Dataset Development of Urban Built Environment Assessment Unit (2022)

Zhang, S. J.  Li, M.^*  Dan, B. Y.  Han, J. B.  Hao, L. Q.

China Academy of Urban Planning & Design, Beijing 100044, China

Abstract: Urban built-up area and core built-up area are the basic units for urban built-up environmental assessment, and their size directly affects the level of built-up environmental indicators. This paper makes full use of the advantages of multi-source big data, formulates a unified delineation method and technical process, and produces a dataset of built-up areas and core built-up areas of 19 cities, including Beijing and Shanghai, etc. (2022). For built-up area, the proportion of impervious surface, road network, POI density and population density are comprehensively considered, and high-resolution imagery are used to form a delineation index system and method for built-up area, and the scope of urban built-up area is scientifically and quickly delineated at the grid scale of 500 m??500 m. For the core built-up area, the scope of the core built-up areas of key cities is delineated through 4 steps: identifying urban centers, identifying high-density streets and towns, verifying the main functional areas and facilities of the city, and deducting the open space of large non-construction land. By unifying the data sources and delineation methods, 2 basic spatial ranges with horizontal comparability in the research of urban built environment assessment are formed. The dataset is archived in .shp format and consists of 16 data files with a data size of 2.23 MB (compressed into 1 file, 1.37 MB).

Keywords: built-up area; core built-up area; big data; urban built environment

DOI: https://doi.org/10.3974/geodp.2025.04.05

Dataset Availability Statement:

The dataset supporting this paper was published and is accessible through the Digital Journal of Global Change Data Repository at: https://doi.org/10.3974/geodb.2025.03.09.V1.

1 Introduction

As a space that centrally carries urban basic public service facilities, the characterization of urban built environment has always attracted the attention of scholars. By calculating the population density, construction intensity, commercial density, transportation accessibility, land use mixing and other indicators within this range, the characteristics of the urban built environment in terms of social and economic development and spatial form can be evaluated, which can provide an objective and accurate understanding of the current situation for urban physical examination and urban renewal, and then promote the city to achieve high-quality development goals.

The urban built-up area is where urban development and construction are concentrated, the area where the main urban functions are arranged, and is also the key area for urban construction and management. It represents the macro level of understanding the overall characteristics of the built environment in the city. The population density and the level of public facility services within the built-up area are significantly higher than those in surrounding rural areas, and the natural environment and land use patterns also differ greatly from rural areas. As a meso-analysis unit, the urban core built-up area is located in the central urban area, with high population density and construction intensity, and accommodating core urban functions and facilities. It aids in the in-depth analysis of built environment characteristics, such as population density and construction intensity. Based on multi-source big data, this paper has produced the boundaries of the built-up areas and core built-up areas for 19 cities in our country.

The traditional method for delineating built-up area is inherently subjective. Urban planners or city construction managers determine the scope of urban built-up area in a balanced manner, considering the actual construction of the city, the service areas of public and utility facilities, and the degree of connection between social and economic activities and the urban center. Population density, a key indicator reflecting the intensity of economic activities, was once a significant criterion for distinguishing towns. However, the statistical units that can publicly access population data are typically large, and the demographic statistics cycle is lengthy, making it challenging to promptly and objectively reflect the dynamic changes in urban areas.

The emergence of remote sensing data and new big data provides new data support and technical ideas for the study of boundary delineation in built-up area. At present, there are 2 main types of built-up area boundaries based on new big data extraction: single-element method and multi-element comprehensive method. The single-element method extracts the built-up area boundary by directly setting a threshold for a determinant or constructing a combined index and then setting the threshold. Night light remote sensing data offer an intuitive and comprehensive reflection of human social and economic activities on the ground, and its advantages of wide format, high temporal resolution, and free access provide a good data basis for the extraction of boundaries in urban built-up area^[1,2]. POI (Point of Interest ) data can spatially reflect the distribution characteristics of urban structure and road skeleton, and can reflect the urban agglomeration effect and scale effect. The difference in the spatial distribution of density reflects different regional development levels and is highly correlated with the distribution of urban built-up area, and the boundaries of urban built-up area can be extracted by setting POI density thresholds^[3]. In order to overcome the problem that a single element can only reflect a certain aspect of the built-up area and cause extraction bias, some scholars have proposed to combine night lighting data with big data such as POI and road network to improve the extraction accuracy of built-up area^[4,5]. However, due to the inevitable ??oversaturation?? and ??spillover?? phenomenon of lighting data at night, and its low spatial resolution, the extraction accuracy is greatly limited in the extraction process of built-up areas. Based on the definition of built-up area, this paper proposes a method to scientifically and quickly extract the boundaries of built-up area at the grid scale by synthesizing multi-dimensional discriminant factors such as statistical data, remote sensing interpretation data, electronic navigation road network data, POI point of interest data, and Baidu LBS (Location Based Services) data.

The concept of core built-up area is close to the concepts of inner city and urban central area, and the research on their delineation methods is even more multi-source. From the perspective of the smallest units delineated, there are blocks, roads, and administrative boundaries, which are mainly used for the planning and control of building forms according to the division of blocks^[6,7], and the division of roads is mainly used for the division of central areas^[8]. Chang, et al. used administrative boundaries as the basis for the circle to compare the scale and density of Beijing and other international cities in each circle^[9]. Wang, et al. delineated the basic units of density distribution based on the boundaries of street and town administrative units^[10]. In terms of delineation factors, Shi analyzed various definition criteria, including spatial texture, road network density, population density, etc., and proposed specific measurement methods such as the ??public service facility index method??^[11,12]. Based on the boundaries of streets and towns, this paper comprehensively uses the determinant factors such as local activity population density, construction intensity, main functional areas such as commercial centers and office centers, and the spatial distribution of large public facilities at the city level to delineate the boundaries of the core built-up areas of 19 cities.

2 Metadata of the Dataset

The metadata for the Dataset of assessment units for 19 urban built environment of China based on multi-source data (2022)^[13], including the title, authors, geographical region, data format, data size, data files, etc., is summarized in Table 1.

Table 1  Summary of metadata of Dataset of assessment units for 19 urban built environment of China based on multi-source data (2022)

3 Methods

Based on the definition of built-up area and core built-up area, this study makes full use of the advantages of multi-source big data, comprehensively considers urban population, buildings, road networks, POI, land use and other factors, formulates a unified delineation method and technical process, and produced the datasets of built-up areas and core built-up areas in 19 cities including Beijing and Shanghai. The specific data sources are shown in Table 2.

Table 2  Data source description

3.1 Algorithm

3.1.1 Delimitation of Built-up Area

According to the Basic terminology standards for urban planning^[17], urban built-up area refer to areas within the urban administrative area that have been completely developed and constructed, with basic municipal and public facilities in place. In view of the 2 important characteristics of ??completely developed and constructed?? and ??basic municipal and public facilities are in place??, multi-source big data is utilized for expression. Among them: (1) Identification of ??fully developed and constructed?? areas: using impervious surface data obtained by interpreting 10-m resolution remote sensing images to extract the impervious surfaces formed by human construction; (2) Judgment of ??basic municipal facilities??: selecting road network data from navigation electronic maps (including roads above the township and village level), and reflecting whether some municipal facilities are available through the road network density factor. At the same time, the spatial distribution of roads can also determine whether the map is connected to the central urban area. (3) Judgment of ??basic availability of public facilities??: selecting POI data from navigation electronic maps, and evaluating the level of public service facility construction in different areas through the analysis of the spatial distribution of major public facilities such as education, medical care, culture, and sports. Finally, local population density data from Baidu LBS and high-resolution imagery are used as verification factors to correct the preliminarily delineated potential construction areas.

3.1.2 Delineation of Core Built-up Area

The core built-up area refers to the central built-up area that undertakes core functions and exhibits a high-density development pattern. Its connotation is mainly reflected in the following 3 points: in terms of location, it is located in the geographical center or the core area of the city that is generally recognized; In terms of morphology, it shows high construction intensity and population density. In terms of function, it brings together core functions such as commerce, finance, administration, history and culture, and is a dense area of high-level public service facilities. When delineating the core built-up area, fully consider the data expression of the key points in its definition. Firstly, the geographical center and cognitive center within the built-up area are identified to form the central area. At the same time, the density of each street and township unit is measured to identify the higher density range; After that, verify whether the core functional areas and main facilities have been covered by the high-density range; Finally, large ecological spaces such as major water bodies and regional parks are identified, which are not included in the calculation scope (Figure 1).

Figure 1  Principle of delineation of core built-up area

3.2 Technical Route

3.2.1 Delineate Technical Routes in Built-up Areas

The technical route of built-up area delineation can be summarized as follows: firstly, the area with surface coverage type impervious surface is extracted and defined as ??developed area??; set a certain threshold in the developed area, and judge the area where the road network density and POI density are higher than the threshold as ??quasi-built-up area??; Based on the population density of Baidu grid and high-resolution imagery, the built-up area is checked and corrected to obtain the boundary of the urban ??built-up area?? (Figure 2).

Figure 2  Flowchart of the built-up area delineation

Taking Beijing as an example, the extraction process of built-up area is as follows:

Step 1: Identify ??developed area??.

Within the jurisdiction of the city, based on the land cover data interpreted by high-resolution imagery, the area of the type impervious surface is regarded as the ??developed area??. According to the Global Land Cover Product (FROM-GLC10) produced by the Department of Earth System Sciences of Tsinghua University using AI technology, Beijing has a total developed area of 2,560 km² (Figure 3).

Figure 3  Maps of land cover type (left) and ??developed area?? (right) in Beijing

Step 2: Identify ??quasi-built-up area??.

(1) Divide the municipal area into uniform 500 m??500 m grids. The size of the grid significantly influences the area and boundary shape of the built-up area, and the large grid can easily lead to the area error of band-shaped and finger-shaped built-up area, as well as the area error of small and medium-sized urban built-up area, or the omission of urban land patches such as development zones and mining areas that are not connected to the urban center group. If the grid is too small, while the resolution can be improved, the fragmentation of the patch will bring unnecessary trouble to the subsequent processing and analysis work, and it is difficult to evaluate the construction of infrastructure and public service facilities in a patch that is too small. In this paper, the smallest unit of the built-up area is set to 500 m grid.

(2) Determine the thresholds of road network density and POI density at urban grid scale, and the size of the threshold directly determines the spatial location and area of the built-up area boundary. Considering that the geographical location, morphological characteristics, urbanization level, socio-economic development level and other characteristics of the 19 cities are quite different, each city should set different thresholds. In this paper, the most widely used mutation detection method in built-up area is used to extract night light remote sensing data to determine the threshold values of road network density and POI density. Based on the morphological characteristics, the urban area will maintain the integrity of the overall geometry, and the urban built-up area boundary will shrink along the edge in the process of gradually increasing the threshold, but when the segmentation threshold reaches a certain point, the built-up area boundary will no longer shrink along the edge, but will break from the inside, resulting in a new smaller polygon, which will lead to a sudden increase in the perimeter of the built-up area boundary, and this point is defined as the threshold point^[1]. This paper argues that there is also a mutation point (i.e., the boundary of urban built-up area) in the transition from the central urban area to the peripheral suburbs, and this mutation point is the threshold of this factor. It is calculated that the road network density threshold in Beijing??s built-up area delineation method is 4.1 km/km², and the POI density threshold is 45 units/km² (Figure 4).

(3) The grid with road network density or POI density higher than its corresponding threshold in the ??developed area?? of the municipal area is determined as ??quasi-built-up area??, and polygon merging, denoising or filling of the ??quasi-built-up area?? is carried out. Firstly, the grid in which any index of road network density and POI density is higher than its corresponding threshold is extracted, and converted into a vector polygon. Then the vector

Figure 4  Maps of areas with road network density above the threshold (left) and areas with POI density above the threshold (right) in Beijing

polygons are merged to form a ??quasi-built-up area?? patch. Then, the fine polygons with an area of less than 2 km² (i.e., less than 8 grids) in the ??quasi-built-up area?? patch are deleted, and the cavities with an area of less than 10 km² (i.e., less than 40 grids) are eliminated, and a relatively regular ??quasi-built-up area?? is obtained (Figure 5).

Step 3: Modify the ??quasi-built-up area?? to form the boundary of the urban built-up area (Figure 6).

(1) Calculate the population density at the urban grid scale. Although the definition of urban built-up area in the ??Basic terminology standard for urban planning?? does not include a description of the urban population, whether it is land developed and constructed in patches or complete municipal public facilities, it serves people. Population is one of the important indicators that reflect the construction status of a city. Based on Baidu LBS data this paper calculates the urban 500 m??500 m grid-scale population density index, and uses it as a verification factor to correct the boundary of urban built-up area.

(2) By integrating high-resolution imagery, Baidu grid population density, urban road network, and ecological corridors, and the boundaries of the quasi-built-up area are refined by removing large water bodies and undeveloped land, such as peripheral mountains to achieve a more precise built-up area delineation.

3.2.2 Delineate the Technical Route of the Core Built-up Area Data Development

In order to ensure the comparability of the built-up environmental indicators between different cities, and draw on similar concepts at home and abroad (such as 92.5 km² in the core area of Beijing^[18], 105 km² in the core area of the Paris metropolitan area^[19]), the area is controlled at about 90?C160 km². The specific delineation process includes the following four steps (Figure 7):

Step 1: Delineation of the urban central area. Urban centers include 2 types, one is the objective geometric center, that is, the geographical center; One is the subjective public perception of the city. For geographic centers, they can be identified by geometric centers within the built-up area of the city (usually the center of a ring road, the intersection of major roads). The cognitive center identifies representative public facilities and public areas that can represent the image of the city center, usually commercial districts, cultural facilities, parks and squares, etc. as the city center.

Step 2: Identification of high-density towns/sub-districts based on population and building data. Based on Baidu grid population density data and building base data, the construction intensity and population density index of each sub-district or township unit within the urban built-up area are calculated. Through the spatial distribution of indicators, the higher density range is identified as the basis for the spatial range of the core built-up area.

Step 3: Verification of the main urban functional zones and facilities. Verify that the city??s main functional areas and major city-level facilities are covered by the high-density range identified in step 2. The layout of commercial centers, business centers, administrative centers, historical districts and large public facilities in the city is generally examined, but urban sub-centers that are far away from the urban center area (such as Beijing Tongzhou sub-center, Tianjin Binhaixinqu, etc.) are not included in the delineation of the core built area of this paper.

Step 4: Exclusion of major open spaces to form the ??core built-up area??. Identify ecological spaces such as large water bodies and country parks with low construction intensity and less accessible and used by people, and deduct them from the scope of the core built area. Among them, urban parks are spaces used by urban residents on a daily basis, which are still included in the calculation scope of urban core built-up area.

Figure 7  Flowchart of the core built-up area delineation

4 Data Results and Validation

4.1 Dataset Composition

Based on multi-source big data, this paper uses unified delineation methods and delineation processes to produce the built-up areas boundaries and core built-up areas boundaries of 19 cities in China. The dataset is archived in .shp format, and the city name, province name and zoning code information of the built-up area and the core built-up area are declared in the attribute table (Table 3). The dataset consists of 2 data files with a total data volume of 2.23 MB (compressed into 1 file, 1.37 MB).

Table 3  Dataset attribute information

4.2 Data Products

The results of the 19 urban built-up areas and core built-up areas delineated in this paper are shown in Figure 8.

This paper analyzes the distribution of built-up areas and core built-up areas in each district and county unit in 19 cities (Figure 9). The analysis results show that there are significant differences in the urbanization level of districts and counties within different cities, which are mainly reflected in the 2 dimensions of built-up area coverage (proportion of built-up area) and urban core function agglomeration degree (proportion of core built-up area).

(1) A highly urbanized core area

Some district and county units have been highly mature in terms of construction form and core functional layout, and their built-up area and core built-up area occupy an absolutely dominant position. Typical representatives include: Beijing Dongcheng District and Xicheng District (100% of the built-up area, nearly 100% of the core built-up area), Tianjin Heping District (100%, 100%), Xi??an Xincheng District (100%, 99.6%), Beilin District (100%, 100%), Shanghai Hongkou District (97.7%, 97.3%), Qingdao Shinan District (96.4%, 95.9%), Shibei District (97.9%, 97.3%). The built-up areas and core built-up areas of these districts and counties generally account for more than 95%, which are key areas for accommodating core urban functions.

(2) Areas with mature built-up area but insufficient core function agglomeration

Some districts and counties show high coverage of built-up area but relatively weak core built-up functions. The proportion of built-up area in Wuhou District of Chengdu is the highest in the city (96.7%), but the proportion of core built-up area (56.9%) is significantly lower. This phenomenon shows that these areas are highly urbanized in terms of spatial construction, but there is still room for development in carrying and agglomerating the core functions of the city.

(3) Areas with a low proportion of built-up area

The proportion of built-up areas in districts and counties in some cities is low (less than 50%), while the proportion of core built-up area is even lower, reflecting the characteristics of the distribution of urban built-up area and core built-up area affected by natural geographical conditions, urban development planning and other factors. Taking Harbin as an example, Nangang District has the highest proportion of built-up area and core built-up area in the city, but only 42.2% and 22.2% (the area includes a large area of farmland); Guandu District, which has the highest proportion of built-up areas in Kunming, is only 25.5% (most of the area is mountainous and watery).

The analysis shows that there are significant hierarchical differences between different district and county units in carrying urban physical space (built-up area) and core functional space (core built-up area). This difference profoundly reflects the uneven development within the city, the gradient of functional distribution and the constraints of the natural environment, and provides an important basis for understanding the urban spatial structure and functional layout.

Figure 8  Maps of the spatial scope of 19 urban built-up areas and core built-up areas of China (2022)

Figure 9  Proportion of built-up areas and core built-up areas of 19 urban districts and counties of China

4.3 Data Validation

In order to verify the accuracy of the built-up area delineation results, this study compared the built-up area data of 19 cities with the built-up area data used in the 2022 third-party urban physical examination of the Ministry of Housing and Urban-Rural Development (Figure 10). The results show that the two are highly consistent in the spatial center position and overall morphology of the built-up area. The area size is basically the same, but different cities show certain differences. Specifically, the difference between the built-up area and the reference data in Tianjin, Nanjing, Changsha, Kunming, Harbin, Shenyang, Jinan and Hangzhou is less than 1%, showing the best coincidence, while the area difference in Wuhan, Qingdao, Chongqing and Zhengzhou is more than 10%, and the difference is relatively significant; The area difference in the remaining cities is between 1% and 10%, and the difference level is moderate. The main reason for the difference in area is that there are differences between the two algorithms in the determination of green space and water bodies with large areas within and at the edges of the city.

Figure 10  Verification of built-up area results

For example, Shanghai and Beijing have built-up area of more than 1,500 km², while Kunming, Harbin, Changsha, etc. are less than 500 km². This significant scale difference leads to unreasonable direct comparison of urban development evaluation indicators such as facility coverage and population density in each built-up area. However, the area difference in urban core areas, especially those with higher density and core functions and facilities, is usually smaller (Figure 11). Therefore, comparing the level of urban development at this scale can effectively make up for the limitations based on the overall comparison of built-up area. Due to the lack of officially published standard datasets for core built-up area, standard deviation is used as an auxiliary criterion to evaluate the accuracy of the delineation results. The delineation results show that the standard deviation of the area of the core built-up areas

Figure 11  Verification of the results of the core built-up areas

of the 19 cases is 22.06, which is much lower than the standard deviation of the built-up areas (394.63), indicating that the delineation results meet the expected spatial characteristics (i.e., the difference in the area of the core area is small), which verifies the effectiveness of the method.

5 Discussion and Conclusion

Urban built-up area and core built-up area are commonly used spatial boundaries for evaluating urban built-up environmental indicators, and they are also the basic spatial boundaries for urban management, planning and construction, urban renewal and urban physical examination. The traditional delimitation method is mainly manual and too subjective. With the emergence of new big data and the advancement of spatial data mining technology, the built-up area delimitation method has been gradually improved and perfected, and the accuracy and precision of extraction are also constantly improving. The multi-source data fusion method proposed in this paper comprehensively uses high-resolut­ion imagery, road networks, points of interest, Baidu LBS population density and other data to efficiently identify the boundaries of urban built-up area and core built-up area at the grid scale. With the advantage of big data, this method provides an accurate spatial foundation for the study of built environmental assessment. The delineation method of urban built-up area and core area based on multi-source data has the same data sources, the same delineation ideas and methods, and unified technical processes, which can ensure that the extracted urban boundaries can be compared horizontally, vertically, and updated annually.

However, there are still some limitations in the method of delineating the boundary of built-up area. The core problem is that it does not fully respond to the key dimension of ??municipal public facilities are basically available?? in the definition of built-up area. Although this paper fully considers the road network, an important municipal infrastructure, it fails to effectively consider equally important municipal pipelines (such as water supply and drainage, gas, electricity, communication pipeline networks, etc.). It is mainly limited by 2 factors: first, the degree of electronic and standardized municipal pipeline data in different cities is significantly different; Second, such data are often sensitive and not publicly accessible. Consequently, the current delineation results do not fully satisfy the strict definition of a built-up area. Besides, the Global Land Cover Product (FROM-GLC10) which was used in this study is needed to be validated.

In the future, the range of data sources will be further expanded, and artificial intelligence technology will be proactively introduced to enhance the automation level of data processing, feature extraction, and boundary delineation, thereby building a more accurate, efficient, and standardized urban built environment assessment data support system.

Author Contributions

Zhang, S. J. designed the overall framework for the dataset development and wrote the main part of the paper; Li, M. extracted the boundaries of urban built-up area and participated in the writing of the paper; Dan, B. Y. processed the basic data in the text and participated in the writing of the paper; Han, J. B. delineated the boundaries of the core built-up area. Hao, L. Q. revised and revised the dataset and paper.

Conflicts of Interest

The authors declare no conflicts of interest.

References

[1]        Shu, S., Yu, B. L., Wu, J. P. Evaluation and application of urban built-up area extraction method based on night light data [J]. Remote Sensing Technology and Application, 2011, 26(2): 169?C176.

[2]        Yang, Y., Sun, W. B., Han, Y. H. Extraction method of urban built-up areas based on DMSP/OLS lighting brightness combination [J]. Remote Sensing for Land & Resources, 2020, 23(3): 39?C48.

[3]        Xu, Z. N., Gao, X. L. Boundary identification method of urban built-up areas based on electronic map points of interest [J]. Acta Geographica Sinica, 2016, 71(6): 928?C939.

[4]        Wang, X. Y., Li, C. M., Chen, X. D., et al. Extraction method of built-up area based on multi-source data fusion of night lights [J]. Remote Sensing Information, 2021, 36(5): 114?C123.

[5]        Zheng, H. D., Gui, Z. P., Li, F., et al. Built-up area extraction method combining night lighting data and point of interest data [J]. Geography and Geographic Information Science, 2019, 35(2): 25?C32.

[6]        Huang, N., Xu, Z. H., Xu, S. S. Empirical research and dynamic optimization of urban construction land intensity control in Wuhan City [J]. Journal of Urban Planning, 2012(3): 96?C101.

[7]        Jin, T. H., Yang, J. Y., Wang, D. From urban density zoning to spatial morphology zoning: evolution and empirical evidence [J]. Journal of Urban Planning, 2018(4): 34?C40.

[8]        Tang, J. X., Long, Y. Measurement of street spatial quality in the central area of megacities: a case study of Beijing??s second and third ring roads and Shanghai inner ring road [J]. Planner, 2017, 33(2): 68?C73.

[9]        Chang, Q., Xu, Q. Z., Yang, C., et al. Reflections and exploration on the reduction and control of construction land in Beijing??s New General Plan [J]. Urban Planning, 2017, 41(11): 33?C40.

[10]     Wang, H. F., Shi, S., Rao, X. J. Spatial distribution logic of urban density: a case study of Shenzhen [J]. Urban Issues, 2015(8): 22?C32.

[11]     Shi, B. X. Quantitative study on the phenomenon of polar core structure in Asian urban central [D]. Nanjing: Southeast University, 2015.

[12]     Imhoff, M. L., Lawrence, W. T., Stutzer, D. C., et al. A technique for using composite DMSP/OLS ??City Lights?? satellite data to accurately map urban areas [J]. Remote Sensing of Environment, 1997, 61(3): 361?C370.

[13]     Zhang, S. J., Li, M., Dan, B. Y., et al. Dataset of assessment units for 19 urban built environment of China based on multi-source data (2022) [J/DB/OL]. Digital Journal of Global Change Data Repository, 2025. https://doi.org/ 10.3974/geodb.2025.03.09.V1.

[14]     GCdataPR Editorial Office. GCdataPR data sharing policy [OL]. https://doi.org/10.3974/dp.policy.2014.05 (Updated 2017).

[15]     Gong, P., Liu, H., Zhang, M., et al. Stable classification with limited sample: transferring a 30-m resolution sample set collected in 2015 to mapping 10-m resolution global land cover in 2017 [J]. Science Bulletin, 2019, 64: 370?C373.

[16]     Che, Y., Li, X., Liu, X., et al. 3D-GloBFP: the first global three-dimensional building footprint dataset [J]. Earth System Science Data Discussions, 2024, 16(11): 18.

[17]     Ministry of Construction of P. R. China. Standard for basic terminology of urban planning (GB/T 50280??98) [S]. Beijing: China Architecture Industry Press, 1998.

[18]     Beijing Municipal Planning and Natural Resources Commission. Beijing Urban Master Plan (2016?C2035) [Z]. Beijing: Beijing Municipal People??s Government, 2017.

[19]     Ni, J. C. The link of urban and suburban integration in the Greater Paris Metropolitan Area: Paris metropolitan (suburban) railway [J]. Urban Rail Transit, 2021(2): 26?C31.