Spatiotemporal Dynamic Analysis and Simulation Prediction of Land Use and Landscape Patterns from the Perspective of Sustainable Development in Tourist Cities

Abstract

The unreasonable development of land use and use of landscape patterns are the primary causes of the unsustainable growth of tourism cities. Dawa District, a well-known tourist destination in China, was chosen as the research object in order to scientifically evaluate and simulate the dynamic changes in land use. Landsat remote sensing images of the Dawa District from 2011, 2016, and 2021 were processed, using the development of regional tourism in 2016 as the dividing point. Using Arcgis10.3, ENVI5.3, and Fragstats tools, GIS spatial analysis and landscape index analysis were used to explore the spatiotemporal dynamic characteristics of land use and landscape patterns. The CA-Markov model was used to simulate and predict the land use landscape pattern in 2026. The results show that the rapid expansion of built-up areas in 2011–2021 and the increasing marginal complexity and fragmentation of landscape patterns are the main evolution trends. In 2016, emphasis was placed on sustainable land resources, and the area of forested water gradually recovered. However, the patch density (PD) and number of patches (NP) indices reflecting fragmentation still showed a 1.5–2 fold increase. It is expected that in 2026, the dual pressure of urban development and ecological sustainable construction will enhance connectivity and aggregation, and the ecological environment will gradually recover. At the same time, the competition for various types of land will become increasingly prominent, and agricultural activities will be affected. In summary, optimization strategies are proposed from the perspectives of government, land, landscape, and industry, aiming to promote the sustainable development of land and landscape resources in tourist cities.

1. Introduction

In 2016, the term “regional tourism” was proposed based on the Chinese context. In the English context, there is no professional term to characterize the concept, but there are similar concepts in the development of tourism [1]. It refers to the spatial panoramic system of tourism, which breaks the limitations of traditional fixed-point tourism and small tourism and emphasizes the concept of point, line, three-dimensional, and all-around sustainable large-scale tourism [2]. In China, the regional tourism policy was fully implemented in 2016, and the tourism industry’s potential was stimulated in order to attain economic and ecological sustainability simultaneously. Simultaneously, linked issues are slowly gaining prominence. Some tourism development is carried out mindlessly in the pursuit of short-term economic gains and policy dividends, without proper study and evaluation of land coverage and landscape resources. This conduct has had an effect on the ecology, resulting in ecological deterioration, landscape pattern fragmentation, an incomplete industrial chain, and other phenomena [3,4]. As a result, evaluating and simulating the evolution trend of land use and landscape patterns is critical in the context of regional tourism strategy implementation.

The evolution of landscape patterns and land use and land cover change (LUCC) can be directly linked to tourism development [5], and it is also an essential way to alter ecosystem economy services [6]. For example, the conservation of wetland and woodland landscapes, the expansion of shopping and other service industry regions, and the expansion and contraction of land of various properties all have an impact on the entire ecosystem. There are numerous classic examples of sustainable tourism in practice. The Swiss Alps tourist area, for example, extends tourism operations to every town and village, and each location serves as a tourism service unit [7]. The French notion of whole-region tourist structure construction of “agricultural production reception + rural vacation + leisure space” represents an early type of all-region tourism [8,9,10]. However, as tourism growth progresses, the unavoidable expansion of tourism construction puts strain on the regional ecology [11], and agricultural arable land is crushed by tourism-related land, posing a serious risk of irreversible exhaustion [3]. This requires research on the evolution of land use and land cover change (LUCC) and landscape patterns in the context of sustainable regional tourism.

Tourism-related research has reached its full maturity. However, regional tourism is in its initial stage due to its particularity of mobilizing regional resources and sustainable development. Existing studies tend to construct regional tourism strategies at the content level. There is a scarcity of studies on landscape resource allocation and sensible spatial layout planning through regional tourist practices. For example, Lars Aronsson, writing from the perspective of Swedish tourism, proposed the concept of sustainable resource production in supply and demand, constituting the preliminary concept of coordinated and sustainable regional tourism [12], and Trono explored tourism while respecting the principles of the environment, culture, society, and economic sustainability [13]. After China formally introduced the concept of regional tourism in 2016, the study of relevant techniques showed methodical growth. Gradually, the professional concept of regional sustainable tourism and numerous strategies, including resources, industry, management, environment, characteristics, and sharing, took form (Zhang (2016) [14], Mao (2019) [15], and Liu (2017)). Liu (2017) used the expert consultation method and the hierarchical analysis method to develop a comprehensive evaluation index system for regional rural tourism and the regional tourism construction evaluation system for the first time [16].

At the level of research methodology, regional tourism focuses primarily on subjective and qualitative investigation and analysis but lacks objective and quantitative investigation. Consequently, in conjunction with the comparatively mature remote sensing technology and the LUCC and landscape pattern research approaches, researchers have applied satellite data to large-scale natural resource evaluation, analysis, and planning, resulting in a relatively mature theoretical system [17]. Some studies used RS and GIS to analyze the changes in land use types and landscape patterns in the research area and conducted studies in a variety of contexts, including spatial and temporal evolution trend analysis and landscape vulnerability assessment (Choudhary (2017) [18], Li (2017) [19], and Min (2021) [20]). Simultaneously, the spatial modeling simulation conceptualizes natural system complexity, intrinsic dynamic features, and evolutionary patterns [21]. The cellular automata (CA)–Markov chain model (Markov) is a more effective method for simulating and predicting spatial phenomena. Combining the capability of complex CA with Markov long-term probability prediction is commonly used to simulate land use space and probability evolution (Singh (2015) [17]), predict future construction expansion, and predict ecological construction trends (Ajeeb (2020) [22], Fu (2022) [23], and Zhou (2016) [24]). Also, in China, the CA–Markov model was applied to simulation-related regional tourism landscape ecological security research (Liu (2022) [25]).

This research is based on a sustainable vision of regional tourism. As an example, consider the Dawa District of Panjin, which was designated as the first group of regional tourism demonstration zones in 2016. Using regional tourism starting in 2016 as the dividing line, the evolution of LUCC and landscape patterns in the study area between 2011 and 2021 were examined. In addition, a CA–Markov simulation was used to fore-cast the land use change in the study area in 2026. Added to the calculation for the landscape pattern index is evaluation at four levels of the internal order of the ecological environment of a landscape, including the area index, boundary shape index, connectivity index, and diversity index. Unlike traditional research on fixed-point tourism modes, this paper examines the evolution of land use and landscape patterns in tourist cities through the lens of the novel idea of regional sustainable tourism. Simultaneously, the objective and quantitative method is used to undertake evidence-based research on regional sustainable tourism practices in China in order to compensate for a lack of existing research on regional tourism. On the one hand, the positive and negative evolution of land use patterns and landscape patterns in regional sustainable tourism are investigated. On the other hand, by driving force analysis and spatial simulation prediction, damaging development processes, environmental degradation, and other issues can be avoided. This promotes the long-term development of the tourist city landscape pattern, the stability of the ecology, and the stable and efficient growth of regional tourism. The findings of this study provide research ideas for supporting the sustainable development of land and landscape resources in the context of regional tourist policy, as well as a reference for the building and development of similar tourism regions.

2. Materials and Methods

2.1. Study Area and Data Sources

The center of Panjin City’s Dawa District is located at 122°07′ E, 40°98′ N (Figure 1). Located in the Liaohe River Delta in the southwest of Liaoning Province, the lower reaches of the Liaohe River and the Daliao River, and the northeast of Liaodong Bay, a coastal plain was formed by the alluvial sediment downstream of the river. There are ecologically livable and flat terrain within China’s largest coastal wetland, red wetland beach wonders, China’s third-largest oil field, and other regional resources [26]. It can develop wetland ecological types, historical and cultural types, pastoral scenery types, and other tourism modes. There exist a 5A scenic spot in Red Wetland Beach, a 4A Panjin Weihai Dingxiang tourism resort, a number of 3A and 2A scenic spots, and minority and cultural tourism resources, in accordance with regional tourism conditions, for vigorously pursuing tourism development. In recent years, however, with the development of land resources, artificial landscape fragmentation has increased year by year. Wetland ecology has been damaged, with reduced area and frequent pollution. The original ecosystem has shown a trend of degradation due to damage to structure, function, and stability. The evolution of this landscape pattern and the ecological environmental effects it brings have received widespread attention from all sectors of society [27].

The Landsat7 ETM SLC-off satellite image data set was searched using the administrative region as the search border on the geospatial data cloud platform launched by the Computer Network Information Center of the Chinese Academy of Sciences. We chose remote sensing pictures from 2011, 2016, and 2021, with less than 50% cloud occlusion and a resolution of 30 m (

Table 1. Details of data sources used in the study.

Satellite	Sensor	Resolution (m)	Path/Row	Acquisition Date	Cloud Cover (%)	Bands
Landsat7	ETM+	30	120/32	2011-05-26	8.03	8
Landsat7	ETM+	30	120/31	2011-10-01	27.05	8
Landsat7	ETM+	30	120/32	2016-08-11	21.1	8
Landsat7	ETM+	30	120/31	2016-08-27	0.97	8
Landsat7	ETM+	30	120/31	2021-06-06	1	8
Landsat7	ETM+	30	120/32	2021-07-24	2	8

).

Due to the lack of vector image data such as county-level water systems and roads, the images provided by the Resource and Environment Science Data Center of the Chinese Academy of Sciences were used as the base map, and the Google satellite map from 2021 was added as a reference in order to improve the accuracy of subsequent land use simulation predictions.

Satellite remote sensing DEM data and slope data (resolution 250 m) were retrieved from the mirror website of the Chinese Academy of Sciences’ Computer Network Information Center’s Geospatial Data Cloud. However, since the lowest elevation in the study area is −9 m, and the highest elevation is 14 m, the difference between the lowest and highest elevations in the entire region is only 25 m, and the average elevation is 2.7 m. The terrain is generally elevated in the northeast and low in the southwest, and the elevation variation is minimal, having little effect on the development of LUCC and landscape patterns. No further analysis of the elevation and slope is necessary.

2.2. Preprocessing and LUCC Classification of Satellite Images

2.2.1. Image Preprocessing

• To begin with, a considerable number of band ripples in the Landsat7 data collected in the study hampered observation and analysis. Therefore, ENVI 5.3 software was used for Landsat strip restoration (Landsat_Gapfill) to create a complete satellite image free of ripple gaps.
• Due to the influence of atmospheric radiation, the velocity angle, and cloud conditions, there will be numerous variations in the initial state of satellite images captured by satellite, which cannot be used directly for research analysis. Before applying the analysis, data are necessary for radiative calibration (RC) and atmospheric correction (QUAC). Thus, the RC and QUAC modules in ENVI 5.3 were utilized to modify and correct the radiation values in remote sensing images as well as eliminate the impact of the atmosphere on remote sensing images. Consequently, more precise surface reflectivity data were obtained.
• Typically, the downloaded satellite images cannot contain the entirety of the study area, necessitating the use of multiple satellite images for composite processing. After applying the aforementioned geometric corrections, the satellite image’s coordinate system unifies them. ENVI 5.3 software was used for the image inlay operation, uniform color fusion, and cutting of the research area, thereby completing the preprocessing of remote sensing images [28]. Finally, a satellite image data collection with numerous bands and the ability to differentiate image content was created.

2.2.2. LUCC Classification of Satellite Images

The research focused primarily on the question of whether the evolution characteristics of land use and landscape patterns, as well as changes in the ecological environment, can be sustained in the context of regional tourism policy in order to create sustainable tourist ecology and economic benefits. Consequently, based on the natural characteristics of the land use, the LUCC classification system of the Chinese Academy of Sciences [27], which is more appropriate for natural ecology research, was chosen.

Table 2. LUCC classification and selection of land use types in the study area.

Land Type	Included Land Types
Built-up area	Urban and rural residential areas and other industrial and mining, transportation, and other land.
Forest	Forestry land that grows trees, vegetation, bamboo, and coastal mangroves.
Farmland	Cropland, including cultivated land, newly reclaimed land, leisure land, grass land, and crops; beach and marine land that has been cultivated for more than 3 years.
Water	Natural waters and land for water conservation facilities.
Unused	Currently undeveloped territory, including land that is difficult to use.

lists the five land categories chosen for this study: built-up area, forest, farmland, water, and unused.

With the aid of the software ENVI 5.3, supervised classification was primarily chosen, and classification samples of each land use type were selected in conjunction with the visual interpretation.

Table 3. Visual interpretation discrimination diagram.

Land Type	Band Combination	Visual Interpretation	Corresponding Remote Sensing Image Samples
Built-up area	6,5,3	Gray-white, color block combination broken, square shape
Forest	6,5,2	Green or dark green, with patches continuous
Farmland	5,4,3	Yellow-green or yellow areas, regular and uniform shape
Water	6,5,3	Large area of dark blue images
Unused	6,5,2	Irregular shaped yellow and white areas, and color messy images

displays the visual interpretation samples for the five land use types used in the study. Multiple samples were selected in conjunction with different bands and land use conditions in the field to assure classification accuracy [29]. The classification of land use was extracted from the satellite images. After selecting samples, the classification accuracy was evaluated to determine if each form of land sample could distinguish each land use type with precision. The data could be distinguished if they were greater than or equal to 1.90; if they were less than 1, then the sample selection was unqualified. The results of the three Landsat7 image classification tests are displayed in

Table 4. The separation calibration of supervised classification.

Land Type Comparison	2011 Image	2016 Image	2021 Image
Built-up area and farmland Built-up area and forest Built-up area and unused	1.6725	1.9873	1.8785
	1.9110	1.9799	1.8478
	1.8477	1.9832	1.9464
Built-up area and water Farmland and forest	1.9960	1.9942	1.9928
	1.8089	1.9598	1.7414
Farmland and unused Farmland and water Forest and water Forest and unused	1.8551	1.9999	1.9999
	1.9999	1.9999	1.9625
	1.9999	2.0000	1.9871
	1.9959	1.9999	1.9999
Water and unused	1.9889	2.0000	2.0000

, indicating that the selection of classified samples could distinguish between each land type that satisfied the conditions of the study.

2.3. Research Methods

2.3.1. Geographic Information System’s Spatial Analysis

The geographic information system’s spatial analysis function was used to combine spatial information data. We analyzed the land use change caused by tourism, as well as the origin and location of each land change. Fundamental quantitative technical support for the alteration of a land use type’s spatial distribution was provided. The acquired land use classification map was imported into Arcgis10.3 software in order to compute the land use type area data for each period. Using the ArcGIS10.3 software’s fusion and intersection tools, the area transfer matrix of the study area for 2011–2016 and 2016–2021 was created. Because of the overlapping content with the subsequent Markov transition probability matrix, the content of the matrix was not repeated. Based on the GIS spatial analysis, document processing was conducted to facilitate the subsequent landscape pattern index analysis and CA–Markov analysis.

2.3.2. Selection of Landscape Indices

The landscape pattern index is a quantitative index of highly concentrated structure organizations and spatial configurations of landscape patterns. The landscape pattern method is a research method for quantifying landscape pattern information using the landscape pattern index and then analyzing the evolution trend of landscape patterns in a specific time dimension [30]. According to the characteristics and objectives of this study, there were 13 indices: landscape type proportion (PLAND), maximum patch index (LPI), patch density (PD), plaque number (NP), perimeter area dimension (PAFRAC), patch shape index (LSI), contag (CONTAG), aggregation (AI), separation (SPLIT), conductance index (COHESION), Shannon diversity index (SHDI), and Shannon evenness index (SHEI). The Arcgis10.3 land use images and data were imported into Fragstats4.2, and the indexes were calculated based on the selected indexes to derive the landscape pattern index analysis results.

2.3.3. CA–Markov Model

The CA–Markov model incorporates the two fundamental principles of cellular automaton (CA) and the transition probability matrix of Markov (Markov). Based on the concept of neighborhood interaction, CA is a motor feature of the self-replication of cell bodies. It is capable of constructing a grid-based dynamic model based on spatial interaction and temporal causation. It is capable of simulating the spatiotemporal evolution of complex systems [31], but the quantitative results are limited. The transformation probability matrix determines the probability of a cell or pixel converting from one land use category to another, where the Markov model complements the absence of CA quantification results [17]. The CA–Markov model is a suitable approach for predicting the spatial and temporal evolution of land use and landscape patterns using simulations.

This research used existing data for data preparation. Vector photos from 2011, 2016, and 2021 were converted into grid files using the Arcgis10.3 conversion tool. To unify the three picture borders, three land use intersection mask images were further generated, and the boundaries were cut according to the mask extraction technique. After processing, we used the conversion tool to convert the file to ascll format. We defined or unified each data set using a definition projection tool or a projection conversion tool. The files were imported into IDRISI17.0 to build the Markov chain by using the ability to predict the land use change in the next time period according to the land use change in continuous time periods. The t in the Markov model represents a certain time, and t1 < t2 < … < tn < tn + 1. The Markov chain process of a random t time period X (t) is as follows [32]:

$$\begin{gathered} F_x\left(X\left(t_{n+1}\right)\le x_{n+1}|X\left(t_n\right)=x_n,X\left(t_{n-1}\right)=x_{n-1},\dots ,X\left(t_1\right)=x_1\right) \\ =F_x\left(X\left(t_{n+1}\right)\le x_{n+1}|X\left(t_n\right)=x_n\right) \end{gathered}$$

(1)

If X (k) is a Markov chain in states {x1, x2, x3... xn}, then the probability of transition to j at state i is as follows:

$$P_{i,j}=P_r(X\left[k+1\right]=j|X\left[k\right]=i)$$

(2)

The transition probability matrix is as follows, where P is the transition probability:

$$\left[\begin{array}{cccc}{P}_{\mathrm{1,1}}& P_{\mathrm{1,2}}& \cdots & P_{1,n}\\ P_{\mathrm{2,1}}& P_{\mathrm{2,2}}& \cdots & P_{2,n}\\ ⋮& ⋮& \ddots & ⋮\\ P_{n,1}& P_{n,2}& \cdots & P_{n,n}\end{array}\right]$$

(3)

The CA model simulation predicts the spatial change processes of land use, including the cell, grid, cell neighborhood, boundary regulation, and conversion rules. This paper set the cell to a 30 m by 30 m grid image element. By analyzing the degree of influence of influencing factors, the propelling mechanism of evolving influencing factors, namely the conversion rules, could be constructed. Image prediction was run as shown in Equation (4) [33]:

$$X_{t+1}=f(X_t,N)$$

(4)

where X is the state of the land use cell, t and t + 1 are any t year and t + 1 year, respectively, N is the cell neighborhood, and f is the conversion rule.

3. Results

3.1. The Spatiotemporal Evolution Characteristics of LUCC

The three satellite image phases in the 5 years prior to and following regional tourism in 2011, 2016, and 2021 were monitored and categorized with the aid of ENVI 5.3 software. According to the LUCC categorization system, five different categories of land—farmland, built-up area, forest, water, and unused—were chosen, and ultimately the land use type photos of Panjin City’s Dawa District were obtained (Figure 2).

A field study was conducted to further compare and check the interpretation image accuracy and division outcomes of land use classification images in order to further clarify the interpretation accuracy of the land use interpretation images. More than 85% of the survey findings were similar, and the classification results were reasonably accessible.

Table 5. Area distribution of land use types in 2011, 2016 and 2021.

Land Type	Area in 2011 (hm²)	Area in 2016 (hm²)	Area in 2021 (hm²)
Built-up area	71,970.01477	76,714.85897	58,605.64768
Forest	23,670.95002	36,928.65714	47,425.50111
Farmland	29,982.20868	22,456.09545	26,337.96976
Water	24,340.67049	19,668.66897	25,028.7234
Unused	9515.706876	3713.39787	2083.836433

depicts the area distribution of land use types in 2011, 2016, and 2021. The data contain a calculation error of 2–3 hectares, although it had no effect on the data’s comparative analysis in large areas. The land use interpretation map was imported into IDRISI for Markov transfer probability matrix calculation in order to further analyze the conversion relationship among various land use types and form an in-depth analysis of the change trend of land use before and after whole-region tourism, as shown in

Table 6. Markov transition probability matrix of LUCC during 2011–2016.

Land Type	Farmland	Built-Up Area	Forest	Water	Unused
Built-up area	67.92%	18.19%	10.56%	2.31%	1.00%
Forest	16.38%	61.89%	16.39%	1.99%	3.35%
Farmland	44.38%	14.00%	38.29%	1.79%	1.55%
Water	12.49%	21.64%	1.59%	58.48%	5.80%
Unused	24.05%	40.88%	12.20%	11.93%	10.94%
Average	33.04%	31.32%	15.81%	15.30%	4.53%
Total transfer probability	165.22%	156.60%	79.03%	76.50%	22.64%

and

Table 7. Markov transition probability matrix of LUCC during 2016–2021.

Land Type	Farmland	Built-Up Area	Forest	Water	Unused
Built-up area	60.10%	18.60%	13.74%	7.16%	0.40%
Forest	7.60%	66.51%	17.42%	6.46%	2.02%
Farmland	12.45%	31.20%	53.47%	2.67%	0.21%
Water	3.51%	10.88%	1.78%	76.65%	7.18%
Unused	0.22%	44.06%	4.28%	34.39%	17.04%
Average	16.78%	34.25%	18.14%	25.47%	5.37%
Total transfer probability	83.88%	171.25%	90.69%	127.33%	26.85%

.

From 2011 to 2016, a comparison of the area data and transfer probability matrix data revealed a rising trend for both the built-up area and farmland. The probability of transferring to built-up area and farmland reached 156.60% and 165.22%, respectively. The forest land, water area, and unused land were progressively diminished, with a 44.38% probability of forest land being converted to farmland and a 14.00% probability of built-up area being converted. Water had a 12.49% probability of becoming farmland and a 21.64% probability of becoming built-up area. This observation demonstrates that during the first 5 years, the region primarily promoted modernization through built-up area and vigorously developed agriculture, as well as developing the natural ecological environment, such as forest land and water areas, and damaging the ecological environment barrier to some degree.

The regional tourism policy was introduced in 2016. From 2016 to 2021, the amount of farmland decreased, while the amount of built-up area increased. Simultaneously, forest land and water areas gradually recovered, and unused land continued to diminish. There is an 18.60% chance that farmland will become built-up area, a 13.74% chance that it will become forest land, and a 7.16% chance that it will become a water area. Built-up area has a 31.20% chance of becoming forest land, but forest land has a 17.42% chance of becoming built-up area. Overall, the probability of transferring built-up area was 172.25%; the probability of transferring forest land rose from 79.03% to 90.69%; and the probability of transferring water rose from 76.5% to 127.3%. This demonstrates that the research area exhibited the phenomenon of farmland reverting to forests and water and that the local construction industry progressively paid attention to the environmental protection and restoration project and took appropriate action. It is important to note that there is a high probability of conversion between forest land and built-up area in the region, which creates relatively unstable conditions.

3.2. Spatiotemporal Simulation of LUCC

3.2.1. Evolutionary Driver Factor Analysis

To further predict the evolution trend of land use and the evolution trend of landscape patterns in 2026, the driving factors of evolution are analyzed and discussed. These factors were divided into subjective qualitative analysis and objective quantitative driving mechanism creation of models. Since the research concentrated on the perspective of regional tourism, the driving factors were analyzed and selected from natural factors, human factors, and development external force factors, as shown in

Table 8. Evolutionary driver factor analysis.

Categories of Drivers	Driving Factors	Influence	Modeling or Subjective Analysis
Natural factors	Elevation or slope	The research area’s lowest and highest points are 25 m apart, and the average elevation is 2.7 m. The land is flat with little impact.	No influence
	Climate	The climate environment has a wide impact, and climate change in a short time and small region is modest. Thus, evolution has minimal impact.	No influence
	River system	Water resource development affects settlement location, habitat construction, and tourism development.	Modeling
	Road	Outside of the road communication research area, it includes the transportation of tourism resources, construction materials, and tourists, all of which have an effect.	Modeling
Humanistic factors	Population and economy	Tourism depends on population migration and economic support, but regional distribution data are hard to obtain, and thus they are subjectively analyzed.	Subjective analysis
	Folk culture	Folk culture is the research area’s tourism content and subject of subjective analysis.	Subjective analysis
	Policies and regulations	Policies and regulations affect tourist construction and macro-control in the research area. Content subjective analysis.	Subjective analysis
Develop external forces	Town center	Development and building focus on urban centers and constantly expand. Urbanization affects regional tourism.	Modeling
	Large scenic area	Small tourist areas can benefit from large picturesque sites.	Modeling

. We selected three humanistic driving factors for subjective analysis. Four quantifiable influencing factors were selected to construct the driving mechanism of influencing factors. The subjective analysis is included below.

1.

Population and Economy

We compared the data from the sixth census conducted by the Chinese Bureau of Statistics in 2010 with the data from the seventh census in 2020. The population of the Dawa District of Panjin City remained stable with a slight decline, dropping from 430,900 in 2010 to 422,800 in 2015. But in 2020, the urban population of the Dawa District would have grown by 58,700, or 62.23%, since 2010. A large number of the population in the region flowed from the rural to the city, and the speed of urbanization construction accelerated, which confirms the acceleration of urbanization construction in the data analysis of land use area analysis.

From 2011 to 2016, the Dawa District vigorously exploited oil fields and minerals, and the industrial added value of the secondary industry was significantly greater than that of the primary and tertiary industries, accounting for more than 50% of the total. Nevertheless, mineral resources are nonrenewable resources. With the continuous exploitation of oil, the output value of the secondary industry continues to decrease, and then the proportion of the industrial added value of the tertiary industry and the primary industry gradually increases. In the Dawa District, tourism development and revenue growth have become a significant source of economic revenue. The output value of the tertiary industry rose from CNY 7.796 billion in 2011 to CNY 14.62 billion in 2020. The region’s economy has benefited greatly from tourism. After 2016, the level of tourism development in the region is comparatively high, which has an effect on the evolution of the landscape pattern.

2.

Folk Culture

The Dawa District is rich in cultural and touristic resources. The extant folk culture and tourism resources in the Dawa region were analyzed in light of the theory of regional coordination and the current research content. The entire Dawa area can be divided into six core scenic spots with six characteristics, namely the central city business and tourism scenic spot, the Red Beach wetland scenic spot, the Guandi Temple, the war site red cultural scenic spot, the North Garden Bay pastoral scenic spot, the characteristic agricultural experience area, and the harbor scenic spot, in order to implement a radiation and diffusion layout. We finalized the sustainable tourism development of the city by maximizing the advantages of the folk culture and landscape functions of the various regions.

3.

Policies and Regulations

Using the longitudinal time comparison of the Dawa District of Panjin’s annual policies during the research period, the primary development policy directions from 2011 to 2021 can be roughly divided into three time periods. The first phase, 2011–2014, focuses on industrial construction and modernization development. It can be confirmed that, of the aforementioned population and economic factors, the industrial economy represents the largest proportion at this juncture, and the increase in construction land in the research area continued to rise. The 2015–2017 period constitutes the second phase, with the regional tourism to drive the modern service industry as the primary objective, the creation of a new tourism card for Liaoning Province, and assistance with ecological environment construction and urban-rural integration. To confirm the previous landscape pattern analysis, the environmental impact of industrial construction was steadily recovering at this stage. The 2018–2021 period is the third phase, which commenced the incremental transformation and upgrading of modern industries and the adjustment of all aspects of the industrial structure, taking comprehensive development as the primary objective, rather than industry or tourism, which were included in the development objective. Consequently, the accelerated development of modernization at this stage had a certain effect on the ecological tourism environment.

3.2.2. MCE-Based Suitability Atlas Production

The four quantifiable variables selected in Section 3.2.1 were applied to the MCE model of IDRISI’s suitability atlas. To construct the CA–Markov model’s conversion principles, preliminary image processing of objective quantitative data was carried out, as shown in Figure 3.

The selection driving factors were imported into ArcGIS 10.3 and analyzed in conjunction with the land use imagery from 2011, 2016, and 2021. Using the buffer zone analysis function of ArcGIS 10.3, data quantification was performed to form the distance-related area change function of various land use types under various drivers, and the function was applied to the production of the MCE model. Each land use type generated the MCE model image corresponding to it, which is the suitability image. The suitability atlas is a compilation of all MCE model images. The atlas depicts the future change suitability of various land use categories; the greater the value, the greater the probability of transferring to that type of land (Figure 4). After numerous tests, the evolution of undeveloped land was typically transferred to other land types, and the undeveloped land was less influenced by various influencing factors in the buffer zone calculation of ArcGIS 10.3. As depicted in Figure 4, the Markov-calculated transfer probability image was selected directly. Then, the suitability atlas applicable to the transfer rules was obtained.

3.2.3. CA–Markov Accuracy Calibration

Kappa is a multivariate statistical method that is frequently employed in land use simulation prediction to calculate the classification accuracy and validate the image accuracy [34]. On the basis of the land use images in 2011 and 2016, using CA–Markov simulation to predict the 2021 land use map in IDRISI17.0, we selected the Kappa coefficient to test the accuracy of the simulated prediction results in 2021 and the actual results in 2021 in order to validate the accuracy and functionality of the suitability atlas application. The results indicate that Kappa = 0.7597 > 0.75 and that the two images were comparable, indicating that the suitability atlas and simulation results were reasonably accessible.

3.2.4. 2026 LUCC Projection and Simulation

Based on the land use images in 2016 and 2021, the simulation prediction results for the 2026 land use images are shown in Figure 5. We obtained the Markov transition probability matrix for the years 2021–2026.

Based on the transition probability matrix data in the simulated prediction outcomes, an analysis was conducted. The area of built-up area expands from 2021 to 2026, assuming no other external influences, based on the present evolution trend, and the trend of returning farmland to forests and returning farmland to water remains unchanged. Therefore, more farmland is being converted into built-up area, forest land, and water area. The probability of transferring other land to built-up area reached 147.82%, followed by 103.15% of the continuously restored water area and 97.11% of forest land. Under this trend of evolution, if the rapid expansion of built-up area is not restrained, then it will pose a threat to the development of agriculture (

Table 9. Markov transition probability matrix of LUCC during 2021–2026.

Land Type	Farmland	Built-Up Area	Forest	Water	Unused
Farmland	58.50%	22.42%	11.74%	7.30%	0.04%
Built-up area	7.14%	70.44%	12.66%	8.74%	1.01%
Forest	4.86%	21.34%	71.63%	2.17%	0.00%
Water	0.01%	14.19%	0.64%	74.02%	11.14%
Unused	0.00%	19.43%	0.44%	10.92%	69.20%
Average	14.10%	29.56%	19.42%	20.63%	16.28%
Total transfer probability	70.51%	147.82%	97.11%	103.15%	81.39%
Area in 2026 (hm²)	43,595.90	54,692.09	31,603.38	27,423.80	2088.00

).

3.3. Analysis of Spatiotemporal Changes of Landscape Pattern Characteristic

3.3.1. Area Characteristics Analysis

The area index can comprehensively reflect the degree of landscape fragmentation, its distribution, and its dominance (Figure 6). The landscape type proportion (PLAND) for the period 2011–2021 was analyzed in the previous area transfer content and will not be analyzed again. However, it is important to note that the proportion of forest and water area in the later period gradually recovered, whereas the number of patches (NP) consistently increased considerably. The plaque density (PD) also proliferated, and the patch density in forested areas increased from 2.4142 to 4.7735 to 8.2502 patches per square meter. The density of the patches varied from 0.7432 to 1.3229 to 1.8871. This demonstrates that although the forest and water area have been partially restored, landscape fragmentation is increasing, and there are still issues with the internal ecological order. In 2026, the forest land and water area will be partially restored to their ideal state, a connected system will be established, and the degree of fragmentation will be reduced. The maximal patch index (LPI) indicates that the built-up area increases from 0.8479 in 2011 to 4.3789 in 2026, displacing farmland as the dominant landscape type in the region over the next 5 years.

3.3.2. Boundary Shape Characteristics

The index of the boundary shape reflects the complexity of the boundary shape, the human perturbation of the landscape, and the stability of the plaque (Figure 7). From 2011 to 2021, with the exception of unused land, the boundary shape index (LSI) of all land use types showed a consistent upward trend. It represents the boundary shape and becomes more complex. It demonstrates that during this time period, each patch was interwoven with one another with considerable human intervention, and the traces of human construction progressively increased. The larger the perimeter area dimension (PAFRAC), the more complex the shape of the plaque, which reflects the degree of interference of human activities with the landscape. The PAFRAC of farmland and unused land increased and then decreased. The built-up area, forest land, and water area all maintained increasing rates, but the increase rates decreased significantly after 2016, generally maintaining values between 1.4 and 1.5. This demonstrates that awareness of environmental protection and restoration is progressively growing, that appropriate measures have been implemented, and that the ecological environment is improving gradually. Nonetheless, it is still significantly impacted by human interference, and its stability is diminished. The ecological environment is still in a fragile state, and the landscape’s problems have not been fully resolved. Until 2026, under the present evolution trend and ideal ecological restoration conditions, the complexity of the boundaries between different patches is reduced, the PAFRAC value is maintained between 1.3 and 1.4, and stability has been somewhat enhanced.

3.3.3. Connectivity Characteristics Analysis

The connectivity index describes the connectivity, distribution, and aggregation of the landscape pattern and reflects the isolation and fragmentation, as well as the advantages of the landscape type (Figure 8). From 2011 to 2016, the vigorous construction of farmland and built-up area had excellent connectivity, and the conductance index (COHESION) for the farmland and built-up area was consistently above 99. The continuity of the forest land and water was poor, and the degree of fragmentation was considerable. Between 2016 and 2021, the separation between forest and water decreased, the degree of fragmentation moderated but remained high, and the aggregation index(AI) continued to fall. Combined with image observation, it is possible to estimate that, as a result of the expansion of built-up area, the farmland, forest, and water areas have become fragmented, resulting in a lack of connectivity. On this basis, it was observed that between 2011 and 2016, the separation of forest and water increased substantially and then decreased gradually from 2016 to 2021, and the local government began to pay attention to the separation and fragmentation of the protected landscape. However, the fragmentation degree of the remaining unused land increased, resulting in an increase in the overall separation degree, and thus the observation that the rational use of unused land is of equal importance. Simultaneously, according to the contag index (CONTAG) and overall separation (SPLIT), the entire research area developed from 2011 to 2016, and the built-up area and farmland developed as the most advantageous types in the research area, thereby reducing the overall spread and separation of the site. With the construction and development of forest land and water area in 2021, the dominant landscape on the site is subdivided, increasing the overall degree of separation and transforming the advantage of the single development of farmland and built-up area into the comprehensive development of multiple types. With the acceleration of the ecological construction process, a new combination of dominant landscape types is anticipated to emerge in 2026, and site connectivity will begin to recover to a certain extent.

3.3.4. Diversity Characteristics Analysis

Indicators of diversity can reflect the landscape’s diversity and abundance (Figure 9). The Shannon diversity index of 2011–2016 (SHDI) decreased from 1.5167 to 1.4569, and the Shannon evenness index(SHEI) decreased from 0.8465 to 0.8131, according to the analysis of the data. This indicates that the landscape type of the research area tended to evolve into a singular state of farmland and built-up area. The forest land and water area progressively recovered after the 2016–2021 implementation of a sustainable tourism policy, and the SHDI and SHEI rose to 1.493 and 0.8333, respectively. According to the present evolution trend, it is anticipated that the diversity index will continue to rise from 2021 to 2026, with the SHDI increasing to 1.5123 and SHEI increasing to 0.844. This indicates that the dominant landscape types have been diversified and landscape diversity has increased.

4. Discussion

4.1. The Spatiotemporal Evolution Characteristics of the Research Area from the Perspective of Regional Tourism

During the research period, studying and investigating the characteristics of the evolution of LUCC and landscape patterns under the direction of regional tourism policy cannot completely exclude the influence of other policies or external forces, such as epidemic factors. However, investigation and research have revealed that there were no major factors influencing large-scale direction guidance during the research period. Therefore, in the process of deliberation, objective factual data are respected, and the relative impact results of implementation of the regional tourism policy are discussed.

During the period from 2011 to 2016, the region focused on modernization and agricultural development, as evidenced by the expansion of built-up area and farmland, the conversion of forest land and water areas to built-up area and farmland, and the conversion of ecological land to urban construction and agricultural production. The forest land and water areas demonstrate the landscape evolution trend of landscape fragmentation, decreased stability, and poor connectivity, which cause numerous environmental issues. From 2016 to 2021, after the beginning of regional tourism, the forest land and water areas increased steadily, while the farmland area decreased significantly and was converted to woodland and water areas, demonstrating the evolution trend of “returning farmland to forest” and “returning farmland to lake”. Simultaneously, the built-up area continues to expand rapidly. Gradually, the landscape advantages of farmland and built-up area are eroded, the water area of forest land is restored, and the overall diversity and richness increase.

It is important to note, however, that the landscape pattern index indicates that the interference with and destruction of the internal ecological order and ecological basis were not mitigated by the rapid increase in forest land and water area. The degree of fragmentation continues to increase, and the stability and connectivity of forest land and water remain in a poor state, creating an imperative issue for sustainable tourism to solve. However, the emergence of these issues is not an isolated occurrence [3]. Due to regional tourism development’s pursuit of short-term policy dividends and economic benefits, it does not conduct reasonable analysis, evaluation, and research of land and landscape resources before blindly constructing in accordance with the policy, resulting in ecological environment degradation, destruction of landscape patterns, and other phenomena [4].

4.2. Simulation and Prediction Analysis of Land Use and Landscape Patterns in 2026

The land use and landscape pattern status of the studied area in 2026 were predicted using the CA–Markov model. The results indicate that the built-up area continues to expand, and the trends of “returning farmland to forest” and “returning farmland to water” persist. The forest land and water area continue to recover. The area of farmland continues to decrease as a result of modernization, construction, and tourism’s ecological restoration. Consequently, it was observed that in the process of constructing sustainable tourism, the versatility of land resources becomes the primary cause of land use conflict, and the transition of the land use mode is a significant manifestation of land use conflict [35]. In this conflict, at the expense of farmland, the overall ecological environment of the study area will be significantly restored by 2026. Forest land and water patches will continue to increase, connecting the fragmented patches into a system, reducing the fragmentation degree, increasing stability, and the dominant landscape in the site tending to diversify.

The prediction results show some ideality; that is, in future construction, the construction of road water conservancy facilities will expand the network, divide the patches, and have an effect on the network’s continuity, but simulation prediction results cannot accurately predict the artificially decisive road water conservancy network’s construction. Therefore, the ecological environment will not necessarily recover to the extent that is optimal. However, it can serve as a guide for the current land use planning by suggesting the artificial construction of an ecological network connected to a system and the integration of fragmented patches into a continuous whole, which can play an important role in restoring the internal ecosystem’s order [36].

4.3. Land Structure Optimization Strategies from the Perspective of Regional Tourism

4.3.1. The Government Regulates the Expansion of Building Land

According to the results of the prediction, the built-up area increased from 47,425.50 hm² to 54,692.09 hm². In the current social structure, there are multiple demands for various forms of competition over various land resources. There is a contradiction between the expansion of the built-up area and the demand for ecological restoration, as well as the preservation of farmland. Blind development and built-up area will not only exert great pressure on the ecological environment and contribute to plaque fragmentation, instability, and other negative effects, but it will also cause some harm to the characteristics of tourist destinations. This establishes new requirements for the government’s construction land planning strategy in an effort to prevent the disorderly expansion of urban and built-up areas [37]. Some studies propose playing the role of policy regulation, encouraging the advent of land expropriation, built-up area collective management, and reform of the homestead management system, stimulating the capitalization potential of land resources and resolving land use conflicts [38].

4.3.2. Protect Farmland and Optimize Land Use

Under the dual pressures of modernization and ecological environment protection, the area of 58,605.65 hm² in 2021 decreased continuously to 43,595.90 hm², fragmentation increased gradually, and stability and connectivity declined continuously, consequently regulating the area of farmland and optimizing the layout of farmland. Several studies have recently explored China’s cultivated land conservation policy [38], emphasizing that farmland is reinforced by separating unoccupied residential land, industrial and mining land, and underutilized land [39]. Given the conflict between farmland and forest in the Atlantic tropical rainforest, several researchers advocated dividing land for agricultural protection while continuing to expand tree planting in select portions of each farm to produce a win-win situation [40,41]. These policies can help to balance land resources, but differentiated land use requires differentiated policies or procedures to assure optimal land resource allocation, coordinate the interaction between people and land, and promote sustainable development [42]. Simultaneously, encouraging the use of modern technology and ways to improve the agricultural efficiency and output capacity should be included in present planning.

4.3.3. Construct the Ecological Corridor Network for Forest Land and Water Areas

The prediction results demonstrate that all types of plaque system construction, as opposed to fragmented construction, can enhance local ecological and environmental benefits to a greater degree. The simulation prediction results have an ideal ecological restoration strength in the absence of a significant external force, but the forest land and water in the region may not be restored to such an ideal state. Therefore, we can learn from the simulation prediction results, select similar concepts and strategies for an “elastic landscape” and “dissolved city” in the field of landscape science, and construct a continuous ecological network to restore the landscape pattern’s order. Then, we can make internal ecological resources realize mutual circulation and conversion and improve the site’s internal ecosystem stability [43].

5. Conclusions

In order to avoid pursuing short-term economic benefits and policy dividends, blind tourism development causes damage to ecosystems. This research examined the evolution trend of land use and landscape patterns in the 5 years before and after the regional tourism in 2011, 2016, and 2021 and used the CA–Markov model to simulate and predict the change in land use and land coverage in 2026. The research findings indicate that between 2011 and 2021, urban development was vigorously pursued in the study area, and that built-up area supplanted agricultural land as the most advantageous landscape type. After the commencement of regional tourism in 2016, farmland was subjected to the dual pressures of built-up area and forest land and water areas, resulting in a significant reduction in area. Despite the fact that forest land and water areas have expanded as a result of the development of tourism in the entire region, there are still problems of increasing fragmentation and severe human disturbance which have not been alleviated by the expansion. In 2026, it is anticipated that competition between various types of land use will continue to deteriorate, with the expansion of built-up area becoming increasingly disordered and the area of farmland continuously declining. The implementation of tourism construction facilitated by governmental policies reveals that while these policies aim to enhance ecological protection, the lack of scientific evaluation of land use and landscape patterns in tourism development and construction often results in a mere expansion of the area without effectively addressing the issue of high internal ecological vulnerability. As the efforts to enhance land functionality progress, the prominence of land rivalry and conflict will inevitably intensify. Hence, it is imperative to conduct planning research on land use and landscape patterns to mitigate the adverse effects of unsustainable development and land conflicts in the context of tourist cities’ sustainable development.

Meanwhile, this research still has some limitations. For example, in the process of research, it was found that in the process of tourism city development, different development directions of tourism resources in different urban areas are different. Although the changes in land use types are similar, they are not completely consistent. This also shows that policy formulation and planning need to be tailored to local conditions. In addition, in the process of driving factor analysis, the policy and other driving factors adopted subjective analysis. However, if the humanistic impact of the policy can be quantitatively evaluated, the accuracy of the simulation prediction can be greatly improved through the introduction of the prediction model. This quantitative method still needs to be explored further in the tests.

Based on the comprehensive study data results, regional development goals, and research limitations, this research study proposes a reference strategy for the existing social and economic development policy in the research area. This strategy involves the implementation of government macro-control measures to regulate the expansion of built-up area. It also emphasizes the need for planning to protect farmland and the establishment of an ecological corridor system encompassing forest land and water areas. Existing research indicates that tourist cities, akin to the research region under investigation, encounter comparable issues and require similar modes of thinking. The long-term objective of sustainable development in tourism cities should involve achieving a balance between grain production and ecological protection, as well as promoting the development of the tourism sector and implementing ecologically sustainable construction practices. Additionally, it is crucial to design management policies that are tailored to the specific local conditions.