摘要详情
552 / 2019-02-28 21:49:24
A Spatial Equilibrium Model for Home-work Separation in Megacity: A Case Study in Beijing
spatial equilibrium modeling,commuting cost,occupational-residential separation
全文录用
A Spatial Equilibrium Model for Home-work Separation in Megacity: A Case Study in Beijing
Zijun Xu1,2, Yuhao Kang1,3, Teng Fei1,*
1School of Resource and Environmental Science, Wuhan University, Wuhan, PR China, 430079
2State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, PR China, 430079
3Geospatial Data Science Lab, Department of Geography, University of Wisconsin-Madison, Madison, Wisconsin, United States 53706
*Corresponding author: feiteng@whu.edu.cn
keywords: spatial equilibrium modeling, commuting cost, occupational-residential separation,
Introduction
Occupational-residential separation phenomenon is closely related to urban sprawl (Brueckner 2000) and can bring high commuting cost. Such phenomenon has drawn much attention from researchers majoring in urban planning and urban design.
Previous researchers used conventional data sources such as travel survey and census to examine if commuting distance is correlated to certain attributes, how they correlated, and to what extent (Gutiérrezipuigarnau, Mulalic, and Ommeren 2016, Liu et al. 2008, Meng, Zheng, and Yu 2011). The emergence of geo-big data provides new data sources and insights for detecting commuting behavior patterns and exploring urban spatial structure (Kang, Liu, and Wu 2015, Long and Thill 2015). The convincing results of theses researches indicate that user generated data can provide a fine, personal granularity-based data source for understanding occupational-residential separation phenomenon.
In this article, we aim to build a gravity-repulsion model and to simulate the separation of workplace and household in Beijing. This model has the following new features: (1) A gravity-repulsive model is developed by aggregating attributes including income, property price, commuting cost and spatial opportunity (Liang 2014) into attractive factor and repulsive factors. (2) We make use of advantages provided by mobile user trajectory data. (3) A field model was employed. Each point in the field is assigned with repulsion and gravity. This model is then validated by the real-world data. The research may offer insights on home-work separation to enrich the understanding of urban spatial structure in urban planning.
Data and Methodology
Multi-source data is used in this research including signaling data made up of records from 3 million users, house price data containing more than 10,000 records and recruitment data from more than 200,000 companies.
This research consists of the following three steps (as shown in Figure 1):
Figure 1 workflow
After data collection, commuting trajectories of mobile phone users were extracted according to the following procedure: first, mobile phone users with excessive or insufficient number of records are filtered out. Then, according to the call frequency during work hours and night-time hours, residence and workplace location of valid users are derived. Finally, commuting distance was measured based on Manhattan distance.
Attractive factors and repulsive factors are involved in the equilibrium model. Attractive factors are represented by spatial opportunity, and we use population density to represent that. Repulsive factors are summarized as commuting cost and housing burden. The functional form of the equilibrium model can be represented as follows:
Where , ,, are parameters; represents population density; P indicates property price; C represents commuting cost; S represents salary and D represents commuting distance.
With kernel density algorithm is approximated using signaling data. Kriging interpolation method was used to simulate P with property information data. Logarithmic normal distribution was applied to simulate salary distribution at each location using recruitment data.
Based on this model, it is assumed that rational individual with workplace location will choose optimal residence with maximum F value. Finally, after parameters in the model are optimized, this model can be used to assist policy formulation concerning urban planning
Result
The simulation result of spatial opportunity is shown in Figure 2 (a). It indicates that spatial opportunity is highly concentrated in urban centers and sub-centers. Figure 2 (b) shows the simulation result of spatial property price distribution in Beijing. Property price varies from 151431 CNY/ to 11234 CNY/
(a) (b)
Figure 2 spatial opportunity simulation result (a) and property price simulation result (b)
The simulation result of income distribution is plotted versus the distribution of the income extracted from recruitment data in Figure 3.
Figure 3 salary simulation result
Figure 4 compares the real-world commuting distance distribution and the modeling result. The correlation coefficient of the two distributions is 0.99, which shows the efficacy of the model.
Figure 4 commuting distance simulation result
Conclusion
In this article, an equilibrium model was established to simulate the home-work separation in the city of Beijing by taking both attractive factors (spatial opportunity) and repulsive factors (relative house price and commuting cost) into consideration. The result of the model is calibrated and validated using multi-source data. The results show that our model can greatly explains the home-work separation phenomenal in Beijing extracted from mobile user trajectory data and promotes our understanding of the phenomenal in a megacity for a more scientific policy support.
Reference
Brueckner, Jan K. 2000. "Urban sprawl: diagnosis and remedies." International regional science review 23 (2):160-171.
Gutiérrezipuigarnau, Eva, Ismir Mulalic, and Jos N. Van Ommeren. 2016. "Do rich households live farther away from their workplaces?" Journal of Economic Geography 16 (1).
Kang, Chaogui, Yu Liu, and Lun Wu. 2015. "Delineating intra-urban spatial connectivity patterns by travel-activities: A case study of Beijing, China." Geoinformatics, 2015 23rd International Conference on.
Liang. 2014. "Re-discussion on "Urban Man": People-oriented Urbanization." Urban Planning 9:64-75.
Liu, Wangbao, SanPei Yan, Zhiping Fang, and Xiaoshu Cao. 2008. "Relevant Characteristics and Formation Mechanism of Over-commuting in Guangzhou." Journal of Geography 63 (10):1085-1096.
Long, Ying, and Jean-Claude Thill. 2015. "Combining smart card data and household travel survey to analyze jobs–housing relationships in Beijing." Computers, Environment and Urban Systems 53:19-35.
Meng, Bin, Limin Zheng, and Huili Yu. 2011. "Commuting Time Change and Influencing Factors of Urban Residents in Beijing." Progress in Geographic Science 30 (10):1218-1224.
Zijun Xu1,2, Yuhao Kang1,3, Teng Fei1,*
1School of Resource and Environmental Science, Wuhan University, Wuhan, PR China, 430079
2State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, PR China, 430079
3Geospatial Data Science Lab, Department of Geography, University of Wisconsin-Madison, Madison, Wisconsin, United States 53706
*Corresponding author: feiteng@whu.edu.cn
keywords: spatial equilibrium modeling, commuting cost, occupational-residential separation,
Introduction
Occupational-residential separation phenomenon is closely related to urban sprawl (Brueckner 2000) and can bring high commuting cost. Such phenomenon has drawn much attention from researchers majoring in urban planning and urban design.
Previous researchers used conventional data sources such as travel survey and census to examine if commuting distance is correlated to certain attributes, how they correlated, and to what extent (Gutiérrezipuigarnau, Mulalic, and Ommeren 2016, Liu et al. 2008, Meng, Zheng, and Yu 2011). The emergence of geo-big data provides new data sources and insights for detecting commuting behavior patterns and exploring urban spatial structure (Kang, Liu, and Wu 2015, Long and Thill 2015). The convincing results of theses researches indicate that user generated data can provide a fine, personal granularity-based data source for understanding occupational-residential separation phenomenon.
In this article, we aim to build a gravity-repulsion model and to simulate the separation of workplace and household in Beijing. This model has the following new features: (1) A gravity-repulsive model is developed by aggregating attributes including income, property price, commuting cost and spatial opportunity (Liang 2014) into attractive factor and repulsive factors. (2) We make use of advantages provided by mobile user trajectory data. (3) A field model was employed. Each point in the field is assigned with repulsion and gravity. This model is then validated by the real-world data. The research may offer insights on home-work separation to enrich the understanding of urban spatial structure in urban planning.
Data and Methodology
Multi-source data is used in this research including signaling data made up of records from 3 million users, house price data containing more than 10,000 records and recruitment data from more than 200,000 companies.
This research consists of the following three steps (as shown in Figure 1):
Figure 1 workflow
After data collection, commuting trajectories of mobile phone users were extracted according to the following procedure: first, mobile phone users with excessive or insufficient number of records are filtered out. Then, according to the call frequency during work hours and night-time hours, residence and workplace location of valid users are derived. Finally, commuting distance was measured based on Manhattan distance.
Attractive factors and repulsive factors are involved in the equilibrium model. Attractive factors are represented by spatial opportunity, and we use population density to represent that. Repulsive factors are summarized as commuting cost and housing burden. The functional form of the equilibrium model can be represented as follows:
Where , ,, are parameters; represents population density; P indicates property price; C represents commuting cost; S represents salary and D represents commuting distance.
With kernel density algorithm is approximated using signaling data. Kriging interpolation method was used to simulate P with property information data. Logarithmic normal distribution was applied to simulate salary distribution at each location using recruitment data.
Based on this model, it is assumed that rational individual with workplace location will choose optimal residence with maximum F value. Finally, after parameters in the model are optimized, this model can be used to assist policy formulation concerning urban planning
Result
The simulation result of spatial opportunity is shown in Figure 2 (a). It indicates that spatial opportunity is highly concentrated in urban centers and sub-centers. Figure 2 (b) shows the simulation result of spatial property price distribution in Beijing. Property price varies from 151431 CNY/ to 11234 CNY/
(a) (b)
Figure 2 spatial opportunity simulation result (a) and property price simulation result (b)
The simulation result of income distribution is plotted versus the distribution of the income extracted from recruitment data in Figure 3.
Figure 3 salary simulation result
Figure 4 compares the real-world commuting distance distribution and the modeling result. The correlation coefficient of the two distributions is 0.99, which shows the efficacy of the model.
Figure 4 commuting distance simulation result
Conclusion
In this article, an equilibrium model was established to simulate the home-work separation in the city of Beijing by taking both attractive factors (spatial opportunity) and repulsive factors (relative house price and commuting cost) into consideration. The result of the model is calibrated and validated using multi-source data. The results show that our model can greatly explains the home-work separation phenomenal in Beijing extracted from mobile user trajectory data and promotes our understanding of the phenomenal in a megacity for a more scientific policy support.
Reference
Brueckner, Jan K. 2000. "Urban sprawl: diagnosis and remedies." International regional science review 23 (2):160-171.
Gutiérrezipuigarnau, Eva, Ismir Mulalic, and Jos N. Van Ommeren. 2016. "Do rich households live farther away from their workplaces?" Journal of Economic Geography 16 (1).
Kang, Chaogui, Yu Liu, and Lun Wu. 2015. "Delineating intra-urban spatial connectivity patterns by travel-activities: A case study of Beijing, China." Geoinformatics, 2015 23rd International Conference on.
Liang. 2014. "Re-discussion on "Urban Man": People-oriented Urbanization." Urban Planning 9:64-75.
Liu, Wangbao, SanPei Yan, Zhiping Fang, and Xiaoshu Cao. 2008. "Relevant Characteristics and Formation Mechanism of Over-commuting in Guangzhou." Journal of Geography 63 (10):1085-1096.
Long, Ying, and Jean-Claude Thill. 2015. "Combining smart card data and household travel survey to analyze jobs–housing relationships in Beijing." Computers, Environment and Urban Systems 53:19-35.
Meng, Bin, Limin Zheng, and Huili Yu. 2011. "Commuting Time Change and Influencing Factors of Urban Residents in Beijing." Progress in Geographic Science 30 (10):1218-1224.
重要日期
-
会议日期
07月08日
2019
至07月12日
2019
-
06月28日 2019
初稿截稿日期
-
07月12日 2019
注册截止日期
联系方式
历届会议
-
2021年06月08日 芬兰 Espoo
第17届计算机在城市规划和城市管理中应用国际会议
