Multi-Scale Remote Sensing Analysis of Terrain–Resilience Coupling in Mountainous Traditional Villages: A Case Study of the Qinba Mountains, China

Abstract

Mountainous traditional villages represent unique socio-ecological systems that have evolved through centuries of adaptation to complex topographies and multi-hazard environments. Understanding their terrain–resilience coupling mechanisms is essential for risk-sensitive planning and heritage preservation in mountainous regions. This study integrates multi-source remote sensing data and GIS spatial analysis to investigate 57 national-level traditional villages in the southern Qinba Mountains, China. Using kernel density estimation (KDE), nearest neighbor index (NNI), and Geodetector modeling, we identify the spatial distribution characteristics and topographic driving forces that shape settlement patterns across macro-meso-micro scales. Results reveal that 83% of the villages are clustered in low-mountain and hilly zones (550–1200 m elevation), preferring slopes below 15° and south-facing aspects. Elevation exerts the strongest influence (q = 0.46), followed by slope (q = 0.32) and aspect (q = 0.29), forming a multi-level adaptation framework of “macro-elevation differentiation, meso-slope constraint, and micro-aspect optimization.” Morphological Spatial Pattern Analysis (MSPA) further indicates that traditional villages achieve ecological balance and disaster avoidance through adaptive spatial strategies such as terrace-based flood prevention, convex-bank stabilization, and platform-based hazard avoidance. These strategies are not merely topographic preferences but natural adaptation mechanisms formed by long-term responses to multi-hazard environments—dynamic adaptation processes that reduce disaster exposure and optimize resource use efficiency through active adjustment of site selection and spatial transformation (the disaster density in the 100m core zone buffer is 0.077 events/km², significantly lower than 0.290 events/km² in peripheral areas). These findings demonstrate that remote sensing techniques can effectively reveal the terrain–resilience coupling of traditional villages, providing quantitative evidence for integrating spatial resilience into cultural landscape conservation, ecological security assessment, and rural revitalization planning. The proposed multi-scale analytical framework offers a transferable approach for evaluating settlement adaptability and resilience in other mountainous cultural heritage regions worldwide.

1. Introduction

Mountainous traditional villages represent unique socio-ecological systems that have evolved through centuries of adaptation to complex topographies and multi-hazard environments [1,2]. Understanding their terrain–resilience coupling mechanisms is essential for risk-sensitive planning and heritage preservation in mountainous regions. The southern Shaanxi Qinba Mountains, located at the transition between northern and southern China, serve as both an important ecological barrier in the upper Yangtze River basin and a key repository of national cultural heritage [3,4,5]. With over 70% of its area covered by mountains, the region is characterized by steep terrain, fragile geological conditions, and frequent natural hazards such as landslides, debris flows, and floods [6,7]. Since the implementation of China’s disaster-avoidance relocation policy in 2011, millions of residents have been resettled from high-risk mountain zones to low-lying dam areas. Although this large-scale relocation has effectively reduced immediate hazard exposure, it has also caused a series of problems, including population loss [8,9], cultural discontinuity [10], and landscape degradation in traditional villages [11]. These challenges highlight the limitations of a single relocation-oriented approach and emphasize the necessity of developing in situ disaster prevention and resilience enhancement strategies that integrate local ecological wisdom and spatial adaptability [12].

Traditional settlements are self-organized spatial adaptation systems formed by long-term human–environment interaction. Their site selection patterns, land use methods, and architectural morphologies not only reflect reasonable preferences for topographic conditions but also embody natural adaptation mechanisms that maintain system stability by proactively responding to disaster risks, which is the core carrier of Traditional Ecological Knowledge (TEK) [10,13,14]. International research from the Alps, Andes, and Japan’s Shikoku region has demonstrated that mountainous settlements sustain resilience through terrain-sensitive design, hydrological management, and adaptive construction practices. However, studies in China have primarily focused on southeastern coastal, loess, and karst regions [15,16,17], while the Qinba Mountains—one of the country’s largest and most typical mountainous cultural regions—have received insufficient attention. Existing research has mostly examined cultural conservation, ecological restoration, or disaster prevention in isolation, without systematically exploring the coupled mechanisms between terrain factors, spatial distribution, and resilience formation. Although there is growing interest in spatial resilience theory, few empirical studies have quantitatively analyzed the relationship between topography and settlement adaptability. Research that integrates remote sensing, GIS, and spatial statistics to reveal how elevation, slope, and aspect jointly influence settlement patterns remains limited. Moreover, while the importance of relocation and ecological restoration has been widely acknowledged, the spatial mechanisms underlying in situ disaster prevention and adaptive planning are still underexplored. Addressing these gaps is essential for linking heritage preservation with risk-sensitive territorial governance and sustainable rural revitalization in mountainous China.

To address these challenges, this study investigates 57 nationally recognized traditional villages in the southern Qinba Mountains to identify their spatial characteristics and terrain–resilience coupling mechanisms at multiple scales. Specifically, the study aims to (1) quantitatively identify spatial clustering patterns and topographic drivers influencing village distribution; (2) analyze terrain-driven adaptive strategies that reduce multi-hazard exposure and enhance spatial resilience; and (3) develop a transferable analytical framework for multi-scale resilience assessment in mountainous heritage regions. Accordingly, this study addresses the following research question: How do terrain factors interact to shape the spatial resilience of traditional villages in mountainous regions? By integrating remote sensing data, GIS spatial modeling, and Geodetector analysis within a social-ecological system (SES) framework, this research investigates the coupling relationships between natural terrain and settlement resilience across macro, meso, and micro scales. The findings are expected to deepen the theoretical understanding of human–environment coevolution and provide spatial evidence to support in situ disaster prevention, ecological security, and adaptive village planning in mountainous regions of China and beyond (Figure 1).

2. Theoretical Basis and Literature Review

2.1. Spatial Patterns of Mountainous Villages

The spatial distribution of traditional villages is shaped by the combined effects of natural environment, economic development, and sociocultural factors [18]. In mountainous regions, topography and disaster risk are among the primary determinants of settlement configuration [19]. Previous studies have shown that mountainous settlements tend to follow the principle of “adapting to terrain and relying on water,” forming several distinctive spatial types such as linear settlements along river valleys, dispersed settlements on terraces, and clustered settlements on gentle hillslopes [20]. Such patterns reflect both a passive adaptation to topographic constraints and an active selection based on agricultural production and livelihood needs.

With the advancement of geographic information science (GIS) and spatial statistics, the spatial pattern of villages has become increasingly amenable to quantitative analysis. For instance, the Nearest Neighbor Index (NNI) has been widely used to assess whether village distributions exhibit random or clustered characteristics [21]; Kernel Density Estimation (KDE) can reveal spatial hotspots of settlement concentration [22]; and measures such as the geographical concentration index and imbalance index are useful for evaluating spatial equity and dispersion [23]. The introduction of landscape ecology has further enriched analytical approaches, employing indicators such as landscape fragmentation, patch density, and connectivity to explore the interaction between settlements and natural landscapes [24]. These methodological advances signify a shift from descriptive analyses to multidimensional quantitative approaches in the study of traditional settlement patterns.

2.2. Natural Adaptability and the Social-Ecological Resilience Framework

The Social-Ecological System (SES) framework provides a robust theoretical foundation for analyzing the natural adaptability of traditional villages. Holling [25] first introduced the concept of resilience to describe an ecosystem’s capacity to maintain its structure and functions after external disturbances. Later, Walker et al. [26] expanded this idea to social–ecological resilience, emphasizing the dynamic feedbacks and co-evolution between human and natural systems.

In the context of traditional villages, natural adaptability manifests primarily in two dimensions: (1) Spatial and architectural adaptation-settlement siting and spatial layout that minimize exposure to hazards; (2) Socio-environmental coordination—resource use, agricultural practices, and social organization that promote long-term environmental harmony. For example, in rural settlements of Xinjiang, China, communities have continuously adjusted their spatial organization and land-use structure to cope with geomorphic and climatic hazards, illustrating the interactive evolution of human and ecological systems [13]; Similarly, traditional engineering practices such as terracing, convex-bank construction, and platform-based settlement building serve as effective measures for flood avoidance, soil retention, and hazard mitigation [27]. These practices embody a form of eco-wisdom, demonstrating how local knowledge fosters long-term sustainability through adaptive design.

The SES framework has also been widely applied in international research on mountainous settlements. Case studies from the Alps, Andes, and Japan have highlighted how traditional communities enhance resilience through socio-ecological interactions. For instance, the concept of urban heritage resilience integrates disaster preparedness into spatial planning and conservation [28] while recent work on seismic retrofit optimization using genetic algorithms provides new approaches for enhancing community resilience [29]. In China, scholars have explored similar mechanisms through examples such as the Hani Rice Terraces, Huizhou villages, and karst mountain settlements, emphasizing how traditional ecological knowledge (TEK) contributes to soil and water conservation, biodiversity maintenance, and disaster response [27]. Therefore, in the context of our research, TEK moves beyond a background concept to become an explanatory framework. It helps us decipher the underlying logic behind the quantified spatial patterns—for instance, why the ‘moderate elevation–gentle slope–south-facing’ configuration emerged as optimal, or why terracing is more than a simple agricultural technique but a refined hydrological adaptation strategy. These insights provide a robust theoretical foundation for examining the villages of the Qinba Mountains.

2.3. Comparative Analysis of Domestic and International Cases

Both domestic and international research on mountain settlements have yielded rich insights into spatial adaptability and disaster-coping mechanisms. Collectively, these studies underscore the co-evolutionary relationship between human settlements and complex mountain environments, providing critical references for sustainable development and resilient community building [30].

Internationally, attention has been given to the intricate linkages between topography, water management, ecological sensitivity, and hazard risk, as well as to the integration of traditional knowledge and modern technology in resilience strategies. For example, studies on climate change adaptation in alpine permafrost zones emphasize the importance of multi-sectoral governance to mitigate climate-related risks [31]. he use of Nature-based Solutions (NbS)—such as ecological hydraulic engineering to manage floods—has also emerged as a promising approach for mountain settlement resilience [32]. In Europe, the Alpine villages of Italy have long relied on terraced agricultural systems to maximize land use while controlling soil erosion and flood hazards [33,34]. In Japan, settlements in Honshu and Shikoku employ a three-tier system of drainage ditches, terraced fields, and slope vegetation to mitigate floods and landslides, maintaining stability even during extreme rainfall events [35]. Similarly, in the Andean region of South America, the “terrace–platform–canal” system efficiently integrates land use, water management, and hazard mitigation to safeguard agricultural productivity and settlement safety [36,37].

In China, analogous practices can be found in the Honghe Hani Rice Terraces, where a “forest–village–terrace–water system” structure balances ecological conservation and agricultural sustainability [38]. In Lishui, Zhejiang, the “low-intervention” planning philosophy emphasizes minimal disturbance and adaptive reuse of historical patterns, ensuring a balance between cultural heritage protection and economic vitality [39]. In the Southwest Karst region, local residents have developed adaptive strategies such as locating villages on gentle slopes and terraces, constructing stone walls, and maintaining forest belts to mitigate landslides and karst collapses [40,41]. These diverse examples, both domestic and international, demonstrate that through context-specific spatial wisdom and the integration of traditional and modern approaches, mountainous settlements can effectively enhance sustainability and resilience in complex environments.

2.4. Research Gaps and Innovations of This Study

A comprehensive review of the literature indicates that although previous studies have explored various environmental, social, and cultural factors influencing traditional village distribution, there remains a lack of systematic analysis integrating these elements into a unified spatial-resilience framework. Furthermore, while multi-factor approaches have emerged in recent years, their application to complex mountain regions such as the Qinba Mountains is still limited. Quantitative and spatially explicit evaluations of natural adaptability remain scarce, and most studies focus on qualitative interpretation rather than data-driven verification.

Therefore, this study adopts a social-ecological system perspective to investigate 57 national-level traditional villages in the southern Qinba Mountains, employing GIS spatial analysis, kernel density estimation, nearest neighbor index, and geographical detectors to examine village distribution and terrain–resilience coupling across macro, meso, and micro scales. Instead of emphasizing a research gap, this study highlights its contribution through methodological innovation and regional application. It establishes a multi-scale analytical framework that integrates remote sensing and spatial statistics to uncover the terrain–resilience mechanisms of mountainous traditional villages. The key innovations include:

(1)

Developing a “macro-meso-micro” nested mechanism framework to reveal the hierarchical coupling between topography and settlement resilience;

(2)

Applying quantitative tools such as Geodetector to assess the explanatory power of multiple terrain factors and their interactions;

(3)

Using typical cases from the Qinba Mountains to illustrate how traditional adaptive wisdom can inform contemporary resilience-oriented rural planning.

3. Research Area and Methods

3.1. Research Area

The southern section of the Qinba Mountains in Shaanxi Province encompasses the cities of Hanzhong, Ankang, and Shangluo, covering an area of approximately 74,000 km². The region features a distinct geomorphological configuration described as “two mountain ranges with one basin in between.” The Qinling Mountains in the north are dominated by fault-block structures, whereas the Daba Mountains in the south are characterized by folded mountains. Between them lies the Han River valley, consisting of interlaced low hills and alluvial basins. The climate is classified as a northern subtropical humid monsoon climate, with an annual precipitation of 800–1200 mm, most of which occurs during the flood season, frequently triggering landslides, debris flows, and floods. The geological structure of the region is highly complex, with dense fault zones, heterogeneous lithology, and low shear strength of rock masses, resulting in high geological hazard susceptibility (Figure 2).

Due to population migration and ecological resettlement programs over recent decades, many high-risk mountain villages have been relocated, while a considerable number of traditional settlements remain distributed in the low-mountain and hilly areas. According to the List of Chinese Traditional Villages issued by the Ministry of Housing and Urban–Rural Development, a total of 57 nationally recognized traditional villages are currently located within the study area. Most of them are found at elevations between 550 and 1200 m, exhibiting a clustered spatial pattern along valley terraces or foothills. These settlements are not only living spaces for mountain residents but also represent valuable living cultural heritage. Building materials, such as timber, stone, and rammed earth, are locally sourced, and village layouts are carefully adapted to the surrounding terrain and hydrological systems. This demonstrates a profound body of place-based knowledge and adaptive wisdom, accumulated through generations of interaction between humans and the mountainous environment.

The selection of these 57 nationally recognized traditional villages was based on the following considerations: (1) their persistence through historical evolution reflects effective natural adaptation mechanisms, closely aligning with the research theme of terrain–resilience coupling; (2) the national list provides authoritative and standardized metadata, ensuring comparability in spatial analysis; (3) a uniform policy framework helps control for variations in local governance; (4) multi-source high-quality data support rigorous quantitative analysis; and (5) the villages represent the main geomorphological units of the Qinba Mountains, ensuring topographic representativeness.

3.2. Data Sources

The datasets used in this study consist of five major categories:

(1)

Village Data: Derived from the List of Chinese Traditional Villages (5th Batch) published by the Ministry of Housing and Urban-Rural Development, supplemented and cross-validated with local gazetteers and field survey data from Shaanxi Province.

(2)

Topographic Data: Digital Elevation Model (DEM) data with a spatial resolution of 30 m were obtained from the Geospatial Data Cloud of the Chinese Academy of Sciences. These data were used to extract elevation, slope, and aspect variables for the entire study area.

(3)

Disaster Data: Collected from the Shaanxi Department of Natural Resources, Shaanxi Provincial Flood and Drought Control Headquarters, and geological disaster yearbooks. The dataset covers nearly 30 years of landslide, debris flow, and flood events, including detailed spatial locations.

(4)

Land Use and Landscape Data: Derived from China’s National Land Cover Dataset (GlobeLand30, 2020) and Landsat imagery, used to analyze land-use composition and landscape indices surrounding each village.

(5)

Socio-economic Data: Extracted from the Shaanxi Statistical Yearbook and official reports of county governments, providing information on population scale and socio-economic context.

All datasets were spatially standardized and projected in ArcGIS 10.7 (Esri, Redlands, CA, USA) and ENVI 5.3 (Harris Geospatial Solutions, Broomfield, CO, USA) with a unified coordinate system and study boundary. Spatial statistical analyses, including Local Indicators of Spatial Association (LISA), were carried out using the GeoDa-based LISA function integrated within the GIS analysis workflow. Morphological Spatial Pattern Analysis (MSPA) was conducted using Guidos Toolbox 3.0 (Joint Research Centre, European Commission, Ispra, Italy). These procedures ensured the construction of a consistent and comprehensive geospatial database for subsequent analysis.

3.3. Methodology

3.3.1. Spatial Pattern Analysis

The overall spatial distribution of traditional villages was examined using the Nearest Neighbor Index (NNI) to determine whether the pattern tends toward clustering, randomness, or dispersion. Specifically, the NNI is defined as:

$$R={\displaystyle \frac{\overline{r_0}}{r_e}}$$

(1)

where $\overline{r_0}={\displaystyle \frac{1}{N}}{\sum }_{i=1}^N{r}_i$ is the observed mean nearest-neighbor distance ($r_i$ is the distance from point $i$ to its nearest neighbor) and $\overline{r_e}={\displaystyle \frac{1}{2\sqrt{\lambda }}}={\displaystyle \frac{1}{2}}\sqrt{{\displaystyle \frac{A}{N}}}$ is the expected mean distance for a random (Poisson) distribution, with N the number of points, A the study area, and λ = N/A the point density. A value R < 1 indicates clustering, R ≈ 1 randomness, and R > 1 dispersion.

For statistical significance, we compute the z-score as:

$$z={\displaystyle \frac{\overline{r_0}-\overline{r_e}}{\mathrm{S}\mathrm{E}\left(\overline{r}\right)}}$$

(2)

where SE$\left(\overline{r}\right)$ denotes the standard error of the mean nearest-neighbor distance (standard formulas for SE of nearest-neighbor statistics were used).

To further visualize clustering intensity and spatial hotspots, Kernel Density Estimation (KDE) was employed. The KDE estimator used is:

$$\widehat{f}\left(\mathrm{x}\right)={\displaystyle \frac{1}{n{h}^2}}{\sum }_{i=1}^{n}K{\displaystyle \frac{x-x_i}{h}}$$

(3)

where x is a two-dimensional location, are observed point locations, n is the number of points, ℎ is the bandwidth (kernel smoothing parameter), and K(·) is the kernel function.

In addition, to quantify spatial concentration and equity across administrative units, we computed the geographic concentration index G and the normalized imbalance index I:

$$G={\sum }_{j=1}^m{p}_j^2$$

(4)

where $p_j{\displaystyle \frac{n_j}{N}}$ (here $n_j$ is the number of villages in administrative unit $j$, $m$ is the number of units, and $N$ the total number of villages).

The imbalance index is calculated as:

$$I={\displaystyle \frac{G-\frac{1}{m}}{1-\frac{1}{m}}}$$

(5)

which rescales $G$ to the interval [0, 1], where $I=0$ indicates perfect evenness and $I=1$ indicates maximal concentration.

In addition, a Local Indicators of Spatial Association (LISA) analysis was conducted to identify local clustering patterns and spatial heterogeneity of elevation, slope, and aspect at the village level. The analysis was based on Anselin’s Local Moran’s I statistic, expressed as:

$$I_i=Z_i{\sum }_j{W}_{ij}{Z}_j$$

(6)

where $Z_i$ and $Z_j$ denote standardized values of the variable (e.g., elevation) for village $i$ and its neighbors $j$; $W_{ij}$ represents the spatial weight matrix defined by inverse distance within a 10 km neighborhood; and $I_i$ measures the extent to which the value at location $i$ is similar or dissimilar to the surrounding values.

Positive $I_i$ values indicate high–high or low–low clusters (spatial similarity), while negative $I_i$ values represent high–low or low–high outliers (spatial dissimilarity). Statistical significance (p < 0.05) was assessed using the LISA statistical function based on the GeoDa module within the GIS analysis workflow, rather than through standalone software. And results were mapped to visualize clusters of elevation, slope, and aspect. This approach allows for fine-scale identification of spatial association patterns beyond global statistics such as the overall Moran’s I.

3.3.2. Geodetector Analysis

To quantitatively assess the explanatory power of topographic factors, the Geodetector model was applied [42]. The method evaluates the spatial consistency between an independent variable and a dependent variable using the q-statistic, defined as:

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

(7)

where $N_h$ and N represent the number of samples in subregion h and the entire study area, respectively, and ${\delta }_h^2$ and ${\delta }^2$ denote their variances. A higher q-value indicates a stronger explanatory power of the factor.

In this study, elevation, slope, and aspect were treated as explanatory variables, while the spatial distribution of villages served as the dependent variable. The model allows for interaction detection among multiple factors, enabling a deeper understanding of compound topographic effects on settlement location.

3.3.3. MSPA Ecological Pattern Analysis

Morphological Spatial Pattern Analysis (MSPA) was used to characterize the ecological structure and its relationship with settlement locations [43]. The analysis consisted of four technical modules: data preprocessing, parameter optimization, spatial interpretation, and correlation analysis. Using the GlobeLand30 (2020) dataset with a spatial resolution of 30 m, forests, grasslands, and water bodies were defined as foreground classes, whereas cropland and built-up land were designated as background classes. The binary raster was generated by assigning a value of 1 to the foreground and 2 to the background. Key parameters were defined based on local geomorphological characteristics and village morphology:

Edge width: 100 m, representing the transition interface between villages and ecological core zones;

Core area threshold: 5 ha, used to filter out fragmented patches and retain contiguous ecological areas;

Connectivity rule: 8-neighbor adjacency, appropriate for maintaining continuity in mountainous terrain.

The results of MSPA were interpreted using Guidos Toolbox 3.0, which classified the landscape into seven morphological categories: core, edge, bridge, perforation, branch, islet, and loop. These spatial components were further analyzed in relation to the distribution of villages and recorded disaster points to quantify risk avoidance patterns and ecological buffering effects.

4. Results and Analysis

4.1. Spatial Distribution Characteristics

Based on the analysis of 57 nationally recognized traditional villages in the southern Qinba Mountains, the Nearest Neighbor Index (NNI) was calculated as R = 0.86 (<1), indicating a significantly clustered distribution pattern. This finding is consistent with previous studies suggesting that settlements in mountainous regions are rarely random, being strongly influenced by both topographic constraints and historical transportation networks. The Kernel Density Estimation (KDE) results further revealed that high-density clusters are concentrated primarily in Liuba County (Hanzhong) and Hanbin District (Ankang) (Figure 3). These areas are characterized by relatively gentle terrain, proximity to river terraces or alluvial basins, and favorable agricultural and transport conditions, which collectively make them preferred sites for traditional settlements.

The geographical concentration index of the villages is 61.60, demonstrating a high degree of spatial concentration, while the imbalance index of 0.26 indicates moderate spatial inequity across administrative units. This unevenness is closely related to topographic variability and the spatial heterogeneity of disaster exposure in the study area. A high-low cluster analysis using the Getis-Ord $G_i^{*}$ statistic indicates that elevation exhibits a pronounced high–high clustering pattern, whereas slope and aspect display random spatial distributions (

Table 1. High-Low Clustering Analysis of Traditional Villages in the Southern Shaanxi Qinba Mountains.

Category	General G Observed Value	z	p	Result	Significance
Elevation	0.000016	2.739471	0.006154	High clustering	Significant
Slope	0.000014	0.393093	0.694250	Random distribution	Not significant
Aspect	0.000013	−0.047718	0.961941	Random distribution	Not significant

). Villages in Liuba County are predominantly situated in high-elevation, low-slope clusters, while those in Hanbin District tend to occupy low-elevation, high-slope areas. Overall, approximately 83% of the traditional villages are located within the 550–1200 m elevation range, 68.4% are distributed on slopes gentler than 15°, and 41% are concentrated on south- and southwest-facing slopes (Figure 4). This composite pattern-characterized by moderate elevation, gentle slopes, and sunny aspects-effectively balances environmental suitability and hazard avoidance. It provides favorable hydrothermal conditions for agricultural production while significantly lowering disaster occurrence. Within a 100 m core buffer zone, the average disaster density decreases to 0.077 events per square kilometer, compared with 0.290 events per square kilometer in surrounding areas. Such terrain-oriented spatial logic embodies the traditional adaptive wisdom of “backing against the wind, facing the sun, and seeking advantages while avoiding disadvantages”, reflecting an environmentally responsive approach to settlement resilience and sustainability.

4.2. Classification of Settlement Types

Based on geomorphological criteria from the Shaanxi Provincial Gazetteer: Geography Volume and the China Landform Classification Code, the study area can be divided into three primary geomorphological units—plains and basins, low mountains and hills, and medium mountains (

Table 2. Classification of Geomorphological Features in the Southern Shaanxi Qinba Mountainous Region.

Classification	Elevation (m)	Relative Height Difference (m)	Slope (°)	Geomorphological Characteristics	Typical Regions
Plain Basin	≤550	≤50	≤5	Flat plains–gentle slopes–low relief	Han River Valley
Low Mountainous Hilly	550–1200	50–300	10–25°	Low mountains–moderate slopes–moderate relief	Southern slopes of the Qinling Mountains
Medium-Mountainous	≥1200	>300	>25°	Medium mountains–steep slopes–high relief	Main ridge of the Qinling Mountains, northern slopes of the Daba Mountains

). Among them, plain basins are areas with an altitude of less than 550 m and a relative height difference of ≥50 m, typically represented by the Han River Valley, with gentle slopes (≤5°), serving as core areas for agricultural and urban development; low mountain and hilly areas are areas with an altitude of 550–1200 m and a relative height difference of 50–300 m, concentrated in the southern foot of the Qinling Mountains, with slopes of 10–25°, spatially interwoven with the transition zone of middle mountain areas, and facing landslide disaster risks; middle mountain areas are areas with an altitude of ≥1200 m and a relative height difference of >300 m, typically represented by the main ridge of the Qinling Mountains, with steep slopes (>25°), mainly functioning as ecological barriers. For the transition zone between the middle mountains and plains (southern foot of the Qinling Mountains), low mountains and hills overlap in terms of altitude (500–1200 m) and topographic relief (50–300 m), so they are merged into a single “low mountain and hilly unit” to avoid fragmentation of spatial heterogeneity.

Furthermore, based on geomorphological characteristics and disaster risks, traditional villages in southern Shaanxi are classified into three typical types (

Table 3. Geographical Distribution of Traditional Villages in Southern Shaanxi.

Classification	Range	Type	Number of Villages
Elevation (m)	142–550	Plain	9
	550–1200	Low Mountain	44
	1200–3052	Medium Mountain	3
Slope (°)	0–15	Gentle Slope	39
	15–30	Moderate Slope	17
	30–76.3	Steep Slope	1
Aspect (°)	22.5–67.5	Northeast	3
	67.5–112.5	East	5
	112.5–157.5	Southeast	5
	157.5–202.5	South	13
	202.5–247.5	Southwest	9
	247.5–292.5	West	10
	292.5–337.5	Northwest	3

). Plain Basin Type: Predominantly located in the Han River Valley and larger tributary floodplains, such as the Hanzhong Basin and Ankang Dam areas, these villages feature flat terrain, convenient transportation, and fertile land suitable for agriculture. However, they are susceptible to flooding. Low Mountainous Hilly Type: The most numerous, these villages are situated on gentle slopes (550–1200 m) in low mountainous hilly areas, balancing favorable agricultural conditions with lower disaster risks. They often utilize gentle slopes for terraced fields and platforms to achieve a balance between production and safety. Medium-Elevation Platform Type: Mainly located in the medium-elevation areas of the Daba and southern Qinling Mountains, these villages are relatively scarce (Figure 5). They mitigate flood and debris flow risks by building on platforms elevated above flood levels but face challenges related to limited transportation and resource accessibility, resulting in lower development levels. The spatial differentiation patterns of these three village types indicate that traditional villages consider both resource acquisition and production conditions, as well as disaster risk avoidance, in their site selection, reflecting a “human-land synergy” approach to natural adaptability.

4.3. Relationship Between Spatial Distribution and Topographic Factors

Geodetector analysis showed that elevation had the strongest explanatory power for village distribution (q = 0.46), indicating that elevation is the dominant factor determining the spatial pattern of villages. The explanatory power of slope was q = 0.32, which mainly constrained village site selection at the meso scale—village distribution decreased significantly when the slope exceeded 25°. The explanatory power of aspect was relatively weak (q = 0.29), but it played an important role in optimizing the microclimate and light conditions of villages (

Table 4. Explanatory Power of Topographic Factors on Village Distribution.

Topographic Factor	q-Value	Significance (p-Value)	Dominant Scale of Action
Elevation	0.46	<0.01	Macro-pattern
Slope	0.32	<0.05	Meso-constraint
Aspect	0.29	>0.05	Micro-optimization

).

Notably, the explanatory power significantly increased when elevation interacted with factors such as slope and aspect. This indicates that the spatial distribution of villages is not determined by a single factor but by the combined effect of multiple topographic factors. For example, south-facing platforms on gentle slopes of low mountains often become the optimal site for villages, as they not only ensure convenient farming and smooth drainage but also reduce the risks of floods and landslides.

4.4. Disaster Avoidance Mechanisms

The Qinba Mountains are characterized by complex geological structures, intense hydrological processes, and multi-hazard chains involving landslides, debris flows, and floods. A typical disaster chain evolves as: intense rainfall → slope failure → debris flow → river blockage → secondary flooding. Studies have shown that villages distributed in river valley plains and first-order terraces are more vulnerable to flood disasters—especially along the main stream of the Han River and major tributaries, villages are often located at the edge of alluvial fans or terraces, which are highly prone to inundation when flood peaks arrive. In contrast, villages distributed in the interlaced zones of low mountains, hills, and gullies are more threatened by landslides and debris flows. These hazards are often triggered by heavy rainfall and rapidly evolve through the convergence of surface runoff, forming typical geological disaster chains. For example, during an extreme rainstorm in Ziyang County (Ankang City) in 2020, multiple traditional villages were hit by landslides and debris flows, with some roads and houses buried, seriously endangering village safety. This disaster chain not only reflects the vulnerability of the natural environment but also indicates a close coupling relationship between village site selection and topographic environment (Figure 6).

Using MSPA to delineate ecological core zones and buffer areas, we identified that core zones, accounting for 65.95% of the landscape, are concentrated in contiguous forested areas along the southern Qinling Mountains and the Han River watershed, forming significant ecological barriers (

Table 5. Seven Landscape Elements Identified by MSPA.

MSPA Category	Foreground Percentage	Global Percentage	Meaning
Core	80.02	65.95	Continuous and unfragmented ecological patches
Edge	6.00	4.95	Transitional zones surrounding core areas
Bridge	1.86	1.53	Narrow zones connecting two core areas
Perforation	8.98	7.40	Non-ecological patches within core areas
Branch	2.12	1.75	Linear structures extending from core areas
Islet	0.25	0.20	Isolated small core areas
Loop	0.77	0.64	Corridors connecting the same core area

). By overlaying village and disaster point locations in ArcGIS, we calculated core zone avoidance rates and disaster risk incidence rates. Among the 57 traditional villages, 83% are located in low mountainous hilly areas (550–1200 m), and 61.4% overlap with core zones. However, their site selection demonstrates selective avoidance of core zones, with disaster densities of 0.077 per km² within 100-m core zone buffers, significantly lower than the 0.290 per km² in peripheral areas, validating the effectiveness of ecological barriers (Figure 7). This spatial association analysis indicates that traditional villages achieve dynamic balance between ecological sensitivity and disaster risks through “edge adherence-core avoidance” strategies, reflecting a synergistic mechanism of natural adaptability and social choice.

5. Terrain-Driven Mechanisms and Natural Adaptation Strategies

5.1. Topographic Driving Mechanisms

Topography is the fundamental determinant shaping the spatial distribution of traditional villages. Results from the Geodetector analysis indicate that elevation, slope, and aspect are the three most influential topographic drivers, with explanatory powers of q = 0.46, 0.32, and 0.29, respectively. Among them, elevation exerts the strongest control, defining the overall vertical framework of settlement distribution (Figure 8).

Elevation-Driven Mechanism: Villages are predominantly concentrated within the low-mountain and hilly zones at elevations of 550–1200 m. This altitudinal range provides moderate relief and abundant water resources, offering favorable conditions for both agricultural production and daily living. In contrast, river valleys below 500 m, though convenient for transportation, are highly susceptible to flooding; while mid- to high-mountain areas above 1500 m are characterized by steep terrain, fragile ecosystems, and limited accessibility, which impose clear constraints on cultivation and habitation, resulting in a sharp decline in village density. By influencing temperature, precipitation redistribution, and the accessibility of human activities, elevation functions as the dominant factor shaping the spatial pattern of villages, defining the macro-scale resilience boundary of settlement distribution.

Slope-Driven Mechanism: Slope directly affects both land suitability and disaster risk. In the study area, village locations fall within a slope range of 1.28–32.04°, with 62% of villages concentrated on gentle slopes between 4° and 18°. Areas with slopes below 8° provide favorable conditions for infrastructure construction and cultivation but are more vulnerable to flooding. Moderate slopes between 8° and 15° offer both safety and effective drainage, making them ideal sites for settlement development. Conversely, steep slopes above 25° are subject to severe soil erosion, leading to sparse settlement distribution. As a meso-scale constraining factor, slope influences land bearing capacity, construction cost, and disaster exposure, guiding villages toward low-hazard and high-suitability areas such as gentle slopes and terraces. This selective adaptation shapes the spatial distribution of villages, achieving a balance between productive needs and resilience safety.

Aspect-Driven Mechanism: Villages are predominantly distributed on south- and southwest-facing slopes, a locational preference closely related to solar radiation, thermal conditions, and moisture availability. In the study area, 41% of villages exhibit a south or southwest orientation, reflecting strong adaptation to the regional climatic context. The Qinba Mountains experience prevailing northwesterly winds in winter, and south-facing slopes maximize solar exposure and improve the local microclimate. Meanwhile, southeast-facing slopes combine the advantages of sheltering from cold winds and facilitating ventilation. Such orientations increase winter sunshine duration by approximately 2–3 h per day compared with north-facing slopes, effectively reducing the risk of frost damage and demonstrating a localized climatic adaptation characterized by “sheltering from wind and facing the sun.”Although the explanatory power of slope aspect (q = 0.29) is weaker than that of elevation and slope gradient, its microclimatic regulation significantly enhances residential resilience, making it a key optimization factor at the micro scale. Its spatial differentiation pattern is influenced by the superimposed effects of macro-topographic structures. For instance, Chejiahe Village in Shangluo, southern Shaanxi, selects diverse layout patterns in response to varying mountain–water configurations. Dispersed layouts are adopted in areas with significant terrain undulations, clustered forms in relatively flat and open areas, and linear layouts in narrow valleys. Village sites are often located along main mountain ridges and away from river channels, harmonizing with natural conditions to reduce ecological risks. By leveraging abundant natural endowments, such spatial strategies strengthen the stability of village systems under disturbance, ultimately embodying the wisdom and redundancy of adaptation to natural foundations.

In summary, the complex landscape pattern of southern Shaanxi exposes traditional settlements to significant uncertainties in ecological risks. To mitigate this uncertainty, traditional settlements in southern Shaanxi have demonstrated unique local wisdom in their adaptation to nature. Village distribution shows obvious topographic adaptability, following the rule of “selecting moderate elevation, choosing gentle slopes, and favoring sunny aspects”. This topographic driving mechanism is common in mountain villages worldwide but is particularly prominent in the Qinba Mountains.

5.2. Natural Adaptation Strategies and Resilience Enhancement Mechanisms

The natural adaptation of traditional villages in the Qinba Mountains is not a passive response to environmental constraints but a proactive process of spatial optimization and environmental regulation. Through targeted strategies that integrate topographic modification and ecological feedback, these settlements have developed systematic mechanisms to enhance resilience against multiple natural hazards. The core strategies and intrinsic mechanisms are illustrated in Figure 9 and elaborated below.

Terrace-Based Flood Prevention Strategy and Hydrological Regulation Mechanism. The Qinba Mountains experience concentrated rainfall during the flood season, with annual precipitation typically ranging from 800 to 1200 mm. Consequently, flooding frequently occurs in low-lying river valley plains. To mitigate such risks, traditional villages tend to be located on secondary or tertiary river terraces situated approximately 3–5 m above flood levels, where the terrain is further reshaped through terracing to enhance safety and usability. It should be noted that the “secondary terraces” referred to in this study are geomorphic units classified based on geomorphological and hydrological characteristics. Specifically, they are mid-slope fluvial terraces formed along secondary river systems or intermountain valleys, generally 400–800 m above the local riverbed, representing relatively stable geomorphic surfaces shaped by long-term river erosion and deposition [44]. This stepped configuration slows surface runoff by about 30–50%, prolongs confluence time, and improves soil infiltration capacity. Such hydrological regulation reduces slope floods and soil erosion while enhancing water retention and agricultural stability. A representative example is Mahe Village in Hanbin District, which is located on the secondary terrace of the Han River. During the 2020 Ziyang heavy rain event, the terrace system effectively protected the village from inundation, while nearly 60% of adjacent plain settlements suffered flood damage. This demonstrates how the terrace-based strategy establishes a flood resilience mechanism through an adaptive chain of terrain transformation, hydrological regulation, and disaster avoidance.

Convex-Bank Stabilization Strategy and Geomorphological Adaptation Mechanism. The fluvial systems of the Qinba Mountains are highly dynamic, characterized by active erosion on concave banks and sediment accumulation on convex banks. Traditional villages consciously locate themselves on convex banks to capitalize on the stable foundations formed by alluvial deposition, where the soil bearing capacity is approximately 20–30% higher than that of concave banks. The convex-bank landform not only reduces the risk of bank collapse but also expands the usable area through its fan-shaped morphology, which promotes effective surface drainage. Empirical observations from Ningqiang County, including Chejiahe Village, show that the riverbank collapse rate of convex-bank settlements is only one-fifth that of concave-bank villages. This pattern reflects a geomorphological adaptation mechanism in which residents utilize the natural evolution of river morphology to achieve a dynamic balance between settlement stability and land-use efficiency.

Platform-Based Hazard Avoidance Strategy and Disaster Chain Interruption Mechanism. In mid-mountain areas above 1200 m in elevation, compound disaster chains—such as “intense rainfall-landslide-debris flow”—pose significant threats to settlements. To mitigate these risks, traditional villages are typically built on natural or artificially leveled platforms located 50–100 m above valley bottoms, often supplemented with surrounding drainage ditches. The elevated platforms provide vertical isolation from debris flows and flash floods, while the drainage system facilitates rapid runoff diversion, thereby lowering the probability of slope saturation and landslide initiation. Monitoring data from Gaoshan Village in Hanbin District indicate that this spatial configuration reduces exposure to debris-flow hazards by approximately 75% and lowers landslide risk by around 60%. By combining elevation isolation with hydrological diversion, the platform-based strategy effectively interrupts the transmission path of disaster chains, forming a robust resilience defense mechanism against geological hazards.

Collectively, these adaptive strategies are deeply rooted in topographic awareness yet go beyond mere site selection preferences. Traditional villages in the Qinba Mountains actively reshape their physical environment and harness natural processes to achieve precise and differentiated responses to multiple hazard types. This proactive engagement with the landscape enhances ecological stability, disaster resilience, and long-term sustainability of settlement systems, aligning closely with the fundamental concept of “natural adaptation mechanisms” in the context of resilient spatial planning.

5.3. Multi-Scale Coupling Mechanisms

The natural adaptation of traditional villages is not the outcome of a single, isolated strategy but a systemic response that has evolved through multi-scale coupling mechanisms. These mechanisms operate across macro–meso–micro spatial scales and exhibit a nested hierarchical structure, in which topographic factors play dominant, secondary, and supplementary roles. Each spatial scale reflects a distinct dimension of human–environment interaction, jointly forming a resilient settlement system (

Table 6. Coupling Mechanism of Spatial Scales and Topographic Factors: Roles, Resilience Contributions, and Intensity Ranks.

Spatial Scale	Topographic Factor	Parameter Range	Role in Settlement Pattern	Resilience Contribution	Intensity Rank
Macro (Regional Level)	Elevation	550–1200 m (low-mountain and hilly zones)	Establishes vertical differentiation and delineates safe settlement boundaries	Establishes vertical differentiation and delineates safe settlement boundaries	Dominant
Meso (Slope Level)	Slope	4–18° (gentle to moderate slopes)	Shapes land-use organization and construction suitability	Shapes land-use organization and construction suitability	Secondary
Micro (Building–Courtyard Level)	Aspect	South- and southwest-facing slopes	Optimizes microclimatic conditions and spatial orientation	Optimizes microclimatic conditions and spatial orientation	Supplementary

).

Macro Scale (Regional Level): At the regional level, altitude serves as the dominant control factor (q = 0.46), creating a vertical differentiation that defines the spatial safety boundaries for settlement distribution. Villages are primarily located within low-mountain and hilly zones (550–1200 m), which provide a balance between environmental safety and resource accessibility. This altitudinal range allows settlements to avoid flood-prone lowlands (<550 m) while also steering clear of the harsh climate and steep terrain above 1200 m. Consequently, altitude establishes the macro-scale foundation of disaster resilience, providing a stable environmental base upon which human settlements can develop and persist.

Altitude-Dominated Foundation for Disaster Resilience. At the regional scale, altitude serves as the dominant controlling factor (q = 0.46), establishing vertical differentiation that defines the safe boundaries for settlement distribution. Villages are mainly concentrated in the low-mountain and hilly zones between 550 and 1200 m above sea level. This altitude range enables them to avoid flood-prone plains below 550 m while also steering clear of the cold climate, steep slopes, and resource scarcity typical of elevations above 1200 m. Such vertical stratification forms the macro-scale foundation of disaster resilience for traditional settlements.

Meso Scale (Slope Level): Slope-Regulated Production–Safety Balance Mechanism. At the meso scale, slope gradient emerges as the secondary influencing factor (q = 0.32), governing the planar configuration and land-use organization of villages. Traditional settlements typically occupy gentle to moderate slopes (4–18°) that ensure both terrain stability and agricultural efficiency. Through practices such as terracing and slope optimization, soil erosion is minimized and the risk of geological hazards is effectively reduced. This creates a production–safety balance mechanism, in which topographic adaptation supports both livelihood productivity and environmental protection. The integration of agricultural terraces and residential areas thus reflects a deliberate spatial strategy to achieve equilibrium between human needs and landscape constraints.

Micro Scale (Building–Courtyard Level): Aspect-Based Livability Optimization Mechanism. At the micro scale, slope aspect functions as a supplementary factor (q = 0.29), enhancing the microclimatic comfort and habitability of the built environment. South- and southwest-facing orientations are preferred, as they provide two to three additional hours of sunlight during winter, mitigating cold stress and improving thermal comfort. Meanwhile, the careful alignment of courtyards and pathways along natural gradients enhances drainage efficiency and prevents water accumulation. These spatial and climatic optimizations together constitute a livability optimization mechanism, reinforcing the everyday resilience and environmental adaptability of traditional dwellings.

In summary, the multi-scale coupling system of traditional villages can be summarized as a nested progression of “altitude differentiation—slope regulation—aspect optimization.” This structure reflects an adaptive logic of “risk identification—spatial response—functional enhancement,” which has gradually evolved through continuous human–environment interactions. By coordinating responses across different spatial levels, traditional villages have not only reduced their exposure to natural hazards but also achieved ecological sustainability and cultural continuity. This multi-scale synergy provides critical insight into the spatial resilience mechanisms underpinning vernacular settlements in mountainous regions.

6. Discussion

6.1. Theoretical Implications: Terrain–Resilience Coupling in Mountainous Settlements

The findings of this study indicate that the spatial distribution characteristics and adaptive mechanisms of traditional villages in the Qinba Mountains are the outcomes of closely intertwined terrain–resilience coupling processes. In this coupling dynamic, terrain not only imposes physical constraints but also shapes a cognitive framework that guides communities in perceiving, interacting with, and utilizing their surrounding environment. This relationship demonstrates that the resilience of mountainous settlements is rooted in the long-standing dynamic feedback between human spatial practices and environmental dynamics.

Essentially, the quantitative patterns revealed in this study—including the hierarchical influences of elevation, slope, and aspect, as well as specific hazard-avoidance spatial strategies—are precisely the quantifiable expressions of Traditional Ecological Knowledge (TEK). In this context, TEK can be defined as a cumulative system of knowledge, practices, and beliefs refined through long-term adaptive processes and passed down through generations via cultural transmission. By quantifying the physical traces of TEK inscribed in the landscape, this study provides a methodological approach to converting such abstract knowledge into concrete, spatially explicit evidence.

From the theoretical perspective of social-ecological systems (SES), resilience denotes a dynamic process wherein systems undergo continuous adjustment and ongoing transformation in response to environmental disturbances. The traditional spatial wisdom documented in this study—preference for moderate elevation zones, prioritization of south-facing slopes, and settlement construction on stable terraces—represents proactive, embodied adaptive strategies rooted in TEK, rather than passive reactions to environmental pressures. These strategies are the tangible outcomes of long-term co-evolution between human activities and mountain landscapes, and are closely aligned with Holling’s adaptive cycle theory and Walker’s conceptualization of “resilience as persistence, adaptability, and transformability.”

Fundamentally, the spatial patterns of traditional villages in the Qinba Mountains can be understood as the “spatial manifestation of resilience”—where form, function, and environment interact to jointly sustain system stability. Additionally, the self-organizing characteristics of these settlements echo the concept of morphological resilience, highlighting that resilience is deeply embedded in the structural and spatial configuration of the built environment.

6.2. Comparative Perspectives: Global Parallels and Regional Specificity

A comparative analysis with other mountainous regions worldwide reveals both universal patterns and regional particularities in the adaptive strategies of the Qinba villages.

Globally, traditional mountain settlements—such as those in the Italian Alps, Japanese Shikoku region, and Andean highlands—share common strategies of terrain adaptation, hydrological management, and ecological integration. These include terraced agriculture, convex-bank siting, and multi-tiered drainage systems that mitigate environmental risks while sustaining livelihoods. The Qinba villages exhibit a similar logic: combining topographic selection with ecological buffering, they represent a form of vernacular resilience rooted in the co-existence between human settlements and natural processes.

However, compared with European and Japanese cases that emphasize heritage preservation and policy-based resilience, Qinba villages face more acute challenges of population outmigration, economic marginalization, and multi-hazard exposure. Their adaptive mechanisms are thus more environmentally driven than institutionally supported. This distinction underscores the need for integrating local ecological knowledge (LEK) with modern spatial governance systems to sustain resilience in less-developed mountainous regions.

While many Western resilience models rely on infrastructural and technological adaptation, the Qinba experience emphasizes spatial wisdom and material adaptation—that is, resilience achieved through the layout, materiality, and morphology of settlements themselves. This endogenous, experience-derived resilience not only enriches the global understanding of how small-scale rural communities adapt to environmental uncertainty without intensive external intervention but also underscores through these findings the dual attributes of the Qinba case: it aligns with the universal logic of ecological adaptation in global mountain settlements and embodying uniqueness in its environment-driven, endogenous resilience model rooted in local spatial wisdom—offering a replicable pathway for resilience building in underdeveloped mountainous regions with limited institutional and technological support.

6.3. Practical Implications: Planning for Spatial Resilience in Mountainous Villages

The protection and development of traditional villages in mountainous regions are not only vital for disaster prevention and ecological security but are also closely aligned with the Rural Revitalization Strategy. Both previous research and the findings of this study show that the spatial configuration and natural adaptation mechanisms of traditional villages are central to strengthening rural resilience. These adaptive principles should therefore be integrated into regional planning and translated into measurable indicators for resilient construction and land management.

Building on the multi-scalar analytical framework developed in this study, three planning implications emerge across regional, landscape, and settlement levels. Together, these insights illustrate how resilience operates through spatial hierarchy and provide a structured foundation for disaster-sensitive planning in mountainous environments.

At the macro scale, planning should follow the principle of topographic suitability. New settlements and relocation projects should avoid floodplains located below the level of secondary river terraces and slopes steeper than 15 degrees, which are highly susceptible to floods and landslides. GIS-based spatial analysis can support the delineation of traditional village protection zones, using elevation thresholds between 400 and 800 m above riverbeds and terrace heights 3 to 5 m above flood levels as practical safety parameters. These measurable indicators offer a scientific reference for risk-sensitive zoning and sustainable land use.

At the meso scale, maintaining landscape connectivity and ecological buffering is crucial. Natural corridors, forest belts, and traditional terraced fields help slow surface runoff, increase infiltration, and reduce hazard transmission. The restoration and adaptive reuse of traditional disaster-prevention structures—such as drainage ditches, retaining walls, and stepped terraces—should be prioritized. These elements provide both ecological protection and cultural continuity, achieving a balance between risk mitigation and heritage conservation.

At the micro scale, design strategies should enhance in situ resilience by adapting traditional wisdom to modern needs. South-facing layouts, tiered foundations, and integrated drainage systems improve safety, thermal comfort, and spatial efficiency. Encouraging community participation in the maintenance of these systems strengthens local self-organization and ensures the long-term sustainability of disaster-prevention measures.

From a policy perspective, the findings offer practical insights for integrating resilience-oriented thinking into rural revitalization. While national initiatives have largely emphasized relocation and infrastructure improvement, local adaptive capacity has often been overlooked. Recognizing traditional settlements as living archives of resilience can inspire governance models that harmonize cultural preservation with disaster risk reduction.

Finally, this research integrates GIS-based terrain analysis, Geodetector modeling, and MSPA-based ecological interpretation to demonstrate the value of multi-source spatial data in identifying settlement vulnerability and optimizing adaptive strategies. By combining spatial analysis with an understanding of traditional ecological knowledge, the methodological framework developed here can serve as a transferable reference for spatial resilience planning in other mountainous regions under changing climatic conditions. In essence, this study bridges traditional adaptive wisdom with modern spatial planning, providing actionable pathways to achieve an integrated vision of disaster prevention, ecological restoration, and cultural heritage conservation in rural mountain landscapes.

7. Conclusions

This study was guided by the central research question: how do traditional villages in the Qinba Mountains spatially adapt to complex topographic and hydrological environments to achieve disaster resilience?

To answer this question, we conducted a multi-scale spatial analysis of 57 national-level traditional villages, integrating GIS-based terrain analysis, kernel density estimation, and geographical detectors to identify the spatial patterns and natural adaptation mechanisms shaping village resilience. The results demonstrate that traditional villages exhibit distinct topographic adaptation strategies—preferring secondary terraces, gentle south-facing slopes, and areas with moderate elevation—to balance safety, accessibility, and agricultural productivity. These findings clarify how long-term human–environment interactions and Traditional Ecological Knowledge (TEK) have co-evolved into a spatially embedded system of in situ disaster prevention. The study thus provides a quantitative and spatially explicit interpretation of resilience mechanisms in mountainous traditional settlements, offering methodological and practical insights for regional disaster prevention, ecological restoration, and rural revitalization.

Despite these contributions, several limitations should be noted. First, this study mainly focuses on topographic and hydrological factors, while socio-economic dynamics such as demographic changes and land-use transitions were not fully considered. Second, the research objects are limited to national-level traditional villages in the Qinba Mountains, which, although representative, do not reflect the full diversity of settlement types. Future studies could expand the sample to include provincial- and local-level traditional villages to enable broader comparative analysis. In addition, the temporal evolution of adaptation could not be quantitatively assessed due to data limitations. Future work should integrate socio-cultural variables and time-series data to capture the dynamic evolution of resilience mechanisms and to validate the proposed multi-scale spatial framework across different mountain regions.