont catégorisées et présentées dans un espace tridimensionnel. Les stratégies sont issues de la connaissance de l'urbanisme et de la conception et adaptées au contexte local et aux normes de construction. Les variables climatiques sont calculées à partir de modèles climatiques et du score de performance OTC à l'aide d'un modèle statistique de gestion des risques. Les directives discutées dans cet article se limitent aux stratégies de conception passive telles que la géométrie urbaine, la signalisation et la végétation pour les villes tropicales à haute densité. Ces lignes directrices peuvent être utiles aux architectes et aux planificateurs dans les premières étapes de la conception d'un nouveau quartier.

Haut de page

Texte intégral

1. Introduction

1The microclimate has been an influential aspect of traditional architecture and town planning. This is however not the case when it comes to modern urban planning (Eliasson, 2000). New building materials, construction methods, and technology allow for indoor climate modification in buildings, making them less dependent on outdoor conditions. As a result, the waste heat released from buildings is increased and subsequently alters the outdoor air temperature (Kikegawa, Genchi, Yoshikado, & Kondo, 2003). On the other hand, built structures hinder the wind flow in the city, which trap the heat and further increase the ambient air temperature (Yuan, 2018). For urban populations, this has serious implications on the thermal comfort and heat stress, which is predicted to increase significantly in the next decades if global warming continues to escalate at current rate (IPCC, 2019). Heat exposure in outdoor spaces can cause thermal discomfort and affect the behavior and productivity of the people living in urban areas. Rising temperatures can affect human health and increase the risks of heat-related morbidity and mortality, especially during the extreme weather conditions (e.g., Patz, Campbell-Lendrum, Holloway, & Foley, 2005; Uejio, Wilhelmi, Golden, Mills, Gulino & Samenow, 2011). While there is the increase of frequency and intensity of extreme weather events due to climate change, both thermal discomfort and heat stress are becoming great concerns for citieslivability.

2Outdoor Thermal Comfort (OTC) is described as the state of mind which expresses the satisfaction with the thermal environment, be it in an indoor or outdoor setting (ASHRAE, 2004). It was first developed for indoor environments by Fanger (1970), who described thermal comfort quantitatively by incorporating physical, physiological and psychological aspects of human biometeorology. The OTC, however, is more difficult to evaluate as compared to similar study done for indoor settings, due to the lack of thermal boundaries, intensive direct radiation, and high variability of wind speed (Höppe, 2002; Nazarian, Acero, & Norford, 2019). Furthermore, the vast range of activities in outdoor space, as well as the physiological and psychological state of each individual, make the OTC evaluation even more complex. Several indices are created in order to assess OTC level, such as the Predicted Mean Vote (PMV), the Physiological Equivalent Temperature (PET), the Standard Effective Temperature (SET), Universal Thermal Climate Index (UTCI), and Thermal Sensation Vote (TSV). Even though all of these indices have the same objective to describe the OTC, each of them has a different approach and level of details (Nazarian, Acero, & Norford, 2019).

3Peoples sensation of OTC is greatly influenced by the local microclimate in the lowest level of atmosphere, so called the Urban Canopy Layer (UCL). However, UCL consists of a variety of microclimates. These microclimates can be managed through purposeful design, with specific goals in mind, such as to provide thermally comfortable urban space (Oke, Mills, Christen, & Voogt, 2017). However, our understanding on how to achieve that is rather limited (de Schiller & Evans, 1990; Eliasson, 2000). To design a climate-responsive city, it requires a good understanding of the local climate and environment as design parameters could help us plan new climatically-sensitive developments or resolve the climate-related problems in existing ones (Looman, 2017). When it comes to urban climate and OTC analysis, both spatial and temporal variabilities need to be considered (Oke, 2006). OTC analysis also needs to take into account the dynamic movement and transient behavior of people outdoors which highly correlate with the social and cultural characteristics of the city residents (Nazarian, Acero, & Norford, 2019). All these complexities create a profusion of information, which can actually impede the process of climate-responsive planning as it can be overwhelming for the planners to handle and make them unable to take proper action in a timely manner (Ng, 2012). Also, as city layout, buildings and urban infrastructures typically last for decades, there is a strong need for a set of comprehensive and well-researched design guidelines.

4The integration of urban climatology in city planning started to emerge in the 1980s. Oke (1984) introduced several methods to establish a general relationship between urban climatology and planning principle, in the form of rules or guidelines. In the 1990s, the climatic analysis began to be implemented as a guide for modern urban planning. Some have been successfully incorporated in the real project by involving the climate scientists since the beginning of the project (e.g., Evans & de Schiller, 1996; Givoni, 1991). However, the use was unsystematic, hence the impact of climate knowledge in the planning process was relatively low (Eliasson, 2000). Eliasson (2000) highlighted the importance of cross-discipline communication techniques in order to transfer scientific knowledge successfully into applicable and user-friendly tools for planning professionals. Several studies emphasized at least three aspects to ensure a good knowledge transfer between the two disciplines (Klemm, 2018). First, the urban climate knowledge must be presented in a comprehensible way, ideally in a clear and simple graphical representation. Second, it must be applicable and easy to use during the design and planning processes. Third, it should be feasible to implement considering a multitude of practical constraints such as time and budget.

5In this paper we present a framework to assess the effects of urban configurations on OTC improvement. To this end, the framework combines climate information, design objectives, and mathematical metrics to quantify and rank each urban configuration. It is applied in the city-state of Singapore and examines the effectiveness of urban geometry and vegetation on the OTC of pedestrians in tropical high-density developments (Ruefenacht & Acero, 2017). Accordingly, the key research objectives are threefold: (1) what metrics can measure the effectiveness of strategies over space and time, (2) how can these metrics be aggregated into a framework to quantify the overall OTC performance of each strategy and scenario, and finally (3) how can these results be translated into easy-to-use guidelines for architects and urban designers.

6The effectiveness of the strategies is calculated by using computer simulations and statistical models, which incorporate concepts of climate science as well as risk and decision theory. Finally, the results are translated into climate-responsive guidelines or recommendations that can be used by decision-makers for planning new districts.

2. Methodology

2.1 Study area

7The study area is located in the city of Singapore in the south Malaysian peninsula. The climate of Singapore is classified as a tropical rainforest (Koppen climate classification Af), which is characterized by high temperatures and humidity during the whole year. The annual mean temperature is close to 27.5°C (with a low diurnal amplitude); the mean relative humidity can reach 80%, and the mean wind speed can reach 3 m/s (Roth & Chow, 2012). Furthermore, Singapores urban environment consists of densely built districts with mainly high-rise buildings. These large and densely built urban surfaces absorb the solar radiation and store it in the form of heat, having a negative impact on the OTC of pedestrians.

8The selected study area is a greenfield near the shore, located in the south of Singapore (Figure 1). The site is part of the existing Central Business District, and should become a high-density, mainly commercial, district. The new district is located near the waterfront, which allows the penetration of the sea breeze without major obstructions, especially during the southwest-monsoon and inter-monsoon seasons when the south-southwest (SSW) oriented winds are more prominent. This microclimatic condition differs from those areas in the north, east, west and centre of the island. Therefore, specific weather conditions (also called weather types) have been identified for the site.

9The site size is approximately 800 by 800 meters, with a predefined Gross Floor Area (GFA) of about 3.3 million square meters. The GFA is defined as the total area inside the building envelope. The maximum allowed building height is 280 meters (approximately 60 floors). The site is composed of buildings, streets, sidewalks, plazas, and parks.

Figure 1. Location of the selected study area: new district south-eastern shore of the Central Business District.

Figure 1. Location of the selected study area: new district south-eastern shore of the Central Business District.

2.2 Weather types

10With the aim of analyzing annual OTC levels, representative weather conditions for the study area were defined. Seven Weather Types (WTs) were calculated following a clustering procedure (statistical k-means method) described in Acero, Koh, Pignatta, & Norford (2020a). This approach used hourly meteorological data from 2016 (i.e., Air Temperature (AT), Relative Humidity (RH), Wind Speed (WS) and incoming solar radiation) from nearby meteorological stations in Singapore. We specify that the results represent the current” weather in Singapore. As an output, for each cluster with similar climate features, a diurnal mean cycle was calculated to represent the specific WT. Hourly series for AT and RH were obtained by averaging data in each cluster. The seven WTs obtained represent fairly well the climate context of Singapore, with small annual variations in weather variables (Chow & Roth, 2012).

11Table 1 shows the meteorological conditions describing the annual WTs for the region of study. These were used as boundary conditions in the modelling approach. WT0, WT1 and WT6 are representative of the northeast-monsoon season, WT2 and WT5 are representative of the SW-monsoon season, and WT3 and WT4 of the inter-monsoon seasons (Acero, Koh, Ruefenacht & Norford, 2021).

Table 1. Description of the model meteorological boundary configuration of each weather type (WT), (Acero, Koh, Ruefenacht, & Norford, 2021).

Weather conditions

Weather type

WT0

WT1

WT2

WT3

WT4

WT5

WT6

Frequency of occurrence

11.2%

15.85%

15.03%

16.12%

16.12%

11.48%

14.21%

Reference date for simulation

December 15th

January 15th

July 15th

April 15th

October 15th

June 15th

February 15th

Wind speed (10 m above ground level)

2.2 m/s

1.7 m/s

1.6 m/s

0.8 m/s

0.3 m/s

2.4 m/s

2.8 m/s

Wind direction

332º

320.7º

228.1º

140.8º

345º

297º

43.8º

Minimum air temperature

26.3ºC (6:00 LAT)

26ºC (6:00 LAT)

28ºC (6:00 LAT)

28.2ºC (6:00 LAT)

27.5ºC (6:00 LAT)

26.8ºC (6:00 LAT)

27ºC (6:00 LAT)

Maximum air temperature

28.1ºC (13:00 LAT)

30.3ºC (13:00 LAT)

31.6ºC (13:00 LAT)

33ºC (13:00 LAT)

30.9ºC (12:00 LAT)

33.9ºC (13:00 LAT)

32ºC (13:00 LAT)

Minimum relative humidity

78.3% (14:00 LAT)

68.6% (13:00 LAT)

62.4% (15:00 LAT)

61.9% (14:00 LAT)

70.4% (12:00 LAT)

56.9% (14:00 LAT)

57.3% (14:00 LAT)

Maximum relative humidity

87.1% (6:00 LAT)

89.2% (6:00 LAT)

82.8% (5:00 LAT)

85.7% (6:00 LAT)

86.9% (6:00 LAT)

83.1% (6:00 LAT)

84.1% (6:00 LAT)

Cloud cover (oktas)

7.5

7

5

4.5

6.5

5

4.5

Soil upper layer (0-20 cm) initial temperature

27.9ºC

28.5ºC

28.3ºC

29.8ºC

27.8ºC

28.6ºC

28.1ºC

Soil middle layer (20-50 cm) initial temperature

27.8ºC

28.5ºC

287C

29.9ºC

28.2ºC

28.8ºC

28.4ºC

Soil deep layer (>50 cm) initial temperature

28.1ºC

28.6ºC

288C

29.9ºC

28.5ºC

28.9ºC

28.5ºC

Soil upper layer (0-20 cm) moisture content

80%

80%

65%

60%

70%

65%

65%

Soil middle layer (20-50 cm) moisture content

70%

70%

55%

50%

60%

55%

55%

Soil deep layer (>50 cm) moisture content

60%

60%

45%

40%

50%

45%

45%

2.3 Strategies and scenarios

12The studied strategies are related to urban geometry and vegetation. In total 39 scenarios were evaluated, including (a) block and street orientation, (b) aspect ratio and street canyon width, (c) building height variation, (d) podium-tower typology, (e) height of elevated podium, (f) number of towers, (g) green coverage, and (h) tree planting location. The selection of these strategies is based on key urban design aspects for planning high-density districts as well as aspects that impact pedestrian OTC. The GFA (3.3 million square meters) and the maximum building height (280 meters) are the main constraints set in the master plan for the development of the selected study area.

2.4 CFD Simulations and validation

13To evaluate the impact of the urban design strategies on the quality of OTC we have used the ENVI-met model. Climatic outputs of the model (i.e., AT, RH, Mean Radiant Temperature (MRT) and WS) were used to calculate the Physiological Equivalent Temperature (PET) index (Höppe, 1999; Chen and Matzarakis, 2014) on an hourly basis. The PET is an index that estimates the thermal condition of a microclimate, and provides the effect of comfort or sensation on people. It is the equivalent air temperature required in an outdoor environment to reproduce a standardized indoor setting (Höppe, 1999), measured in degrees Celsius [°C]. The human metabolic rate and other human parameters were standardized (i.e., age: 35 years; height: 1.74 m.; metabolic rate: 80 w/m2; clothing: 0.9; weight: 75 kg; sex: male).

14ENVI-met model (version 4.3), with certain approximations, considers a complete radiation budget (i.e., direct, reflected, and diffuse solar radiation, as well as long-wave radiation). The airflow is solved with the standard k-turbulence closure model through the Reynolds Averaged Navier–Stokes (RANS) equations (Huttner, 2012; Bruse & Fleer, 1998). The model allows forcing climate variables along the simulation to account for the variation of meteorological variables along the day. However, only boundary conditions of AT and RH were updated during the diurnal simulation. WS, Wind Direction (WD) and cloudiness remained constant. In our study, a 3 by 3 meters horizontal resolution has been selected. On the vertical, the grid size was 2 meters in the five lowest levels (i.e., up to 10 meters above ground level). From this height upwards the grid increased in the vertical resolution with a specific telescoping factor. The vertical extension of the model depended on the scenario.

15The urban design strategies consider different building heights, street canyon widths, and sizes of open or green areas, while keeping the GFA constant. Each urban design scenario was run with the seven clustered WTs described in section 2.2. The results for each WT were factored and annual OTC levels were estimated. Figure 2 shows the definition of the study area. The size of study area (800 x 800 meters) described in section 2.1 remains constant in all strategies and scenarios. For each urban design scenario (see section 3) the study area corresponds to the inner 75% of the area covered by the simulated buildings (Figure 2) and is used to compare the results of each scenario. This way, we avoid the effects of the border of the domain and increase representativeness of the results (Acero, Koh, Ruefenacht & Norford, 2021).

16Although a specific and detailed analysis of the performance of the model was not undertaken, sensor measurements were carried out in a nearby high-rise area (Acero, Koh & Tan, 2020b). The analysis of Acero (2020b) showed that the results provided by the model in high-rise areas were comparable to the ones registered by the measurements.

Figure 2. Example of a simulated domain covered by buildings, and the study area (800 x 800 meters) defined by the inner part of the dotted square (corresponding to the 75% of the area covered by buildings), (modified from Acero, Koh, Ruefenacht & Norford, 2021).

Figure 2. Example of a simulated domain covered by buildings, and the study area (800 x 800 meters) defined by the inner part of the dotted square (corresponding to the 75% of the area covered by buildings), (modified from Acero, Koh, Ruefenacht & Norford, 2021).

2.5 Performance Metrics

17In order to evaluate the quality (or goodness”) of urban design strategies from an OTC perspective, we developed a mathematical framework. The framework (Figure 3) is based on aggregating a performance metric over the spatial-temporal domains (Nevat, Ruefenacht & Aydt, 2020a; Nevat, Pignatta, Ruefenacht & Acero, 2020b; Nevat, 2021). The framework combines various design criteria (i.e., urban geometry), climate conditions (i.e., weather types), and the specific usage of urban areas (i.e., exposure map and time interval) in order to quantify the overall OTC performance (i.e., PET) of a district. The OTC assessment results in aggregated spatial-temporal metrics, which help evaluate the impact of individual design strategies on OTC. Finally, each design strategy receives an overall score, which reflects its overall quality. The design with the lowest score has the highest quality, and thus, ranked as the best solution.

18To this end, three different intuitive and easy-to-understand performance metrics are introduced. These metrics provide a summary of the OTC performance in both space and time. These are the mean, upper-semi-deviation, and coverage of values metrics. In addition, in order to better understand the contribution of each of the climatic variables to the score, a marginal analysis is presented, which measures the contribution of each climate variable (i.e., MRT, AT, RH, and WS) to the final score. This analysis allows the decision-maker to better understand, and isolate the impact of different climate variables on the score.

19For the spatial-temporal analysis two criteria were used to define the specific usage of urban areas. The first is the time interval, which represents the relevance of each location in time. The selected time interval for this analysis is the afternoon (11-16h), when people are more exposed to heat stress. The second is the exposure map, which represents the importance” of each location in space. The term importance” means how each location is valued (or used) by the user. For example, streets and parking lots are considered as unimportant since people are not expected to spend time there. In contrast, plazas, parks, sports fields or playgrounds are considered by the user as having high importance. Walkways or sidewalks are considered by the user as medium important.

20The metrics used for the assessment of the OTC performance are defined below.

Figure 3. A flowchart of the framework

Figure 3. A flowchart of the framework

2.5.1 Mean performance metric

21The mean performance metric captures the central tendency of the spatial-temporal PET of a specific scenario. The lower the mean PET, the better is the OTC in space and time.

22This average performance metric μ for the i-th climate variable, j-th exposure value, and w-th weather type, is given by:

where Z is evaluation of the climate variable (e.g., PET), over the whole spatial domain X, for a single weather type W, within the predefined time interval T, and for each exposure area E.

2.5.2 Semi-deviation metric

23The upper/lower semi-deviation metrics capture the variation of PET above/below the mean performance of a specific scenario. The smaller the deviation, the lower is the spread of PET. This metric is more sensible than the commonly used variance metric which captures the spread of the whole data-set. In contrast, semi-deviation captures the deviation above/below a reference point (e.g., the mean). In our case, the deviation above the mean value should be discouraged and therefore our metric penalizes for large values of upper semi-deviation.

24The upper semi-deviation metric σ+ for the i-th climate variable, j-th exposure value, and w-th weather type, is given by:

where Z is evaluation of the climate variable (e.g., PET), over the whole spatial domain X, for a single weather type W, within the predefined time interval T, and for each exposure area E.

2.5.3 Coverage of values metric

25The coverage metric captures the range of PET in which 99% falls inside a specific interval. The smaller the coverage, the lower is the span between the maximum and minimum PET.

26The coverage metric Δ for the i-th climate variable, j-th exposure value, and w-th weather type, is given by:

where ρ(Zi|E(x,t) = j, τh) is the τ -th quantile level of dataset Zi and extract only the values corresponding to the j-th exposure values. The high quantile is denoted τh (0.95), and the low quantile is denoted τl (0.05).

2.5.4 Performance score

27The overall performance score S combines the mean performance μ, the semi-deviation σ+, and the coverage Δ metrics of the PET into a single score. The lower the metrics, and thus, the lower the overall score, the higher is the scenario's OTC ranking.

28The score S for the i-th climate variable, and w-th weather type, is given by:

29On the one hand, the exposure of streets receives a weight of 0.0 (β1), and the exposure of sidewalks receives 1.0 (β2). For strategies / scenarios that also have a plaza, park or green scape, the exposure of sidewalks receives a weight of 0.5 (β2), and the exposure of plazas, parks, and green spaces receives 0.5 (β3). On the other hand, the metric of mean performance μ receives a weight of 0.5 (ω1), the metric of semi-deviation σ+ receives 0.25 (ω2), and the metric of coverage Δ receives 0.25 (ω3).

2.5.5 Performance ranking

30The ranking portrays the overall quality of the OTC performance (PET is defined in section 2.3). If the performance score is below 21 the scenario is ranked "good", between 21 and 22 it is ranked "ok", and above 22 it is ranked "bad".

2.5.6 Sensitivity analysis of the PET index

31Sensitivity analysis is a process that quantifies the impact of each climate variable separately (marginally) on the PET value. This analysis is important, in order to understand the impact that each climate variable has on the resulting PET in a simple and intuitive way. Therefore, the marginal impact factor of each climate variable is calculated by finding the correlation between the PET value, and the values of the different climate variables.

32To achieve that we used a multiple linear regression model, where each climate variable acts as a covariate (explanatory variable) and multiplied by a corresponding weight. The weights are estimated via Least Squares (LS) where the ENVI-met PET values (response variables) are used. Then the impact of each climatic variable on the PET value is captured by the corresponding marginal impact factor (weights). Based on the ENVI-met data, the resulting marginal impact factor for PET is 1.0, MRT is 0.4, AT is 0.9, RH is 0.1, and WS is -2.7. This type of analysis is important to decision makers as it informs them how to better allocate their resources in order to improve the PET levels.

2.5.7 Excess score

33The excess score captures the difference of the mean OTC performance of a scenario in comparison to the baseline. This score shows if the PET or any other climate variable has improved or deteriorated compared to a baseline.

34This analysis is used based on the sensitivity analysis, thus allowing one to understand how each climatic variable improves or deteriorates in comparison to the baseline. If the excess score is positive the average performance has improved. If the excess score is negative the average performance has deteriorated. This analysis provides the decision maker the ability to understand how each urban design performs (better or worse) for each of the climatic variables separately.

35The excess score C for the i-th climate variable, j-th exposure value, and w-th weather type, is given by:

where αi is the marginal factor of the i-th climate variable, and μ is the mean performance metric of the i-th climate variable, j-th exposure value, and w-th weather type. μbl is the performance metric of the baseline, while μsc of the comparison scenario. Each performance μ is weighted by a marginal impact factor α. The factor for PET (αpet) is 1.0, MRT (αmrt) is 0.4, AT (αat) is 0.9, RH (αrh) is 0.1, and WS (αws) is -2.7.

3. Results and discussion

36The urban design strategies presented in this paper are limited to urban geometry and vegetation, which are relevant to the planning and design of high-density districts, and have potential impacts on pedestrian OTC. Each strategy consist of a baseline and a set of possible scenarios. The baseline and each scenario are an alternative application of a strategy. The baseline is used as the initial scenario for comparison. The GFA used for the scenario comparison is kept constant. The results presented are based on the analysis of the hottest period of the year and day, where extreme heat stress is more likely to occur. That is the month of April (also called weather type 3), and the afternoon hours between 11h and 16h. Three exposure areas are considered in this framework, such as sidewalks, plazas, and green areas/spaces.

3.1 Urban geometry

3.1.1 Block and street orientation

37The orientation of building blocks and streets mainly influences the wind speed and direction, and the period of maximum solar access (Acero, Koh, Ruefenacht, & Norford, 2021). In order to maximize the wind penetration, buildings should be arranged parallel or oblique to the prevailing wind direction. In this paper, block and street orientation refers to the position of buildings and streets in relation to a specific geographic direction. The building and street dimensions used in this study remain constant, as well as the aspect ratio of 2.5.

38As shown in Figure 5, we analyzed the results of four scenarios with different orientations: north-south (N-S), northeast-southwest (NE-SW), northwest-southeast (NW-SE), and east-west (E-W). The N-S orientation is chosen as the baseline.

39The N-S orientation is the recommended orientation in terms of the overall OTC performance score during hot days of the inter-monsoon season (WT3). This orientation shows up to 0.8 m/s higher mean WS along the street canyons, hence 0.9°C lower mean PET than the NE-SW orientation, which is perpendicular to the prevailing wind direction in WT3 (Table 1). Figure 4 shows that 42% of the sidewalk area along the long street canyon (i.e., north-south axis) has the lowest PET of 34.3°C. Slightly different tendency is observed for scenarios with NE-SW, NW-SE, and E-S orientation. The lowest PET of 34.3°C for NW-SE orientation is found also at the sidewalk along the long street canyon, however, only in 28% of the area.

40The N-S orientation is recommended, however, from a mean annual perspective (factoring the influence of all WTs) there is little difference between the different orientations. The overall OTC score of the N-S orientation is 20.3, thus classified as good performance, and of the NW-SE orientation is 22.3, thus classified as bad performance. Other literature examined the effects block and street orientation, and found that the N-S orientation has better OTC than the NW-SE; however, the differences between the different orientations are small (Yang, Wong & Li, 2016).

Figure 4. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., north-south orientation) and scenarios (i.e., northeast-southwest, northwest-southeast, and east-west orientations).

Figure 4. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., north-south orientation) and scenarios (i.e., northeast-southwest, northwest-southeast, and east-west orientations).

Figure 5. Scenarios with different orientations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 5. Scenarios with different orientations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

3.1.2 Aspect ratio and street canyon width

41The aspect ratio and street canyon width influence the solar access into the canyon and the capacity of heat storage (Acero, Koh, Ruefenacht, & Norford, 2021). For high-rise districts, a high aspect ratio and narrow canyon width is preferred in general as it reduces the incoming solar radiation. In this paper, aspect ratio refers to the ratio between the canyon height (i.e., building height) and the canyon width (i.e., street width); while the street canyon width refers to the maximum horizontal distance between two buildings. The building and street dimensions used in this study vary; a street width that ranges from 15 meters to 45 meters, and the aspect ratio that ranges from 1.5 (i.e., low aspect ratio) to 4.5 (i.e., high aspect ratio).

42As shown in Figure 7, we analyzed the results of four scenarios with different aspect ratios: low aspect ratio (1.5) and high aspect ratio (4.5) along the long street canyon (i.e., north-south axis), and low aspect ratio (1.5) and high aspect ratio (4.5) along the short street canyon (i.e., east-west axis). Low aspect ratio along the long street canyon is chosen as the baseline.

43High aspect ratio (4.5) along both short (i.e., E-W oriented) and long street (i.e., N-S oriented) canyons are the recommended ratios in terms of the overall OTC performance score. Also, low aspect ratio (1.5) along the long street canyon (i.e., N-S oriented) provides good OTC mean results due to the effect of building shade throughout the midday afternoon hours. High aspect ratio can reduce mean MRT and PET (between 11-16h) up to 1.6°C and 1.5°C respectively with respect to low aspect ratio (1.5) E-W oriented. For the scenario of high aspect ratio (4.5) along the long canyon, Figure 6 shows that 39% of the sidewalk area along the long street canyon has the lowest PET of 34.5°C. Low aspect ratio (1.5) along the short street canyon (i.e., E-W oriented) is not recommended. For this scenario, Figure 6 shows that only 11% of the sidewalk area along the short street canyon has the lowest PET of 35.1°C. The overall PET score of high aspect ratio (4.5) along the long street canyon is 20.4, thus classified as good performance, and of low aspect ratio (1.5) along the short street canyon is 22.1, thus classified as bad performance.

44To assess the effects of the street width, a scenario with a wide street (45 meters) and a narrow street (15 meters) are compared. In terms of PET, the narrow street canyon performs better as it blocks the high incoming solar radiation available in Singapore. Even though the mean WS of a narrow street canyon reduces up to 0.5 m/s, the mean MRT is up to 1.6°C lower than of a wide street canyon (Figure 8). For long street canyons that are oriented N-S, both a narrow and wide street width are recommended; while for short street canyons that are oriented E-W, the wide street width should be avoided.

45In general, high aspect ratio (4.5) along both short (i.e., E-W oriented) and long street (i.e., N-S oriented) canyons, as well as narrow street canyons, are recommended. These results coincide with other studies conducted in Singapore (Yang, Wong & Lin, 2015) as well as in other latitudes (Ali-Toudert & Mayer, 2006; Muniz-Gäal, Pezzuto, Carvalho & Mota, 2020).

Figure 6. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., low aspect ratio along the long street canyon) and scenarios (i.e., high aspect ratio along the long street canyon, low aspect ratio along the short street canyon, and high aspect ratio along the short street canyon).

Figure 6. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., low aspect ratio along the long street canyon) and scenarios (i.e., high aspect ratio along the long street canyon, low aspect ratio along the short street canyon, and high aspect ratio along the short street canyon).

Figure 7. Scenarios with different aspect ratios (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 7. Scenarios with different aspect ratios (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 8. Scenarios with different street canyon widths (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 8. Scenarios with different street canyon widths (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

3.1.3 Building height variation

46The building height has a decisive influence on the shading that reaches pedestrian areas (Acero, Koh, Ruefenacht, & Norford, 2021). High-rise buildings can maximize shading, while low-rise buildings minimize wind-blocking. In this paper, building height refers to the maximum vertical point of a building, and building height variation refers to the elevation profile of multiple building heights, ranging from low- to high-rise. In the symmetric height variation profile, the buildings along the long street canyon (i.e., north-south axis) have the same height, while in the asymmetric the buildings have different heights. The building and street dimensions used in this study vary, with constant street width, and a building height that ranges from lower than 35 meters to higher than 60 meters.

47As shown in Figure 9, we analyzed the results of four scenarios with different height variations and profile symmetries: symmetric low-high-low and high-low-high, and asymmetric low-high-low and high-low-high. The symmetric high-low-high is chosen as the baseline.

48In Singapore, all building height variations provide good OTC performance. Also, from an annual perspective there is not much difference between the different height profiles. For the four scenarios, Figure 11 shows that between 21% and 36% of the sidewalk area along the long street canyon has the lowest PET between 34.3°C and 34.5°C. The overall PET score of all four scenarios is similar (between 20.4 and 20.6), thus classified as good performance. Results are highly conditioned by the shadows cast produced by the high-rise buildings.

49To assess the effects of the building height, high-rise (>60 meters), mid-rise (35-60 meters) and low-rise (<35 meters) buildings are compared. In terms of PET, high-rise buildings perform better as the shade coverage is larger, hence the mean MRT is up to 1.6°C lower and the mean PET is up to 2.3°C lower than for low-rise buildings (Figure 10). However, a good balance between high-rise and low-rise buildings should be achieved to maximize both shade and the entry of sea-breezes. High-rise buildings located along the coast and waterfront should be avoided.

50In general, high-rise buildings and all building height variations are recommended. Other literature evaluated the effects of building height variability, and concluded that high variability of building heights within the urban canyon can create a network of shaded spaces, improve the wind velocity, and ultimately reduce the temperature (Sharmin, Steemers & Matzarakis, 2015; Priyadarsini, Wong & Cheong, 2008).

Figure 9. Scenarios with different height variations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 9. Scenarios with different height variations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 10. Scenarios with different building heights (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 10. Scenarios with different building heights (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 11. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., symmetric high low-high) and scenarios (i.e., symmetric low-high-low, asymmetric high-low-high, and asymmetric low-high-low).

Figure 11. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., symmetric high low-high) and scenarios (i.e., symmetric low-high-low, asymmetric high-low-high, and asymmetric low-high-low).

3.1.4 Podium-tower typology

51A building composed of a tower on a podium is a commonly used building typology in high-rise cities, such as Hong Kong, New York and Singapore (Lau & Zhang, 2015; Chang & Winter, 2015). The large footprint of a podium combined with a smaller footprint of a tower provide sufficient shading of pedestrian areas. In this paper, a tower-podium building refers to a building typology of a podium located on the ground and one or more towers arranged on the podium. The building and podium dimension as well as the street width used in this study remains constant. The podium height is 15 meters and the tower height is 150 meters.

52As shown in Figure 12, we analyzed the results of three scenarios with different podium configurations: podium with single tower, stepped podium with single tower, single tower without podium, and elevated podium with single tower. The podium with a single tower is chosen as the baseline.

53The podium with a single tower is recommended in terms of the overall OTC performance score. This scenario shows up to 3.4°C lower mean MRT along the street canyons, hence 4.3°C lower mean PET than the single tower without podium. Also, it shows up to 1.7°C lower PET than the stepped podium. However, the best OTC is provided by the single tower with an elevated podium. For the podium with a single tower, Figure 13 shows that 11% of the sidewalk area along the long street canyon (i.e., north-south axis) has the lowest PET of 34.8°C. While for the stepped podium and tower without podium, 9% and 7% of the sidewalk area has the lowest PET of 35.5°C and 37.5°C respectively. The overall PET score of the tower-podium typology is 21.1, thus classified as good performance; while of the stepped podium and no podium is 22.0 and 23.7 respectively, thus classified as bad performance.

54The podium-tower typology is recommended; however, the stepped or terraced podium should be avoided. Other studies on podium-tower building typologies and their impact on OTC were carried out in Hong Kong (Wai, Yuan, Lai & Yu, 2020; Chen & Mak, 2021). Similarly, they concluded that the podium-tower typology with the narrow side facing the prevailing wind direction, and elevated (or lift-up”) from the ground provide good OTC, mainly due to the increase in porosity and shading.

Figure 12. Scenarios with different podium-tower typologies (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 12. Scenarios with different podium-tower typologies (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 13. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with single tower) and scenarios (i.e., stepped podium with single tower, and single tower without podium).

Figure 13. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with single tower) and scenarios (i.e., stepped podium with single tower, and single tower without podium).

3.1.5 Height of elevated podium

55Elevating the podium from the ground creates additional semi-outdoor spaces at pedestrian level, allowing higher wind penetration. As presented in different studies, semi-outdoor spaces play an important role to improve thermal comfort in tropical cities like Singapore (e.g., Gamero-Salinas, Kishnani, Monge-Barrio, Lopez-Fidalgo, & Sanchez-Ostiz, 2021). In this paper, an elevated tower-podium refers to a building typology of a podium elevated from the ground. The podium, tower and street dimensions used in this study remain constant. The podium height is 15 meters and the tower height is 150 meters. The podium is elevated between 5 to 20 meters from the ground.

56As shown in Figure 14, we analyzed the results of four scenarios with different podium elevations: podium 5 meters, 10 meters, 15 meters and 20 meters elevated from the ground. The podium 5 meters elevated from the ground is chosen as the baseline.

57All four scenarios provide good OTC performance. However, compared to a podium on the ground the elevated podium shows up to 2.9°C higher mean PET. For the four scenarios, Figure 15 shows that between 24% and 29% of the covered plaza has the lowest PET between 31.5°C and 32.3°C. Also, between 12% and 30% of the sidewalk area along the long street canyon has the lowest PET between 33.0°C and 33.2°C. The overall PET score of all four scenarios is between 18.6 and 18.9, thus classified as good performance. Even though the OTC performance of all four scenarios is good, the 20 meters elevation provides up to 0.7°C lower mean PET on the sidewalks and covered plazas compared to the 5 meters elevation. Elevated structures allow higher penetration of wind, therefore large obstructions at the pedestrian level should be avoided.

58All podium elevations are recommended. Other literature quantified the benefits of semi-outdoor spaces by comparing them with conventional outdoors spaces in Singapore, and found that the PET differences can reach 10-16°C (Acero, Ruefenacht, Koh, Tan, & Norford, 2022).

Figure 14. Scenarios with different podium elevations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 14. Scenarios with different podium elevations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 15. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 5 meter elevated podium) and scenarios (i.e., 10 meter elevated podium, 15 meter elevated podium, and 20 meter elevated podium).

Figure 15. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 5 meter elevated podium) and scenarios (i.e., 10 meter elevated podium, 15 meter elevated podium, and 20 meter elevated podium).

3.1.6 Number of towers

59As discussed in the previous point, the elevated podium-tower typology provides good levels of OTC. However, multiple towers instead of a single tower on an elevated podium can provide additional shade, especially on the sidewalks. In this paper, the multiple tower-podium typology refers to one to four towers arranged at the center or at the edge of an elevated podium. The podium dimension and street width used in this study remains constant, while the tower height and footprint size vary. The podium height is 15 meters and the tower height ranges between 150 and 160 meters.

60As shown in Figure 17, we also analyzed the results of three scenarios with elevated podium and different tower configurations: podium with one tower, podium with two towers, and podium with four towers. The podium with one tower is chosen as the baseline.

61The elevated podium with four towers is recommended in terms of the overall OTC performance score. This scenario shows up to 3.8°C lower mean MRT along the street canyons, hence 4.0°C lower mean PET than the podium with one tower. Figure 16 shows that 55% of the covered plaza under the podium has the lowest PET of 29.0°C. Slightly different tendency is observed for the scenario with two towers, where 44% of the covered plaza has the lowest PET of 31.8 °C. The overall PET score of all scenarios is between 16.1 for the podium with four towers, and 18.9 for the podium with one tower, this classified as good performance. Even though all scenarios are recommended, the larger the number of towers on a podium, and the smaller the tower footprint, the better the OTC.

62To assess the effects of the tower arrangement, a scenario of one tower located at the podium centre is compared with a scenario of two towers located at the podium edge. In terms of PET, the second scenario performs slightly better as the shade coverage is larger, hence the mean MRT is up to 1.1°C lower and the mean PET is up to 0.5°C lower than for the first scenario (Figure 17).

63Multiple towers with a small footprint arranged at the edge of the podium are recommended. It creates a semi-outdoor space with a more uniform shade distribution than a single-tower configuration, up to a point so that the breezeway is not blocked. The importance of these semi-outdoor spaces in a tropical city and the metrics to evaluate their impact on OTC are described in other studies that align well with our finding (Tao, Lau, Gou, Zhang & Tablada, 2019; Gamero-Salinas, Kishnani, Monge-Barrio, Lopez-Fidalgo & Sanchez-Ostiz, 2021).

Figure 16. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with one tower) and scenarios (i.e., podium with two towers, and podium with four towers).

Figure 16. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with one tower) and scenarios (i.e., podium with two towers, and podium with four towers).

Figure 17. Scenarios with different tower configurations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 17. Scenarios with different tower configurations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

3.2 Vegetation

3.2.1 Green coverage

64Parks or green open spaces in the city are known to have positive impacts on providing a pleasant outdoor experience at precinct level and larger (e.g., Oke, Mills, Christen & Voogt, 2017). Green coverage is one of the common metrics used to describe the size of green open space in design practice. In this paper, green coverage refers to the plan area fraction of open space covered with trees over the total site area. All the trees used in this study are identical, with uniform height of 8 meters and crown diameter of 9 meters, which represent one of the common tree typologies found in Singapore. Table 2 shows the tree configurations and properties used in this study.

Table 2. Tree properties and settings.

Tree properties

Settings

Height

8 meters

Width

9 meters

Leaf type

Deciduous leaf

Foliage shortwave albedo

0.15

Leaf weight

100 gr/m2

Depth of roots

2.33 meters

Diameter of roots

10 meters

65As shown in Figure 19, we analyzed the results of four scenarios with different green coverage ratios: 0%, 40%, 60% and 85% coverage. The green cover of 40% presents the current green cover percentage of Singapore in overall, and it is chosen as the baseline.

66The green coverage ratio of 40% to 60% is the recommended ratio in terms of the overall OTC performance score. This ratio provides a good balance between the shading created by the trees and buildings, causing up to 0.9°C lower mean MRT, hence up to 1.6 °C lower mean PET than the green coverage ratio of 85%. This can be seen on Figure 18, 32% of the green area along the long street canyon (i.e., north-south axis) and 28% of the same location have the lowest PET of 34.5°C for the 40% and 60% green coverage scenario respectively. Slightly different tendency is observed for scenarios with 85% green coverage. The lowest PET is also found in the green area along the long axis, however, there is 1°C increase of PET and it covers only 11% of the area. As the building footprint is reduced drastically, the building height increases, creating a different building typology of a point block structure. Unlike in the previous scenarios, as the footprint of the point block is smaller, the shaded area during the mid-day reduces as compared to the previous scenarios. This resulted in a rise in the overall PET score of 22.1, higher than green coverage of 40% and 60%, with the overall PET score 20.5 and 20.8 respectively.

67To assess the effects of trees, scenario 0% is tested. In this scenario, the building footprint covers 60% of the site, while the remaining 40% is open space without trees. It is then compared with a scenario with 60% built-up area and 40% green coverage with trees. Without the shading from the trees, the mean MRT is increased up to 1.8°C and the mean PET up to 1.7°C at the green areas, both along the north-east and east-west axis. Even though the wind speed is improving due to less obstruction by the trees, the PET overall score is as high as 22.0, almost the same as green coverage of 85%.

68In terms of providing shade, the effect of trees can reach up to 1.7°C PET while buildings can reach up to 6.0°C PET. However, the OTC performance varies between tree species. The shade of trees can compensate for areas that are not covered by the building shadow cast. Therefore, achieving a good balance of large elements like buildings and natural elements like trees and vegetation in a design is necessary to create a good OTC.

69Other studies also concluded that street trees and urban parks in dense developments provide local cooling at pedestrian level (Jamei, Rajagopalan, Seyedmahmoudian & Jamei, 2016; Alexandri & Jones, 2008; Kong, Lau, Yuan, Chen, Xu, Ren & Ng, 2017). Melnikov, Christopoulos, Krzhizhanovskaya, Lees & Sloot (2022) investigated the effects of sun exposure on path choice in Singapore, and found that pedestrians preferred the shade cast by buildings to that cast by trees.

Figure 18. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 40% green coverage) and scenarios (i.e., 0% green coverage, 60% green coverage, and 85% green coverage).

Figure 18. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 40% green coverage) and scenarios (i.e., 0% green coverage, 60% green coverage, and 85% green coverage).

Figure 19. Scenarios with different ratios of green coverage and building footprint (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 19. Scenarios with different ratios of green coverage and building footprint (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

3.2.2 Tree planting location

70As discussed in the previous point, trees can provide sufficient shade for areas that are not getting sufficient shade by the buildings. However, adding too many trees can also create some disadvantages in some cases, such as wind and visual blockage, higher levels of humidity, and high maintenance cost. Hence, it is important to plant the trees strategically to avoid such negative effects.

71As shown in Figure 20, we tested and compared four different scenarios with different trees planting locations: north, south, east and west frontage. The ratio of building coverage, tree species and size remain constant in all scenarios.

72Amongst all locations, trees located along the east façade showed the best OTC performance score with 19.5, followed by west and south façade, while the trees at the north façade are the worst with 20.3 in terms of OTC performance score. With the trees lined at the east frontage, the PET can be reduced as much as 1.1°C, as compared to trees lined at the north side at 11-16h in the afternoon, due to lower MRT. This result might be varied for different seasons and times of the day as the sun position changes.

73Similar conclusion is obtained from the vertical greenery study done by Acero, Koh, Li, Ruefenacht, Pignatta, & Norford (2019). It is reported that the highest PET reduction occurs in the area near the east-facing facade with a 10-meter vertical green wall (< 3 meters radius) during daytime. This experiment, however, is limited to a single building arrangement and plant configuration. Certainly, it needs further analysis and explorations with different building and plant placement permutations.

Figure 20. Scenarios with different ratios of tree locations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

Figure 20. Scenarios with different ratios of tree locations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).

4. Conclusion

74The evaluation of various passive strategies to improve the OTC in tropical cities with high density has been translated into applicable climate-informed design guidelines. These guidelines aim to help architects and urban planners to design and evaluate new districts with climate as a central design criterion. There are examples in the literature and in practice of integrating urban climatology into urban planning. Compared to existing literature, this paper presents a systematic and comprehensive approach in terms of (1) evaluation of multiple passive strategies, (2) aggregation of OTC performance metrics into a single score, and (3) translation of results into easy-to-use guidelines.

75The design guidelines were developed through the application of a mathematical framework based on spatial and temporal analysis as well as risk metrics. The framework evaluates the potential effects of design strategies on multiple climate variables (i.e., MRT, AT, RH, WS), and applies the specific usage of urban areas to quantify an overall OTC performance score. The score incorporates the mean, positive deviation, and coverage of every climate variable. The lowest score represents the best solution, while the highest score represents the worst solution. In addition, this paper presents a marginal analysis of each climate variable to better understand their impact on the overall score.

76The results presented in this paper are tailored to the urban context and climate conditions of the southern shore of Singapore. The results are also based on the analysis of the hottest period of the year and day, that is the month of April (also called weather type 3), and the afternoon hours between 11h and 16h. The exposure areas considered in this study are the sidewalks, plazas, and green areas. Below, we summarize the effectiveness of each studied strategy.

(a) Block and street orientation

77Four strategy scenarios with different orientations were studied, such as north-south (N-S), northeast-southwest (NE-SW), northwest-southeast (NW-SE), and east-west (E-W). There is no significant difference between the various orientations in terms of the OTC performance. However, for the southern shore area of Singapore, the N-S orientation is recommended and the NW-SE orientation is not recommended.

(b) Aspect ratio and street canyon width

78Four strategy scenarios with different aspect ratios were studied, these are low aspect ratio (1.5) and high aspect ratio (4.5) along the long street canyon (i.e., north-south axis), and low aspect ratio (1.5) and high aspect ratio (4.5) along the short street canyon (i.e., east-west axis). For the southern shore area of Singapore, high aspect ratio (4.5) along both short and long street canyons provide the best OTC performance. However, low aspect ratio (1.5) along the long street canyon oriented north-south could also be considered due to higher levels of shadow from buildings. The low aspect ratio (1.5) along the short street canyon oriented east-west is not recommended. In terms of street canyon width, narrow street canyons (15-25 meters) are recommended. Wide street canyons (35-45 meters) oriented east-west should be avoided.

(c) Building height variation

79Four strategy scenarios with different height variations and profile symmetries were studied, such as asymmetric low-high-low and high-low-high, and symmetric low-high-low and high-low-high. There is no significant difference between the various height variations in terms of the OTC performance. For the southern shore area of Singapore, the symmetric low-high-low height profile is recommended. In terms of building height, high-rise buildings provide the best OTC performance. However, a good balance between high-rise, mid-rise, and low-rise buildings should be achieved to maximize both shade and the entry of sea-breezes. High-rise buildings located along the shore should be avoided.

(d) Podium-tower typology

80Four strategy scenarios with different podium configurations were studied. These are podium with a single tower, stepped podium with a single tower, single tower without a podium, and elevated podium with single tower. For the southern shore area of Singapore, the podium-tower typology is in general recommended. However, the stepped or terraced podium form should be avoided. The elevated podium with a single tower provides the best OTC performance.

(e) Height of elevated podium

81Four strategy scenarios with different podium elevations were studied, such as a podium 5 meters, 10 meters, 15 meters, and 20 meters elevated from the ground. There is no significant difference between the various elevations in terms of the OTC performance. However, the 20 meters elevation is recommended as the wind penetration at pedestrian areas is higher. In comparison to the podium on the ground, the elevated podium provides the best OTC performance. Therefore, large obstructions at the pedestrian level should be avoided.

(f) Number of towers

82Three tower configurations were studied, such as podium with one tower, podium with two towers, and podium with four towers. For the shore area, multiple towers (i.e., two or four towers) with a small footprint on a podium provide the best OTC performance. In terms of tower arrangement, it is recommended to arrange the towers at the edge or corner instead of at the center of the podium, as the shade coverage at pedestrian areas is larger.

(g) Green coverage

83Four strategy scenarios with different green coverage ratios were studied, such as 0%, 40%, 60% and 85% coverage. For the southern shore of Singapore, 40% to 60% green coverage is recommended as it creates a good balance between the shading created by trees and buildings. Higher ratios, such as 85% green coverage, is not recommended. In terms of providing shade, trees can improve the OTC by approximately 2°C PET (based on the physiology of the species selected), while buildings by 6.0°C (as mean afternoon value). Therefore, achieving a good balance of large elements like buildings and natural elements like trees in a design is necessary to create a good OTC at pedestrian level.

(h) Tree planting location

84Four strategy scenarios with different tree planting locations were studied, such as north, south, east and west frontage. There is no significant difference between the various locations in terms of the OTC performance. However, trees located along the east façade show the best OTC performance score, followed by west, south, and north façades. The highest cooling effect of trees can be perceived in close proximity to them.

85Figure 21 and Annex 1 (Table 3) present the mean, semi-deviation, and coverage metric values, as well as the OTC performance score and ranking of each strategy and scenario presented in this paper. A technical report has been published presenting the results of this study in detail (Ruefenacht, Adelia, Acero & Nevat, 2020).

86This set of guidelines is a practical tool for improving OTC of existing neighborhoods, and for planning new districts, especially in urban areas with tropical climates, with high density and high human activity.

Figure 21. The PET performance score of each strategy and scenario.

Figure 21. The PET performance score of each strategy and scenario.

5. Future Work

87The guidelines presented are limited to tropical cities, like Singapore, and more specifically to microclimatic conditions of the southern shore area of Singapore. Future research will address other microclimatic locations in Singapore, such as central areas, as well as northern, eastern, and western shore areas. However, the framework presented can be applied to develop climate-informed design guidelines for different cities around the world.

88The results are limited to urban geometry and vegetation and do not consider the effects of traffic or other anthropogenic heat sources. However, there is evidence that the influence of the urban form on OTC is relatively large because of the high capacity of urban surfaces to absorb solar radiation and store it as heat (Singh, Mughal, Martilli, Acero, Ivanchev & Norford, 2022).

89This paper focused on the analysis of the impact of single strategies and scenarios. In the future, the effects of combined strategies will be assessed in order to take into account the synergies and trade-offs of the strategies in terms of OTC. Furthermore, this paper presented passive strategies related to urban geometry and vegetation. However, further passive strategies related to urban geometry (e.g., wind corridors), vegetation (e.g., large scale vs small scale urban parks), and surface materials, as well as active strategies (e.g., electric vehicles, district cooling plants) will be assessed.

90In addition, future publications will present the analysis of each strategy presented in this paper in-depth. For example, the strategies related to block orientation, aspect ratio, and building height have been published in Acero, Koh, Ruefenacht, & Norford (2021).

6. Acknowledgements

91The research was conducted under the Cooling Singapore project, funded by Singapores National Research Foundation (NRF) under the Virtual Singapore programme. Cooling Singapore is a collaborative project led by the Singapore-ETH Centre (SEC), with the Singapore-MIT Alliance for Research and Technology (SMART), TUMCREATE (established by the Technical University of Munich), the National University of Singapore (NUS), the Singapore Management University (SMU), and the Agency for Science, Technology and Research (A*STAR). The authors would like to thank the Meteorological Service Singapore (MSS) for providing the climatic/meteorological data. Special thanks to Prof. Gerhard Schmitt, Prof. Leslie Norford, Dr. Jimeno Fonseca, Shiying Li, and Luis Santos for supporting the research, providing ideas, and reviewing the methods and results.

Haut de page

Bibliographie

Acero J.A., Koh E.J., Li X., Ruefenacht L.A., Pignatta G., Norford L.K., 2019, "Thermal impact of the orientation and height of vertical greenery on pedestrians in a tropical area", Building Simulation, Vol.12, No.6, 973-984.

Acero J.A., Koh E.J., Pignatta G., Norford L.K., 2020a, "Clustering weather types for urban outdoor thermal comfort evaluation in a tropical area", Theoretical and Applied Climatology, Vol.139, No.1-2, 659-675.

Acero J.A., Koh E.J.Y., Tan Y.S., 2020b, "Analysis of climatic variables in different urban sites of Singapore and evaluation of strategies to improve the outdoor thermal environment", Technical Report.

Acero J.A., Koh E.J.Y., Ruefenacht L.A., Norford L.K., 2021, "Modelling the influence of high-rise urban geometry on outdoor thermal comfort in Singapore", Urban Climate, Vol.36, 100775.

Acero J.A., Ruefenacht L.A., Koh E.J.Y., Tan Y.S., Norford L.K., 2022, "Measuring and comparing thermal comfort in outdoor and semi-outdoor spaces in tropical Singapore", Urban Climate, Vol.42, 101122.

Alexandri E., Jones P., 2008, "Temperature decreases in an urban canyon due to green walls and green roofs in diverse climates", Building and Environment, Vol.43, 480-493.

Ali-Toudert F., Mayer H., 2006, "Numerical study on the effects of aspect ratio and orientation of an urban street canyon on outdoor thermal comfort in hot and dry climate", Building and Environment, Vol.41, 94-108.

ASHRAE A., 2004, Standard 55-2004: Thermal environmental conditions for human occupancy, ASHRAE: American Society of Heating, Refrigeration and Airconditioning Engineers, Atlanta, [en ligne]. URL : http://arco-hvac.ir/wp-content/uploads/2015/11/ASHRAE_Thermal_Comfort_Standard.pdf

Bruse M., Fleer H., 1998, "Simulating surface-plant-air interactions inside urban environments with a three-dimensional numerical model", Environmental modelling & software, Vol.13, No.3-4, 373-384.

Chang J.H., Winter T., 2015, Thermal modernity and architecture”, The Journal of Architecture, Vol.20, No.1, 92-121.

Chen L., Mak C.M., 2021, Integrated impacts of building height and upstream building on pedestrian comfort around ideal lift-up buildings in a weak wind environment”, Building and Environment, Vol.200, 107963.

Chen Y.C., Matzarakis A., 2014, "Modification of physiologically equivalent temperature", Journal of Heat Island Institute International, Vol.9, No.2.

de Schiller S., Evans J.M., 1990, "Bridging the gap between climate and design at the urban and building scale: research and application", Energy and Buildings, Vol.15, No.1-2, 51-55.

Eliasson I., 2000, "The use of climate knowledge in urban planning", Landscape and Urban Planning, Vol.48, No.1-2, 31-44.

Evans J.M., De Schiller S., 1996, "Application of microclimate studies in town planning: a new capital city, an existing urban district and urban river front development", Atmospheric Environment, Vol.30, No.3, 361-364.

Fanger P.O., 1970, "Thermal comfort. Analysis and applications in environmental engineering", Thermal comfort. Analysis and applications in environmental engineering.

Gamero-Salinas J., Kishnani N., Monge-Barrio A., López-Fidalgo J., Sánchez-Ostiz A., 2021, "The influence of building form variables on the environmental performance of semi-outdoor spaces. A study in mid-rise and high-rise buildings of Singapore", Energy and Buildings, Vol.230, 110544.

Givoni B., 1991, "Impact of planted areas on urban environmental quality: A review", Atmospheric Environment, Urban Atmosphere, Vol.25, No.3, 289-299.

Höppe P., 1999, "The physiological equivalent temperature - a universal index for the biometeorological assessment of the thermal environment", International Journal of Biometeorology, Vol.43, No.2, 71-75.

Höppe P., 2002, "Different aspects of assessing indoor and outdoor thermal comfort", Energy and Buildings, Vol.34, No.6, 661-665.

Huttner S., 2012, Further development and application of the 3D microclimate simulation ENVI-met, Doctoral dissertation, Universitätsbibliothek Mainz.

IPCC, 2019, Climate Change and Land: an IPCC special report on climate change, desertification, land degradation, sustainable land management, food security, and greenhouse gas fluxes in terrestrial ecosystems.

Jamei E., Rajagopalan P., Seyedmahmoudian M., Jamei Y., 2016, "Review on the impact of urban geometry and pedestrian level greening on outdoor thermal comfort", Renewable and Sustainable Energy Reviews, Vol.54, 1002-1017.

Kikegawa Y., Genchi Y., Yoshikado H., Kondo H., 2003, "Development of a numerical simulation system toward comprehensive assessments of urban warming countermeasures including their impacts upon the urban buildings' energy-demands", Applied Energy, Vol.76, No.4, 449-466.

Klemm W., 2018, Clever and cool: generating design guidelines for climate-responsive urban green infrastructure, Doctoral dissertation, Wageningen University.

Kong L., Lau K.K.L., Yuan C., Chen, Y., Xu Y., Ren C., Ng E., 2017, "Regulation of Outdoor Thermal Comfort by Trees in Hong Kon", Sustainable Cities and Society, Vol.31, 12-25.

Lau S.S., Zhang Q., 2015, "Genesis of a vertical city in Hong Kong", International Journal of High-Rise Buildings, Vol.4, No.2, 117-125.

Looman R., 2017, "Climate-responsive design: A framework for an energy concept design-decision support tool for architects using principles of climate-responsive design", A+ BE| Architecture and the Built Environment, Vol.1, 1-282.

Melnikov V.R., Christopoulos G.I., Krzhizhanovskaya V.V., Lees M.H., Sloot P., 2022, "Behavioural thermal regulation explains pedestrian path choices in hot urban environments", Scientific Reports, Vol.12, No.1, 1-11.

Muniz-Gäal L.P., Pezzuto C.C., Carvalho M.F.H. de, Mota L.T.M., 2020, "Urban geometry and the microclimate of street canyons in tropical climate", Building and Environment, Vol.169, 106547.

Nazarian N., Acero J.A., Norford L., 2019, "Outdoor thermal comfort autonomy: Performance metrics for climate-conscious urban design", Building and Environment, Vol.155, 145-160.

Nevat I., Ruefenacht L., Aydt H., 2020a, "Recommendation system for climate informed urban design under model uncertainty", Urban Climate, Vol.31, 100524.

Nevat I., Pignatta G., Ruefenacht L.A., Acero J.A., 2020b, "A decision support tool for climate-informed and socioeconomic urban design", Environment, Development and Sustainability, 1-25.

Nevat I., 2021, "Climate-informed urban design via probabilistic acceptability criterion and Sharpe ratio selection", Environment, Development and Sustainability, 1-29.

Ng E., 2012, "Towards planning and practical understanding of the need for meteorological and climatic information in the design of high‐density cities: A case‐based study of Hong Kong", International Journal of Climatology, Vol.32, No.4, 582-598.

Oke T.R., 1984, "Towards a prescription for the greater use of climatic principles in settlement planning", Energy and Buildings, Vol.7, No.1, 1-10.

Oke T.R., 2006, "Towards better scientific communication in urban climate", Theoretical and Applied Climatology, Vol.84, No.1, 179-190.

Oke T.R., Mills G., Christen A., Voogt J.A., 2017, Urban Climates, Cambridge: Cambridge University Press.

Patz J.A., Campbell-Lendrum D., Holloway T., Foley J.A., 2005, "Impact of regional climate change on human health", Nature, Vol.438, No.7066, 310-317.

Priyadarsini R., Womg N.H., Cheong K.W.D., 2008, "Microclimatic modeling of the urban thermal environment of Singapore to mitigate urban heat island", Solar energy, Vol.82, No.8, 727-745.

Roth M., Chow W.T., 2012, "A historical review and assessment of urban heat island research in Singapore", Singapore Journal of Tropical Geography, Vol.33, No.3, 381-397.

Ruefenacht L., Acero J.A., 2017, Strategies for cooling Singapore: A catalogue of 80+ measures to mitigate urban heat Island and improve outdoor thermal comfort, Technical Report.

Ruefenacht L., Sukma A., Acero J.A., Nevat I., 2020, Climate-responsive design guidelines: Urban design guidelines to improve outdoor thermal comfort in the southern shore area of Singapore, Technical Report.

Sharmin T., Steemers K., Matzarakis A., 2015, "Analysis of microclimatic diversity and outdoor thermal comfort perceptions in the tropical megacity Dhaka, Bangladesh", Building and Environment, Vol.94, 734-750.

Singh V.K., Mughal M.O., Martilli A., Acero J.A., Ivanchev J., Norford L.K., 2022, "Numerical analysis of the impact of anthropogenic emissions on the urban environment of Singapore", Science of The Total Environment, Vol.806, 150534.

Tao Y., Lau S.S.Y., Gou Z., Zhang J., Tablada A., 2019, "An investigation of semi-outdoor learning spaces in the tropics: Spatial settings, thermal environments and user perceptions", Indoor and Built Environment, Vol.28, No.10, 1368-1382.

Uejio C.K., Wilhelmi O.V., Golden J.S., Mills D.M., Gulino S.P., Samenow J.P., 2011, "Intra-urban societal vulnerability to extreme heat: the role of heat exposure and the built environment, socioeconomics, and neighborhood stability", Health & Place, Vol.17, No.2, 498-507.

Wai K.M., Yuan C., Lai A., Peter K.N., 2020, "Relationship between pedestrian-level outdoor thermal comfort and building morphology in a high-density city", Science of the total environment, Vol.708, 134516.

Yang W., Wong N.H., Li C.Q., 2016, "Effect of street design on outdoor thermal comfort in an urban street in Singapore", Journal of Urban Planning and Development, Vol.142, No.1.

Yang W., Wong N.H., Lin Y., 2015, "Thermal comfort in high-rise urban environments in Singapore", Procedia Engineering, Vol.121, 2125-2131.

Yuan C., 2018, Urban wind environment: Integrated climate-sensitive planning and design, Springer.

Haut de page

Annexe

Annex 1

Table 3. Table with individual metric values, and the PET performance score of each strategy and scenario.

Strategy set

Baseline / Scenarios

Mean performance metric

Semi-deviation metric

Coverage of value metric

Performance score

Performance ranking

Block and street orientation

north-south orientation

34.9

1.7

9.7

20.3

Good

northeast-southwest orientation

35.7

2.2

14.8

22.1

Bad

northwest-southeast orientation

35.2

2.7

16.2

22.3

Bad

east-west orientation

35.4

1.6

9.2

20.4

Good

Aspect ratio

low aspect ratio (1.5) along N-S axis

35.4

1.8

9.4

20.5

Good

high aspect ratio (4.5) along N-S axis

34.8

1.8

10.2

20.4

Good

low aspect ratio (1.5) along E-W axis

37.3

1.9

11.9

22.1

Bad

high aspect ratio (4.5) along E-W axis

35.8

1.9

12.5

21.5

Ok

Street canyon width

wide street canyon (45m) along N-S axis

35.4

1.8

9.4

20.5

Good

narrow street canyon (15m) along N-S axis

34.8

1.8

10.2

20.4

Good

wide street canyon (45m) along E-W axis

37.3

1.9

11.9

22.1

Bad

narrow street canyon (15m) along E-W axis

35.8

1.9

12.5

21.5

Ok

Building height variation

symmetric high-low-high

35.4

1.8

9.8

20.6

Good

symmetric low-high-low

34.9

1.9

10.9

20.6

Good

asymmetric high-low-high

35.0

1.8

10.7

20.4

Good

asymmetric low-high-low

35.2

1.9

10.0

20.6

Good

Building height

high-rise building

34.3

1.7

8.9

19.7

Good

mid-rise building

35.9

1.8

9.4

20.8

Good

low-rise building

36.6

1.6

10.4

21.3

Ok

no building

40.3

1.5

12.4

23.7

Bad

Podium-tower typology

podium with single tower

36.2

1.7

10.3

21.1

Ok

stepped podium with single tower

37.9

1.6

10.8

22.0

Bad

5m elevated podium with single tower

33.3

1.3

7.7

18.9

Good

no podium

40.3

1.5

12.4

23.7

Bad

Height of elevated podium

5m elevated podium with single tower

33.3

1.3

7.7

18.9

Good

10m elevated podium with single tower

33.2

1.4

8.6

18.9

Good

15m elevated podium with single tower

33.1

1.4

8.6

18.9

Good

20m elevated podium with single tower

32.6

1.4

8.0

18.6

Good

Number of towers

one tower with 0.15 ratio

33.1

1.4

8.6

18.9

Good

two towers with 0.25 ratio

32.5

1.3

8.9

18.8

Good

four towers with 0.30 ratio

29.1

1.3

4.8

16.1

Good

Green coverage

0% green coverage

37.4

1.8

10.4

22.0

Bad

40% green coverage

35.7

1.5

8.9

20.5

Good

60% green coverage

36.1

1.6

9.6

20.8

Good

85% green coverage

37.2

1.6

12.4

22.1

Bad

Tree planting location

trees along north frontage

35.8

1.4

8.4

20.3

Good

trees along south frontage

35.0

1.3

8.2

20.0

Good

trees along east frontage

34.7

1.2

7.0

19.5

Good

trees along west frontage

34.9

1.5

8.6

20.0

Good

Haut de page

Table des illustrations

Titre Figure 1. Location of the selected study area: new district south-eastern shore of the Central Business District.
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-1.jpg
Fichier image/jpeg, 660k
Titre Figure 2. Example of a simulated domain covered by buildings, and the study area (800 x 800 meters) defined by the inner part of the dotted square (corresponding to the 75% of the area covered by buildings), (modified from Acero, Koh, Ruefenacht & Norford, 2021).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-2.png
Fichier image/png, 108k
Titre Figure 3. A flowchart of the framework
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-3.jpg
Fichier image/jpeg, 176k
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-4.png
Fichier image/png, 6,8k
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-5.png
Fichier image/png, 13k
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-6.png
Fichier image/png, 5,8k
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-7.png
Fichier image/png, 6,6k
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-8.png
Fichier image/png, 3,3k
Titre Figure 4. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., north-south orientation) and scenarios (i.e., northeast-southwest, northwest-southeast, and east-west orientations).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-9.png
Fichier image/png, 278k
Titre Figure 5. Scenarios with different orientations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-10.jpg
Fichier image/jpeg, 268k
Titre Figure 6. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., low aspect ratio along the long street canyon) and scenarios (i.e., high aspect ratio along the long street canyon, low aspect ratio along the short street canyon, and high aspect ratio along the short street canyon).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-11.png
Fichier image/png, 169k
Titre Figure 7. Scenarios with different aspect ratios (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-12.jpg
Fichier image/jpeg, 224k
Titre Figure 8. Scenarios with different street canyon widths (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-13.jpg
Fichier image/jpeg, 220k
Titre Figure 9. Scenarios with different height variations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-14.jpg
Fichier image/jpeg, 252k
Titre Figure 10. Scenarios with different building heights (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-15.jpg
Fichier image/jpeg, 208k
Titre Figure 11. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., symmetric high low-high) and scenarios (i.e., symmetric low-high-low, asymmetric high-low-high, and asymmetric low-high-low).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-16.png
Fichier image/png, 303k
Titre Figure 12. Scenarios with different podium-tower typologies (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-17.jpg
Fichier image/jpeg, 244k
Titre Figure 13. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with single tower) and scenarios (i.e., stepped podium with single tower, and single tower without podium).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-18.png
Fichier image/png, 271k
Titre Figure 14. Scenarios with different podium elevations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-19.jpg
Fichier image/jpeg, 228k
Titre Figure 15. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 5 meter elevated podium) and scenarios (i.e., 10 meter elevated podium, 15 meter elevated podium, and 20 meter elevated podium).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-20.png
Fichier image/png, 281k
Titre Figure 16. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., podium with one tower) and scenarios (i.e., podium with two towers, and podium with four towers).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-21.png
Fichier image/png, 205k
Titre Figure 17. Scenarios with different tower configurations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-22.jpg
Fichier image/jpeg, 164k
Titre Figure 18. Histogram of PET coverage per exposure area for April (WT 3) in the afternoon (11-16h) for baseline (i.e., 40% green coverage) and scenarios (i.e., 0% green coverage, 60% green coverage, and 85% green coverage).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-23.png
Fichier image/png, 253k
Titre Figure 19. Scenarios with different ratios of green coverage and building footprint (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-24.jpg
Fichier image/jpeg, 232k
Titre Figure 20. Scenarios with different ratios of tree locations (top). Bar charts of the excess score of each scenario with respect to the baseline (bottom).
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-25.jpg
Fichier image/jpeg, 232k
Titre Figure 21. The PET performance score of each strategy and scenario.
URL http://journals.openedition.org/cybergeo/docannexe/image/39608/img-26.jpg
Fichier image/jpeg, 344k
Haut de page

Pour citer cet article

Référence électronique

Lea A. Ruefenacht, Ayu S. Adelia, Juan A. Acero et Ido Nevat, « Urban Design Guidelines for Improvement of Outdoor Thermal Comfort in Tropical Cities », Cybergeo: European Journal of Geography [En ligne], Cartographie, Imagerie, SIG, document 1032, mis en ligne le 30 septembre 2022, consulté le 27 novembre 2022. URL : http://journals.openedition.org/cybergeo/39608 ; DOI : https://doi.org/10.4000/cybergeo.39608

Haut de page

Auteurs

Lea A. Ruefenacht

Singapore-ETH Centre, 138602, Singapore
Corresponding author
E-mail address: lea.ruefenacht@esc.ethz.ch

Ayu S. Adelia

Singapore-ETH Centre, 138602, Singapore

Juan A. Acero

Singapore-MIT Alliance for Research and Technology, 138602, Singapore

Ido Nevat

TUMCREATE, 138602, Singapore

Haut de page

Droits d’auteur

Licence Creative Commons
La revue Cybergeo est mise à disposition selon les termes de la Licence Creative Commons Attribution 4.0 International.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search