Urban Form and Energy Demand

Abstract

The implications of urban form on energy have long been present in international debate, whether considering travel patterns or thermal comfort in buildings. The urban environment is a result of a set of intertwined attributes, the understanding of which is often unclear. The energy trade-offs between urban form attributes haven’t received proper attention. Research remains sectorial, considering buildings and transport in isolation. In order to allow for a comprehensive analysis of this relationship, this article reviews urban attributes with energy relevance. A collection of attributes and metrics is gathered from the literature for incorporating urban form in urban energy analysis.

Keywords

urban form buildings transports urban planning energy sustainability metric

In recent decades, particular attention has been drawn to cities and urban areas under the framework of the sustainability and the climate change agenda. The recently published Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Seto et al. 2014) identifies four clusters of drivers for greenhouse gas (GHG) emissions—mostly energy-driven: economic geography and income, sociodemographic factors, technology, and infrastructure and urban form. Urban form is ranked in the report as a driver of high importance in mature cities. Such importance is due to its influence on mobility patterns and other energy uses, notably heating and cooling in buildings.

The relationship between urban form and energy has long been present in the international debate. It is not the only driver of energy demand and arguably not the most important. While there is scientific evidence that urban form, or some of its attributes, has an impact on the demand for this resource, the characterization of the effect of urban form per se remains a major challenge. This can be attributed to several reasons (M. Silva, Oliveira, and Leal 2016): (i) it is difficult to isolate urban form from other drivers of energy demand, (ii) there are many variables of urban form to be considered, (iii) the degree of interaction among each of the variables is not fully defined, and (iv) two key sectors often addressed in isolation (buildings and transport) should be simultaneously considered.

Despite the fact that there is a wide array of studies emphasizing the relationship between urban form and energy, not all factors influencing this relationship are simultaneously addressed nor are they fully quantified. It has been argued that the effect of each variable of urban form has a relatively small effect on the overall urban energy demand. Their combined effect, however, is expected to be more significant and worth controlling for in statistical analysis (Ewing and Cervero 2010). There is a broad set of urban form attributes with relevance for energy conservation in urban planning. However, the existing research has been sectorial and focusing only in one or few variables at a time (Naess 2003; Rickwood, Glazebrook, and Searle 2008). The energy trade-offs between the different urban form attributes haven’t been duly investigated.

In order to allow for a comprehensive analysis of the effect of urban form on energy demand, this article reviews the literature considering effects both in buildings and in transportation. It will focus on the attributes of the physical form of cities and their energy implications throughout different scales. The metrics used for characterizing these urban attributes are reviewed as well. Whereas the effect of specific parameters of urban form has been widely studied, the suitability of the metrics employed has been subject to limited discussion. The relevance of the physical attributes to urban planning lies on the ability to thoroughly characterize and quantify them, so that increasingly more objective and tailored policies can be adopted toward energy conservation in cities. Although the connections between land use and energy supply have recently been brought to the fore (e.g., Kaza and Curtis 2014), the scope of this review is placed on the demand-side.

The first section presents an overview of how the body of research on urban form and energy has developed over the years. The second section reviews the attributes of urban form with energy relevance, discussing how a given attribute affects energy demand, and presents metrics that may be applied in an energy analysis. The article concludes with the lessons learnt, providing some suggestions for future work.

Urban Form and Energy Demand: An Overview

The fields of urban form and energy merged over two decades ago. Although some earlier work has been identified, a seminal study in the United States in the late 80s explores how the built environment affects energy demand (P. W. G. Newman and Kenworthy 1989), probably driven by the acknowledgment that the American dispersed development model (urban sprawl) leads to higher fuel costs and congestion. The authors analyze the consumption of gasoline in ten US cities, concluding that a significant share of the variation could be attributed to land use and transport planning choices (instead of price or income). Density of urban activities is the planning factor considered. Despite some criticism invoking the validity of global comparisons or the interaction between transportation, labor, housing, and land (Gordon and Richardson 1989), this study is undeniably a landmark on the analysis of the energy demand derived from the physical structure of the city. Since then, it has been generally accepted that higher urban densities lead to energy savings for travel purposes (Banister, Watson, and Wood 1997; P. Newman and Kenworthy 2006; Glaeser and Kahn 2010). In the same year, Cervero (1989) points out the jobs–housing balance as a factor with an impact on mobility patterns; and a few years later, Frank and Pivo (1994) explore several features of the urban environment simultaneously, reinforcing the correlation between density and mix of land uses and mobility patterns (and indirectly with energy, as mobility patterns affect transport-related energy demand). Mixed land uses have increasingly been advocated to reduce motorized travel (Ewing, Haliyur, and Page 1994; Cervero 1996; Dieleman, Dijst, and Burghouwt 2002; C. Lee and Moudon 2006a). Nevertheless, the analysis of the urban environment could still benefit from a more comprehensive description. The effects of other urban form attributes on travel, such as accessibility, connectivity, and design, have also been explored, although not as extensively as density and diversity (S. Handy 1993; Ryan and McNally 1995; Naess 2003). Few studies consider a broad set of attributes simultaneously to describe the urban environment, although there are exceptions (Ewing and Cervero 2001, 2010).

At present, urban form and travel is a vast field of research. While there is still disagreement on the role of urban form, especially in relation to behavioral or economic factors (Crane 2000), the importance of scale when addressing travel patterns is widely acknowledged (Crane 2000; Stead and Marshall 2001; Ewing and Cervero 2001; J. E. Anderson, Wulfhorst, and Lang 2015), since the attributes at stake and their effects depend on the level of detail considered. Stead and Marshall (2001) argue that the literature has been focusing on macro features, such as settlement size, mix of land uses, and density, while lower scale features, such as the proximity to transport networks, network features, and neighborhood type, may also be relevant but have been overlooked. The effect of microurban attributes on travel has not been widely investigated (Soltani and Allan 2006).

This debate has moved from the transportation field to effects within the building sector. Buildings (households and services) are the largest energy consumers and GHG emitters in the European Union (European Environment Agency 2015). Some urban form attributes influencing travel patterns (e.g., density and the existence of green spaces) have also been linked to the urban heat island (UHI) effect (Oke 1982, 1988) as well as to building thermal comfort, and thus energy conservation. The effect of urban form on the UHI is, in fact, a vast field of research per se. Nevertheless, the implications on energy are not always explored.

While the patterns of energy use in nonresidential buildings are more complex, urban form is expected to be particularly relevant in dwellings. Ewing and Rong (2008) argue that after controlling for economic factors, the physical aspects of housing significantly influence energy use. Anisimova (2011) claims that the type of construction largely determines energy demand, with compact structures and the passive use of solar energy as the basis of low-energy design. Ko (2013) reviews the body of research on the effects of urban form on residential energy demand, considering house size and type, density, community layout, planting, and surface coverage. The estimation of the overall magnitude of the effect of urban form on building’s energy demand is still underexplored (S. Lee and Lee 2014) when compared to that on the transportation field. Nevertheless, for both sectors, the combined influence of urban form is still unclear. Comprehensive studies are still a few.

The implications of urban form in buildings are mostly concerned with building geometry and surroundings, and the UHI, although electric transmission and distribution losses have also been pointed as a causal pathway for this link (Ewing and Rong 2008). “Macro” features of the urban environment, such as density or compactness (usually evidenced through clustered building patterns and increased heights), are often associated with lower heating needs in winter (Høyer and Holden 2003; Rode et al. 2014). However, Kaza (2010) points that the focus of densification should be on transportation and argues that only dramatic changes in the housing type mix could lead to sizable effects in energy conservation. Lower scale characteristics of urban texture, notably building shading and passive gains, have also been investigated (Baker and Steemers 2000; Ratti, Baker, and Steemers 2005). Baker and Steemers (2000) estimate the weight of different factors in the energy performance of buildings. Unlike building design (leading to a variation of 2.5×), systems’ design (2×), and occupant’s behavior (2×), the contribution of the urban context is still unquantified.

Two additional studies attempt at estimating the effect of urban form on heating. From an analysis of three urban tissues in Paris, Salat (2009) concludes that urban form has an effect of a factor 1.8. The main features considered are density and a shape factor (defined by an index relating the external surface of the building and its volume). Rode et al. (2014) consider different indicators of building density and compactness in four cities and find that the parameters analyzed can lead to a variation of sixfold. Still, both approaches disregard relevant features of the urban environment, notably those related to transportation.

Despite significant progress in characterizing the relationship between the physical environment and energy demand, it is widely agreed that further analysis and more robust conclusions are needed. Table 1 summarizes the urban attributes addressed in existing review papers and their unique contributions. It evidences that sectorial research still dominates. This article attempts to summarize what is known on the attributes of urban form with energy relevance, belonging to the built environment and to the urban networks (Figure 1). It explores how they are expected to influence energy demand, accounting for potential trade-offs, and suggests metrics to promote the development of a coherent analysis framework in the planning practice. The article will address two key questions: (i) how do urban attributes influence energy demand and (ii) how to describe and measure each attribute of urban form with energy relevance. It argues that a straightforward assessment of the impact of urban form on energy could contribute to defining more informed planning strategies as well as to prioritizing the physical aspects that can be intervened for more efficient and sustainable urban planning.

Table 1.

Summary of Key Reviews Published on Urban Form and Energy Demand.

Author (Year)	Sector	Urban Attributes Addressed	Contribution/Key Findings
Anderson, Kanaroglou, and Miller (1996)	Essentially travel	Mostly macro attributes (density, diversity, distance to CBD, and accessibility)	- Urban form influences energy demand
			- Further research is needed to improve knowledge of existing relationships
			- Suggestions for future models
Anderson, Wulfhorst, and Lang (2015)	B/T (as a category within urban scale)	Orientation, shape, compactness, shading, passive condition (within a category named architectural design)/density (single category), and mix of uses (urban form category)	- Urban form influences energy demand
			- Review focuses on two scales (the building and the urban scale, each with several subcategories)
			- Considering only one is insufficient to capture existing interactions
Crane (2000)	T	Density (different measures), land-use mixing, traffic calming, street and circulation pattern, and pedestrian features	- Urban form and travel represents a complicated interaction of several factors, including economic and attitudinal ones, which should be included in the analysis
Crane (2000)	T		- Acknowledges the importance of scale
Ewing and Cervero (2001)	T	Density; land-use mix; accessibility (within land-use patterns), connectivity, directedness of routing, block sizes, sidewalk continuity (within urban networks), building orientation, pedestrian amenities, and parking (as urban design features)	- Urban form influences energy demand
			- Derived composite land use–travel elasticities for VMT and trip frequency
			- Trip frequencies are more influenced by socioeconomic factors, while VMT are more dependent on the built environment.
			- Land use–travel elasticities are advocated to be descriptors of the link between travel and the built environment, able to be transferred between different regions.
Ewing and Cervero (2010)	T	“Six Ds”: density, diversity, and design + destination accessibility and distance to transit + demand management (parking)	- Urban form influences energy demand
			- Performed a meta-analysis arriving at weighted averages of elasticities
			- Walking and transit use are new trip outcomes considered
			- Population and job densities are weakly associated with travel behavior once other variables are controlled for
Ewing and Hamidi (2015)	T	Sprawl (density, diversity, connectivity, centrality, clustering, shape, and size)	- Urban form influences energy demand
			- Sprawl is a multidimensional issue
			- Discusses costs and benefits of sprawl (including energy)
			- Conclude that disaggregating the different elements of urban form to study the effects of sprawl may lead to more precise conclusions
Ko (2013)	B	House size, housing type, density, community layout, planting, and surface coverage	- Urban form influences energy demand
			- Bridges design based and planning based research
			- Research on urban form and residential energy is scarcer than that on travel
			- Existing research is insufficient to understand the full impact of urban form on energy efficiency
			- It is unlikely the existence of an ideal urban form due to trade-offs across urban forms and climates
Rickwood, Glazebrook, and Searle (2008)	B/T	Dwelling type (construction, design, size, and orientation) and density	- Urban form influences energy demand
			- Consider the effect of urban structure on embodied and operational energy
			- Literature lacks studies combining travel and in-house energy use, resulting on inaccurate estimates of the overall effect
Stead and Marshall (2001)	T	Distance to CBD, settlement size, mix of land uses, provision of local facilities, density, proximity to transport networks, parking, road network type, and neighborhood type	- Urban form influences energy demand
			- Reviews research across local to regional scales
			- Interactions between socioeconomic factors, urban form, and travel patterns complicate the analysis of existing relationships.
			- Different ways of expressing the same variable make interpretations more difficult
Steiner (1994)	T	Density (including mixed uses and proximity to public transport)	- Addresses the need to include socioeconomic factors along with urban form studies
Steiner (1994)	T		- Review categorizes research into empirical and policy formulation studies

Note. B = Buildings; CBD = central business district; VMT = vehicle miles traveled; T = Travel.

Figure 1.

Urban form attributes, their outcome factors, and expected effects on overall urban energy demand.

The second section reviews the attributes of urban form with energy relevance, discussing how these attributes may determine energy needs in buildings and in transport, and collects a set of metrics for describing the urban environment to be used in future analytical studies.

A Review of Urban Form Attributes Influencing Energy Demand

Physical attributes of the urban environment affecting energy demand have been pinpointed and discussed in the literature in the last decades. The degree up to which these attributes are explored in the existing research varies greatly and comprehensive studies are still few. This raises the need to broaden the scope of the physical dimension in urban energy analysis throughout different scales. Although microscale attributes may play an important role in reducing energy demand at the local scale, they have been underexplored (Soltani and Allan 2006). Additionally, the two most important form-related sectors consuming energy in cities (buildings and transportation) are repeatedly treated separately, although it is acknowledged that they are interconnected, entailing energy trade-offs (S. Lee and Lee 2014), and that they should be analyzed together whenever possible (O’Brien et al. 2010).

The following review identifies a set of energy-relevant urban attributes and presents measures to quantify them. The purpose is not to get a single metric for each urban attribute. It is here advocated that the urban environment should be described in a comprehensive way to better capture existing variations. This article presents different perspectives and discusses their suitability for an effective urban energy analysis.

The attributes of urban form are included within two major focus areas: the built environment and urban networks. The built environment refers to the urban settings that are effectively built-up (i.e., the building stock), whether considering urban structure or urban form, as well as those areas which are not built-up, but that are intimately related and shaped by the former, like green spaces. Urban networks refer to the characteristics of urban transport infrastructure, independently of which transport mode is considered.

The attributes of urban form reviewed range from the city scale to the scale of individual urban elements, such as buildings. Urban features at a higher level of detail than the single building (like construction materials or fenestration aspects) are disregarded, as they are not considered to belong to the urban form debate. For each urban attribute, a set of indicators and metrics is gathered (Table 2). It is acknowledged that the classification and the metrics proposed are not definitive. Some attributes may have an ambiguous definition. The studies examined explore urban attributes considering their impact on energy demand. Research unrelated with energy or focused on energy supply is not considered.

Table 2.

Factors and Metrics of Urban Form Influencing Energy Demand in Cities.

Focus Area	Factor	Indicator	Relevance Scale	Metric	Sources
Built environment	Density	Intensity of the built environment	Bl, N, and C	Population/households/jobs per land area	P. W. G. Newman and Kenworthy (1989), Frank and Pivo (1994), and Porter et al. (2013)
		Floor space index	S, Bl, N, and C	Gross floor area/total ground area	Pont and Haupt (2005)
		Ground space index	S, Bl, N, and C	Building coverage area/total ground area	Pont and Haupt, (2005)
		Building height	All	Number of floors	Pont and Haupt, (2005)
		Subdivision indicator	S, Bl, N, and C	Number of parcels/total ground area (square meter)	Bourdic, Salat, and Nowacki (2012)
		Housing size	S, Bl, N, and C	Number of bedrooms	Wilson (2013)
	Diversity	Dissimilarity index	N and C	${[\sum_{j}^{k} \sum_{l}^{8} (X l / 8)] / K},$ where K = number of developed hectare grid cells in tract and Xl = 1 if land-use category of neighboring hectare grid cell differs from hectare grid cell j (0 otherwise)	Cervero and Kockelman (1997)
		Entropy	N and C	${\sum_{k} [\sum_{j} P_{j k} \ln (p_{j k})] / \ln (J)} / K$ , where p_jk = proportion of land-use category j within a radius of the developed area surrounding grid cell k, j = number of land-use categories, and K = number of actively developed hectares in tract	Cervero and Kockelman (1997)
		Structural diversity	N and C	$I = \frac{1}{cat} {\sum_{i = 1}^{cat} [1 - \frac{N {entities}_{cat i}}{N {entities}_{cat i}^{obj}}]}^{2}$	Bourdic, Salat, and Nowacki (2012)
		Mixed-use index	Bl, N, and C	Residential gross floor area/gross floor area	van den Hoek (2008)
	Green areas	Spatial distribution index	N and C	$I s d = \frac{Q}{Q - 1} [1 - \sum_{district = 1}^{Q} {(\frac{A_{i}}{S_{i}})}^{2}]$ , where Q districts exist in the city, each with an overall area of S_i (square meter) and A_i of green areas (square meter)	Bourdic, Salat, and Nowacki (2012)
		Ratio of green space	Bl, N, and C	Percentage of green space area in the total area of the parcel	Ko and Radke (2014)
		Green area geometry	Bl, N, and C	Perimeter/area	Vaz Monteiro et al. (2016)
	Compactness	Surface-to-volume ratio	Bu, Bl, N, and C	$C = S / V$	Ratti, Baker, and Steemers (2005)
		Size factor (SF)	Bu, Bl, N, and C	$SF = V^{1 / 3}$	Bourdic, Salat, and Nowacki (2012)
		Form factor (FF)	Bu, Bl, N, and C	$FF = S / V^{2 / 3}$	Bourdic, Salat, and Nowacki (2012)
	Shading	UHA	Bu and S	$tan (UHA) = H / W$	Baker, Hoch, and Steemers (1992)
		OSV	Bu and S	$cos (OSV) = H / W$	Ratti, Baker, and Steemers (2005)
		Sky view factor	Bl, N, and C	Proportion of unobstructed sky at a given location	Coseo and Larsen, (2014)
	Passivity	Passive/nonpassive ratio	All	Passive areas/nonpassive areas	Ratti, Baker, and Steemers (2005)
	Passivity	Plan depth	All	Volume/exposed wall area	Evans, Liddiard, and Steadman (2017)
	Orientation	Orientation	Bu and S	Degrees, radians	—
Urban networks	Connectivity	Intensity of network intersections	N and C	Number of intersections per area unit; average distance between intersections	Bourdic, Salat, and Nowacki (2012)
		Cyclomatic number	N and C	Number of primary loops in a network: µ = L − N + 1, where L is the number of links, corresponding to the different sections of streets between every intersection, and N is the number of nodes, corresponding in road networks to intersections	Bourdic, Salat, and Nowacki (2012) and Chen, Claramunt, and Ray (2014)
		Connectivity	N and C	Percentage of T-intersections	Baran, Rodríguez, and Khattak (2008)
		β Index	S, Bl, N, and C	$β = e / n (0 \leq β \leq (n - 1) / 2)$ , where β is the average number of edges (e) per node (n) in a given street network	Chen, Claramunt, and Ray (2014)
		α Index	N and C	Alpha index (α) is defined as the proportion between the actual and maximal number of circuits in a fully connecting planar network $α = 2 µ / (n - 1) (n - 2) = (e - n + q) / (2 n - 5 q) (0 \leq α \leq 1)$	Chen, Claramunt, and Ray (2014)
	Accessibility	Active accessibility	Bl, N, and C	$A_{act, i} = \sum_{j} g (W_{j}) f (c_{i j})$ , where W_j is the activity to reach in zone j and c_ij is the generalized cost of reaching zone j from zone i	Papa and Coppola (2012)
		Passive accessibility	Bl, N, and C	$A_{pas}_{, i} = \sum_{j} g (W_{j}) f (c_{i j})$ , where W_j are the potential “consumers” of the activity to be reached in zone i and c_ij is the generalized cost of reaching zone i from reach in zone j	Papa and Coppola (2012)
		DivAct	Bl, N, and C	$DivAct = \frac{\sum_{y} ({Act}_{y} * f_{y})}{\sum_{y} f_{y}}$ , where y is the activity type, Act _y represents the existence or not of the activity type y inside accessibility boundaries (Act _y {0; 1}) and f_y the potential use frequency of the activity type	C. Silva (2013)
		Reach	S, Bl, N, and C	$Reach [i] = \sum_{j \in G - {i}; d [i, j] \leq r} W [j]$ , where d[i, j] is the shortest path distance between nodes and in G, and W[j] is the weight of a destination node j	Sevtsuk and Mekonnen (2012)
	Centrality	Proximity	Bl, N, and C	Linear distance (meter) from a tract to CBD	Ewing and Cervero (2010)
	Centrality	Jobs–housing dispersal index	N and C	Ratio of the average distance of jobs from the CBD (jobs dispersal) to the average distance of workers’ households from the CBD (housing dispersal)	Brotchie et al., (1996)
	Proximity to PT	Distance to PT	Bl, N, and C	Average shortest path from residences or work to nearest PT station or stop	Ewing and Cervero (2010)
	Proximity to PT	PT route density	Bl, N, and C	Length of PT routes per land area	Ewing and Cervero (2010)
	Design	Pedestrian provisions	S, Bl, N, and C	Proportion of blocks with sidewalks, sidewalk width	Cervero and Kockelman (1997)
		Provision of trees	S, Bl, N, and C	Proportion of blocks with planting strips and trees	Cervero and Kockelman (1997)
		Shading effect of vegetation	Bu, S, Bl, and N	Sum of tree heights with different orientations	Ko and Radke (2014)
		Cycling provisions	S, Bl, N, and C	Proportion of blocks with cycling lanes, bicycle lanes per developed acre	Cervero and Kockelman (1997)
		Parking lots provisions	S, Bl, N, and C	Proportion of nonresidential parcels with off-street parking, on-street and side parking	Cervero and Kockelman (1997)

Note. CBD = central business district; PT = public transport; UHA = Urban Horizon Angle; OSV = Obstruction sky view. Key: Bu – Building, S – Street, Bl – Block; N – Neighborhood, C – City

Legend: Bu – Building, S – Street, Bl – Block; N – Neighborhood, C – City; DivAct = diversity of activity.

Built Environment

Density

Density is likely the most prominent and frequently cited urban attribute acknowledged to influence energy demand in cities, feeding a large debate on its benefits and dangers (Jenks, Burton, and Williams 1996; Williams, Jenks, and Burton 2000). Density can be defined in many ways. Two types of definitions can be considered: density of the built environment and density of people living or working in a given area (Porter et al. 2013). In all, it is a variable of interest per unit of area. However, density described as population per area unit may refer to different urban environments (Pont and Haupt 2005; Salat 2009). The distribution of a fixed amount of residents or households within an urban area may vary greatly, depending on the configuration of the built structures, specifically building footprints and number of floors (Casas Castro Marins and Andrade Roméro 2013). As a result, population or housing density alone will hardly capture these different configurations and the corresponding energy behavior.

Density of urban activities was pointed back in 1989, in the seminal study by Newman and Kenworthy, as one of the most important factors influencing gasoline consumption in cities. In this case, density was measured as persons or jobs per unit of area. These authors also used an “urban intensity” indicator as jobs and people per unit of area (P. Newman and Kenworthy 2006). Cervero and Kockelman (1997) considered density as one of the “3Ds” affecting travel demand (along with diversity and design, and more recently extended to six Ds, including destination accessibility, distance to transit, and (parking) demand management (Ewing and Cervero 2010)). It was expressed under three metrics: (i) population density, (ii) employment density, and (iii) accessibility to jobs.

Density (or compactness¹) is expected to influence travel patterns because it potentially brings urban activities closer and shortens distances to be traveled, thus reducing motorized travel needs (Kanaroglou and South 2001; Cervero and Murakami 2010). Denser urban areas also promote more reliable public transport (PT) means (Dieleman, Dijst, and Burghouwt 2002). Frank and Pivo (1994) analyze the effect of population and jobs density (and mix of land uses) on mode choice both for work and shopping purposes. The strongest correlation between mode choice and density was verified for average gross population density at trip origins and destinations for shopping trips. However, the net benefits of density are far from being consensual. It has been questioned whether density is the driver of lower energy intensities or if it works as a proxy for other variables often found in dense urban areas, such as proximity to PT or accessibility to activities (S. Handy 1996b; Ewing et al. 2008; Ewing and Cervero 2010). Critics of density point it as a cause of traffic congestion, thus increasing energy needs, air pollution and noise (Gordon and Richardson 1997; Nijkamp and Rienstra 1996), as well as crowding and lower housing availability (Echenique et al. 2012). In addition, while density may decrease everyday travel needs, it has been linked to higher levels of out-of-city leisure travel by plane (Holden and Norland 2005) as a compensation for lower access to local green spaces.

In addition to effects on travel, density has been treated as a driver of energy demand for the built environment. Density (or compactness) is accredited to influence the UHI effect, as extensive impervious urban surfaces may increase local temperatures and thus cooling loads (Taha 1997). Compact and dense urban forms (Giridharan, Ganesan, and Lau 2004; Chun and Guldmann 2014) and development patterns like the infill development (Tran et al. 2017) have been associated with higher urban temperatures. Conversely, Stone and Rodgers (2001) affirm that sprawled patterns emit more radiant heat energy per parcel than high-density urban patterns. Density has also been associated with lower thermal losses to the outside environment in the cold season, due to more clustered urban fabrics with fewer exposed surfaces, promoted by multifamily housing (Ewing and Rong 2008). However, while minimizing heat losses, denser urban fabrics have limited solar gains (Steemers 2003). Kanters and Wall (2014) argue that density is the most influential urban attribute on the solar potential of building blocks, whereas Hachem, Athienitis, and Fazio (2012) claim that higher densities, obtained by attaching housing units, reduces both heating and cooling loads by up to 30 percent and 50 percent, respectively, when compared to detached configurations.

Due to the wide array of perspectives on the effects of density, the variables used to describe it become of major relevance. Pont and Haupt (2005) argue that if density is expressed through one metric only it may be referring to very different urban environments. In order to better understand the urban typologies associated to a certain density, the authors resort to four different variables: floor space index (intensity), ground space index (often considered under different designations, like the building ground footprint area [Chun and Guldmann 2014]), number of floors (height), and open space ratio (pressure on the unbuilt space). The authors develop a diagram to illustrate different aspects of urban density—the Spacemate, claiming that this tool may help identifying efficient urban forms. Figure 2 illustrates different urban configurations with similar floor area, evidencing different facets of density. The degree of dispersion of buildings (Zhao 2011) also correlates to average building energy consumption in passive zones. The dispersion degree is measured as follows: $T = L / 4 \sqrt{S}$ , where L is the perimeter of the facades and S is the gross floor area.

Figure 2.

Different urban configurations with different heights and with the same floor space index (app. 0,7). Source: Pont and Haupt (2005).

Behnisch et al. (2012) use three different density metrics: the so-called building load (persons/hectare), building density (structural urban density), and street network density. Finally, Bourdic, Salat, and Nowacki (2012) propose a simple metric that accounts for the number of parcels per unit of area. While this may not be a conventional indicator for measuring density, this subdivision index provides a sense of the granularity of the urban environment, which is expected to affect the energy behavior by influencing the spatial configuration of urban elements.

It has been shown that the effects of density are complicated and are not fully understood. While higher densities are mostly associated to higher energy efficiencies in mobility, their net effect in buildings is not evident. A denser pattern may be associated to lower thermal losses through the building envelope but also to lower solar gains due to shading from surrounding buildings. Trade-offs should be considered, and solutions may depend on the geographies at stake. In order to capture the different facets of density, the use of a diversified set of metrics is advocated. With a view to better understanding the effect of urban form on energy, metrics should consider elements of the built environment in addition to population ratios.

Diversity

After density, diversity (referring to the mix of the different land uses or urban functions) may be the second most cited urban attribute affecting energy demand, especially for travel. It is worth noting that some dissonance may be found between diversity and mix of land uses. Jabareen (2006) claims that diversity is a wider concept that promotes more desirable urban features, concerning not only a variety of land uses but also different building types, household sizes, and cultural urban environments. This article refers to diversity as specifically applied to land uses.

Diversity is claimed to decrease motorized needs (vehicle miles traveled [VMT]) by bringing housing and urban activities closer, thus shortening travel distances (Cervero and Kockelman 1997; Jabareen 2006; C. Lee and Moudon 2006b; Baran, Rodríguez, and Khattak 2008). In addition, it creates more vibrant and interesting urban environments, and it is positively associated to the adoption of soft modes, notably walking (Kenworthy 2006; Baran, Rodríguez, and Khattak 2008; Ewing and Cervero 2010). Diversity may be defined as a measure of the spatial distribution of urban uses or of how heterogeneous the unit of analysis is. It is often called a measure of entropy.

Diversity is the second D affecting travel demand (Cervero and Kockelman 1997), characterized by different metrics that vary in complexity. The dissimilarity index, entropy, vertical mixture, intensities of land uses, activity center mixture, and commercial intensities may be considered as more elaborate measures, whereas proximities to commercial retail uses are simpler metrics. Frank and Pivo (1994) conclude that having retail activities within neighborhoods is closely associated to mode choice for work trips, probably because people may use the “way to/from work” to do shopping and other daily activities.

Behnisch et al. (2012) use a contagion index to express heterogeneity of land uses following O’Neill et al. (1988) and Li and Reynolds (1993). This is somehow similar to the entropy index proposed by Cervero and Kockelman (1997). Entropy is also used by O’Kelly and Niedzielski (2009) to describe the relative location of jobs and housing with respect to excess commuting. Another frequently used metric for diversity is the Gini coefficient, which measures the inequality of distribution of an attribute in defined spatial units. Here, attributes refer to urban functions (population as proxy for residential areas and jobs as proxy for commerce and services).

Straightforward metrics of diversity are the percentage of the plot area occupied by major land uses (Baran, Rodríguez, and Khattak 2008) and the job–housing ratio (Cervero 1989). Cervero points that jobs–housing ratios only indicate a potential of community balance. It may happen that jobs in a certain area are not held by local residents. In addition, the mixed-use index (MXI; van den Hoek 2008) is an easy-to-measure tool for assessing the diversity of a given urban project, weighting residential floor area against gross floor area. The author argues that a desirable mix of residential and nonresidential areas is often 50/50. Bourdic, Salat, and Nowacki (2012) refer to the Simpson index for analyzing the diversity of urban projects, although admitting that it may not be appropriated for measuring cases where an even distribution of urban elements is not the objective. They propose an alternative for such cases called structural diversity. It analyzes the distance from a given share of current land uses to a target. Moreover, it is claimed that diversity indicators focus on the proportion of different land uses but not on their spatial distribution. In order to overcome this, they propose another index for analyzing the spatial distribution of urban elements.

There is a wide array of metrics to define diversity. The choice may depend on the level of complexity but also on the scale of analysis. Some metrics are more appropriate for the city scale, such as entropy or the Gini coefficient, while others may better capture lower scale aspects, like the MXI. Accordingly, diversity indicators that consider the number of different activities in a given area make more sense at city or district scales, where representation gains importance (for instance, it is not expected that every urban block has a pharmacy, however at the district scale, the number of pharmacies matters); while at lower scales, indicators should be targeted at measuring the balance between the main land uses. Whatever the scale considered, diversity is a desirable feature of the urban environment. Regarding the proportion between residential and nonresidential uses, it is widely agreed that a certain degree of mixing has benefits mostly in reducing motorized travel (VMT).

Green areas

Green areas or green infrastructure may influence energy demand in different ways. Green infrastructure, such as urban parks and trees, is advocated to help saving energy through maintaining the amenity of the urban climate (Giridharan, Ganesan, and Lau 2004; Taha 1997). It is acknowledged to help preventing the UHI effect (Ko and Radke 2014; Tran et al. 2017) and is associated to lower cooling needs (Vaz Monteiro et al. 2016). The UHI is characterized by increased temperatures in urban areas in relation to their rural hinterlands. This is due to differences in land cover, that is, to the lack of green areas (Stone and Rodgers 2001). Relevant physical characteristics of green areas include their size, width, and geometry. Small- to medium-sized green areas provide a cooling service that extends beyond their boundaries, and regular geometries seem to deliver higher cooling intensities (Vaz Monteiro et al. 2016). Nevertheless, although the size of parks is a relevant feature for urban cooling, large green areas may also imply longer distances to other urban amenities. It is important to consider the trade-off between these two effects. Other aspects like the type of vegetation cover (captured, for instance, by J.-P. Kim and Guldmann (2014) through the normalized difference vegetation index [NDVI]) also affect the UHI. These, however, are not explored here, as they are not considered urban form attributes—the same NDVI may refer to different land uses, as it reflects spectral imaging bands.

In addition to parks, Cervero and Kockelman (1997) consider that the presence of trees in streets provides a sense of a sheltered corridor for walking—trees provide shade in summer and some shelter from rain and wind (Forsyth et al. 2008). Trees are considered an element of urban design, promoting walking-friendly environments and more pleasant urban places (Forsyth et al. 2008; S. Kim, Park, and Lee 2014). At lower densities, they are also claimed to define space both at a horizontal and a vertical level, providing a sense of enclosure that it pleasant and encourages walking (Ewing and Handy 2009). With regard to buildings, if properly located, deciduous trees contribute to reduce excessive solar gains by providing shading in summer and enabling solar gains in winter. In the northern hemisphere, trees should be located South in relation to the building (Ko 2013). They may also block unwanted winter winds (United Nations-Habitat 1990). Nevertheless, the preferred location is geography dependent.

The existence of green areas is not a common feature of urban form to be analyzed, especially in a quantitative way. It may be measured in absolute terms—for example, number of parks or total green area, or in terms of green density—green area per total urban area or green area per inhabitant (e.g., Urban Audit). Distance to parks and other open spaces has been used as an accessibility indicator when analyzing travel implications (Kitamura, Mokhtarian, and Laidet 1997; Lund 2003), and park geometry has been captured by the perimeter/area ratio (Vaz Monteiro et al. 2016). Bourdic, Salat, and Nowacki (2012) suggest the use of the spatial distribution index, referred to above, to understand whether green areas are more or less evenly distributed. Their spatial distribution is claimed to affect the equity of the urban environment and also the influence of these areas on the whole city. It is expected that the existence of small- to medium-sized parks located within relatively small distances from each other brings larger benefits than a single large park (Vaz Monteiro et al. 2016). This applies to the UHI effect and is expected to work similarly regarding the effects on travel.

Cervero and Kockelman (1997) use the proportion of blocks with trees, as an element of the third D—design, to analyze the effect of tree presence on travel. Forsyth et al. (2008) measure trees (within a distance buffer) per length of road. Under a different scope, Bremer et al. (2016) measure the vegetation volume in order to build a detailed 3D model of solar irradiation. Both green areas and the effect of trees deserve further attention in the urban form–energy research. In this case, simple indexes may be useful, such as the proportion of land covered by green areas, or the proportion of street length with trees.

Compactness

Compactness of the urban tissue refers to how clustered built structures are (Ewing and Rong 2008). This was addressed in part in the density section. However, this concept also refers to building geometry, with significant energy relevance. The energy demand of a building depends on its exposed surfaces, which directly affect heat flows between the inside and the outside environment, as well as on its access to natural light. This indicator is strongly dependent on building types (Ko 2013) and is influenced by building allometry. According to Batty et al. (2008), as buildings grow in size, their shape has to change to enable them to function properly. That is when scaling is linked to allometry. Buildings cannot sustain their volume through increasing their floor areas if they cannot make use of natural light and other forms of externally supplied energy. This helps to explain why buildings tend to increase in height instead of horizontally.

Compactness may be seen as a proxy for building geometry. The quest for the most efficient building form has been taking place for quite a long time. Martin and March (1975) find that a perfect rectangular parallelepiped (half cube) would be the most efficient shape, targeted at minimizing heat losses. Ratti, Baker, and Steemers (2005) propose a method accounting for ground losses, under which the optimum shape would be a cube. Recent findings from Hachem, Athienitis, and Fazio (2012) confirm that deviations of building shape from the rectangle involve increasing heating loads. Steadman et al. (2000) categorize nondomestic built forms for investigating energy implications and Steadman, Evans, and Batty (2009) point out building depth as an important feature for energy use, as it is related to the limits of passive zones.

Metrics of building compactness are quite simple to define. They are usually a ratio of the envelope surface to the building volume—surface-to-volume (STV; Ratti, Baker, and Steemers 2005; Rode et al. 2014). The STV is significantly correlated (R ² >.75) with building energy performance (Fichera et al. 2016).

Bourdic, Salat, and Nowacki (2012) define building compactness through three different metrics: volumetric compactness, size factor, and form factor (Table 2). The form factor is useful to eliminate the effect of size and to keep only the effect of form. Nevertheless, this expression assumes that the building area grows at a rate of two-thirds of the rate of the volume growth. Volumetric compactness is equivalent to the STV ratio (Ratti, Baker, and Steemers 2005). The STV is claimed to be the most useful indicator of compactness for estimating heating and cooling needs, as a function of the shape of the building. Anisimova (2011) suggests that STV should be lower than .8. The effect of compactness is inverse for heating and cooling (typically, compact buildings have lower needs for heating but higher ones for cooling), although not necessarily in in the same magnitude. Compactness metrics measure the potential for heat transfer through the envelope, which also depends on other factors not linked to the building form, notably construction materials.

Passivity

The passive or nonpassive condition is often referred to as a factor of urban texture (Ratti, Baker, and Steemers 2005). For differentiating form and texture, see Lynch (1981). Solar passive energy use in buildings is more widespread than passive cooling. However, the reason why the latter is not further explored is because it largely depends on smaller scale building elements (Kisilewicz 2015), and thus it is not directly related to urban form. Passive zones are defined as those which can be naturally lit and ventilated and that make use of solar gains for heating (Baker, Hoch, and Steemers 1992). These areas are those within six meters (or twice the ceiling height) from the façade (Figure 3). The size of passive areas is deeply dependent on building geometry. Passivity strategies tend to conflict with those for improving building compactness, as there are trade-offs between these two effects (Ratti, Baker, and Steemers 2005). While increasing passive zones may increase solar gains, it potentially leads to higher thermal losses due to larger exposed surfaces. The prevailing effect on the building thermal balance and the choice for the best strategy depends on the local climate.

Figure 3.

A passive zone (transversal cut). Source: Ratti, Baker, and Steemers (2005).

The broader use of the passive/nonpassive ratio has occurred after Ratti, Baker, and Steemers (2005), who claim that the relation between passive and nonpassive areas in buildings is a better indicator for energy consumption than the STV ratio for the latitudes considered. Steadman, Evans, and Batty (2009) argue that building depth, largely determining the limits of passive zones, is an important feature for energy use. Evans, Liddiard, and Steadman (2017) consider the ratio of volume to exposed wall area as an approximation of the average depth of a plan (the ratio is equivalent to half plan depth). They claim that this indicator is likely to be related with the need for air conditioning. This is supported by the findings from Casas Castro Marins and Andrade Roméro (2013).

Bourdic, Salat, and Nowacki (2012) measure this property at the block or neighborhood scale and call it rate of passive volume. Metrics employed may be applied across different scales. However, this property also indicates a potential because it does not consider aspects such as the shadowing caused by urban obstructions.

Shading

As building passivity does not consider obstructions, it should be complemented by a shading indicator. Shading significantly affects the energy balance of buildings. Baker, Hoch, and Steemers (1992) and Ratti, Baker, and Steemers (2005) suggest the urban horizon angle (UHA), which refers to sky obstructions and is used to evaluate the effects of overshadowing from adjacent buildings. Baker and Steemers (2000) define it as the average elevation of the skyline from the center of the facade being considered (Figure 4). It is measured for each facade, as the height (H) of the opposite buildings divided by the canyon width (W): H/W = tan(UHA). Ratti, Baker, and Steemers (2005) apply an algorithm to measure the UHA in complex urban geometries, using the perpendicular line to the facade and a weighted average of six more directions in the range [−67.5°, +67.5°] from the perpendicular line. These authors also use the obstruction sky view (OSV) angle, which quantifies the luminance of the obstructing facades. It is defined as H/W = cos(OSV). These metrics involve the relation between street width and building height (H/W), also called urban canyon ratio (Coseo and Larsen 2014). Street width alone significantly influences the global radiation yield of the urban canyon, with larger widths leading to larger yields (van Esch, Looman, and de Bruin-Hordijk 2012). Alternatively, the depth ratio is used to consider the effect of mutual shading between building facades (Hachem, Athienitis, and Fazio 2011, 2012). Considering an L geometry, it is measured by the ratio of branch to main wing lengths. In addition, the sky view factor (SVF) describes the openness of the sky to radiative transport at a given ground location (J.-P. Kim and Guldmann 2014) and may be considered a proxy for density (Giridharan, Ganesan, and Lau 2004). The SVF significantly influences the urban air temperature and the UHI (Oke 1981) with lower SVF leading to increased temperatures (Chun and Guldmann 2014).

Figure 4.

Urban horizon angle. Source: Adapted from Baker and Steemers (2000).

Shading may ultimately be considered an effect of density (resulting from building’s heights and proximity). It is often treated separately due to its lower scale implications on individual buildings and on the urban microclimate. It is likely that trade-offs exist between cooling loads in buildings and energy needs for mobility. In addition, shading (especially from trees) may contribute to more pleasant walking environments in summer.

Robinson (2006) discusses the effectiveness of shading indicators. They are usually calculated for a single building. Alternative metrics, applicable at the urban block or neighborhood scales, would constitute interesting inputs for an energy analysis. For instance, the average UHA of an urban frontage along a street could provide a useful picture of the shading effect derived from the built environment.

Knowles (2003) presents the concept of solar envelope to translate the boundaries under which buildings will not shadow their surroundings. Also, Vermeulen et al. (2015) develop a framework for optimizing the urban layout to maximize direct solar irradiation. These studies, however, rely on more complex techniques than the metrics presented above to explore the effect of shading on energy demand. In practice, urban layouts are frequently constrained by existing structures, where optimum solutions may not always be possible.

Orientation

The orientation of a building determines the amount of solar radiation incident on each of its facades and consequently the requirements for space heating and cooling (Steemers 2003; Ratti, Baker, and Steemers 2005). In the northern hemisphere, south-facing facades should be favored to maximize solar gains in winter (Littlefair 1998; Hachem, Athienitis, and Fazio 2011). East and west orientations gather excessive solar gains in summer, and north-oriented facades gather the lowest solar gains, thus implying higher heating needs (Hachem, Athienitis, and Fazio 2012). Orientation is widely incorporated in several modeling platforms for building simulation—for instance, the renowned lighting and thermal method (Baker, Hoch, and Steemers 1992). While it is pointed as a low cost and easily addressed building feature to achieve passive solar design (Morrissey, Moore, and Horne 2011), it has been argued that its effect on building solar potential is not that straightforward (Kanters and Wall 2014). The optimization of a single building orientation is estimated to have minimal reductions on annual energy use and costs. However, for a whole community or urban area, important savings may be achieved by improving building orientation (Hemsath 2016). Additionally, street orientation is claimed to influence the UHI due to the fact that it determines the shading and ventilation of the urban canyon. However, it does not seem to significantly explain urban air temperatures (Coseo and Larsen 2014).

There is not much innovation on measuring orientation. It is usually measured in degrees or radians, representing the azimuth (e.g., Okeil 2010; Wilson 2013). This measure applies to each exposed building facade. Simplifications are needed in order to get a dominant orientation for a single building, such as the longest building axis (Hemsath 2016) or the alignment in relation to the street, with an east–west axis being preferred for maximizing buildings facing south, and thus increasing solar gains in winter (Ko 2013; Ko and Radke 2014; Kanters and Wall 2016). Streets with an east–west direction are also claimed to potentiate solar gains from increased street widths (van Esch, Looman, and de Bruin-Hordijk 2012). However, while using street orientation instead of building orientation may simplify the analysis, it may happen that buildings’ main façades are not aligned with the streets. In the cases where buildings are not aligned with the street, street orientation is unlikely to be a suitable indicator. Although the orientation of a building or group of buildings is a low-scale structural variable, it has been pointed out that it can be measured at the neighborhood scale (Mitchell 2005).

Transport Networks

The following sections review urban attributes that are not included in the “built environment” category. They draw on features of urban networks and on their ability to promote specific transport means. The urban transport network may be characterized in different ways, referring to the network itself or to “where” the network leads.

Connectivity

Connectivity is largely influenced by the spatial configuration of the urban network and is a widely acknowledged urban property influencing travel patterns (e.g., Ewing and Cervero 2001; Baran, Rodríguez, and Khattak 2008). Topologically, it can be interpreted as the degree to which two points communicate with each other. It depends on the number of intersections and on the spatial arrangement of the network edges (Cervero and Kockelman 1997), but it is also dependent on block size (Forsyth et al. 2008). Marshall and Gong (2009) compare two archetypal network structures, tree and grid, pointing out that connectivity can be understood and exist at different scales (macro and micro). At a macrolevel, it is generally accepted that a greater connectivity shortens distances to be traveled and potentially leads to reduced energy demand (Litman and Steele 2005). This, however, is not the consensus. Crane (2000) argues that shorter trips may lead to increased trip frequency. At a microlevel, a higher connectivity may encourage walking and other soft modes, making urban activities more accessible by these modes (Ewing and Cervero 2010). Cervero and Kockleman (1997) point out that gridded networks can work at a lower level to improve pedestrian movements. However, at a higher scale, they are associated to higher levels of road traffic, as in the case of superblocks.

Simple metrics of connectivity are the percentage of grid streets within a radius from a given point (Boarnet and Sarmiento 1998) or the percentage of four-way intersections or the percent of cul-de-sacs (S. Handy 1996a). Bourdic, Salat, and Nowacki (2012) describe connectivity as the number of different ways of going from one point to another. They propose a set of three indicators: (i) the intensity of network intersections (number of intersections per area unit); (ii) the average distance between intersections, which can be seen as a proxy of how walking-friendly a city is; and (iii) the cyclomatic number, which is in line with the definition of connectivity presented by these authors. The cyclomatic number represents the number of primary loops in a network and is given by: µ = L − N + 1, where L is the number of links, corresponding to the different sections of streets between every intersections, and N is the number of nodes, corresponding to the intersections in a road network. In order to be comparable, this number should refer to an area unit.

Chen, Claramunt, and Ray (2014) present a set of four indexes for measuring overall network connectivity: (i) the beta index (β) gives the average number of edges (e) per node (n) in a given network, that is, $β = e / n (0 \leq β \leq (n - 1) / 2)$ ; (ii) the cyclomatic number provides the number of circuits to indicate a gap between e and n, while counting the number of subnetworks q (q = 1 for a fully connected network), that is, $µ = e - n + q (0 \leq µ \leq (n - 1) (n - 2) / 2)$ ; (iii) the alpha index (α) is defined as the proportion between the actual and maximal number of circuits in a fully connecting planar network: it is given as $α = 2 µ / (n - 1) (n - 2) = (e - n + q) / (2 n - 5 q) (0 \leq α \leq 1)$ ; and finally (iv) the gamma index (γ) is the ratio between the actual and maximal number of edges: $γ = 2 e / n (n - 1) = e / [3 * (n - 2)] (0 \leq γ \leq 1)$ .

Connectivity is also a measure provided by the space syntax technique² (Hillier and Hanson 1984). In this context, it is described as the number of lines that are connected to a certain line (Baran, Rodríguez, and Khattak 2008). This technique is grounded in graph theory and assumes that better connected areas attract a higher density of flows. Nevertheless, the concept of connectivity may sometimes be fuzzy. The so-called syntactical measures (under a metameasure of accessibility) found in the space syntax are closer to the concept of connectivity than to the concept of accessibility, as defined below. For instance, the integration measure corresponds to how easy it is, from each line, to reach all other lines of the urban system. Nevertheless, the space syntax may provide significant advantages over existing methods to measure street connectivity and syntactical accessibility and to describe part-to-whole relationships of street networks (Baran, Rodríguez, and Khattak 2008).

At a microscale, Moudon et al. (1997) focus on the connectivity and safety of pedestrian facilities when analyzing the effect of site design on walking trips. Here, (pedestrian) connectivity is characterized by how well a network links land-use parcels or activities within a given area. This is claimed to be a function of route directness and the completeness of (pedestrian) facilities, and to be closely linked to the concept of accessibility. The authors measure route directness as (i) the ratio of actual route distance traveled to a straight-line distance and (ii) the walking distance contour, which plots the area from which a pedestrian can reach the center within a 800 meters walk or less. Completeness of pedestrian facilities is computed by the total length of the sidewalk to the total length of block (or street) frontage.

Connectivity metrics should focus on the configuration of the urban network and avoid mixing other attributes, such as the density of network nodes, or the β index. More empirical knowledge is needed on the effects of connectivity at different scales. In all, pedestrian connectivity is to be favored in urban areas, while large-scale connectivity has mixed effects.

Accessibility

Accessibility has no single definition. Gould (1969) calls it a “slippery notion” that is difficult to measure. Here, accessibility relates to the ease of reaching desired destinations or opportunities (aligned with Levine and Garb [2002] and Geurs and Ritsema van Eck [2001]). The implications on energy demand depend, to a great extent, on the transport mode considered. Pedestrian and transit accessibility should, thus, be increased. S. Handy (1993) claims that both local and regional accessibility should be increased, as they are associated to lower levels of nonwork travel. Accessibility is usually linked to a type of travel “cost,” typically distance or time. S. L. Handy and Niemeier (1997) consider three types of accessibility measures: the simplest refers to cumulative opportunities, whereas gravity-based and utility-based measures are more complex. Most measures consist of two parts: one concerning impedance and the other concerning a given activity or trip attraction.

A different classification is proposed by Papa and Coppola (2012), who consider two types of measures: active and passive accessibility. The former is defined as a proxy for the ease of reaching activities located in different zones j of the study area for a certain purpose. It is given by the following expression: $A_{act, i} = \sum_{j} g (W_{j}) f (c_{i j})$ , where W_j is the activity to reach in zone j and c_ij is the generalized cost of reaching zone j from zone i. Passive accessibility is seen as a proxy for the opportunity of an activity located in a zone i to be reached by potential consumers coming from all the other zones j of the study area for a given purpose. It is defined as: $A_{pas}_{, i} = \sum_{j} g (W_{j}) f (c_{i j})$ , where W_j are the potential “consumers” of the activity to be reached in zone i and c_ij is the generalized cost of reaching zone i from zone j. Likewise, Geurs and Ritsema van Eck (2001) distinguish between activity-based (availability of opportunities to satisfy individual needs) and utility-based accessibility (benefits that individuals may retrieve from the land use and transport system). Activity-based measures include, among others, distance measures, contour measures, and potential measures.

Bourdic, Salat, and Nowacki (2012) simply measure the number of activities within 500 meters from a PT station in relation to the total number of activities from a given area. Additionally, C. Silva (2013) considers distance limitations in characterizing potential accessibility measures, that is, farther away places from a certain origin are less accessible. She proposes a method—the Structural Accessibility Layer (SAL)—for comparing accessibility levels in the territory. SAL is applied by transport mode to different types of opportunities generating travel. It comprises two accessibility measures: the diversity of activity DivAct index (Table 2) and the accessibility cluster.

Also with a geographical perspective, Sevtsuk and Mekonnen (2012) propose five “centrality” indexes, while establishing a parallel between these indexes and other existent accessibility measures: Reach, Gravity Index, Betweenness, Closeness, and Straightness indices. Despite their usefulness in characterizing the urban environment, the fact that these were classified as centrality indexes shows that accessibility is entangled with other concepts, notably diversity (Ewing et al. 2016).

Selecting an accessibility metric is not an easy task. The choice may depend on the desired level of detail and degree of complexity. In the case of accessibility, contour measures may not properly characterize the urban environment. The DivAct (C. Silva and Pinho 2010) and the Reach measure (Sevtsuk and Mekonnen 2012) have produced very interesting results. Utility-based metrics are dependent on individual preferences and thus are less linked to the urban environment in a stricter sense. In the case of an urban form–energy analysis, they may not be the most suitable choice.

Distance to central business district (CBD)/centrality

The distance (of housing) to CBD is often used to determine energy demand for travel purposes. However, this distance is linked to other concepts already presented in this article (e.g., density, diversity, and accessibility). The CBD is the area with the highest employment density and number of trip ends (W. P. Anderson, Kanaroglou, and Miller 1996). Although distances traveled are directly influenced by other attributes of the urban area, notably density, mix of land uses, or network design, the distance to the CBD is often seen as an evaluation criterion by itself. For instance, in centralized systems, areas closer to the CBD are more likely to have higher accessibility indexes. As such, this indicator provides an incomplete description of the urban area because it is dependent on a set of other factors, and so, it should not be analyzed in isolation to draw conclusions on urban performance (this is also true for other attributes). However, maybe because it is so intuitive, it is still applied in current research.

Greater distances to the CBD are often associated to an increased use of energy for commuting (Alford and Whiteman 2009; Naess 2005; Holden and Norland 2005). This is particularly true for monocentric cities. However, the definition of the “center” is an intrinsic issue. While this indicator may be quite easily measured in a concentric or radial city (W. P. Anderson, Kanaroglou, and Miller 1996), it may not be so evident in polycentric cites and neither are the energy implications. S. Lee and Lee (2014) argue that a polycentric city may have lower commuting distances, despite being less favorable to PT commuting. For a comprehensive review on polycentric patterns, see Marshall and Gong (2009).

Ewing and Cervero (2010) consider the distance to downtown as a built environment variable falling under the category “destination accessibility.” Frost, Linneker, and Spence (1998) estimate average travel distances in order to assess the so-called excess commuting in several urban areas. The authors concluded that urban form is a determinant of observed increases in trip length. However, they point out that analyzing the commuting efficiency of a city based only in one measure may be misleading. Following Frost, Linneker, and Spence (1998) approach, Chowdhury, Scott, and Kanaroglou (2013) evaluate commuting efficiency through commuting distances (actual, minimum, and maximum) against urban form criteria, typically, the jobs–housing balance.

Interestingly, Brotchie et al. (1996) combine in a single indicator the distance of amenities to CBD and a metric of diversity. The jobs–housing dispersal index (x) is one of the variables taken from Brotchie’s triangle and is the ratio of the average distance of jobs from the CBD (jobs dispersal) to the average distance of workers’ households from the CBD (housing dispersal). Mathematically: $x = (\frac{1}{E} \sum_{j} d_{j} e_{j}) / (\frac{1}{H} \sum_{j} d_{j} h_{j})$ , where E and H are the total number of metropolitan jobs and workers’ households, respectively; e_j and h_j are, respectively, the number of jobs and workers’ households in zone j; and d_j is the network distance from the CBD to zone j. For a monocentric city, the index is 0.

Under a slightly different perspective, B. Lee (2007) collects a set of centralization indexes, which define the extent of concentration of employment near the CBD. Additional centralization measures may be found in S. Lee and Lee (2014), where they use a principal component analysis to derive a centrality index. Centrality is found to have a significant negative association to VMT. In all, despite having an urban center that is well-defined, walkable and well served by PT may reduce motorized needs, it is important to keep some smaller clusters of urban services close to housing (as described in the diversity section). The right balance, though, is not clear.

Proximity to PT

After the identification of three D’s influencing travel demand (Cervero and Kockleman 1997), three more were added thereafter. Distance to PT is referred by Ewing and Cervero (2010) as the fifth D (the fourth being destination accessibility and the sixth is referred to as demand management, involving parking issues). Proximity to PT is acknowledged to promote the adoption of this transport mode (Kitamura, Mokhtarian, and Laidet 1997; Dieleman, Dijst, and Burghouwt 2002). This may be related to the travel time budget theory, which states that people are willing to spend a certain daily amount of time for travel purposes. Even if the transit system has very good frequencies and is very reliable, if the bus or train stop is not reachable within a short walk from home (or work), it is not likely to be successfully adopted. While door-to-door service may address this issue, the transit system may become slower. Clearly, this is not a straightforward point.

Proximity to PT relies on simple metrics, such as the distance to the nearest station or stop (Kitamura, Mokhtarian, and Laidet 1997). Ewing and Cervero (2010) claim that it is commonly measured as (i) the average of the shortest street routes from the residences or workplaces in an area to the nearest station or stop, (ii) PT route density, (iii) distance between stops, or (iv) the number of stations per unit area.

Corroborating the importance of the location of transport infrastructure, Marshall and Gong (2009) propose the spinality concept for classifying urban areas according to the spatial configuration of settlements with respect to a strategic transport network. Spinality is described as the extent to which an urban area is aligned with a transport axis. Two indicators are proposed for capturing spinality: (i) the buffer ratio or B-ratio, which is the proportion of the built-up area within a given distance of strategic routes and (ii) the route length ratio or A-ratio, which is the proportion of strategic route length to total route length.

While the distance to transport infrastructure is not the only factor affecting the adoption of this transport mode, it may certainly play an important role. Indicators are generally simple and should be considered in the urban form–energy analysis. More complex indicators resorting to location of land uses may fall under the domain of accessibility measures.

Design

Design is related to low-scale features of the urban environment that often have a subjective nature (Ewing and Handy 2009). Design features are related to how pleasant the urban environment is, thus indicating the potential for using soft modes instead of motorized ones. Design stands for the third D influencing travel, which has not been explored enough (Cervero and Kockelman 1997). Design measures applied by these authors cover a wide spectrum and can be included in some of the attributes considered earlier. For instance, the “proportion of intersections that are four-way” or the “number of dead ends and cul-de-sacs” may be considered as indicators of connectivity. Similarly, the existence of trees may fit in the design attribute.

Ewing and Cervero (2001) reinforce the importance of network design in encouraging the use of alternative modes of transport. They claim that “grids with skinny streets, short blocks, and traffic-calming measures are hardly conducive to long distance car travel. Conversely, grids with six lanes of fast-moving traffic, long blocks, and no medians or pedestrian refuge islands are no panacea for pedestrians” (p. 100).

Apart from the configuration of the street network and the presence of trees, additional frequently referred design features are sidewalks and parking. The first is associated to pedestrian-friendly environments (Southworth 1997) and recognized to be positively correlated with the frequency of nonmotorized trips (Kitamura, Mokhtarian, and Laidet 1997). Sidewalk length and width are both negatively correlated with VMT (Salon et al. 2012).

Parking supply is often assumed to promoting driving (Cervero and Gorham 1995) and creating obstructed spaces, through displacing active land use occupation (Ewing and Cervero 2001). These authors add that parking affects travel behavior twofold, in terms of supply and location in relation to streets and buildings. The relation between parking and VMT may also be examined using parking prices (Salon et al. 2012). Willson (2005) claims that parking supply and policy may improve the effects of transit-oriented development. If properly located, parking may work as a dissuasive factor for driving to final destinations (usually CBDs) and reduce congestion and air emissions.

Existing studies exploring urban design usually find low correlations between the metrics selected and travel demand. However, it is claimed that the effect of urban design on travel is likely to be a collective and cumulative one, involving multiple design features (Ewing and Cervero 2010) that may have a significant weight together.

Hierarchy

Bourdic, Salat, and Nowacki (2012) suggest that the scale hierarchy of urban elements is a significant attribute influencing energy demand. This is based on the idea that hierarchic structures are associated to a higher urban structural efficiency (Salingaros and West 1999). Salat and Bourdic (2011) argue that hierarchically organized urban systems follow power laws, with an important impact on their efficiency: “the most energy efficient structure for a complex flow-driven system is a highly organized state, based on power law distributions.” They add that “in a highly dense and connected city with high levels of complexity, functional mix allows sparing significant amounts of inputs (materials, energy…)” (p. 1196).

These authors propose a metric for measuring the hierarchy of the street network that is based on the width and on the classification of streets, applied at city or district scales. This concept is also implicit on the design D of Cervero and Kockelman (1997), where they consider arterial speed limits and street widths as properties of urban space influencing travel demand. Alternatively, Casas Castro Marins and Andrade Roméro (2013) use the share of road area (road area/total urban area) to describe the road network.

Batty (2012) claims that urban development patterns may be classified through their fractal dimension, where network structure may be the link for existing agglomeration economies and scaling effects. The effect of hierarchy on urban efficiency is still underexplored. More empirical research on this topic could help uncovering desirable scale efficiencies of energy demand. Additional metrics could be useful for exploring the existing relationships.

Summary and Discussion

Cities have a complex nature. There are several attributes of urban form acknowledged to influence energy demand. This influence takes place at different scales (the building, the block, the neighborhood, and the whole city). Urban form influences two major urban sectors: housing and mobility. In the first case, urban form mainly determines the building type and the UHI effect, influencing thermal needs (heating and cooling) and lighting. In regard to transport, urban form influences mode choice, distances traveled, and trip frequency, with an impact on the overall energy needs. While microattributes are most important in buildings (except for density), for transports, it is at the meso- to the macroscale that effects of urban form are mostly felt.

Conceptually, the literature is consistent in identifying the urban attributes with energy relevance. However, there are no clear boundaries on the understanding of such attributes. While these may be invoked with different meanings (e.g., compactness has a different meaning depending on the scale of analysis), it may also happen that urban attributes which often occur together are taken as synonyms (for instance, accessibility and connectivity). This might be explained by the fact that a single urban attribute may bring others attached. The consequence is the persistence of some ambiguity when dealing with urban indicators, which complicates conclusions on the effect of urban form, particularly on energy demand. An adequate definition of urban attributes is essential for advancing the existing knowledge, as the results of urban analysis may vary depending on the choice of the indicators and metrics. The literature on urban form and energy has not carefully considered the impact of such choice.

Existing metrics are not evenly distributed over different urban form attributes. There are many ways of measuring an urban property, and a single urban attribute may require more than an indicator to be properly described. A study of the effects of urban form on energy (whatever source or vector of energy is considered) should use variables that describe the object under analysis, that is, the physical environment. This article argues that urban metrics should be as simple as possible to be implemented in planning practice. Simple metrics can be as informative as very sophisticated formulas, as long as they are targeted at effectively describing the urban environment. Table 2 summarizes the metrics for the urban attributes considered. Such a collection is expected to guide future research on the topic.

Moreover, research has been placing its emphasis on a limited set of attributes at a time. This article argues that the influence of urban form on energy is a result of a set of intertwined variables, which should be analyzed at the same time under a comprehensive analysis framework. Their overall effect is a result of intra- and intersector trade-offs. This article goes across the literature, identifying attributes of urban form with energy relevance at different scales, describing how each urban attribute is expected to place its influence. Table 3 illustrates the variation (increase or decrease) in the energy needs that each attribute is expected to create, for each the three end uses (heating, cooling, and mobility). While it is not possible, under the scope of a qualitative review, to produce a robust prescription of how the urban environment should be organized, the main conclusions are summarized as follows:

Table 3.

Impact of Each Attribute of the Urban Environment on Energy Demand.

Urban Attribute	Buildings— Heating Needs	Buildings—Cooling Needs	Travel— Energy Needs
↑Density	↓/↑	↓/↑	↓/↑
↑Diversity	NA/ND	NA/ND	↓
↑Green areas	↓/↑	↓	↓/↑
↑Compactness	↓	↑	NA/ND
↑Passivity	↓/↑	↓/↑	NA/ND
↑Shading	↑	↓	↓/↑
↑Orientation(preferred)	↓	↓	NA/ND
↑Connectivity	NA/ND	NA/ND	↓/↑
↑Accessibility	NA/ND	NA/ND	↓
↑Distance to CBD	NA/ND	NA/ND	↓
↑Design	↓/↑	↓/↑	↓
↑Proximity to PT(preferred)	NA/ND	NA/ND	↓
↑Hierarchy	NA/ND	NA/ND	↓

Note: Legend: ↑—increases; ↓—decreases; ↓/↑—possible trade-offs; NA/ND—not applicable/not defined; PT = public transport.

Density (residential) should be promoted near PT and clustered around commerce and services for reducing private motorized trips (affects diversity too). In latitudes where heating is an important energy use, and wherever possible, infill development should be promoted and buildings heights should be carefully considered in order to prevent blocking solar access.

The desired diversity mix may not necessarily be 50/50. Nevertheless, ratios close to 0 and 1 in each neighborhood should be avoided. In areas that are dominantly residential, creating daily-need services (e.g., a coffee shop, a bakery, a bank…) not only reduces trip needs but also creates more lively and vibrant urban environments.

Keeping lower scale and connected green infrastructure within the urban core may create pleasant walking environments as well as regulating the urban microclimate, while larger green areas in the periphery may work as a green belts preventing urban dispersion along with other ecological benefits—the city of Toronto being a good example (Green Infrastructure Ontario Coalition 2014).

At the building scale, wherever heating is a dominant use, compact geometries should be favored, minimizing exposed facades for preventing heat losses to the outside. This may be largely achieved by promoting nondetached housing. A favorable orientation should be promoted (south-facing facades in the northern hemisphere), while trees represent low-cost amenities to regulate excessive solar gains and reduce cooling needs in summer.

Lower scale connectivity regarding pedestrian paths should be encouraged (for instance, by reducing the size of blocks). Higher-scale street connectivity should be considered carefully, as it is often associated with motorized travel. In order to enable this distinction, urban elements such as streets should be hierarchically arranged. Increased levels of desirable accessibility may be attained, depending on the mode, by improving the proximity to PT infrastructure or by bringing activities closer to inhabitants, in the case of pedestrians (as a combination of density and diversity mentioned earlier).

Finally, lower scale design elements, such as trees, sidewalks, and parking, also exert an important influence. Parking availability in the urban core should be reduced (for instance, replaced by trees), while parking located in the urban boundaries next to key PT hubs (e.g., park and ride stations) could promote the shift to more sustainable modes. Sidewalks should be proportional to street width, whereas narrower and shaded streets are usually more pedestrian-friendly than large boulevards.

The large bulk of existing studies provides useful insights on how the different elements of urban form affect energy for different urban purposes. However, the estimation of the weight of each indicator, as well as the overall balance of trade-offs, may only be achieved by a comprehensive quantitative framework and may vary from city to city. The indicators and metrics gathered here are expected to contribute twofold for the state of the art: (i) supporting the emergence of sound innovative analytical frameworks by collecting a set of urban form attributes with energy relevance and (ii) providing planners with systematic and quantifiable means to design more energy-efficient urban areas.

Conclusion and Future Research

The growing acknowledgment of the relationship between urban form and energy sets urban planning as a relevant field for addressing the sustainability and the climate change agendas. There is a wide array of studies exploring the influence of the physical attributes of urban areas on energy. However, the existing knowledge is sectorial and dispersed. A better understanding of existing links and a systematic characterization of urban form attributes with energy relevance would be of major importance for new and existing cities. This article reviews the literature on two main urban sectors (buildings and transports) and collects (i) evidence of how urban form affects energy for three main end uses (heating, cooling, and mobility) and (ii) urban form indicators and metrics to address this link.

Comprehensive research is needed considering housing and mobility features. The urban attributes identified, and the indicators and metrics gathered, represent a cornerstone to be tested and explored in future quantitative research on urban form and energy. It is anticipated that the effects of urban attributes may not be additive, but synergistic. The ability to uncover such synergies may open new grounds to reshape our cities, with significant potential for energy conservation. A careful characterization of urban attributes will enable the assessment of their individual contribution and also the estimation of the overall energy balance.

Moreover, the description of the impact of each attribute offers urban planners a framework for addressing energy in urban areas and some guidance regarding the impacts of their choices. The different attributes of the urban environment should be simultaneously considered. Measures range from defining building’s orientation or planting trees to defining density boundaries or specifying urban functions. The choice may depend on context features such as the local climate, as well as on the key issues to be tackled (residential energy vs. mobility). Understanding the existing trade-offs between urban variables and how they impact the size of energy flows within the urban system is key to prioritize action and energy-efficient urban planning policies.

Footnotes

Acknowledgment

The first author gratefully thanks FCT—Fundação para a Ciência e a Tecnologia, for the financial support of her PhD studies SFRH/BD/52305/2013.

Declaration of Conflicting Interests

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

Funding

The authors received financial support for the research, authorship, and/or publication of this article. The first author received funding from FCT—Fundação para a Ciência e a Tecnologia, for the development of her PhD studies (BD/52305/2013).

Notes

References

Alford

Whiteman

2009. “Macro-urban Form and Transport Energy Outcomes: Investigations for Melbourne.” Road & Transport Research: A Journal of Australian and New Zealand Research and Practice 18:53.

Anderson

J. E.

Wulfhorst

Lang

2015. “Energy Analysis of the Built Environment—A Review and Outlook.” Renewable and Sustainable Energy Reviews 44:149–58.

Anderson

W. P.

Kanaroglou

P. S.

Miller

E. J.

1996. “Urban Form, Energy and the Environment: A Review of Issues, Evidence and Policy.” Urban Studies 33:7–35.

Anisimova

2011. “The Capability to Reduce Primary Energy Demand in EU Housing.” Energy and Buildings 43:2747–51.

Baker

Hoch

Steemers

1992. “The LT Method.” Version 1.2. Cambridge, MA: Cambridge Architectural Research and Martin Centre for Architectural and Urban Studies, University of Cambridge.

Baker

Steemers

2000. Energy and Environment in Architecture: A Technical Design Guide. London, UK: E & FN Spon.

Banister

Watson

Wood

1997. “Sustainable Cities: Transport, Energy, and Urban Form.” Environment and Planning B 24:125–43.

Baran

P. K.

Rodríguez

D. A.

Khattak

A. J.

2008. “Space Syntax and Walking in a New Urbanist and Suburban Neighbourhoods.” Journal of Urban Design 13:5–28.

Batty

2012. “Building a Science of Cities.” Cities 29:S9–16.

10.

Batty

Carvalho

Hudson-Smith

Milton

Smith

Steadman

2008. “Scaling and Allometry in the Building Geometries of Greater London.” The European Physical Journal B 63:303–14.

11.

Behnisch

Bräuer

Schumacher

Deilmann

2012. “Urban Pattern Specification using Spatial Metrics and Density Values.” In Proceedings of the 26th International Conference “Informatics for Environmental Protection,” edited by Arndt

Knetsch

Pillman

, 319–29. Dessau, Germany: Shaker Verlag Aachen.

12.

Boarnet

M. G.

Sarmiento

1998. “Can Land-use Policy Really Affect Travel Behaviour? A Study of the Link between Non-work Travel and Land-use Characteristics.” Urban Studies 35:1155–69.

13.

Bourdic

Salat

Nowacki

2012. “Assessing Cities: A New System of Cross-scale Spatial Indicators.” Building Research & Information 40:592–605.

14.

Bremer

Mayr

Wichmann

Schmidtner

Rutzinger

2016. “A New Multi-scale 3D-GIS-approach for the Assessment and Dissemination of Solar Income of Digital City Models.” Computers, Environment and Urban Systems 57:144–54.

15.

Brotchie

J. F.

Anderson

Gipps

P. G.

McNamara

1996. “Urban Productivity and Sustainability—Impacts of Technological Change.” In Transport, Land-Use and the Environment, edited by Hayashi

Roy

, 81–99. Dordrecht, the Netherlands: Kluwer Academic.

16.

Casas Castro Marins

K. R.

Andrade Roméro

2013. “Urban and Energy Assessment from a Systemic Approach of Urban Morphology, Urban Mobility, and Buildings: Case Study of Agua Branca in Sao Paulo.” Journal of Urban Planning and Development 139:280–91.

17.

Cervero

1989. “Jobs-housing Balancing and Regional Mobility.” Journal of the American Planning Association 55:136–50.

18.

Cervero

1996. “Mixed Land-uses and Commuting: Evidence from the American Housing Survey.” Transportation Research Part A: Policy and Practice 30:361–77.

19.

Cervero

Gorham

1995. “Commuting in Transit versus Automobile Neighborhoods.” Journal of the American Planning Association 61:210–25.

20.

Cervero

Kockelman

1997. “Travel Demand and the 3Ds: Density, Diversity, and Design.” Transportation Research Part D: Transport and Environment 2:199–219.

21.

Cervero

Murakami

2010. “Effects of Built Environments on Vehicle Miles Traveled: Evidence from 370 US Urbanized Areas.” Environment and Planning A 42:400–18.

22.

Chen

Claramunt

Ray

2014. “A Spatio-temporal Modelling Approach for the Study of the Connectivity and Accessibility of the Guangzhou Metropolitan Network.” Journal of Transport Geography 36:12–23.

23.

Chowdhury

T. A.

Scott

D. M.

Kanaroglou

P. S.

2013. “Urban Form and Commuting Efficiency: A Comparative Analysis across Time and Space.” Urban Studies 50:191–207.

24.

Chun

Guldmann

J. M.

2014. “Spatial Statistical Analysis and Simulation of the Urban Heat Island in High-density Central Cities.” Landscape and Urban Planning 125:76–88.

25.

Coseo

Larsen

2014. “How Factors of Land Use/Land Cover, Building Configuration, and Adjacent Heat Sources and Sinks Explain Urban Heat Islands in Chicago.” Landscape and Urban Planning 125:117–29.

26.

Crane

2000. “The Influence of Urban Form on Travel: An Interpretive Review.” Journal of Planning Literature 15:3–23.

27.

Dieleman

F. M.

Dijst

Burghouwt

2002. “Urban Form and Travel Behaviour: Micro-level Household Attributes and Residential Context.” Urban Studies 39:507–27.

28.

Echenique

M. H.

Hargreaves

A. J.

Mitchell

Namdeo

2012. “Growing Cities Sustainably: Does Urban Form Really Matter?” Journal of the American Planning Association 78:121–37.

29.

European Environment Agency. 2015. Trends and Projections in Europe 2015. Tracking Progress towards Europe’s Climate and Energy Targets. Luxembourg: European Environment Agency.

30.

Evans

Liddiard

Steadman

2017. “3DStock: A New Kind of Three-dimensional Model of the Building Stock of England and Wales, for Use in Energy Analysis.” Environment and Planning B: Planning and Design 44:227–55.

31.

Ewing

Bartholomew

Winkelman

Walters

Chen

2008. Growing Cooler: The Evidence on Urban Development and Climate Change. Washington, DC: Urban Land Institute.

32.

Ewing

Cervero

2001. “Travel and the Built Environment: A Synthesis.” Transportation Research Record 1780:87–114.

33.

Ewing

Cervero

2010. “Travel and the Built Environment. A Meta-Analysis.” Journal of the American Planning Association 76:265–94.

34.

Ewing

Hajrasouliha

Neckerman

K. M.

Purciel-Hill

Greene

2016. “Streetscape Features Related to Pedestrian Activity.” Journal of Planning Education and Research 36 (1), 5–15.

35.

Ewing

Haliyur

Page

G. W.

1994. “Getting around a Traditional City, a Suburban Planned Unit Development, and Everything in between.” Transportation Research Record 1466:53–62.

36.

Ewing

Hamidi

2015. “Compactness versus Sprawl A Review of Recent Evidence from the United States.” Journal of Planning Literature 30:413–32.

37.

Ewing

Handy

2009. “Measuring the Unmeasurable: Urban Design Qualities Related to Walkability.” Journal of Urban Design 14:65–84.

38.

Ewing

Rong

2008. “The Impact of Urban Form on U.S. Residential Energy Use.” Housing Policy Debate 19:1–30.

39.

Fichera

Inturri

La Greca

Palermo

2016. “A Model for Mapping the Energy Consumption of Buildings, Transport and Outdoor Lighting of Neighbourhoods.” Cities 55:49–60.

40.

Forsyth

Hearst

Oakes

J. M.

Schmitz

K. H.

2008. “Design and Destinations: Factors Influencing Walking and Total Physical Activity.” Urban Studies 45:1973–96.

41.

Frank

L. D.

Pivo

1994. “Impacts of Mixed Use and Density on Utilization of Three Modes of Travel: Single-occupant Vehicle, Transit, Walking.” Transportation Research Record 1466:44–52.

42.

Frost

Linneker

Spence

1998. “Excess or Wasteful Commuting in a Selection of British Cities.” Transportation Research Part A: Policy and Practice 32:529–38.

43.

Geurs

Ritsema van Eck

2001. Accessibility Measures: Review and Applications. Evaluation of Accessibility Impacts of Land-use Transportation Scenarios, and Related Social and Economic Impact. Bilthoven, the Netherlands: National Institute of Public Health and the Environment.

44.

Giridharan

Ganesan

Lau

S. S. Y.

2004. “Daytime Urban Heat Island Effect in High-rise and High-density Residential Developments in Hong Kong.” Energy and Buildings 36:525–34.

45.

Glaeser

E. L.

Kahn

M. E.

2010. “The Greenness of Cities: Carbon Dioxide Emissions and Urban Development.” Journal of Urban Economics 67:404–18.

46.

Gordon

Richardson

H. W.

1989. “Gasoline Consumption and Cities: A Reply.” Journal of the American Planning Association 55:342–46.

47.

Gordon

Richardson

H. W.

1997. “Are Compact Cities a Desirable Planning Goal?” Journal of the American Planning Association 63:95–106.

48.

Gould

1969. Spatial Diffusion, Resource Paper. Washington, DC: Association of American Geographers.

49.

Green Infrastructure Ontario Coalition. 2014. Health, Prosperity and Sustainability: The Case for Green Infrastructure in Ontario. Toronto, Canada: Green Infrastructure Ontario Coalition.

50.

Hachem

Athienitis

Fazio

2011. “Investigation of Solar Potential of Housing Units in Different Neighborhood Designs.” Energy and Buildings 43:2262–73.

51.

Hachem

Athienitis

Fazio

2012. “Evaluation of Energy Supply and Demand in Solar Neighborhood.” Energy and Buildings 49:335–47.

52.

Handy

1993. Regional Versus Local Accessibility: Implications for Nonwork Travel. Berkeley: University of California Transportation Center.

53.

Handy

1996a. “Methodologies for Exploring the Link between Urban Form and Travel Behavior.” Transportation Research Part D: Transport and Environment 1:151–65.

54.

Handy

1996b. “Travel Behaviour Issues Related to Neo-traditional Developments—A Review of the Research.” Presented at TMIP Conference on Urban Design, Telecommuting, and Travel Behavior, US Department of Transportation, FHWA, Washington, DC.

55.

Handy

S. L.

Niemeier

D. A.

1997. “Measuring Accessibility: An Exploration of Issues and Alternatives.” Environment and Planning A 29:1175–94.

56.

Hemsath

T. L.

2016. “Housing Orientation’s Effect on Energy Use in Suburban Developments.” Energy and Buildings 122:98–106.

57.

Hillier

Hanson

1984. The Social Logic of Space. Cambridge, MA: Cambridge University Press.

58.

Holden

Norland

I. T.

2005. “Three Challenges for the Compact City as a Sustainable Urban Form: Household Consumption of Energy and Transport in Eight Residential Areas in the Greater Oslo Region.” Urban Studies 42:2145–66.

59.

Høyer

K. G.

Holden

2003. “Household Consumption and Ecological Footprints in Norway—Does Urban Form Matter?” Journal of Consumer Policy 26:327–49.

60.

Jabareen

Y. R.

2006. “Sustainable Urban Forms Their Typologies, Models, and Concepts.” Journal of Planning Education and Research 26:38–52.

61.

Jenks

Burton

Williams

1996. The Compact City: A Sustainable Urban Form? London, UK: E & FN Spon.

62.

Kanaroglou

South

2001. “Can Urban Form Affect Transportation Energy Use and Emissions?” Energy Studies Review 9:22–40.

63.

Kanters

Wall

2014. “The Impact of Urban Design Decisions on Net Zero Energy Solar Buildings in Sweden.” Urban, Planning and Transport Research 2:312–32.

64.

Kanters

Wall

2016. “A Planning Process Map for Solar Buildings in Urban Environments.” Renewable and Sustainable Energy Reviews 57:173–85.

65.

Kaza

2010. “Understanding the Spectrum of Residential Energy Consumption: A Quantile Regression Approach.” Energy Policy 38:6574–85.

66.

Kaza

Curtis

M. P.

2014. “The Land Use Energy Connection.” Journal of Planning Literature 29:355–69.

67.

Kenworthy

J. R.

2006. “The Eco-city: Ten Key Transport and Planning Dimensions for Sustainable City Development.” Environment and Urbanization 18:67–85.

68.

Kim

J.-P.

Guldmann

J. M.

2014. “Land-use Planning and the Urban Heat Island.” Environment and Planning B: Planning and Design 41:1077–99.

69.

Kim

Park

Lee

J. S.

2014. “Meso- or Micro-scale? Environmental Factors Influencing Pedestrian Satisfaction.” Transportation Research Part D: Transport and Environment 30:10–20.

70.

Kisilewicz

2015. “Passive Control of Indoor Climate Conditions in Low Energy Buildings.” Energy Procedia 78:49–54.

71.

Kitamura

Mokhtarian

P. L.

Laidet

1997. “A Micro-analysis of Land Use and Travel in Five Neighborhoods in the San Francisco Bay Area.” Transportation 24:125–58.

72.

Knowles

R. L.

2003. “The Solar Envelope: Its Meaning for Energy and Buildings.” Energy and Buildings 35:15–25.

73.

2013. “Urban Form and Residential Energy Use: A Review of Design Principles and Research Findings.” Journal of Planning Literature 28:327–51.

74.

Radke

J. D.

2014. “The Effect of Urban Form and Residential Cooling Energy Use in Sacramento, California.” Environment and Planning B 41:573–93.

75.

Lee

2007. ““Edge” or “Edgeless” Cities? Urban Spatial Structure in U.S. Metropolitan Areas, 1980 to 2000*.” Journal of Regional Science 47:479–515.

76.

Lee

Moudon

A. V.

2006a. “The 3Ds + R: Quantifying Land Use and Urban Form Correlates of Walking.” Transportation Research Part D: Transport and Environment 11:204–15.

77.

Lee

Moudon

A. V.

2006b. “Correlates of Walking for Transportation or Recreation Purposes.” Journal of Physical Activity & Health 3:S77.

78.

Lee

2014. “The Influence of Urban Form on GHG Emissions in the U.S. Household Sector.” Energy Policy 68:534–49.

79.

Levine

Garb

2002. “Congestion Pricing’s Conditional Promise: Promotion of Accessibility or Mobility?” Transport Policy 9:179–88.

80.

Reynolds

J. F.

1993. “A New Contagion Index to Quantify Spatial Patterns of Landscapes.” Landscape Ecology 8:155–62.

81.

Litman

Steele

2005. Land Use Impacts on Transport. Victoria, Canada: Victoria Transport Policy Institute.

82.

Littlefair

1998. “Passive Solar Urban Design: Ensuring the Penetration of Solar Energy into the City.” Renewable and Sustainable Energy Reviews 2:303–26.

83.

Lund

2003. “Testing the Claims of New Urbanism: Local Access, Pedestrian Travel, and Neighboring Behaviors.” Journal of the American Planning Association 69:414–29.

84.

Lynch

1981. Good City Form. Cambridge, MA: MIT Press.

85.

Marshall

Gong

2009. WP4 Deliverable Report: Urban Pattern Specification. London, UK: University College London.

86.

Martin

March

1975. Urban Space and Structures. Cambridge, UK: Cambridge University Press.

87.

Mitchell

2005. Urban Development, Form and Energy Use in Buildings: A Review for the Solutions Project. Leeds, UK: The University of Leeds.

88.

Morrissey

Moore

Horne

R. E.

2011. “Affordable Passive Solar Design in a Temperate Climate: An Experiment in Residential Building Orientation.” Renewable Energy 36:568–77.

89.

Moudon

Hess

Snyder

Stanilov

1997. “Effects of Site Design on Pedestrian Travel in Mixed-use, Medium-density Environments.” Transportation Research Record 1578:48–55.

90.

Naess

2003. “Urban Structures and Travel Behaviour. Experiences from Empirical Research in Norway and Denmark.” European Journal of Transport and Infrastructure Research 3:155–78.

91.

Naess

2005. “Residential Location Affects Travel Behavior—But How and Why? The Case of Copenhagen Metropolitan Area.” Progress in Planning 63:167–257.

92.

Newman

Kenworthy

2006. “Urban Design to Reduce Automobile Dependence.” Opolis 2:35–52.

93.

Newman

P. W. G.

Kenworthy

J. R.

1989. “Gasoline Consumption and Cities.” Journal of the American Planning Association 55:24–37.

94.

Nijkamp

Rienstra

S. A.

1996. “Sustainable Transport in a Compact City.” In The Compact City; A Sustainable Urban Form? edited by Jenks

Burton

Williams

, 190–99. London, UK: E and FN Spon.

95.

O’Brien

W. T.

Kennedy

C. A.

Athienitis

A. K.

Kesik

T. J.

2010. “The Relationship between Net Energy Use and the Urban Density of Solar Buildings.” Environment and Planning B: Planning and Design 37:1002–21.

96.

Oke

T. R.

1981. “Canyon Geometry and the Nocturnal Urban Heat Island: Comparison of Scale Model and Field Observations.” Journal of Climatology 1:237–54.

97.

Oke

T. R.

1982. “The Energetic Basis of the Urban Heat Island.” Quarterly Journal of the Royal Meteorological Society 108:1–24.

98.

Oke

T. R.

1988. “Street Design and Urban Canopy Layer Climate.” Energy and Buildings 11:103–13.

99.

Okeil

2010. “A Holistic Approach to Energy Efficient Building Forms.” Energy and Buildings 42:1437–44.

100.

O’Kelly

M. E.

Niedzielski

M. A.

2009. “Are Long Commute Distances Inefficient and Disorderly?” Environment and Planning A 41:2741–59.

101.

O’Neill

R. V.

Krummel

J. R.

Gardner

R. H.

Sugihara

Jackson

DeAngelis

D. L.

Milne

B. T.

Turner

M. G.

Zygmunt

Christensen

S. W.

Dale

V. H.

Graham

R. L.

1988. “Indices of Landscape Pattern.” Landscape Ecology 1:153–62.

102.

Papa

Coppola

2012. Gravity-based Accessibility Measures for Integrated Transport-land Use Planning (GraBAM). Brussels, Belgium: COST Office.

103.

Pont

M. B.

Haupt

2005. “The Spacemate: Density and the Typomorphology of the Urban Fabric.” Nordisk Arkitekturforskning 4:55–68.

104.

Porter

C. D.

Brown

Dunphy

R. T.

Vimmerstedt

2013. Effects of the Built Environment on Transportation: Energy Use, Greenhouse Gas Emissions, and Other Factors. Washington, DC: US Department of Energy.

105.

Ratti

Baker

Steemers

2005. “Energy Consumption and Urban Texture.” Energy and Buildings 37:762–76.

106.

Rickwood

Glazebrook

Searle

2008. “Urban Structure and Energy—A Review.” Urban Policy and Research 26:57–81.

107.

Robinson

2006. “Urban Morphology and Indicators of Radiation Availability.” Solar Energy 80:1643–48.

108.

Rode

Keim

Robazza

Viejo

Schofield

2014. “Cities and Energy: Urban Morphology and Residential Heat-energy Demand.” Environment and Planning B: Planning and Design 41:138–62.

109.

Ryan

McNally

M. G.

1995. “Accessibility of Neotraditional Neighborhoods: A Review of Design Concepts, Policies, and Recent Literature.” Transportation Research Part A: Policy and Practice 29:87–105.

110.

Salat

2009. “Energy Loads, CO2 Emissions and Building Stocks: Morphologies, Typologies, Energy Systems and Behaviour.” Building Research & Information 37:598–609.

111.

Salat

Bourdic

2011. “Power Laws for Energy Efficient and Resilient Cities.” Procedia Engineering 21:1193–98.

112.

Salingaros

N. A.

West

B. J.

1999. “A Universal Rule for the Distribution of Sizes.” Environment and Planning B: Planning and Design 26:909–23.

113.

Salon

Boarnet

M. G.

Handy

Spears

Tal

2012. “How Do Local Actions Affect VMT? A Critical Review of the Empirical Evidence.” Transportation Research Part D: Transport and Environment 17:495–508.

114.

Seto

Dhakal

Bigio

Blanco

Delgado

Dewar

Huang

Inaba

Kansal

Lwasa

McMahon

Müller

Murakami

Nagendra

Ramaswami

2014. “Human Settlements, Infrastructure and Spatial Planning.” In Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by Edenhofer

Pichs-Madruga

Sokona

Farahani

Kadner

Seyboth

Adler

Baum

Brunner

Eickemeier

Kriemann

Savolainen

Schlömer

von Stechow

Zwickel

Minx

J. C.

, 923–1000. Cambridge, UK: Cambridge University Press.

115.

Sevtsuk

Mekonnen

2012. “Urban Network Analysis. A New Toolbox for ArcGIS.” Revue Internationale de Géomatique 22:287–305.

116.

Silva

2013. “Structural Accessibility for Mobility Management.” Progress in Planning 81:1–49.

117.

Silva

Pinho

2010. “The Structural Accessibility Layer (SAL): Revealing How Urban Structure Constrains Travel Choice.” Environment and Planning A 42:2735–52.

118.

Silva

Oliveira

Leal

2016. “Urban Morphology and Energy: Progress and Prospects.” Urban Morphology 20:72–73.

119.

Soltani

Allan

2006. “Analyzing the Impacts of Microscale Urban Attributes on Travel: Evidence from Suburban Adelaide, Australia.” Journal of Urban Planning and Development 132:132–37.

120.

Southworth

1997. “Walkable Suburbs?: An Evaluation of Neotraditional Communities at the Urban Edge.” Journal of the American Planning Association 63:28–44.

121.

Stead

Marshall

2001. “The Relationships between Urban Form and Travel Patterns: An International Review and Evaluation.” European Journal of Transport and Infrastructure Research 1:113–41.

122.

Steadman

Bruhns

H. R.

Holtier

Gakovic

Rickaby

P. A.

Brown

F. E.

2000. “A Classification of Built Forms.” Environment and Planning B: Planning and Design 27:73–91.

123.

Steadman

Evans

Batty

2009. “Wall Area, Volume and Plan Depth in the Building Stock.” Building Research & Information 37:455–67.

124.

Steemers

2003. “Energy and the City: Density, Buildings and Transport.” Energy and Buildings 35:3–14.

125.

Steiner

R. L.

1994. “Residential Density and Travel Patterns: Review of the Literature.” Transportation Research Record 1466:37–43.

126.

Stone

Rodgers

M. O.

2001. “Urban Form and Thermal Efficiency: How the Design of Cities Influences the Urban Heat Island Effect.” Journal of the American Planning Association 67:186–98.

127.

Taha

1997. “Urban Climates and Heat Islands: Albedo, Evapotranspiration, and Anthropogenic Heat.” Energy and Buildings 25:99–103.

128.

Tran

D. X.

Pla

Latorre-Carmona

Myint

S. W.

Caetano

Kieu

H. V.

2017. “Characterizing the Relationship between Land Use Land Cover Change and Land Surface Temperature.” ISPRS Journal of Photogrammetry and Remote Sensing 124:119–32.

129.

United Nations-Habitat. 1990. National Design Handbook Prototype on Passive Solar Heating and Natural Cooling of Buildings. Nairobi, Kenya: United Nations Centre for Human Settlements (UN-Habitat).

130.

van den Hoek

2008. The MXI (Mixed-use Index) as Tool for Urban Planning and Analysis, Corporations and Cities: Envisioning Corporate Real Estate in the Urban Future. Brussels, Belgium: TU Delft.

131.

van Esch

M. M. E.

Looman

R. H. J.

de Bruin-Hordijk

G. J.

2012. “The Effects of Urban and Building Design Parameters on Solar Access to the Urban Canyon and the Potential for Direct Passive Solar Heating Strategies.” Energy and Buildings 47:189–200.

132.

Vaz Monteiro

Doick

K. J.

Handley

Peace

2016. “The Impact of Greenspace Size on the Extent of Local Nocturnal Air Temperature Cooling in London.” Urban Forestry & Urban Greening 16:160–69.

133.

Vermeulen

Knopf-Lenoir

Villon

Beckers

2015. “Urban Layout Optimization Framework to Maximize Direct Solar Irradiation.” Computers, Environment and Urban Systems 51:1–12.

134.

Williams

Jenks

Burton

2000. Achieving Sustainable Urban Form. London, UK: E & FN Spon.

135.

Willson

2005. “Parking Policy for Transit-oriented Development: Lessons for Cities, Transit Agencies, and Developers.” Journal of Public Transportation 8:79–94.

136.

Wilson

2013. “Urban Form and Residential Electricity Consumption: Evidence from Illinois, USA.” Landscape and Urban Planning 115:62–71.

137.

Zhao

2011. “Energy Consumption Based Urban Texture Analysis.” Presented at the Eighteenth International Seminar on Urban Form. Urban morphology and the post-carbon city, 26-29 August, Montreal, Canada.