Multi-Scenario Dynamic Simulation of Urban Agglomeration Development on the Northern Slope of the Tianshan Mountains in Xinjiang, China, with the Goal of High-Quality Urban Construction

Abstract

The construction of high-quality urban agglomeration has become a guiding strategy for future urban development. Based on the current development status of urban agglomeration on the northern slope of the Tianshan Mountains, the concepts of environmental protection, harmonious coexistence, and sustainable development were combined in the present study. Land cover data for 2010 and 2020 as well as data on various driving factors and limiting factors were selected to simulate and forecast the land change of the urban agglomeration under environmental constraints. At the same time, to simulate the natural development scenario, farmland protection scenario, and ecological protection scenario for the land development of urban agglomeration on the northern slope of the Tianshan Mountains in 2030, the future land use simulation and Markov (FLUS-Markov) model and the urban growth boundary (UGB) model were combined. The following conclusions may be drawn from the results. (1) Using the land cover in 2010 to simulate the land cover in 2020, the kappa value was 0.724, the overall accuracy was 82.9%, and the FOM value was 0.245, exhibiting a high accuracy. (2) Under the three scenarios, the degree of expansion varied significantly from 2020 to 2030, but the proportion of construction area remained stable at 3%. Under the natural development scenario, urban land expansion was the most obvious, followed by the farmland protection scenario, while under the ecological protection scenario, construction land expansion was the least obvious. (3) Under the three scenarios, the expansion of construction land was mainly dominated by the encroachment of grassland, and the edge expansion mode was characterized by concentrated contiguous land. (4) The kernel density results show that the urban area exhibited a year-by-year expansion, and the best suitable development area was the surrounding farmland. (5) Under the three scenarios, the delineation of UGB in urban agglomeration at the northern slope of the Tianshan Mountains was reasonable and effective, and it can provide a relevant reference for the government’s future urban development and layout planning.

1. Introduction

The world’s population is growing, and human activities are mainly concentrated in cities. It has been predicted that the world urban population will exceed 5 billion by 2030 [1], which will undoubtedly cause a series of problems. The Pearl River Delta is a testament to the rapid rise in Chinese cities, while intensive economic development and urban sprawl pose a significant threat to the surrounding environment [2]. Strengthening regional development planning and pursuing positive economic, social, and environmental linkages between cities, peri-urban areas, and rural areas are vital objectives for sustainable development [3].

As urbanization continues to progress, China’s urbanization rate has reached close to 60% [4]. Most cities have transitioned to higher-quality development, and the main driving force of future urban planning will be to seek high-quality urban construction concepts. An urban agglomeration is a highly developed form of integrated urban space. Cities are highly integrated with urban agglomerations, and are significant vehicles for global economic development [5,6]. In response to China’s long-standing problem of a lack of precision and difficulty in highlighting the characteristics of urban agglomerations through comprehensive policies, a “four special + comprehensive” approach was proposed in the Fourteenth Five-Year Plan period. The development of urban agglomerations is therefore of great scientific significance [7]. A vibrant city needs to be well managed and consistently improved, and attention needs to be paid to the balance between economic development and the capture of environmental resources. High-quality urban agglomeration is a kind of urban development model with a benign interaction that needs not only to coordinate the relationship with the environment but also to promote regional economic development in the process of urban agglomeration development. Thus, urban sustainability is a notable concern in city development [8,9,10,11].

The overall urban sustainability in China is increasing, but the regional distribution is inconsistent, and the sustainability in western China is poor [12]. Building a beautiful new China has become the current development focus in China. In this process, a new era of ecological social development characterized by “respecting and protecting nature” must be established [13]. The urban agglomeration on the northern slope of the Tianshan Mountains, as the key road of the Northwest Silk Road, has facilitated the cultural exchange between the East and the West and the circulation of goods. Despite such cultural exchange, the arid climate environment and rapid economic growth have increased the level of environmental stress on urban development [14,15,16]. Several scholars have explored the relevant factors affecting urban land cover development [17,18,19]; urban planning and policy orientation are dominant factors in sustainable urban construction [20,21]. Urban expansion leads to significant changes in inland cover in metropolitan areas, and the growth of marginal cities is a significant factor in the whole region [22]. In recent years, scenario analysis has been extensively adopted in the process of future land changes. Dynamic simulations of future land with different constraints can be effectively utilized to analyze future land changes and the optimal allocation of resources. As an example, the function of ecological services is significantly involved in the rational layout or planning of urban land and the optimization of UGS (urban green space) [23]. The intervention of different policy scenarios has an obvious positive effect on sustainable urban development [24]. By using the optimized Markov-FLUS model, the simulated multi-scenario results can be coupled with production, living, and ecological demands to optimize spatial pattern and resource allocation [25].

Common land future simulation and prediction software include FLUS, SD, CA, PLUS, CLUE-S, and others [26,27,28,29]. SD is a feedback mechanism model based on the mutual causality of the internal components of a system [30]. The CA model is a model that simulates the spatiotemporal evolution of complex systems [31,32]. In the FLUS model, the SD model and CA model are combined, and ANN (artificial neural network) is further introduced, thereby allowing for human and natural effects to be coupled in order to meet future land demands [33,34]. The combination of the FLUS and CA models can allow for a high accuracy to be obtained in future land cover simulations [35]. At present, the FLUS model is widely used in land expansion prediction and optimization strategies [36,37], and the combination of the FLUS model with other models has been extensively adopted. By introducing spatial autocorrelation factors into an ANN to establish the FLUS model, the probability distribution of local classes has also obtained good prediction ability [38]. Urban growth boundaries (UGBs) are also used to coordinate urban development, manage regional urban sprawl, and optimize the future land demand of a country [39]. The demarcation of urban development boundaries based on ecological protection and cultivated land protection can meet the requirements of permanent cultivated land and ecological land protection, while also reflecting the status of urban development [40].

In the present study, the concepts of ecological development, sustainable development and human–land coordination were combined, and the FLUS-Markov model, UGB model, and nuclear density analysis were introduced to simulate and forecast the land use and spatial and temporal dynamics of the urban cluster on the northern slope of the Tianshan Mountains in the arid zone of Xinjiang, China, in 2030. The simulations were performed under the natural development scenario, arable land conservation scenario, and ecological conservation scenario, and the urban land development and development boundary delineation of the urban cluster on the northern slope of Tianshan Mountains under the three scenarios were compared and analyzed.

The overview of the study area and data sources of the present study are discussed in the second part, and the methodology is introduced in the third part. The focus of most previous research methods has been on single simulation prediction, rather than a combination with simulation prediction, urban land space-time comparative analysis, or urban development boundary delimitation methods. As such, in the present study, the FLUS-Markov model was used to predict different scenarios, while also combining Kernel density analysis to compare the spatial and temporal changes of urban land, and using the UGB model to delineate urban boundaries. In the fourth part, the results and analysis are presented. There is a scarcity of simulations of arid regions and the instability factors there interfere with the simulation results; thus, the main focus of traditional multi-scenario simulations has been on the economically developed regions in eastern China. As such, the changes in urban land use in the urban agglomeration on the northern slope of the Tianshan Mountains in Xinjiang, China, in 2030 under multiple scenarios were explored by combining the models, and the kernel density was used to compare the urban land use between 2010 and 2030. The spatial and temporal dynamics of urban land use from 2010 to 2030 were compared using the kernel density, and the urban boundaries from 2010 to 2030 were compared with the boundaries of urban land use in 2030 under multiple scenarios. In the fifth section, the reliability and limitations of the model are discussed, a comparison with previous studies is provided, and the basic concept of UGB delineation is given. In the present study, values and concepts for guiding cities to follow the concept of sustainable development in the process of urban construction were proposed. The simulated changes in the future urban land status can clearly show the direction of urban development, which can provide a reference for the local government’s future urban planning decisions and can also provide a reference for the urban development of similar regions.

2. Materials and Methods

2.1. Research Area

The urban agglomeration on the northern slope of the Tianshan Mountains are located in the Junggar Basin to the north of the Tianshan Mountains and south of the Altay Mountains in Xinjiang, China, with a longitude of 84°14′ E−89°55′ E and a latitude of 42° 55′ N−46° 15′ N. The northern slopes mainly include includes Urumqi, Karamay, Shihezi, Shawan, Kuitun, Wusu, Changji, Fukang, Wujiaqu, Manas County, Hutubi County, and Jimusar County, with a total area of about 95,000 ${\mathrm{km}}^2$, accounting for about 6.62% of the total area of Xinjiang. The population is approximately over 5 million, accounting for 24% of the total population of Xinjiang. The arid inland climate makes its landform type relatively simple, mainly desert Gobi and oasis, with the highest elevation of 5197 m and the lowest of 165 m. The region is an economic corridor for the development of Xinjiang, and also a significant location for the Belt and Road initiative. The geographical location is of considerable import, of which Urumqi is the political, economic, and cultural center of Xinjiang. The region contains rich land, climate, mineral, and tourism resources, rich products, and a wide variety of cultural landscapes, as shown in Figure 1.

2.2. Data Sources

The land cover data set was mainly taken from GlobeLand 30 with high classification accuracy. Since GlobeLand 30 land data have multiple types, the land cover of multiple types in the research area were redivided into 6 types, including cultivated land, forest land, grassland, water body, construction land, and other land, as shown in Table 1. The data on driving factors were mainly obtained from Geospatial Data Cloud, World Population, Earth Observation, and the National Geomatics Center of China. The data were classified as natural driving factors and social economic driving factors, and the required data could be obtained by the European distance of social economic driving factors in the study area, as shown in Table 2. All data were resampled to 500 m × 500 m separations and then normalized, so as to maintain the uniformity of data.

2.3. Research Method

The FIUS model is used to predict the development of land use, and it usually simulates different scenarios of land cover in combination with the impact of human beings and the natural environment. The model is mainly composed of two modules. The first module is the bottom-up CA model, while the second module is the top-down SD model. The combination of the two modules can simulate the future development of land use. The scenario analysis is mainly divided into two steps. The first step is to obtain a suitable development probability of land prediction through an ANN, and the second step is to simulate and predict different future scenarios based on the land cover and the obtained development probability at time t, as shown in Figure 2.

2.3.1. The Land Type Development Probability Generated by the ANN Model

ANNs are used in machine learning, and mainly consist of three layers: an input layer, a hidden layer, and an output layer. By inputting different types of variables (the variables in the present study were mainly composed of four parts, namely, social and economic factors, natural factors, limiting factors, and land cover dates), transformation among cells of each variable is conducted through hidden layers, and finally the development probability of geo-type cells of each variable is output [41].

(1)$P\left(p,k,t\right)={\displaystyle \sum }_j{w}_{j,k}\times sigmoid\left(ne{t}_j\left(p,t\right)\right)$

where $w_{j,k}$ is the adaptive weight between the hidden layer and the output layer, and the output layer is similar to $w_{i,j}$, which conducts the calibration process during training. After the training and calibration of $w_i$, j, $w_j$, and k, a neural network model is established based on the training data set, which can be used to estimate the probability of each land cover type in a specific grid.

2.3.2. Adaptive Coefficient of Inertia

The adaptive inertia coefficient in the FLUS model is used to adjust the amount of current land cover to make the simulated land development according to the actual demand. To be specific, in one iteration, when there is a large gap between the number of a certain ground class and the expected demand, the change amount of the raster in the next iteration should be increased to ensure that the development trend of the ground class is consistent with the expected trend. The calculation formula is as follows:

(2)$Inerti{a}_k^t=\left\{\begin{array}{ll}Inerti{a}_k^{t-1}& if \left|D_k^{t-1}|\le |D_k^{t-2}\right|\\ Inerti{a}_k^{t-1}\times \frac{D_k^{t-2}}{D_k^{t-1}} & if D_k^{t-1}<D_k^{t-2}<0\\ Inerti{a}_k^{t-1}\times \frac{D_k^{t-1}}{D_k^{t-2}}& if 0<D_k^{t-1}<D_k^{t-1}\end{array}\right\}$

where $Inerti{a}_k^t$ represents the inertia coefficient of land type k at the t iteration, $D_k^{t-1} \mathrm{and} D_k^{t-2}$ represent the difference between the quantity of land type k and the demand at t − 1 and t − 2, respectively.

2.3.3. Neighborhood Effect

Neighborhood influence refers to the interaction between land cover change around the cell and land cover in the cell center, similar to the CA model. In the present study, the 3 × 3 neighborhood influence mechanism was adopted to analyze the influence of land units and urban land units among different types. The formula is as follows:

(3)${\Omega }_{p,k}^t=\frac{{\sum }_{N\times N}con\left(c_p^{t-1}=k\right)}{N\times N-1}\times w_k$

The formula after conversion is as follows:

(4)${\Omega }_{p,k}^t=\frac{{\sum }_{3\times 3}con\left(c_p^{t-1}=k\right)}{3\times 3-1}\times w_k$

where t is the number of iterations; p is a cellular unit; K is the land cover type; ${\Omega }_{p,k}^t$ is the neighborhood action of land type K on cell P at iteration t; ${\sum }_{N\times N}con\left(c_p^{t-1}=k\right)$ is the number of land classes k in the N × N Moore window in t − 1 iterations, with N taken as 3 in the present study; and $W_k$ is the weight of neighborhood effects of different land cover types, with larger $W_k$ values indicating stronger neighborhood effects of land cover types.

2.3.4. Overall Probability of Development

To obtain the overall development probability of each cell land, the overall development probability combines the conversion probability, conversion cost, field influence, and adaptive inertia coefficient of each region, and the calculation formula is as follows:

(5)$T{P}_{p,k}^t=P_{p,k}\times {\Omega }_{p,k}^t\times I{nertia}_k^t\times \left(1-s{c}_{c\to k}\right)$

where $T{P}_{p,k}^t$ is the total probability of transformation from original land type to land type k on cell P at the t iteration; $P_{p,k}$ represents the occurrence probability of land type k on cell P; ${\Omega }_{p,k}^t$ is the neighborhood action of land type k on cell p at t iteration; $Inerti{a}_k^t$ is the inertia coefficient of land type k at t iteration; and $s{c}_c$→k is the cost of converting from original land type c to land type k.

2.3.5. Model Testing

In the present study, the land cover in 2010 was taken as the initial data to verify the accuracy of the land cover prediction data in 2020. The Kappa coefficient and FOM (Figure of Merit) coefficient were used to evaluate the accuracy. The formula is:

(6)$Kappa=\frac{N{\sum }_i^r{x}_{ii}-\sum \left(x_{i+}\cdot x_{+i}\right)}{N^2-\sum \left(x_{i+}\cdot x_{+i}\right)}$

where N is the total number of cells; r is the number of land types; $x_{ii}$ represents the correct number of the ith land type in the simulation results; and $x_{i+}$ and $x_{+i}$ represent the pixel number of the its-land cover type in actual and simulated land patterns, respectively.

(7)$FOM=\frac{Hits}{Hits+Misses+False alarms}\times 100\%$

where Hit means that the model correctly simulates urban growth units, Miss means that the model misses real urban growth units, and False alarm means that the model incorrectly simulates non-urban units as urban units.

2.3.6. Scenario Setting

In the present study, three common scenarios were set up, namely, a natural development scenario, a farmland protection scenario, and an ecological protection scenario. In order to better simulate the urban development of the northern slope of Tianshan Mountain urban agglomeration under different scenarios under the objective of high-quality construction, combined with the relevant national policies on the construction of high-quality urban agglomeration, the concepts of environmental protection, ecological security, and human–land coordination should be considered to the maximum extent in the process of urban development, on the basis of which the relevant norms of land conversion in various regions should be conducted. In the present study, related studies in the literature were referred to [42,43,44,45], and natural development scenarios, farmland protection scenarios, and ecological protection scenarios were set up for each class of meta-cell conversion based on the land development pattern from 2010 to 2020 to predict the development of urban land in 2030.

• (1). The natural development scenario is involved considering less human intervention in land development and letting each land use develop in its natural form.

• (2). The farmland protection scenario involved limiting the encroachment of other land uses on arable land to the maximum extent. The construction of high-quality urban clusters should comply with the important policy that basic farmland is the essence of arable land and the most basic reliance for maintaining national food security. In order to improve the protection of arable land, the urban development of the urban agglomeration on the northern slope of the Tianshan Mountains was reasonably planned and the land expansion pattern in 2030 was simulated.

• (3). The ecological protection scenario involved simulating and forecasting the land expansion in 2030 under the concept of adequate protection of ecological security. Due to the large scope of the study area, in order to obtain better accuracy in the simulation process, the parameters for trial and error needed to be adjusted several times, so as to ultimately obtain simulation results with better accuracy. The transfer cost matrix of its scenario development was shown in Table 3.

2.3.7. Kernel Density Analysis

Kernel density analysis is a nonparametric test method for estimating unknown density functions in probability theory. The spatial properties of data are used to study the distribution characteristics to avoid the error caused by the specific distribution in advance. In the present study, the default bandwidth was used to further analyze the density of the built-up area, and the formula is as follows:

(8)$SearRadius=0.9 \ast min\left(SD,\sqrt{\frac{1}{In\left(2\right)}} \ast D_m\right) \ast n^{-0.2}$

where $D_m$ is the (weighted) median distance of the (weighted) mean center; n is the number of points for which the population field is not used, or the sum of the population field values if one is provided; and SD is the standard distance. The min part indicates that the smaller of the two options (SD or $\sqrt{\frac{1}{I{n}_{\left(2\right)}}}$) will be used.

2.3.8. Delineation of UGB Based on Morphology

Commonly used methods for planning urban boundaries are corrosion and expansion algorithms [46,47]. Erosion is the shrinking of urban boundaries, mainly to eliminate meaningless objects. Expansion causes the boundaries of the city to expand outwards to fill in voids in the target area or to eliminate some small noise. The formulae are as follows:

• (1). The corrosion algorithm formula is:

  (9)$A\ominus B=\left\{x,y|{\left(B\right)}_{xy}\subseteq A\right\}$

  In the formula, structure B is used to corrode A. Notably, an origin needs to be defined in B, and the movement process of B is consistent with the movement process of the convolution kernel, which is the same as the calculation after the convolution kernel overlaps with the image. When the origin of B is shifted to the pixel (x, y) of image A, if B is completely included in the overlapping area of image A at (x, y), the corresponding pixel (x, y) of the output image is assigned to 1; otherwise, the value is assigned as 0.

• (2). The formula of the expansion algorithm is:

  (10)$A\oplus B=\left\{x,y|{(B)}_{xy}\cap A\ne \varnothing \right\}$

  In the formula, A is expanded by structure B, and the origin of structural element B is shifted to the position of image pixel (x, y). If the intersection of B and A at image pixel (x, y) is not empty (that is, at least one image value of A corresponding to the position of element 1 in B is 1), then the pixel (x, y) corresponding to the output image is assigned the value of 1; otherwise, it is assigned the value of 0.

3. Results

3.1. Dynamic Change in Land Cover and Effectiveness Evaluation of Validation Results

In the present study, the land cover data in 2010 were used as the initial data to predict the simulated land use status in 2020, as shown in Figure 3. The kappa coefficient was 0.725 and the overall accuracy was 82.9%. The FOM value was 0.235, and the 2020 accuracy of the simulation was good. As such, the land cover data in 2010 could be used as the initial data of land use development for the next 10 years. Between 2010 and 2020, the simulation accuracy of each type of land is shown in Table 4, and the corresponding land cover transfer matrix is shown in Table 5.

The conversion of grassland to other land was the most obvious, 12,541 km², followed by 4020.75 km². Then, 2201 km² was also converted from other land to grassland, 1763 km² was converted from cultivated land to grassland, and nearly 1023.5 km² was converted from farmland to construction land.

3.2. Characteristics of Multi-Scenario Development of Urban Agglomeration under the Goal of High-Quality Construction

The FLUS model was used to predict the relative changes in various land covers in 2030 under the natural development scenario, the farmland protection scenario, and the ecological protection scenario. As shown in Figure 4, other land cover increased from 20% to 30% between 2010 and 2030, while the change in construction land after 2020 was not particularly obvious and remained at about 3%. Since the construction of urbanization is a slow process, considering that the urbanization rate of the urban agglomeration reached a high level before 2020, the construction quantity of the city was standardized under the goal of high-quality urban agglomeration construction. The change in water body was stable at about 1%, and the relative change was not obvious. The grass land decreased from 56% in 2010 to 40% in 2020, and the decrease was most obvious between 2010 and 2020. After 2020, the growth rate remained at about 2%. The change in forestland and agricultural land was relatively not obvious, forest land was stable at about 3%, showing an upward trend, while agricultural land was stable at about 21%, also showing an upward trend. However, under the scenario of ecological protection in 2030, the farmland was predicted to decrease. Considering the importance of ecological security, the original cultivated land area was replanned.

3.2.1. Urban Land Development under the Natural Development Scenario

The natural development scenario involved free development without the interference of human factors on urban planning. Taking into account the construction of high-quality urban agglomeration, the urban land cell was reduced less or not enough. Considering the current development of urban agglomeration, the urban land was simulated. Based on the number of construction land patches in 2030 simulated by Markov, the expansion degree of construction area to other land pairs was slightly reduced. Based on the natural development scenario, land cover changes in three Chinese regions (Urumqi, Shihezi, and Wusu-Kuitun) were selected to analyze the urban expansion land. Firstly, the increase in plaques in Urumqi was obvious, as shown in Figure 5. And most of the plaques were concentrated in the southern region, followed by the number of plaques in the northeast region. The quantity of the patch increase was evenly distributed in Shihezi. The patch distribution of Wusu city and Kuitun city was mainly concentrated in the southern region, and the increase was more obvious. In the process of natural development, urban land expansion mainly involved excessive encroachment on grassland and farmland, mainly on grassland. In the process of expansion, the internal filling type of development was not obvious, while the edge expansion development mode was very obvious.

3.2.2. Urban Land Development under the Farmland Protection Scenario

Farmland protection is an overall strategic issue related to the sustainable development of China’s economy and society. “Rational use of land and practical protection of farmland” is a basic state policy that has been adhered to for a long time [48,49]. Under the high-quality construction of urban agglomeration, as an important basic policy, attention must be paid to the protection of cultivated land. As shown in Figure 6, Compared with the natural development scenario, the simulation of land cover status in 2030 under the protection of farmland shows that the development pattern of construction land occupying cultivated land decreased, and cities mostly expanded towards grassland. Three regions in the urban agglomeration, Urumqi, Shihezi and Wusu were selected. For example, the developments of Urumqi city and Wusu were more obvious, while the expansion degree of Shihezi was not obvious.

3.2.3. Urban Land Development under Ecological Protection Scenario

An environment-friendly development mode was adopted in the ecological protection scenario, so as to minimize the damage to the surrounding environment in the process of urban development. With the goal of high-quality urban agglomeration construction, urban expansion not only occupied less green space, but also a facilitated an increase in the green space area to form a state of harmonious development between humans and the environment. Figure 7 shows the prediction of urban development under the ecological protection scenario. The results show that cities in the three regions expanded significantly in cultivated land, the new patches developed to the urban edge, and the expansion was mainly concentrated in farmland areas. In other areas, urban land use exhibited a trend of local growth, which was characterized by the relatively saturated development of urban construction land. Such development did not destroy the natural characteristics to a large extent. Obviously, under the ecological protection scenario, the ecological conditions were seldom broken and there was little deviation from the state of urban natural development, but there is a certain degree of reference significance for people who want to adjust the urban development at a later period.

3.3. Comparison of Temporal and Spatial Characteristics of Urban Land Development

The urban land development was compared by means of the kernel density method, as shown in Figure 8. The results show that between 2010 and 2030, the four cities with the highest density of urban land use were Urumqi, Wusu-Kuitun, Shihezi, and Karamay. Under the natural development scenario, the urban land expansion was the most obvious, and the surrounding farmland was the most suitable area for future development. Under the natural development scenario, the urban density range increased, and the Wusu-Kuitun area was predicted to change significantly from 2010 to 2030. Under the scenario of cultivated land protection, some areas in south China may become land for urban development in the future. Under the ecological protection scenario, the scope of surrounding urban expansion was not predicted to change significantly, and the expansion of urban land was not obvious compared with 2020. In the future, major urban land expansion will mainly be concentrated in several areas in the north.

3.4. UGB Delineation of Urban Agglomeration on the Northern Slope of the Tianshan Mountains under Multiple Scenarios

According to FLUS-UGB, in the present study, 3 × 3, 5 × 5, and 7 × 7 neighborhood windows were adopted to establish the planning boundary (Figure 9). The results show that the 3 × 3 neighborhood analysis planning city had a better boundary effect and could display the urban contour in a relatively fine manner. Therefore, the 3 × 3 neighborhood window was adopted to establish the growth boundary of urban agglomeration on the northern slope of the Tianshan Mountains.

The demarcation of an urban growth boundary (UGB) can provide relevant policy norms for the future development of cities. In the present study, based on the morphological erosion and expansion method, a UGB for cities in the urban agglomeration on the northern slope of the Tianshan Mountains in 2030 was demarcated under the natural development scenario, the cultivated land protection scenario, and the ecological protection scenario. The 3 × 3 neighborhood window retained the urban results of the previous multi-scenario simulation well, and the effects of nature and the social economy comprehensively. The simulated UGB results under multiple scenarios are shown in Figure 10 and Figure 11. Under the natural development scenario, the urban development boundary range was larger, followed by the cultivated land protection scenario, and finally the ecological protection scenario. Based on the goal of high-quality construction, the concept of ecological protection, harmonious coexistence between man and nature and sustainable development will be adhered to in the future development of urban land. On the basis of protecting cultivated land, water areas, and ecological protection areas, reasonable UGB planning policies can be formulated. UGBs effectively show the planning problems of future urban development boundaries, and have relevant reference value for future urban development.

4. Discussion

High-quality development encompasses the concepts of ecological protection, human–land harmony, and sustainability. By increasing the extent of green space and slowing down urban land expansion, the urban land conversion process fully follows the concept of urban expansion. In the simulated land use development process, relative deviations were observed in each land use conversion category, which were caused by not fully considering the role of constraints. When the constraints were added and simulated again, the accuracy of the simulations was improved. When the results of UGB delineation were simulated, the small-scale urban land boundaries were not well shown or were missing, but the 3 × 3 domain window used in the simulation of the large-scale urban land boundaries shows the urban boundary contours better, which is consistent with previous studies [50].

In the present study, the changes in urban land use were analyzed under various scenarios. In addition, urban boundaries were set under corresponding scenarios for a simulated urban cluster on the northern slope of the Tianshan Mountains in Xinjiang, China, under the goal of high-quality construction. The FLUS model was validated as a reliable model for future land change simulation, and in previous studies, most studies in which the FLUS model is used have focused on the more developed regions in eastern China [51,52,53,54]. As such, there is a scarcity of research on cities in the western arid zone. Urban land expansion is caused by multiple factors [55,56], and any one of such factors can influence changes in urban land use. In arid zone studies, the restricted area has a significant role in future land use changes in urban clusters on the northern slopes of Tianshan. In the present study, the permanent snow area and water bodies in Tianshan were classified as limiting factors and also considered as drivers of land use change. The results confirm the reliability of the simulation results. However, differing from previous studies, the land cover change and urban land dynamics of the urban cluster on the northern slopes of Tianshan were analyzed under multiple scenarios. At the same time, the urban land growth boundaries were limited under multiple scenarios, indicating that UGB designation is reasonable.

High-quality development covers ecological protection, man–land coordination, and sustainability [57,58,59]. Urban land conversion should fully follow the increase in green space and slow down the development of urban land, so as to conform to the concept of high-quality urban development. In the process of simulating land development, there was a relative deviation in the conversion of different types of land cover. The main reason for the deviation was that the role of limiting factors was not fully considered. The accuracy of the simulation was improved when the constraint factor was added. In the setting of UGB, the boundary of small land construction land was not well displayed or was even missing, but the 3 × 3 field window was used in the boundaries of large cities to better display the outline of the urban boundaries.

Although multiple types of factors for urban land change were fully considered in the present study, a regional land change is disturbed by multiple types of factors. Thus, the setting of the driving factors is also a significant indication of the accuracy of the simulation results. The area selected for the simulation is mainly in the central region of Asia and Europe, and the arid climate environment also has a considerable influence on the direction of urban development. With the advancement of technology, the new generation of urban land change simulation will be more accurate, which will increase the scientific level and feasibility of the simulation for urban planning to a large extent [60,61].

5. Conclusions

In the present study, the FLUS-Markov model, the UGB model, and kernel density analysis were combined to simulate the urban agglomeration on the northern slope of the Tianshan Mountains under multiple scenarios. At the same time, the spatial-temporal dynamics of urban land were analyzed and the urban development boundary was defined. In the process, the land cover data of 2010 were used to simulate the land cover data of 2020. The results show that the kappa value was 0.725, the overall accuracy was 82.9%, and the FOM value was 0.245, exhibiting good accuracy. Under multiple scenarios, the expansion degrees of grassland, other land, and cultivated land varied significantly between 2020 and 2030, but the proportions of construction area to total regional area remained at about 3%. Under the natural development scenario, urban land expansion was the most obvious, followed by the farmland protection scenario, while under the ecological protection scenario, construction land expansion was the least significant. Under the three scenarios, the expansion of construction land was mainly dominated by the occupation of grassland and cultivated land, showing the expansion of urban concentrated contiguous edges. At the same time, the changes in urban land use under the natural development scenario were relatively in line with the reality. The results of the kernel density show a year-by-year expansion of the urban area, and the best suitable development area should be the surrounding farmland. The city boundary information in 2030 was displayed according to the delineation of UGB in the corresponding scenario. By comparing the urban boundaries between 2010 and 2030, the dynamic changes of the boundaries were clearly depicted. The dynamic changes and UGB delineation of the urban agglomeration on the northern slope of the Tianshan Mountains under multiple scenarios can provide effective reference suggestions for future city planning.

Author Contributions

Conceptualization, D.L. and Z.S.; methodology, D.L.; software, D.L.; validation, D.L. and Y.C.; formal analysis, D.L.; data curation, D.L.; writing—original draft preparation, D.L.; writing—review and editing, D.L.; project administration, Z.S.; funding acquisition, S.Y. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. (a) The position of Xinjiang Uygur Autonomous Region in China; (b) the position of the Tianshan northern slope urban agglomeration in Xinjiang; (c) the study area.

Figure 2. The flow of the FLUS model.

Figure 3. Comparison of simulation results.

Figure 4. Temporal change in land area in different periods.

Figure 5. Change in urban land under the natural development scenario.

Figure 6. Changes in urban land under the farmland protection scenario.

Figure 7. Change in urban land under the ecological protection scenario.

Figure 8. Temporal and spatial variation characteristics of urban land use.

Figure 9. Urban development boundaries of neighborhood windows at different scales.

Figure 10. Urban growth boundary demarcation under three scenarios (Take Urumqi, Xinjiang, China, as an example).

Figure 11. Differences in UGBs in different periods: (a) the real city boundary in 2010; (b) the real city boundary in 2020; (c) the city boundary under the natural development scenario in 2030; (d) the urban boundary of the farmland protection scenario in 2030; (e) the city boundary under the ecological protection scenario in 2030.

Table 1

Land cover classification.

Types	After Classification (Name)
Farmland	Farmland
Grass	Grass
Forest	Forest
Shrubland
Wetlands	Water body
Water body
Contribution land	Contribution land
Bare ground	Other lands
Permafrost and Glacier Snow

Table 2

Data sources.

Data Categories	The Required Data	Source	Use
Land use data	2010	GlobeLand 30 (accessed on 3 May 2020)(http://www.globallandcover.com/)	Initial data for the model
Land use data	2020	GlobeLand 30(http://www.globallandcover.com/)	Validation data for the model
Natural factors	DemSlopeAspect	Geospatial data cloud(accessed on 22 June 2013)(http://www.gscloud.cn/)	Consider topography
The limiting factors	Water bodymountains	Geospatial data cloud(http://www.gscloud.cn/)	Limiting conversion factor
Socioeconomic factors	Population raster data	World Pop (accessed on 3 October 2013)(https://www.worldpop.org/)	Consider the socioeconomic impact
	GDP raster data	Earth Observation Group (accessed on 3 May 2021)(https://eogdata.mines.edu/products/vnl/)
	Distance to railway	NCSG (accessed on 3 May 2021)(https://www.webmap.cn)
	Distance to high speed
	Distance to dart
	Distance to the city center
	Distance to town center

Table 3

Transfer cost matrix.

2010–2030	Natural Development Scenario						Farmland Protection Scenario						Ecological Protection Scenario
	Farmland	Forest	Grass	Water	Construction Land	Other Land	Farmland	Forest	Gras	Water	Construction Land	Other Land	Farmland	Forest	Grass	Water	Construction Land	Other Land
Farmland	1	1	1	0	1	0	1	0	0	0	1	0	1	1	1	1	0	0
Forest	1	1	1	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0
Grass	0	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	0
Water	1	0	1	1	0	1	0	0	0	1	0	0	0	0	0	1	0	0
Construction land	1	0	1	1	1	0	1	0	0	0	1	0	0	0	0	0	1	0
Other lands	1	1	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1

Note: 1 means that the transformation of the land class can occur; 0 means that the transformation of the land class cannot occur.

Table 4

Conversion accuracy of each land type.

Land Type	Farmland	Forest	Grass	Water	Construction Land	Other Land
Accuracy	0.79	0.69	0.80	0.71	0.72	0.83

Table 5

Land cover transfer matrix.

2010–2020 (km2)	Farmland	Forest	Grass	Waterbody	Construction Land	Other Land
Farmland	15,172	6.75	1763	98.5	1023.5	141
Forest	0.5	794.5	688	0.5	10.2	5.2
Grass	4020.7	701	34,558.2	249.2	505.7	12,541
Water body	137.2	2.75	284.5	377	13	63.7
Construction land	699.5	9.5	198.25	15.7	845.7	14.7
Other lands	857.5	4	2201	101	155.2	15,468.5

References

1. United Nations Population Fund. Available online: https://www.unfpa.org/ (accessed on 18 August 2021).

2. Fang, C.; Yu, D. Urban Agglomeration: An Evolving Concept of an Emerging Phenomenon. Landsc. Urban Plan.; 2017; 162, pp. 126-136. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2017.02.014]

3. Report on the Sustainable Development Goals 2021. Available online: https://unstats.un.org/sdgs/report/2021/ (accessed on 6 July 2021).

4. DESA U N. World Urbanization Prospects: The 2018 Revision (ST/ESA/SER. A/420); United Nations: New York, NY, USA, 2019.

5. Fang, C. Development status quo and key points of construction of urban agglomerations in west regions of China. J. Arid Land.; 2010; 33, pp. 667-675.

6. Surya, B.; Hamsina, H.; Ridwan, R.; Baharuddin, B.; Menne, F.; Fitriyah, A.T.; Rasyidi, E.S. The Complexity of Space Utiliza-tion and Environmental Pollution Control in the Main Corridor of Makassar City, South Sulawesi, Indonesia. Sustainability; 2020; 12, 9244. [DOI: https://dx.doi.org/10.3390/su12219244]

7. Guo, R.; Sun, Y.; Fan, J. Policies for classified governance of Chinese urban agglomerations during the 14th Five-Year Plan period. Proc. Chin. Acad. Sci.; 2020; 35, pp. 844-854. (In Chinese) [DOI: https://dx.doi.org/10.16418/j.issn.1000-3045.20200408001]

8. Chen, S.; Huang, Q.; Liu, Z.; Meng, S.; Yin, D.; Zhu, L.; He, C. Assessing the Regional Sustainability of the Beijing-Tianjin-Hebei Urban Agglomeration from 2000 to 2015 Using the Human Sustainable Development Index. Sustainability; 2019; 11, 3160. [DOI: https://dx.doi.org/10.3390/su11113160]

9. Tang, M.; Liang, C.; Zeng, J.; Gen, Y. Development characteristics and transformation paths of towns in Fujian Triangle ur-ban agglomeration under ecological constraints. Acta Ecol. Sin.; 2021; 41, pp. 6038-6050. (In Chinese)

10. Gao, Q.; Fang, C.; Liu, H.; Zhang, L. Conjugate evaluation of sustainable carrying capacity of urban agglomeration and multi-scenario policy regulation. Sci. Total Environ.; 2021; 785, 147373. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2021.147373]

11. Surya, B.; Salim, A.; Hernita, H.; Suriani, S.; Menne, F.; Rasyidi, E.S. Land Use Change, Urban Agglomeration, and Urban Sprawl: A Sustainable Development Perspective of Makassar City, Indonesia. Land; 2021; 10, 556. [DOI: https://dx.doi.org/10.3390/land10060556]

12. Yang, B.; Xu, T.; Shi, L. Analysis on sustainable urban development levels and trends in China’s cities. J. Clean. Prod.; 2017; 141, pp. 868-880. [DOI: https://dx.doi.org/10.1016/j.jclepro.2016.09.121]

13. Cui, Y. Moving Towards a Greener and More Beautiful Society. Life Expectations of the People; Springer: Singapore, 2021; pp. 199-217.

14. Khan, A.A.; Zhao, Y.; Rahman, G.; Rafiq, M.; Moazzam, M.F.U. Spatial and temporal analysis of rainfall and drought condition in southwest Xinjiang in Northwest China, using various climate indices. Earth Syst. Environ.; 2021; 5, pp. 201-216. [DOI: https://dx.doi.org/10.1007/s41748-021-00226-5]

15. Liu, Y.; Hu, W.; Wang, S.; Sun, L. Eco-environmental effects of urban expansion in Xinjiang and the corresponding mechanisms. Eur. J. Remote Sens.; 2021; 54, pp. 132-144. [DOI: https://dx.doi.org/10.1080/22797254.2020.1803768]

16. Fang, C. China’s urban agglomeration and metropolitan area construction under the new development pattern. Econ. Geogr.; 2021; 41, pp. 1-7.

17. Chakraborty, S.; Maity, I.; Patel, P.P.; Dadashpoor, H.; Pramanik, S.; Follmann, A.; Novotný, J.; Roy, U. Spatio-temporal patterns of urbanization in the Kolkata Urban Agglomeration: A dynamic spatial territory-based approach. Sustain. Cities Soc.; 2021; 67, 102715. [DOI: https://dx.doi.org/10.1016/j.scs.2021.102715]

18. Mallick, S.K.; Das, P.; Maity, B.; Rudra, S.; Pramanik, M.; Pradhan, B.; Sahana, M. Understanding future urban growth, urban resilience and sustainable development of small cities using prediction-adaptation-resilience (PAR) approach. Sustain. Cities Soc.; 2021; 74, 103196. [DOI: https://dx.doi.org/10.1016/j.scs.2021.103196]

19. Huang, Y.; Liao, T.; Fu, W.; Zhou, Z. Study on Dynamic Change of Regional land cover based on GIS. J. Soil Water Convers.; 2002; 16, pp. 54-57. (In Chinese)

20. Tonne, C.; Adair, L.; Adlakha, D.; Anguelovski, I.; Belesova, K.; Berger, M.; Brelsford, C.; Dadvand, P.; Dimitrova, A.; Giles-Corti, B. et al. Defining Pathways to Healthy Sustainable Urban Development. Environ. Int.; 2021; 146, 106236. [DOI: https://dx.doi.org/10.1016/j.envint.2020.106236] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33161201]

21. Zhou, L.; Dang, X.; Sun, Q.; Wang, S. Multi-scenario simulation of urban land change in Shanghai by random forest and CA-Markov model. Sustain. Cities Soc.; 2020; 55, 102045. [DOI: https://dx.doi.org/10.1016/j.scs.2020.102045]

22. Dadashpoor, H.; Azizi, P.; Moghadasi, M. Analyzing Spatial Patterns, Driving Forces and Predicting Future Growth Scenarios for Supporting Sustainable Urban Growth: Evidence from Tabriz Metropolitan Area, Iran. Sustain. Cities Soc.; 2019; 47, 101502. [DOI: https://dx.doi.org/10.1016/j.scs.2019.101502]

23. Liu, J.; Zhang, L.; Zhang, Q.; Zhang, G.; Teng, J. Predicting the surface urban heat island intensity of future urban green space development using a multi-scenario simulation. Sustain. Cities Soc.; 2021; 66, 102698. [DOI: https://dx.doi.org/10.1016/j.scs.2020.102698]

24. Yang, Z.; Yang, H.; Wang, H. Evaluating Urban Sustainability under Different Development Pathways: A Case Study of the Beijing-Tianjin-Hebei Region. Sustain. Cities Soc.; 2020; 61, 102226. [DOI: https://dx.doi.org/10.1016/j.scs.2020.102226]

25. Huo, J.; Shi, Z.; Zhu, W.; Xue, H.; Chen, X. A Multi-Scenario Simulation and Optimization of Land Use with a Markov–FLUS Coupling Model: A Case Study in Xiong’an New Area, China. Sustainability; 2022; 14, 2425. [DOI: https://dx.doi.org/10.3390/su14042425]

26. Guo, H.; Cai, Y.; Yang, Z.; Zhu, Z.; Ouyang, Y. Dynamic simulation of coastal wetlands for Guangdong-Hong Kong-Macao Greater Bay area based on multi-temporal Landsat images and FLUS model. Ecol. Indic.; 2021; 125, 107559. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107559]

27. Shen, Q.; Chen, Q.; sin Tang, B.; Yeung, S.; Hu, Y.; Cheung, G. A System Dynamics Model for the Sustainableland coverPlanning and Development. Habitat Int.; 2009; 33, pp. 15-25. [DOI: https://dx.doi.org/10.1016/j.habitatint.2008.02.004]

28. Liang, X.; Guan, Q.; Clarke, K.C.; Liu, S.; Wang, B.; Yao, Y. Understanding the Drivers of Sustainable Land Expansion Using a Patch-Generatingland coverSimulation (PLUS) Model: A Case Study in Wuhan, China. Comput. Environ. Urban Syst.; 2021; 85, 101569. [DOI: https://dx.doi.org/10.1016/j.compenvurbsys.2020.101569]

29. Li, Y.; Ye, C.; Hang, X. Temporal-Sptional Evolution and Scenario simulation of Production-Living-Ecological space in Nanchang City based on CLUE-S model. Res. Soil Water Conserv.; 2021; 28, pp. 326-331. (In Chinese)

30. Wang, H.; Bao, C. Scenario modeling of ecological security index using system dynamics in Beijing-Tianjin-Hebei urban agglomeration. Ecol. Indic.; 2021; 125, 107613. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107613]

31. Li, X.; Chen, Y.; Liu, X.; Xu, X.; Chen, G. Experiences and issues of using cellular automata for assisting urban and regional planning in China. Int. J. Geogr. Inf. Sci.; 2017; 31, pp. 1606-1629. [DOI: https://dx.doi.org/10.1080/13658816.2017.1301457]

32. Aburas, M.M.; Ho, Y.M.; Ramli, M.F.; Ash’aari, Z.H. The simulation and prediction of spatio-temporal urban growth trends using cellular automata models: A review. Int. J. Appl. Earth Obs.; 2016; 52, pp. 380-389. [DOI: https://dx.doi.org/10.1016/j.jag.2016.07.007]

33. Liu, X.; Liang, X.; Li, X.; Xu, X.; Ou, J.; Chen, Y.; Li, S.; Wang, S.; Pei, F. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landsc. Urban Plan; 2017; 168, pp. 94-116. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2017.09.019]

34. Maithani, S. A Neural Network Based Urban Growth Model of an Indian City. J. Indian Soc. RS; 2009; 37, pp. 363-376. [DOI: https://dx.doi.org/10.1007/s12524-009-0041-7]

35. Wang, Q.Z.; Guan, Q.; Lin, J.; Luo, H.; Tan, Z.; Ma, Y. Simulating Land Use/Land Cover Change in an Arid Region with the Coupling Models. Ecol. Indic.; 2021; 122, 10723. [DOI: https://dx.doi.org/10.1016/j.ecolind.2020.107231]

36. Mansour, S.; Al-Belushi, M.; Al-Awadhi, T. Monitoring land use and land cover changes in the mountainous cities of Oman using GIS and CA-Markov modelling techniques. Land Use Policy; 2020; 91, 104414. [DOI: https://dx.doi.org/10.1016/j.landusepol.2019.104414]

37. Zhang, C.; Wang, P.; Xiong, P.; Li, C.; Quan, B. Spatial Pattern Simulation ofland coverBased on FLUS Model under Ecological Protection: A Case Study of Hengyang City. Sustainability; 2021; 13, 10458. [DOI: https://dx.doi.org/10.3390/su131810458]

38. Zhang, J.; Mei, Z.; Lv, J.; Chen, J. Simulating Multipleland coverScenarios based on the FLUS Model Considering Considering Spatial Auto correlation. J. Geo-Inf. Sci.; 2020; 22, pp. 532-542. (In Chinese)

39. Zhang, D.; Liu, X.; Lin, Z.; Zhang, X.; Zhang, H. The Delineation of Urban Growth Boundaries in Complex Ecological Environment Areas by Using Cellular Automata and a Dual-Environmental Evaluation. J. Clean. Prod.; 2020; 256, 120361. [DOI: https://dx.doi.org/10.1016/j.jclepro.2020.120361]

40. Zhang, S.; Wei, L.; Jin, X.; Lu, Y. Theland coverSimulation and Delimitation of Urban Development Boundary in County Area based on FLUS-UGB. J. Geo-Inf. Sci.; 2020; 22, pp. 1849-1859. (In Chinese)

41. Chettry, V.; Surawar, M. Delineating Urban Growth Boundary Using Remote sensing, ANN-MLP and CA model: A Case Study of Thiruvananthapuram Urban Agglomeration, India. India. J. Indian Soc. Remote Sens.; 2021; 49, pp. 2437-2450. [DOI: https://dx.doi.org/10.1007/s12524-021-01401-x]

42. Wang, B.; Liao, J.; Zhu, W.; Qiu, Q.; Wang, L.; Tang, L. Setting of neighborhood weights in FLUS model based on historical scenarios: A case study ofland coversimulation of Golden Triangle Urban agglomeration in Southern Fujian in 2030. Acta Ecol. Sin.; 2019; 12, pp. 1-13. (In Chinese)

43. Liu, X.; Ming, W.; Li, Z.G.; Zeng, J. Multi-scenario simulation of urban growth boundaries with an ESP-FLUS model: A case study of the Min Delta region. China Ecol. Indic.; 2022; 135, 108538. [DOI: https://dx.doi.org/10.1016/j.ecolind.2022.108538]

44. Liang, X.; Liu, X.; Li, X.; Chen, Y.; Tian, H.; Yao, Y. Delineating multi-scenario urban growth boundaries with a CA-based FLUS model and morphological method. Landsc. Urban Plan.; 2018; 177, pp. 47-63. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2018.04.016]

45. Li, Q.; Feng, Y.; Tong, X.; Zhou, Y.; Wu, P.; Xie, H.; Jin, Y.; Chen, P.; Liu, S.; Xv, X. et al. Firefly Algorithm-Based Cellular Automata for Reproducing Urban Growth and Predicting Future Scenarios. Sustain. Cities Soc.; 2022; 76, 103444. [DOI: https://dx.doi.org/10.1016/j.scs.2021.103444]

46. Wang, W.; Jiao, L.; Zhang, W.; Jia, Q.; Su, F.; Xu, G.; Ma, S. Delineating urban growth boundaries under multi-objective and constraints. Sustain. Cities Soc.; 2020; 61, 102279. [DOI: https://dx.doi.org/10.1016/j.scs.2020.102279]

47. Zhao, Z.; Guan, D.; Du, C. Urban growth boundaries delineation coupling ecological constraints with a growth-driven model for the main urban area of Chongqing, China. GeoJournal; 2020; 85, pp. 1115-1131. [DOI: https://dx.doi.org/10.1007/s10708-019-10014-4]

48. Zhou, Y.; Li, X.; Liu, Y. Cultivated land protection and rational use in China. Land Use Policy; 2021; 106, 105454. [DOI: https://dx.doi.org/10.1016/j.landusepol.2021.105454]

49. Sun, Q.; Wu, M.; Du, P.; Qi, W.; Yu, X. Spatial Layout Optimization and Simulation of Cultivated Land Based on the Life Community Theory in a Mountainous and Hilly Area of China. Sustainability; 2022; 14, 3821. [DOI: https://dx.doi.org/10.3390/su14073821]

50. Wu, X.; Liu, X.; Liang, X.; Chen, G. Delineation of urban growth boundaries in the Pearl River Delta based on FLUS-UGB multi-scenario simulation. GSIS; 2018; 20, pp. 532-542. (In Chinese)

51. Zhang, J.; Liu, Z.; Li, S. Research on Land Use Simulation of Incorporating Historical Information into the FLUS Model—Setting Songyuan City as an Example. Sustainability; 2022; 14, 3828. [DOI: https://dx.doi.org/10.3390/su14073828]

52. Wang, Y.; Shen, J.; Yan, W.; Chen, C. Backcasting approach with multi-scenario simulation for assessing effects of land use policy using GeoSOS-FLUS software. MethodsX; 2021; 6, pp. 1384-1397. [DOI: https://dx.doi.org/10.1016/j.mex.2019.05.007]

53. Liu, P.; Hu, Y.; Jia, W. Land use optimization research based on FLUS model and ecosystem services–setting Jinan City as an example. Urban Clim.; 2021; 40, 100984. [DOI: https://dx.doi.org/10.1016/j.uclim.2021.100984]

54. Chen, Z.; Huang, M.; Zhu, D.; Altan, O. Integrating remote sensing and a markov-FLUS model to simulate future land use changes in Hokkaido. Remote Sens.; 2021; 13, 2621. [DOI: https://dx.doi.org/10.3390/rs13132621]

55. Sahana, M.; Hong, H.; Sajjad, H. Analyzing urban spatial patterns and trend of urban growth using urban sprawl matrix: A study on Kolkata urban agglomeration, India. Sci. Total Environ.; 2018; 628, pp. 1557-1566. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2018.02.170] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30045573]

56. Li, G.; Li, F. Urban sprawl in China: Differences and socioeconomic drivers. Sci. Total Environ.; 2019; 673, pp. 367-377. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2019.04.080] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30991326]

57. Sachs, J.; Kroll, C.; Lafortune, G.; Fuller, G.; Woelm, F. Sustainable Development Report; Cambridge University Press: Cambridge, CA, USA, 2021.

58. Hasan, S.S.; Deng, X.; Li, Z.; Chen, D. Projections of Future Land Use in Bangladesh under the Background of Baseline, Ecological Protection and Economic Development. Sustainability; 2017; 9, 505. [DOI: https://dx.doi.org/10.3390/su9040505]

59. Cai, B.; Shao, Z.; Fang, S.; Huang, X.; Huq, M.E.; Tang, Y.; Li, Y.; Zhuang, Q. Finer-scale spatiotemporal coupling coordination model between socioeconomic activity and eco-environment: A case study of Beijing, China. Ecol. Indic.; 2021; 131, 108165. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.108165]

60. Naghibi, F.; Delavar, M.R.; Pijanowski, B. Urban growth modeling using cellular automata with multi-temporal remote sensing images calibrated by the artificial bee colony optimization algorithm. Sensors; 2016; 16, 2122. [DOI: https://dx.doi.org/10.3390/s16122122]

61. Zhuang, H.; Liu, X.; Yan, Y.; Zhang, D.; He, J.; He, J.; Zhang, X.; Zhang, H.; Li, M. Integrating a deep forest algorithm with vector-based cellular automata for urban land change simulation. Trans. GIS.; 2022; 1, pp. 1-25. [DOI: https://dx.doi.org/10.1111/tgis.12935]