Coupling between Population and Construction Land Changes in the Beijing–Tianjin–Hebei (BTH) Region: Residential and Employment Perspectives

Abstract

To gain a deeper understanding of the human–land coupling relationship, this study analyzes the coupling relationships with the spatial distribution of construction land from two perspectives: the residential population and the employment population, exploring the similarities and differences in coupling relationships among different subsystems. The Beijing–Tianjin–Hebei region of China is selected as the study area, covering the period from 2000 to 2020. An analytical framework is proposed, encompassing three approaches: coupling analysis based on county-level spatial units; mean center position analysis based on construction land grids; and regression fitting and residual analysis based on homogeneous grid units. The analysis results indicate: (1) the coupling between the employment population and construction land shows a significant advantage; (2) the coupling between the residential population and construction land has improved faster in recent years; (3) factors such as location, development level, and strategic opportunities have an important influence on the spatial and temporal changes in the coupling relationship. The study further discusses the trade-off relationship between different subsystems, key measures to enhance coupling degree, and the application pathways of this analytical framework at various stages of planning. Considering the limitations of industry sector differences, spatial unit precision, and construction land development intensity, this paper also outlines future research directions.

1. Introduction

Human activities have a profound impact on the land use status of the Earth’s surface [1]. It is difficult for human activities, such as living and production, to avoid causing changes in land cover on the Earth’s surface [2]. Since the Industrial Revolution, the demand for land due to industrial production has significantly increased, leading to a rapid expansion of construction land. Simultaneously, industrial development has created numerous job opportunities, attracting a large population to urban areas for employment, thus forming the employment population. Consequently, more people have chosen to reside in cities long-term, becoming the residential population. Construction land typically includes residential, public service, industrial, commercial, storage, and transportation areas. The residential population refers to those who live regularly in a certain area (in China, this is defined as residing for more than six months). The employment population generally refers to individuals who have reached the legal working age, are engaged in social labor, and earn wages or business income. Changes in the population are reflected in the increase or decrease in total numbers, changes in structure, and changes in spatial distribution (such as places of residence or employment). The prosperity and development of a city and the healthy growth of construction land require the support of an appropriate scale of residential and employment populations.

In the early stages of urban expansion, economic benefits are often prioritized, with land resources primarily considered as factors contributing to economic growth. As civilization has progressed, human understanding of development has become more scientific, and the harmony and sustainable development of the human–land relationship have gradually been given more attention [3]. In China, the concept of ecological civilization is gradually receiving more attention in territorial spatial planning and governance. As the proportion of the urban population exceeded 50% of the total population around 2010, China’s urbanization has transitioned from rapid growth to smart growth [4]. The slowdown in urbanization implies that the incremental demand for construction land due to population growth is gradually decreasing. Consequently, many cities are beginning to control the disorderly expansion of construction land and impose strict constraints on the development boundaries of construction land in overall urban planning. Some cities, such as Shanghai, have even set the goal of zero growth in construction land, attempting to meet the city’s development needs for spatial resources by tapping into the potential of existing construction land [5]. In the current process of promoting high-quality development of construction land, the coordination between human living, production, and construction land has gradually become a focal point of attention. Construction land should not expand in a disorderly manner, leading to the idle or inefficient use of large areas of land, which would result in a waste of resources. At the same time, certain areas should avoid the high concentration of populations on construction land. Therefore, it is particularly important to discuss the coupling relationship between population and land use. This paper aims to explore the coupling relationship between construction land and population distribution from the perspectives of the residential population and the employment population. The literature review will primarily focus on three aspects: changes in land use, changes in population distribution, and the study of the human–land coupling relationship.

The development of remote sensing technologies has provided strong support for the study of land use/land cover change (LUCC) [6,7]. The basic approach involves capturing images of the Earth’s surface over time using aerial and satellite sensors, collecting data from different time points for band recognition, pixel classification, and data correction, thereby creating relatively reliable datasets to support related research [1,8,9]. The accuracy of the identification results directly affects the reliability of the analysis, making the optimization of modeling methods very important [2,10,11]. For instance, García-Álvarez and other researchers explored several paths to enhance LUCC reliability from the perspectives of map data input, resolution, and model construction [12,13,14]. Empirical analyses based on reliable LUCC data are a global research hotspot [15,16], encompassing fields such as ecological risk impacts [17], meteorological conditions [18], water resource system impacts [19], and land productivity changes [20]. In addition, LUCC is also a key concern in the field of territorial spatial planning, and some studies attempt to support planning decisions based on quantitative analysis results. For example, Wei et al. (2021) used LUCC data to identify urban, agricultural, and ecological spaces in the Yangtze River Economic Belt, analyzed spatial pattern evolution from 1980 to 2018, and proposed planning recommendations [21]. Therefore, this study can use LUCC data to identify the change characteristics of construction land.

The characteristics of changes in the spatial distribution of populations are also a focus of this study. Census data and other standardized statistics typically include social attributes of populations, such as age, education level, and occupation. Combined with administrative boundary GIS data, it becomes relatively easy to analyze their spatiotemporal variation characteristics. These studies typically address different scales, such as employment distribution across 145 regions in the European Union [22], spatial distribution characteristics of residential and employment populations in China [23,24], and employment spatial structures in metropolitan areas like Toronto and Tehran [25,26]. Some studies combine other data sources, such as remote sensing, to estimate the spatial distribution of the population [27]. Many studies have found that the spatial distribution of populations in different regions exhibits distinct characteristics, including a concentric structure [28], spatial agglomeration [29], and suburbanization [30]. Additionally, some studies have indicated that there may be significant differences in the spatial distribution characteristics of residential and employment populations within the same region [31,32]. Therefore, it is necessary for this study to examine the human–land coupling relationship from the perspectives of both residential and employment populations.

Research on human–land coupling relationships primarily focuses on the interaction between human spatial activities and land use changes [33,34]. These studies mainly use land use data and population distribution characteristics, among other data, to conduct analyses by establishing corresponding model frameworks and indicator systems [35,36,37]. Such studies are also conducted across various spatial scales. At the national level, Li et al. (2020) analyzed the coordination between population and urban construction land in Chinese provinces and provincial capitals, finding that the expansion rate of urban construction land is generally faster than the population growth rate, resulting in an overall inharmonious human–land relationship [38]. Wu et al. (2018) focused on 636 cities in China and analyzed the macro pattern of urbanization development and the changing trend of the coupling between population and land urbanization in China [39]. Liu et al. (2018) studied the spatiotemporal coupling relationship between rural residential land and rural population in China [40]. At the regional level, the research includes urban agglomerations [41], provinces [42], cities [43], counties [44], and other types, with a focus on the coordination of human–land coupling relationships. By analyzing the influencing mechanisms of human–land systems and other factors, many scholars have identified several driving factors contributing to the current situation, such as unsustainable land use practices [45], industrial structure characteristics [46], geographic location conditions [39], and government decision-making behavior [41]. Some scholars have proposed pathways to promote a more harmonious development of the human–land relationship, tailored to the needs of planning and decision-making [47].

Although extensive research has been conducted on the human–land coupling relationship, there is still a lack of in-depth studies that separately examine the spatial coordination from the perspectives of residential and employment populations. Existing research has already demonstrated that there can be significant differences between the spatial distribution characteristics of residential and employment populations. Therefore, it can be hypothesized that the coupling relationship between residential populations and construction land, as well as employment populations and construction land, may vary significantly across different regions and development periods. This study aims to explore this difference by addressing the question: “Which is More Closely Coupled with Changes in Construction Land: Residence or Employment?”

Therefore, this study aims to further investigate the spatial distribution relationship between population and construction land by separately analyzing the human–land coupling relationship from the perspectives of residential and employment populations, using the Beijing–Tianjin–Hebei (BTH) region in China as a case study. The BTH region, chosen for this study, is a representative example among China’s urban agglomerations. Its empirical analysis conclusions and policy recommendations have broader reference value.

The specific objectives of this research include (1) exploring the mechanisms and patterns of the impact of spatial distribution differences between residential and employment populations on the human–land coupling relationship; (2) analyzing the differences in the two types of human–land coupling relationships across different cities and regions from a spatial perspective; (3) examining the characteristics and trends of changes in these two coupling relationships over time; and (4) proposing strategies to enhance the human–land coupling relationship based on the findings from the empirical analysis. This study contributes to the theoretical exploration and empirical analysis of human–land coupling relationships to some extent. By examining the coupling relationship with construction land from the perspectives of both residential and employment populations, the research provides significant theoretical insights into the characteristics and development patterns of human–land coupling relationships.

2. Materials and Methods

2.1. Study Area and Data

The spatial scope of this study is the Beijing–Tianjin–Hebei (BTH) region in China, which includes two municipalities (Beijing and Tianjin) and one province (Hebei). Hebei Province is composed of 11 prefecture-level cities. The BTH region covers approximately 216,000 square kilometers, with a total population of about 110 million (as of 2020), accounting for 7.65% of the national population. The Beijing–Tianjin–Hebei (BTH) region is one of the five key urban agglomerations in China and is home to the nation’s capital, containing two of China’s four municipalities. The BTH region is also a major hub for employment in the tertiary sector in China. According to the seventh national census, this region accounts for 14.07% of the employment in scientific research and technical services, 12.97% in information transmission, software, and information technology services, 11.14% in cultural, sports, and entertainment industries, and 10.65% in the financial sector in mainland China. In 2020, the region’s GDP totaled 8.6393 trillion yuan, representing 8.50% of the national GDP, with the tertiary sector contributing 10.48% to the national total. Therefore, the BTH region holds a significant position in China’s urban and economic development. Given its representative role in China’s urban agglomeration development, selecting this region for a case study ensures that the findings will have considerable reference value and broader applicability.

Land use change data are the core data required for this study. This study utilizes the China National Land Use/Cover Change (CNLUCC) dataset, sourced from the Resource and Environment Science Data Registration and Publication System, which has been widely used in recent years [21,48,49]. The CNLUCC dataset is a national-scale multi-temporal database constructed through visual interpretation of Landsat satellite images from the United States, serving as the main source of information [50]. The CNLUCC dataset adopts a two-level classification system, with the first level including categories such as farmland, forest land, grassland, water areas, construction land, and unused land, where construction land includes built-up areas of cities and towns, rural settlements independent of urban areas, as well as land used for industrial purposes, mining, transportation roads, airports, and other similar uses [50]. The data on the residential population and the employment population needed for this study are sourced from county-level information from the last three national censuses of China (2000, 2010, and 2020) [51,52,53]. It is important to note that the residential population data represent the complete dataset, while the employment population comes from a 10% sample of the long-form data. To ensure comparability, the employment population data will be multiplied by 10 in subsequent analyses. The study also requires collecting administrative division data for the corresponding years at the county level (including counties, county-level cities, and districts) and taking into account the administrative division adjustments that occurred between 2000 and 2020.

This study includes data from the years 2000, 2010, and 2020. There are several considerations for selecting these specific time points: (1) The study focuses on the 20-year period since the beginning of the 21st century. During this period, China’s urbanization rate increased rapidly, reaching over 50% around 2010 [4]. The characteristics of urbanization in the first decade differ noticeably from those in the second decade. (2) The selection of these years takes into account the limitations of data availability and other relevant data conditions.

The land use/cover situation in the BTH region in 2020 is shown in Figure 1a. The residential and employment population densities in the BTH region for 2020, calculated based on the above data, are shown in Figure 1b,c. Overall, the population and construction land are relatively concentrated in the southeastern region, which is closely related to topographical factors. The western region is occupied by the Taihang Mountains, the northern region by the Yanshan Mountains, while the southeastern region is the vast North China Plain, which has historically been a densely populated area. Major cities such as Beijing, Tianjin, Shijiazhuang, and Tangshan are the primary hubs of construction land and population. Some areas distant from city centers also have significant clusters of construction land but with sparser populations, primarily consisting of large port areas and industrial parks.

2.2. Methods

This study approaches the spatial distribution coupling relationships between construction land and the residential population, as well as the employment population, from three perspectives and paths. The flowchart of the methodology is shown in Figure 2. The three approaches each have their own advantages, so this study needs to conduct a comprehensive discussion of the results from the three approaches in order to gain a more thorough understanding. The spatial analysis methods and techniques involved in this study are implemented based on the ArcGIS platform.

The first path takes the perspective of the county-level spatial unit scale. Since the population data are already at the county-level spatial scale, it is only necessary to aggregate the construction land data into the vector polygons representing the counties. Each county polygon’s attribute Table should include at least three fields: construction land area, the residential population count, and the employment population count. Based on these field data values, the coupling coordination degree of the human–land relationship can be calculated and analyzed. The formula for calculating the coupling degree C is [54] as follows:

$$C={\left[{\displaystyle \frac{{\prod }_{i=1}^n{U}_i}{{\left(\frac{1}{n}{\sum }_{i=1}^n{U}_i\right)}^n}}\right]}^{{\displaystyle \frac{1}{n}}}$$

(1)

where n represents the number of subsystems, and this study involves three subsystems (construction land, the residential population, and the employment population). This study conducts a coupling analysis of construction land and the residential population distribution, as well as construction land and the employment population distribution. Each analysis includes two data systems, hence n = 2. U_i represents the values of each subsystem, and the original values need to be normalized to have a distribution range from 0 to 1. According to the principle of the formula, the range of the coupling degree C is [0, 1], where higher values indicate a stronger coupling between the subsystems.

The second path takes the perspective of the construction land grid cell scale. Since the majority of residential and employment populations are distributed within the construction land areas, population data for each county can be allocated to the construction land grid cells within it, allowing for more accurate identification of population spatial distribution characteristics. On this basis, the mean center positions of construction land, the residential population, and the employment population for different years are calculated, which can identify the migration of each subsystem’s center points. The population density of each grid cell is used as the weight value for calculating the mean center position. This part of the research is conducted at two scales: the overall scale of the BTH region and the scale within the boundaries of each city.

The third path takes the perspective of homogeneous grids. Using homogeneous grids can partially avoid interference caused by differences in county-level spatial unit scales or adjustments to administrative divisions. Considering data accuracy and the size of the study area, a 10 km by 10 km grid is chosen as the basic spatial unit. The construction land area, residential population, and employment population values are calculated for each basic spatial unit. This analysis step can be performed using ArcGIS through regional raster value analysis. Since each basic spatial unit stores values for the three subsystems, linear regression analysis can be performed on each unit to calculate the coefficient of determination (R²), which helps analyze the coordination between subsystems. The residual values between predicted values and actual values can identify the coupling degree of different subsystems.

3. Results

3.1. Subsystem Coupling Analysis Based on County Spatial Units

3.1.1. Subsystem Data Presentation Based on County Units

Figure 3 shows the proportion of construction land to the total land area within the county-level units in the BTH region for the years 2000, 2010, and 2020. It can be observed that the proportion of construction land is relatively high in the central areas of major cities such as Beijing, Tianjin, Shijiazhuang, Tangshan, Baoding, and Cangzhou. Some regions along Bohai Bay also have a relatively high proportion of construction land, such as Tianjin’s Binhai New Area, Tangshan’s Caofeidian area, and Cangzhou’s Huanghua Port area, among others. In contrast, the northern and western areas of the BTH region have large county-level areas with a relatively low proportion of construction land, generally below 10%.

Figure 4 shows the results of per capita construction land area calculated based on the residential and the employment population within county-level units in the BTH region for the years 2000, 2010, and 2020. It can be observed that the per capita construction land area in the central areas of Beijing, Tianjin, and almost all prefecture-level cities is relatively low. However, some coastal counties and a few western counties have higher per capita land area. It is worth noting that the coastal regions with higher per capita land area include Caofeidian, Tianjin Xinhai New Area, and Huanghua Port in Cangzhou, which are industrial areas relying on seaport development. This indicates that these areas have more extensive land use characteristics.

3.1.2. Analysis of Coupling between Subsystems

Figure 5 shows the results of the coupling degree C_R between construction land and the residential population distribution, and the coupling degree C_E between construction land and the employment population distribution within county-level units in the BTH region for the years 2000, 2010, and 2020. It can be observed that regions with significantly low coupling degrees include areas along Bohai Bay and mountainous regions in the northwest (e.g., Zhangjiakou). Additionally, several city centers also exhibit relatively low coupling degrees. Through the analysis of initial data for the subsystems, it is found that these two types of low coupling degrees correspond to different patterns. The low coupling degrees in the Bohai Bay area and the mountainous regions in the northwest indicate that the construction land is relatively abundant compared to the population. On the other hand, in some city centers, the low coupling degree indicates a scarcity of construction land relative to the population.

The differences between C_R and C_E can reveal more information, as shown in Figure 6 for the years 2000, 2010, and 2020. The green-colored areas represent regions where the differences between the two coupling degrees are relatively small, including most city centers and county-level areas in the central region. The purple-colored areas indicate regions where the coupling degree between the employment population and construction land has a significant advantage, mainly including coastal counties and peripheral areas of some central cities, which are often dominated by industrial development. The orange-colored areas represent regions where the coupling degree between the residential population and construction land has a significant advantage, mainly distributed in the western and northern regions of the BTH area, with sporadic distribution in other areas.

The summarized results of the C_E and C_R values for the entire BTH region and its 13 cities for the years 2000, 2010, and 2020, based on county-level spatial units, are shown in

Table A1. C_R and C_E values for different regions in different years.

Area	CR			CE
	2000	2010	2020	2000	2010	2020
Total	0.939 [0.748, 1.000]	0.952 [0.622, 1.000]	0.960 [0.601, 1.000]	0.953 [0.808, 1.000]	0.962 [0.649, 1.000]	0.959 [0.663, 1.000]
Beijing	0.911 [0.814, 1.000]	0.923 [0.747, 0.996]	0.949 [0.754, 1.000]	0.918 [0.839, 1.000]	0.933 [0.780, 0.992]	0.948 [0.754, 1.000]
Tianjin	0.901 [0.748, 0.990]	0.892 [0.730, 0.989]	0.884 [0.601, 1.000]	0.918 [0.808, 0.983]	0.920 [0.838, 0.976]	0.918 [0.771, 0.995]
Shijiazhuang	0.945 [0.806, 0.994]	0.972 [0.923, 1.000]	0.966 [0.836, 1.000]	0.955 [0.820, 0.999]	0.982 [0.932, 1.000]	0.966 [0.855, 0.999]
Tangshan	0.941 [0.897, 1.000]	0.921 [0.622, 0.978]	0.944 [0.761, 0.999]	0.955 [0.916, 1.000]	0.926 [0.649, 0.995]	0.949 [0.752, 1.000]
Qinhuangdao	0.964 [0.933, 1.000]	0.965 [0.947, 0.999]	0.982 [0.960, 0.999]	0.976 [0.934, 1.000]	0.965 [0.917, 0.989]	0.977 [0.925, 0.999]
Handan	0.968 [0.918, 1.000]	0.979 [0.900, 1.000]	0.991 [0.934, 1.000]	0.981 [0.934, 1.000]	0.987 [0.944, 0.999]	0.992 [0.961, 1.000]
Xingtai	0.955 [0.844, 0.998]	0.972 [0.872, 0.997]	0.984 [0.910, 1.000]	0.971 [0.880, 0.999]	0.985 [0.898, 0.999]	0.979 [0.903, 1.000]
Baoding	0.940 [0.898, 0.997]	0.966 [0.912, 1.000]	0.982 [0.918, 1.000]	0.955 [0.906, 1.000]	0.974 [0.877, 1.000]	0.981 [0.875, 1.000]
Zhangjiakou	0.927 [0.820, 1.000]	0.923 [0.849, 0.997]	0.913 [0.700, 1.000]	0.946 [0.871, 0.999]	0.921 [0.846, 1.000]	0.880 [0.696, 0.991]
Chengde	0.966 [0.933, 1.000]	0.975 [0.956, 1.000]	0.979 [0.927, 0.999]	0.971 [0.943, 0.997]	0.964 [0.913, 1.000]	0.966 [0.926, 0.999]
Cangzhou	0.951 [0.891, 0.991]	0.975 [0.917, 1.000]	0.958 [0.713, 1.000]	0.967 [0.890, 0.988]	0.985 [0.917, 0.999]	0.955 [0.663, 1.000]
Langfang	0.929 [0.872, 0.962]	0.957 [0.895, 0.983]	0.991 [0.964, 1.000]	0.939 [0.889, 0.975]	0.968 [0.922, 0.993]	0.990 [0.946, 0.999]
Hengshui	0.925 [0.893, 0.952]	0.953 [0.928, 0.983]	0.968 [0.934, 0.995]	0.955 [0.924, 0.978]	0.977 [0.965, 0.993]	0.972 [0.923, 0.994]

Note: In each cell, the first row of data represents the overall average value, and the second row of data represents the range of values for county-level spatial units within the region.

. Looking at the overall situation in the BTH region, the coupling degree (C_E) between the employment population and construction land was significantly higher in the early years (2000, 2010), but in recent years, the coupling degree (C_R) between the residential population and construction land has surpassed it. This trend is evident in almost all cities. In 2000, the C_E values of all 13 cities were higher than the C_R values. However, by 2020, the C_R values for Beijing, Shijiazhuang, Qinhuangdao, Xingtai, Baoding, Zhangjiakou, Chengde, Cangzhou, and Langfang had either surpassed or equaled the C_E values. This changing trend reflects the shift in construction land utilization and human–land coupling in the BTH region. It is inferred that in recent years, the utilization of construction land in this region has become more focused on meeting the residential needs of the population. In addition, from 2000 to 2020, there is a general trend of increasing C_R values at both the overall level of the BTH region and individual cities, further confirming the continuous improvement and enhancement of the coupling degree between the residential population and construction land in this area.

3.2. Changes in the Mean Center Positions Based on the Land Use Grids

3.2.1. Mean Center Change Characteristics of the Entire BTH Region

Figure 7 shows the results of the population density calculated based on the construction land raster data in the BTH region for the year 2020. The mean center positions of the three subsystems for the years 2000, 2010, and 2020, calculated based on the construction land grid, are shown in Figure 8.

At the overall level of the Beijing–Tianjin–Hebei region, the mean centers of the three subsystems (construction land, the residential population, and the employment population) are roughly located near the junction of Baoding, Cangzhou, and Langfang. The mean center of construction land tends to be more northern and eastern. Analyzing the shifting characteristics of the mean centers for the three subsystems from 2000 to 2020, it can be observed that they generally exhibit a trend of moving from southwest to northeast. The northeast direction roughly corresponds to the direction of Beijing, Tianjin, and Tangshan. This indicates a continuous aggregation of both construction land and population towards the core cities in this region.

Upon closer examination, distinct differences in the shifting characteristics of the mean centers for the three subsystems can be observed. (1) Regarding the direction of movement, the mean center of construction land tends to shift more towards the east, and even in the period from 2010 to 2020, it showed a southeastward movement. The location of the mean center of construction land lies east of the Bohai Sea, indicating a strong correlation between the recent growth of construction land in the Beijing–Tianjin–Hebei region and the areas along the Bohai Sea. For instance, the Tianjin Binhai New Area and Tangshan Caofeidian area were hotspots for construction during the period from 2000 to 2010, while the Cangzhou Bohai New Area and other regions became hotspots for construction during the period from 2010 to 2020. (2) In terms of the distance of movement, the mean centers of construction land and the residential population had larger shifting amplitudes during the first decade (2000–2010) than during the second decade (2010–2020), whereas the employment population showed the opposite pattern, with larger shifting amplitudes during the second decade (2010–2020). This feature indicates that the spatial pattern of construction land and the residential population has stabilized in recent years, while the employment population is still accelerating its concentration in dominant areas.

3.2.2. Mean Center Change Characteristics of the 13 Cities

The movement of the mean center of the three subsystems within the municipal areas of 13 cities from 2000 to 2020 is shown in Figure 9. The changes in the distances between the mean centers of construction land and population distribution (distinguishing between residential and employment populations) are shown in

Table A2. Changes in the mean center distance between construction land and population distribution in different regions.

Area	Distance between Mean Center of Construction Land and the Residential Population (m)			Distance between Mean Center of Construction Land and the Employment Population (m)
	2000	2010	2020	2000	2010	2020
Beijing	3036	3056	2569	2717	2037	2163
Tianjin	7153	6024	5910	4528	3210	4406
Shijiazhuang	276	518	1584	419	1341	409
Tangshan	967	6134	8033	1471	5225	6865
Qinhuangdao	4685	3397	6430	2576	818	3809
Handan	2395	1365	772	1786	3028	1470
Xingtai	1446	2275	1730	3290	5106	1515
Baoding	2210	1978	2461	2142	2671	3395
Zhangjiakou	11,986	9818	12147	9210	7772	15,282
Chengde	4867	4814	3078	1627	1475	3994
Cangzhou	2788	3499	9045	3632	4421	9124
Langfang	2532	848	2733	2264	2041	2437
Hengshui	1485	1432	2709	963	883	2442
Average	3525	3474	4554	2817	3079	4409

. By combining the results of Figure 9 and Table A2, some commonalities can be identified: (1) Except for a few cities, the overall movement directions of the mean centers for the three subsystems in most cities are consistent. (2) The distances between construction land and population centers have slightly expanded within a small range, possibly related to the overall urban expansion. (3) Compared to the residential population, the distance between the employment population and the mean center of construction land is shorter, but the gap is continuously narrowing.

Additionally, through in-depth analysis of individual cases, some distinct characteristics of different cities can be observed: (1) Cities around Beijing show a significant trend of mean center movement towards Beijing, particularly evident in cities like Langfang, where all three subsystems consistently move towards Beijing. The population center of cities like Baoding and Zhangjiakou also notably shifts in the direction of Beijing. This phenomenon reflects the urgent need for coordinated development in the Beijing surrounding area. (2) The mean center movements of the three subsystems are roughly similar in most cities, but some cities exhibit a situation where the mean center of construction land moves significantly more than the mean center of the population, such as Tangshan and Cangzhou, which may be related to the construction of large-scale industrial parks. (3) While most cities show relatively similar mean center movement directions for the three subsystems, some cities exhibit significant differences. For example, in Zhangjiakou, although the population center continues to move southeastward (towards Beijing) over the years, the mean center of construction land reversed its direction towards the north from 2010 to 2020. This may be related to Zhangjiakou’s role as a co-host city of the Beijing Winter Olympics and the construction of large-scale event venues in the northern region. (4) The periods of active mean center movements vary among different cities. The mean center of construction land in most cities, including Beijing, Tianjin, Shijiazhuang, and Tangshan, experienced significant movement in the first decade (2000–2010), while Cangzhou showed more significant movement in the second decade (2010–2020). This may be related to the developmental stages of the cities and the strategic timing for the development of the Bohai New Area.

3.3. Regression Analysis Based on Homogeneous Grids

3.3.1. Regression Analysis

Figure 10 shows the construction land area values for each 10 km by 10 km homogeneous grid in the years 2000, 2010, and 2020. The residential population and the employment population are also calculated based on these homogeneous grids.

The linear regression analysis based on the data values of the three subsystems stored in each homogeneous grid is shown in Figure 11 and Figure 12. The horizontal axis represents the construction land area, while the vertical axis represents the residential population and the employment population, respectively. The distribution of sample values in the two-dimensional geometric plane exhibits a clear linear fitting trend, with coefficients of determination (R²) generally exceeding 0.6.

Furthermore, by comparing the six plots in Figure 11 and Figure 12, the following observations can be made: (1) The linear regression fitting of the employment population and construction land has a clear advantage in terms of the coefficient of determination, but the relative advantage is decreasing, as evidenced by the narrowing gap in R² from 2000 to 2020. (2) The coordination between the residential population and construction land has improved, as indicated by the increase in the coefficient of determination from 0.6843 in 2000 to 0.6917 in 2020. (3) The coordination between the employment population and construction land has continuously decreased, reflected by a continuous decline in the coefficient of determination from 0.7731 in 2000 to 0.7266 in 2020.

3.3.2. Residual Analysis Based on the Method of Ordinary Least Squares

The results of the regression analysis reflect the overall coordination level between different subsystems. The residuals between the predicted and actual values for each homogeneous grid can be used to determine the differences in coordination levels across different regions. Figure 13 shows the results of the ordinary least squares analysis using ArcGIS, including residual analysis results for both construction land and the residential population values, and construction land and the employment population values.

Overall, the residual results for these two categories show similar characteristics. The light yellow areas represent regions with relatively small absolute values of residuals, covering the vast majority of the Beijing–Tianjin–Hebei region. The red areas indicate regions where the actual population values are significantly higher than the predicted values, suggesting higher population pressure on the construction land in these areas. The blue areas represent regions where the actual population values are significantly lower than the predicted values, indicating surplus construction land supply relative to the population scale in these areas.

The differences are also quite noticeable. Taking 2020 as an example, (1) In the grid areas near the central cities of Chengde and Zhangjiakou, although the actual residential population values are significantly higher than the predicted values, the absolute values of residuals for the employment population are within the smallest range. (2) The areas where the actual employment population values in Beijing are significantly higher than the predicted values are more extensive compared to the areas where the actual residential population values are significantly higher than the predicted values. These differences actually reflect variations in the development of work and residence functions in different regions to some extent.

4. Discussion

4.1. The Coordination with Construction Land Varies, with Residential and Employment Populations Exhibiting a Trade-Off Relationship

The empirical analysis based on the BTH region shows that there are significant differences in the coupling relationships between the residential population and construction land, as well as between the employment population and construction land, within a certain range. Compared to general studies on human–land relationships, this research provides more in-depth findings.

From the comparative perspective of the residential population and the employment population, the coupling degree between the employment population and construction land shows an overall advantage. This point is confirmed by the results from the three perspectives and paths of analysis. Compared to the residential population, the spatial distribution of the employment population is more closely related to the industrial layout. In the context of economic development, the layout of construction land needs to meet the needs of industrial development. Almost every city in the Beijing–Tianjin–Hebei region constructs various industrial parks as the focus of development, whether in coastal or inland cities. This can be considered the main reason for the significant advantage in the coupling degree of the employment population. It should be noted that in China, construction land indicators are closely related to population indicators. Therefore, the differences in coupling relationships between the residential and the employment population are reflected within a relatively narrow range.

In terms of change characteristics, the growth of the coupling degree between the residential population and construction land has an advantage. Although the coupling degree of the employment population is generally higher, its growth rate is limited and even shows a decrease in many areas during the period from 2010 to 2020. Meanwhile, the coupling degree of the residential population with construction land has grown rapidly, and in many areas, it has even surpassed the coupling degree of the employment population. This result indicates that the utilization pattern of construction land in this region is undergoing a transformation, shifting from a focus on meeting industrial layout needs in the past to a greater emphasis on meeting the needs of a better living environment in recent years.

Different cities exhibit certain differences due to their varying stages of development. For example, differences between more developed major cities (e.g., Beijing) and other cities, differences between cities around Beijing (e.g., Langfang, Zhangjiakou) and other cities, and differences between coastal port cities (e.g., Tianjin, Tangshan, Cangzhou) and inland cities.

4.2. The Key Measures to Improve the Coordination between Construction Land and Population Distribution

The empirical analysis of the Beijing–Tianjin–Hebei region shows that there are significant differences in the coupling relationships between the residential population and the employment population with respect to the spatial distribution of construction land. Key measures to improve the coupling degree of human–land relationships are proposed based on the analysis results of this study.

Promoting the compound utilization of construction land is crucial. The relatively low coupling degree of human–land relationships in the early stages is partly attributed to the single-type usage of construction land. Many industrial parks in the Beijing–Tianjin–Hebei region, such as Tianjin Binhai and Caofeidian, were primarily focused on industrial production, leading to a lack of residential, public service, and recreational facilities in these areas. This contributed to the high coupling degree of the employment population with construction land. In recent years, the approach of multifunctional use of construction land has gradually replaced single-use types, and this trend will continue to improve the coupling degree of human–land relationships.

Strict limits on the development of construction land in ecologically sensitive areas, such as mountainous regions, should be implemented. The western and northern parts of the Beijing–Tianjin–Hebei region consist of mountainous terrain, and there are wetlands in some areas of the central plain and ecologically sensitive zones in coastal areas. The analysis results indicated that in some regions of the northwestern mountainous areas, the coupling degree of human–land relationships was relatively low, as indicated by significantly higher per capita construction land indicators. Therefore, this region still needs to strengthen the constraints on construction land development in ecologically sensitive areas.

Efforts should be made to decongest public resources in the central areas of large cities and encourage population migration to the periphery. The central areas of most cities in the Beijing–Tianjin–Hebei region show a relatively low coupling degree of human–land relationships, suggesting significant pressure on construction land to accommodate residential and employment activities. This is often related to the high concentration of public service resources in these central areas. Therefore, improving the coupling degree of human–land relationships in the central areas can be achieved by optimizing the spatial distribution of public resources.

Establishing a sound mechanism for construction land withdrawal is essential. The aggregation of populations in areas with better living conditions and more job opportunities is a common pattern. In some areas with continuous outflow of population, if construction land is still retained, the coupling degree of human–land relationships will continue to decrease. Hence, it is necessary to establish a mechanism for construction land withdrawal based on the situation of population migration to promote the intensive use of land resources.

Giving attention to the sustainable utilization of construction land after major events is crucial. Hosting major events like the Olympics can significantly accelerate the development and construction of an area in a short period. For example, due to the preparation for the Winter Olympics, Zhangjiakou experienced significant construction of venues and supporting facilities from 2010 to 2020, which even influenced the change in the mean center of construction land. However, it is essential to consider how to promote the sustainable development and utilization of the relevant construction land after such events, as the direction of the mean center of construction land in Zhangjiakou seemed to be in reverse correlation with the center of the population’s movement. Therefore, after hosting major events, it is crucial to address the issue of sustainable development and utilization of related construction land.

4.3. The Applicability of the Analysis Methods Used in This Study to Urban Planning and Decision-Making

As the coordinated development relationship between people and land becomes increasingly important in national spatial planning, the analytical methods used in this study can be applied in the process of related planning and decision-making. This paper proposes three potential application paths for the framework.

Firstly, it can be applied during the current status assessment stage. The proposed analysis framework can be used to analyze the coupling relationship between the spatial patterns of population and construction land distribution, identify areas with low coupling coordination in the current status, and further investigate the specific reasons for the weak coordination by comparing the results of the residential population and the employment population analysis. The relevant analysis results can support the evaluation work in territorial spatial planning.

Secondly, it can be applied during the formulation of planning schemes. During the phase of planning scheme compilation in territorial spatial planning, this framework can be applied, together with future population distribution predictions, to evaluate the construction land planning and layout schemes. The analysis results can identify weak areas in the coupling relationship between the human–land system in the planning schemes, which can facilitate further optimization and improvement of the planning schemes.

Thirdly, it can be applied during the planning implementation phase. In recent years, Chinese cities have attached great importance to digital construction and have developed comprehensive database systems that integrate multiple layers of data sources, including land, urban construction, population census, and social governance. During the implementation phase of territorial spatial planning, if the analytical framework proposed in this study can be integrated with updated database systems, it will enable real-time dynamic monitoring of the coupling and coordination between human and land relationships.

5. Conclusions

This study establishes an analytical framework and explores from three perspectives and pathways in an attempt to answer the core question of “Which is More Closely Coupled with Changes in Construction Land: Residence or Employment?” Based on the analysis results from the Beijing–Tianjin–Hebei region during the period from 2000 to 2020, it shows that the employment population exhibits a significantly stronger coupling with construction land, while the coupling between the residential population and construction land has improved rapidly in recent years. Different cities display certain spatial and temporal characteristics in coupling relationships due to factors such as natural conditions, location, development level, and strategic opportunities. Building on the analysis findings, this study further discusses the trade-off relationship between different subsystems, key measures to enhance coordination, and potential applications of the analysis framework in planning and decision-making processes. Overall, this empirical study based on the analytical framework to a certain extent successfully addresses the above-mentioned question and essentially achieves the expected research objectives.

Of course, this study still has limitations. First, compared to the service industry, the manufacturing industry may have a greater demand for construction land. This study did not differentiate between employment populations in manufacturing and service industries. Second, within county areas, there are still significant differences in population density. Due to data limitations, this study could not further use population data at the township level. Third, this study did not consider variations in the intensity of construction land development.

As the stages of economic and social development continue to be iteratively renewed, human settlement and employment patterns, as well as the land-use pattern, also continue to change. The trend of these changes will affect the development of the future human–land coupling relationship. Therefore, the topics of concern in this study will continue to have greater academic and applied value in the future. To further enhance the scientific rigor of the analytical framework and address the limitations identified in this study, future research could optimize several aspects, such as (1) establishing specialized analytical frameworks for the needs of regional and urban research; (2) adopting higher precision data sources, and using population data of township or community units when possible, in order to better match with the precision of the data on construction land; (3) introducing the parameter of land-use intensity to deeply analyze the coupling relationship with population density.