Mapping Spatial Interconnections with Distances for Evaluating the Development Value of Eco-Tourism Resources

Abstract

The sustainable development of eco-tourism is significantly influenced by multiple conditions within spatiotemporally continuous geographic scenarios. However, existing evaluations of the development value of eco-tourism resources (Eco-TRDVs) are non-spatial and do not sensitively represent their complex relationships. This study proposed a GIS approach for evaluating regional Eco-TRDVs by mapping the complex interconnections with spatial distances. Inherent and external conditions for evaluating Eco-TRDVs were classified under three indicators and digitized using GIS and remote sensing technologies. Then, the analytic hierarchy process and GIS cost distance analysis were introduced to define the initial values and cumulate Eco-TRDVs with distances. Taking the Taihang Honggu National Forest Park, China, as the case area, the Eco-TRDVs over the entire area in 2017 and 2020 were mapped. The results present a continuous spatial variability of Eco-TRDVs and comprehensively reflect the complex interconnections of constraint elements with spatial distances. The evaluation is sensitive to the intrinsic value of poles, as evidenced by the high development values and high-density distribution of their contours. Source additions improve the evaluation considerably, with transportation networks having a greater impact than economic development zones and urban elements. Furthermore, aggravated fragmentation of the price flow field increases spatial heterogeneity. The development value shows a negative linear correlation with distance. The proposed approach handles the spatially oriented relationships of the multi-conditions, and supports future planning and monitoring of spatial-temporal changes in eco-tourism development.

1. Introduction

Eco-tourism resources provide important contributions to the sustainable development of mountain regions with relatively underdeveloped economies [1,2]. It is important to comprehensively understand the development value of eco-tourism resources (Eco-TRDVs) for eco-tourism planning and sustainable management [3,4]. However, the traditional linear scoring methods are non-spatial and do not sufficiently consider the interconnection and cumulative effects of the influencing factors of Eco-TRDVs. Eco-TRDVs exhibit not only significant spatiotemporal heterogeneity but also spatial aggregation because of the complicated relationships between eco-attractions and the natural, social, and human systems [5]. As a result, mapping spatial interconnections with distances and evaluating Eco-TRDVs at the regional scale are challenging. These basic issues need to be resolved in order to formulate scientific eco-tourism policies and marketing strategies, and improve tourism management and, in turn, the healthy development of mountain areas.

Eco-TRDVs refer to the level of investment costs required for developing eco-tourist attractions, and it is linked to multiple conditions. Underdeveloped eco-tourism resources can be significantly enhanced by the addition of transportation infrastructure, human settlements, tourism support facilities, and urbanization. These conditions, together with eco-tourism resources, are the constraints of Eco-TRDVs. Accordingly, the evaluation criteria for Eco-TRDVs are fundamentally categorized into two interrelated types: inherent and external conditions. The inherent condition involves eco-tourism resources, which are determined by their endowment landscapes [6]. The external condition encompasses socioeconomic, human, and natural environmental factors. On the one hand, inherent conditions generate extremely high Eco-TRDVs. On the other hand, improved socioeconomic and human conditions will save costs and enhance Eco-TRDVs [7], but natural factors bring about impediments that would decrease the development value [8]. In addition, the distances to eco-attractions affect tourist choices and corresponding costs for development, thereby significantly influencing Eco-TRDVs [9,10,11]. These conditions are not independent, but rather contribute together to the evaluation [12]. This interconnection with distance defines their spatial configuration, and thus the spatial variability of Eco-TRDVs.

Many studies have employed linear weighted methods that presuppose a linear relationship among evaluation indicators and oversimplify spatial configurations to a number [13]. They overlook the non-linear relationships of evaluation indicators, which interact with each other and vary with locations. Some linear weighted methods rely on expert judgment to determine weights, which introduces personal bias. Many studies use mathematical methods, such as the analytic hierarchy process (AHP) and fuzzy mathematics, to objectively weigh the evaluation variables and perform an advanced evaluation [14,15]. However, such methods still do not facilitate regional mapping and they cannot respond to the inquiry of “where”. Furthermore, when eco-attractions are distributed in a concentrated manner, the spatially overlapping Eco-TRDVs and their diverse responses to the shared constraint conditions with distance are not given adequate consideration [16]. It is thus necessary to employ a geographic information system (GIS) spatial analysis to model these complex relationships with distance constraints.

Some studies have presented regional eco-tourism resources through thematic maps [17,18,19,20]. They interpreted evaluation indicators using remote sensing and processed their patterns using GIS digitization. This GIS mapping framework provides a reference for evaluation mapping. However, their key models still employ linear superposition, ignoring non-linear interconnections in geographical scenes. In addition, the impact of spatial distance on the evaluation still lacks characterization. For example, distance-dependent differences in tourists’ willingness and the required development cost are key considerations yet to be adopted. Furthermore, the state of infrastructure such as different transportation network conditions will exacerbate these differences. Together with external conditions, the distance to scenic spots inherently exhibits spatial variability and accumulation effects on Eco-TRDVs. It is a key issue that needs to be addressed in evaluation mapping.

The nonlinear interconnections with distances can be precisely addressed using the GIS cost distance analysis. It calculates the accumulative cost for each cell from or to the source over a cost surface [21]. In this context, the movement of a traveler is explored over a geographical scene and an accumulative cost–distance raster would be created using graph theory. Taking the cost–distance analysis as reference, the Eco-TRDVs negatively correlated with potential costs can be mapped by simulating tourist movement pathways. These pathways would determine the expenditure involved during the development of eco-attractions. In the calculation of Eco-TRDVs, the accumulation of the value with distance is calculated, which originates from multi-sources, traverses potential cost pathways, and reaches the eco-tourism destinations. The scarcer the infrastructure and supporting conditions, and the farther the distance to scenic spots, the higher the development costs, and the lower the Eco-TRDVs. Therefore, both tourist sources and eco-tourism resources can enhance Eco-TRDVs, whereas potential costs associated with pathways and distance undermine Eco-TRDVs.

There are three basic parameters of the cost–distance analysis: source raster, target points, and the travel cost raster. According to their roles in the Eco-TRDV evaluation, three indicators, namely, the source, pole, and flow field, are proposed. First, the source refers to infrastructure or population centers, and it represents the initial values for external Eco-TRDV accumulation. Second, the pole refers to tourism destinations, representing the target points of the cost–distance analysis. As the pole is endowed with substantial intrinsic values, it has extremely high inherent values. The term “pole” originates from the concept of a “pole” in graph theory. Third, the flow field refers to the cost raster, representing the travel cost associated with potential routes from sources to poles. Its existing impedance exerts negative impacts on external Eco-TRDVs. Therefore, the negative relationship between the cost and Eco-TRDVs necessitates the use of inverted values in the cost–distance analysis for the cumulative evaluation of Eco-TRDVs. In addition, each indicator is spatially composed of multiple components. Their contributions to Eco-TRDVs can be scored using the AHP method. On the whole, Eco-TRDVs are evaluated by accumulating index values, which begins with the source elements, cumulates with distance through the raster of flow field, and reaches pole landscapes.

This study aimed to evaluate Eco-TRDVs by mapping spatial interconnections with distances. The three indicators of the pole, source, and flow field were defined by the constraint roles for Eco-TRDVs. Then, evaluation maps of the development value of eco-tourism resources were conducted based on the AHP method and GIS cost–distance analysis. AHP was used to assign the weight and define initial index value for each element of the three indicators. The cost–distance analysis was performed to obtain the external cumulative value of development with distance. The final Eco-TRDVs are composed of the external Eco-TRDVs and the internal Eco-TRDVs from the initial pole value assigned via AHP in the previous step.

2. Study Area

The Taihang Honggu National Forest Park in Qinshui County of Shanxi Province, China, was selected as the case area, and its location is shown in Figure 1. The park encompasses the Taihang Honggu Block, Shanchanyan Block, and Longgang Block, which are located between 112°2′15″ E–112°13′08″ E and 35°26′32″ N–35°44′13″ N, covering a total area of approximately 269 km² [22]. The park features three geomorphological landscapes, namely, temperate karst, precipitous cliffs, and deep secluded valleys. Various hydrological resources are distributed across streams and valleys, with ponds, underground rivers, and suspended rivers representing the three primary types. The forest ecosystem exhibits typical warm temperate zone characteristics and occupies a unique position in the North China flora system. Unique microclimates have created distinctive meteorological landscapes, including cloud, fog, snow, rain, and sunrise landscapes. Among these, the cloud, rime, and ice waterfall landscapes are particularly extraordinary. Additionally, being located in the middle reaches of the Yellow River, one of the cradles of Chinese civilization, the park is rich in cultural heritage and has numerous cultural landscapes. In this study, the Taihang Honggu Block and Shanchanyan Block were selected as evaluation areas, whereas the Longgang Block was not considered because of its close proximity to Qinshui County.

3. Data and Methods

3.1. Data

Scenic spot elements of the pole indicator were firstly collected from historical national park planning documents. Then, an on-site visit and investigation was conducted with the park managers’ assistance in 2020, through which the landscapes of scenic spots for the study were finally verified. In 2017, the Taihang Honggu Block included six scenic spots (Figure 2): Karst Hole, Shenzheng Mountain, House of Jing Hao, Rime of Honggu Mountain, Honggu River, and Tree of Chinese Sophora japonica. In 2020, three new eco-attractions were added in the Taihang Honggu Block (Figure 3), namely, Wanggu Mountain, Shigu Rock, and Thirteen River, and two in the Shanchanyan Block, namely, Natural Bridge and Leopard.

Social and human elements of the source indicator were digitized from historical planning documents in 2017 and 2020. In 2017, some towns, villages, and one rural road were distributed in the southern region. In 2020, one new rural road was added, one economic development zone was upgraded, and one town was expanded, as shown in Figure 3.

Land use elements of the flow field indicator were obtained through the interpretation of 4 m resolution images in 2017 and 2020 released by Google Maps. The natural geographical factors for the flow field indicator, including altitude and slope, were obtained from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Model (GDEM) with a 30 m resolution (https://www.gscloud.cn). From 2017 to 2020, the most significant change in area within the southern part was the addition of forest land, which was primarily converted from cultivated land. These new forest land patches were mainly distributed in the northern and eastern parts of the Taihang Honggu Block, and in the surrounding areas of villages and farmlands over the Shanchanyan Block, as shown in Figure 3.

3.2. Methods

3.2.1. Methodological Framework

The procedure of the proposed evaluation includes the categorization of the evaluation indicators, acquisition of their element rasters, allocation of the weights by AHP, determination of the initial index values, accumulation of the extrinsic development value using the GIS cost–distance analysis and finally the calculation of the total development value. The methodological framework for the evaluation of Eco-TRDVs is shown in Figure 4.

In the above approach, three indicators, i.e., the pole, source, and flow field, are proposed. For each indicator, the elements were identified according to the effects of their constraints on development costs and their subsequent influence on development values. Firstly, the elements of the pole indicator are mainly eco-attractions with natural and cultural landscapes on which the eco-tourism resources depend. Target points for the cost–distance analysis were identified, including the determination of the internal Eco-TRDVs through their inherent value of landscapes. In addition, the elements of the source indicator involve the source locations of tourists, such as surrounding cities, towns, and communities, and all traffic networks and other facilities for transporting tourists. The more superior the available source elements, the higher the contribution of the development value to the poles. Finally, the elements of the flow field indicator refer to land use covers and physical geographic factors, which influence the difficulty and expenses of development. In the cost–value-based quantification, the accumulations start from the source elements, move through potential flow field elements in natural geographical scenes, and accumulate cell-by-cell until reaching the pole elements, i.e., scenic spots, as shown in Figure 5.

3.2.2. Categorization of the Evaluation Indicators

(1)

Pole

The pole indicator refers to objects in a geographical scene that have extremely high land prices at a specific time. As shown in

Table 1. Basic information of the evaluation indicators.

Indicators	Referenced Contents	Composition Elements	Type of Conditions	Role
Pole	Any scenic spot, referring to all the natural and cultural landscapes	Geomorphological resources, hydrological resources, biological resources, cultural resources, astronomical resources	Internal	Extremely high land prices and high Eco-TRDVs
Source	Sources of tourists and other sources	Distribution centers or places of origin of visitors, channels that transport tourists to scenic spots	External	Provide initial values for accumulation
Flow Field	Any geographical unit or potential route	Land uses, geographical factors	External	Determine investment difficulties and costs

, any scenic spot with endogenous aesthetic value, ornamental value, recreational value, or other specialized ecological values, such as national parks, nature reserves, and other protected areas [23,24,25], can be categorized as a pole element, including all the natural and cultural landscapes. Therefore, the pole indicator determines the internal Eco-TRDVs with their inherent conditions. The value of poles far exceeds that of any surrounding landscapes and resembles a mathematical extremum in Euclidean geometry. Based on the standard [26] and combined with field surveys, the intrinsic pole resources of the case area were classified into five major categories: geomorphological resources, hydrological resources, biological resources, cultural resources, and astronomical resources. Six representative scenic spots, including the Karst Hole, Shenzheng Mountain, Old Residence of Jinghao, Rime of Honggu Mountain, Honggu River, and Tree of Chinese Sophora japonica, corresponded to geomorphological, geomorphological, cultural, astronomical, hydrological, and biological landscape types, respectively.

(2)

Source

The source indicator refers to the sources of tourists and other sources that affect tourists at a certain time in the geographical scenarios, as shown in Table 1. On one hand, tourist sources indicate the distribution centers or places of origin of visitors, including surrounding cities and towns, economic development zones, and geographical areas with special tourism-supporting projects and facilities [27,28]. On the other hand, the source also refers to channels that transport tourists to scenic spots, that is, the traffic network [29]. The traffic network can be further subdivided into various units, such as expressways, national highways, provincial highways, county roads, township roads, dedicated tourist routes, and railways. These source elements determine the feasibility of developing eco-tourism resources, influence the investment cost for eco-tourism, and are attributed to one of the evaluation indicators for the external conditions of eco-attractions [30]. The larger the number of tourist sources near eco-attractions and the more convenient the traffic network, the lower the required investment cost, and the greater the Eco-TRDVs. According to the traffic planning and economic development statistics of Qinshui County and the forest park from 2017 to 2020, five types of elements were set for the source indicator, namely, expressways, arterial roads, rural roads, cities and towns, and economic development zones. Population size and economic development level, indicating the ability of the source to supply visitors, were characterized by the size of the population settlement.

(3)

Flow Field

The flow field indicator refers to any geographical unit or potential route that tourist sources pass through to reach poles, as shown in Table 1. It refers to the fundamental elements of tourism, which should be concentrated on spatial patterns [31,32]. The indicator is divided into two categories: land use and natural geographical elements. The flow field determines the investment cost and difficulty in developing eco-tourism resources and is another indicator for the external conditions of eco-attractions. Land use elements determine the compensation cost for land development and use, or determine the resistance due to protection policies [33]. The higher the land prices or the greater the policy constraints, the higher the cost of developing eco-tourism resources and the lower the development value. Furthermore, natural geographical elements determine the difficulty of tourists passing through natural geographic areas [34]. The higher the altitude and the steeper the slope, the stronger the resistance and the smaller the development value. The element of land use in the case park is divided into six classes: forest land, shrubland, grassland, water, residential land and farmland. In addition, the natural geographical elements include altitude and slope factors.

3.2.3. Scoring the Importance of the Elements of the Evaluation Indicators

The importance of the categorized elements of the three indicators should be firstly quantified so as to allocate their weights via AHP [14]. The relative measure applied by Saaty [15] was adopted to score the importance of evaluated elements. Saaty [15] suggested 9 importance levels and assigned values from 1 to 9 through a pairwise comparison of evaluation elements. From 1 to 9, the importance increases progressively, of which 9 represents extreme importance. The results constitute the judgment matrix of AHP. If the importance score of one element is rated as ${\mathrm{a}}_{\mathrm{i}\mathrm{j}}$ when compared with another element, then the importance score of the latter element relative to the former shall be ${\displaystyle \frac{1}{{\mathrm{a}}_{\mathrm{j}\mathrm{i}}}}$. The scoring basis for importance is shown in

Table 2. Scoring basis of the evaluation elements in terms of importance.

Evaluation Indicator	Scoring Basis
Pole	Criteria or standards of tourism resource classification and experts’ experience
Source	Statistical reports of the tourism industry released by authorities and experts’ experience
Flow field	Local policies of current natural resource protection and utilization, as well as compensation standards for expropriation or requisition for non-freehold estates during the evaluation period

.

Initially, the pole elements and their relative importance were determined by one of the criteria or standard of tourism resource classification and expert knowledge of their intrinsic value. According to the standard of [26], the pole elements in the case park can be optionally divided into five categories, namely, geomorphological landscape, cultural landscape, hydrological landscape, astronomical landscape, and biological landscape. Considering the ecological significance of different types of eco-tourism resources and their attractions to tourists, their importance can be quantified as 9, 7, 5, 3, and 1, respectively (

Table 3. AHP judgment matrix table of the pole indicator.

Element	Biological Landscape	Astronomical Landscape	Hydrological Landscape	Cultural Landscape	Geomorphological Landscape	Weight
Biological landscape	1	3	5	7	9	0.035
Astronomical landscape	0.333	1	3	5	7	0.068
Hydrological landscape	0.200	0.333	1	3	5	0.134
Cultural landscape	0.143	0.200	0.333	1	3	0.260
Geomorphological landscape	0.111	0.143	0.200	0.333	1	0.503

CR = 0.054, which is less than 0.1, passing the test.

).

Secondly, the influence magnitudes of each element contained in the source indicator were identified using statistical reports of the tourism industry released by authorities and expert knowledge. The larger the traffic system near scenic spots, the more the visitor transportation and the higher the score. In addition, areas with a larger resident population and a higher economic level have higher visitor attraction and hence a higher score. Between the traffic system and the human system, traffic networks are considered to have a stronger impact on the development value than human residences. In the case area, five elements of the source indicator are distributed, namely, economic development zones, cities and towns, rural roads, expressways, and arterial roads. Arterial roads refer to provincial and national roads. This element is assigned the highest score because of its maximum transportation capacity for short-haul tourism in mountain areas. Therefore, based on the above scoring rule, their scores are accordingly defined as 1, 3, 5, 7, and 9, as shown in

Table 4. AHP judgment matrix table of the source indicator.

Element	Economic Development Zones	Cities and Towns	Rural Roads	Expressways	Arterial Roads	Weight
Economic development zones	1	3	5	7	9	0.035
Cities and towns	0.333	1	3	5	7	0.068
Rural roads	0.200	0.333	1	3	5	0.134
Expressways	0.143	0.200	0.333	1	3	0.260
Arterial roads	0.111	0.143	0.200	0.333	1	0.503

CR = 0.054, which is less than 0.1, passing the test.

.

Thirdly, the relative importance of different elements of the flow field indicator was defined using the compensation standards of expropriation or requisition for non-freehold estates during the evaluation period, as well as policy impediments for environmental protection [35]. The cost and resistance of developing lands for eco-tourism were determined by natural resource protection, land utilization policies, and compensation standards [36]. Hence, forest land has the highest development difficulty and cost, followed by shrubland, and then grassland. In villages, the compensation price per acre for residential land is slightly lower than that for a water body or woodland but higher than that for farmland. The flow field indicator in the case area contains six land use elements, namely, forest land, shrubland, grassland, water body, residential land, and farmland. Forest land has the highest difficulty and cost of development, followed by shrubland and grassland. In terms of value accumulation with reference to the inverse correlation of the development cost, these elements can be ranked as follows: forest land < shrubland < grassland < water body < residential land < farmland, corresponding to importance scores of 9, 7, 5, 4, 3 and 1, respectively, as shown in

Table 5. AHP judgment matrix table of the flow field indicator.

Category	Element	Forest Land	Shrubland	Grassland	Water	Residential Land	Farm Land	Weight
Land use	Forest land	1	3	4	5	7	9	0.029
	Shrubland	0.333	1	3	4	5	7	0.053
	Grassland	0.250	0.333	1	3	4	5	0.091
	Water	0.200	0.250	0.333	1	3	4	0.149
	Residential land	0.143	0.200	0.250	0.333	1	3	0.248
	Farmland	0.111	0.143	0.200	0.250	0.333	1	0.430
Natural geographical factor	Altitude and slope							DDC’

CR = 0.064, which is less than 0.1, passing the test.

. For example, forest land poses the greatest resistance to development, incurring the highest costs and thus contributing the least to Eco-TRDVs.

3.2.4. Weight Calculation for the Evaluation Elements of the Three Indicators

The AHP procedure [14] was introduced to calculate the weights for assigning the initial index value of each element. Larger weights indicate lower initial index values and then high development values, and vice versa. Based on the above importance scoring of the evaluation elements, AHP judgment matrices were first established for elements of the pole, source, and flow field indicators, respectively, as shown in Table 3, Table 4 and Table 5. Subsequently, the approximate values of the eigenvectors were calculated by the root method. The weight of each element was calculated through standardized calculations, in which the sum of all element weights for each indicator equals 1.

For the pole indicator, the weights of biological landscape, astronomical landscape, hydrological landscape, cultural landscape, and geomorphological landscape were 0.035, 0.068, 0.134, 0.260, and 0.503, respectively, as shown in Table 3. For the source indicator, the weights of economic development zones, cities and towns, rural roads, expressways, and arterial roads were 0.035, 0.068, 0.134, 0.260, and 0.503, respectively, as shown in Table 4. For the flow field indicator, the weights of forest land, shrubland, grassland, water, residential land, and farmland were 0.029, 0.053, 0.091, 0.149, 0.248, and 0.430, respectively, as shown in Table 5.

The above process reflects hierarchical single ranking for determining weights, which requires a consistency check. The consistency ratio (CR) of the judgment matrices was tested using the consistency index (CI) and the average random consistency index (RI). The CR calculation formula is as follows [37]:

$$\mathrm{C}\mathrm{R}={\displaystyle \frac{\mathrm{C}\mathrm{I}}{\mathrm{R}\mathrm{I}}}$$

(1)

where CR is the consistency ratio, CI is the consistency index, and RI is the consistency index. Generally, the judgment matrix is considered to have passed the consistency test if CR < 0.1, suggesting that the above results are reliable. The AHP-derived CRs for the pole and source indicators were identical at 0.054, while the CR for the flow field indicator was 0.064. With CR values less than 0.1, all indicators passed the consistency test.

For the natural geographical elements of the flow field indicator, their weights were calculated by the development difficulty coefficient (DDC). It was calculated by combining the slope and altitude as follows [38]:

DDC = (sin(SP) × H^0.3 + 0.1)^0.3

(2)

where DDC is the difficulty raster, SP is the slope, and H is the altitude, extracted from ASTER GDEM. The higher the DDC value, the higher the resistance to development and the higher the weight. When considering value accumulation in the context of the inverse correlation with development costs, DDC should be inverted to change high cost values to low index values, and the inverted result of DDC is denoted as DDC’, as shown in Function (3):

DDC’= −1 × DDC + m

(3)

where DDC’ is the weight of natural geographical factors, and m is the maximum value of DDC. In the study area, the value of m was 1.90.

3.2.5. Definition of the Initial Index Value for Evaluation Elements

The initial index values of the pole, source, and flow field indicators were determined on the basis of the abovementioned weights. Following the principle of greater weights corresponding to higher development values, the index value for the following distance accumulation was scaled to 1–100 by multiplying by 100 and integer rounding. According to

Table 6. Initial index values of the three indicators.

Indicator	Element Index		Index Value
Pole	Biological landscape		4
	Astronomical landscape		7
	Hydrological landscape		13
	Cultural landscape		26
	Geomorphological landscape		50
Source	Economic development zones		4
	Cities and towns		7
	Rural roads		13
	Expressways		26
	Arterial roads		50
Flow field	Land use	Forest land	3
		Shrubland	5
		Grassland	9
		Water	15
		Residential land	25
		Farmland	43
	Natural geographical factor	Altitude and slope	100 $\times$ DDC’

, for the pole elements, the initial index values for biological landscape, astronomical landscape, hydrological landscape, cultural landscape, and geomorphological landscape were 4, 7, 13, 26, and 50, respectively. Regarding the source elements, the initial index values for economic development zones, cities and towns, rural roads, expressways and arterial roads were defined as 4, 7, 13, 26, and 50, respectively. For the land use elements of the flow field indicator, the initial index values for forest land, shrubland, grassland, water, village, and farmland were 3, 5, 9, 15, 25, and 43, respectively. In addition, the initial index value for the natural geographical factor was defined as $100\times \mathrm{D}\mathrm{D}\mathrm{C}\mathrm{’}$, as shown in Table 6.

3.2.6. Accumulation of the Development Value for Evaluation Mapping

The GIS cost–distance accumulation algorithm was used to evaluate Eco-TRDVs at the regional scale in the geographic scenario. It simulates the accumulation process of the development value from the sources to the poles through grid cells of the flow fields [39]. With the pole points as the target points and the above rasters of the initial index value of the source and flow field, the accumulated external Eco-TRDVs $V\_external$ from source to pole grid cells were calculated. The raster of the source index value is taken as the starting value of Eco-TRDV accumulation. Meanwhile, the raster of the flow field index value serves as the cumulative value for Eco-TRDV accumulation, which is the summation of the land use index value and the natural geographic index value. Then, the total Eco-TRDVs were determined by adding the accumulated external Eco-TRDVs and the internal Eco-TRDVs. This can be expressed as follows:

$$V=V\_internal+V\_external$$

(4)

where V is the final Eco-TRDVs; V_internal is the internal Eco-TRDV raster, which is built using the above initial index values of pole elements; V_external is the external Eco-TRDV raster.

The impact of distance on the accumulated value is fully expressed in the GIS cost–distance analysis model, which considers spatial distance as one of the modeling parameters. The accumulated value of each cell was calculated by considering not only the path cost from the current cell to the adjacent previous cell but also the distance between two adjacent cells. The core function of the GIS cost distance accumulation is shown in Function (5):

$$\mathrm{a}\mathrm{c}\mathrm{c}\mathrm{u}\mathrm{m}\_\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}\_\mathrm{b}=\mathrm{a}\mathrm{c}\mathrm{c}\mathrm{u}\mathrm{m}\_\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}\_\mathrm{a}+\left(\right(\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}\_\mathrm{a}+\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}\_\mathrm{b})/2)\times \mathrm{d}\mathrm{i}\mathrm{s}\_\mathrm{a}\mathrm{b}$$

(5)

where accum_cost _b is the accumulated path cost of the current cell b; accum_cost_a is the accumulated path cost of the adjacent previous cell a; cost_a is the path cost of cell a; cost_b is the path cost of cell b; and dis_ab is the distance between two adjacent cells a and b, which is calculated by the centers of the two raster cells. If the length and width of a cell are D, then the distance between adjacent cells to the left, right, top, and bottom is D, and the diagonal distance between adjacent cells is 2D. In particular, if the cost value is NODATA, then it would be automatically avoided during the cost–distance analysis. In this study, cost is inversely related to the weight and index value. During the value accumulation process, the cost raster in the GIS cost–distance analysis is replaced by the index value rasters of the flow field indicator that has a positive relationship with Eco-TRDVs.

4. Results

4.1. Mapping of Eco-Tourism Development Value

Images of the development values in 2017 and 2020 with a 30 m × 30 m cell size showed a continuous spatial distribution, with variations over or near different eco-attractions, as shown in Figure 6a and 6b, respectively. The evaluations started with the tourist sources, including county towns, economic development zones, and roads, passed through the flow field with cell-by-cell accumulations, and then reached the poles where the scenic spots are located. Higher development values were observed in the Taihang Honggu Block and Shanchanyan Block, where scenic spots are concentrated, and they gradually decreased from these blocks toward the surrounding areas and then nearby county towns and economic development zones. Regions far from scenic spots, such as Qinshui County and its surrounding densely populated areas, showed the lowest development value owing to their poor eco-tourism resources. For two spatially adjacent scenic spots, the development value from one scenic spot to another generally showed a trend of a gradual decrease followed by an increase.

With the improvement of infrastructure and economic development, the eco-tourism resources of the study area were significantly enhanced from 2017 to 2020. Areas with an increased development value covered 259.16 km², accounting for 96% of the total area, and areas with a decreased development value covered only 10.08 km², accounting for 4% of the total area. Therefore, the development value increased in the majority of the study area. The average development values in 2017 and 2020 were 316 and 357, respectively, with standard deviations of 140 and 165. The overall development value in 2020 was 13% higher than that in 2017, and the standard deviation of the development value in 2020 was 18% higher than that in 2017.

4.2. Effects of the Evaluation Indicators

4.2.1. Pole

Pole cells have higher development values compared to other geological raster points over the entire region owing to their endogenous values of eco-tourism resources. The average development values of major scenic spots in the Taihang Honggu Block were 543 and 642, respectively, which were 72% and 80% higher than those of the entire region for the same periods. The regional statistics also showed that the development values over the Taihang Honggu Block and Shanchanyan Block and their surrounding areas were significantly higher than those of the entire region. Here, in 2020, the minimum, maximum, and average development values in the Taihang Honggu Block were 506, 654, and 571, respectively; the minimum, maximum, and average development values in the Shanchanyan Block were 559, 642, and 612, respectively. In 2020, two new scenic spots, named Natural Bridge and Leopard, were added in the Shanchanyan Block, which improved its development value. The highest development value of scenic spots suggests that poles significantly affect the evaluation of the development value.

Poles not only have high development values but also drive the enhancement of development values for their surrounding areas. This enhancement is better represented by the trend in contour lines extracted from development value raster at an interval of 10, as shown in Figure 7. First, the contour lines of development values clustered around scenic spots, with dramatic changes near the scenic spots. Second, the contour lines grew gradually sparser from the densest spots towards surrounding areas, and the farther from the scenic spots, the lines were more widely spaced. In 2017, the development value dropped sharply from 550 to 330 in a query along the vertical direction of the contour line near the Rime of Honggu Mountain and Honggu River scenic spots, as shown in Figure 7a. In 2020, the development value also rapidly decreased from 640 to 470 in a query near the Honggu River, as shown in Figure 7b. Lastly, the values for contour lines from one spot to another neighboring spot typically first decreased and then increased. For example, in 2017, the values attributed to the contour lines first gradually decreased from 540 to 470 and then increased up to 560 from the Rime of Honggu Mountain to Honggu River. The distribution characteristics of the contour lines indicate that the development value changes drastically near the poles and smooths out further away from the poles.

4.2.2. Source

While poles play a significant role, source changes also remarkably affect the distribution of local Eco-TRDVs. Eco-TRDVs presented a significant increase with the construction of traffic network elements. In 2020, a new rural road was built in the southern part of the Taihang Honggu Block. This new source element runs from east to north, passing successively through the Tree of Chinese Sophora japonica, Honggu River, Rime of Honggu Mountain, and House of Jing Hao. Thereafter, the development value showed increases of over 100 in scenic spots in the Taihang Honggu Block, such as the Rime of Honggu Mountain, Karst Hole, Shenzheng Mountain, House of Jing Hao, Thirteen River, and Wanggu Mountain. Taking the scenic spots in the Taihang Honggu Block as examples, their development values and changes are shown in Figure 8. As observed, the scenic spot with the most noticeable increase in development value was the Rime of Honggu Mountain, with an increase of 117; followed by Karst Hole, with an increase of 105; Shenzheng Mountain, Honggu River, and House of Jing Hao, with increases of around 100; and finally, the Tree of Chinese Sophora japonica, with an increase of 84. This indicates that even a rural road with the lowest weight in the traffic network can significantly enhance the Eco-TRDVs in the area, assuming that other conditions are relatively stable. The evaluation is sensitive to changes in source elements.

The results reflect a comprehensive non-linear relationship among the evaluation elements, which contradicts the simple linearity employed in previous approaches with linear weighted calculations. The distances from the new road to scenic spots and their evaluated development values do not exhibit a linear relationship of an increase or decrease. Moreover, the increase in development value between the two years also did not show a linear relationship, as shown in Figure 8. Among the six scenic spots, the Rime of Honggu Mountain showed the shortest distance at only 40 m away but the largest increase in development value from 2017 to 2020. The attribute of its closest distance compensated for its low scoring of landscape as an astronomical landscape. On the contrary, the Karst Hole was the farthest from the new road at 1176 m, but the increase in its development value was second only to the Rime of Honggu Mountain. This could be attributed to the highest intrinsic value and the increased value derived from the surrounding source of visitors. Next, Shenzheng Mountain, the House of Jing Hao, and the Honggu River showed similar changes in development values under the combined effects of distances from the new road and endogenous resources. Finally, the distance of the Tree of Chinese Sophora japonica from the new road was smaller than that of the Rime of Honggu Mountain at 50 m. However, as a biological landscape, it had the smallest endogenous value of only 4 and thus the smallest increase in development value. As evident, while the addition of source elements can enhance the development value of eco-attractions, the final magnitude of increase is a result of a combination of multiple elements, including the distance from the constraint element to its scenic spots.

In addition to the new rural road, one economic development zone was also developed in the study area and the area of one town was expanded in 2020. A line buffer analysis with a radius of 500 m around the new rural road revealed that the increases in the development value in the buffer zone had a mean of 85, a minimum of 39, a maximum of 126, and a standard deviation of 25. On the other hand, a polygon buffer with the same radius around the regions of increasing economic development and expanded town led to smaller increases with a mean of 42, a minimum of 35, a maximum of 54, and a standard deviation of four. These statistical indices showed that the addition of the new traffic road more significantly enhanced the development value compared to the addition of the economic development zone and town expansion with other conditions being equivalent. Therefore, the source indicator is highly sensitive to traffic network elements.

4.2.3. Flow Field

In addition to pole and source indicators, flow field elements serve as circulation pathways for estimation, determining the spatial distribution in areas distant from both poles and sources. The above new source elements were transformed from the flow field elements that have high values, which improved the entire development value of eco-tourism resources, but the broken land use elements intensified the spatial heterogeneity of the distribution of the development value. The standard deviation of the development value over the study area was higher by 18% in 2020 than in 2017. In the study area, the total number of land use patches increased from 370 in 2017 to 480 in 2020, with the total perimeter increasing from 1,432,150 m in 2017 to 1,456,639 m in 2020. Furthermore, in the southern part of the study area, the total number of land use patches increased from 54 in 2017 to 79 in 2020, with the total perimeter increasing from 105,738 m in 2017 to 150,467 m in 2020. During the two years, some areas of forest land were converted into road and rural areas, while others were interconverted between forest land and farmland. The decentralized transformation led to an increase in the number of element patches and exacerbated the fragmentation of land use elements, which would increase the spatial heterogeneity of the regional development value.

Unlike land use elements, natural geographical elements were relatively stable and remained unchanged during the evaluation period. Changes in the development value and its increase with slope and elevation distinctly differed in the northern and southern parts, as shown in Figure 9a, b, separately. In the northern part, where the terrain is relatively flat, the development value and its increase showed a slight linear decrease with slope and altitude. At a significance level of p = 0.05, the development values in 2017 and 2020 and the increases showed Pearson’s correlation coefficients (r) of −0.32, −0.32, and −0.29 with slope, and −0.34, −0.39, and −0.53 with altitude, respectively. In the southern part with higher slopes and altitudes, the development value and its increase showed a slightly pronounced linear increase with increasing slope and altitude. At a significance level of p = 0.05, the Pearson correlation coefficients (r) between the slope and the development values in 2017 and 2020 and their increases were 0.07, 0.22, and 0.36, and for altitude, the coefficients were 0.12, 0.33, and 0.53, respectively. Owing to the intrinsic value of poles, the development value of these samples tended to trend upward with DDC’, as shown in Figure 9c, although poles were generally distributed in areas with higher slopes, higher altitudes, and lower DDC’. The influence of natural geographical elements is more pronounced on natural geographical areas far away from eco-attractions, whereas it is not significant in areas where eco-attractions are concentrated.

4.2.4. Spatial Distance

The development value is essentially a result of dynamic changes in and the spatial accumulation of evaluation indicators with the Euclidean distance. The correlation between the development value and distance from scenic spots and the new road were ascertained from the scatter plots. Firstly, as shown in Figure 10a, the development values in both 2017 and 2020 showed clear downward linear trends with distance from the scenic spot. They had an r² of 0.96, indicating a high correlation. The closer the distance from the scenic spots, the larger the development value. Moreover, the increase from 2017 to 2020 also showed an overall downward linear trend with distance, and their r² was 0.68, indicating a moderate correlation. The farther the distance from the poles, the larger the decrease in the development value. Secondly, the distances of the randomly selected 211 sampling points to the new road were further calculated. Scatter plots with the distance and the development value in 2017 and 2020 are shown in Figure 10b. They also revealed negative linear correlations with r² values of 0.95 in 2017 and 0.94 in 2020, both of which indicate strong correlations. Furthermore, the r² for the increase was 0.64, indicating a moderate correlation. As described, the two types of distance had linear effects on the development value and its dynamic changes. In addition, the distance to poles had a slightly stronger effect than the distance to new sources.

It is noteworthy that the good linear relationship across the full range in Figure 10a,b considerably weakened in the enlarged scatter plots, which were within a distance range of 2500 m. In this case, the r² of the linear relationship between the distance to poles and the development value in 2017 and 2020 decreased to 0.58 and 0.65, respectively, and the R² for the increase decreased to 0.34. Meanwhile, the r² of the linear relationship between the distance to the new road and the development value in 2017 and 2020 decreased to 0.29 and 0.19, respectively, and its r² for the increase decreased to 0.01. A second-order curve fitted to the scatter points revealed that when the scatter points started from the scenic spots or new road, the development value initially showed a clear downward trend, but as the points moved away from the scenic spots or the new road to around 1500 m, the original linear downward trend gradually flattened and then transformed into an upward trend. It can be ascertained that the spatial change in the development value is not affected only by distance but it is instead a spatial accumulation of the value of the poles and other elements with distance.

5. Discussion

5.1. Effectiveness of the Presented GIS Approach in a Regional Evaluation

Varying natural, social, human, and economic conditions would control the spatial heterogeneity of Eco-TRDVs. This study evaluated eco-tourism resources on a regional scale rather than a single number, which is the result of the traditional non-spatial scoring methodology. According to the classical scoring assessment for the case park [40], the development value had a final score of only six in 2017. This singular value represents the development value of the entire region and thus cannot indicate regions with a high value for eco-tourism resource development and those with a low value for development. It also cannot sharply respond to the regional differences in development value over areas with changes in economic conditions and infrastructure facilities. In addition, it is difficult to answer the ‘how’ and ‘why’ questions, i.e., not supporting a comprehensive interpretation. Aiming to overcome these shortcomings, this study used GIS techniques, defined evaluation variables in geographical and temporal processes, and quantitatively expressed the continuous changes in the development value over the entire area. The grid data and map algebra in the GIS domain were employed to map the development value with a unit of 30 m of geographical points. Therefore, the evaluation can support any regional statistics with upward scales, such as regional evaluations with administrative units or other zonal units of any arbitrary boundary.

This study evaluated the development value in terms of both the intrinsic value of eco-tourism resources and the value determined by the costs incurred by external elements. As mentioned by previous studies [5,12], important aspects of rural tourism development include ecological resources, tourism, traffic networks, environmental conditions, and supporting facilities. The complicated nonlinearity of their relationships, the interactions of their effects, and the accumulation of their contributions with geographical distance can only be incorporated into the evaluation by relying on GIS raster data, raster calculation, and spatial analysis. Therefore, this study adopted existing research ideas [12,33,41,42] focused on the “how”, “why”, and “where” questions of tourism geography, and realized the spatial-temporal evaluation in geographic scenarios. The proposed approach is superior in dynamic evaluations for temporal processes because changes in elements in economic, social, cultural, and natural conditions could be simply integrated using GIS spatial updating and merging functions. The cost–distance analysis is not constrained by the input data. It is compatible with any evaluation variable that can be mapped using GIS and remote sensing techniques. A computer-executable and generalized module allows the rapid realization of regional evaluations by simply transforming the input raster datasets and AHP scoring parameters. When switching to new geographical regions or scenic spots, it only requires input data to be updated and the discriminant matrix to be updated. The main body of the method remains unaffected by these changes.

5.2. Interpretation of Variables’ Influences on the Evaluation

The relationships between eco-tourism resources and economic, social, and human systems are complex and are often overlapping and entangled [43,44]. At first, new source elements clearly improved the development value for nearby areas, especially in eco-attractions. According to previous studies, the traffic network is very important for mountain tourism in sparsely populated areas. Therefore, this study set the initial value of roads to be greater than that of economic development areas and townships. Secondly, a large amount of eco-tourism resources are concentrated in existing scenic spots. As destinations, these areas have much lower development costs than those rich in natural resources but require new landscapes to be converted to tourism destinations. Therefore, this study defined existing attractions as poles with the highest development value. Thirdly, the land for new pole or source elements comes from individual, collective, or state lands, which require purchasing, compensating, and breaking protection policies. Protection policies of forest land in most of the national parks pose the biggest challenges to development compared to other land use types. Therefore, the cost of expropriating forest land for new pole or source elements is the highest, and, accordingly, it carries the lowest index value.

With increasing distance from the scenic spots, the development value shows a downward trend, which is consistent with the law that tourism benefits are negatively correlated with distance [45]. Spatial distance is a crucial aspect of tourism demand, influencing the travel intention of tourists. In addition, the convenience of transportation is another significant aspect affecting tourist decisions in mountain tourism. Thus, in the investment and development of tourism resources for a certain tourist attraction, the distances to source and flow field elements are crucial aspects for consideration [46,47]. Accordingly, the distance from social and human systems (i.e., the source and flow field elements) to tourist attractions (i.e., the pole elements) should be a parameter for modeling the evaluation of the development value. Furthermore, the superposition of multiple pole, source, and flow field elements can be addressed by adopting the AHP method. The spatial variation in these accumulated effects of pole, source, and flow field elements with distance can only be obtained by relying on the GIS spatial cost–distance analysis. Such efforts are necessary to support the evaluation of the development value at the regional scale.

5.3. Authenticity Test Using Contemporary Land Prices

The evaluation results for the development value were verified with land price changes [48,49]. Some studies have proven that tourism development can significantly promote the rise of land prices. According to this finding, increases in the tourism value of the study area can drive up the land prices of surrounding areas in the southern part, including the Taihang Honggu Block and Shanchanyan Block. Spatial statistics reveal that the average development value in the southern part was 460 in 2017 and 539 in 2020. The average development value in 2020 increased by 17% over that in 2017, which agrees with the increase in land price in the area over the same period of time. According to data from the Chinese land market website, the land price in 2020 increased by nearly 11% compared to the average price in 2016. In detail, five land transactions in 2016 had an average price of approximately 219 CNY/km², including land prices of 220 CNY/km², 218 CNY/km², 218 CNY/km², 217 CNY/km², and 220 CNY/km², respectively, in Xiangyang Village, Shangchuan Village, and Xiachuan Village, Zhongcun County in the southern part and the surrounding areas. Then, in 2020, one land transaction was performed in Tuwo Village, at a price of 242 CNY/km². Additionally, no other large-scale investment projects significantly increased land prices in the area from 2016 to 2020. Hence, the upward trends with similar increase ratios indicate that the evaluation coincides with the actual situation and it is thus rational.

5.4. Clarification of the Relationship Between Conventional Understanding and This Study

The cost in the cost–distance analysis is replaced by the index value. The conventional cost refers to the consumption of time, money, and human resources [50], while the path grid in the cost–distance analysis refers to the index value of external conditions. The cumulative effects of the source, pole, flow field, and distance on the development value be unified only in this manner. If the cumulative number is defined according to the original meaning of “cost”, contradictions will arise. For instance, suppose that source elements at the same distance from a pole traverse forestland and agricultural land. If the cumulative value is calculated based on the high development cost imposed by forestland, it will exceed that of agricultural land. Evidently, this presents a contradiction with the inherent high development value of agricultural land.

In conventional understanding, transportation networks are often regarded as paths to target points. However, this study deviates from this consideration in that transportation networks are classified into the category of sources. This is because transportation networks, as developed infrastructure, serve as the origin of tourists. Development costs are incurred when new tourism-supporting facilities are developed from existing transportation networks, thereby influencing the development value.

5.5. Uncertainty and Limitations

The practical significance of the evaluation relies on the spatiotemporal differences in evaluation elements. The proposed evaluation methodology is only applicable to the evaluation of eco-tourism resources in non-urban areas. In urban areas, traffic networks are well-developed, and the development investment is minimal for areas meeting basic accessibility conditions. In general, after establishing different land use types (e.g., residential and industrial zones), corresponding travel characteristics and traffic demands will be generated and the spatial location and scope of various economic activities will be determined, thereby causing changes in spatial accessibility. The more reasonable the land use planning, the higher the concentration of land use and the more places that can be reached within a certain distance (time). As a result, more opportunities increase accessibility and reduce the investment cost for development, and they would ultimately lead to a high development value. This is clearly not consistent with the reality of eco-tourism. Therefore, the proposed method is only suitable for evaluating tourism resources in non-urban areas and not urbanized areas. Furthermore, varying AHP scoring criteria may yield distinct numerical outcomes for Eco-TRDVs. Nevertheless, the ordinal relationships inferred from these values hold greater relevance for mapping applications than their absolute magnitudes. Hence, this study refrained from performing a sensitivity analysis of the scoring criteria.

6. Conclusions

A scientific, comprehensive, and accurate evaluation of the Eco-TRDVs is an important prerequisite for formulating reasonable development planning and making facility development decisions for scenic spots. This study proposed a methodology for the evaluation mapping of Eco-TRDVs based on the GIS cost–distance accumulation algorithm and AHP. The “non-linear” spatial accumulation process was taken into account on a regional scale, and the evaluation results could suitably reflect the cross-reactive effects of distance and different constraint conditions.

The evaluation results, using geographical grid cells as units, reflect a nonlinear spatial cumulative dynamic process from sources, through potential paths, to scenic spots. The Eco-TRDV raster was derived from the interplay of evaluation indicators with spatial distance. Its spatial pattern is significantly influenced by the intrinsic resources of scenic spots. In particular, when other conditions are relatively stable, the addition of transportation networks within the source indicator notably enhances the development value, outperforming economic zones and urban elements. Moreover, the spatial heterogeneity of the development value is greatly affected by the fragmentation degree of land use patches. At the regional scale, Eco-TRDVs exhibit strong negative linear correlations with distances to scenic spots and new sources.

The evaluated grid data conform to requirements for the refined spatialization of eco-tourism, which has significance as scientific reference for government and industry departments to formulate reasonable eco-tourism development plans and regional development plans. If the eco-attractions change with time, more dynamic elements can be input into the evaluation, which is facilitated by the generalizability of the proposed approach. In addition to the evaluation of the development value, the proposed method can provide technical insights for assessing the development conditions of ecological resources, determining the resource protection value for planning targeted tourism investment, land-use zoning, infrastructure prioritization, and even priority areas in nature conservation.

The proposed methodology is more suitable for evaluating the development value of eco-tourism resources in underdeveloped areas, but it is inapplicable to urban regions with a highly developed infrastructure. In future research, data such as the road network density and nighttime light data can be employed to differentiate between rural and urban areas and establish separate evaluation technical frameworks. With the spatial resolution of GIS and remote sensing data improving continuously, the discriminative power for mountain microtopography and traffic networks is being enhanced. The proposed method thereby enables evaluation mapping with higher precision, offering greater significance for guiding rural development planning at finer spatial scales. Integrating high-resolution remote sensing imagery and advancing dynamic evaluations to finer spatiotemporal scales are also promising directions for future research. Nevertheless, finding a balance between multiple conditions remains an unresolved challenge, especially for remote areas that have abundant resources but are difficult to access. The findings of this study can provide a spatial reference for identifying areas suitable for development.