Pre-Assessment Research of Regional Spatial Planning from the Perspective of Spatial Evolution

Abstract

As an important policy tool for guiding the development and protection of territorial space, the specific impact of spatial planning on the evolution of territorial space and the effects of its implementation have not yet been fully recognized. At the planning formulation stage, the feedback cycle of the implementation effect of planning is too long, and the adjustment is too slow to take effect. This directly affects the effectiveness and relevance of planning implementation. In this study, we propose a framework for pre-assessment of regional spatial planning based on the evolution of territorial space. We construct an indicator system from four dimensions to pre-assess the effectiveness of territorial spatial planning. The results show that the land use change simulation model, based on historical data coupled with planning elements, achieves an accuracy of 0.8457, which can accurately reflect the impact of planning on the evolution of land space. The results of the evaluation show that: (1) Planning elements such as concentrated urban construction areas, schools, and other public service facilities are the main driving forces behind the future evolution of territorial space. (2) Regional spatial planning has a significant effect on adjusting the spatial layout and optimizing the spatial form, and it has a certain effect on restricting the total scale of the construction area, improving the efficiency of land use, and lowering carbon emissions. (3) Existing planning strategies are still too rigid, so further enhancement of “flexibility” and “blank space” in planning is necessary. Additionally, guidance for improving the efficiency of land development needs to be further strengthened. The main contribution of this study is to provide a reliable pre-evaluation framework for policymakers and scholars. This will help identify problems and shortcomings at the planning formulation stage, allowing them to be optimized and adjusted in a timely manner. Under the guidance of scientific and reasonable planning, it will further promote the green and high-quality development and protection of territorial space.

1. Introduction

As a policy tool for managing urban development and shaping urban form, regional spatial planning plays an important role in urban development and the evolution of spatial patterns [1]. The essence of spatial planning is to predict and control the future development of a city through clear objectives and guidance, which is important for the efficient allocation of natural resources and high-quality socioeconomic development [2]. Since the 1980s, in order to better guide the development, protection, and utilization of territorial land space, China has issued a series of plans, such as regional planning and master planning for national land resources [3]. However, problems such as inconsistent planning technical standards, significant differences in spatial layout, and conflicting sectoral plans have persisted over a long period of time, severely restricting the effectiveness of their implementation and affecting the intensive and efficient utilization of regional land resources [4].

To address these issues, China has been in the process of establishing a new spatial planning system since 2014. The new spatial planning system integrates the main functional zone, land use, urban–rural development, and eco-environmental protection into a unified spatial planning framework to achieve the integration of multiple plans. Since new spatial planning is a spatial and temporal arrangement for the development and protection of the entire territory, it is named as territorial spatial planning [5]. The formulation of territorial spatial planning needs to be based on three fundamental bottom lines: the red line of ecological protection, the red line of permanent basic farmland, and the urban development boundary. The red line of ecological protection refers to areas with important ecological functions that need to be strictly protected. The red line of permanent basic farmland designates farmland under special protection to ensure the supply of important agricultural products and guarantee food security. The urban development boundary is established to prevent uncontrolled urban sprawl, and all construction and development activities within the planning period shall not exceed its spatial limitations [5,6]. In addition, in terms of the scale and rate of change for each land use type, the planning also imposes strict regulations and restrictions using quantitative planning targets. Based on pilot experiences, the first round of territorial spatial planning was officially launched in 2019. However, as of 2024, county-level spatial planning is still not complete in most counties. Thus, there is still not enough practical experience and relevant research to determine whether the implementation of land spatial planning oriented in this way has achieved the expected results.

According to the timeline, research on territorial spatial planning can be categorized into preparation- and implementation-stage research. In the preparation stage, scholars focus more on the theoretical discussion of planning and technical research on specific planning aspects. At the theoretical level, how to achieve the concepts of ecological civilization, regional synergy, intensive efficiency, carbon neutrality, and high-quality development in planning has been the focus of recent scholarly attention [7,8,9,10]. At the technical level, scholars are increasingly inclined to investigate proactive spatial optimization strategies to improve planning preparation methodologies. For example, to achieve the synergistic integration of planning objectives, methods such as the IPSO algorithm, multi-objective spatial optimization, and other spatial analysis models are widely used [11,12,13,14]. Alternatively, spatial methods such as machine learning models (MLMs) and multi-intelligence agent decision-making models can be used to optimize the layout of important planning areas, control baselines, or infrastructure, resulting in environmental protection and intensive and efficient territorial space development [15,16,17,18,19]. The application of spatial modeling techniques, particularly Cellular Automata (CA), to define growth boundaries for urban expansion or ecological area evolution has emerged as a prominent research focus in recent years [20,21]. In addition, how to effectively avoid ecological and social risks, such as ecosystem degradation, food security risks, and unbalanced urban–rural development, through planning remains a key area of study in recent research.

During the planning implementation stage, relevant studies focus more on summarizing and reflecting on planning outcomes after implementation [22]. For instance, by assessing the extent of planning completion and evaluating implementation outcomes, insights applicable to future planning can be synthesized [23,24]; or by consolidating and reflecting on management and institutional levels, it is feasible to evaluate potential deficiencies in the implementation of planning at the operational and supervisory levels, resulting in planning outcomes that fall short of expectations [25,26,27]. In terms of technology, some scholars develop theoretical frameworks to evaluate the effectiveness of planning implementation [28]. Examples include the theoretical model of data analysis grounded in social data, the literature-derived MEAS model, and the framework for assessing planning compliance based on interviews, observations, and planning documents, among others [29,30,31]. Numerous studies have also assessed the efficacy of different types of planning in areas such as economic efficiency, climate control, land development efficiency, performance standards, rural rehabilitation, and ecological services using mathematical models and indicator systems [32,33]. From a spatial perspective, predominant research involves comparing land use patterns in the target planning year with the original planning objectives, or further selecting pertinent indicators for appropriate metrics, and assessing the implementation of the finalized spatial planning [34]. The discussion of management systems and planning implementation strategies, on the other hand, focuses more on how to utilize natural means and economic policies to improve the effectiveness of planning implementation [35,36]. However, most of these studies are conducted after planning has been implemented or during its final stages. Therefore, the results and experiences of these studies can only be used to guide future planning efforts. Since there is usually a long period of time between two rounds of planning, and considering that relevant socioeconomic backgrounds, planning systems, and technical standards may undergo substantial changes, the experience of the previous round of planning offers only limited applicability for the next round of planning.

In this study, we systematically analyze the functional logic of planning elements in generating effectiveness and examine the driving mechanisms through which planning influences the evolution of regional spatial patterns. We also construct a land use change model that integrates key planning elements to simulate the spatial pattern of national territory by the final year of territorial spatial planning (2035). Then, we construct a planning pre-assessment system on this basis to quantitatively analyze the effectiveness of planning implementation. This research enables the prediction and evaluation of planning outcomes across multiple dimensions, while also providing practical and feasible suggestions for the formulation of territorial spatial planning.

2. Materials and Methods

2.1. Study Area

This study selected Datong County, China, as the main study area (Figure 1). Datong County is located in Qinghai Province in northwestern China, serving as a transition zone between the Tibetan Plateau and the Loess Plateau. The topography of the county is dominated by highland mountains, with abundant mountainous water systems and a wide distribution of ecological land and arable land. The forest coverage of the county reaches 43.3%, and the county is responsible for protecting more than 460 km² of farmland. Therefore, the urban development of the county is subject to multiple constraints of natural conditions, ecological protection and farmland protection. However, as an important node city within the Lanzhou–Xining urban agglomeration and an important member of Xining–Haidong Metropolitan Area, Datong County has urgent construction and development needs at the same time. In 2020, the country’s per capita GDP surpassed USD 4300, the urban population reached 206,000, and the urbanization rate exceeded 51%. Therefore, how to improve the land use efficiency and optimize the land development layout to further promote the economic development of the county through reasonable land space planning, without compromising ecological protection and maintaining the protection of arable land, requires urgent research and resolution.

2.2. Data Source

The 2009–2020 land use change data, permanent basic farmland data, DEM data, land use planning, urban general planning, and spatial master planning of land in Datong County used in this study were obtained from the Datong Natural Resources Department. Meteorological monitoring data are from the Meteorological and Water Resources Department of Datong County; socioeconomic data are from the Statistical Department of Datong County. In this research, the land use data of Datong County in 2020 is used as the base data, and the land use change model is constructed by combining relevant natural, socioeconomic data and planning elements. The simulation results are compared and analyzed with the relevant target indicators of the land spatial planning to conduct the pre-assessment study of the implementation of territorial spatial planning (

Table 1. Table of data types and sources.

Data Types	Data Contents	Data Access
Land use Data	Land use change data (2009–2020)	Datong Natural Resources Department
Natural Data	DEM data	Datong Natural Resources Department
	Meteorological monitoring data	Meteorological and Water Resources Department of Datong
Socioeconomic data	Population statistical data	Statistical Department of Datong
	Layout of urban functional facilities	Datong Natural Resources Department
Planning Data	Territorial spatial planning	Datong Natural Resources Department
	Land use planning
	Urban general planning
	Spatial master planning of land
	Layout of future urban functional facilities

).

2.3. Methods

This study initially computed the probability of land use type conversion across various distance intervals from the driving factors, utilizing historical planning elements with natural and socioeconomic driving factors and land use change data. Consequently, we formulated and utilized the functional connection between each driving force element and the probability of land use type conversion at specified distance intervals to determine the probability of future land use development. This process is primarily obtained by calculating the proportion of the area where land use type change occurs to the total area of the region at each 1000 m distance from each driving factor (e.g., 0–1000 m, 1000–2000 m, and so on). Afterward, we used the PLUS model to simulate the future spatial evolution of the national territory and identify the different characteristics of the spatial evolution of the national territory with or without planning guidance. This step is achieved mainly by comparing the differences in land use change between the planning-guidance scenario and natural development scenario. The two scenarios differ only in the choice of driving factors, with the planning guidance scenario considering the planning objectives as well as the guidance of other planning elements. However, the choice of other parameters such as the range of the model neighborhood, the number and manner of sampling, and the number of iterations was kept consistent. We also assessed the effectiveness or redundancy of the planning elements based on the extent of their impact on the evolution of the territorial spatial pattern. Secondly, utilizing the projected territorial spatial configuration for the planning target year and the set of planning objectives, we developed an indicator system including four dimensions: the completion of planning targets, the spatial conformity with the planning, the efficiency of space development, and the territorial spatial patterns, to evaluate the achievement of planning objectives. This method will help us to quantitatively measure the effectiveness and inadequacy of planning in various aspects. Finally, based on the effectiveness of the planning elements and the accomplishment of the planning objectives, we propose reasonable planning optimization recommendations to ensure reasonable and efficient implementation of future planning. The research technology roadmap is shown as Figure 2.

2.3.1. Driving Factors Selection and Driving Force Calculation

In this study, we divided the driving factors into three categories based on existing research. The first part is natural factors that objectively affect land use change, such as temperature, precipitation, and slope [37]. The second part is socioeconomic factors. It includes population distributions and existing urban functional facilities, such as point-like commercial, educational, scientific, and technological, administrative, and public service facilities, line-like roads and railroads, and face-like development areas and protected areas [38,39,40]. The third part is planning elements such as the centralized urban area [41], and the layout of future urban functional facilities based on the planning. Existing research results have confirmed that the impact of planning elements on land use changes tends to be regionally heterogeneous [42,43]. Therefore, we first calculate the probability of land use change occurring within different distances by using the land change data from 2010 to 2020 and each driving factor so as to analyze the driving force of each factor on land use change. The formula is as follows:

$${PC}_{n-m}={\displaystyle \frac{{SC}_{n-m}}{S_{n-m}}}$$

(1)

where ${PC}_{n-m}$ is the proportion of land use types that undergo land use conversion within the distance $n-m$ from the driving factor. The ${SC}_{n-m}$ is the area of land that undergoes land use conversion within the distance of $n-m$ from the driving factor. $S_{n-m}$ is the total area of land undergoing land change under the influence of this driving factor. Based on the ratio of the area of land use change patches within each distance to the total area of land use change patches within the distance for the years 2010–2020, the trend was fitted to obtain the functional relationship between each driving factor and land use change, as shown in

Table 2. Driving factors selection and functional relationships.

Type	Driving Factors	Functional Equation	${\mathit{R}}^2$
Natural factors	Annual precipitation	Normalize data values as driving force values
	Annual average temperature
	Terrain slope
	Population
Socioeconomic factors	Infrastructure	$y=0.0189e^{-0.127x}$	0.8602
	Transportation facilities	$y=0.022e^{-0.113x}$	0.7295
	Education	$y=0.038e^{-0.132x}$	0.7681
	Enterprise and factories	$y={\displaystyle \frac{1}{0.1667+31.601\times {1.049}^x}}$	0.613
	Business center	$y=0.0229e^{-0.147x}$	0.7231
	Medical and public health	$y=0.0325e^{-0.126x}$	0.8014
	Government departments	$y=0.0324e^{-0.117x}$	0.8076
	Residential community	$y=0.022e^{-0.153x}$	0.9046
Planning elements	Centralized urban area	$y=0.044e^{-0.156x}$	0.7263
	Village construction area	$y=0.006e^{-0.625x}$	0.8776
	General farmland area	$y=0.0138e^{-0.57x}$	0.9636
	Ecological control area	$y=0.00005e^{0.0945x}$	0.6461

. According to the actual change trend of the driving force of each driving factor with distance, the functional relationship between the two can be expressed by the following equation:

$$y={ae}^{-bx}$$

(2)

where y is the driving force of each driving factor; a is the driving force of the driving factor at a distance of 0 (normalized value of 0–1); b is the adaptive coefficient; and x is the distance from the driving factor.

The impact of different driving factors on land use change may be positive or negative. Therefore, we normalize the magnitude of the driving force generated by all the factors in order to facilitate the next step of the calculation. The formula is as follows:

$$Y_{ij}={\displaystyle \frac{X_{ij}-\mathit{Min}\left(X_{ij}\right)}{Max\left(X_{ij}\right)-Min\left(X_{ij}\right)}}$$

(3)

$$Y_{ij}={\displaystyle \frac{\mathit{Max}\left(X_{ij}\right){-X}_{ij}}{Max\left(X_{ij}\right)-Min\left(X_{ij}\right)}}$$

(4)

In the indicator calculation process, positive indicators are standardized using Equation (3), and negative indicators are standardized using Equation (4). Where $Y_{ij}$ is the standardized indicator value and $X_{ij}$ is the jth indicator of year i. $Max\left(X_{ij}\right)$ is the maximum value of the jth indicator of year i and $Min\left(X_{ij}\right)$ is the minimum value of the jth indicator of year i.

2.3.2. Land Use Change Simulation

To better explain the effects of spatial planning elements and other driving factors on land use change, we use the patch-generated land use simulation (PLUS) model to simulate future land use patterns. The PLUS model integrates the Land Expansion Analysis Strategy (LEAS) model and a cellular automata (CA) model based on multiple types of stochastic patch seeds, and is able to accurately simulate future land use patterns based on drivers and evolutionary patterns. The module allows the superposition of land use data from two periods, with patches of land use change at the end of the period representing the area of each land type transition, and with special markers for the different transition types. The markers are then used to extract multiple driving force values at the same location to construct the training dataset. This analysis method pays more attention to the type evolution of each patch, eliminating the influence of factors such as the number of land types on future land use changes, and it has high accuracy. In addition, LEAS adopts a binary classification to calculate the land type conversion rules, which makes the model highly compatible with other statistical calculation methods. Finally, the model extracts random samples from each training dataset through the random forest algorithm, mines to solve the land conversion data and driving factors data, and finally obtains the conversion law of each land use type. The formula is as follows:

$$P_{i,k}^d\left(x\right)={\displaystyle \frac{\sum _{n=1}^{M}I(h_n\left(x\right)=d)}{M}}$$

(5)

where $P_{i,k}^d$ is the probability of growth of land use type $k$ on patch $i$. The value of $d$ is 0 or 1, which denotes the conversion to other land use types or to land use type $k$, respectively. $x$ is a vector of multiple drivers. $h_n\left(x\right)$ is the type of prediction for the nth decision tree of the vector $x$, and $I$ is the indicator function for the set of decision trees. $M$ is the total number of decision trees. In the spatial simulation of land use change, PLUS uses a stochastic seed (CARS) model based on multiple types of land use. The model takes more account of the relationship between land demand, development objectives and land competition. Technically, the model constructs a roulette system based on the overall probability of all land use types for selecting the land use state in the next iteration and thus is more adaptive to land use changes guided by planning objectives. More details of the model can be found in Liang’s research [44].

2.3.3. Spatial Evolution Scenario Setting

As the main objective of this study is to construct an initial framework for planning pre-assessment, the study focuses on assessing the scientific validity and effectiveness of planning. Therefore, in order to more accurately identify the overall impacts on the region under the macro-guidelines of the planning, we set up a natural urban growth scenario and a planning guidance scenario in this study to conduct a comparative research [45]. In the natural urban growth scenario, the expansion and evolution of territorial space is driven by its own objective natural conditions and existing socioeconomic factors. It develops in a completely natural expansion mode and is not guided or constrained by policies and plannings. In the planning-guided scenario, the expansion and evolution of territorial space is more driven by planning elements. In this research, the driving factor data of the natural development scenario mainly come from the data of natural conditions such as temperature, precipitation, elevation, slope, etc., of the current year, as well as the data of various socioeconomic facilities such as commerce, education, medical care, etc., that have been built. In the planning guidance scenario, key development areas, public service facilities planned to be built in the future, and key construction projects such as industries are added to the simulation according to the local territorial spatial planning. In addition, the red line of the ecological protection area and the red line of permanent basic farmland protection in the planning are also included in the model in the form of prohibited construction areas to reflect the seriousness of the planning red line. That is, within the red line area, the existing land use status must be maintained, and any development, construction, and other activities are prohibited.

In the setting of the simulation. For the future land use change targets, the development targets of the natural development scenario are calculated based on the land use data of the past years using the Markov chain prediction. In the planning-guided scenario, the future development targets of the region are calculated entirely based on the targets set by territorial spatial planning. And for the parameters that need to be set for the simulation, we uniformly set the neighborhood size of the model to 3, the expansion coefficient to 0.1, and the percentage of random seeds to 0.0001. These settings are intended to more realistically simulate the effects of changes in each land use type as influenced by the current state of the surrounding land use, and to reduce the occurrence of random irregular land use changes. In terms of the feasibility of inter-conversion of land use types, we have prohibited the possibility of conversion from urban to agricultural and ecological land. This is because it hardly ever happens in cities nowadays. The initial conversion probability between land use types is calculated from historical land use change data.

2.3.4. Construction of the Evaluation Indicator System

In order to more accurately evaluate the effect of planning implementation, and based on existing research, we constructed an evaluation indicator system from four dimensions: the completion of planning targets [46,47], the spatial conformity with the planning [48,49], the efficiency of space development [50,51], and territorial spatial patterns [52,53] (

Table 3. Evaluation indicator selection and calculation formula.

Evaluation Dimensions	Evaluation Indicators	Calculation Formula	Formula Explanation
Degree of completion of planning targets	Total scale of construction land	$C={\displaystyle \frac{\left|S-P\right|}{P}}$	$C$: Degree of completion of the planning target $S$: Actual indicator value $P$: Planning target value
	Scale of new construction land
	Forest cover rate
	Farmland area
	Water surface rate
	Per capita urban construction land area
Degree of spatial conformity with the planning	Total construction area in the planning area	$M={\displaystyle \frac{P_i}{P_t}}$	$M$: Degree of spatial conformity with planning $P_i$: Number of raster patches in the planning area $P_t$: Total number of raster patches for a given land use type
	Percentage of construction land in suitable urban construction area
	Percentage of construction land in ecological protection areas
	Percentage of ecological land in ecological protection areas
Efficiency of space development	Intensity of expansion of construction land	$I={\displaystyle \frac{A_b-A_a}{TA}}\times {\displaystyle \frac{1}{T}}$	$I$: Intensity of expansion of construction land $A_b$: Construction land area at the end year of the study $A_a$: Construction land area at the first year of the study $TA$: Total land area $T$: Time interval (years)
	Utilization rate of stock land	$R_s={\displaystyle \frac{S_i}{S_t}}$	$S_i$: Stock land in construction land at the end year of the study $S_t$: Total stock land area
	Average land GDP	$C_{GDP}={\displaystyle \frac{GDP}{A}}$	$GDP$: GDP of the study area $A$: Total area of construction land of the study area
	Average annual carbon dioxide emissions change *	$\begin{array}{l}{C}_c\\ =E_{b,i}+E_{b-a}-S_{a,i}{-S}_{b-a}\end{array}$	$C_c$: Change in carbon emissions $E_{b,i}$: Carbon emissions from land category $i$ in year b $E_{b-a}$: Carbon emissions from land use change from year $a$ to year $b$ $S_{a,i}$: Carbon sequestration for land category $i$ in year $b$. $S_{b-a}$: Carbon sequestration from land use change from year $a$ to year $b$
Urban spatial pattern	PLAND index	$\begin{array}{l}PLAND\\ ={\displaystyle \frac{\sum _{j=1}^n{a}_{ij}}{A}}\times 100\end{array}$	$a_{ij}$: Area of landscape type $i$raster j $A$: Total area of landscape
	AWMPFD index	$\begin{array}{l}AWMPFD\\ ={\displaystyle \frac{\sum _{i=1}^n\left(\frac{2\ln \left(0.25p_j\right)}{\ln a_j}\right)}{N}}\end{array}$	$a_i$: Area of raster patch i $p_i$: Perimeter of raster patch $i$
	MNN index	$MNN={\displaystyle \sum _{i=1}^n}{h}_i\times {\displaystyle \frac{1}{n}}$	$h_i$: Distance from raster patch $i$to the nearest construction land patch

Note*: The land use carbon balance coefficients used in this study were derived from the relevant studies [7,54].

). Based on the historical land use data and simulation data of the study area, we calculate the value of each indicator. The entropy weight method is used to calculate the weight of each indicator and evaluate the score of each indicator so as to realize the quantitative evaluation of the effect of future planning implementation.

The indicators at the level of planning indicator completion are mainly related to the land use pattern explicitly set in territorial spatial planning, such as the total scale of construction land and forest coverage rate. The initial value of this type of indicator is determined by directly calculating the area of each land type for each year. The closer the calculated value of the indicator is to 0, the closer the indicator’s actual value is to the targets set by the planning. As a result, after calculating each indicator value, they must be standardized using Equation (4).

The degree of spatial conformity of the planning evaluates whether the evolution of the regional spatial pattern of land is in line with the planning expectations. These indicators are calculated by the percentage of some land use types that correspond to the planning area or the percentage of land use types that should not be present in the planning area. Of these indicators, the total construction area in the planning area, percentage of construction land in suitable urban construction area, percentage of ecological land in ecological protection areas represents the consistency between the reality and the planning area. Hence, these are positive indicators, which need to be standardized using Equation (3). At the same time, the larger the percentage of construction land in ecological protection areas, the greater the discrepancy between reality and the planning area. Therefore, this indicator needs to be standardized using Equation (4).

Land development efficiency is a representative of the quality of urban development. In this study, in addition to selecting traditional indicators such as the intensity of land expansion, changes in carbon emissions in the conversion of land categories are also considered to analyze the efficiency of urban territorial spatial development under the guidance of planning in a more comprehensive way. Among these indicators, the intensity of expansion of construction land is a negative indicator, and the others are positive. They also require comparable standardized computations.

In addition, the urban spatial pattern also reflects the scientific and rationality of the spatial layout of future land use. In this study, we utilize Fragstats v4.2 software to calculate the landscape pattern indicators to reflect the land use change trends guided by future planning. Among them, the PLAND and AWMPFD are positive indicators, while the MNN is the negative indicator. The calculation method of specific indicators is shown in Table 3.

2.3.5. Validation of Model Accuracy

In order to assess the feasibility and accuracy of the land use change simulation model, we initially simulated the land use pattern of the study area for 2020, utilizing the land use change data from 2010 along with relevant driving and planning factors from that period. We then compared the 2020 simulation results with the actual land use pattern of 2020 to evaluate the accuracy and validity of the PLUS model. To quantitatively evaluate the model, we use the Kappa coefficient to reflect the accuracy of the model. The Kappa coefficient is often used to evaluate the consistency and reliability of classification results and is also commonly used to analyze the accuracy of raster data. The formula for calculating the Kappa coefficient for real and simulated raster plots can be expressed as:

$$k={\displaystyle \frac{P_o-P_c}{1-P_c}}$$

(6)

$$P_o={\displaystyle \frac{S}{n}}$$

(7)

$$P_c={\displaystyle \frac{a_1\times b_1+a_2\times b_2+\dots +a_n\times b_n}{n\times n}}$$

(8)

where $S$ is the number of rasters for which the simulated value matches the true value, $n$ is the total number of rasters, $a_n$ is the number of rasters for which the true raster value is n, and $b_n$ is the number of rasters for which the simulated raster value is $n$. The Kappa value is between 0 and 1. The closer it is to 1, the higher the consistency of the simulation.

3. Results

3.1. Driving Factors Analysis

As mentioned above, different types of driving factors have their unique driving patterns for urban spatial evolution, and the distance factor can be utilized to connect the relationship between driving factors and land use change. In this study, based on the land use data of 2010 and 2020 and the current land use master planning, we calculated and analyzed the driving force equations of each driving factor.

Figure 3 shows the change in the proportion of land use type conversions per unit distance with respect to the distance of each driving factor. It should be noted that the analysis here only covers the external conditions of each driving factor and does not include the land use changes inside the faceted driving factors. This is a result of the fact that the probability of construction and development is the highest of all areas in the urban development areas and village construction areas, which are the key construction areas in the future. In the protection areas, construction activities are strictly constrained and limited, and the probability of land construction changes is 0.

From the analysis of the results, the impact of driving factors such as infrastructure, transportation facilities, education, enterprises, factories, businesses, medical and public health services, government departments, residential communities, centralized urban areas, village construction areas, general farmland area plays a positive role in land use change; while the impact of ecologically controlled areas on land use change is negative. Specifically, the driving factors that undertake major urban functions, such as infrastructure, have played a positive role in driving urban development and construction. Urban construction and expansion are extremely active around these elements. However, outside the top 10% of the total distance range of these factors, the probability of construction and development activities occurring suddenly drops to an extremely low level and remains constant. Although the basic farmland protection area is the bottom-bound factor, it is surrounded by general farmland, which is the main source of land converted to construction land. Therefore, construction and development activities are still very active in the surrounding general farmland areas close to the basic farmland protection areas. On the other hand, the ecological protection area, which is also a bottom-line constraint factor, is generally located in the core area of the regional ecological space, and there are few construction and development activities even outside the protection area. The regions surrounding urban areas and villages are the primary regions for development and construction activities, with a particular concentration of these activities occurring within 4 km of existing urban construction land. In contrast, the development around villages is relatively loose, with active development activities within 10 km. In general, though, with the exception of ecological control areas, all the other driving factors have a positive effect on urban construction and development, which diminishes rapidly with distance (Figure 3).

3.2. Model Parameters Analysis

3.2.1. Model Accuracy Validation

In this study, we use the KAPPA coefficient to calculate and test the accuracy of the model. Since the preparation of China’s territorial spatial planning had not yet begun in 2010, the planning data used in the validation phase of the model came from the land use general planning and urban planning of that year. After calculation, the simulation accuracy of urban construction land expansion in the planning-guided scenario reaches 0.8457, and in the natural development scenario, the simulation accuracy of urban construction land reaches 0.8636, which indicates that the quantification method of driving factors and the land use change simulation model adopted in this paper have high simulation accuracy and can meet the needs of the study.

3.2.2. Quantification of Driving Factors

According to the calculation methodology in Section 2.3.1, we first calculated the distance between each patch and each driving factor in the study area. The percentage segments of the distances were then calculated by substituting them into the functions corresponding to each factor in Table 3 and obtaining the values of the driving influence of each factor in each raster cell. The larger the number, the stronger the driving force of that factor on land use change. From the results, we can see that spatially, the high driving influence of each factor is distributed in a relatively concentrated area. After exceeding a certain distance range, the driving influence of each factor stabilizes at a very low level, which is consistent with the influence range of the relevant factors. The results of the calculation of the future driving factors are shown in Figure 4.

3.2.3. Conversion Probability Distributions for Each Land Use Type

Based on the calculation results of each driving factor, we applied the LEAS module to explore the relationship between each driving factor and land use change in order to obtain the probability of each land use type conversion (Figure 5). Before the simulation analysis, the total probability of land conversion occurring in each patch is first compared and analyzed. Figure 6 shows the sum of the number of rasters in the study area with land use changes for all land use types as a proportion of the total number of rasters in the study area. From the results, the probability of land use type conversion occurring in the parcel under the natural development scenario is more distributed in the low probability range of 0.05~0.4. The number of rasters in this interval reaches 1,476,170, which accounts for 86.70% of the total number of all rasters with the possibility of development. Under planning guidance, the number of rasters in this range only accounts for 71.28% of the total number of rasters. Correspondingly, under the planning guided scenario, the number of rasters in the range of 0.5 to 1.0 where land use type conversion occurs is relatively higher, amounting to 364,037 rasters, accounting for 25.98% of all patches with development possibilities. In the natural development scenario, the proportion of rasters within the same range is only 11.63%. This shows that under planning guidance, the influence of various driving factors is more specific and the location of areas for potential development in the future is more concentrated.

Conversely, in the natural development scenario, the quantity of rasters subject to change is larger and more broadly distributed, but the disparity in development the probability across rasters is less. This will lead to a more dispersed and fragmented distribution of urban construction and ecological land in the future, and a more disordered land use pattern. In addition, the number of rasters with 0 development probability under the planning guided scenario, which means the number of rasters that will obviously not be developed, reaches 204,363,939, which is significantly more than the 173,600,001 rasters under the natural development scenario. It shows the compact and intensive characteristics of the planning guided scenario. In terms of the match between high development probability areas and planning construction areas, the high probability areas of urban development under the planning guided scenario are more concentrated around the planning construction areas, and the remaining externally high-probability areas are similarly located near the existing developed areas. However, in the natural development scenario, the match between urban development and planning is weaker, and there are still large-scale high potential areas for construction away from the city center.

3.3. PLUS-Based Land Use Change Simulation

Based on the probability distribution of the conversion of each land use type under different scenarios, we use the PLUS model to simulate the future spatial pattern of the country. The simulation results are shown in Figure 7 and

Table 4. Land use transfer matrix for both scenarios. Unit: hectare.

		2035	Urban	Village	Farmland	Forest	Grassland	Water Bodies	Other Land	Total land Area
	2018
Planning Guidance Scenario	Urban		2999.16	0.00	0.00	0.00	0.00	0.00	0.00	2999.16
	Village		6.03	7254.63	0.00	5.94	0.00	0.00	0.00	7266.60
	Farmland		30.60	0.00	68,096.61	317.79	0.00	0.00	0.00	68,445.00
	Forest		776.97	0.00	0.00	136,617.30	0.00	0.00	4.50	137,398.77
	Grassland		0.00	0.00	0.00	0.00	86,069.61	0.00	0.00	86,069.61
	Water bodies		0.00	0.00	0.00	0.00	0.00	668.07	0.00	668.07
	Other land		86.40	0.00	0.00	60.75	0.00	0.27	7052.67	7200.09
	Total land area		3899.16	7254.63	68,096.61	137,001.78	86,069.61	668.34	7057.17	310,047.30
Natural Development Scenario	Urban		2999.16	0.00	0.00	0.00	0.00	0.00	0.00	2999.16
	Village		5.13	7188.48	60.39	12.33	0.27	0.00	0.00	7266.60
	Farmland		218.61	485.28	66,586.77	1151.82	2.52	0.00	0.00	68,445.00
	Forest		697.68	103.14	764.91	135,821.07	11.97	0.00	0.00	137,398.77
	Grassland		15.84	33.21	10.35	54.45	85,955.76	0.00	0.00	86,069.61
	Water bodies		0.00	0.00	0.00	0.00	0.00	668.07	0.00	668.07
	Other land		8.37	0.81	0.00	0.00	0.00	0.00	7190.91	7200.09
	Total land area		3944.79	7810.92	67,422.42	137,039.67	85,970.52	668.07	7190.91	310,047.30

. As can be seen from the figure, the simulation results of the two scenarios have some differences in the quantity, spatial distribution and spatial pattern of each land use type. In terms of quantity structure, the urban land, village, forest, and unutilized land in the planning-guided scenario have a smaller area than the natural development scenario, while the area of farmland, grassland, and waterbodies is significantly larger than the area of the same land type in the natural development scenario. This result is not only due to the fact that, under the guidance of planning, the speed of urban development and construction, as well as the total area available for new construction, are strictly limited, but also because the objectives of ecological and farmland protection are clearly put forward. More importantly, urban construction is spatially limited to a specific area by the urban development boundary, and its occupation of farmland and other ecological land is strictly controlled. In the natural development scenario, since there are no quantitative or spatial limits, the development of construction land is quick and non-aggregated, and the emergence of other land use types is similarly highly random. As a result, the natural development scenario exhibits more disorderly traits. Furthermore, agricultural and ecological land is being occupied more extensively. As in the natural development scenario, the size of farmland is 674.19 hectares smaller than in the planning guidance scenario.

Overlaying the 2035 simulation results with the current land use situation in 2020 reveals that the main source of new construction land under both scenarios is different. In the planning guided scenario, the additional urban land is primarily converted from forest land, totaling 776.97 hectares or 86.33% of all new urban land. At the same time, the effect of farmland protection is more obvious, only 30.6 hectares of farmland are occupied. In the natural development scenario, the construction of urban and villages consumes more farmland, resulting in a significant reduction in the amount of farmland. In contrast, the natural development scenario consumes 764.91 hectares more forest and 10.35 more grassland to replenish farmland, leading to regional land use chaos. Nevertheless, the planning guided scenario utilizes additional general farmland outside the planning basic farmland protection areas, requiring an additional 73.44 hectares to restore farmland in order to satisfy the stringent morphological requirements of the planning area (Table 4).

As for the spatial distribution and spatial pattern, from Figure 7, it can be clearly seen that under the guidance of planning, the distribution of patches within each land use type is more clustered. The boundaries of towns, villages, and other various land use types are straighter and clearer. In the case of natural development, the distribution of patches of each land use type is much more sporadic, with relatively small unit patches and more irregular shapes. Most of its borders are also more curved and tortuous. As a result, the functions of each land use type will not be completely exhibited, and land use efficiency will be significantly lowered.

3.4. Contribution Analysis of Driving Factors

By analyzing the degree of contribution of each driving factor under different scenarios, we can obtain the role of each planning factor in future land use change. This calculation can be realized in the Land Expansion Analysis Strategy (LEAS) module of the PLUS model. The results show that under the planning guidance scenario, urban development is more guided by factors such as centralized urban area, temperature and layout of science, education, culture, and medical facilities, and this development pattern is in line with the actual law of urban development. Under the guidance of the centralized urban area, the urban development is further clustered in an orderly manner. The planned enterprises, factories, and other public infrastructures also greatly influence the layout of urban development. In addition, as the study area is located in the plateau region, the cities are more inclined to develop towards areas with higher average annual temperatures. In contrast, in the natural development scenario, natural factors such as temperature and precipitation dominate urban land use changes. Among existing infrastructures, only the commercial facilities have a more significant role in guiding urban development. This also reflects the fact that planned infrastructure is more attractive to developers and land users than existing infrastructure, and that planning scenarios for future urban development have a more significant impact on urban development (Figure 8).

4. Planning Rationality Analysis

In order to more accurately analyze the rationality and effectiveness of territorial spatial planning. This study constructs an evaluation indicator system from four dimensions: the completion of planning targets, the spatial conformity with the planning, the efficiency of space development, and territorial spatial patterns. Through the comparison of indicator values under different scenarios, we analyze the possible positive or negative impacts of spatial planning on future land use patterns. The indicator system is shown in Table 3. This study adopts the entropy weight method to calculate the scoring weight of each indicator. The score of each indicator value is shown in

Table 5. Evaluation of the rationality of territorial spatial development.

Evaluation Dimensions	Evaluation Indicators	2010	2012	2014	2015	2016	2017	2018	2019	2020	Planning Guided Scenario (2035)	Natural Development Scenario (2035)
Degree of completion of planning targets	Total scale of construction land	0.0443	0.0291	0.0237	0.0208	0.0202	0.0174	0.0104	0.0073	0.0044	0.0027	0.0000
	Scale of new construction land	0.0128	0.0000	0.0132	0.0164	0.0194	0.0166	0.0110	0.0161	0.0164	0.0073	0.0067
	Forest cover rate	0.0239	0.0239	0.0238	0.0238	0.0238	0.0236	0.0236	0.0236	0.0235	0.0014	0.0000
	Farmland area	0.0111	0.0075	0.0055	0.0044	0.0026	0.0034	0.0015	0.0004	0.0000	0.1743	0.1758
	Water surface rate	0.0270	0.0269	0.0269	0.0269	0.0268	0.0268	0.0267	0.0267	0.0264	0.0001	0.0000
	Per capita urban construction land area	0.0057	0.0015	0.0000	0.0230	0.0310	0.0318	0.0310	0.0306	0.0326	0.0325	0.0321
	Total score for the Dimension	0.1247	0.0888	0.0932	0.1154	0.1238	0.1196	0.1042	0.1045	0.1033	0.2183	0.2146
Degree of spatial conformity with the planning	Total construction area in the planning area	0.0000	0.0073	0.0084	0.0092	0.0096	0.0184	0.0302	0.0307	0.0176	0.0187	0.0095
	Percentage of construction land in suitable urban construction area	0.0000	0.0027	0.0043	0.0067	0.0072	0.0108	0.0728	0.0742	0.0158	0.0497	0.0406
	Percentage of construction land in ecological protection areas	0.0172	0.0030	0.0000	0.0030	0.0035	0.0000	0.1178	0.1192	0.0031	0.0970	0.0200
	Percentage of ecological land in ecological protection areas	0.0000	0.0009	0.0013	0.0017	0.0017	0.0051	0.1484	0.1486	0.0070	0.0199	0.0278
	Total score for the Dimension	0.0172	0.0139	0.0140	0.0205	0.0221	0.0343	0.3693	0.3727	0.0435	0.1854	0.0980
Efficiency of space development	Intensity of expansion of construction land	0.0096	0.0000	0.0099	0.0123	0.0146	0.0125	0.0083	0.0121	0.0123	0.0113	0.0111
	Utilization rate of stock land	0.0408	0.0427	0.0000	0.0031	0.0067	0.0085	0.0131	0.0180	0.0201	0.0355	0.0335
	Average land GDP	0.0017	0.0067	0.0119	0.0185	0.0185	0.0085	0.0059	0.0000	0.0053	0.1135	0.1112
	Average annual carbon dioxide emissions change	0.0184	0.0078	0.0172	0.0217	0.0193	0.0190	0.0225	0.0211	0.0230	0.0044	0.0000
	Total score for the Dimension	0.0797	0.0572	0.0490	0.0679	0.0734	0.0610	0.0587	0.0634	0.0732	0.1763	0.1673
Urban spatial pattern	PLAND index	0.0257	0.0224	0.0212	0.0206	0.0204	0.0198	0.0183	0.0176	0.0170	0.0009	0.0000
	AWMPFD index	0.0073	0.0083	0.0092	0.0092	0.0085	0.0000	0.0078	0.0124	0.0124	0.0211	0.0146
	MNN index	0.0000	0.0034	0.0036	0.0136	0.0150	0.0170	0.0317	0.0469	0.0493	0.0363	0.0105
	Total score for the Dimension	0.0330	0.0340	0.0341	0.0434	0.0439	0.0368	0.0578	0.0769	0.0787	0.0584	0.0250
Total Score		0.2546	0.1939	0.1903	0.2473	0.2632	0.2517	0.5899	0.6175	0.2987	0.6384	0.5049

.

From the total score of the evaluation of the rationality of territorial spatial development, it can be seen that the total score from 2010 to 2017 has always stabilized at a relatively low level. This indicates a suboptimal alignment between spatial development and planning at this period, as planning approaches fail to effectively direct the progression and development of the space pattern, resulting in disorderly development across different land use types. When the plannings are rigorously enforced, by 2018–2020, the end of the last round of planning, the total score is significantly higher than those of other years, except for the score in 2035 in the planning guided scenario. As traditional planning is also effective during 2010–2020, this proves that traditional plannings have some positive effect on enhancing the rationality and effectiveness of spatial development. However, compared with more unified territorial spatial planning, there is still a certain gap in its planning and guidance effect.

From the different evaluation dimensions, for the degree of completion of planning targets, the national spatial pattern of the study area under the planning guidance scenario in 2035 better accomplished the target indicators set by territorial spatial planning. In particular, the effect is more obvious than the natural development scenario in limiting the scale of urban construction land and forest protection. Based on the analysis of historical data from 2010 to 2020, the overall trend of the total completion scores has been decreasing year by year. Among the factors affecting this indicator, the scores of the total scale of construction land and the forest coverage rate have declined most significantly. The role of planning in controlling the spatial pattern of the territory space has gradually weakened over time. Especially at the end of the planning period, construction activities became more intense, occupying a considerable amount of forest land and other land. However, it cannot be ignored that the completion of farmland protection under planning guidance is worse than the natural development scenario. Spatially, this part of over-occupied farmland is concentrated in the area around the city, especially the current farmland in the centralized construction area of the city. This indicates that under the combined regulatory limitations of the urban development boundary and the red line of basic farmland protection, general farmland located outside the basic farmland protection areas, especially farmland located in urban areas, will have a greater possibility of being occupied.

In terms of the spatial conformity with the planning the spatial pattern under the guidance of planning aligns significantly greater than in other scenarios and years. This shows that planning has strong spatial constraints on land use. Urban development is also able to focus on suitable areas for urban construction and reasonably avoid unsuitable construction areas and important ecological protection regions. However, under the natural development scenario, the encroachment of urban development on the ecological space is particularly obvious, which further highlights the importance of the role of planning in guiding land use changes. In the historical data, the spatial control effect of planning is not obvious due to the wide range of planning areas delineated in the original land use planning. This suggests that we need to improve the relevance of control measures and instruments in the preparation stage of territorial spatial planning.

Regarding the efficiency of space development, the score in the planning guidance scenario also improves significantly due to the clear restrictions on the utilization of stock land in the planning. However, as with the natural development scenario, the improvement in this score is also influenced by better prospects for regional economic development and higher expected GDP. Although the speed and intensity of land development decrease under both scenarios, it continues to remain at a high level. As the scale of construction land continues to expand, the area’s carbon emissions are still rising annually, resulting in a much lower score for the carbon emissions indicator in 2035 for both scenarios than in previous years. Therefore, the impact of this planning on enhancing land development efficiency is not obvious, especially in the elimination of the stock land and the control of new construction land, there is still much room for optimization.

Finally, in terms of the territorial spatial patterns, the score under the natural development scenario is significantly lower than that of the historical years, indicating that the construction land under this scenario. This suggests that the construction land in this scenario is clearly experiencing disorderly expansion, with significant variations in the size, density, and configuration of land use patches, which are unevenly distributed. These problems have been solved in territorial spatial planning to a certain extent. Especially the spatial density and scale of the city under the guidance of the planning have been kept within a reasonable range. And in the extensive metropolitan areas, appropriate vacant spaces have emerged, enhancing the form and morphology of the urban environment.

5. Discussion

5.1. Viability of Employing Land Use Change Simulation Models for Planning Evaluations

In previous planning evaluation research, since the effects of planning need to undergo a period of implementation before they can be realized, most of the studies were conducted after the completion of planning implementation. Although relevant experiences can be summarized from historical planning studies for future planning, these experiences are often macroscopic and principled, and still lack relevance and timeliness for addressing the preparation, implementation, and optimization adjustment of territorial spatial planning. The planning implementation pre-assessment framework proposed in this study uses a land use change model to simulate the future spatial pattern of the national territory to realize the pre-assessment of the planning implementation effect. From the historical data, there is an obvious positive or negative correlation between different types of land use changes and the driving factor distance, and the experimental results of this land use change model have a high precision, which also proves that it is feasible to quantify the spatial planning elements using distance as a variable. The analysis results also show that the evaluation system in this study can help to identify the possible problems and defects in planning in a timely manner and provide policy suggestions for timely adjustment and optimization of planning. This is particularly beneficial for spatial indicators such as spatial pattern and land use structure, where the model is better adapted. However, the model is still unable to accurately predict and analyze the micro-indicators within the city, such as the coverage of green space and the average commuting time of residents.

5.2. Impact of Planning Elements on the Evolution of Territorial Space

The results of the study show that the evolution of spatial patterns by territorial spatial planning is mainly reflected in three aspects: quantitative structure, spatial layout, and spatial morphology. In terms of quantity, the planning has changed the long-term development pattern of the study area, which was based on rapid expansion, to one based on the pursuit of efficiency. Although the average growth rate of 49 hectares of new construction land per year is not the lowest level in history, compared with the rate of 63 hectares per year from 2010 to 2020, the trend of slowing down is still obvious. The required ratio of new construction land to stored land, as specified in the planning, decreases the necessity for new construction to extend into farmland, facilitating the attainment of farmland conservation objectives.

In terms of spatial layout, the planning has strictly delineated centralized urban construction areas and clearly defined the key construction and development areas of the future city, which not only objectively meets the requirements of urban construction suitability and land carrying capacity but also better brings into play the agglomeration effect of urban space. While there are not any rigid requirements or instructions in the planning that directly affect the city’s shape, the layout of planning areas and infrastructure, like putting together centralized urban construction areas, does have some effect on the direction, location, and size of urban growth. This is performed so that the city’s final shape is more in line with the real needs of the regional development. However, overly strict boundary control makes it inevitable that cities will occupy a certain amount of farmland, forest land, and grassland in their development. Although it does not break the bottom line of ecological and farmland protection, it will still have certain negative impacts on the ecological environment and food security. And since Datong County is still in the primary stage of urban development, the demand for new urban construction land is high. Despite planning restrictions on the extent of new construction land, the pace and intensity of new urban growth remain elevated, while the utilization of existing stock land remains insufficient. In summary, national spatial planning has positively influenced the optimization of regional spatial patterns.

5.3. Inspiration for the Future Development of Datong County

In the introduction to the research area, we indicated that Datong County is responsible for ecological and protection, food security, and regional socioeconomic growth. Therefore, in order to further co-ordinate the two major aspects of development and conservation, the objectives and positioning of the future development of the area should be rationally determined in the first place when the planning is formulated. This positioning should be jointly proposed from spatial, economic, environmental, ecological, social, and cultural considerations. In terms of development, the development pattern of different areas should be rationally defined using spatial boundaries through the delineation of functional zones such as ecological conservation zones, food-growing zones and urban development zones. And the relationship between development and conservation can be further coordinated through the constraints of specific quantitative indicators, such as the scale of farmland protection area and the scale of new construction land. Furthermore, initiatives such as ecological protection grants and construction land use quotas can be utilized to encourage areas to prioritize environmental conservation while also improving land development and utilization efficiency.

5.4. Optimized Adjustment Paths for Spatial Planning

From the experimental results, it is not difficult to see that the problems existing in territorial spatial planning mainly focus on the following two aspects. First, the planning boundary is too rigid, which has certain negative impacts on the protection of arable land. The second is that the improvement of land development efficiency is not obvious, and there is still room for increasing the level of saving and intensification of land use. For better control of boundaries, planning has made it clearer where and how cities can grow. To meet the needs of cities developing, more farmland, forest land, and grassland will have to be taken over. This will probably lead to conflicts between construction land, basic farmland, and the ecological protection red line. Therefore, more flexible blank space should be reserved in the planning to reconcile the contradiction between urban development, farmland, and ecological protection [55,56]. A lot of important steps and schedules for protecting farmland can also be seen in spatial planning. These include reclaiming abandoned mine lands, stopping people from illegally living on farmland, integrating, and improving village layouts, and improving the quality of farmland [57]. The aim is to achieve double improvement in the quantity and quality of farmland protection and to dissipate the impacts of urban development on farmland.

As for the improvement of land use efficiency, it is too simple to rely solely on boundary control and target setting. In spatial planning, it is still necessary to support relevant policy initiatives and implementation paths to promote the realization of planning objectives. For example, a comprehensive guarantee mechanism of “Integrated spatial planning + Accurate rationing of land use quotas” could be adopted [58,59]. A spatial security system for important land use projects could be established to guarantee the demand for essential land use in diverse specialized sectors in a systematic and classified approach. The structure, scale, and schedule of land supply should be systematically planned, and a framework could be established for the coordinated distribution of land use quotas by county and for interregional mobility. Improving the allocation of new construction quotas with “basic quota + transfer quota” is also necessary. The basic quotas are based on the basic idea that people and land should each have the right role. The transfer quotas make it easier for counties to coordinate and control time and space, especially when it comes to grants and other help for developing suitable areas. For the construction needs of major infrastructure projects or other key projects, they can apply for transfer quotas upon approval by the county government. We could also improve the mechanism for revitalizing the stock of inefficient construction land connected to the increase in stock. According to the scale of stock revitalization in the previous year in each region, the amount of new construction land transfer quotas could be approved in a reasonable proportion of stock and increment. For regions where the disposal of stock, idleness,3 and inefficiency has not been completed, the basic quotas will be reduced proportionately, and the transfer quotas will be launched in a prudent manner [60]. Through the parallel cooperation of various initiatives, high-quality development and protection of land space under the guidance of planning will eventually be realized.

6. Conclusions

This study focuses on the scientific quantification of planning elements by means of rational land use change simulation. Through the dynamic simulation of the territorial spatial pattern for the target planning year, we construct an evaluation indicator system to calculate the scores of each dimension of planning in order to realize the pre-assessment of its implementation effects. Historical data reveal a clear positive or negative correlation between the evolution of different types of land use and driving factor distance. The probability of land use change, obtained by coupling the distance factor with historical data, demonstrates high simulation accuracy. This suggests that it is feasible to quantify the planning elements under the guidance of planning to simulate the evolution of spatial patterns of land use. In terms of the practical effects of planning, the most prominent effect of territorial spatial planning on regional development is to limit the boundaries of urban development and optimize urban form. In addition, it has had some positive effects on reducing urban carbon emissions and other aspects. However, from the perspective of urban development, overly rigid planning boundary control will still have some negative impact on farmland protection and the ecological environment. In particular, the amount of farmland occupation around the city is still at a high level. As the city is still in the stage of rapid development, it is imperative to enhance the effective development and utilization of land resources through boundary control, indicator regulation, and other planning strategies that require further optimization.

For the planning pre-evaluation framework designed in this study, the framework enables the evaluation of planning effects based on the simulation of territorial spatial evolution. The evaluation of spatial dimensions, such as urban expansion speed and urban landscape patterns, demonstrates a high degree of accuracy. However, the evaluation framework is more suitable for macro-scale and overall regional spatial pattern related evaluations. Relatively microscopic planning objectives and indicators within cities, such as easing traffic congestion and improving infrastructure, cannot be accurately evaluated using this approach. In addition, the quantitative approach to planning elements proposed in this study relies primarily on historical regional evolution data and planning data for mathematical calculations.

In terms of scenario setting, since the planning used in this study is comprehensive planning, its influence on spatial evolution of the national territory is shaped by a combination of spatial boundaries, quantitative development objectives, and other planning elements. As a result, it is not feasible for us to isolate specific elements that promote socioeconomic development or environmental protection. This is the specific reason why we did not set up more scenarios for comparative analysis. However, according to China’s spatial planning system, under the guidance of territorial spatial planning, specialized planning and detailed planning will be prepared for specific aspects such as industrial development and environmental protection. Therefore, in the future, further research can incorporate these specialized plans to analyze the specific impacts of different planning measures on regional development and environmental protection, allowing for further optimization of the planning pre-assessment framework. In addition, the completeness of research data directly affects assessment accuracy. Therefore, whether it is possible to find a universal method of quantification and simulation of planning elements to achieve fast and accurate evaluation based on existing data is the next research direction to focus on.