Evaluation of Geotourism Potential Based on Spatial Pattern Analysis in Jiangxi Province, China

Abstract

To provide essential information on geoheritage and geotourism potential in Jiangxi Province—a key region for geoheritage distribution in China—this study summarizes and categorizes the types, grades, and distribution characteristics of geoheritage within local communities. The primary analytical methods included average nearest neighbour analysis, kernel density estimation, and spatial autocorrelation to explore spatial distribution patterns. A total of 202 significant geoheritage sites were identified in Jiangxi Province. Furthermore, an evaluation index system was established using the entropy weight TOPSIS model to assess the geotourism potential of each city. The findings reveal the following: (1) Geoheritage sites in Jiangxi Province exhibit an overall aggregated spatial distribution, although clustering intensity varies among different geoheritage types and grades. (2) Considering both grade and category, the core distribution area of geoheritage is located in eastern Shangrao City, while global-level geoheritage sites are mainly concentrated in the Poyang Lake Plain. (3) Spatial autocorrelation analysis indicates that, except for global-level geoheritage sites, other geoheritage sites display significant spatial agglomeration with positive spatial correlation. Moreover, local-scale spatial association characteristics differ notably according to geoheritage type and grade. (4) The geotourism development potential across Jiangxi Province shows clear spatial differentiation, with higher potential concentrated in the eastern and southern regions.

1. Introduction

In recent years, global awareness of geoheritage conservation has increased markedly. As an important approach to “development underpinned by conservation”, geotourism promotes the sustainable utilization of geoheritage resources through well-managed tourism activities. This approach not only enables the public to experience and understand the value of geoheritage conservation but also generates stable financial support for ongoing protection, thereby forming a virtuous cycle of protection–development–benefit–renewed protection.

Jiangxi Province is characterized by favorable natural geographical conditions, abundant geological resources, diverse geoheritage resources, and a rich geological landscape shaped by long-term geological process. In recent years, several Chinese geoscience researchers have begun to investigate geoheritage in Jiangxi Province, mainly focusing on micro-scale geoheritage resources and tourists’ perceptions within geoparks [1,2,3,4,5]. However, existing evaluation methods remain insufficient for analyzing the spatial distribution of geoheritage. Moreover, they are inadequate for assessing geotourism development potential across Jiangxi Province. This paper focuses on the spatial pattern of geoheritage and the potential for geotourism development in Jiangxi Province. It addresses the following three core research questions:

(1)

What multi-scale patterns and clustering characteristics characterize the spatial distribution of geoheritage in Jiangxi Province?

(2)

How does geotourism potential differ among the various prefecture-level cities?

(3)

How can differentiated strategies be proposed, based on spatial analysis and potential assessment, to promote the sustainable development of geotourism?

To address these questions, this study takes the 11 prefecture-level cities of Jiangxi Province as the basic research units. Spatial distribution patterns of geoheritage are systematically analyzed using geographical spatial analysis methods such as the nearest neighbour index, kernel density estimation, and global/local spatial autocorrelation. In addition, a multidimensional evaluation indicator system is established, incorporating geological endowment, tourism infrastructure and accessibility, socio-economic conditions, and ecological environment. The entropy-weighted TOPSIS model is then applied to comprehensively assess the geotourism development potential of each city. Finally, based on quantitative findings, this study proposes region-specific recommendations for geotourism that balance conservation with use. These recommendations aim to support high-level geoheritage protection and promote quality geotourism development in Jiangxi Province.

2. Literature Review

The concept of geotourism has undergone continuous refinement in recent decades [6]. Early definitions developed in the United Kingdom and Australia emphasized geology and landscape as the basic elements of geotourism [7,8,9,10]. For example, Newsome and Dowling defined geotourism as “a form of natural area tourism that specifically focuses on landscape and geology” [11]. National Geographic in the United States proposed a broader definition, describing geotourism as “tourism that sustains or enhances the geographical character of the place being visited, including its environment, culture, aesthetics, heritage, and the well-being of its residents” [12].

Geotourism is widely regarded as a form of sustainable tourism that is fundamentally based on the existence of geoheritage [13,14]. Geoheritage comprises valuable and non-renewable geological legacies that have formed, evolved, and been preserved throughout Earth’s long geological history through various endogenic and exogenic processes. Owing to the diversity of geological processes, many geoheritage sites possess significant scientific, educational, and esthetic value. These geoheritage landscapes are frequently developed for geotourism purposes, thereby contributing to local economic development [15,16,17]. Additionally, conservation generally refers to a range of actions applied to heritage sites, including preservation, restoration, rehabilitation, reconstruction, and adaptive reuse, either individually or in combination [18]. With the continuous advancement of the national geopark system and the growing importance of tourism within the national economy, geotourism—centered on geoheritage resources—has increasingly emerged as an important component of the sustainable tourism sector. Its primary objective is to achieve the coordinated protection and rational utilization of geoheritage resources [16,19].

The evaluation of tourism development potential is a crucial component of destination planning and management [20,21]. One of the primary motivations for assessing tourism resources is financial constraint, as municipal budgets are often limited. Therefore, evaluating the potential of regional tourism resources is essential for supporting government decision-making regarding the efficient allocation of funds and the sustainable development of tourism destinations [22]. A widely adopted approach in existing studies involves integrating multiple attributes of tourism potential into a single composite indicator using multi-criteria decision-making (MCDM) methods [23,24]. Numerous studies have applied MCDM approaches. For example, Iatu and Bulai employed a multiple linear regression model incorporating four variables, natural resources, cultural resources, tourism infrastructure, and general infrastructure, to predict tourist arrivals [25]. Mazanec et al. defined and measured tourism destination competitiveness based on a set of indicators, including heritage and culture, infrastructure, communication facilities, social competitiveness, environmental preservation, tourism prices, openness, and education levels [26].

Geographic Information Systems (GISs) are also widely used in tourism potential assessment and spatial analysis. Effat and Hegazy integrated MCDM with multiple spatial layers, which were appropriately overlaid to derive rankings and weights for tourist attractions and infrastructure [27]. González-Ramiro et al. evaluated rural tourism potential using a combined GIS and Analytic Hierarchy Process (AHP) approach, considering indicators such as accommodation availability, tourism activities, and natural environmental conditions [28]. To provide a spatial evaluation of geotourism potential, De Sena conducted a quantitative spatial assessment of the Lund Warming Ramsar Site (LWRS) and its geotourism prospects based on four variables, including the geological diversity index, visibility, capillarity, and the vulnerability of abiotic systems [29].

3. Methods and Data Sources

3.1. Study Area

Jiangxi Province is situated in central China, on the southern bank of the middle and lower reaches of the Yangtze River. It lies between 24°29′ N to 30°04′ N latitude and 113°34′ E to 118°28′ E longitude (Figure 1). The province is characterized predominantly by mountainous and hilly terrain. Situated within the northern subtropical zone, Jiangxi experiences a typical monsoon climate with distinct seasonal variations. From a geomorphological perspective, Jiangxi constitutes a major part of the Jiangnan Hills and is bounded by mountain ranges to the east, west, and south. The central region exhibits an interlaced pattern of hills and river valleys, whereas the northern part is dominated by the Poyang Lake Plain. Covering a total area of approximately 166,900 km², Jiangxi Province had an estimated population of 45.15 million at the end of 2023. Administratively, the province comprises 11 prefecture-level divisions and 100 county-level divisions.

3.2. Study Methods

The methods of research comprise two main components: (1) spatial pattern analysis of geoheritage, including average nearest neighbour, kernel density, and spatial autocorrelation analyses, and (2) evaluation of geotourism development potential, based primarily on the entropy-weighted TOPSIS model.

3.2.1. Average Nearest Neighbour

Average Nearest Neighbour (ANN) analysis is a spatial statistical method used to assess the proximity relationships and distribution patterns of point features within a given geographic area. By calculating the ratio between the observed mean nearest neighbour distance and the expected mean distance under a random distribution, ANN categorizes spatial patterns into three types: uniform, random, and clustered (aggregated) distributions [30]. In this study, ANN analysis was applied to identify the spatial distribution characteristics of significant geoheritage sites in Jiangxi Province.

$$R={\displaystyle \frac{r_i}{r_E}}$$

(1)

$$r_E={\displaystyle \frac{1}{2\sqrt{\frac{n}{A}}}}$$

(2)

In the formula, R represents the nearest neighbour index; $r_i$ is the average observed distance between geoheritages sites; $r_E$ is the expected observed distance of geoheritages; n is the number of geoheritage sites; and A is the area of Jiangxi Province. The interpretation of R is as follows: R > 1 indicates that the point features of geoheritage sites tend to be uniformly distributed, R = 1 shows that the geoheritages sites are randomly distributed, and R < 1 suggests an agglomerated distribution.

3.2.2. Kernel Density Estimation

Kernel density estimation is a non-parametric smoothing method for spatial point patterns, used to reveal variations in the density distribution of point features across continuous space [31,32]. Its core principle is to construct a smooth, bell-shaped surface (the kernel function) centered on each observed point with a certain bandwidth as the radius. By superimposing the kernel surfaces of all points within the study area, a continuous spatial density surface is formed, which visually reflects the spatial clustering and diffusion patterns of point features. Taking geoheritage as the analysis object, the kernel density estimation method was employed to analyze the spatial density characteristics of their distribution in Jiangxi Province:

$$f\left(x\right)={\displaystyle \frac{1}{nh}}\sum _{i=1}^{n}K\left({\displaystyle \frac{x-X_i}{h}}\right)$$

(3)

In the formula, f(x) represents the kernel density value of the geoheritage x to be evaluated, h is the search radius of point x and $X_i$ is the geoheritage within the search radius. In this study, the Leave-One-Out Cross Validation (LOOCV) method was employed to optimize this parameter; x − X_i is the relative distance from point x to $X_i$, $K$ is the kernel density coefficient, and n is the number of geoheritage sites. The larger the value of f(x), the higher the density of geoheritage sites.

By applying the kernel density analysis to the density estimation of significant geoheritage in the Jiangxi Province, when classified into seven levels using the geometric interval method, the spatial density visualization of significant geoheritage in Jiangxi Province becomes clearer.

3.2.3. Spatial Autocorrelation

Spatial autocorrelation analysis is a fundamental approach for identifying and quantifying the dependence of attribute values across geographic space. In this study, both global and local spatial autocorrelation analyses were conducted to investigate the distribution patterns of geoheritage site counts across county-level units in Jiangxi Province.

Global spatial autocorrelation reflects the overall degree of spatial association across the study area and is commonly measured using the Global Moran’s I statistic [33]. The statistic is defined as follows:

$$I={\displaystyle \frac{n\sum _{i=1}^n\sum _{j=1}^n{W}_{ij}(x_i-\overline{x})(x_j-\overline{x})}{{\sum }_{i=1}^n{(x_i-\overline{x})}^2{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}}}$$

(4)

In the formula, I is the Global Moran’s I index; $n$ is the number of county-level units in Jiangxi Province; $x_i$ and $x_j$ are the number of geoheritages sites in the i and j units, respectively; $\overline{x}$ is the mean value of the number of geoheritages in each unit; and $W_{ij}$ is the element of the spatial weight matrix constructed based on the Queen contiguity rule: $W_{ij}$ = 1 if units i and j share a boundary or vertex, otherwise 0. The spatial weight matrix is row-standardized. The value of I ranges from –1 to 1, where I > 0 indicates positive spatial autocorrelation, I < 0 indicates negative spatial autocorrelation, and I = 0 indicates a random spatial distribution.

Local Spatial Autocorrelation. To detect local spatial heterogeneity and identify clusters and outliers, the Local Moran’s I statistic was employed [34]. The statistic is calculated as follows:

$$I_i={\displaystyle \frac{n(x_i-\overline{x})\sum _{j=1}^n{W}_{ij}(x_j-\overline{x})}{{\sum }_{i=1}^n{(x_i-\overline{x})}^2}}$$

(5)

$I_i$ denotes the Local Moran’s I index for unit i and $W_{ij}$ is the row-standardized spatial weight matrix element; other symbols retain the same meanings as in Equation (4). Based on the sign of $I_i$ and its comparison with neighbouring units, four types of local spatial associations can be identified:

High–High (HH) cluster: high values in both the unit and its neighbours;

Low–Low (LL) cluster: low values in both the unit and its neighbours;

High–Low (HL) outlier: high value in the unit but low values in neighbouring units;

Low–High (LH) outlier: low value in the unit but high values in neighbouring units.

3.2.4. Entropy Weight TOPSIS Model

This study employs the entropy-weighted TOPSIS model, an improved version of the traditional TOPSIS method. The approach first uses the entropy weight method to determine the relative weights of each indicator and then applies the TOPSIS method for comprehensive evaluation [35]. By combining the objective weighting of the entropy method with the approximation to the ideal solution in TOPSIS, this model effectively reduces the influence of subjective factors and provides more robust results compared with using the entropy weight or TOPSIS methods alone. The main process of the model is shown below:

Firstly, the entropy weight method is used to calculate the weights.

The calculation steps are as follows.

First, convert negative indicators to positive ones, then standardize all indicators to eliminate dimension differences for data standardization processing. All indicators in the calculation process are positive indicators, as can be indicated by applying the following formulas:

$$Y_{\mathrm{ij}}={\displaystyle \frac{x_{ij}-min\left(x_j\right)}{max\left(x_j\right)-min\left(x_j\right)}}$$

(6)

In the formula,$Y_{\mathrm{ij}}$ represents the standardized data, and $x_{ij}$ represents the $j$ indicator value in the $i$ year.

Calculate the information entropy of each indicator. The formula is below.

$$E_j={\displaystyle \frac{-1}{\mathrm{l}\mathrm{n}n}}\times \sum _{i=1}^m{P}_{ij}\mathrm{l}\mathrm{n}{P}_{ij}$$

(7)

$$P_{ij}={\displaystyle \frac{Y_{ij}}{\sum _{i=1}^m{Y}_{ij}}}$$

(8)

where $E_j$ is the information entropy of the $j$ index. If $P_{ij}$ = 0, the ln($P_{ij}$) = 0 is defined.

Calculate the weight $W_j$ of each indicator. The formula is below.

$$W_j={\displaystyle \frac{1-E_j}{\sum _{i=1}^m{h}_j}}$$

(9)

Secondly, use TOPSIS to calculate the relative proximity. The calculation steps are as follows: Form a weighted matrix. First, form a standardized matrix, and then weight the indicators within it. The formula is below.

$$v_{ij}=b_{ij}\times W_i$$

(10)

Determine positive and negative ideal solutions.

$$v_j^{+}=\mathrm{m}\mathrm{a}\mathrm{x}\left(v_{ij}\right)$$

(11)

$$v_j^{-}=\mathrm{m}\mathrm{i}\mathrm{n}\left(v_{ij}\right)$$

(12)

Calculate the distance from each evaluated object to positive and negative ideal solutions.

$$d_i^{+}=\sqrt{\sum _{j-1}^n{\left({\nu }_{ij}-{\nu }_j^{+}\right)}^2}\left(i=1,2,3,\cdots ,m\right)$$

(13)

$$d_i^{-}=\sqrt{\sum _{j-1}^n{\left({\nu }_{ij}-{\nu }_j^{-}\right)}^2}\left(i=1,2,3,\cdots ,m\right)$$

(14)

Calculate the relative closeness of each subject to be evaluated. The formula is as follows.

$$c_i={\displaystyle \frac{d_i^{-}}{d_i^{-}+d_i^{+}}}(i=\mathrm{1,2},3,\cdots ,m)$$

(15)

3.3. Data Source

The research data used in this study comprise two main components: (1) spatial point data of geoheritage sites in the Jiangxi Province and (2) evaluation data for assessing the geotourism development potential of each administrative unit (

Table 1. Data types and sources used for evaluating geotourism potential in Jiangxi Province.

Type	Source
Geoheritage sites	GeoCloud (https://geocloud.cgs.gov.cn/)
The base map of the administrative region of Jiangxi Province and the DEM data with 30 m resolution	Geospatial Data Cloud (https://www.gscloud.cn/)
Number of National Geoparks	China Geological Survey (https://www.cgs.gov.cn/)
Number of A-level scenic spots	Jiangxi Department of Culture and Tourism (http://dct.jiangxi.gov.cn/)
Number of Star-Rated Hotels	China Tourist Hotel Association (http://www.ctha.com.cn/)
Online Travel Asset Index	China Tourism Academy (Data Center of the Ministry of Culture and Tourism) (https://www.ctaweb.org.cn/)
Domestic tourism revenue, domestic tourist arrivals, added value of the tertiary industry, per capita GDP, per capita disposable income, highway mileage data, permanent population, green coverage area, waste gas treatment facility treatment capacity, sewage treatment rate, among which the number of inbound tourists and tourism foreign exchange income are taken as an example in 2019	Jiangxi Statistical Yearbook (2019–2023) (Jiangxi Provincial Bureau of Statistics, 2019–2023), or Statistical Bulletin on National Economic and Social Development of Jiangxi Province (https://www.jiangxi.gov.cn/)
The number of geological hazard hidden danger points	Resource and Environmental Science Data Platform (https://www.resdc.cn/)

).

4. Analysis of Multi-Scale Spatial Pattern of Geoheritage

The widespread distribution and comprehensive variety of geoheritage in Jiangxi Province are attributed to its active tectonic setting, intensive endogenic and exogenic geological processes, distinctive geomorphic evolution, and complex patterns of hazard development. A total of 298 geoheritage sites have been documented within the study area, spanning two major categories, eight classes, and twenty-one subclasses [36]. Among these, 202 are identified as significant geoheritage sites. According to the classification system outlined in the Specification for Geoheritage Investigation (DZ/T 0303-2017) [37], 125 sites fall under the major category of geomorphic landscapes, while 77 belong to basic geology. In terms of hierarchical classification, the study area contains 19 global-level geoheritage sites, 62 national-level geoheritage sites, and 111 provincial-level geoheritage sites (Figure 2).

4.1. Spatial Equilibrium Characteristics

From an overall perspective, the average observed distance and expected observed distance for the 202 significant geoheritage sites are 12,571.35 m and 17,778.07 m, respectively. The nearest neighbour index is 0.71 (<1), with a Z-score of −7.96 (<−2.58), indicating statistical significance at the 99% confidence level and confirming a clustered distribution pattern (

Table 2. Calculation results of mean nearest neighbour of geoheritage in Jiangxi Province.

Category	Quantity	Average Observed Distance/m	Predicted Mean Distance/m	R	Z	p	Spatial Distribution Type
Total	202	12,571.35	17,778.07	0.71	−7.96	0.00	Clustering type
Geomorphic landscapes	125	17,475.05	22,165.13	0.79	−4.53	0.00	Clustering type
Basic geology	77	22,254.04	26,535.58	0.84	−2.71	0.00	Clustering type
Global-level	19	53,260.66	45,678.06	1.17	1.38	0.16	Random type
National-level	62	31,701.18	31,455.66	1.01	0.12	0.91	Random type
Provincial-level	111	17,352.26	22,713.92	0.76	−4.99	0.00	Clustering type

).

In terms of category, the geoheritage sites in Jiangxi Province are predominantly classified into two types, geomorphic landscapes and basic geology, both of which exhibit clustered distributions. The former comprises 125 sites and the latter 77. Their nearest neighbour ratios are 0.79 and 0.84, respectively, suggesting that geomorphic landscapes are slightly more aggregated than basic geology sites.

Considering hierarchical grade, the study area includes 19 globally significant and 62 nationally significant geoheritage sites. Their nearest neighbour indices are 1.17 and 1.01 (both >1), with Z-scores below 1.65 and p-values greater than 0.1, indicating that the null hypothesis of spatial randomness cannot be rejected. This reflects a tendency toward a random spatial distribution. In contrast, the 111 provincially significant sites exhibit a nearest neighbour index of 0.76, demonstrating a clearly clustered pattern.

4.2. Spatial Density Characteristics

The spatial density of geoheritage sites in Jiangxi Province exhibits pronounced differentiation. Overall, the distribution is centered in northeastern Jiangxi, with a secondary concentration in the northwestern region. These high-density areas primarily follow the Gan River in a belt-shaped pattern (Figure 3A). In the Ganbei region, geoheritage sites are densely clustered in Shangrao, Yingtan, and Jingdezhen, with secondary concentrations in Jiujiang, Yichun, and Xinyu. The central region (Ganzhong) shows a concentration along the fringe areas of Ji’an and Fuzhou, while Gannan is predominantly concentrated in southern Ganzhou.

This study analyzes the spatial density patterns of geoheritage sites by category and hierarchical grade (Figure 3). Notable differences emerge between the two major geoheritage types. The geomorphic landscape category exhibits four core density zones, primarily located in the Jiujiang–Yichun region, the Shangrao–Fuzhou area, Ji’an, and southern Ganzhou, with density diminishing radially from these centers. In contrast, the basic geology category displays dense clustering mainly around the Jingdezhen–Shangrao–Yingtan junction, along with two distinct locations in Xinyu City.

From a hierarchical perspective, the core distribution of global-level geoheritage sites is situated around the Jingdezhen–Shangrao–Yingtan junction and the northern part of Jiujiang. National-level geoheritage sites are primarily distributed around the Jingdezhen–Shangrao–Yingtan junction and the border area of Jiujiang, Yichun, Fuzhou, and Ganzhou, forming the main distribution cores. Provincial-level geoheritage sites are concentrated in eastern Shangrao City, forming an extreme-density nucleus.

In summary, the eastern part of Shangrao City emerges as the absolute core area of geoheritage distribution, while global-level geoheritage sites are mainly located in the Poyang Lake Plain.

4.3. Spatial Correlation Characteristics

4.3.1. Global Spatial Autocorrelation

Global spatial autocorrelation analysis was conducted using Moran’s I to assess the spatial correlation of significant geoheritage sites at the county level in Jiangxi Province. The results indicate a Moran’s I of 0.30, with a Z-score of 5.21 (exceeding the 99% confidence critical value of 2.58) and a p-value < 0.01. These results demonstrate significant positive spatial autocorrelation, indicating that geoheritage sites are spatially clustered and the pattern is statistically significant at the 1% level.

Analyzing by geoheritage category, both geomorphic landscapes and basic geology sites exhibit significantly positive Moran’s I indices (p < 0.01), confirming spatial clustering and marked aggregation.

Considering hierarchical grade, global-level geoheritage sites show a relatively low Moran’s I of 0.05, suggesting weak spatial clustering. In contrast, nationally and provincially significant sites display significantly positive Moran’s I values (p < 0.01), indicating pronounced spatial autocorrelation and strong aggregation across Jiangxi Province.

4.3.2. Local Spatial Autocorrelation

The clustering and outlier analysis tool in ArcGIS 10.8 was used to further explore the distribution patterns of geoheritage sites at the local scale in Jiangxi Province. The analysis categorizes local spatial associations into four statistically significant types: HH (high–high clusters), HL (high–low outlier), LL (low–low clusters) and LH (low–high outlier).

While geoheritage sites in Jiangxi Province display an overall clustered pattern, significant local spatial differentiation is observed (Figure 4). Specifically, HH clusters are located in southern Ganzhou, eastern Shangrao, and the western border area between Yichun and Jiujiang; LL clusters are distributed around Nanchang City, reflecting generally low site density in this region; an HL outlier occurs in Ji’an County; and LH outliers are found in Hengfeng County (Shangrao) and Dingnan County (Ganzhou). In terms of category, geomorphic landscape sites are mainly concentrated in the border area between Jiujiang and Yichun in northwestern Jiangxi, while fewer sites are distributed in the Poyang Lake area and Ji’an City. For basic geology sites, HH clusters are concentrated in eastern Shangrao, LL clusters are observed in Nancheng County of Fuzhou, and a large number of LH outliers are distributed in Ganzhou City.

Regarding hierarchical classification, global-level geoheritage sites show HL outliers in Shicheng County of Ganzhou, Jinggangshan City of Ji’an, and Ruichang City of Jiujiang, with LH outlier areas formed in northeastern and western Jiangxi. National-level sites exhibit significant spatial differentiation, including HH clusters and LL clusters. Provincial-level sites display HH clusters in locations such as Longnan County and Quannan County of Ganzhou, Wuning County and Xiushui County of Jiujiang, and Jing’an County and Tonggu County of Yichun, while LL clusters are distributed in Jishui County of Ji’an, Qingshanhu District and Nanchang County of Nanchang, and Yugan County of Shangrao.

4.4. Analysis of Spatial Pattern of Geoheritage

From the perspective of spatial equilibrium, geoheritage sites in Jiangxi Province generally exhibit an aggregated distribution, with geomorphic landscapes displaying a higher degree of clustering than basic geology sites. In terms of hierarchical classification, globally and nationally significant geoheritage sites tend to be spatially random due to their limited numbers and wide dispersion, whereas provincially significant sites show distinct clustering patterns.

Regarding spatial density, high-density cores are predominantly located along watersheds and within mountainous regions, reflecting an intrinsic relationship between site distribution and natural watershed boundaries, as well as the topographic and geological characteristics of surrounding geomorphological units.

With respect to spatial correlation, global Moran’s I analysis indicates pronounced positive spatial autocorrelation and clustering for geoheritage across Jiangxi Province. Both geomorphic landscapes and basic geology sites, as well as national-level and provincial-level sites, exhibit significant spatial autocorrelation, while globally significant sites show only weak clustering. Local autocorrelation analysis reveals HH clusters in southern Ganzhou, eastern Shangrao, and the western border area between Yichun and Jiujiang, with LL clusters concentrated near Nanchang City. Geomorphic landscapes and provincial-level geoheritage predominantly form HH clusters around the western junction of Jiujiang and Yichun, whereas basic geology and national-level sites form HH clusters in eastern Shangrao. Notably, globally significant geoheritage sites do not exhibit HH clustering.

5. Evaluation of Geotourism Development Potential

5.1. Indicator System Construction

Based on an understanding of geoheritage tourism resources and a review of the relevant literature on tourism development potential, this study considers the current state of tourism industry development and the construction of cultural tourism and heritage initiatives in Jiangxi Province. Adhering to the principles of scientific rigor, systematicity, and data accessibility, an evaluation index system for geotourism development potential was constructed across four dimensions: geological endowment, tourism infrastructure and accessibility, socio-economic conditions, and ecological environment [38,39,40,41,42,43] (

Table 3. Evaluation index system of geotourism development potential in Jiangxi Province.

Criterion Layer	Indicator Layer	Metric Weights	Attribute	Explanation and References
Geological endowment 0.2066	Number of geoheritage sites	0.0512	+	Number of geoheritage in the region
	Density of geoheritage sites	0.0181	+	The ratio of the number of geoheritage in a region to the area
	Geoheritage value	0.0470	+	5 × number of global-level geoheritage + 3 × number of national-level geoheritage + 1 × number of provincial-level geoheritage
	Resource combination advantage	0.0903	+	The number of national geoparks in the region
Tourism infrastructure and accessibility 0.4615	Tourist attraction	0.0773	+	5 × the number of 5A-level scenic spots + 3 × 4A-level scenic spots
	Tourist reception capacity	0.1587	+	5 × number of five-star hotels + 3 × number of four-star hotels
	Foreign exchange earnings from tourism	0.0439	+	—
	Domestic tourism revenue	0.0804	+	—
	Number of inbound tourists	0.0386	+	—
	Domestic tourism	0.0374	+	—
	Online Travel Asset Index	0.0252	+	The online asset performance index of destination tour operators published by the China Tourism Academy
Socio-economic conditions 0.2058	Added value of the tertiary industry	0.0249	+	—
	GDP per capita	0.0468	+	—
	Disposable income per capita	0.0364	+	—
	Number of permanent residents	0.0461	+	—
	Highway mileage	0.0516	+	—
Ecological environment 0.1261	Green coverage area	0.0392	+	The area of urban green space
	Waste gas treatment facility treatment capacity	0.0555	+	—
	Sewage treatment rate	0.0175	+	—
	Geological setting	0.0139	–	Number of geological hazard risk points in the area

+: Represents a positive indicator; –: Represents a negative indicator; —: Indicates no relevant explanation

).

Among these dimensions, geological endowment directly determines the uniqueness, scientific value, and market appeal of geotourism, forming the fundamental basis for tourism products. Tourism infrastructure and accessibility determine whether geological resources can be smoothly, comfortably, and safely transformed into tourist experiences, serving as the key to converting potential into reality [15]. Socio-economic conditions provide market demand, capital investment, intellectual support, and policy guarantees for geotourism development. The ecological environment constitutes the setting in which geological resources exist and tourism activities take place, influencing the quality of tourist experiences and the sustainability of the industry.

5.2. Evaluation Results

The geotourism potential evaluation indicators were first normalized. Subsequently, the entropy-weighted TOPSIS model was applied to assess the development potential of the 11 prefecture-level cities in Jiangxi Province. Based on the calculated scores, the potential levels were classified into five categories using the natural break method: high (0.412–0.482), middle high (0.343–0.411), middle (0.261–0.342), middle low (0.153–0.260), and low (0–0.152) (Figure 5).

In summary, the Gannan and Gandong regions are identified as the areas with the highest geotourism development potential in Jiangxi Province (Figure 5). Among the prefecture-level cities, Ganzhou exhibits the highest potential, with a score approximately three times greater than that of the lowest-ranked city, Jingdezhen (

Table 4. Evaluation results and ranking of geotourism development potential in Jiangxi Province.

Prefecture-Level Divisions	Ideal Positive Distance d+	Ideal Negative Distance d−	Relative Proximity Ci	Rank
Ganzhou	0.189	0.176	0.482	1
Fuzhou	0.187	0.166	0.471	2
Shangrao	0.191	0.154	0.447	3
Jiujiang	0.198	0.138	0.411	4
Yichun	0.190	0.112	0.370	5
Nanchang	0.219	0.114	0.342	6
Pingxiang	0.223	0.099	0.309	7
Yingtan	0.234	0.104	0.308	8
Xinyu	0.221	0.078	0.260	9
Ji’an	0.227	0.073	0.243	10
Jingdezhen	0.249	0.045	0.152	11

).

5.3. Analysis of Evaluation Results

The spatial differentiation of geotourism development potential in Jiangxi Province exhibits a distinct hierarchical pattern, which results from the combined effects of geological endowment, tourism infrastructure and accessibility, socio-economic conditions, and ecological environment.

As high-potential regions, Ganzhou, Fuzhou, and Shangrao possess core advantages over other cities in the province, thanks to the synergistic empowerment of high-quality geological resources and robust socio-economic foundations. Ganzhou, currently with the second-highest GDP in the province, leverages its large consumer market and strategic location as a regional transportation hub to efficiently connect with the tourist source market of the Guangdong–Hong Kong–Macao Greater Bay Area. Furthermore, its integrated development model of “red tourism + eco-tourism + geotourism” has fostered differentiated competitiveness. As for Shangrao, which ranked first in total tourism revenue in Jiangxi in 2023, it capitalizes on the first-mover advantage of its mature tourism cluster—the “Sanqing Mountain–Wuyuan–Guifeng” golden tourism route—seamlessly integrating geotourism into the existing tourism system, thereby achieving resource, infrastructure, and tourist sharing.

In contrast, regions with moderate geotourism potential (e.g., Nanchang, Pingxiang, and Yingtan) face disparate development bottlenecks, showing striking differences from both high- and low-potential areas. Although Nanchang is an economically strong city on par with Ganzhou, it is constrained by weak and scattered geological resource endowment. The geological landscapes within its territory, such as small karst caves and gentle hills, have relatively low scientific and ornamental value, and pale in comparison to global-level geoheritage sites like Sanqing Mountain and Longhu Mountain. This constitutes the core gap between Nanchang and high-potential regions. As for Pingxiang and Yingtan, despite hosting world-class geoparks and boasting a better resource base than Nanchang and Jingdezhen, their potential remains far from fully realized compared to Ganzhou and Shangrao, due to challenges such as low-level tourism product development and insufficient integration of geological and cultural elements. These issues represent their key weaknesses, distinguishing them from high-potential regions.

Jingdezhen represents the region with the lowest geotourism potential. It lacks both the economic strength of Nanchang and the world-class geopark resources of Pingxiang and Yingtan. Its tourism industry has long been dominated by ceramic culture, leading to the marginalization of geotourism in regional development strategies. Additionally, the single economic structure centered on the ceramic industry has resulted in a severe shortage of investment in geotourism. Coupled with problems such as lagging infrastructure construction and a scarcity of professional talent, Jingdezhen faces a development predicament far more severe than that of other regions.

The present study reveals significant spatial heterogeneity in the distribution of geotourism development potential across Jiangxi Province. Regions with relatively high to high potential account for 45% of the total, primarily clustered in the eastern and southern parts of the province, where they form contiguous zones. In contrast, areas with relatively low to low potential make up about 27% and are more spatially scattered. Overall, the province possesses favorable conditions for geotourism development, laying a solid foundation for its sustainable advancement.

6. Discussion

6.1. Comparison with Related Research Conclusions

This study integrates data on geoheritage and development potential for Jiangxi Province, applying spatial analysis and an entropy-weighted TOPSIS model to systematically assess the geotourism development potential and its spatial differentiation across the province. The results show that the overall geotourism development potential of Jiangxi Province is relatively high. This confirms that the province possesses a solid foundation and favorable resource endowment for geotourism development, providing core support for the large-scale and high-quality growth of the regional geotourism industry. Furthermore, the study reveals that the spatial distribution of geoheritage in Jiangxi is significantly constrained by natural factors such as topography. This finding aligns closely with the work of Luan and Wang [44], further corroborating the role of natural foundations in shaping the spatial patterns of geoheritage. Terrain and river systems influence the formation, preservation, and accessibility of geoheritage, thereby indirectly shaping the spatial framework of regional geotourism potential.

In summary, disparities in geotourism development potential among cities in Jiangxi Province stem primarily from the degree of match between geological endowment and socio-economic support capacity. High-potential regions have achieved a positive coupling of core development elements; moderate-potential regions are constrained by deficiencies in a single element; and low-potential regions face the dual challenges of both limited resources and economic weakness.

6.2. Recommendations for Geotourism Development

Based on the fundamental principle of integrating geotourism with the conservation and utilization of geoheritage, differentiated development strategies should be formulated according to local conditions. Specific recommendations are as follows: First, local distinctive geoheritage should be utilized to foster product innovation and specialization—for instance, by designing family-oriented geoscience educational tours. Second, geoheritage resources ought to be integrated in a scientific and effective manner to amplify the radiating and leading effects of well-known geosites. Finally, publicity, education, and public engagement efforts should be strengthened to raise societal awareness and build consensus regarding geoheritage conservation.

6.3. Limitations and Future Research Directions

This study has several limitations. First, the analysis is based on data from a single time period and may not fully reflect long-term trends in the geotourism development potential of Jiangxi Province. Second, although the evaluation indicators have been refined as much as possible, the influence of external variables cannot be fully eliminated. Third, due to constraints of time and resources, no qualitative investigation was conducted (e.g., regarding local community perceptions or the development status of specific sites). Finally, at the methodological level, this study employs the entropy weight method for objective weighting, which helps reduce subjective bias. However, no sensitivity analysis has been performed on changes in weights or data structure, which may affect the absolute interpretability of the final ranking scores and the robustness of the model. Future research could expand the data sample, incorporate insights from geotourism experts, and apply multiple weighting methods (e.g., Analytic Hierarchy Process or CRITIC method) or conduct perturbation analyses. This would help examine and optimize the evaluation results, thereby enhancing the study’s representativeness and timeliness.

7. Conclusions

Based on multi-source data and GIS technology, this study represents the first systematic integration of geospatial pattern analysis with a multi-criteria evaluation using an entropy-weighted TOPSIS model at the provincial scale in Jiangxi. The identified spatial differentiation patterns provide direct scientific support for formulating differentiated geotourism development strategies. Practically, the findings can guide the optimization of the geopark network, inform tourism infrastructure investment, and promote cross-regional collaborative marketing, thereby contributing to the sustainable development of geotourism in the province.

The main conclusions are as follows:

(1)

The geoheritage sites in Jiangxi Province exhibit an overall aggregated distribution pattern.

(2)

Significant differences exist in the degree of clustering among sites of different types and grades.

(3)

Key geoheritage sites are primarily distributed along the Gan River, its tributaries, and major mountainous terrains.

(4)

The eastern part of Shangrao City is identified as the absolute core distribution area for geoheritage.

(5)

Global spatial autocorrelation analysis indicates that, with the exception of global-level geoheritage sites, all other geoheritage sites in Jiangxi Province demonstrate significant spatial clustering and exhibit positive spatial correlation.

(6)

Local spatial autocorrelation analysis reveals high–high clustering areas in southern Ganzhou, eastern Shangrao, and the western border region between Yichun and Jiujiang, while low–low clustering areas are found around Nanchang. There are notable differences in spatial association characteristics among geoheritages of various grades and types.

(7)

The overall geotourism development potential of Jiangxi Province is relatively high. Cities with high and relatively high potential account for 45% of the total and are mainly concentrated in the eastern and southern regions, forming contiguous spatial patterns. In contrast, cities with low and relatively low potential represent about 27% of the total, are fewer in number, and display a more scattered distribution.