Landscape Visual Affordance Evaluation at a Regional Scale in National Parks: A Case Study of the Changhong Area in Qianjiangyuan National Park

Abstract

National parks play a vital role in safeguarding natural scenery, maintaining ecological integrity, and preserving cultural heritage, while simultaneously offering valuable opportunities for recreation and education. Among the diverse resources provided by national parks, visual landscape resources hold particular significance due to their capacity to inspire, educate, and enhance aesthetic appreciation. However, assessing and managing these resources remain challenging, as they span both the physical attributes of the landscape and the human visual perception process. This study aims to develop a theoretical and practical framework for evaluating the “landscape visual affordance” of national parks. Grounded in ecological psychology’s affordance theory, the proposed approach integrates physical affordance and sensory affordance, encompassing both the objective physical attributes of the landscape and the subjective processes of human perception. Drawing on a multi-dimensional set of indicators, the research quantifies physical features—such as topography, land use, vegetation cover, and landscape structure—as well as sensory dimensions, including visibility, visual prominence, and viewing frequency. These elements are synthesized into a landscape visibility assessment model built upon the affordance theory framework. The results demonstrate that landscape visual affordance effectively identifies landscape patches with varying degrees of visual quality and importance within national parks and other protected areas. By providing robust support for management decisions—such as zoned protection, optimizing recreational facilities, and evaluating visitor carrying capacity—this model offers new insights and practical guidance for the sustainable planning and management of landscapes in national parks and other ecologically critical regions.

1. Introduction

Landscape visual resources encompass all visible natural and anthropogenic elements in a landscape [1], functioning as a vital carrier for ecological, scientific, cultural, historical, recreational, and aesthetic values [2]. They integrate the objective attributes of landscape resources with subjective visual appreciation, underscoring the interactive characteristics of scenic aesthetics. Landscape visual resources play an important role in ecological and natural resource conservation, enhancing visitor experiences [3], promoting cultural and historical values [4], fostering sustainable tourism development [5], and strengthening community involvement and communication [6]. Recognizing visual landscapes as a resource necessitates their measurement and management [7].

Since the 1960s, numerous studies have evaluated landscape visual quality using methods that can be classified into three paradigms: (1) an objective paradigm emphasizing the physical landscape, typically led by experts; (2) a subjective paradigm grounded in public perception and judgment [8]; and (3) a combined (objective–subjective) paradigm that applies expert assessments and perceptual methods in tandem [9]. In the context of national parks and nature reserves, several approaches have been developed:

• Objective Paradigm (Direct Perception Based on Photographs): This method delineates spatial units with clear boundaries derived from geographic features. Visual landscape assessments (based on the naked eye) are then performed at selected vantage points within each unit, producing an integrated visual quality distribution map for the entire area [10,11]. Examples include the Visual Resource Management (VRM) system administered by the U.S. Bureau of Land Management and the Scenic Quality Rating (SQR) used by the National Park Service [4]. This approach combines physical landscape attributes (features within the field of view) with sensory elements (e.g., visibility and sensitivity). However, it emphasizes static evaluations of visual resources and depends on the selection and accessibility of viewpoints, rendering it time-consuming and labor-intensive for large-scale assessments [12,13].
• Objective Paradigm (Indirect Perception Based on Geospatial Data): Experts select appropriate indicators from the literature, utilizing data such as land use, elevation, and vegetation coverage (rather than photographs). These indicators are calculated and overlaid using GISs and other software. This approach primarily examines the physical properties of the visual landscape and presents findings in map form, as illustrated by the Landscape Attraction Indicator [14] and the Landscape Aesthetic Quality studies in Germany [15], the Lithuanian landscape aesthetic quality gradation [16], the assessment of visual landscape features in Helan Mountain National Nature Reserve in China [17], the quality evaluation of the Baikal Lake landscape [18], and the Sonoyta Plain visual landscape assessment in the U.S. [19]. Because it views the landscape from an indirect perception perspective, this paradigm is suited for large-scale assessments, focusing on the physical attributes that can predict scenic beauty [13]. Nevertheless, it overlooks the dynamic and interactive aspects of human perception and lacks integration with real-time human sensory experiences.
• Subjective Paradigm (Based on Social Media Data): Leveraging large-scale social media data and photo archives, researchers analyze how public aesthetic preferences map onto a given area. This approach can be applied at both regional and local scales; however, the evaluation’s accuracy is contingent on the volume of available data. It provides valuable insights into public preferences and perceptions, yet it often suffers from spatial bias and uneven data coverage [7].
• Combined (Objective–Subjective) Paradigm (Addressing Both Direct and Indirect Perception): This approach synthesizes wide-ranging public participation with photographic datasets to derive values for visual landscape preferences. It then computes landscape indicators within the field of view and constructs regression models to explore the relationship between physical attributes and subjective preferences, offering a holistic assessment of the setting and sensory experience. Typical applications include scenic beauty assessments in the Alps [20], Germany-wide landscape aesthetic studies [21], research in Saxony, Germany [22], and visual landscape evaluations in Queensland, Australia [23]. While these methods help bridge the gap between objective data and subjective assessments (e.g., photo archives or crowdsourced data), they remain reliant on comprehensive photo datasets and face challenges in quantifying the dynamic interactions between physical landscape properties and human sensory perceptions [24,25].

Compared with the above research paradigms, the introduction of affordance theory offers new possibilities for more deeply elucidating human–environment interactions. Originally proposed by James J. Gibson in ecological psychology, affordance theory posits that the environment can “provide” specific behavioral opportunities for an organism, and these possibilities not only depend on the organism’s perceptual capacities but also unfold through mutual interaction [26]. From a design and human–computer interaction perspective, Norman further emphasizes that an “affordance” is neither a purely objective property of the environment nor a subjective projection of the individual, but rather a potential function formed through the convergence of both in a given context. Although affordance theory has seen initial applications in fields such as ecological perception, urban landscape interaction, and elderly-friendly space design [27,28,29,30], it has thus far been less frequently employed in the evaluation of landscape visual resources. Existing studies often address questions such as “Which areas exhibit high visual sensitivity?”, “Which places possess high landscape aesthetic quality?”, or “How do people perceive a particular landscape?” but generally fail to tackle more profound managerial challenges—specifically, assessing the potential for landscape visual resources to deliver aesthetic functions and to be captured by visual senses (for instance, identifying areas with vast visual fields but insufficient scenic quality, or areas that are aesthetically appealing yet difficult to be seen or used). In other words, a comprehensive discussion that integrates both the “physical attributes and aesthetic functions” of landscapes with the “ability of the scenery to be captured by the visual senses” is still lacking.

To bridge this gap, the present study adopts an affordance-based perspective, integrating both “physical landscape affordance” and “visual sensory affordance” into a more holistic evaluation framework. This enables us to reveal the potential visual resources and experiential value landscapes may offer society from both physical and sensory standpoints, ultimately providing practical guidance for future landscape planning and management strategies. Specifically, this research focuses on the following objectives:

• Clarify the Concept and Indicators of Landscape Visual Affordances: Grounded in ecological psychology and visual perception theory, we articulate a holistic definition of landscape visual affordances. By integrating physical landscape affordances with visual–sensory affordances, we present a comprehensive set of indicators to quantify the capacity of landscape resources to supply visual experiences, thereby filling a critical theoretical gap.
• Conduct Integrated Analyses Using Geographic Information Systems (GISs): Employing elevation data, land use data, and remote sensing imagery, we analyze the physical characteristics and aesthetic potential of landscapes to identify visual hotspots and key landscape elements. Our quantitative method ensures scientific rigor and practicality.
• Apply the Theoretical Framework to the Changhong Region of Qianjiangyuan National Park to Validate Its Practicality and Scientific Value: Through a case study in an actual protected area, we demonstrate how this framework can inform conservation planning and the management of visual resources. We also provide concrete recommendations for visual landscape management, serving as a reference for subsequent research and practice.

2. Materials and Methods

2.1. Theoretical Framework: Integrating Affordances and Visual Landscapes

To evaluate the potential for using scenic aesthetic resources, this study adopts affordance theory and proposes a “landscape visual affordance (LVA)” framework (Figure 1). First, building on the premise of affordance theory, we recognize that whether a landscape is truly “seen” or “used” depends not only on its physical attributes (e.g., terrain morphology, land use, and vegetation coverage) but also on human visual–sensory processes in real environments. Focusing solely on physical factors makes it difficult to explain how well landscape visual resources are perceived, whereas concentrating only on sensory dimensions may overlook how spatial structures and aesthetic elements underpin the act of appreciation. Therefore, LVA is divided into two core dimensions:

2.1.1. Physical Landscape Affordance (PLA)

PLA refers to the potential aesthetic and functional conditions a landscape provides for visual experiences, based on its physical attributes, such as topography, land use, and vegetation coverage. Its key components include:

• Landscape visual characteristics, formed by combining the landform (slope, aspect, and general curvature), vegetation coverage (trees, shrubs, bamboo forests, nurseries, etc.), and land use types (farmland, forest land, grassland, villages, roads, water bodies, etc.) into distinct landscape patch types [31].
• Aesthetic features derived from these landscape patch types, such as naturalness, diversity, historicity, seasonal variation, harmony, and visual scale [32,33,34]. Essentially, PLA classification is grounded in classifying these aesthetic features.

These factors are chosen because they capture both the morphological and functional dimensions of how landscapes can support human activities and aesthetic experiences at the physical level. Numerous studies in scenic evaluation, ecology, and land use research have confirmed that variables such as landform, land use types, and vegetation coverage are central for depicting landscape structure and morphology. Meanwhile, “aesthetic features” (incorporating concepts of diversity, seasonal variation, historic elements, etc.) gauge the landscape’s potential aesthetic functionality. By integrating these aesthetic metrics with physical attributes such as terrain and land use, we can assess both the physical qualities of the environment and its capacity to deliver aesthetic enjoyment. Moreover, these factors can typically be quantified via GISs, remote sensing data, and digital elevation models (DEMs), facilitating their application at large scales and across multiple regions.

2.1.2. Visual Sensory Affordance (VSA)

VSA examines the potential for a landscape to be received by human vision; only when the scenery is “seen” can it generate aesthetic or sociocultural value. Hence, VSA focuses on how a landscape can enter one’s field of view and attract attention during the viewing process, centering on the factors that influence “being observed” and “being perceived”. These usually involve the viewer’s visible range and attention focal points, mediated by location, time, view corridors, dynamic changes, and overall landscape appeal. Areas high in VSA can either be more susceptible to disturbance or possess strong visual allure, making them critical in conservation and management. Key elements include:

• Location factors such as the layering of the foreground, middle ground, and background, viewpoint position and elevation, and the spatial interplay of landscape elements and potential obstructions. Research in visual ecology and environmental psychology indicates that different vantage points and distance gradients profoundly affect individuals’ sensory stimulation and aesthetic judgments [35]. Terrain, viewing distance, and barriers determine “what can be seen” and “how it is seen”, thus shaping the overall capacity for visual reception.
• Temporal factors such as viewing duration, viewing frequency, and seasonal variation. Because lighting conditions differ between day and night or during dawn and dusk, a landscape’s color, brightness, and layering are directly affected; viewing the same spot repeatedly or only briefly can produce varying degrees of visual impact and memorability. Seasonal shifts in leaf color, light intensity, and weather conditions can also alter a landscape’s appearance and aesthetic value.
• Visual salience such as color contrasts, terrain ruggedness, the diversity and seasonal dynamics of vegetation, and prominent landscape elements. The human visual system is highly sensitive to color contrasts, abrupt shapes, and distinctive landmarks. In a scenic context, striking color differences or unique landforms are more readily perceived.

When PLA and VSA jointly take effect, they can encourage various visitor activities (e.g., sightseeing, photography, and cultural exploration), thereby generating diverse aesthetic or cultural benefits. In national parks or protected areas, the attributes of a given landscape can inspire or constrain different types of behavior, while also shaping visitors’ psychological experiences. The framework of “high PLA + high VSA” highlights places that offer both strong aesthetic appeal and ease of visibility. Conversely, areas with high PLA but low VSA may be visually striking but remain overlooked due to limited visibility or accessibility. Hence, within the LVA framework, it is critical to go beyond the mere inventory and scoring of scenic resources. A more refined approach involves identifying which landscapes merit targeted protection and management, and how best to guide visitor experiences. Below are illustrative categories of “potential behaviors and psychological experiences” that landscapes may evoke, corresponding to variations in PLA and VSA:

• Stimulating Curiosity and a Desire to Explore: When a landscape contains multiple elements and varied boundaries (high complexity and abundant edges), visitors tend to linger, take photographs, and engage in exploration. Such visual richness can prompt extended stays, group discussions, and shared experiences.
• Immersive Experiences: Areas with high ecological integrity foster a sense of relaxation and eco-friendly tourism, enabling visitors to “escape” from urban life.
• Science Popularization and Environmental Education: Rich biodiversity or striking natural features may encourage guided tours, interpretive signage, and conservation-oriented activities.
• Contemplation and Healing: quiet, unspoiled environments and natural landscapes with minimal human interference can support meditation, health-oriented tourism, and restorative recreation.
• Cultural Tourism: Historical buildings, traditional farmland, seasonal flower fields, and distinct cultural or seasonal attributes can appeal to culture enthusiasts or “Instagram tourists”, turning certain sites into “hotspots”.
• Emotional Attachment and Local Identity: Traditional scenic resources and unique regional landscapes can enhance a sense of cultural belonging and collective memory, prompting visitors to revisit or develop a stronger place identity.

2.2. Study Area Overview

This study focuses on the Changhong region of Qianjiangyuan National Park, encompassing approximately 138.4 km². The area exhibits a typical “mountain–water–forest–farmland–village” spatial pattern. In addition to its abundant natural features, such as evergreen broad-leaved forests and river valleys, the region preserves a relatively intact agricultural landscape. Characterized predominantly by mountainous and hilly terrain, the deeply incised valleys are mostly “V”-shaped with narrow gorges and clearly defined ridgelines. The main watercourse is Changhong Creek, running north–south through the region, and the extensive hydrological network displays a dendritic pattern. Due to hydropower stations, reservoirs, and other water-related infrastructure, the water flow varies significantly across different creek sections: some segments have become almost dry, exposing gravel-filled riverbeds overgrown with weeds.

With a forest coverage rate of 85.1%, the Changhong region offers a secluded natural environment, and its evergreen broad-leaved forests retain a high visual quality value throughout all seasons. Rural farmland covers 8287 mu, consisting primarily of terraced fields and level farmland situated between valleys; the fields are cultivated in rotation. From March to April, blooming rapeseed (canola) flowers provide optimal viewing conditions for the agricultural landscape. Village settlements are clustered along the watercourses, and most houses have been rebuilt, leaving only a few traditional dwellings (Figure 2).

2.3. Research Framework

This study employs a multidimensional assessment framework to evaluate landscape visual affordance, integrating physical landscape affordance (PLA) and visual sensory affordance (VSA) (Figure 3). The framework establishes a systematic process for identifying, classifying, and categorizing landscape characteristics, offering valuable insights for conservation and management.

The research consists of two primary components. The first part addresses measurable physical attributes in the landscape that facilitate the realization of its visual quality potential. Key factors include slope and aspect, general curvature, vegetation coverage, and land use types. These features are quantified, classified into visual landscape types, and then linked with landscape metrics to evaluate visual quality functions. The second part targets the visual sensory affordance, examining how a landscape can be perceived by observers, with an emphasis on integrating sensory and spatial dimensions. Core elements include relative slope, visibility, viewing probability, relative distance, and degree of visual prominence [36,37,38]. PLA highlights the landscape’s inherent potential to provide visual quality, whereas VSA focuses on whether and how these qualities are perceived by the human eye. The combination of these two dimensions reflects the comprehensive capacity of a landscape’s visual affordance, offering guidance on preserving and managing visual resources in national parks.

Both dimensions carry an equal weight of 0.5, reflecting their balanced contribution to overall visual affordance and ensuring that objective physical attributes and subjective perceptual factors are regarded as equally important. This equal weighting serves as a general starting point, minimizing additional biases or complexities that might arise from context-specific adjustments [15,16]. The digital elevation model, land use data, and vegetation coverage data used in this study were provided by the Qianjiangyuan National Park Administration (QNPA). Radiometric correction was carried out in line with the protocols described by Chander and Markham [39], which involved adjustments for sensor calibration and post-calibration dynamic ranges. Furthermore, topographic correction methods, as evaluated by Hantson and Chuvieco [40], were applied to reduce illumination effects caused by terrain variations. Techniques were also incorporated to assess and correct for vegetation variability and change over time [41]. To ensure the geospatial accuracy of the remote sensing data, GPS field point validation was conducted by QNPA staff, thereby confirming that the satellite-derived observations aligned closely with ground truth measurements.

2.4. Aessessment of Physical Landscape Affordance

2.4.1. Assessment of Landscape Visual Characteristics

The evaluation of landscape visual characteristics employs an overlay method in conjunction with a clustering method. The overlay method integrates various spatial factors—such as soil, slope, and land use attributes—into a comprehensive scenic information map. Although this approach produces a large number of polygons sharing common landscape attributes, these polygons lack clear categorization, requiring further integration and optimization. Clustering analysis, on the other hand, uses statistical methods to group landscape visual characteristic factors and is currently the primary technique for identifying landscape character areas (LCAs); however, it also needs to clarify specific landscape attributes [42]. Accordingly, this study combines both approaches—conducting an overlay analysis followed by clustering—to leverage the strengths of each method and yield a refined classification of landscape visual types. Indicator calculations are then performed to quantify the characteristics of these landscape types.

On a spatial scale, landscape visual characteristics can be represented by landscape patches or types. According to the analysis by [42], topographic and land use factors are most commonly chosen because they directly affect how a given area’s visual landscape is distinct from others. In the mountainous and forested Changhong region, we selected topography, geomorphology, land use, and vegetation cover to classify the landscape’s visual characteristics. Using the Surface Analysis tools in ArcMap 10.8 and a digital elevation model (DEM), we derived three geomorphological indicators—slope, aspect, and curvature—and then applied a reclassification process based on visual similarity and differentiation [17] (

Table 1. Classification of Landscape Visual Characteristics.

Category	Classification Basis	Classification
Slope	Natural Breaks	Gentle Slope (0°~5°), Mild Slope (5°~10°), Moderate Slope (10°~15°), Medium Slope (15°~25°), Steep Slope (25°~30°), Steeper Slope (30°~45°), Extremely Steep Slope (>45°)
Aspect	Sun Position	Flat (−1°), Shadowed Area (0°~45°, 315°~360°), Semi-Shadowed Area (45°~135°), Sunlit Area (135°~225°), Semi-Sunlit Area (225°~315°)
General Curvature	Topography	Valley (<0 m−1), Range (>0 m−1)
Land Use	Visual Characteristics	Forest, Grassland, Built Area, Farmland, Orchard, Water Area, Swamp, Forest Land, Bamboo, Tea Garden
Vegetation Cover	Type Name	Forest, Unplanted Area, Special Woodland, Agricultural Plantation, Bamboo, Shrubs, General Timberland, Nursery, Burned Area, Other

). We subsequently overlaid the reclassified slope, aspect, land use, and vegetation cover data. The overlaid output was then subjected to k-means clustering in R (version 4.4.3).

The underlying principle of k-means clustering begins by randomly selecting K. K objects as initial cluster centers. Next, the distance between each object and each potential cluster center is calculated, assigning each object to the cluster center closest to it. In detail, the process entails the following five steps:

(1) Select K random samples from the dataset as initial cluster centers (z$z_1,z_2,z_3\dots \dots ,z_k$);

(2) For each sample $x_i$, find the nearest cluster center $z_{\nu }$, and assign $z_{\nu }$ to the corresponding cluster $u_{\nu }$;

(3) Recalculate the cluster centers using the mean of all samples in each cluster;

(4) Calculate

$$D={\Sigma }_i=1^n\left[{\mathrm{m}\mathrm{i}\mathrm{n}}_{r=1,\mathcal{L},k}d{\left(x_i,z_r\right)}^2\right]$$

(1)

(5) If the value of DDD converges, the algorithm terminates; otherwise, repeat from step 2.

Before running k-means clustering, data standardization is necessary to eliminate biases stemming from different scales of the attributes. The elbow method is employed to determine the optimal number of clusters K, based on computing the total within-cluster sum of squares (WSS) for different values of K. Following the selection of an optimal K, the k-means algorithm is executed, and the resulting landscape types serve as the fundamental units for computing and evaluating landscape visual characteristics.

2.4.2. Aesthetic Quality Evaluation

Aesthetic quality encompasses previously established conceptualizations of the landscape’s visual characteristics. Due to computational limitations, visual proportion cannot be quantified [15,16]; however, the remaining concepts are paired with corresponding evaluation metrics [34,43] and supported by theoretical foundations as follows:

(1) Complexity is defined as the diversity and richness of landscape element types and boundaries. Essentially, it constitutes a core dimension of visual diversity and can be measured by Shannon’s diversity index (SHDI), edge density (ED), and patch richness (PR).

• SHDI captures the diversity (i.e., both richness and evenness) of different patch types within a given area [44]. A higher SHDI indicates a more diversified composition of patches—such as forests, water bodies, agricultural lands, and built-up areas—and a more uniform distribution of land cover types. From a visual perspective, greater diversity in patch types tends to attract interest and engender a richer visual field. Moreover, “visual complexity” is also associated with the number of perceived edges or transitions [33]. More frequent edges or transitions can enhance the visual interest and complexity of a scene.
• ED measures the total length of patch boundaries per unit area. A higher ED implies that the spatial structure contains more patch edges, transitions, and fine-scale contrasts. Visually, abundant edges require the eye to process more “breaks” or “boundaries”, thereby contributing to the overall perception of spatial complexity [34].
• PR reflects the number of different land use types, vegetation types, or other functional units within the landscape. A higher PR indicates the presence of a greater variety of patches, and this heterogeneity directly increases the variety and stratification of visual elements, resulting in a more diverse and dynamic visual appearance that enhances visual diversity. Additionally, PR is positively correlated with a “sense of wonder”, which is a core psychological mechanism underlying imageability [11].

(2) Imageability is defined as the quality presented by the overall pattern or constituent elements of a landscape that creates a strong visual image in the observer’s mind, thereby endowing the landscape with a distinctive identity. Patch density (PD) and the landscape shape index (LSI), as important quantitative indicators of the landscape’s spatial structure, provide theoretical support for imageability:

• Imageability requires that the landscape contains visually salient and memorable components. A higher PD generally indicates that there are more elements of differing functions or types present, offering observers a wealth of visual information that can enhance the sense of identity and uniqueness of the landscape. When PD is high, it suggests that the number of patches is large and their distribution is dense, which may lead certain patches—due to their close proximity or specific arrangement—to form distinct visual focal points or landmark features, thereby enhancing the vividness and memorability of the landscape [22].
• The LSI primarily reflects the complexity of patch boundaries and the irregularity of their shapes. The “panoramic” characteristics and “spectacular elements” of a landscape are often closely related to the complexity of its edges and the uniqueness of its shapes. Complex and distinctive forms tend to capture visual attention, forming prominent landmarks or recognizable images. A higher LSI indicates that patch shapes are complex and variable; this irregularity and complexity may enhance the visual impact of certain patches within the overall landscape, making them stand out and thus increasing the landscape’s imageability. Furthermore, complex boundaries and shapes help construct a richer visual texture, ensuring that the landscape presents a unique and attractive image from different viewing angles and distances [31,43,45].

(3) Coherence captures the degree of coordination and consistency among different landscape elements and plays an organizational role in balancing visual diversity [46]. It is measured using the aggregation index (AI) and the Shannon evenness index (SHEI). The AI indicates whether landscape elements cluster into cohesive patches, reflecting structural uniformity and organizational clarity that shape an observer’s overall impression. Meanwhile, the SHEI addresses the balance among various patch types. From a perceptual and psychological standpoint, a more balanced distribution of patch types enhances harmony and predictability—key qualities of a coherent landscape.

(4) Perceived naturalness is represented by the area-weighted naturalness (AWN), which is derived from land use data that assign each landscape feature type a specific naturalness score. These scores reflect a gradient of human intervention, ranging from fully undisturbed (very high naturalness) to heavily anthropogenic (very low naturalness). The classification includes five levels:

• Very high naturalness (score 5): primarily tree-covered areas (e.g., broad-leaved forests and bamboo groves) and water bodies with minimal human interference.
• High naturalness (score 4): other forested areas with limited human activity.
• Moderate naturalness (score 3): shrublands, grasslands, and dry farmland.
• Low naturalness (score 2): irrigated paddy fields, tea plantations, orchards, and managed woodlands under more intense cultivation.
• Very low naturalness (score 1): construction land, strongly dominated by human interventions.

This classification aligns with the continuum of nearly undisturbed ecosystems to heavily modified environments.

(5) Uniqueness is computed through the area-weighted uniqueness (AWU) metric, based on landscape ecology’s conceptualization of distinctiveness [44]. AWU ensures that features with larger areas make proportionally greater contributions to the region’s overall uniqueness. Both land use type and area proportion jointly determine this score. In this study, uniqueness was assigned in a dichotomous manner, with only two possible values: 0 (non-unique) and 1 (unique).

In the Changhong region of Qianjiangyuan National Park, locally adapted, seasonally cultivated paddy fields—arranged in a “village–paddy–forest” pattern—are highly representative of the regional cultural landscape and strongly aligned with the concept of visual uniqueness [34,43]. Hence, paddy fields are designated as a unique landscape type. Moreover, given that the primary goal of establishing Qianjiangyuan National Park is to protect extensive evergreen broad-leaved forests, tree-covered areas also qualify as a unique landscape type. By employing AWU, large forest patches can appropriately influence the overall assessment of uniqueness for the entire region (

Table 2. Landscape visual character indicators.

Visual Character Concept	Definition [15,16]	Indicators	Calculation Method	Potential Behaviors/Psychological Experiences
Complexity	Diversity and richness of landscape element types and boundaries.	SHDI	$SHDI=\sum _{i=1}^R{P}_i\times \ln P_i$ where Pi is the proportion of each land cover type, and R is the total number of land cover types.	Stimulating Curiosity and Exploration. When a landscape has highly diverse elements and more complex boundaries, visitors tend to linger, delve into deeper exploration, and spend more time taking photos or chatting. An expansive field of view often increases their willingness to stay longer. Promoting Photography or Filming Activities: A multilayered or dynamically changing scene is easier to frame and more likely to satisfy visitors’ desire for aesthetically pleasing photographs.
		ED	$ED={\displaystyle \frac{E}{A}}\times 100$ where E is the total length of edges between different patch types, and A is the total landscape area.
		PR	The total number of different patch types within a landscape.
Imageability	Strong visual imagery conveyed by a combination of one or more landscape elements, making the landscape identifiable.	PD	$PD={\displaystyle \frac{PN}{A}}\times \left(\mathrm{10,000}\right)\times 100$ where PN is the number of patches, and A is the total area.	Shaping Identity and “Brand” Recognition: Landscapes with a high shape index or visually distinctive patch structures can become recognizable landmarks, guiding visitors to specific sites or “check in” points. Enhancing Memorability: Fresh, striking visual images help visitors remember the location and increase the likelihood of return visits.
		LSI	$LSI={\displaystyle \frac{P}{2\sqrt{\pi A}}}$ where P is the perimeter of the patch, and A is the area of the patch.
Coherence	Aggregation of similar landscape elements in space and the balance between different elements. Coherence involves not only aggregation but also coordination and balance.	AI	$AI=\left({\displaystyle \frac{g_{ij}}{max\left(g_{ij}\right)}}\right)\times 100$ where $g_{i}j$ is the number of like adjacencies (joins) between pixels of patch type i, and $max\left(g_{ij}\right)$ is the maximum possible number of like adjacencies.	Comfort and Relaxation: Landscapes with a high degree of coherence or balance can foster a sense of “visual calm” and emotional relief, encouraging visitors to rest or daydream.
		SHEI	$SHEI={\displaystyle \frac{-\sum _{i=1}^R{P}_i\times \mathrm{l}\mathrm{n}{P}_i}{\mathrm{l}\mathrm{n}R}}$ Where $P_i$ is the proportion of the i-th type (e.g, species or land cover type) relative to the total number, and R is the total number of types (e.g., total number of species or patch types).
Stewardship and Disturbance and naturalness	The gradient of natural perception resulting from human intervention and management reflects a continuum ranging from entirely natural to highly artificial states.	AWN	$\mathrm{A}\mathrm{W}\mathrm{N}={\displaystyle \frac{\sum _{j=1}^n a_{ij}}{A}}\times N_i$ where $P_i$ is the percentage of type $i$ in the landscape, $N_i$ is the assigned naturalness value of type $i$, and ${\alpha }_{ij}$ is the area of patch $ij$.	Immersive Experiences: Highly natural environments can promote mindfulness, ecotourism, and a sense of escape from urban life. Knowledge and Environmental Education: Pristine ecological sites are valuable for interpretive walks, guided tours, and raising awareness of conservation. Contemplation and Healing: Undisturbed landscapes can support meditative or health-oriented tourism, fostering recovery and mental well-being.
Historicityand Ephemera	Assessment of cultural elements with varying temporal scales, as well as elements and land cover types that change with seasonal and weather variations.	AWU	$\mathrm{A}\mathrm{W}\mathrm{U}=(1-P_i)\times R_i$ where AWU is the Area-weighted uniqueness; $R_i$ is the Uniqueness value assigned to type $i$; and $P_i$ is the Proportion of type $i$ within the landscape.	Cultural Tourism: Historic buildings, traditional farmland, or seasonal flower fields often appeal to culture enthusiasts or social media-oriented travelers, turning certain places into “viral” destinations. Emotional Attachment and Local Identity: Traditional or seasonal scenery can strongly resonate with cultural memories, evoking nostalgia or reinforcing place-based identity.

).

The LVCA map is imported into FRAGSTATS 4.3 to compute the aforementioned indices. The resulting outputs are then visualized in ArcMap 10.8 and processed using the Raster Calculator according to Formula 2. Subsequently, the landscape’s aesthetic quality is classified into five levels (1–5) using the natural breaks method, which serves as the final PLA grading. A PLA level of 5 indicates that the landscape possesses a high potential for aesthetic value, whereas a level of 1 signifies that the inherent quality of the landscape’s aesthetic is low.

PLA = w₁ED_Norm + w₂LSI_Norm + w₃PR_Norm + w₄PD_Norm + w₅AI_Norm + w₆SHEI_Norm + w₇SHDI_Norm + w₈AWN_Norm + w₉AWU_Norm

(2)

Here, ED = edge density, PD = patch density, PR = patch richness, LSI = landscape shape index, AI = aggregation index, SHEI = evenness index, AWN = area-weighted naturalness, and AWU = area-weighted uniqueness. Principal component analysis (PCA) was used to determine indicator weights. PCA is a dimensionality reduction technique that transforms high-dimensional data into lower dimensions while preserving variance. It converts original indicators into a set of new variables (principal components), with each component being a linear combination of the original indicators.

2.5. Assessment of Visual Sensory Affordance

2.5.1. Visual Salience Factors: Relative Slope and Prominence

The U-shaped cross-section of the Changhong area’s valleys makes the slope a critical factor influencing the visual sensitivity of landscapes. Observers typically view the landscape from valley bottoms or nearby trails, gazing upward. Steeper slopes increase the visible area and the likelihood of being noticed. Using DEM data, slope distribution was analyzed with GIS tools, classifying slopes into five categories: gentle slopes (0°–5°), mild slopes (5°–15°), moderate slopes (15°–30°), steep slopes (30°–45°), and extremely steep slopes (>45°). Prominence refers to the human tendency to focus on areas with noticeable differences from their surroundings. For larger regions, the terrain ruggedness index (TRI) is used to measure prominence:

$$TRI=\sqrt{{\displaystyle \frac{\sum _{i=1}^n (H_i-H_c{)}^2}{n}}}$$

(3)

where $H_i:\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{v}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{o}\mathrm{f} \mathrm{n}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{b}\mathrm{o}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{p}\mathrm{i}\mathrm{x}\mathrm{e}\mathrm{l}\mathrm{s}$, $H_c:\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{v}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{o}\mathrm{f} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{l} \mathrm{p}\mathrm{i}\mathrm{x}\mathrm{e}\mathrm{l},\mathrm{a}\mathrm{n}\mathrm{d} n:$ number of pixels in the neighborhood.

2.5.2. Location Factors: Visibility and Visual Distance

The sensory affordance of a landscape is closely tied to the observer’s spatial relationship relative to the landscape itself. Generally, the closer the landscape is to the observer or travel route, the greater its potential to be perceived. Based on existing studies, distance ranges can be divided into four zones: foreground (0–400 m), middle ground (400–800 m), background (800–1600 m), and distant view (>1600 m) [10,35].

In ArcMap 10.8, road data from land use records were loaded, and observer points were generated at 30 m intervals. To enhance analysis accuracy, specific height values were assigned to different vegetation types and buildings. For instance, grasslands and areas without vegetation were set to 0 m, forests to 20 m, buildings to 10 m, bamboo groves to 15 m, tea plantations to 5 m, and shrubs to 3 m [10,47]. These assigned heights were then overlaid onto the original DEM to produce the elevation data required for viewshed analyses. Finally, multi-ring buffering and the Viewshed tool were used to calculate the fields of view at each distance zone.

2.5.3. Temporal Factors: Visual Probability

Time factors encompass elements such as viewing duration and viewing frequency (visual probability). Whether a particular location is revisited multiple times or viewed only briefly can substantially affect the degree of visual impact it imparts on observers, as well as the memorability of the experience. Visual probability refers to the likelihood that a particular portion of the landscape can be observed within the study region, and it is positively correlated with sensory affordance. By calculating the cumulative visibility in ArcGIS, a raster layer representing the visible surface can be generated. This approach not only depicts the line-of-sight accessibility under specific viewpoint conditions but also indicates, for each raster cell, how many times that location can be observed. Compared with simply determining visual probability from a topographic map, quantifying visibility through the cumulative values on a raster surface proves more effective.

The procedure begins by using the viewshed analysis tool to compute individual viewsheds for all observation points across the study area, resulting in single-point viewshed raster layers. These layers are then overlaid to create a composite viewshed raster whose pixel values represent the total number of observation points from which each pixel is visible. Finally, the pixel values are normalized to derive the relative probability of being viewed for each cell, and the results are classified into five categories for mapping purposes.

2.5.4. Integrated Visual Sensory Affordance

These factors frequently interact and jointly shape visual affordance. Therefore, after evaluating each factor individually, a comprehensive overlay assessment is carried out. This process essentially entails the conjunction (∧) of three components—“relative slope”, “visual distance”, and “visual probability”—along with the disjunction (∨) of “prominence”. Finally, a composite sensory affordance score is derived using Formula (4).

$$VSA=\left(Relative Slope\wedge Visual Distance\wedge Observation Probability\right)\vee Prominence$$

(4)

Finally, regions are classified into five categories—extremely low, low, medium, high, and extremely high.

3. Results

3.1. Results of Physical Landscape Affordance

Using ArcMap 10.8 software, the entire study area was divided into 758 distinct landscape visual character areas (LVCAs). These LVCAs were further integrated into 20 significantly heterogeneous visual landscape characteristic types through clustering analysis. The classification results reveal that the landscape visual characteristics of the study area are primarily composed of the following types: the three most dominant types are “steep, semi-shaded valley broad-leaved forests”, “moderate, semi-shaded valley forests”, and “steep, semi-sunny valley broad-leaved forests”. The distribution of these types indicates that the region’s visual landscapes are dominated by arboreal forests, with slope and aspect playing significant roles in shaping them. Combinations of different aspects (semi-shaded and semi-sunny) and terrain characteristics not only determine vegetation diversity but also shape the layering and richness of visual landscapes (Figure 4).

Using the landscape visual characteristic type map, landscape indicators were quantitatively analyzed for each characteristic type. The comprehensive PLA distribution (

Table 3. The proportions of each PLA level.

PLA Level	Count	Area (km2)	Account (%)
1	13,722	32.9617	23.92
2	8485	20.3818	14.79
3	12,868	30.9103	22.43
4	14,903	35.7986	25.98
5	7387	17.7443	12.88

) was derived from these indicators, providing a classification basis for the visual quality of the study area (Figure 5a). Among the PLA classifications, the five types with the highest visual quality include “moderate, semi-sunny slope valley shrubland villages”, “gentle shaded slope valley broad-leaved forests“, “gentle semi-shaded slope valley water bodies”, and “gentle semi-sunny slope valley water bodies”. These areas exhibit high terrain complexity, with unique combinations of steep slopes and semi-sunny aspects, especially moderate and gentle slopes, creating diverse visual experiences for observers. In semi-shaded and shaded areas, alternating distributions of broad-leaved forests and trees contribute to high visual diversity. Additionally, “gentle semi-shaded slope valley water bodies” and “gentle semi-sunny slope valley water bodies” scored highly in PLA, highlighting the significant role of water bodies in enhancing regional visual quality. Some landscapes with human activities, such as “moderate semi-sunny slope valley shrubland villages”, also exhibited high visual quality. The integration of villages with natural surroundings in these areas reflects a blend of cultural and natural aesthetics, creating a rural characteristic with strong visual appeal.

Overall, the PLA distribution exhibits spatial heterogeneity. High-quality visual areas (PLA levels 4 and 5) account for a relatively large proportion of the area, while medium- and lower-quality areas (PLA levels 3 and below) comprise approximately 60% of the total area. This indicates that the physical landscape affordance is relatively evenly distributed across the region.

3.2. Results of Visual Sensory Affordance

Based on the relative slope factor affordance classification shown in Figure 6a and

Table 4. Relative Slope Affordance.

Classification	Relative Slope	Account (%)	Area (km2)
1	Gentle slopes (0°–5°)	13.93	18.5703
2	Mild slopes (5°–15°)	26.48	35.2973
3	Moderate slopes (15°–30°)	28.45	37.9252
4	Steep slopes (30°–45°)	20.88	27.8250
5	Extremely Steep Slopes (>45°)	10.26	13.6725

, the affordance of the relative slope in the Changhong region closely aligns with the actual terrain slope. Steep terrain areas at moderate (level 3) or higher affordance levels are more likely to draw attention, covering approximately 60% of the total area; these zones are primarily located in the central portion and along the peripheral boundaries of the study region. In contrast, areas with medium–low (level 2) or low (level 1) affordance, where the terrain is comparatively gentle, tend to attract less visual attention and are mainly concentrated near the central roadway corridors. As indicated by the prominence classification in Figure 6b and

Table 5. Landscape Prominence Classification.

Classification	Account (%)	Area (km2)
1	00.01	0.020312
2	00.11	0.14980
3	00.31	0.4163
4	50.58	68.7945
5	48.99	66.6388

, the Changhong region exhibits high visual prominence due to its dramatic elevation changes, steep slopes, and rugged landforms, which readily capture observers’ attention and become focal points in the field of view. Additionally, patches with high prominence are both numerous and spatially diverse, highlighting the expansive and heterogeneous nature of this rugged terrain.

According to the visibility classification in Figure 6d, most of the Changhong region lies within the visible range, with only 9.86% of the area remaining out of sight. These invisible sections are mainly located in secluded valleys with limited road accessibility and certain core protected zones. From the relative distance classification shown in Figure 6c and

Table 6. Visibility and Visual distance Classification.

Classification	Visibility and Visual Distance	Account (%)	AREA (km2)
1	Invisibility	9.86	13.3831
2	Distant background (>1600 m)	6.18	8.3904
3	Background (800–1600 m)	12.94	17.5636
4	Middle ground (400–800 m),	26.47	35.9322
5	Foreground (0–400 m)	44.55	60.4683

, the concept of “relative distance affordance” derives from the theoretical basis of normal human visual capability and the framework of visual landscape scale, manifesting as a series of buffer zones along viewing routes. The closer the distance between observers and visible focal points, the clearer the observed scene becomes. High-affordance areas (level 4) are primarily distributed within near-view zones flanking major roads, occupying 46.41% of the total Changhong region and drawing substantial visual attention. Mid-range view distances span 400–800 m, enabling relatively clear observation by the human eye and accounting for 26.47% of the total area. Far-view distances range from 800 to 1600 m, in which visibility is moderately discernible, covering 20.94% of the entire study area.

The visual probability analysis depicted in Figure 6e indicates that the more frequently an area is observed, the higher its sensory visual affordance. Within the study region, the number of times a landscape can be seen from designated observation points ranges between 0 and 2809. Based on these values, a classification of natural breaks in ArcMap 10.8 was applied to categorize visual probability sensitivity into five levels. Low-occurrence areas (0–120 views) dominate the landscape, making up 91.38% of the total area (1240.64 ha). Medium-occurrence zones (120–310 views) account for 7.63% (103.62 ha). Areas with the highest visibility (1000–2809 views) encompass only 0.03% of the region (0.36 ha). This distribution pattern suggests that most of the study area is characterized by limited viewing frequency, with only a small proportion exhibiting distinctly high visibility (

Table 7. Frequency of Being Seen (Visual Probability Classification).

Classification	Frequency of Being Seen	Account (%)	AREA (km2)
1	0–120	91.38	124.0636
2	120–310	7.63	10.3617
3	310–580	0.84	1.1450
4	580–1000	0.11	0.1548
5	1000–2809	00.03	0.0355

).

Based on the composite VSA classification results (Figure 5b,

Table 8. Visual Sensory Affordance Classification.

VSA-Level	Account (%)	AREA (km2)
1	6.86	9.1494
2	4.61	6.1485
3	13.61	18.1453
4	30.81	41.0923
5	44.11	58.8309

), five classification levels were identified along with their respective proportions and areas. High-VSA regions (Class 5) account for 44.11% (58.8309 km²) of the total area, followed by class 4 at 30.81% (41.0923 km²). These high-VSA zones indicate landscapes that can be readily perceived by observers. Class 3, representing moderate VSA, occupies 13.61% (18.1453 km²), whereas low-VSA classes 1 and 2 cover 6.86% (9.1494 km²) and 4.61% (6.1485 km²), respectively. Overall, these findings suggest that most of the study area demonstrates relatively high visual sensory affordance, effectively attracting visitors’ attention and enhancing their travel experiences. The broad distribution of high-VSA regions highlights their priority in landscape planning and ecological conservation, as they not only offer high-quality visual experiences but may also serve as popular attractions, necessitating prudent visitor capacity management to prevent excessive disturbance.

3.3. Comprehensive Landscape Visual Affordance

Based on the quantitative analysis of PLA and VSA for different LVCTs, the overall LVA for each type was calculated (Figure 7,

Table 9. PLA, VSA, and LVA in each Landscape Visual Characteristic Type.

	Account (%)	Area (km2)	LVCT	PLA-MEAN	VSA-MEAN	LVA-MEAN
1	0.0240	0.0330	Gentle Slope Valley Broadleaf Forest	3.0714	4.3333	3.1667
2	0.0848	0.1169	Mild Slope Shaded Valley Broadleaf Forest	3.2500	4.2381	3.6111
3	0.4590	0.6328	Mild Slope Shaded Valley Shrubland	2.8197	4.2621	3.6387
4	5.3955	7.4388	Moderate Slope Shaded Valley Broadleaf Forest	2.8983	4.2165	3.5691
5	25.3369	34.9321	Moderately Steep Slope Semi-Shaded Valley Broadleaf Forest	2.7718	4.2226	3.5030
6	37.2394	51.3423	Steep Slope Semi-Sunny Valley Broadleaf Forest	2.9745	4.2937	3.6378
7	18.9294	26.0981	Steep Slope Semi-Sunny Ridge Broadleaf Forest	2.9135	4.5236	3.6981
8	7.2149	9.9472	Moderately Steep Slope Sunny Valley Broadleaf Forest with Paddy Fields	2.8587	4.7500	3.7927
9	2.6268	3.6215	Moderately Steep Slope Semi-Sunny Valley Special Shrubland with Paddy Fields	2.9220	4.6870	3.8160
10	2.2305	3.0751	Moderately Steep Slope Sunny Valley Broadleaf Forest with Other Woodland	2.9443	4.7937	3.8691
11	0.0903	0.1245	Moderately Steep Slope Semi-Sunny Valley Shrubland with Villages	3.2745	5.0000	3.8214
12	0.0277	0.0381	Mild Slope Sunny Valley Bamboo Forest with Villages	2.8667	5.0000	3.9545
13	0.0406	0.0559	Mild Slope Sunny Valley Broadleaf Forest with Villages	3.0000	5.0000	4.1000
14	0.0774	0.1067	Moderate Slope Sunny Valley Shrubland with Villages	2.8095	5.0000	3.9545
15	0.0719	0.0991	Moderately Steep Slope Semi-Sunny Valley Broadleaf Forest with Villages	2.7879	5.0000	4.0526
16	0.0811	0.1118	Mild Slope Semi-Sunny Valley Water Area	3.3415	5.0000	4.1765
17	0.0350	0.0482	Mild Slope Semi-Shaded Valley Water Area	2.5882	5.0000	4.0000
18	0.0018	0.0025	Mild Slope Semi-Sunny Ridge Water Area	2.5000	5.0000	3.0000
19	0.0111	0.0152	Moderately Steep Slope Semi-Sunny Valley Water Area	1.8333	5.0000	4.5000
20	0.0221	0.0304	Moderately Steep Slope Semi-Sunny Valley Broadleaf Forest with Water Area	3.0833	5.0000	3.7500

). The differences between PLA and VSA across various types exhibited gradient variations. For instance, “gentle sunny slope valley broad-leaved forests”, with relatively gentle slopes and intact vegetation, often had high PLA and moderate VSA, resulting in an overall LVA in the upper-middle range. Conversely, “steep semi-shaded slope valley forest-shrubland-water bodies”, influenced by pronounced terrain characteristics and water visibility, scored highly for both PLA and VSA, achieving a high overall LVA. Comparisons between “moderate sunny slope valley broad-leaved forests” and “gentle slope valley broad-leaved forests” reveal that although both possess strong aesthetic potential (PLA), the latter’s limited terrain variation and reduced visibility corridors result in lower VSA, leading to a slightly lower overall LVA.

The classification results (

Table 10. Area and Patch Count for Each LVA Classification Level.

LVA	Account (%)	Count	Area (km2)
1	11.00	2857	14.8792
2	22.47	5838	30.4043
3	18.78	4879	25.4099
4	22.16	5757	29.9825
5	25.60	6650	34.6332

) show that the overall LVA of the study area is skewed towards higher levels: LVA levels 4 and 5 account for 22.16% and 25.60% of the area, respectively, comprising nearly half of the total area (49%). These high-LVA regions are typically concentrated in areas with layered terrain, intertwined vegetation and water bodies, and high accessibility, reflecting excellent ecological and landscape resource endowments. Conversely, LVA level 1 patches, though only accounting for 11% of the area, indicate that certain regions have significant shortcomings in physical attributes or visibility, such as overly gentle slopes, monotonous landscape elements, insufficient access, or high environmental disturbances. This differentiation highlights the critical role of spatial heterogeneity and vegetation–terrain coupling in shaping landscape visual affordance. It underscores the necessity of prioritizing protection and moderate management of highly visible and aesthetically valuable key areas.

Overall, the comprehensive LVA classification offers multiple insights for landscape planning and ecological management. On the one hand, attention should focus on the potential value of high-LVA areas in enhancing aesthetic experiences, cultural cohesion, and ecological benefits while avoiding overdevelopment or landscape degradation. On the other hand, for low-LVA areas, measures such as improving the structure of vegetation, adding landscape corridors, or optimizing management practices can enhance visual appeal and environmental quality, achieving overall sustainable optimization of regional landscapes. This comprehensive evaluation system not only quantitatively identifies key areas for protection and potential improvement but also provides scientific support for balancing ecological value and societal needs in landscape planning and land use decisions.

4. Discussion

4.1. Methodological Characteristics

The LVA framework proposed in this study represents a methodological advancement by integrating PLA and VSA. This holistic perspective bridges landscape properties with human perception and interaction, providing a comprehensive evaluation of how well landscape visual resources serve human society. PLA focuses on objective physical attributes—such as topographical relief, land use, vegetation cover, and spatial configurations—that form the landscape’s fundamental aesthetic potential. The indicators employed capture multiple dimensions of visual quality, including complexity (via the SHDI, ED, and PR), coherence (AI and SHEI), and imageability (PD and LSI). This multidimensional approach ensures that the structural and perceptual intricacies of the landscape are accurately reflected [48,49,50].

Perceived naturalness, meanwhile, considers the degree of naturalness of various landscape elements, along with management intensity and human disturbance [14,33,34]. By integrating ecological management considerations with perceptual responses, this method delivers a balanced view, helping to quantify how “natural” a landscape feels to observers [51]. In addition, historical and seasonal variations are used as measures of uniqueness, enabling the framework to include cultural and historical attributes rather than merely ecological and visual factors [52].

VSA stresses the fundamental principle that scenic value emerges only once a landscape is actually perceived, analyzing whether and how the landscape is “seen”, along with factors such as visibility, visual salience, and frequency of viewing. Such perceptual aspects are critical for understanding how landscapes engage human observers, especially regarding visibility and sensory participation. By unifying PLA and VSA, LVA offers a consolidated platform that merges landscape ecology approaches with environmental psychology perspectives. Compared with conventional “landscape visual quality” models, LVA devotes greater attention to the intrinsic attributes of the landscape itself and how they translate into visual significance, thereby enhancing the practical utility of the results for planning and decision-making.

4.2. Implications for National Park Landscape Planning and Management

By emphasizing the reciprocal relationship between the natural features of a landscape and human perception, LVA furnishes practical insights for landscape planning and conservation. It pinpoints areas with high visual and experiential value, guiding resource allocation and management priorities. This is aligned with the broader aims of promoting sustainable development and enhancing visitor experiences in both natural and cultural landscapes.

(1)

Identifying Priority Areas for Conservation and Management. By quantifying PLA and VSA indicators, this method can effectively detect the most representative yet vulnerable sites within a national park. Regions with high LVA scores are often visually prominent and hold substantial aesthetic value, but they may also be at higher risk from ecological pressures and human disturbance. Targeted zoning measures, visitor capacity controls, and well-designed recreational infrastructure can be implemented in these “landscape hotspots” to alleviate pressures on fragile ecosystems.

(2)

The LVA approach enables planners to gain deeper insights into potential visual focal points. When a national park is slated to build or upgrade recreational facilities, transportation routes, or visitor centers, management must not only consider land use and ecological impacts but also evaluate how such developments might affect critical scenic corridors or key landscape nodes. LVA provides a technical basis for identifying core viewsheds and distinctive aesthetic features, informing early-stage assessments and guiding architectural siting, overall design, and harmony with the surrounding environment. In certain areas with high visual affordance, LVA results can even justify rejecting or strictly limiting construction. Different combinations of PLA and VSA not only reflect the landscape’s inherent aesthetic potential and degree of visibility but also shape visitors’ behavior and psychological experiences. By comparing PLA and VSA assessments across diverse zones and taking into account common visitor activities in national parks, we can derive the following major insights and recommendations (

Table 11. The application of LVA in national park management and planning.

Category	Landscape Characteristics	Potential Visitor Behaviors/Psychological Experiences	Planning and Management Strategies
High PLA and High VSA	- Diverse terrain and land cover, offering rich scenic elements and strong aesthetic value. - Easily accessible, with wide visibility corridors.	- Frequent photography and social media sharing (owing to high aesthetic appeal). - Longer dwell times; visitors may immerse themselves in the surroundings. - Greater potential for crowding and resource impact due to popularity.	- Provide well-designed viewing platforms or designated viewpoints to accommodate demand. - Implement visitor capacity controls or zoning to prevent overuse and ecological damage. - Enhance interpretive signage or guided tours to enrich visitor experiences while ensuring resource protection.
High PLA and Low VSA	- High scenic or ecological value but limited visibility or accessibility (e.g., obstructed views, remote locations). - “Hidden gems” not on main tourist routes.	- Niche or specialized tourism (e.g., eco-tours, photography expeditions). - Strong sense of discovery and exclusivity for those who venture there. - Potential for extended, in-depth engagement by small groups.	- Improve access selectively (e.g., new trails, small-scale viewing spots) to disperse visitor pressure from popular sites. - Maintain ecological integrity by limiting large-scale development. - Emphasize guided, low-impact tours for research, education, or special interest visitors.
Low PLA and High VSA	- Highly visible or located near main roads but lacking in significant ecological or aesthetic features. - Often monotonous landscapes or heavily altered areas.	- Potential for “visual fatigue” if too many such areas dominate main routes. - May be used primarily for functional purposes (e.g., rest areas, parking).	- Consider targeted ecological restoration or aesthetic enhancement to raise landscape appeal. - Integrate service facilities (e.g., rest stops, visitor centers) in a way that does not undermine the surrounding visual environment. - Provide basic interpretive information to encourage some level of visitor appreciation, if feasible.
Low PLA and Low VSA	- Minimal aesthetic/ecological value, often peripheral or transitional zones. - Generally not visible or appealing to the average visitor. - May serve as ecological buffers or undeveloped areas.	- Little to no direct visitation; mostly bypassed by mainstream tourists. - Potential for scientific monitoring or habitat conservation. - Few immediate psychological impacts on casual visitors.	- Maintain natural processes and restrict development, allowing these areas to function as buffers. - If ecological or cultural resources exist, consider selective enhancement or interpretation. - Otherwise, keep disturbances minimal to preserve habitat connectivity and protect biodiversity.

).

(3)

Facilitating Multi-Objective Management and Stakeholder Coordination. Owing to their multifunctional nature, national parks must balance ecological conservation, recreational development, and community interests. By emphasizing a landscape’s visual dimension, LVA underscores the importance of scenic value for public perception and cultural identity, offering a transparent, data-driven basis for participatory decision-making. For instance, when assessing whether a new visitor center or scenic corridor should be built, LVA analyses can visually illustrate potential impacts on the park’s overall scenic appeal and ecological sensitivity, thus helping to devise spatial layouts that respect both biodiversity conservation and social demand.

(4)

Balancing Visitor Perceptions with Ecological Protection. National parks are often associated with public aspirations for pristine natural environments, yet excessive visitor traffic can disrupt ecological balance. Incorporating LVA into regular monitoring and decision-making can reveal impending overloads on environmental capacities—such as trampling or landscape degradation—and trigger timely remedial measures such as dispersion or access restrictions. Through interpretive signage and visitor education, park authorities can also foster a public understanding of conservation imperatives.

4.3. Limitations and Future Directions

This study has several limitations:

• Data availability: Compared to certain European nations, comprehensive datasets on heritage sites, natural monuments, scenic viewpoints, or abandoned sites may be lacking [13,16]. This scarcity of data could impact the precision of evaluating the aesthetic functions of the physical landscape.
• Subjectivity in indicator selection: The choice of indicators stems from different conceptual frameworks and theories of visual quality, varying according to cultural norms and values [16]. As such, adopting different indicators might alter the evaluation outcomes.
• Potential discrepancies in perceived vs. potential visual quality: PLA assessments, based on geospatial data, primarily reveal the possibility or potential for a landscape to exhibit high visual quality. VSA captures the significance of a landscape’s visibility within the perception process but is not directly based on photographic interpretation. Therefore, a high-PLA area might still yield lower real-world visual quality if obstructed by elements such as unsightly power lines, discordant building colors, or excessive heights [14]. Hence, the current method is well-suited to macro-level regional planning, but local-scale planning may additionally require photo-based evaluations [4].
• Lack of dynamic and long-term monitoring: Landscape characteristics and visual elements within national parks often exhibit pronounced seasonal or climatic variability, potentially affecting LVA metrics over annual or inter-annual cycles. Future research should integrate extended temporal monitoring via remote sensing, field surveys, and visitor behavior data, allowing for the accurate assessment of LVA as landscapes evolve under environmental succession and climate change.

With the rapid advancement of remote sensing big data, AI-powered image recognition, and social media text mining, the LVA framework can be further extended to multi-context, multi-scale, and multi-audience interactive research. A more comprehensive integration of theory and practice will shift landscape planning and resource management decisions from a “thing-centered” to a “people-centered” paradigm, thereby driving deeper development of landscape studies and affordance theory in the modern era. In the future, researchers could utilize mobile devices or surveys to record visitors’ actual movement trajectories and psychological evaluations, quantifying correlations between “high patch diversity” and “number of photos taken” or “length of stay”. Such measurements would more accurately capture the potential behaviors and psychological experiences being shaped.

5. Conclusions

In this study, we define LVA as a theoretical and methodological advancement that evaluates a landscape’s ability to provide aesthetic functions while simultaneously being observable by the human eye. This framework offers a comprehensive perspective linking landscape attributes with human perception and interaction, thereby holistically assessing the capacity of landscape visual resources to serve society. On the one hand, it is related to the physical attributes and intrinsic aesthetic quality of the landscape (referred to in this study as PLA), addressing the question of whether a landscape is objectively utilizable and possesses an aesthetic foundation. On the other hand, it pertains to the human visual sensory system: for instance, some landscapes, despite their inherent beauty, cannot be seen; some areas are visible from various locations; some regions remain visible for extended periods; and some are located in the visual background with reduced clarity. We define this capacity of a landscape to be captured by the visual senses as VSA. Consequently, landscapes that are not readily visible or are embedded within a blurred background will have low VSA. By combining PLA with VSA, we are able to identify landscape areas that not only exhibit a high aesthetic quality but also possess high visibility.

In national park management practice, LVA can assist decision-makers in identifying “high-LVA areas” for prioritized conservation or zonal management, guiding the siting of recreational facilities and the design of landscape corridors, thereby playing a crucial role in balancing ecological protection and tourism development. Nevertheless, LVA still faces challenges such as data gaps, a limited capacity to capture temporal and dynamic changes, and insufficient compatibility with subjective cultural assessments. Future work could integrate multi-source data—ranging from continuous remote sensing observations to social surveys and big data mining—and employ more adaptive spatiotemporal modeling and interdisciplinary approaches for deepening and dynamically monitoring LVA. Additionally, engaging local communities, visitors, and other stakeholders in participatory decision-making and educational outreach will further enhance the relevance and inclusivity of LVA assessments. As the framework and methods continue to evolve, LVA holds great promise for guiding sustainable landscape planning, national park management, and the advancement of ecological civilization, thereby driving deeper integration between landscape research and affordance theory.