Development of Evaluation Indicators and Weight Estimation for Restoration Projects in Degraded Areas of National Parks
Article information
Abstract
Background and objective
National parks are increasingly vulnerable to ecological degradation caused by excessive visitor use, development pressures from local economies, and the intensifying effects of climate change, including wildfires and landslides. These threats underscore the urgent need for active ecological restoration efforts. However, there is a lack of a standardized framework to assess the progress of restoration sites toward recovery. This study aims to establish a set of ecological indicators to evaluate the effectiveness of restoration in national parks and to determine the relative importance of these indicators using the Analytic Hierarchy Process (AHP).
Methods
To develop a comprehensive evaluation system, the study used a multi-step approach combining a literature review, expert consultations, and open-ended surveys. Through this process, 18 indicators were selected: 8 related to vegetation (e.g., invasive species coverage, tree volume), 7 to soil (e.g., electrical conductivity, soil hardness), and 3 to landscape characteristics (e.g., visual continuity). The AHP method was then applied to assign relative weights to each indicator based on expert input and pairwise comparisons.
Results
The AHP analysis revealed that the indicators in the landscape category received the highest overall weight (0.640), likely due to the smaller number of indicators in this category, which may have led to disproportionate weighting. Among the indicators in the vegetation category, the invasive species coverage and simplified tree volume were considered most significant. In the soil category, electrical conductivity (EC) and soil hardness emerged as key indicators.
Conclusion
This study presents a structured and weighted indicator framework for assessing ecological recovery in restoration sites in national parks. While the framework offers valuable insights, it also has certain limitations, such as limited discussion time and a one-directional decision-making process. Future research should adopt more iterative and participatory approaches to improve indicator selection and weighting.
Introduction
National parks are representative protected areas in South Korea and play a vital role in conserving both natural and cultural ecosystems. Within each national park, land is zoned into four distinct districts: Nature Conservation Districts, Natural Environment Districts, Village Districts, and Cultural Heritage Districts. This zoning system reflects the concept of biosphere reserves under UNESCO’s Man and the Biosphere (MAB) Programme, where Nature Conservation Districts function as core areas, Natural Environment Districts as buffer zones, and Village and Cultural Heritage Districts as transition areas. According to the International Union for Conservation of Nature (IUCN) protected area categories, South Korea’s national parks are generally classified as Category II (National Park). However, when considering the functional characteristics of each district, Nature Conservation Districts correspond to Category Ia (Strict Nature Reserve), Natural Environment Districts to Category Ib (Wilderness Area), and Village and Cultural Heritage Districts to Category VI (Protected Area with Sustainable Use of Natural Resources; Lee and Kil, 2024).
As a means of safeguarding South Korea’s key natural resources, national parks are legally designated and managed under the Natural Parks Act. They have fulfilled a range of environmental roles, including the preservation of biodiversity, the maintenance of ecosystem stability, and the enhancement of climate regulation. National parks have also played a critical role in conserving the habitats of endangered species and protecting natural areas from indiscriminate development. Consequently, they have provided stable habitats for various rare plant and animal species and have laid a foundation of academic research. Furthermore, national parks offer opportunities for the public to engage with nature and learn about ecological systems, thereby raising awareness of environmental conservation and contributing to the development of sustainable tourism models. However, growing demand for access to and use of natural and cultural resources has led to a steady increase in visitor numbers. At the same time, the frequency of natural disasters such as forest fires and landslides—exacerbated by climate change—has also increased, resulting in the gradual expansion of degraded areas (or damaged areas) within the parks.
Since a comprehensive survey of degraded areas in national parks was conducted in 1995, restoration projects have been implemented, focusing primarily on severely degraded sites (National Park Service, 2019). The degraded areas were categorized into nine types: independent degraded areas; coastal sand dunes and coastal degraded areas; valley degraded areas; non-statutory trails (side roads); degraded areas of abandoned roads; disconnected ecological axes; invasive species coverage; disconnected coastal ecological axes; and damaged island areas. Restoration efforts have been tailored to the specific characteristics of each type. These projects have been continuously refined and revised based on regular field surveys. However, during the implementation of restoration projects, the absence of a systematic post-project evaluation method underscored the necessity of establishing an approach to determine the effectiveness of the restoration efforts. Establishing a comprehensive evaluation process is crucial, as it can inform the management of future restoration projects and guide the direction of initial planning and implementation efforts.
Regarding forest restoration evaluation, the Korea Forest Service proposed five criteria for assessing the effectiveness of restoration projects: sustainability, resistance, productivity, circulation, and diversity (Korea Forest Service, 2009). However, these criteria serve only as general guidelines and are limited by the absence of specific indicators or clearly defined grading scales (Lim, 2016). Kang (2014) examined changes in forest ecosystems by utilizing soil and vegetation survey indicators to evaluate the effects of forest ecosystem restoration. Jung (2013) assessed the appropriateness of ecological restoration conducted on an abandoned road site using detailed evaluation indicators, including harmony with the surrounding landscape, vegetation stratification, and species diversity. Lee et al. (2015) developed an evaluation method by identifying 40 indicators for assessing ecological restoration in degraded environments near the DMZ in South Korea, an area of high natural resource value. Kim (2013) explored a variety of indicators to develop biotope evaluation indicators for urban ecosystem conservation. However, these studies have limitations in establishing a universal evaluation system that can be directly applied to restoration projects in degraded areas of national parks, as the applicable indicators and measurement methods vary depending on the specific evaluation target.
To systematically evaluate restoration projects in degraded areas of national parks, this study sought to examine essential evaluation indicators using the Delphi technique and to extract key indicators from among them. In particular, since the importance of these key indicators may vary in assessing restoration project performance, the study employed the Analytic Hierarchy Process (AHP) to analyze their relative importance and estimate the weights. The Delphi technique plays a crucial role in validating evaluation indicators by iteratively collecting expert opinions and reaching a consensus through the feedback process. The technique has also been widely adopted in ecological restoration research to ensure reliable outcomes by incorporating expert insights during the indicator selection process (Sung et al., 2022). AHP is a useful tool for systematically analyzing the relative importance of multiple evaluation indicators and estimating their weights. It has the advantage of structurally incorporating expert opinions into complex decision-making processes and has been widely applied in the field of ecological restoration to clarify relationships among evaluation indicators and scientifically support decision-making (Lee et al., 2015; Song et al., 2020). The objective of this study was to develop an objective and practical evaluation system for restoration projects in degraded areas of national parks. To this end, we selected evaluation indicators using the Delphi technique and analyzed their relative importance through AHP. These results are expected to provide foundational data for planning future restoration projects and establishing effective post-management strategies.
Research Methods
Derivation of Restoration Evaluation Indicators
A Delphi analysis was conducted to identify evaluation indicators for restoration projects in degraded areas of national parks. Initially, a preliminary set of indicators was compiled through a review of relevant literature, studies, and reports to identify key factors affecting the degree of restoration. Particular attention was given to soil and vegetation, which are significantly affected by underlying geological conditions. Their interaction plays a critical role in determining both the speed of restoration and the resilience of ecological recovery, and was therefore considered a core element in the evaluation process. To this end, 50 domestic and international academic papers were analyzed using “degraded area type” as the primary search keyword, resulting in the identification of 95 potential evaluation indicators.
In line with previous studies suggesting that evaluation indicators should be simplified as much as possible to address a broad range of evaluation objectives with a limited set of indicators (Lee & Yang, 2001), the number of indicators was progressively reduced through two rounds of expert group discussions. Practical applicability—such as redundancy and the feasibility of comparing the degree of restoration in degraded areas with that of a reference ecosystem (control plot)—was carefully considered throughout the process. In the first round, the initial set was narrowed down to 86 indicators, and in the second round, it was further refined to 57 indicators.
To finalize the selection of the final indicator set and corresponding evaluation indicators, an open-ended survey and expert group meetings were conducted. The survey was administered through interviews and email, targeting a total of 58 experts, including 14 forestry specialists, 39 landscape specialists, 2 geology specialists, 2 soil specialists, and 1 expert from a related field. Based on the survey results, indicators with a valid response rate of 60% or higher were initially selected. Subsequently, an expert group meeting was held to synthesize essential variables and incorporate open-ended feedback. Three rounds of in-depth discussions were conducted, involving five hands-on practitioners from the Korea National Park Service and three professors specializing in ecological restoration. As a result, 18 key indicators were finalized. These indicators were then appropriately categorized into three domains: ecological environment, ecology-based environment, and landscape-related factors.
Finally, based on the selected indicators, an expert survey was conducted to apply AHP, and a weight analysis was performed to assess the relative importance of each indicator.
AHP Analysis
In decision-making processes, it is often necessary to consider multiple criteria in order to identify the optimal alternative. This type of approach is known as multi-criteria decision-making (MCDM), with the Analytic Hierarchy Process (AHP) being one of the representative techniques. While the importance of decision-making factors can sometimes be quantified using objective data, in some cases, it is necessary to rely on subjective judgment (Hwang et al., 2011).
AHP provides a structured methodology for quantifying subjective judgments and offers the advantage of enabling the evaluation of qualitative factors that are otherwise difficult to measure, using a ratio scale (Jeong et al., 2019). In this study, AHP was employed, based on an expert survey, to estimate the relative weights of evaluation indicators for restoration projects derived through the Delphi technique.
In the survey, respondents were asked to evaluate the relative importance of each pair of indicators using a 9-point scale (1 to 9 points). For example, they were asked to determine which indicator is more important for assessing restoration performance: “vegetation” or “soil.” Pairwise comparisons were independently constructed for each indicator within a given domain, and the survey included n(n–1)/2 pairwise comparison questions per domain.
The AHP analysis process centers on the pairwise comparison matrix and its corresponding eigenvector (Hwang et al., 2011). Each element ai/aj in the pairwise comparison matrix A below represents the relative importance of indicator i compared to indicator j, and satisfies the reciprocal property aji=1/aij. The relative importance (or weight) of each indicator was estimated by computing the geometric mean of each row and normalizing the results to form an eigenvector. When evaluations were made by multiple experts, the representative value for each item in the pairwise comparison matrix was calculated using the geometric mean, in order to satisfy the reciprocal matrix condition and ensure statistical validity (Hwang et al., 2011).
Moreover, to assess the consistency of the pairwise comparison responses, the consistency index (CI) and the random index (RI) were used to estimate the consistency ratio (CR; Equations 1 and 2). The maximum eigenvalue (λmax) is approximated by averaging the values obtained by multiplying the pairwise comparison matrix by the weight vector and dividing the resulting vector by the weight vector. Here, n denotes the number of indicators within a given domain (Hwang et al., 2011). The random index (RI) is determined based on the number of indicators (n), following the criteria provided in a table by Saaty and Kearns (1985; Table 1).
Random index (RI) used for estimating the consistency index (CI) in AHP
The survey was conducted with 38 experts from related fields, including landscape architecture, civil engineering, forestry, and environmental engineering (Table 2). Survey response consistency is generally considered as reasonable when the consistency ratio (CR) is below 0.1, tolerable when below 0.2, and inadequate when 0.2 or higher, requiring a resurvey (Park & Sung, 2013; Saaty & Vargas, 2001). In this study, all expert responses had a CR below 0.2, and thus were deemed valid and included in the final analysis for weight calculation.
Status of experts participating in AHP
Results and Discussion
Selection of Evaluation Indicators
A total of 95 candidate indicators were identified through a review of existing literature (Table 3). Among them, 11 indicators were related to topography and 6 to geology. Vegetation was represented by 28 indicators, including 12 related to species, 3 addressing vegetation attributes such as crown width, and 13 categorized as “others” due to diverse interpretations. Soil indicators totaled 20, divided into 10 physical and 10 chemical indicators. Additionally, 30 indicators were associated with the growth environment, encompassing aspects such as landscape ecology and disaster safety. While most indicators were quantitative in nature, some, such as joint presence and landscape types, were qualitative.
List of candidate indicators extracted from previous studies on ecological restoration
Among the 95 indicators identified through the literature review, 86 were selected during the first round of expert discussions, taking into account factors such as duplication, validity, and importance. Subsequently, a second round of expert discussions resulted in the selection of 57 candidate indicators (Table 4). These 57 indicators were categorized as follows: 38 appropriate indicators (○), 19 potentially appropriate indicators (△), and 30 inappropriate indicators (×). The 38 appropriate and 19 potentially appropriate indicators—totaling 57—were then included in the first open-ended survey. These indicators were also selected through expert group discussions to establish criteria for indicator selection. Specifically, priority was given to the following types of indicators, while others such as index-based indicators were designated for indoor analysis: ① indicators that allow for mutual comparison between restoration target sites and their control plots; ② indicators that clearly represent the restoration status; and ③ in cases of overlapping characteristics, indicators that are feasible for on-site measurement.
Status of candidate evaluation indicators pre-selected through literature review and subsequently screened via expert discussions
The categories of candidate indicators, organized through expert group discussions, comprise a total of 57 indicators, classified as follows: 4 topography-related indicators, 17 vegetation-related indicators, 9 related to soil physical properties, 10 related to soil chemical properties, 13 landscape indicators associated with the growing environment, and 3 related to surrounding environmental factors (Table 5).
Categories of candidate indicators derived from expert group discussions
An open-ended survey was conducted with 58 experts to identify key indicators for evaluating national park restoration from among the 57 candidate indicators. Respondents were asked to select key indicators based on four criteria: (1) quantifiability—whether the indicators can be measured and monitored regularly; (2) objectivity—whether the indicator values can be derived from official statistics and surveys; (3) clarity—whether the indicator values and their trends can be clearly interpreted; and (4) representativeness—whether the indicators appropriately reflect ecological restoration in degraded areas.
Fifteen indicators ranked within the top 60% in terms of response frequency among survey participants. When the threshold was expanded to the top 50%, a total of 33 indicators were included (Table 6). Conversely, 15 indicators ranked within the bottom 60% were not selected, and this number increased to 25 when expanded to the bottom 50%. In this study, expert group discussions focused on the 15 indicators that ranked within the top 60%.
Results of Indicator Preference Based on the Open-Ended Survey
Key indicators for evaluating restoration projects in degraded areas within national parks were determined through discussions conducted by a panel of experts. During this process, certain indicators that were either difficult to interpret in the context of restoration outcomes or required simplification in their evaluation methods were revised accordingly. Furthermore, additional indicators considered essential based on survey results were incorporated into the final set.
Notably, within the category of landscape, it was pointed out that structural indicators—such as the number and size of patches, which are quantitative metrics rooted in conventional landscape ecology—are difficult to intuitively interpret in terms of their relevance to restoration performance, and that their applicability to on-site evaluation and policy implementation remains limited. In this regard, experts suggested that cognitive perceptions—such as how natural and acceptable the restored landscape appears—can be regarded as meaningful outcomes. They further noted that visual harmony with the surrounding landscape and the degree of landscape heterogeneity are more directly associated with user satisfaction than simple structural indicators. Based on these discussions, additional qualitative indicators frequently employed in practical landscape assessments and ecological restoration projects—namely, psychological satisfaction, landscape heterogeneity, and naturalness—were incorporated into the evaluation framework.
In addition, the complementary relationships among vegetation, soil, and landscape indicators were discussed by experts. For instance, it was noted that vegetation indicators—such as the ratio of native species and the species diversity index—are affected by underlying conditions like soil organic matter and moisture content. Accordingly, each indicator was categorized to allow for individual measurement while accounting for their causal relationships.
The 18 finalized evaluation indicators comprise items that can be practically measured using various methods, including on-site measurements, review of design documents, and expert visual assessments (Table 7). These indicators were organized to be applicable across the entire process of planning, design review, construction, and post-evaluation phases of future restoration projects.
List of evaluation indicators finalized based on consensus of expert group
Analytic Hierarchy Process (AHP) Results
An analysis of the relative importance of evaluation indicators applicable to national park restoration projects, conducted using the analytic hierarchy process (AHP), revealed the following weighting order among the major indicator categories: landscape (0.640) > soil (0.214) > vegetation (0.146; Table 8). These results suggest that the ultimate objective of restoration in degraded areas extends beyond ecological recovery to include visual and psychological integration. In other words, while the restoration of vegetation or soil presupposes the recovery of fundamental ecological functions, the findings indicate a broader recognition of the importance of achieving social acceptability and long-term sustainability through landscape harmony with the surrounding environment. Especially, a visually cohesive landscape is more likely to receive favorable evaluations of restoration outcomes from the public and key stakeholders, thereby functioning as a key indicator for assessing the success of restoration projects through the significant weighting assigned to the landscape category (Kaplan and Kaplan, 1989).
AHP-Derived weights and rankings of detailed indicators by category: vegetation, soil, and landscape
Among the detailed indicators within the vegetation category, the item with the highest importance was the invasive species coverage (0.222), followed by simplified tree volume (0.217), native species coverage (0.090), species diversity index (0.084), number of native species (0.081), similarity to surrounding vegetation (0.068), tree cover (0.133), and vegetation cover rate (0.015). The highest importance assigned to the invasive species coverage can be attributed to the tendency of invasive species to rapidly colonize and spread in disturbed areas, thereby impeding the introduction and long-term settlement of native vegetation (Song et al., 2020). In particular, the high weighting of this indicator likely reflects the recognition that effective invasive species management is not achievable through short-term or one-off interventions, but rather requires ongoing and systematic efforts to ensure successful ecological restoration (Pyšek and Richardson, 2010). This underscores the practical insight of field experts that early-stage management of invasive species is critical in determining the overall success or failure of restoration initiatives. The simplified tree volume also exhibited a high importance score, which is justifiable considering that the structural characteristics of vegetation at restoration sites—particularly the physical size of existing trees, such as their volume—can serve as a direct indicator of biomass recovery, as opposed to solely relying on metrics like vegetation cover rate or the number of species (Holl and Zahawi, 2014). The native species coverage, species diversity index, and number of native species also represent key biodiversity indicators for assessing the stability and sustainability of restored ecosystems, and their relative importance cannot be disregarded. Nonetheless, their relatively lower weights, compared to invasive species control or tree volume, are likely attributable to their nature of yielding measurable outcomes over the mid- to long-term rather than offering immediate guidance for on-the-ground management actions.
Among the detailed indicators within the soil category, electrical conductivity (EC, 0.198) exhibited the highest importance, followed by soil hardness (0.190), cation exchange capacity (CEC, 0.149), pH (0.147), soil moisture content (0.130), amount of erosion (0.100), and organic matter content (0.086). EC, which reflects the degree of salt accumulation in soil, is considered a critical evaluation item, as excessive salt concentrations can inhibit plant growth and lead to secondary soil pollution (Rhoades et al., 1992). The importance of soil hardness was highlighted as a major physical constraint, as elevated hardness levels can impede seed germination and root development, thereby limiting the stable establishment and spread of vegetation (Marsden and Jones, 1992). Notably, when soil hardness is high, additional restoration measures—such as backfilling and plowing—are often required, linking this indicator directly to actual project implementation costs. Meanwhile, the importance of organic matter content was found to be the lowest, which may reflect the perception that increases in organic matter tend to occur naturally over time, allowing for gradual improvement without the need for immediate intervention or artificial measures. Furthermore, since long-term monitoring is required until the soil’s organic matter content reaches a desirable level, this indicator was likely given a relatively low weight in early-stage evaluations.
The landscape category was assigned the highest importance across the entire AHP analysis. Among its detailed indicators, psychological satisfaction (0.490) had the greatest weight, followed by landscape heterogeneity (0.317) and naturalness (0.193). Although psychological satisfaction is a subjective indicator based on the perceived experience of site users or local residents, it ultimately serves as a key determinant of social acceptance and overall satisfaction with restoration projects. Therefore, this finding aligns with the recent trend emphasizing not only ecological outcomes but also human-centered approaches in restoration and evaluation (Nassauer, 1995; Gobster et al., 2007). landscape heterogeneity and naturalness were also identified as important factors. While perceived disharmony reflects the degree of visual disconnection or imbalance between restored sites and their surroundings, naturalness serves as an indicator of the extent to which artificial elements are excluded and ecological integrity is maintained. These indicators provide practical criteria for assessing restoration success from a visual and spatial perspective and are directly linked to the creation of a resilient landscape as experienced by users (Carrus et al., 2015).
According to the importance rankings of all indicators derived through AHP, the item/indicators with the highest weight was psychological satisfaction (0.490), followed by landscape heterogeneity (0.317) and naturalness (0.193), ranking second and third, respectively. These three items/indicators all fall under the landscape category, suggesting that landscape-related elements serve as key criteria that directly affect both the perceived performance of restoration projects and their overall quality. In the soil category, EC (0.198) ranked fourth overall, followed by soil hardness (0.190) in fifth place. This reflects the importance of the physical and chemical conditions that underpin vegetation growth. In the vegetation category, the invasive species coverage (0.222) did not rank within the top five overall but was the highest-ranking indicator within the vegetation domain, placing eleventh overall. This indicates that vegetation-related factors were perceived as relatively less important compared to landscape and soil elements. On the other hand, indicators such as the number of native species and similarity to surrounding vegetation ranked below 15th overall, suggesting that greater importance is placed on practical management factors—such as the difficulty of managing and removing undesirable factors—rather than on ecological diversity in restoration sites. This overall ranking analysis helps clarify not only the relative importance of each category but also the prioritization of evaluation factors that should be considered first in practical restoration sites.
In summary, the above results suggest that successful restoration sites should integrate not only ecological functions but also visual coherence and social acceptance. Accordingly, the evaluation indicator system for restoration projects should be designed in a balanced manner, incorporating psychological and landscape indicators alongside ecological and physical ones. Furthermore, the weights quantified through AHP can serve as foundational data for developing future evaluation models for restoration projects, and may also be expanded to analyze differences in indicator weighting among experts or to conduct comparative assessments involving multiple stakeholders, including general users.
In particular, the overall ranking analysis offers practical criteria for identifying the key factors that should be prioritized in restoration sites. The high ranking of landscape- related indicators reflects the recognition that social and psychological satisfaction, alongside ecological resilience, is essential to restoration performance. This implies that management strategies for restoration sites should extend beyond ecological recovery to include user-centered landscape harmony and improvements in perceived quality. Future restoration projects should establish target levels and priorities for indicators within each category based on these findings, while also developing a participatory evaluation system that incorporates the needs of diverse stakeholders. Through this approach, it is expected that a multi-dimensional evaluation of restoration performance, including qualitative aspects, can be achieved.
Conclusion
National parks are increasingly exposed to various risks of degradation due to rising visitor numbers, development pressures aimed at stimulating local economies, and natural disasters such as wildfires and landslides, which are exacerbated by climate change. The frequency and spatial extent of degraded areas are expected to continue growing as a result of these environmental and social factors. Restoration of degraded areas typically involves either artificial restoration techniques or the facilitation of natural regeneration. These efforts constitute a series of processes intended to recover the natural functions of the ecosystem. However, to assess whether restoration projects are actually achieving the intended direction and level of ecological recovery, a scientific and quantitative evaluation system is imperative.
This study aimed to select indicators for evaluating the performance of restoration projects in degraded areas within national parks and to estimate the relative importance of these indicators using the Analytic Hierarchy Process (AHP), thereby providing foundational data for developing an evaluation system for future restoration projects. From an initial pool of 95 indicators derived through a literature review, 18 were selected based on criteria such as practical applicability and clarity of evaluative interpretation, as determined through expert group discussions and open-ended surveys. The final set of selected indicators comprised eight in the vegetation category, seven in the soil category, and three in the landscape category.
According to the AHP analysis, landscape indicators received the highest weight among the indicator categories, followed by soil and vegetation. This suggests that in restoration projects of degraded areas within national parks, not only the recovery of ecological functions but also the degree of landscape integration and the psychological satisfaction of users serve as key criteria for evaluating restoration performance. In particular, landscape-related indicators—such as psychological satisfaction, landscape heterogeneity, and naturalness—ranked among the highest overall. These findings suggest that restoration efforts are evolving beyond a narrow focus on ecosystem recovery to include aspects of social acceptance and landscape quality. In the vegetation category, the invasive species coverage and the simplified tree volume received high weights, while in the soil category, electrical conductivity (EC) and soil hardness emerged as key indicators. This indicates that the removal of factors impeding vegetation establishment, the recovery of biomass, and the stabilization of soil—both chemically and physically—are critical determinants of the initial success of restoration sites. In contrast, some indicators, such as organic matter content, were deemed relevant primarily to the long-term stage of restoration, resulting in lower relative importance.
However, this study has the following limitations. First, the interpretation and discussion of each indicator during the decision-making process were not sufficiently repeated, and the consensus formed through iterative expert discussions was limited. Second, due to the structural characteristics of the AHP method, the number of indicators in the landscape category was relatively small, which may have introduced a structural bias leading to disproportionately high weights. This implies that, in addition to differences in expert perception, the imbalance in the number of indicators may have affected the results. In future research, it is necessary to enhance the validity and reliability of the weight estimation by conducting sensitivity analyses or experiments to test for structural bias, ideally under conditions where the number of indicators is more evenly distributed. Third, since the evaluation guidelines for qualitative indicators related to landscapes have not been sufficiently specified, there is a possibility that subjectivity may be introduced by evaluators during on-site application. Therefore, future research should develop methods for quantifying qualitative indicators and clarify the criteria for judgment. Fourth, the evaluation indicators established in this study were designed to be universally applicable across all nine types of degraded areas in national parks. However, in practice, the importance and applicability of each indicator may vary depending on the specific characteristics of each damage type and the stage of restoration. For instance, the primary indicators for assessment may differ between forest-damaged and wetland-degraded areas, and the evaluation criteria may also change between the initial restoration phase and the stabilization phase. Nevertheless, as this study focused on the universality and consistency of indicators, it has limitations in providing tailored indicators for each type and stage. Future research should develop a flexible evaluation framework that allows for adaptive application of indicators based on degraded area characteristics and restoration phases.
Despite these limitations, this study holds significance in that it contributes to the specification and systematization of evaluation criteria for restoration performance by employing an empirical approach to indicator selection and weight analysis for the restoration of degraded areas within national parks. Moving forward, the restoration evaluation system should be further refined by clarifying the estimation methods for each indicator, assessing the potential for conversion into monitoring indicators, and quantifying the evaluation criteria. Furthermore, in order to develop an evaluation system that incorporates the perspectives of diverse stakeholders, it is necessary to adopt methodological approaches that reflect the perceptions and values of non-expert groups (e.g., general visitors, local residents) in the processes of indicator selection and weight estimation. For example, non-expert input can be systematically gathered through participatory AHP techniques conducted alongside expert-driven AHP, public evaluation workshops, and surveys reflecting public opinion. Establishing such a multi-layered, participatory indicator system will not only enhance the validity and public acceptability of restoration projects but also increase their effectiveness in being reflected in future policy-making processes.
Ultimately, this study holds both academic and policy significance as it provides a scientific foundation for the sustainable management of national parks and the restoration of ecosystem health, while also offering a practical direction for establishing an evaluation system for restoration sites. In the future, by advancing indicator selection methodologies and systematizing weighting techniques, the evaluation system should evolve into a more refined framework that accurately reflects the estimation methods for each indicator and the specific characteristics of restoration sites within national parks. Notably, the 18 evaluation indicators proposed in this study are primarily composed of items that can be measured through field surveys or literature-based assessments. Moreover, some qualitative indicators may also be quantifiable in the future through the use of visual data collection, perception surveys, and citizen participation-based evaluations. Therefore, this indicator system is considered to have both practicality and scalability, making it applicable across the planning, monitoring, and post-evaluation stages of actual restoration projects.