Understanding the compactness of employment activities in high-density cities through cellphone location data

Abstract

Urban development in many cities worldwide presents a dispersed development with urban sprawl and excessive land use. Compact city development facilitates an efficient use of land that aims to reduce excessive land consumption and commuting. However, the ignorance of the relationships between land use and the compactness of employment activities has resulted in excessive commuting. The separation of jobs and housing due to functional zoning has caused criticism in compact city development. This research aims to examine the compactness of workers’ employment activities and its relationship with land use to enrich our knowledge on compact city development. The relationship between urban and employment activity compactness is examined through cellphone location data in Shanghai, China. A set of indicators, including intrazonal employment ratio and interzonal commuting distance, are examined to measure the compactness of employment activities. Land-use compactness is measured through building density, land-use diversity, and accessibility. The effect of land-use compactness on employment activity compactness is analyzed through structural equation modeling. A high proportion of residential land use reduces the intrazonal employment ratio and increases the average interzonal commuting distance. A high proportion of industrial and commercial land use increases the intrazonal employment ratio. These are the spatial characteristics to improve urban compactness. This study contributes to decision-making for compact development, which is of great significance to promoting compact activities. It can also provide spatial planning policies that facilitate workers’ access to local employment opportunities with short travel distances through land use planning for mixed land use.

Keywords

Compact city employment activities cellphone data planning policy

Introduction

Compact city development has been regarded as a sustainable paradigm for large cities for urban and regional planning worldwide (Angel et al., 2020). A compact city aims at a high-density, mixed-use, and intensified urban form to reduce trip distances (Vallance et al., 2005). The exploration of the spatial characteristics of a compact city has been receiving great interest to reduce land-use consumption, increase urban density, and lessen urban expansion (Burgess, 2000). Building a compact city is closely linked to the characteristics of high density, mixed land use, and public transport (Howley, 2009), in which compact cities are usually defined as a means to achieve compactness (Adolphson, 2010). The compactness of a city can be attained in terms of density, transport and service accessibility, and mixed use (Jabareen, 2006). Galster et al. (2001) explore compactness with regard to density, continuity, concentration, clustering, centrality, mixed use, and proximity. For example, Hong Kong values high-density development, transit-oriented development, and mixed use as the essence of developing a compact city (Chen et al., 2019; Lang et al., 2019, 2020). Mixed use, road network density, and accessibility are the fundamental factors for developing a vibrant and compact city, in which small blocks, a high mixture of land use, and easily acquired transit and amenities contribute to its vitality (Long and Huang, 2019).

A compact city reinstates a high-density urban-built environment as a sustainable urban form. The basis for compact city development is achieved by building high-density cities, which in turn is the way out for high-density cities, especially for highly populous and developing Asian cities. Studies have investigated the spatial features of building a compact city through diverse perspectives. For example, Chen et al. (2015) measured urban expansion in terms of density gradient, population size, and urban structures by using remote sensing data in the Yangtze River Delta Region of China and found that this area is compact and centralized but is undergoing fast expansion. Zhang et al. (2018) investigated the relations between the urban built environment and the loss of green space and determined that the compact urban form associated with concentrated agglomerations is conducive to growth and protects urban green space. Unraveling the mechanisms underlying compactness is vital for measuring a compact city and its spatial characteristics.

A compact city has various aspects to determine urban compactness (Burton, 2002), but the lack of compactness measurement of employment activities has led to difficulties in investigating the effects of compactness, e.g., regarding the adjacency of employment and residency to achieve compactness (Zhou et al., 2018). The relationship between urban employment activities and the built environment has drawn great attention (Ewing et al., 2016). However, few studies have investigated the performance of employment activities in relation to the built environment of a compact city. Such a relationship is complex and calls for a valid measurement at a fine-grained scale to understand urban compactness. The key research questions of this paper are (1) how does land use influence the compactness of employment activities and the journey-to-work travel? and (2) what spatial planning strategies can promote the compactness of employment activities?

This study aims to examine how urban land-use compactness, including land-use density and diversity, affects the compactness of employment activities by using cellphone location big data and its planning implications. The remainder of this paper is organized as follows. In the literature review, we briefly retrospect the compact city, land use, and commuting patterns for exploring employment activities. In the case study section, we present the study site, data source, and cellphone location data applied in this study. We also indicate an investigation method on urban compactness through mobile phone data. In Built environment compactness section, we provide an overview of the results of the spatial analysis on urban land-use compactness by exploring mixed use, building density, and road accessibility. In Compactness of workers’ employment activities in Shanghai section, we demonstrate and address the results of the analyses. In the final section, we conclude the findings with contributions to planning policies and suggest future studies.

Literature review

Compact city

Since the early 1990s, the concept of the compact city has drawn growing interest from urban researchers. The spatial features of compact cities are associated with a mixed land-use, high-density urban form, and an efficient public transport system. Burton (2002) emphasizes that density is a multifaceted concept, and indicators for measuring density include population density, dwelling density, land use density, and floor area ratio. Jacobs (1961) showed that mixed use, which refers to the mixture of residence, employment, and services, creates successful and livable neighborhoods. A compact city usually has relatively high intraurban daily mobility with an efficient public transport system. Compact development can help the city become more livable and sustainable. It can reduce developmental and transport costs and decrease reliance on private cars (Burton, 2002). Compact city policies are implemented in many cities to reduce car dependence and pollution. Compact development can control urban expansion and protect the open countryside. Land use and transport can be better integrated, and the geographical separation of daily activities will be decreased in compact cities (Howley, 2009). The utility of the compact city is challenged by some other researchers. For instance, Ferreira and Batey (2011) claim that urban densification leads to considerable time wasted because of traffic congestion. The focus of these opposing ideas is the traffic congestion caused by urban compaction owing to the mismatch between urban structure and residents’ activities.

Most studies on the compact city have focused on population distribution, land use, and building structure (Burton, 2002; Chen et al., 2015; Ewing et al., 2016) and ignored the daily activities of residents. Meanwhile, research on the compactness of workers’ employment activities is lacking. Employment activities refer to work or work-related activities in this study. The existing compact development emphasizes mainly land-use intensification and transport infrastructure without considering workers’ employment activities because of the lack of detailed spatiotemporal mobility data. Compact employment activities mean that workers need to travel relatively short distances to accomplish their employment activities, which can reduce motorized commuting. The “spatial mismatch” between land-use patterns and workers’ employment activities has increased travel demand and caused many problems, such as traffic congestion and overconcentration of urban activities (Yang et al., 2012). While the literature has indicated that locating jobs and retail and public services in close proximity to residential areas can reduce travel demand considerably, few studies have suggested which forms of mixed use lead to the greatest dividends (Cervero and Duncan, 2006).

Land use and commuting patterns

The compact city is valid in reducing commuting distances (Boussauw et al., 2010). Many empirical studies have shown that commuting distance is affected by job accessibility, which is determined by land-use characteristics. Residential and employment land-use distributions and development density affect job accessibility and commuting patterns. Compact development, instead of sprawl development, can reduce long-distance commuting (Zhao et al., 2010). Antipova et al. (2011) suggest that land-use patterns are significant in reducing commuting times.

Lin and Yang (2009) found that land-use density and diversity affected travel demand directly and indirectly, and mixed-use development was related to considerable walking, cycling, and public transport usage. Wang et al. (2011) analyzed the relationship between land-use and activity–travel patterns at the household level. Most of these studies are aggregate analyses at the district or neighborhood level based on limited samples of travel survey data. However, how the elements of land-use patterns affect commuting patterns at a large scale from an individuals’ perspective remains to be explored (Handy, 2005).

The unit of analysis also affects the commuting analysis because the intrazonal commute would differ for different analysis units. Most existing studies on land use and commuting patterns are at the city level and use statistical yearbook data, or they are at the neighborhood level and use GPS data or travel diary data (Zhang et al., 2017). With residents’ microlevel mobility data of large sample size, exploring urban commuting from the individuals’ perspective is possible (Zhou and Long, 2016). The challenge is that obtaining individual or household travel diary data with a large sample size is labor-intensive and time-consuming.

The research methods in spatial analysis by using structural equation modeling (SEM) in SPSS AMOS provide a robust tool to build simple models theoretically and operationally to investigate multiple factors’ relations in understanding land use and commuting patterns in an urban built environment. Thus, the complex relationships between the built environment and activity–travel patterns can be explored via SEM (Lu and Pas, 1999; Song et al., 2016). For example, Lin and Yang (2009) analyzed 9 latent variables with 26 observed variables by using SEM. Ewing et al. (2016) assessed the characteristics of a built environment via SEM and determined that it has a strong influence on people’s travel through direct and indirect ways.

Exploring employment activities from cellphone big data

Uncovering urban mobility patterns that characterize residents’ activities is of great importance for urban and transport planning (Chen et al., 2020; Zhou et al., 2019). Huang et al. (2019) evaluated and characterized the urban vibrancy of Shanghai through spatial big data. The availability of cellphone location data enables us to explore commuting patterns with a large sample size. Cellphone location data have higher spatial accuracy and a larger sample size than a household travel survey (Zhou et al., 2018). Previous studies have shown that cellphone location data represent an appropriate proxy for individual mobility and a promising complement to the conventional travel survey (Calabrese et al., 2013). For instance, Diao et al. (2016) incorporated spatiotemporal activities by using mobile phone data and urban space to understand land use and assist in decision making for planning and design. Bokányi et al. (2019) provided a comparative study of three cities by identifying activity types and activity-driven land use by using mobile phone data. Most previous studies on commuting and employment activities have been based on traditional travel survey data, and the investigation of employment compactness by using fine-scale data is needed. On top of the traditional analysis of home-to-work travels based on travel survey data, journey-to-work trips and employment activities extracted from mobile phone data enable exploring the compactness of workers’ employment activities at a fine scale.

Case study

Study site

This study used Shanghai, China as the case city. Urban expansion has caused long-distance commuting. Shanghai is one of the central international metropolises; it is a typical city in China experiencing rapid urbanization, whose land-use expansion and demands for planning a compact city are prominent in China and Asia. The study area is the central-city area with the greatest high-density development in Shanghai, as shown in Figure 1. The Shanghai Municipal Government has established and supervises the Shanghai Municipal Territorial Spatial Planning System to implement the Shanghai Master Plan (2017–2035). Shanghai’s spatial planning promotes compact and intensive land use. New urban development and industrial projects should be built within the urban development boundary to reduce further land use. Shanghai territorial spatial planning strengthens the guidance for promoting integrated construction and intensive land use to form a functionally complex and compact city.

Figure 1.

Case study area (central-city area of Shanghai City, China).

Cellphone location data

The cellphone location data of two weeks in October 2014 were collected from a Chinese mobile company, which is the largest cellphone operator in Shanghai. The number of cellphone users in this dataset was 16 million and could represent a high percentage of the entire population in Shanghai. The IDs of cellphone users, date, time, and the coordinates of the cellphone towers were recorded in the time span of 2 weeks as long as the phone was in service. The location of the cellphone user was estimated using the coverage area of the nearest cellphone tower. The positioning accuracy of cellphones was approximately 100–300 m in the central-city areas.

Journey-to-work trips from cellphone big data

The users’ home and work locations were identified from the cellphone location data in accordance with the methods of Calabrese et al. (2013) and Zhou et al. (2018). Initially, the cumulative stay duration was calculated. The locations where the user had the longest total stay time and stayed at that location for not less than 2 h during the night (20:00–06:00) and day (9:00–18:00) were recognized as the most possible home and job locations of the cellphone users, respectively (equations (1) and (2)). Then, the frequency of repeated occurrences of each user in the same spatial location was calculated. If the cellphone user’s nighttime residency appeared more than five times within 10 normal working days of the two weeks, then it was considered the user’s home location. If the cellphone user’s daytime residency appeared more than five times within 10 normal working days of the two weeks, then it was the user’s workplace. Students and retired seniors who live and work at the same location were excluded, and the remaining workers were examined in this study.

{PO}_{-} Hom e_{i} = \frac{{Duration}_{i j}}{\sum {Duration}_{i j}} from 20 : 00 to 06 : 00

(1)

{{PO}_{-} Work}_{i} = \frac{{Duration}_{i j}}{\sum {Duration}_{i j}} from 9 : 00 to 18 : 00

(2)

where

{{PO}_{-} Home}_{i}

and

{{PO}_{-} Work}_{i}

reflect the possibility of home and work locations of cellphone user i, respectively;

{Duration}_{i j}

is the accumulated time that cellphone user i spent at tower j during daytime and nighttime; and

\sum {Duration}_{i j}

is the total time that cellphone user i spent at different towers during daytime and nighttime.

Analytical framework

Urban compact development emphasizes the links between jobs and housing as a means to increase the intrazonal employment ratio and reduce the interzonal commuting distance. In this study, compactness of the built environment was characterized by mixed use and density. Mixed use was measured using the proportion of residential, commercial, and public services, and industrial land use. Density was measured using the density of residential, commercial and public services, and industrial land use. The land-use mix was calculated through the collation of residence and employment in the three land-use types. The compactness of workers’ employment activities was measured through the intrazonal employment ratio and the interzonal commuting distance by using cellphone location data. The relations between the built environment compactness and compactness of employment activities were then analyzed using SEM in IBM® SPSS® AMOS, an SPSS module. SEM was used to analyze the spatial relationships between the built environment compactness and compactness of employment activities at the 2 km level. The results of the standardized analysis were visualized in AMOS to reflect complex relationships and display the contributions. AMOS graphic was used to draw the SEM model with a model sample size of 229 grids to represent the central urban area of Shanghai City. This analysis model contained the observed variables, including the intrazonal employment ratio, interzonal commuting distance, proportion of commercial land use, proportion of industrial land use, proportion of residential land use, land use density, and road network density (km/km square). The estimate yielded results for standardized regression weights, correlations, covariances, and variances. The analytical framework of SEM is shown in Figure 2.

Figure 2.

Analytical framework of SEM.

Built environment compactness

Some researchers have developed composite indicators to measure compactness (Burton, 2002). The present study measured urban compactness from the perspectives of mixed use, building density, and road accessibility.

Mixed use

Land-use diversity represents the degree to which various land uses exist within an urbanized area. Lands with residential and employment functions were extracted from the land use map to calculate the land-use mix. The residential and employment lands were classified into three types: (1) lands with commercial function, public services, and office usages were classified as workplaces of the service and office workers; (2) lands with industrial function were classified as the workplaces of manufacturing workers; and (3) lands with residence usage were classified as residential areas.

In terms of employment activities, mixed use reflects how people use urban space for residence and employment. In this study, a mixed-use index was developed to measure the colocation of various land uses for residence and employment. The total area in a grid was the sum of commercial and public service, industrial, and residential areas. The percentage of commercial and public service area (Comm%) was the commercial and public service area over the total area. The percentage of the industrial area (Indu%) for a particular zone was the industrial area over the total area. The Comm% of the grid cells within the inner ring was higher than that in other areas (Figure 3(b)), whereas the Indu% of the grid cells within the inner ring was lower than that in other areas (Figure 3(c)). The percentage of residential area (Resi%) was the residential area over the total area (Figure 3(d)).

Figure 3.

(a) Land use patterns, (b) percentage of commercial area (Comm%), (c) percentage of industrial area (Indu%), and (d) percentage of residential area (Resi%).

Building and road densities

High-density development can reduce the reliance on vehicle trips, protect rural land, and reduce energy consumption. The floor area ratio was used to reflect building density and was calculated using the building information data provided by the Shanghai City Planning and Land Resources Bureau, which included the building’s name, address, main function, number of stories, and footprint area of each building. The floor area ratio was relatively high within the inner ring, as shown in Figure 4(a).

Figure 4.

(a) Building and (b) road densities.

High transport accessibility might worsen the compactness of workers’ employment activities because workers may prefer to reside far from their workplaces where housing would be relatively inexpensive. Road density (DEN_Road) was calculated using the road area over the area of urban land to reflect road accessibility. The density of the road network was higher within the inner ring than that in other central-city areas (Figure 4(b)).

Compactness of workers’ employment activities in shanghai

Spatial pattern of employment activity compactness

The compactness of workers’ employment activities measured from the perspectives of the intrazonal employment ratio and interzonal commuting distance is shown in Figure 5. The average intrazonal employment ratio in the central city was 15.72%. The average ratio within the inner ring (18.16%) was higher than that in other areas (15.35%). The average interzonal commuting distance in the central city was 9.62 km. The average interzonal commuting distance in the inner ring (9.46 km) was shorter than that in other areas (9.64 km).

Figure 5.

Compactness of workers’ employment activities: (a) intrazonal employment ratio and (b) interzonal commuting distance.

Relationships between built environment and employment activity compactness

The SEM built in SPSS AMOS was applied to analyze five multiple independent variables with two dependent variables (Tables 1 and 2). The data of nine variables were assessed in the SEM analysis, in which seven were tested as valid and input into the model analysis. The model was created with standardized regression that estimated the values of standardized regression weights, covariances, and variances, in which the predicted variance in the two dependent variables was explained by the collective proportions of the five independent variables. The covariance values showed the correlations among the independent variables, and the standardized regression weights indicated the proportion of changes in the independent variables attributed to a singular standard deviation unit of the changes in the dependent variables (Figure 6). In the standardized regression shown in Table 1, a statistically significant negative relationship was found between the average distance of travel and road network density. Residential land use proportion contributed to the decrease in the intrazonal employment ratio, and commercial land use proportion contributed to the decrease in the average distance of travel. The result also indicated a strong relationship between the intrazonal employment ratio and industrial land use proportion and the average distance of travel and residential land use proportion. Similarly, the relationship between the intrazonal employment ratio and commercial land use proportion was also strong and positive, in contrast to its negative relation with the average distance of interzonal commuting. Road density demonstrated a strong and positive relationship with the intrazonal employment ratio but a strong and negative relationship with the average distance of interzonal commuting.

Table 1.

Effect of the built environment on urban employment activities.

Standardized regression weights			Estimate	S.E.	C.R.	P
Dis_Comm	<---	DEN_Land	.093	271.229	1.199	.230
Intra_Empl	<---	DEN_Land	−.069	.010	−.847	.397
Intra_Empl	<---	Resi%	−.145	.029	−2.226	.026*
Dis_Comm	<---	Indu%	−.060	1,096.829	−.927	.354
Intra_Empl	<---	Indu%	.307	.040	4.565	***
Intra_Empl	<---	Comm%	.291	.057	4.724	***
Dis_Comm	<---	Comm%	−.160	1,558.492	−2.719	.007**
Intra_Empl	<---	DEN_Road	.462	.003	5.541	***
Dis_Comm	<---	DEN_Road	−.301	89.403	−3.776	***
Dis_Comm	<---	Resi%	.574	795.198	9.186	***

Note: The probability test P values indicate the significance level of all reported coefficients, where ***p < 0.001. **p < 0.01. *p < 0.05; −insignificant. S.E. denotes the standard error, and C.R. denotes the critical ratio that is the estimate divided by the standard error. Comm% refers to the commercial land use proportion, Indu% refers to the industrial land use proportion, Resi% refers to the residential land use proportion, DEN_land refers to the land use density, DEN_Road refers to the road network density, Intra_Empl refers to the intrazonal employment ratio, and Dis_Comm refers to the average interzonal commuting distance.

Table 2.

Covariances and variances of the structural equation modeling (SEM) analysis results.

Covariances			Estimate	S.E.	C.R.	P
Comm%	<-->	Indu%	−.004	.001	−5.124	***
Indu%	<-->	Resi%	−.011	.002	−6.619	***
Comm%	<-->	Resi%	.005	.001	4.993	***
DEN_Land	<-->	DEN_Road	.864	.097	8.901	***
Resi%	<-->	DEN_Road	.033	.016	2.149	.032
Indu%	<-->	DEN_Road	−.049	.012	−4.116	***
Comm%	<-->	DEN_Road	.024	.008	3.161	.002
Variances			Estimate	S.E.	C.R.	P
Comm%			.007	.001	10.677	***
Indu%			.016	.002	10.677	***
Resi%			.029	.003	10.677	***
DEN_Land			.390	.037	10.677	***
DEN_Road			3.768	.344	10.940	***
Error 1			.004	.000	10.677	***
Error 2			3,059,161.583	286,516.733	10.677	***

Note: The probability test P values indicate the significance level of all reported coefficients, where ***p < 0.001. **p < 0.01. *p < 0.05; −insignificant. Model fit summary, CFI = 1.000, TLI = 0.950, RMSEA = 0.045 [0.039, 0.050] 90% C.I. NFI = 0.918. NCP = 180.802 [139.970, 229.051] 90% C.I. model fit indices indicate a good fit of the model results. S.E. denotes the standard error, and C.R. denotes the critical ratio that is the estimate divided by the standard error. Comm% refers to the commercial land use proportion, Indu% refers to the industrial land use proportion, Resi% refers to the residential land use proportion, DEN_land refers to the land use density, DEN_Road refers to the road network density, Intra_Empl refers to the intrazonal employment ratio, and Dis_Comm refers to the average interzonal commuting distance.

Figure 6.

Standardized estimates and model output in AMOS graphic.

The standardized estimates revealed the relative contributions of independent variables to dependent variables shown in Table 1. The estimates reflected the proportion of one-unit changes in the independent variables contributing to the predicted one-unit change in the dependent variables. For example, the residential land use proportion decreased by −.145 for each unit increase in the intrazonal employment ratio. The significance tests of individual parameters in standardized regression coefficients are displayed below. The structural equation models fit the data well and provided a substantive interpretation of the results and findings. The analytic model also fit well and was statistically significant. Land uses were interconnected at the neighborhood level with a statistically significant result, in which commercial land use positively correlated with residential land use, and industrial land use passively corresponded to residential and commercial land use. Land use density was correlated positively with road density, but industrial land use was correlated negatively with road density (Table 2). Figure 6 illustrates the direct effects of road density, commercial land use, and industrial and residential land use density on the intrazonal employment ratio and its average distance. It also demonstrates the indirect effect via road network density (commercial, industrial, and residential land uses affected road network density in addition to the effects on the intrazonal employment ratio and its average distance). The accumulation of direct and indirect effects was the entire effect of the five endogenous variables on the two exogenous variables in the multi-relationships in SEM.

The results also showed that residential and commercial land uses were likely to mix with each other, whereas industrial land use was spontaneously exclusive with commercial and residential land uses. That is, residential and commercial land uses coexisted readily in mixed forms at the neighborhood scale to a large extent and were relatively isolated from industrial sites, which was unconducive to the mixed-use urban form in reality. In this study, high land-use density was accompanied by a dense road network, and the two facilitated each other as the premise. A negative correlation was also observed between industrial land use and road network density, i.e., the industrial land was characterized by large land parcels and accompanied by a small number of road networks only for freight and logistic function. In reality, industrial land use is usually separated from other land uses and is mainly a production space with few road networks connecting to the rest of the areas in cities.

Conclusions

This study investigated the compactness of employment activities in high-density cities by examining the relation between employment activities and the spatial characteristics of an urban built environment through cellphone location data in Shanghai, China. The intrazonal employment ratio and interzonal commuting distance were used to assess the compactness of workers’ employment activities in the central urban area of Shanghai City. Density, land-use diversity, and public transport system denoted built environment compactness. The relationship between land use and employment activities was analyzed with SEM.

The general findings can be summarized as follows. The compactness of workers’ employment activities was examined from the perspectives of the intrazonal employment ratio and interzonal commuting distance through cellphone location data. Large cellphone data provided a supplement to travel survey data because they can be used to explore the spatial variation of workers’ employment activities with a large sample size from the individual’s perspective. Individuals’ employment activities were found to be associated highly with the built environment. A high proportion of residential land use reduced the intrazonal employment ratio and increased the interzonal commuting distance. A high proportion of industrial and commercial land uses collectively resulted in increasing intrazonal employment ratio in the neighborhoods.

The existing urban built environment of the case study of Shanghai was characterized by a high-density urban form and a dense road network. These characteristics were conducive to improving urban compactness and promoting compact urban planning and development. Therefore, from the case study, spatial planning, strategies, and policies should support a continuously and strongly mixed land use and ensure public and green space under the condition of high building and road network densities. Mixed land-use functions and the public transportation system in high-density cities should be promoted. In addition to the adjustment in zoning, urban regeneration allows for the inclusion of diverse residences within the vicinity of work sites.

Developing a compact city needs spatial policies for planning vibrant places, impermanent use of spaces, and flexible time with round-the-clock activities and diverse functions. This study provides planning and design guidelines: (1) maintain high density and land-use intensity; (2) reduce block size and create a dense street network; (3) promote mixed use and walkable neighborhoods within walking or cycling distance. It paves the way for the long-term formation of a compact, high-density, and livable urban built environment.

This study shows that the land uses of a compact city affect employment activities in the urban built environment directly and indirectly. High-density mixed-use development provides its residents with convenient accessibility to workplaces through efficient transport systems, which is essential in a compact city. This study provides a vital lesson in high-density and compact development. A compact city should emphasize mixed land use rather than traditional functional zoning of land use, with industrial, residential, and commercial lands isolated in parcels. The integral components of urban planning with balanced land use functions contribute to compact employment activities. In the course of building a compact city, residential, commercial, and industrial land uses can be balanced progressively through infill, intensification, function replacement, and mixed land-use planning.

This study enriches our understanding of compactness in high-density cities and compact city development. It can help formulate urban planning policies for density and mixed land use with respect to the intrazonal employment ratio and interzonal commuting distance in relation to land use, density, and the road network. We hope to contribute to the improvement of urban form and spatial structure in urban planning and in the formulation of future planning policies. Although the spatial characteristics of an urban built environment were found to be closely correlated with employment activities, this analysis is still limited. The limitations of this study include the study on employment activities in relation to socioeconomic preferences and changes in time series. The possible socioeconomic effects were not included in this analysis. These issues will be explored in future studies.

Footnotes

Acknowledgements

Thank Dr. Cheng Shi at Tongji University for help with the mobile phone data analysis.

Declaration of conflicting interests

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

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Natural Science Foundation of China (41801147, 41801163, 51878457) and Shanghai Pujiang Program.

ORCID iDs

Xingang Zhou

Wei Lang

Xingang Zhou is currently working in the Department of Urban Planning, College of Architecture and Urban Planning at Tongji University. He obtained his Ph.D. degree in the Department of Urban Planning and Design of The University of Hong Kong and worked as a Post-Doctoral Fellow in the same department later on. His research interests include the application of GIS in urban planning. He is also interested in applying big data in urban planning and transport planning.

Wei Lang received the B. Eng. Degree in urban planning in 2007 from Harbin Institute of Technology, China; the M.S. degree in urban planning and policy design in 2009 from Politecnico di Milano, Italy; and a Ph.D. degree in urban planning and spatial analysis in 2016 from the Hong Kong Polytechnic University. He is currently working as an associate professor of urban and rural planning in the Department of Urban and Regional Planning at Sun Yat-sen University, Guangzhou, China. His research interest includes spatial structure and urban morphology, regional development and planning, spatial analysis, community participatory planning.

Anthony GO Yeh is a Chair Professor in Urban Planning and Geographic Information System (GIS) at the Department of Urban Planning and Design and the Director of the GIS Research Centre of the University of Hong Kong. He is an Academician of the Chinese Academy of Sciences and the Academy of Social Sciences in the U.K. He is also the President of the Asia GIS Association and was the Secretary General of the Asian Planning Schools Association.

Xinyi Niu is currently working as a professor in the Department of Urban Planning, College of Architecture and Urban Planning, Tongji University. He received a Bachelor of Engineering degree in urban planning in 1993, a master of Engineering degree in Urban Planning and Design in 1999, and a doctorate in urban planning and Design in 2008 from Tongji University, respectively. His main research interests include urban planning technology and methods, urban spatial information analysis and visualization, planning decision support system, etc

References

Adolphson

(2010) Kernel densities and mixed functionality in a multicentred urban region. Environment and Planning B: Planning and Design 37: 550–566.

Angel

Arango Franco

Liu

, et al. (2020) The shape compactness of urban footprints. Progress in Planning 139: 100429.

Antipova

Wang

Wilmot

(2011) Urban land uses, socio-demographic attributes and commuting: A multilevel modeling approach. Applied Geography 31: 1010–1018.

Bokányi

Kallus

Gódor

(2019) Collective sensing of evolving urban structures: From activity-based to content-aware social monitoring. Environment and Planning B: Urban Analytics and City Science 0: 2399808319848760.

Boussauw

Neutens

Witlox

(2010) Relationship between spatial proximity and travel-to-work distance: The effect of the compact city. Regional Studies 46: 687–706.

Burgess

(2000) The compact city debate: A global perspective. In: Jenks

Burgess

(eds) Compact Cities: Sustainable Urban Forms for Developing Countries. London: Routledge, pp. 9–24.

Burton

(2002) Measuring urban compactness in UK towns and cities. Environment and Planning B: Planning and Design 29: 219–250.

Calabrese

Diao

Di Lorenzo

, et al. (2013) Understanding individual mobility patterns from urban sensing data: A mobile phone trace example. Transportation Research Part C: Emerging Technologies 26: 301–313.

Cervero

Duncan

(2006) Which reduces vehicle travel more: Jobs-housing balance or retail-housing mixing.? Journal of the American Planning Association 72: 475–490.

10.

Chen

Yan

Gao

, et al. (2015) Quantifying circular urban expansion patterns of compact Chinese cities: The case of Yangtze river Delta. Environment and Planning B: Planning and Design 42: 279–299.

11.

Chen

Hui

, et al. (2019) Identifying urban spatial structure and urban vibrancy in highly dense cities using georeferenced social media data. Habitat International 89: 102005.

12.

Chen

Zhou

Yeh

(2020) Spatial accessibility to kindergartens using a spectrum combinational approach: Case study of Shanghai using cellphone data. Environment and Planning B: Urban Analytics and City Science. https://journals-sagepub-com-s.web.bisu.edu.cn/doi/abs/10.1177/2399808320954221.

13.

Diao

Zhu

Ferreira

, et al. (2016) Inferring individual daily activities from mobile phone traces: A Boston example. Environment and Planning B: Planning and Design 43(5): 920–940.

14.

Ewing

Hamidi

Grace

(2016) Compact development and VMT—Environmental determinism, self-selection, or some of both? Environment and Planning B: Planning and Design 43: 737–755.

15.

Ferreira

Batey

(2011) On why planning should not reinforce self-reinforcing trends: A cautionary analysis of the compact-city proposal applied to large cities. Environment and Planning B: Planning and Design 38(2): 231–247.

16.

Galster

Hanson

Ratcliffe

, et al. (2001) Wrestling sprawl to the ground: Defining and measuring an elusive concept. Housing Policy Debate 12: 681–717.

17.

Handy S (2005) Smart Growth and the Transportation-Land Use Connection: What Does the Research Tell Us? International Regional Science Review 28(2): 146–167.

18.

Howley

(2009) Attitudes towards compact city living: Towards a greater understanding of residential behaviour. Land Use Policy 26: 792–798.

19.

Huang

Zhou

, et al. (2019) Evaluating and characterizing urban vibrancy using spatial big data: Shanghai as a case study. Environment and Planning B: Urban Analytics and City Science 0: 2399808319828730.

20.

Jabareen

(2006) Sustainable urban forms: Their typologies, models, and concepts. Journal of Planning Education and Research 26: 38–52.

21.

Jacobs

(1961) The Death and Life of Great American Cities. New York: Random House.

22.

Lang

Chen

Chan

, et al. (2019) Understanding livable dense urban form for shaping the landscape of community facilities in Hong Kong using fine-scale measurements. Cities 84: 34–45.

23.

Lang

Hui

ECM

Chen

, et al. (2020) Understanding livable dense urban form for social activities in transit-oriented development through human-scale measurements. Habitat International 104: 102238.

24.

Lin

J-J

Yang

A-T

(2009) Structural analysis of how urban form impacts travel demand: Evidence from Taipei. Urban Studies 46: 1951–1967.

25.

Long

Huang

(2019) Does block size matter? The impact of urban design on economic vitality for Chinese cities. Environment and Planning B: Urban Analytics and City Science 46: 406–422.

26.

Pas

(1999) Socio-demographics, activity participation and travel behavior. Transportation Research Part A: Policy and Practice 33: 1–18.

27.

Song

Diao

Feng

(2016) Individual transport emissions and the built environment: A structural equation modelling approach. Transportation Research Part A: Policy and Practice 92: 206–219.

28.

Vallance

Perkins

Moore

(2005) The results of making a city more compact: Neighbours’ interpretation of urban infill. Environment and Planning B: Planning and Design 32: 715–733.

29.

Wang

Chai

(2011) Built environment diversities and activity–travel behaviour variations in Beijing. Journal of Transport Geography 19: 1173–1186.

30.

Yang

Shen

, et al. (2012) Transport impacts of clustered development in Beijing: Compact development versus overconcentration. Urban Studies 49: 1315–1331.

31.

Zhang

Zhou

Zhang

(2017) Quantifying and visualizing jobs-housing balance with big data: A case study of shanghai. Cities 66: 10–22.

32.

Zhang

Chen

(2018) Exploring the relationships between urban form metrics and the vegetation biomass loss under urban expansion in China. Environment and Planning B 0: 2399808318816993.

33.

Zhao

Lü

de Roo

(2010) Urban expansion and transportation: The impact of urban form on commuting patterns on the city fringe of Beijing. Environment and Planning A: Economy and Space 42: 2467.

34.

Zhou

Long

(2016) Losers and pareto optimality in optimising commuting patterns. Urban Studies 53: 2511–2529.

35.

Zhou

Sipe

, et al. (2019) Monitoring transit-served areas with smartcard data: A Brisbane case study. Journal of Transport Geography 76: 265–275.

36.

Zhou

Chen

Yeh

, et al. (2019) Workplace segregation of rural migrants in urban China: A case study of Shenzhen using cellphone big data. Environment and Planning B: Urban Analytics and City Science. https://journals-sagepub-com-s.web.bisu.edu.cn/doi/abs/10.1177/2399808320954221.

37.

Zhou

Yeh

AGO

, et al. (2018) A commuting spectrum analysis of the jobs–housing balance and self-containment of employment with mobile phone location big data. Environment and Planning B: Urban Analytics and City Science 45: 434–451.