Land Use Conflict Under Different Scenarios Based on the PLUS Model: A Case Study of the Development Pilot Zone in Jilin, China

Abstract

In rapidly urbanizing regions, escalating land use conflicts have raised concerns over sustainable development and ecological security. This study focuses on the Chang-Ji-Tu Development and Opening Pilot Zone in Jilin Province, aiming to reveal the spatiotemporal evolution of land use conflicts and identify their driving factors, based on land use data from 2000 to 2023. The study employs land use data, the PLUS model, SCCI, and the geographic detector to analyze conflict dynamics and influencing factors. Cropland and forest land have steadily declined, while construction land has expanded. Conflicts exhibit a spatial gradient of “western pressure, central alleviation, and eastern stability,” with hotspots in Changchun, Jilin, and Yanji. Conflict evolution is categorized into three phases: intensification (2000–2010), peak (2010–2015), and mitigation (2015–2023), as shaped by industrialization and later policy interventions. Among four simulated scenarios, the Sustainable Development (SD) scenario most effectively postpones conflict escalation. Population density and DEM emerged as dominant driving factors. Natural factors have greater explanatory power for land use conflicts than do socio-economic or locational factors. Strengthening spatial planning coordination and refining conflict governance are key to balancing human–environment interactions in the region.

1. Introduction

With the rapid advancement of global urbanization and industrialization, coupled with the evolving concept of ecological civilization, land use conflicts—regarded as a critical issue within human–environment systems—have increasingly emerged as a major obstacle to regional sustainable development, significantly affecting progress toward regional sustainability [1,2,3]. Theoretical understanding of these conflicts has continuously deepened. The scope of academic inquiry has expanded from initial concerns over ownership disputes and confrontational dynamics to broader systemic competitions and contentious interactions among multiple stakeholders. These conflicts arise from the scarcity of land resources, the multifunctionality of land, and divergent objectives in land use approaches, scales, and configurations [4,5]. The issues involve a wide range of stakeholders, including governments, corporations, and farmers, as well as complex dimensions such as ownership tensions, functional trade-offs, and spatial competitiveness, all of which are continually shaped by policy frameworks and socio-economic dynamics [6]. A central consensus in the field holds that “conflicts originate from the divergence of interests among stakeholders regarding land resources.” In recent years, the development of the “functional land use conflict” framework has further advanced this understanding, emphasizing the interconnected processes of land supply, benefit distribution, and stakeholder strategies, offering a novel perspective for conflict resolution [7].

“Land use conflict” is a multidisciplinary concept first introduced by Bryant and Bailey in 1997 within the fields of geography and resource management, and used to describe conflicts arising from competition for spatial resources among different land use demands [8]. As research has progressed, the concept has gradually acquired multidimensional connotations. From a natural-science perspective, Liu et al. (2017) define land use conflict as spatial overlap and mutual exclusion between land use types caused by functional incompatibility or disruption of ecological processes [9]. In studies related to management and planning, land use conflict emphasizes institutional tensions among resource allocation, land use regulation, and spatial governance. Within the frameworks of political ecology and conflict sociology, Robbins (2019) argues that land use conflict reflects power asymmetry, interest struggles, and deficiencies in environmental justice, highlighting the socially constructed nature of conflicts and the mechanisms of stakeholder participation [10]. Conflicts often emerge among stakeholders during land use processes. Taking farmland protection as an example, government policies typically prioritize food security and ecological boundaries, aiming to preserve both the quantity and the quality of cropland to safeguard farmers’ interests [11]. However, for many farmers, cropland serves not only as a livelihood foundation but also as a key asset for economic returns. In the context of urban expansion and industrial transformation, some farmers prefer to convert cropland into non-agricultural land to achieve higher economic benefits, which contradicts governmental protection goals [12]. Therefore, farmland protection must strike a balance between national interests and individual rights. Overall, land use conflict analysis encompasses both functionalist and technical rationalist approaches, focusing on conflict identification, quantitative assessment, and optimization strategies at the operational level. It also involves the interactions among social structures, power relations, and various actors. However, this study adopts a landscape ecology perspective and narrowly defines “land use conflict” as structural dissonance in the spatial distribution of different land use types, emphasizing boundary compression and spatial tension within landscape patterns, without delving into broader social conflict mechanisms.

The term “landscape,” also an interdisciplinary concept, has continuously evolved over the past fifty years across disciplines such as ecology, geography, and landscape ecology. Troll (1966) was the first to define it as a surface complex shaped by the combined actions of natural and human forces [13]. Forman and Godron (1986) introduced the patch–corridor–matrix model, highlighting the spatial structures of landscapes [14]. In 2001, Turner et al. defined landscape as a spatially heterogeneous area composed of multiple ecosystems, characterized by three key attributes: structure, function, and dynamic processes [15]. In summary, landscapes are broadly recognized as complex systems formed by the interactions of various natural and anthropogenic elements within a defined spatial scale and exhibiting features of spatial heterogeneity and ecological functionality. In this study, the concept of “landscape” is interpreted with a specific emphasis on its characteristic of “spatial heterogeneity,” and is narrowly defined as the spatial structural patterns formed by different land use types—cropland, forest land, grassland, other land, water bodies, and settlements—which are used to analyze land use conflicts and the evolution of landscape patterns.

The diversity of research methods has been observed to increase. Land use conflict analysis has evolved from qualitative descriptions to quantitative spatial analyses and multi-scenario simulations, encompassing scales that range from national to local levels, including metropolitan regions, river basins, provincial areas, cities, counties, towns, and grid structures [16,17]. Comprehensive evaluation approaches that employ multiple indices—such as the Pressure–State–Response (PSR) model—facilitate the establishment of macro-level evaluation frameworks; however, their spatial accuracy is constrained by reliance on statistical data and the application of subjective index weighting [18]. Landscape pattern indices utilize the “vulnerability–stability–complexity” model for quantification but provide insufficient analysis of socio-economic drivers [19]. Geographic detectors excel in discerning interactions between natural and policy factors; however, they are not well-equipped for forecasting future developments [20]. Conventional simulation models like CA-Markov face challenges due to inflexible transition rules and are poorly suited for adaptable policies. In contrast, the FLUS model demonstrates limited precision in simulating complex spatial dynamics such as those involved in cropland conversion into settlements and forest fragmentation [19].

To address the aforementioned limitations, this study employs the PLUS model as its primary analytical tool, supplemented by geographic detectors for auxiliary analysis. The PLUS model integrates the Land Expansion Analysis Strategy (LEAS) with a multi-type random patch seed Cellular Automata mechanism (CARS) [21,22], introducing a “patch generation mechanism” and factor analysis based on random forests. This framework supports multi-scenario simulations and policy intervention analyses, thereby enhancing its ability to capture spatial heterogeneity and policy sensitivity regarding land use conflicts. Consequently, it improves adaptability to flexible policies while effectively overcoming the rigidity inherent in traditional Cellular Automata models. Furthermore, the PLUS model demonstrates superior capability in generating the irregular patch shapes that more accurately reflect real-world scenarios. In contrast to the conventional FLUS model—which relies exclusively on suitability probabilities—the PLUS model significantly enhances simulation accuracy by extracting driving rules from historical land-expansion data. Additionally, it can simulate the natural expansion processes of land use patches, ensuring continuity and integrity at the boundaries between built-up areas and ecological zones. This makes it particularly suitable for transition areas characterized by complex terrain and ecological sensitivity—especially relevant in contexts such as urban expansion, land use conflicts, and ecological protection efforts. The PLUS model adeptly manages boundary changes among various land use types—including urban growth, ecological red lines, and comprehensive land use planning—thereby addressing the issues related to weak spatial expression found in the FLUS model. Moreover, high-resolution raster data generated by the PLUS model can directly facilitate spatial calculations of conflict indices while seamlessly integrating with other modeling frameworks. This integration fosters a “simulation–assessment–optimization” closed loop [23,24]. The geographic detector is adept at analyzing the explanatory power and interaction degrees of various driving factors in the context of existing land use conflicts, thereby elucidating the influence exerted by each factor [25].

The PLUS model, when applied to simulate real-world scenarios, is based on several key assumptions. These include the premise that land use change adheres to specific spatial expansion patterns and is influenced by a variety of driving factors such as economic development, population density, and topography. Furthermore, the PLUS model posits a nonlinear relationship between these driving factors and land use, which can be effectively modeled using machine learning algorithms [26]. Additionally, it assumes that land use expansion occurs progressively and is subject to particular policy or ecological constraints.

However, the PLUS model itself has certain limitations. Its simulation results are sensitive to parameter settings; improper configuration of these parameters may lead to significant errors, which necessitates a thorough tuning during the setup process. In addition to parameter configuration, the spatial resolution and quality of the data play crucial roles in determining the accuracy of simulation outcomes. Insufficient spatial resolution or low-quality data can result in precision errors. Furthermore, the assumptions underlying the model impose constraints when simulating dynamic changes within ecosystems and nonlinear feedback mechanisms, potentially leading to inaccuracies in modeling complex processes. The model primarily emphasizes spatial pattern evolution, but lacks the capability to integrate dynamic feedback processes from ecosystems, hydrology, and other influencing factors. Additionally, it exhibits limited responsiveness to sudden policy changes or unforeseen human behaviors. The construction of simulated scenarios also introduces an element of subjectivity. These factors collectively impact both the stability and the broader applicability of the model’s predictions, indicating that a degree of uncertainty remains inherent within the PLUS model. Therefore, it is essential to combine the PLUS model with other process models for validating simulation results, thereby enhancing both comprehensiveness and reliability in simulations.

Despite significant advancements, current research on land use conflicts continues to exhibit several notable limitations. Firstly, the exploration of spatial differentiation mechanisms within complex terrain areas—such as mountain-to-hill transition zones—and the stringent constraints that such terrains impose on conflict diffusion remain insufficiently addressed. The majority of studies tend to focus predominantly on plains regions or coastal urbanized regions [27,28,29]. Secondly, some investigations prioritize merely describing existing patterns or assessing conditions at a single time point, thereby neglecting a comprehensive analysis of the long-term and multi-staged evolution of conflicts, along with their changing driving forces, particularly the impacts stemming from critical policy shifts [30,31]. Thirdly, scenario simulations frequently overlook the significance of policy intervention factors; they predominantly rely on simplistic adjustments to land use transfer probabilities while disregarding how policies influence land use dynamics [32,33]. Consequently, outcomes from multi-scenario simulations typically emphasize changes in land use quantity and structure but provide minimal detailed analysis regarding the evolution of spatial patterns resulting from these alterations. As a result, there is inadequate development of spatially differentiated governance strategies based on such simulations, which limits the practical applicability of simulation results for optimizing land use.

To address these issues, this study investigates the Chang-Ji-Tu Development and Opening Pilot Zone in Jilin Province, China. As a core area for international cooperation in Northeast Asia and a representative terrain transition zone, it serves as an ideal case for exploring the dynamics of land use conflicts. The objectives of this research are to elucidate the gradient patterns of land use conflicts within this unique region, analyze their stage-wise evolution as influenced by policy interventions, and develop an innovative simulation framework aimed at optimizing the spatial patterns of such conflicts. This research integrates the PLUS model, SCCI, and geographic detectors to examine the spatiotemporal evolution of land use conflicts in the Chang-Ji-Tu area from 2000 to 2023 at a grid scale. Furthermore, it assesses the driving factors behind these conflicts in 2023 and simulates future trends under four policy scenarios for 2030: Natural Development (ND), Cultivated Land Protection (CP), Economic Development (ED), and Sustainable Development (SD). The study identifies key characteristics and transition patterns of the land use conflicts prevalent in economically developed areas within China’s major grain production zones. Additionally, it proposes strategic pathways for achieving a balance between economic growth, ecological preservation, and cropland protection. The findings provide theoretical insights into land use optimization applicable to similar regions across Northeast Asia while supporting a shift from “passive control” to “proactive transformation” in governance related to land conflict management.

2. Study Area Overview, Data Sources, and Research Methods

2.1. Study Area Overview

The Chang-Ji-Tu Development and Opening Pilot Zone (hereinafter referred to as “Chang-Ji-Tu”) is situated in northeastern China, covering an area of approximately 70,000 km². Geographically, it spans longitudes 124°33′36″ E to 131°08′24″ E and latitudes 42°12′00″ N to 45°01′05″ N. The region is characterized by a temperate continental monsoon climate, with distinct seasonal variations and a summer pattern of concurrent heat and rainfall (Figure 1). The Chang-Ji-Tu area includes parts of Changchun City (central urban area, Jiutai District, Dehui City, and Nong’an County), Jilin City (central urban area, Jiaohe City, and Yongji County), as well as the Yanbian Korean Autonomous Prefecture. It has a population of approximately 7.7 million. Despite accounting for about one-third of Jilin Province’s land area and population, the region contributes nearly half of the province’s total economic output. As a central node in China’s Tumen River Area Regional Cooperation and Development Strategy, it holds significant strategic importance. According to the China Tumen River Area Regional Cooperation and Development Planning Outline (2009–2020), Chang-Ji-Tu was designated to become a key economic growth center in Northeast China by 2020. It was envisioned as a modern industrial base, a demonstration zone for advanced agriculture, a hub for science and technology innovation, a logistics center, and an international business service base in Northeast Asia. Land use in Chang-Ji-Tu is predominantly composed of cropland and forest land. However, rapid economic development and accelerating urbanization have led to continuous urban expansion, intensifying land use conflicts. The high proportion of cropland, coupled with the increasing demand for high-quality land for urban construction, has heightened tensions over increasingly scarce land resources. These land use conflicts now pose significant constraints on regional sustainable development. Therefore, the selection of Chang-Ji-Tu as the study area provides a representative and typical case for analyzing land use conflicts in rapidly urbanizing regions.

2.2. Data Sources

This study covers the period from 2000 to 2023, utilizing six sets of land use remote-sensing image data with a spatial resolution of 30 m. These datasets encompass a total of 2,010,511,647 grid cells. Based on the GB/T 21010–2017; Land Use Current Status Classification standard. Publisher: Jilin Province, China, 2017, land use is categorized into six major classes: cropland, forest land, grassland, water bodies, settlements, and other land.

The primary source of land use data is the Resource and Environmental Science Data Center of the Chinese Academy of Sciences (https://www.resdc.cn (accessed on 20 April 2025)). The datasets are derived from Landsat satellite imagery, processed using ENVI 5.6, and interpreted through manual interactive visual interpretation at a spatial resolution of 30 m, ensuring high accuracy and reliability. To support land use simulation, this study also incorporates natural geographic data, socio-economic data, and development restriction datasets. Socio-economic indicators such as population and GDP are sourced from the Resource and Environmental Science Data Center, and the spatial resolution is 1 km. Road network data is obtained from the National Geographic Information Resources Catalog Service System (https://www.webmap.cn (accessed on 20 April 2025)) with a data accuracy of 10 m. Natural condition data, including temperature and precipitation, with a data accuracy of 0.01°, are retrieved from the National Earth System Science Data Center. Elevation and slope data, along with soil type data, with a data accuracy of 30 m, are provided by the Resource and Environmental Science Data Center. Development restriction data, including nature reserve boundaries, is sourced from the Chinese Academy of Sciences, and the spatial resolution is 1 km. River systems data, with a data accuracy of 10 m, is sourced from the National Geographic Information Resources Catalog Service System. Soil data is based on 1995 records, while the nature reserve dataset reflects the status in 2018. The remaining data values are taken for the years 2000, 2005, 2010, 2015, 2020, and 2023. All raster data were resampled to a 100 m resolution prior to being imported into the PLUS model (

Table 1. Data types and sources.

Date Type	Date	Date Source
Land Use Data	Land Use Data	Resource and Environmental Science Data Center, Chinese Academy of Sciences
Socio-economic Factors	Population Density	Resource and Environmental Science Data Center, Chinese Academy of Sciences
	GDP
Distance Factors	Distance to Main Roads	National Geographic Information Resources Catalog Service System
	Distance to Secondary Roads
	Distance to Tertiary Roads
	Distance to Rivers
	Distance to Railways
	Distance to Nature Reserves	Resource and Environmental Science Data Center, Chinese Academy of Sciences
Natural Factors	Soil Type	Resource and Environmental Science Data Center, Chinese Academy of Sciences
	NDVI
	DEM
	Slope	Derived from DEM
	Annual Average Temperature	National Earth System Science Data Center
	Annual Precipitation

).

2.3. Methodology

2.3.1. PLUS Model

• Core algorithm of the PLUS model;

The PLUS (Patch-generating Land Use Simulation) model is a land use change simulation framework developed by the High-Performance Spatial Computing and Intelligent Laboratory at China University of Geosciences (Wuhan) (https://github.com/HPSCIL (accessed on 25 April 2025)). This model integrates a rule-mining module based on the Land Expansion Analysis Strategy (LEAS) with a Cellular Automata (CA) simulation component driven by multi-type random patch seeds (CARS). It enables the effective identification of driving forces behind land expansion and landscape transformation and has been extensively applied in land use simulation research [34].

2.

Scenarios setting

The multi-scenario design aims to simulate land use change trends under different future development pathways, enabling a comprehensive analysis of the spatiotemporal evolution of land use and the dynamics of land use conflicts in the Harbin–Changchun urban agglomeration under various projected scenarios [30,31]. In this study, land use patterns in the Chang-Ji-Tu region are simulated based on the Tumen River Regional Cooperation and Development Plan (2009–2020) and the 14th Five-Year Plan for National Economic and Social Development, along with the Vision 2035 Outline for Jilin Province. Four scenarios are established: the Natural Development Scenario, the Cultivated Land Protection Scenario, the Economic Development Scenario, and the Sustainable Development Scenario. The detailed descriptions of these scenarios are as follows:

① Scenario A: Natural Development Scenario (ND)

This scenario continues the historical trend of land use change from 2010 to 2023, without modifying any model parameters or incorporating planning policies that could constrain land use transitions. Based on a 10-year projection interval, the PLUS model, integrated with the Markov Chain method, is employed to estimate land use demand for the year 2030. Key parameters such as land-expansion capacity, land use transition matrix, domain factor weights, and transition probabilities among land use types are kept consistent with those of the 2010–2020 baseline model [34].

② Scenario B: Cultivated Land Protection Scenario (CP)

The “Jilin Province Land Spatial Planning (2021–2035)” proposes to continuously promote the construction of high-standard cropland and the conservation of black soil resources. Under this policy constraint, a Cultivated Land Protection scenario is established. This scenario emphasizes the protection of stable and high-quality cropland in the Chang-Ji-Tu region. Cropland layers from the years 2000, 2010, and 2020 are overlaid to identify long-term stable cropland. Based on the Regulations for Agricultural Land Classification and previous studies, croplands with slopes of less than 6° are categorized as high-quality cropland. These two categories are combined to delineate a restricted conversion zone. Compared to the Natural Development Scenario, the land use transition matrix and transition probabilities among land use types are modified. Specifically, the probability of converting forest land and grassland into settlements is reduced by 20%, while their conversion into cropland is increased by 30%. Furthermore, the probability of converting cropland into settlements is reduced by 60%. These adjustments ensure that high-quality cropland is preserved and protected in alignment with national land management policies.

③ Scenario C: Economic Development Scenario (ED)

The “Jilin Province Land Spatial Planning (2021–2035)” emphasizes strengthening regional open cooperation and enhancing spatial strategic guidance. It advocates for the full utilization of the region’s locational advantages, the deep implementation of the Changchun–Jilin–Tumen Development and Opening-up Strategy, and the expansion of external connectivity. This policy framework provides substantial support for positioning Jilin as a key northern gateway for national openness and a strategic hub for regional cooperation in Northeast Asia. Under this policy guidance, the Changchun–Jilin–Tumen region faces significant pressure from rapid urbanization; therefore, an Economic Development Scenario is established. In this scenario, the region seizes opportunities presented by national development strategies such as the Belt and Road Initiative and the coordinated development of the Yanji–Longjing–Tumen area. According to the Tumen River Regional Cooperation and Development Plan (2009–2020), the region continues to expand its openness and strengthen its role as a key development zone, aiming to establish a core area for international cooperation and an outward-oriented urban economic belt in Northeast China. No restricted conversion zones are designated under this scenario. The probability of converting cropland, forest land, or grassland into settlements is increased by 20%, while the conversion of settlements into other land types is constrained.

④ Scenario D: Sustainable Development Scenario (SD)

The “Jilin Province Land Spatial Planning (2021–2035)” emphasizes the principle of harmonious coexistence between humans and nature, strictly adheres to ecological safety boundaries, and aims to enhance the stability and biodiversity of ecosystems. Based on the region’s actual conditions, it recognizes that economic development and ecological conservation are not mutually exclusive but can be mutually reinforcing. Therefore, guided by the planning objectives, a Sustainable Development Scenario is established. To achieve a balance between economic development and cropland protection, the Sustainable Development Scenario integrates both economic and ecological objectives to guide land use policies toward a sustainable trajectory. In this scenario, the probability of converting cropland and forest land into settlements is reduced by 10%, the probability of converting grassland and water bodies into settlements is reduced by 20%, and the probability of settlements being converted into forest land, grassland, water bodies, or other lands is also reduced by 10%. These adjustments are designed to mitigate land use conflicts, improve the efficiency of land resource allocation, and promote sustainable regional development. The land use transition matrix for the four scenarios is presented in

Table 2. Transition matrix under the four scenarios.

	Scenario ND						Scenario CP						Scenario ED						Scenario SD
	a	b	c	d	e	f	a	b	c	d	e	f	a	b	c	d	e	f	a	b	c	d	e	f
a *	1	1	1	1	1	1	1	0	0	0	0	0	1	1	1	1	1	1	1	0	0	0	1	0
b *	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	0	0	0	0
c *	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0
d *	1	1	1	1	1	1	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0	1	0	0
e *	1	1	1	1	1	1	0	0	0	0	1	0	0	0	0	0	1	0	0	1	1	1	1	0
f *	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1

* a, b, c, d, e and f represent cropland, forest land, grassland, water bodies, settlements, and other land.

.

3.

Determination of neighborhood weight parameters

Neighborhood weight ($W$) is an indicator used to represent the relative strength of each land use type in terms of its capacity to expand or transform into other land use types [18]. Some scholars suggest that changes in the total area ($\Delta TA$) of each land use category over the same time scale can more effectively reflect the intensity of expansion [18]. The dimensionless change in $\Delta TA$ aligns with both the semantic meaning and the structural requirements of the data, thereby fulfilling the parameter criteria for neighborhood weight in the model. Accordingly, the formula used to calculate neighborhood weight in the PLUS model for this study is as follows:

$$W={\displaystyle \frac{\Delta TAi-\Delta TAmin}{\Delta TAmax-\Delta TAmin}}$$

(1)

Here, $\Delta TAi$ represents the amount of $TA$ change for the land type, which is the change in total area for a specific land use type; $\Delta TAmin$ represents the land type with the smallest amount of change; $\Delta TAmax$ represents the land type with the largest amount of change.Based on the total area changes associated with cropland, forest land, grassland, water bodies, settlements, and other lands in the Chang-Ji-Tu region from 2000 to 2023, the corresponding neighborhood weights were calculated. These values, as shown in

Table 3. Neighborhood Weight Parameters for Different Land Use Types.

Type	Cropland	Forest Land	Grassland	Water	Settlements	Other Land
2030ND	0.97	0.01	0.43	0.66	1	0.62
2030CP	0.77	0.01	0.39	0.60	1	0.56
2030ED	1	0.01	0.35	0.49	0.49	0.47
2030SD	1	0.01	0.42	0.65	0.92	0.61

, were used as neighborhood weight parameters in the PLUS model for simulating land use transformation dynamics.

4.

Accuracy Verification

① Kappa Coefficient

The Kappa coefficient is a widely used metric for assessing simulation accuracy. It is calculated using the following formula:

$$Kappa={\displaystyle \frac{P_a-P_b}{1-P_b}}$$

(2)

In this formula, $P_a$ denotes the proportion of correctly simulated grid cells, while $P_b$ represents the proportion of correctly simulated grid cells expected by random chance. $\mathrm{A}$ $Kappa$ value of 1 indicates perfect agreement between the simulated and reference data. The coefficient ranges from 0 to 1, with higher values reflecting greater simulation accuracy [35]. The corresponding interpretation of simulation accuracy levels based on $Kappa$ values is presented in

Table 4. Kappa coefficient classification and evaluation standards.

Kappa Coefficient	<0.00	0.00~0.20	0.21~0.40	0.41~0.60	0.61~0.80	0.81~1.00
Degree	Very Poor	Weak	Low	Moderate	Significant	Best

.

② FOM Coefficient

The Figure of Merit (FoM) coefficient is used to quantitatively evaluate the accuracy of land use simulation at the grid cell scale. A higher FoM value indicates better simulation performance; however, it typically ranges between 0.01 and 0.25 [35]. The formula for calculating the FoM coefficient is as follows:

$$FoM={\displaystyle \frac{B}{A+B+C+D}}$$

(3)

In this equation, $A$ denotes the area of the error region in which land use actually changed but was predicted to remain unchanged; $B$ denotes the area in which the simulation correctly predicted land use change; $C$ refers to the area of misclassified land use change; and $D$ represents the area in which no actual land use change occurred, but the simulation incorrectly predicted a change.

2.3.2. Spatial Comprehensive Conflict Index ($SCCI$)

Landscapes are the direct subjects of human resource development and utilization, making the landscape scale particularly suitable for examining the environmental impacts of human activities. Landscape ecological risk is widely recognized as an effective indicator of absolute land use conflict, and it exhibits a close relationship with land use conflict. The evaluation dimensions of landscape ecological risk and land use conflict are often aligned in the existing research frameworks [36,37,38]. Absolute conflict evaluates the capacity of land systems to withstand external pressure, thereby indicating their overall stability. Its magnitude can be reflected through the level of regional ecological security risk. In this study, the landscape ecological index is adopted as the core indicator of land use conflict. Drawing upon the existing literature, we construct the $SCCI$ to quantify the absolute conflict intensity of regional land use. The $SCCI$ integrates three components—landscape complexity, vulnerability, and stability—into a unified assessment framework. The specific calculation of $SCCI$ is performed on the ArcGIS (Version 10.8) software. The relevant formulas and interpretations are provided below:

(1)

Spatial comprehensive conflict index ($SCCI$)

$$SCCI=CI+FI-SI$$

(4)

where $CI$, $FI$, and $SI$ denote the Complexity Index, Fragility (Vulnerability) Index, and Stability Index, respectively. The final value is normalized to the range (0, 1) using the general normalization formula (Formula (8)).

(2)

Complex index ($CI$)

$$CI={\displaystyle \sum }_{i=1}^n{\displaystyle \sum }_{j=1}^n\left[{\displaystyle \frac{2\ln \left(0.25P_{ij}\right)}{\ln a_{ij}}}{\displaystyle \frac{a_{ij}}{A}}\right]$$

(5)

where $P_{ij}$ is the perimeter of the j-th patch within the i-th land use type, $a_{ij}$ is its corresponding area, and $A$ is the total landscape area. This is equivalent to the Area-Weighted Mean Patch Fractal Dimension (AWMPFD), an index that effectively describes landscape complexity under anthropogenic disturbance. A higher $CI$ value indicates a more fragmented and complex landscape, which is often associated with intensified land use conflicts.

(3)

Fragility index ($FI$)

$$FI={\displaystyle \sum }_{i=1}^n{F}_i\times {\displaystyle \frac{a_i}{S}}$$

(6)

where $F_i$ represents the vulnerability score assigned to a land use type, $a_i$ is the area of that land use type, and $S$ is the total area. Vulnerability reflects the spatial exposure and sensitivity of landscape types to internal and external pressures [39]. A higher $FI$ value indicates greater susceptibility to land use conflicts. Based on the characteristics of land use change in the Chang-Ji-Tu region and previous research, the vulnerability scores assigned to different land use types are as follows: settlements (6), other land (5), water bodies (4), cropland (3), grassland (2), and forest land (1) [40].

(4)

Stability index ($SI$)

$$SI=1-PD=1-{\displaystyle \frac{N_i}{A}}$$

(7)

where $N_i$ is the number of patches in a landscape unit, and $A$ is the area of a landscape unit.

The fragmentation degree of regional landscapes is one of the typical manifestations of land use conflicts. The higher the degree of spatial fragmentation, the lower the landscape stability and the more intense the conflict, because fragmented landscapes indicate high levels of competition between different land use stakeholders [41]. Therefore, patch density ($PD$) is used to negatively reflect the $SI$ of the regional ecosystem. The larger the $PD$ value, the higher the spatial fragmentation degree of the land system, and the lower the stability.

To ensure comparability, all values in Formulas (4)–(7) are normalized to the [0, 1] interval using the following formula:

$$N={\displaystyle \frac{A-A_{min}}{A_{max}-A_{min}}}$$

(8)

where $N$ is the normalized value, $A$ is the original value, $A_{max}$ is the maximum observed value, and $A_{min}$ is the minimum observed value across all samples.

All indicators are standardized through normalization. It is important to note that landscape patterns are highly sensitive to spatial scale. Based on previous research findings [16,17,40], a spatial resolution of 5 km has been identified as the optimal scale for $SCCI$ calculation in this study.

2.3.3. Geographic Detector

The geographic detector is an analytical tool grounded in the theory of spatial stratified heterogeneity [42,43]. It employs spatial statistical principles to assess the driving mechanisms of spatial patterns and consists of four core sub-models: the factor detector, ecological detector, risk detector, and interaction detector. Among these, the factor detector and interaction detector are the most widely used and are applied in this study to explore the influence and interactive effects of various driving factors on land use conflicts in the Chang-Ji-Tu region.

• Factor Detector.

The factor detector is a tool used to analyze the degree of influence of a single influencing factor on the spatial distribution of a specific geographic phenomenon. It quantifies the explanatory power of one independent variable factor for the dependent variable by calculating the factor detector index (q-value) [44,45]. The q-value ranges from 0 to 1, and a higher q-value indicates that the factor has a stronger explanatory power for the spatial distribution of the target phenomenon. This study selected three types of factors, including socio-economic factors, distance factors, and natural factors, along with fourteen driving factors: population density, GDP, distance to main roads, distance to secondary roads, distance to tertiary roads, distance to rivers, distance to railways, distance to nature reserves, soil type, NDVI, DEM, slope, average annual temperature, and annual precipitation. The results of the factor detector can explain the impacts of socio-economic, distance-related, and natural factors on the degree of land-use conflict. The calculation formula is as follows:

$$q=1-{\displaystyle \sum }_{h=1}^L{\displaystyle \frac{N_h{\sigma }_h^2}{N{\sigma }^2}}$$

(9)

In Formula (9), $N_h$ represents the number of samples in the h-th geographic unit, ${\sigma }_h^2$ represents the variance of the h-th geographic unit,$N$ represents the total number of samples, ${\sigma }^2$ represents the variance for the entire study area, and $L$ represents the number of geographic units.

2.

Interaction Detector

The interaction detector is a tool used to evaluate the extent to which the interactions between two or more influencing factors affect the spatial distribution of a specific geographic phenomenon [46]. The interaction detection process first calculates the influence of two independent variables (factors $X_1$ and $X_2$) on the dependent variable, denoted as q($X_1$) and q($X_2$). Then, it calculates the interaction effect of both factors $X_1$ and $X_2$ together, denoted as q($X_1\cap X_2$). Finally, the values of q($X_1$), q($X_2$) and q($X_1\cap X_2$) are compared to assess whether the combined effect of the driving factors strengthens or weakens the explanatory power of the analysis variable.

3.

Optimal Parameter Selection

A critical step in applying the geographic detector is the selection of optimal parameters for spatial discretization and spatial scale determination [47,48]. The q-value derived from the factor detector serves as an indicator of discretization effectiveness. A larger q-value indicates a more effective classification, stronger spatial differentiation of land use conflicts, and the greater explanatory power of the corresponding driving factor. Conversely, a smaller q-value suggests weaker differentiation and less explanatory capacity. When q = 0, it indicates complete spatial homogeneity, and no spatial heterogeneity; when q = 1, it implies perfect spatial heterogeneity. To define the classification scheme for spatial discretization, various methods can be employed, including equal interval, natural breaks, quantile breaks, and geometric interval classification. These approaches assist in determining the interval thresholds for discretizing continuous variables. This study employs the Optimal Parameter GeoDetector (OPGD) model, which selects the best combination of discretization method and spatial scale by maximizing the q-value. This process ensures that the spatial heterogeneity of land use conflicts and the explanatory power of driving factors are accurately identified and optimized.

3. Results Analysis

3.1. Analysis of Land Use Evolution Characteristics in the Chang-Ji-Tu Region from 2000 to 2023

According to the Current Land Use Classification Standard (GB/T 21010–2017), land in the Chang-Ji-Tu region is categorized into six types: cropland, forest land, grassland, water bodies, settlements, and other land. Based on this classification system, the spatial distribution patterns of land use were derived for the years 2000, 2005, 2010, 2015, 2020, and 2023 (Figure 2).

3.1.1. Spatiotemporal Evolution Characteristics of Land Use in the Chang-Ji-Tu Region

As illustrated in Figure 2, forest land and cropland represent the most dominant land use types in the Chang-Ji-Tu region. Cropland is predominantly concentrated in Changchun, followed by the peri-urban areas of Dunhua City and the Yanbian Korean Autonomous Prefecture. Forest land is mainly distributed across the central and eastern parts of the region, particularly in the less-developed zones of Jilin City and Yanbian Prefecture. Grassland exhibits a relatively sparse spatial distribution, whereas water bodies are primarily located along the Tumen and Songhua Rivers, with higher concentrations observed in the mountainous eastern areas of Yanbian Prefecture and the river valley regions between Jilin and Changchun.

Settlements are predominantly concentrated in the central urban areas of Changchun and Jilin, as well as in the city center of Dunhua and the border regions connecting Yanji, Helong, and Longjing. These areas demonstrate a high level of urban agglomeration, identifying them as core development zones.

To further highlight changes among different land use types, the area and proportion of each category were calculated (

Table 5. Land use area and proportion by type in the Chang-Ji-Tu Region (2000–2023) (hm²).

Type Year	2000	2005	2010	2015	2020	2023
Cropland	456,275.3	2,383,533	2,370,565	2,389,942	2,366,291	2,352,584
	35.74%	33.19%	33.01%	33.28%	32.95%	32.76%
Forest Land	764,862.5	4,512,226	4,500,040	4,452,037	4,450,245	4,449,599
	59.92%	62.83%	62.66%	61.99%	61.96%	61.95%
Grassland	914.49	29,631.69	23,569.38	22,664.25	22,762.53	22,884.66
	0.07%	0.41%	0.33%	0.32%	0.32%	0.32%
Water Bodies	23,999.58	66,026.88	69,697.26	64,238.67	67,401.27	70,321.59
	1.88%	0.92%	0.97%	0.89%	0.94%	0.98%
Other Land	30,435.75	187,209.6	2773.35	1625.85	1036.44	640.17
	2.38%	2.61%	0.04%	0.02%	0.01%	0.01%
Settlements	11.25	3575.25	215,558.5	251,694.5	274,466.8	286,173.1
	0.00%	0.05%	3.00%	3.50%	3.82%	3.98%

). The results show that from 2000 to 2023, the proportion of cropland gradually decreased, while forest land exhibited a trend of initial slight increase followed by a decline. Grassland first expanded and then contracted. The proportion of water bodies fluctuated over time but remained relatively stable overall. The share of other lands continuously declined, whereas the proportion of settlements steadily increased, reflecting continuous urban expansion and the transformation of land use patterns in the region.

Figure 2. Land use patterns in the Chang-Ji-Tu region from 2000 to 2023.

Figure 2. Land use patterns in the Chang-Ji-Tu region from 2000 to 2023.

3.1.2. Characteristics of Land Use Transformation in the Chang-Ji-Tu Region

This research makes use of land-use data from 2000, 2010, and 2023 to conduct an investigation into the transformation of land-use patterns in the Chang-Ji-Tu region over the past two decades. Based on the structural changes that took place during the period from 2010 to 2023, a land-use transition Sankey diagram was developed. This diagram serves to visually illustrate the quantity and direction of land-use conversions (Figure 3). Furthermore, land-use transition matrices for the time spans of 2000–2010 and 2010–2023 were constructed. The purpose of these matrices is to quantitatively capture the types and magnitudes of land-use changes (

Table 6. Land use transition matrix from 2000 to 2010 (Unit: km²).

2010	2000
	Cropland	Forest Land	Grassland	Water Bodies	Other Land	Settlements	Total
Cropland	——	976.95	95.54	22.95	2.04	29.48	1126.96
Forest Land	1021.45	——	12.72	5.84	0	0.66	1040.67
Grassland	71.52	2.37	——	1.98	4.97	1.25	82.09
Water Bodies	174.04	2.4	2.66	——	0.72	32.44	212.26
Other Land	0.52	0.01	14.35	0.16	——	0.15	15.19
Settlements	562.13	10.14	27.61	12.57	2.1	——	614.55
Total	1829.66	991.87	152.88	43.52	9.83	63.98	3091.74

and

Table 7. Land use transition matrix from 2010 to 2023 (Unit: km²).

2023	2010
	Cropland	Forest Land	Grassland	Water Bodies	Other Land	Settlements	Total
Cropland	——	1188.18	117.80	54.41	8.33	32.06	1400.77
Forest Land	728.96	——	12.94	3.13	0.01	1.59	746.64
Grassland	106.24	24.28	——	0.89	11.68	1.21	144.29
Water Bodies	57.22	5.71	3.55	——	0.23	28.13	94.84
Other Land	0.48	0.02	1.26	1.18	——	0.20	3.13
Settlements	684.58	31.62	19.07	28.09	4.45	——	767.80
Total	1577.49	1249.80	154.61	87.69	24.69	63.19	3157.47

).

According to the data presented in Table 6, from 2000 to 2010, cropland underwent the most significant conversion to other land use types, with a total area of 1829.66 km² being transformed. The majority of this area was converted into forest land and settlements. The second-largest land use change occurred in forest land, with 991.87 km² converted to other types, primarily to cropland. Overall, during this decade, cropland experienced the greatest reduction; grassland showed a slight decrease; forest land remained largely stable; water bodies increased substantially; other types of lands exhibited a minor increase; and settlements underwent the most pronounced expansion.

These transformations were profoundly influenced by the period of rapid industrialization and urbanization under the national “Revitalization of Northeast China” strategy. (This policy aims to promote economic growth and regional development in northeastern China, including the provinces of Liaoning, Jilin, and Heilongjiang. Once a major industrial base, the region has experienced economic decline, prompting an initiative to rejuvenate industries, enhance infrastructure, and attract investment). The rapid expansion of settlements reflects the region’s strong emphasis on economic growth and served as the primary driver of land use transformation during this period. The substantial net reduction in cropland resulted from both urban and industrial encroachment and the implementation of national ecological policies, such as the Grain-for-Green program. (This initiative encourages farmers to cease cultivation on ecologically sensitive lands, such as hills and slopes, and instead engage in afforestation. Its objective is to restore forest cover, reduce soil erosion, and protect the environment, with farmers receiving compensation for transitioning from agricultural to forestry activities). The relative stability of forest land suggests a dynamic equilibrium between ecological restoration efforts and developmental pressures. The increase in water bodies was primarily attributed to the construction of water infrastructure and ecological engineering projects, reflecting heightened environmental awareness and the initial effectiveness of policy interventions. The slight decline in grassland indicates persistent ecological pressures, while the rise in the collective area of other lands may reflect land disturbances caused by overdevelopment and expansion activities in specific areas.

This ten-year period represented a phase of intensive spatial expansion and heightened development demand in the Chang-Ji-Tu region, driven by the broader national revitalization strategy. The land use changes observed during this time reflect both the large-scale consumption of land resources for economic growth—such as the conversion of cropland into settlements—and the initial achievements of ecological initiatives, including reforestation and wetland conservation. This transformation pattern has presented both challenges and a foundation for future land use planning, ecological protection, and sustainable development in the region.

According to Figure 3 and Table 7, between 2010 and 2023, cropland again experienced the most significant conversion, with 1577.49 km² converted to other land types—primarily into forest land (728.96 km²) and settlements (684.58 km²)—resulting in a net decrease of 176.72 km². Forest land followed, with 1249.80 km² converted to other types, chiefly into cropland (1188.18 km²), leading to a net loss of 503.16 km². Grassland saw 154.61 km² converted, mostly into cropland (54.41 km²) and settlements (28.09 km²), resulting in a net decrease of 10.32 km². Water bodies increased by a net of 7.15 km², with 43.52 km² transferred in and out. Other lands decreased by 21.56 km², with areas converted into cropland (8.33 km²), grassland (11.68 km²), and settlements (4.45 km²). Despite 63.19 km² of settlements being converted back—mainly into cropland (32.06 km²) and water bodies (28.13 km²)—settlements still experienced a net increase of 704.61 km², mostly at the expense of cropland.

During the 2010–2023 period, land use change in the Chang-Ji-Tu region was characterized by a dual trend of development-driven transformation and policy-guided ecological governance. The cropland protection “redline” functioned as a binding constraint, facilitating mutual conversions between cropland and forest land while achieving a basic balance between losses and compensations, thereby effectively limiting the net loss of cropland. The substantial increase in settlements was driven by new urbanization strategies and regional development initiatives, with expansion primarily occurring in central urban areas and largely encroaching on cropland. The conversion of other land types into productive or ecologically managed areas reflects the progress of land consolidation efforts. The slight increase in water bodies, despite the competing pressures from cropland protection and rising demand for settlements, illustrates the implementation of the national “ecological protection first” principle. (This principle emphasizes that environmental protection should be the top priority in development decisions, meaning that before any land is allocated for urban, agricultural, or industrial use, its ecological value and potential environmental impacts must be assessed. The goal is to ensure sustainable development while preserving natural ecosystems). Spatially, land consolidation not only replenished cropland but also optimized its spatial configuration—for example, by converting forest land in central cluster areas into cropland—thus enhancing cropland use efficiency while supporting both food security and sustainable land management.

This thirteen-year period (2010–2023) represented a pivotal phase in which the Chang-Ji-Tu region sought to balance the competing objectives of revitalizing Northeast China, ensuring national food security, and advancing ecological restoration, all within the broader framework of ecological civilization. Land use dynamics during this time revealed the inherent tensions in achieving this balance—for instance, the notable decline in forest land and the rapid expansion of settlements—while also highlighting the positive outcomes of governance measures, such as the reduced cropland loss, the significant decrease in other lands, and the modest increase in water bodies. This transformation pattern introduces new challenges for future territorial spatial planning in the region. The central task ahead is to reconcile the spatial demands of high-quality development with the stringent protection of cropland redlines and the safeguarding of ecological security, particularly the conservation of forest land and water bodies.

3.2. Land Use Prediction Results for the Chang-Ji-Tu Region, Based on the PLUS Model

3.2.1. Accuracy Verification

To ensure the accuracy and practical applicability of this study, model validation was conducted prior to using the PLUS model to project the 2030 land use pattern in the Chang-Ji-Tu region. Specifically, land use data from 2000 and 2010 were utilized to simulate the land use distribution for 2020, and the simulated results were then compared with the actual land use data from that year. To evaluate the reliability of the simulation, two widely recognized performance metrics—the Kappa coefficient and the Figure of Merit (FoM)—were employed. The simulation produced a Kappa value of 0.85 and an FoM value of 0.12, both of which are within acceptable ranges. These results indicate a high level of simulation accuracy and confirm the robustness and reliability of the PLUS model in generating land use projections.

• Kappa Coefficient.

Due to the functional constraints of the PLUS software (Version 1.40), land use simulations require datasets with equal time intervals. As this study aims to project land use patterns for the year 2030, data from 2000 and 2010 were selected to simulate the 2020 land use pattern. The simulated results were then compared with the actual 2020 land use data, using the Kappa coefficient as the accuracy metric. To ensure both representativeness and computational efficiency during the validation process, a 5% sampling rate was applied. Detailed validation results are presented in

Table 8. Confusion matrix of actual vs. predicted land use patterns in 2020.

2010
	Land Type	Cropland	Forest Land	Grassland	Water Bodies	Other Land	Settlements	Total
2020 Predicted	Cropland	96,804	4216	958	437	3166	630	106,211
	Forest Land	10,082	207,751	997	171	312	795	220,108
	Grassland	1457	2103	4772	56	61	407	8856
	Water Bodies	246	93	12	5242	59	27	5679
	Other Land	1245	73	22	29	12,174	8	13,551
	Settlements	721	939	55	104	80	2430	4329
	Total	110,555	215,175	6816	6039	15,852	4297	358,734

.

Based on the analysis of the confusion matrix and calculations derived from the corresponding formulas, the Kappa coefficient was calculated to be 0.85, and the overall accuracy reached 0.92. According to the classification criteria outlined in the referenced table, a Kappa value exceeding 0.8 indicates that the simulation results are highly accurate. Therefore, the model demonstrates robust performance and high reliability in the simulation of land use patterns.

2.

FoM Coefficient.

By comparing the simulated and actual land use patterns for the year 2020 in the study area, the following area values were obtained: A = 334,372 hm², representing areas in which land use actually changed but was incorrectly predicted to have remained unchanged; B = 79,759 hm², representing areas in which the prediction was correct; C = 5786 hm², referring to areas incorrectly classified; and D = 251,651 hm², representing areas in which land use remained unchanged but was incorrectly predicted to have changed.

According to Formula (3), the Figure of Merit (FoM) coefficient was calculated to be 0.12. Given that FoM values generally fall within the acceptable range of 0.01 to 0.25, this result indicates a satisfactory level of simulation accuracy. Therefore, the PLUS model is deemed reliable for simulating land use changes in the Chang-Ji-Tu region.

3.2.2. Land Use Prediction Results Under Four Scenarios in the Chang-Ji-Tu Region

By inputting predefined land demand, land use transition cost matrices, and neighborhood weights, the PLUS model was utilized to simulate land use patterns in the Chang-Ji-Tu region for the year 2030 under four distinct scenarios: Natural Development (ND), Cultivated Land Protection (CP), Economic Development (ED), and Sustainable Development (SD). The resulting simulated land use structure maps are presented in Figure 4.

A comparative analysis of the land use structures across the four scenarios reveals the following patterns:

Under the CP scenario, the extent of cropland is significantly higher than that under the ND scenario, particularly in the southwestern part of the region, where settlements would otherwise tend to expand. Consequently, the area of settlements is relatively smaller under the CP scenario. In the ED scenario, compared to the ND scenario, there is a notable expansion of settlements, especially within the southwestern economic development zone and the eastern forest land areas. This reflects intensified urbanization and increased pressure on land conversion in both industrial and ecologically sensitive regions. Under the SD scenario, the overall land use structure remains largely similar to that under the ND scenario. However, a slight increase in forest land and a modest decrease in cropland suggest a more balanced approach between ecological conservation and development.

By comparing the proportions of the six land use types in 2030 under the four simulated development scenarios with the 2020 baseline, a comprehensive analysis of land use area changes was obtained, as presented in

Table 9. Comparison of land use area by type in 2020, and under different development scenarios for 2030 (km²).

Year	Scenario	Land Type	Cropland	Forest Land	Grassland	Water Bodies	Settlements	Year
2020		21,349.17	44,016.5	1761.84	1141.39	2749.41	859.62	21,349.17
2030	ND	22,673.15	42,235.88	1117.41	1276.95	3704.21	870.33	ND
	CP	23,535.34	41,879.3	1068.34	1277.86	3250.35	866.74	CP
	ED	22,531.04	42,221.01	1115.61	1276.38	3864.19	869.7	ED
	SD	22,139.65	43,097.13	1118.39	1229.86	3625.95	666.95	SD
Area Change (2020–2030)	ND	1323.98	−1780.62	−644.43	135.56	954.8	10.71	ND
	CP	2186.17	−2137.2	−693.5	136.47	500.94	7.12	CP
	ED	1181.87	−1795.49	−646.23	134.99	1114.78	10.08	ED
	SD	790.48	−919.37	−643.45	88.47	876.54	−192.67	SD

.

Under the ND scenario, cropland, water bodies, and settlements all exhibit slight increases. In contrast, forest land and grassland show a declining trend, while the area composed of other lands remains essentially unchanged. Most conversions involve forest land and grassland being transformed into cropland and settlements. Cropland experiences the largest expansion, increasing by approximately 1324 km², followed by settlements, which expand by around 955 km². In this context, urbanization and economic growth continue to drive settlement expansion, exerting pressure on forest- and grass-covered areas and posing challenges to ecological security and environmental protection.

Under the CP scenario, cropland, settlements, and water bodies all increase compared to 2020. Cropland shows the most significant growth, expanding by approximately 2186 km², while settlements increase by about 501 km². Meanwhile, forest land and grassland both decrease, with forest land declining more sharply. Most of the forest land and grassland is converted into cropland and settlements, with the area of forest land converted being roughly three times greater than that of the converted grassland. Compared with the ND scenario, cropland expands at a faster rate, while the reductions in forest land and grassland are more pronounced. Settlement expansion slows slightly, and water bodies increase modestly. Although cropland protection policies are strictly enforced, this scenario still results in certain negative ecological consequences due to the accelerated conversion of forest land.

In the ED scenario, cropland, water bodies, settlements, and other lands all increase in area, whereas forest land and grassland decline. Cropland shows the largest expansion, increasing by about 1182 km², followed closely by settlements, which grow by approximately 1115 km². Forest land experiences the most substantial decline. Forest land and grassland are primarily converted into cropland and settlements, with the area of forest land converted being nearly 2.8 times greater than that of the converted grassland. Compared with the ND scenario, cropland expands at a slower pace, but forest land loss becomes more severe, and the decrease in grassland is slightly larger. Settlements expand more rapidly, while water bodies experience a slight reduction. This pattern reflects a policy focus on economic development, which intensifies land consumption and ecological pressure.

Under the SD scenario, cropland, water bodies, and settlements all increase relative to 2020. Settlements undergo the most notable growth, expanding by approximately 877 km², followed by cropland, which increases by around 790 km². Meanwhile, forest land, grassland, and other lands all decrease, with forest land showing the greatest reduction. Forest land and grassland are mainly converted into settlements and cropland, with the area of forest land converted being approximately 1.4 times greater than that of grassland. Compared with the ND scenario, the rate of cropland expansion slows, the decline in forest land is less severe, the reduction in grassland is comparable, settlement expansion is more moderate, and water bodies fluctuate slightly. The collective area of other lands decreases significantly. This scenario balances development needs and ecological protection, fully reflecting the principles of sustainable development. It supports environmental improvement and contributes to long-term land use equilibrium.

3.3. Land Use Conflicts in the Chang-Ji-Tu Region (2000–2023)

Spatiotemporal Evolutionary Characteristics of Land Use Conflicts (2000–2023)

At the grid scale, the cumulative frequency curve of the land use conflict index displays an inverted “U” shape, reflecting the evolutionary trend of spatial land use conflicts. Based on this pattern and using the natural breaks method in ArcGIS, the degree of land use conflict in the Chang-Ji-Tu region from 2000 to 2023 is categorized into four levels: Level I: Severe Conflict: [0.55, 1.00); Level II: Moderate Conflict: [0.35, 0.55); Level III: Mild Conflict: [0.15, 0.35); and Level IV: Low Conflict: (0.00, 0.15) (Figure 5 and Figure 6).

Overall, at the grid scale, land use conflicts in the Chang-Ji-Tu region from 2000 to 2023 exhibited a spatial gradient characterized by “high pressure in the west, moderate conditions in the center, and stability in the east.” The region has experienced relatively severe conflict conditions, with areas classified as severe conflict zones consistently accounting for over 20% of the total area. These zones are primarily concentrated in and around the central urban areas of Changchun, Jilin, and Yanbian. Despite minor fluctuations in the proportions of different conflict levels over time, the overall intensity of land use conflict has shown a slight decline, and this has been accompanied by a modest reduction in the extent of severe conflict zones (Figure 5). In the core urban areas of Changchun, Jilin, and Yanbian, high population density, concentrated industrial activity, and expanding urban footprints have intensified spatial competition between settlements and permanent basic cropland. Consequently, the degree of spatial aggregation of land use conflicts has been increasing.

According to the histogram data (Figure 5), between 2000 and 2023, the number of severe and moderate conflict zones in the Chang-Ji-Tu region experienced fluctuations, albeit with minimal variation in amplitude. This suggests overall stability with a slight downward trend. Similarly, the number of mild conflict zones has remained relatively stable over time. In contrast, the number of low conflict zones has shown a gradual increase despite minor fluctuations. Collectively, the land use conflict pattern in the region remains significant and persistent, reflecting a stable yet entrenched trend that calls for targeted mitigation strategies and coordinated management efforts.

The areas experiencing the most pronounced changes are concentrated in the central portion of the Chang-Ji-Tu region, particularly within the central urban area of Jilin City, as well as in Jiaohe and Dunhua (Figure 6). These areas have undergone relatively structured expansion along transportation corridor nodes, resulting in a significant reduction in land use conflict intensity. In several locations, zones previously classified as severe conflict areas have transitioned into moderate conflict zones. This transformation can be attributed to localized interventions such as the relocation of traditional industries from city centers, the implementation of hillside cluster development strategies, and the promotion of circular economic models. Additionally, ecological restoration efforts along the Songhua River and the revitalization of resource-dependent cities have played a crucial role in alleviating land use conflicts, facilitating the shift from high-intensity pressures to more moderate land use pressures.

In the western part of the region—specifically in the urban agglomerations of Changchun, Nong’an County, Dehui City, and Jiutai District—land use conflicts are most acute. The western region is predominantly characterized by severe and moderate conflict zones, with both the extent and the intensity of the severe conflict areas showing a gradual increase. As the economic core of the region, Changchun contributes over 50% of Jilin Province’s GDP and is under increasing developmental pressure. The city’s reliance on land-based finance and an output-driven land supply system has weakened the effectiveness of urban growth boundary controls. Consequently, Changchun, Nong’an, and Dehui have experienced persistently high and intensifying land use conflicts. This escalation is further fueled by the concentration of national-level strategic platforms. Zones such as the Changchun New Area and Xinglong Comprehensive Bonded Zone have driven large-scale industrial development, particularly in the automobile manufacturing and biopharmaceutical sectors. Concurrently, efforts to establish a transportation hub in Northeast Asia have led to rapid settlement expansion into Jiutai and Nong’an, often encroaching upon high-quality black soil cropland. This trend has exacerbated land use conflict intensity and contributed to the ongoing spread of severe conflict zones across the western region.

In contrast, land use conflicts in the eastern part of the region, particularly in Yanbian Prefecture, have remained relatively stable. Severe conflict zones are concentrated in cities such as Yanji, Helong, Longjing, and Tumen, although the intensity of conflict in these areas has decreased slightly in recent years. Yanbian Prefecture has controlled development intensity by implementing functional zoning of forest land areas in the Changbai Mountain region and enforcing regulatory constraints. At the same time, the construction of open-access corridors and the “Borrowing Ports to Access the Sea” strategy have facilitated the development of an export-oriented economy. The region has also promoted integrated urban cluster development through the Yanji–Longjing–Tumen city group and leveraged tourism—particularly border cultural and ecological tourism—to reduce reliance on spatial expansion. Although cities such as Yanji and Helong remain conflict hotspots, the strategic shift toward a corridor-based economy, as opposed to land-intensive development, has contributed to an overall stabilization and modest reduction in land use conflicts.

3.4. Land Use Conflict Results in the Chang-Ji-Tu Region Under Different Development Scenarios

Based on the size of the study area and informed by previous research findings [16,17], a 5 km grid was selected as the analytical scale. This resolution effectively balances the need to capture spatial heterogeneity in land use conditions with the requirement for computational efficiency. In this study, simulations of the spatial distribution patterns of land use conflicts in the Chang-Ji-Tu region under different future development scenarios were conducted at the 5 km grid scale. The results indicate that the overall spatial pattern of land use conflicts in 2030—across all scenarios—remains largely consistent with the pattern observed from 2000 to 2023, preserving the characteristic gradient of “high pressure in the west, moderate conditions in the center, and stability in the east.” However, the severity of conflicts is projected to increase significantly. Notably, there are evident signs of conflict intensification and spatial degradation, suggesting that the challenges associated with land use conflicts are becoming increasingly severe and urgent (Figure 7).

In 2030, under the different development scenarios, low-conflict land use areas are primarily located in non-central development zones and their surrounding regions within Jilin City, Dunhua, and Yanbian Prefecture. These areas are predominantly covered by forest land—in Yanbian, they are mainly distributed across steep mountainous and valley terrain, while in Jilin, they are found in low mountains, hills, and river valleys. Severe land use conflict zones remain concentrated in the central urban areas and adjacent settlements of Changchun, Jilin, and Yanbian. The situation is particularly severe in Changchun, where nearly the entire area has transitioned into a moderate to severe conflict zone. This spatial pattern underscores a strong overlap between land use conflicts and socio-economic development hubs, with the spatial diffusion of conflicts constrained by topographic features. Consequently, conflicts have intensified in central urban plains but have not extended into higher-altitude regions. While land use conflicts are fundamentally driven by competition among stakeholders, natural barriers such as elevation and terrain inhibit their expansion into mountainous zones. Instead, conflicts tend to spread into adjacent, flatter areas near the original high-conflict zones.

Based on the simulated land use conflict patterns in 2030 across various development scenarios, the Chang-Ji-Tu region exhibits clear differences in both spatial distribution and severity of conflict. Under the ND scenario—which follows the historical trajectory of socio-economic development—land use conflict displays a distinct spatial pattern. Low-conflict areas gradually extend outward, although their total area decreases. At the same time, severe conflict zones also expand outward and increase in area, while moderate conflict zones contract and mild conflict zones experience a slight expansion. As economic and social systems continue to evolve alongside historical inertia under the ND scenario (

Table 10. Land use structure in 2030 under different development scenarios (%).

Land Use	2020	ND	CP	ED	SD
Cropland	29.70	31.55	32.74	31.35	30.80
Forest Land	61.24	58.76	58.26	58.74	59.96
Grassland	2.45	1.55	1.49	1.55	1.56
Water Bodies	1.59	1.78	1.78	1.78	1.71
Settlements	3.83	5.15	4.52	5.38	5.04
Other Land	1.20	1.21	1.21	1.21	0.93

), the polarization of development centers becomes more pronounced. Constrained by regional topography, severe conflict zones expand slowly from core areas toward the periphery. However, increased landscape vulnerability contributes to higher conflict intensity. This trend is particularly evident in the central urban and peripheral zones of Changchun, Jilin, and Yanbian. In Changchun—located in the western part of the region—rapid economic development, supported by national policy, has further intensified land use conflict, with conflict levels rising steadily. Overall, the ND scenario reflects a persistent and escalating pattern of land use conflict characterized by the gradual outward expansion of both low-conflict and high-conflict zones (Figure 7).

Under the CP scenario, there is a noticeable expansion in both the extent and the number of low-conflict land use zones, particularly in the western part of the region, such as within Changchun. Compared to the ND scenario, the area of severe conflict zones in Changchun is reduced in both quantity and intensity. However, in other regional development hubs—including Jilin City, Dunhua, and Yanji, along with their surrounding areas—the severe conflict zones expand in both spatial extent and number. This contrast reflects differing development priorities among cities. While Changchun’s focus on rapid economic growth benefits from the alleviation of land use pressures through cropland protection policies, other urban centers may encounter emerging tensions between ecological conservation and agricultural expansion. For instance, regions with a higher proportion of forest land are more likely to experience intensified land use conflicts under cropland protection measures, as efforts to preserve farmland may encroach upon forested areas. This trade-off increases spatial competition and aggravates land use conflict in ecologically sensitive zones.

Under the ED scenario, the evolution of land use conflicts closely mirrors that of the ND scenario, indicating a strong alignment between the region’s current socio-economic trajectory and an economy-centered development strategy. This similarity suggests that the ongoing growth in the Chang-Ji-Tu region is predominantly driven by economic expansion and often occurring at the expense of ecological protection. Consequently, land use conflicts are projected to intensify further, posing significant risks to both long-term ecological sustainability and the stability of socio-economic development.

In contrast, under the SD scenario, a more balanced approach is adopted—one that simultaneously integrates economic growth, ecological conservation, and cropland protection. Consequently, the number of severe, moderate, and mild land use conflict zones all decrease, while low-conflict zones expand. This shift demonstrates the effectiveness of the SD scenario in curbing the expansion of high-conflict areas and reducing overall conflict intensity. Ultimately, it facilitates the formation of a more stable and sustainable spatial development framework for the region.

3.5. Analysis of the Land Use Conflict Driving Factors in the Chang-Ji-Tu Region Based on the Optimal Parameter Geographic Detector

A single-factor detection analysis of land use conflicts in the Chang-Ji-Tu region for the year 2023 was conducted using the Optimal Parameter Geographic Detector model (

Table 11. Geographic detector results and classification of land use conflict driving factors in the Chang-Ji-Tu Region, 2023.

Factor Classification	Factor Detector:	Variable	qv	p	Discretization Method	Level
Economic and Social	Population Density	X1	0.62889974	0.00	Quantile	4
	GDP	X2	0.3177068	0.00	Equal	2
Distance	Dis to Primary Roads	X3	0.27500837	0.00	Quantile	6
	Dis to Secondary Roads	X4	0.31513733	0.00	Quantile	6
	Dis to Secondary Roads	X5	0.27129711	0.00	sd	6
	Dis to River	X6	0.09912196	0.00	natural	6
	Dis to Railway	X7	0.27235302	0.00	Quantile	6
	Dis to Reserve	X8	0.26092267	0.00	Quantile	6
Natural	Soil Type	X9	0.34395547	0.00	Quantile	6
	NDVI	X10	0.39706256	0.00	sd	5
	DEM	X11	0.61758434	0.00	sd	6
	Slope	X12	0.25579578	0.00	Quantile	5
	Tem	X13	0.40122987	0.00	Quantile	6
	Rain	X14	0.49290027	0.00	Quantile	5

). Quantile break classification was applied to discretize the explanatory variables, with most variables categorized into five to six classes. The results indicate that the factors with the highest explanatory power—represented by the highest q-values—are population density and DEM, suggesting that both human population distribution and topographic variation are key determinants in shaping the spatial pattern of land use conflicts. Among socio-economic variables, GDP also exhibits strong explanatory capacity. Within the category of distance-based factors, the explanatory power is relatively balanced. Specifically, distances to major roads, secondary roads, tertiary roads, railways, and nature reserves all demonstrate notable influence, while distance to rivers shows the weakest effect. This implies that land use conflicts in the region are only weakly associated with river systems and that regional development is less reliant on water networks. Regarding natural environmental factors, average annual temperature and annual precipitation exhibit strong explanatory power, whereas slope has a relatively limited influence. Overall, population density and DEM emerge as the most influential drivers, highlighting their critical roles in the emergence and intensification of land use conflicts. Among socio-economic factors, both population density and GDP significantly influence the spatial distribution and intensity of conflicts, reflecting the substantial impact of demographic and economic dynamics. Given that the Chang-Ji-Tu region encompasses both plains and mountainous areas and contains a high proportion of cropland and forest land, natural factors such as elevation, temperature, precipitation, NDVI, soil type, and slope demonstrate moderate to high explanatory strength.

In contrast, distance-based factors generally demonstrate lower explanatory power compared to socio-economic and natural variables. Among these, distance to rivers exhibits the weakest influence, further supporting the conclusion that river systems play a minimal role in shaping land use conflicts in the region. Meanwhile, distances to roads and nature reserves show moderate and comparable levels of influence, indicating that land use conflicts are more likely to concentrate within urban settlements, as well as in transitional zones where urban, agricultural, and ecological spaces intersect. These areas represent the primary hotspots for land use conflicts in the Chang-Ji-Tu region.

Based on the interaction detection results for land use conflicts in the Chang-Ji-Tu region in 2023, a hotspot interaction map was generated (Figure 8). The results show that most interactions between influencing factors exhibit a bi-factor enhancement effect, indicating that the combined influence of two variables is greater than their individual effects. Among these, interactions involving population density, DEM, average annual temperature, and annual precipitation demonstrate particularly high explanatory power, with population density being the most significant. While population density alone already exhibits strong explanatory strength, its interaction with other variables further enhances this effect. This highlights that population density is a core driver of current land use conflicts and is likely to remain a dominant influence in future conflict dynamics. Similarly, DEM, temperature, and precipitation also show strong interaction effects with other variables, underscoring the significant roles that natural geographic conditions play in shaping the spatial distribution of land use conflicts. This is consistent with the land use composition of the Chang-Ji-Tu region, in which cropland and forest land dominate the landscape. These natural factors exert considerable influence on both the formation and the intensity of land use conflicts, particularly in areas where agricultural and ecological values overlap. In contrast, distance-related variables, such as distances to roads, railways, and nature reserves, exhibit moderate interaction effects and possess intermediate explanatory power when considered individually. This indicates that while these distance factors contribute to conflict formation and spatial differentiation, they do not serve as primary drivers. This finding aligns with the region’s land use structure, in which settlements and urbanized areas are relatively limited in scale, and land development pressures are more localized. Consequently, the impact of proximity-based variables is spatially constrained and less pronounced, compared to demographic or natural factors.

In summary, natural factors and population density have been identified as the most influential drivers of land use conflicts in the Chang-Ji-Tu region. Natural variables primarily influence conflicts associated with cropland and forest land, particularly in areas where ecological and agricultural interests overlap. Meanwhile, socio-economic factors—especially population density—have a significant impact on both the intensity and the spatial distribution of conflicts. Distance-based factors are more influential in urbanized zones, although their overall effect is secondary and largely confined to specific localized contexts.

4. Strategies for Managing Land Use Conflicts in the Chang-Ji-Tu Region

To evaluate the effectiveness of the mitigation of land use conflicts under different development scenarios, the proportions of land areas under various conflict levels were calculated for the period from 2000 to 2023, as well as projected values for 2030 under each scenario. Based on these data, trend graphs were generated to illustrate the proportional changes in land classified as experiencing severe conflict and the combined value for lands associated with severe and moderate conflict, as well as the overall conflict intensity across the scenarios (Figure 9). The corresponding yearly proportions for each conflict level are summarized in

Table 12. Proportional analysis of land use conflict by severity level, for 2030, under different development scenarios (%).

Year	2000	2005	2010	2015	2020	2023	2030
Scenario							ND	CP	ED	SD
I	21.92	24.97	21.86	24.72	22.55	19.65	31.27	32.56	30.67	30.10
II	28.96	26.32	27.89	26.13	28.02	29.37	24.57	22.67	25.99	22.16
III	17.45	16.42	16.73	16.89	17.74	17.39	22.76	18.59	22.79	19.92
IV	31.67	32.30	33.52	32.26	31.70	33.58	21.40	26.18	20.55	27.82

.

From 2000 to 2023, the average proportion of severe land use conflict areas in the Chang-Ji-Tu region was approximately 22.6% and showed periodic fluctuations. Although the share declined slightly from 21.9% in 2000 to 19.7% in 2023, the long-term average remained above the 2000 baseline, indicating a gradual increase in conflict severity over time. These areas have expanded primarily within the core development zones of major cities, reflecting growing tensions between human activities and land availability. A turning point occurred around 2015, when the extent of the severe conflict zones began to decrease. This change was closely linked to a series of policy interventions, including the implementation of the “Three Zones and Three Lines” delineation, the adoption of ecological-priority land use strategies, spatial layout optimization through regional development planning, industrial restructuring, and more intensive land utilization policies. (The “Three Zones and Three Lines” delineation refers to a land use planning strategy in China that divides land into three zones and draws three lines to guide development. Three Zones: Development Zone (for urbanization and infrastructure), Ecological Protection Zone (for conservation), and the Agricultural Production Zone (for farming). Three Lines: Ecological Red Line (areas that must be protected for environmental reasons), Agricultural Protection Line (areas essential for food production), and the Urban Expansion Line (limits the areas that can be developed for cities)). In addition, ecological restoration projects have gradually taken effect. These coordinated measures have contributed to the alleviation of land use conflicts. Fundamentally, this improvement reflects the combined effects of strict regulatory planning, ecological restoration priorities, spatial functional reconfiguration, and more efficient land resource allocation. Looking ahead to the 2030 simulation projections, all development scenarios indicate a resurgence in severe land use conflicts, though to varying degrees. Both the ND and the SD scenarios follow the historical trajectory, marked by continued outward expansion of severe conflict zones. The CP scenario results in the highest share of severe conflict areas, making it the most intense among the projections. In contrast, the SD scenario shows the slowest increase in severe conflict zones and the lowest overall share, demonstrating its relative effectiveness in mitigating conflict intensity.

From 2000 to 2023, the combined proportions of severe and moderate land use conflict zones remained relatively stable, with only minor fluctuations, and with a slight decrease around 2010. The simulation results suggest that all four development scenarios project continued increases in these zones by 2030, primarily driven by the expansion of severe conflict areas. The ED scenario results in the most pronounced intensification, as it prioritizes rapid economic growth while placing less emphasis on ecological protection. Urban expansion, the outward spread of settlements, and the migration of rural labor into cities contribute to de-agriculturalization and the conversion of cropland to non-agricultural uses. These changes reduce land use efficiency and intensify conflict. The ND scenario exhibits a lower rate of escalation, reflecting the region’s historical development trajectory. The CP scenario performs slightly better than ND by safeguarding high-quality cropland, thereby moderately alleviating conflict intensity. However, the SD scenario proves to be the most effective in reducing overall land use conflict, due to its integrated approach that balances economic development, ecological restoration, and farmland preservation.

Between 2000 and 2023, the proportion of general land use conflict zones in the Chang-Ji-Tu region exhibited a gradual upward trend with minor fluctuations. The average proportion of low-conflict areas was approximately 32.50%, an increase from 31.67% in 2000, indicating a steady, albeit modest, improvement. This trend suggests that regional development adjustments, structural optimization of land use and the economy, and ecological restoration initiatives have had positive impacts on the mitigation of land use conflicts. The 2030 simulation results project continued growth in general conflict zones under all scenarios. The ED and ND scenarios yield nearly identical outcomes, while the SD scenario results in a slightly lower proportion. The CP scenario produces the smallest share of general conflict areas, reflecting the varying effects of different development priorities. Scenarios with more balanced or conservation-oriented strategies demonstrate greater potential in reducing the spatial extent of moderate land use tensions.

Overall, the evolution of land use conflict across the four development scenarios demonstrates varying degrees of mitigation and intensification. Among them, the SD scenario—balancing economic growth with ecological protection and cropland preservation—proves to be the most favorable. It effectively reduces conflict intensity while promoting coordinated outcomes across economic, environmental, and agricultural dimensions. This scenario maximizes regional social welfare and provides the most sustainable development pathway for the Chang-Ji-Tu region. Looking ahead, further research is needed to explore innovative models that can more comprehensively and sustainably resolve tensions relating to human use of the land and support coordinated spatial governance to alleviate regional land use conflicts.

5. Discussion

In terms of land use simulation methodology, the PLUS model has demonstrated significant effectiveness in supporting multi-scenario projections [49,50]. This effectiveness is primarily due to its integrated LEAS rule-mining framework and the CARS patch-generation algorithm, both of which enhance the model’s capacity to interpret the driving forces behind land expansion [34,51]. The accuracy of the simulation conducted in this study was evaluated, returning a Kappa coefficient of 0.85 and a Figure of Merit (FoM) value of 0.12. These metrics meet the reliability criteria established in the mainstream research, confirming that the model accurately captures the spatial and temporal patterns of land use change within the Chang-Ji-Tu region. The Spatial Conflict Coordination Index (SCCI), which incorporates dimensions such as landscape complexity, vulnerability, and stability, provides a robust and comprehensive framework for quantifying the intensity of land use conflicts. Furthermore, the geographic detector method was employed to assess both the individual effects and the interactive effects of multiple driving factors on land use conflict. Together, these analytical tools offer a holistic understanding of the factors influencing conflict distribution and intensification [52,53].

Nevertheless, the model still presents three primary limitations that warrant further consideration. First, although a 5 km grid scale achieves optimal performance in capturing the general spatial patterns of land use conflict and demonstrates a satisfactory model fit, it may compromise the detection of localized conflict details—such as the fragmentation of cropland along urban peripheries. This limitation may lead to the underrepresentation of micro-scale conflict phenomena. Therefore, if the objective is to emphasize fine-grained spatial conflicts and land fragmentation issues, further refinement and validation of the grid resolution are required. This could involve adopting a smaller grid size and managing larger data volumes to enhance the model’s precision in simulating land use conflicts. The key challenge lies in identifying a spatial resolution that maximizes information retention while minimizing computational costs and avoiding spatial distortion. Second, the model currently lacks region-specific differentiation in the detection of driving factors across areas with varying developmental characteristics. Although the Changchun–Jilin–Tumen region generally exhibits a degree of homogeneity in natural, social, and economic conditions—resulting in strong correlations among driving factors—subregional differences in development priorities remain notable. If the model aggregates administrative units with agriculture-oriented development together with those focused on industrial or economic growth when assessing the influence of driving factors on land use conflict, it risks introducing analytical bias. This approach may obscure the distinct ways in which various driving forces contribute to land use conflicts under different developmental contexts, potentially leading to significant discrepancies between simulated and actual outcomes. To address this issue, future improvements could involve partitioning the study area into zones based on dominant development types—such as distinguishing agricultural zones from industrial zones—and conducting separate analyses of the weights assigned to driving factors. This would enhance the alignment between model outputs and real-world conditions. Third, the model does not currently incorporate dynamic socio-economic feedback mechanisms, such as population migration patterns. This limitation may introduce forecasting inaccuracies in the projection of settlement expansion under the economic development (ED) scenario. To mitigate this issue, it would be beneficial to integrate dynamic datasets—such as trends in population movement and shifts in industrial spatial configurations—into the model framework. These variables could be introduced through time series analysis or predictive modeling techniques to enhance the internal logic and temporal realism of scenario simulations. Collectively, these limitations highlight the need for future modeling efforts to incorporate dynamic digital technologies aimed at strengthening the representation of human–environment feedback mechanisms and improving the model’s robustness and applicability.

Focusing on the period from 2000 to 2023, the land use dynamics in the Chang-Ji-Tu region have been characterized by a consistent decline in cropland and the rigid expansion of settlements. This transformation has been primarily driven by rapid urbanization and industrialization, resulting in a significant increase in settlement areas at the expense of agricultural and ecological land. The proportion of cropland decreased from 35.74% to 32.76%, largely due to urban encroachment. Concurrently, forest land underwent a bidirectional transformation influenced by evolving policy priorities: the early implementation of the “Grain-to-Green” program contributed to a modest increase in forest cover, whereas post-2010 cropland protection policies reversed this trend, leading to the conversion of forest land back into cropland and resulting in a substantial net reduction in forest resources.

These land use changes reveal two fundamental structural challenges. First, a significant spatial imbalance exists, as settlements are predominantly concentrated in central urban areas such as Changchun and Jilin, thereby exacerbating regional development disparities. Second, latent land use conflicts are becoming increasingly evident. The “cropland occupation–compensation balance” strategy has largely depended on the conversion of forest land to maintain cropland quantity targets, often at the expense of ecological integrity. Consequently, land use conflicts in the region exhibit pronounced spatiotemporal differentiation. Spatially, a distinct gradient is observed: high conflict pressure in the west (Changchun), moderate alleviation in the central region (Jilin), and relative stability in the east (Yanbian). Severe conflicts are primarily concentrated in urban cores and their surrounding suburban belts, where direct competition between settlements and high-quality black soil cropland is most acute. Temporally, conflict intensity has shown an overall upward trend. The average proportion of severe conflict zones increased to 22.6%, surpassing the 21.9% recorded in 2000. This trend is primarily driven by the outward expansion of settlements due to socio-economic growth. Although the implementation of policies such as the “Three Zones and Three Lines” framework and various ecological restoration initiatives after 2015 contributed to localized mitigation—temporarily reducing the proportion of severe conflict zones to 19.7%—land use governance continues to be constrained by structural contradictions. Specifically, land allocation remains influenced by an economy-centered supply mechanism and insufficient interdepartmental coordination, both of which undermine the long-term effectiveness of conflict mitigation strategies. Natural constraints also significantly shape land use conflicts. Topographic barriers restrict development in the lowland plains, where conflict zones tend to remain spatially persistent yet intensify over time. In contrast, higher-altitude regions exhibit relatively stable or even declining conflict levels, reflecting the limiting influence of natural geographic conditions. Therefore, a critical future challenge lies in how to internalize and effectively address land use conflicts within these lowland plains, given the fixed nature of the natural constraints. Moreover, in scenario-based simulations, the modeling of policy intervention mechanisms remains relatively simplistic. Most existing models rely on basic adjustments to land use transition probabilities and lack integration with spatially detailed regulatory frameworks. This limitation hampers their ability to capture the spatial heterogeneity and enforcement intensity of real-world policy interventions, thereby diminishing the utility of these models in supporting precise and effective land governance. By comparatively analyzing the status of land use conflict research in the Changchun–Jilin–Tumen region and other global regions, it becomes apparent that land use conflict studies outside China predominantly focus on large-scale driving forces such as climate change, global resource distribution, and social stability [20,22,32,33]. This trend is particularly evident in developing countries and regions, where issues such as water resource conflicts and food security often constitute the core of research agendas. In contrast, land use conflict research in China tends to emphasize localized and context-specific challenges, including urban–rural expansion, cropland conservation, and ecological protection. A substantial body of Chinese research literature focuses on the pressures associated with rapid urbanization, particularly concerning the protection of cultivated land and the implementation of ecological redlines—an emphasis that closely aligns with the focus of the present study on cropland preservation under urban expansion pressures [54,55]. In terms of methodological approaches, while traditional statistical methods remain prevalent in China, there is a growing tendency to adopt advanced spatial analysis tools and modeling techniques developed internationally. For instance, this study utilizes the PLUS (Patch-generating Land Use Simulation) model, a spatially explicit simulation tool that exemplifies this evolving methodological direction. Alongside PLUS, other land use change models are increasingly being incorporated into Chinese land system research, reflecting a broader effort to harmonize domestic analytical frameworks with international standards in spatial simulation and scenario-based analysis [56].

Based on the land use conflict simulation results under the four distinct scenarios generated by the PLUS model, the spatial distribution patterns of land use conflicts were systematically analyzed. The study indicates that conflicts are predominantly concentrated in three key areas: urban expansion fringe zones, high-quality cropland regions, and transitional zones between urban development and ecological protection [57]. These spatial tensions are most pronounced in the peri-urban areas of core cities such as Changchun and Jilin, where the encroachment of settlements on cropland and forest land is most severe. Under the Cultivated Land Protection (CP) scenario, it is feasible to effectively safeguard both the quantity and the quality of cropland during future regional development. The probability adjustment strategy for cropland conversion within this scenario aligns closely with the targets outlined in the “Jilin Province Land Spatial Planning (2021–2035),” including maintaining total cropland area above 109.44 million mu, creating cropland of a high standard, and reinforcing the grain security industrial belt. Therefore, the CP scenario provides essential spatial support for the government to delineate stable cropland protection boundaries and identify core zones of high-quality cropland. In agriculture-dominant areas, in which challenges such as population outflow, low land development demand, and land abandonment or inefficient use are prevalent, the study recommends implementing land reclamation and cropland consolidation programs. These initiatives aim to enhance the comprehensive productivity of cropland while promoting large-scale and intensive agricultural practices to strengthen the foundation of food security. In contrast, the Economic Development (ED) scenario promotes future regional economic growth and accelerates urbanization. However, it also results in extensive cropland occupation, ecological degradation, and an intensification of land use conflicts. The ED scenario corresponds with the policy direction of the Changchun–Jilin–Tumen Development and Opening-Up Strategy and the establishment of an outward-oriented economic belt. It facilitates the identification of zones with high urban development potential and supports the rational allocation of resources toward core urban centers.

In rapidly expanding central urban areas, in which high population density and increasing demand for construction land contribute to heightened risks of cropland and ecological land encroachment, spatial conflicts are most intense. To address this challenge, the study recommends promoting compact urban development, encouraging vertical urban expansion and urban renewal, and restricting uncontrolled urban sprawl. It further proposes the establishment of a dual spatial constraint mechanism integrating “ecological boundaries” and “cropland protection redlines” to prevent unauthorized development in cropland and ecological zones. The delineation of urban development boundaries is also recommended, directing the expansion of settlements toward low-conflict areas and thereby alleviating conflict pressures in central urban districts. Building on this foundation, the Sustainable Development (SD) scenario represents an intermediate pathway that seeks to balance economic growth with ecological conservation, aiming to harmonize the development of social, economic, and environmental systems to maximize overall regional benefits. The SD scenario incorporates key policy objectives such as “enhancing ecosystem carrying capacity,” “establishing an ecological protection network,” and “achieving harmonious coexistence between humans and nature.” It reflects a governance strategy that prioritizes ecological integrity while promoting the efficient and intensive use of land resources. In peri-urban transitional zones and urban–rural interface areas, in which land use types are highly mixed and frequent conversions occur between settlements and cropland, conflicts tend to be recurrent, yet exhibit greater spatial flexibility. In response, the study proposes establishing a tiered cropland management system that prioritizes the protection of high-quality permanent basic cropland in areas with dynamic land use transitions. Additionally, it recommends the implementation of regulatory policies governing land use conversion, with clearly defined conditions and procedures to guide such transitions. Low-intensity construction projects that integrate ecological agriculture are encouraged, promoting multifunctional land use while ensuring environmental sustainability. For ecologically sensitive areas, strict compliance with the “Three Zones and Three Lines” regulatory framework—particularly the enforcement of the “ecological conservation redline”—is essential. On this basis, modern technologies such as unmanned aerial vehicles and internet-based monitoring systems can be deployed to enhance the visual supervision and automated inspection of ecological land, thereby improving monitoring efficiency while reducing labor costs. The simulation results under the SD scenario also provide technical support for the implementation of the “multi-plan integration” initiative and the enforcement of the “land use regulation system.” By constructing differentiated conflict governance pathways driven by alternative scenarios, the study offers quantifiable and actionable spatial decision-making frameworks for policymakers. These tools can assist in navigating the spatial tensions between urban expansion and cropland preservation, thereby facilitating territorial development strategies that are more balanced and informed.

The multi-scenario simulations further reveal the distinct impacts of various development pathways on the evolution of land use conflict in the Chang-Ji-Tu region. Under the ND scenario, the continued reduction of forest land, combined with the expansion of settlements, intensifies land use conflict, increasing the proportion of severe conflict zones to 31.27%. In the CP scenario, although the area of cropland expands significantly through the reconversion of forest land into farmland, this expansion occurs at the expense of ecological quality. The resulting land use transformation intensifies latent ecological conflicts, particularly in central areas such as Jilin, where development and conservation interests intersect. The ED scenario produces the most severe outcome. The unregulated expansion of settlements under this scenario leads to excessive resource consumption and intense spatial competition, elevating the proportion of severe conflict zones to 30.67%, the highest among all scenarios. In contrast, the SD scenario imposes stricter constraints on the probabilities associated with the conversion of cropland and forest land into settlements, facilitating the more balanced development of both agricultural and urban land uses. This scenario also promotes the ecological restoration of other lands, encouraging their reallocation for environmental functions. Consequently, the proportion of severe conflict zones is effectively held at 30.10%, making the SD scenario the most effective pathway for conflict mitigation. These findings indicate that single-objective policies, such as those emphasized in the CP or ED scenarios, tend to exacerbate spatial conflicts due to their failure to account for systemic interdependencies. In contrast, the integrated implementation of economic intensification, ecological restoration, and cropland quality improvement offers a more comprehensive strategy for addressing land use conflicts at their root. Only by balancing development and conservation priorities can long-term spatial sustainability be achieved in regions facing complex land use pressures.

At its core, land use conflict in the Chang-Ji-Tu region constitutes a tripartite tension between economic development, ecological conservation, and food security. To effectively mitigate the pressures between human activities and land resources, it is essential to move beyond traditional path dependencies and explore innovative, context-specific development models. On one hand, strategies should be adapted to local natural conditions and stages of socio-economic development, promoting regionally differentiated compensation mechanisms that enable each area to capitalize on its comparative advantages. In high-pressure zones such as the urban core of Changchun, increasing the intensity and efficiency of settlement use can help conserve horizontal land resources, while the establishment of interregional ecological compensation markets can contribute to spatial equilibrium. In ecologically rich areas like Yanbian Prefecture, development should prioritize under-forest economies and pilot mechanisms such as ecological banking, reducing dependence on land expansion and enhancing the value of ecosystem services. On the other hand, an intensity-based, fine-grained governance framework is urgently required. In severe conflict zones such as urban Changchun and Yanji, strict growth boundaries should be imposed on settlement expansion, alongside the permanent protection of high-quality black soil cropland. In moderate conflict zones, such as the outskirts of Jilin and Dunhua, ecological buffer zones can serve to delineate urban–rural boundaries and facilitate rational industrial clustering. In low-conflict areas, such as forest land in Yanbian or the hilly zones of Jiaohe, a conservation-first approach is recommended, leveraging sustainable forestry and moderate ecotourism to ensure long-term ecological stability. To dynamically support these strategies, technological innovation plays a pivotal role. For instance, integrating the PLUS model with real-time remote sensing can enable the development of a conflict early-warning system and policy-response feedback mechanism. In parallel, digital platforms can enhance evidence-based decision-making, supporting more responsive and adaptive land use governance systems.

The challenges encountered by the Chang-Ji-Tu region reflect broader trends in China’s rapid urbanization, and the path toward sustainable development necessitates a strategic transition from “conflict control” to “conflict transformation.” Achieving a stable equilibrium among economic growth, ecological conservation, and food security can only be realized through the coordinated integration of spatial restructuring, institutional innovation, and technological advancement. This comprehensive framework not only presents a viable sustainable development model for Northeast China but also offers actionable insights for other emerging economies grappling with similar pressures arising from urban expansion, agricultural land protection, and environmental sustainability.

6. Conclusions

This study established a land use conflict assessment framework based on landscape pattern metrics and applied grid-based analysis to explore the spatiotemporal evolution of land use conflicts in the Chang-Ji-Tu region from 2000 to 2023. Furthermore, the PLUS model was utilized to simulate conflict evolution trends under four development scenarios for the year 2030. The key findings are summarized as follows:

(1)

Between 2000 and 2023, land use in the Chang-Ji-Tu region displayed a clear overall trend marked by the continuous decline of cropland and the substantial expansion of settlements. The proportion of cropland decreased from 35.74% to 32.76%, primarily due to encroachment by settlements. Forest land initially increased slightly, driven by the implementation of the “Grain-to-Green” policy; however, after 2010, it began to decline as cropland protection policies prompted the reconversion of forest land back into cropland. The share of settlements rose from nearly 0% to 3.98%, with expansion concentrated in central urban areas such as Changchun and Jilin.

(2)

Between 2000 and 2023, land use conflicts in the Chang-Ji-Tu region displayed significant spatiotemporal differentiation. Spatially, a clear gradient pattern emerged, one characterized by high conflict pressure in the west, moderate levels in the central areas, and relatively stable conditions in the east. Core urban areas such as Changchun and Yanji remained major hotspots of severe conflict, while western Yanbian Prefecture experienced generally low conflict levels. Temporally, conflict intensity showed an overall increasing trend. Although the proportion of severe conflict zones decreased from 21.9% in 2000 to 19.65% in 2023, the average share over the entire period reached 22.7%, surpassing the 2000 baseline. This indicates a gradual intensification of land use tensions over time. In 2023, population density and DEM were identified as the most influential driving factors, both in terms of single-factor explanatory power and in interaction with other variables. Natural geographic factors generally exhibited high and stable explanatory strength, followed by socio-economic variables, while distance-based factors had the lowest level of explanatory influence.

(3)

The scenario simulations indicate that development pathways have a substantial impact on land use conflicts. The Natural Development Scenario follows historical trends and exacerbates existing conflict conditions. The Cultivated Land Protection scenario increases cropland area but generates new ecological pressures. The Economic Development scenario results in the most severe conflicts due to uncontrolled expansion of settlements. In contrast, the Sustainable Development Scenario achieves a more balanced approach relative to conflicts between economic growth and environmental conservation, effectively reducing severe conflicts while improving the overall land use structure.

(4)

Based on the above findings, the Sustainable Development Scenario is recommended as the most viable pathway. By integrating economic development, ecological conservation, and cropland protection, it reduces the proportion of severe conflict zones to 30.10% by 2030, which is significantly lower than the values under the ND (31.27%) and ED (30.67%) scenarios. The SD scenario also achieves the slowest rate of conflict increase and the smallest share of moderate-to-severe conflict zones and demonstrates notable success in reclaiming and transforming other lands. These outcomes are well-aligned with the national “ecological priority” strategy.