Application of a DPSIR-Based Causal Framework for Sustainable Urban Riparian Forests: Insights from Text Mining and a Case Study in Seoul

Abstract

As urbanization accelerates and climate change intensifies, the ecological integrity of urban riparian forests faces growing threats, underscoring the need for a systematic framework to guide their sustainable management. To address this gap, we developed a causal framework by applying text mining and sentence classification to 1001 abstracts from previous studies, structured within the DPSIR (Driver–Pressure–State–Impact–Response) model. The analysis identified six dominant thematic clusters—water quality, ecosystem services, basin and land use management, climate-related stressors, anthropogenic impacts, and greenhouse gas emissions—which reflect the multifaceted concerns surrounding urban riparian forest research. These themes were synthesized into a structured causal model that illustrates how urbanization, land use, and pollution contribute to ecological degradation, while also suggesting potential restoration pathways. To validate its applicability, the framework was applied to four major urban streams in Seoul, where indicator-based analysis and correlation mapping revealed meaningful linkages among urban drivers, biodiversity, air quality, and civic engagement. Ultimately, by integrating large-scale text mining with causal inference modeling, this study offers a transferable approach to support adaptive planning and evidence-based decision-making under the uncertainties posed by climate change.

1. Introduction

Historically, human societies and urban settlements have emerged along riverbanks, driven by economic opportunities, the need for protection, and access to water resources and fertile land [1]. Riverine regions have played a pivotal role in enhancing agricultural productivity, facilitating trade, and providing water for daily use, thereby shaping the emergence of human societies and urban settlements along riverbanks [2]. However, with the Industrial Revolution and subsequent urbanization, the primary functions of riverine areas shifted from agricultural and water resource management to residential and commercial expansions [3]. This transformation has led to significant ecological degradation, including soil erosion, altered river channels, and the construction of artificial structures [3,4]. Such changes have reduced habitat quality, diminished biodiversity, and posed substantial challenges to ecosystem health [5,6]. Furthermore, the increase in impervious surfaces has hindered rainwater absorption, altered runoff patterns, and exacerbated water pollution and flood risk [7]. These issues present significant barriers to sustainable urban development, particularly in the context of the ongoing climate crisis [8].

In large cities such as Seoul, riverine environments face numerous ecological and social challenges as they adapt to urban expansion [9,10]. Urban riparian forest buffers are critical in addressing these challenges by providing essential ecosystem services, including water management, pollution reduction, and wildlife habitat provision [5]. These forests filter agricultural runoff, stabilize eroding soils, and provide shade, shelter, and nourishment for aquatic life [11,12]. Additionally, they offer recreational opportunities, promote physical and mental well-being, and mitigate the urban heat island effect, contributing to the overall health of urban residents [13,14]. In the context of climate change, urban riparian forest buffers are increasingly recognized as valuable carbon sinks, supporting carbon neutrality efforts [15].

Despite their importance, rapid urbanization continues to impact the structure and functionality of urban riparian forests [16]. The alteration or destruction of these ecosystems due to riverside development undermines the ecosystem services they provide, disrupts wildlife habitats, and degrades water quality [17,18]. The expansion of riverside development has led to habitat fragmentation, biodiversity loss, and water quality degradation, significantly disrupting the ecological balance of urban riparian forests [19]. Unlike their natural counterparts, which evolve through natural processes with minimal human intervention, urban riparian forests are subject to extensive modifications due to land use changes, pollution, invasive species spread, and active management practices [6,12,16,20]. Consequently, they often follow ecological trajectories that diverge from those of natural riparian forests, leading to shifts in vegetation composition and changes in ecological processes that undermine their ability to provide essential ecosystem services [21]. Understanding these divergent ecological pathways is essential for formulating effective strategies for the sustainable management and restoration of urban riparian forests.

In response to these challenges, several studies have proposed sustainable management strategies aimed at mitigating the adverse effects of urbanization on riparian forests. Goldman et al. (2008) examined various conservation approaches, emphasizing their role in preserving protected areas and enhancing ecosystem services [22]. Likewise, Newham et al. (2011) investigated the ecosystem functions provided by riparian forests and found that during baseflow periods, a forested riparian corridor provided energy subsidies to the stream through litterfall and had a controlling influence on instream production through shading [21]. Singh et al. (2021) explored policy-driven conservation initiatives, highlighting the potential of integrated management frameworks in strengthening the ecological resilience of urban riparian forests [23].

Despite these efforts, previous studies have exhibited certain limitations. Guida-Johnson and Zuleta (2019) noted that much of the existing research has predominantly focused on ecological aspects while overlooking the socio-political dimensions of urban riparian forest management [24]. Similarly, Pollard et al. (2014) criticized the narrow scope of prior studies, emphasizing that many assessments lacked a systematic approach to analyzing interactions among ecological, social, and policy-related factors [25]. Domínguez-López and Ortega-Álvarez (2014) further pointed out the limited generalizability of case-specific studies, arguing that a more comprehensive framework is required to develop adaptable and scalable management strategies [26].

This study aims to address these gaps by developing a causal framework for urban riparian forest management, incorporating text-mining techniques to systematically extract key variables from a broad range of literature. By structuring these variables within the DPSIR (Driver–Pressure–State–Impact–Response) framework—which is particularly well-suited for analyzing complex socio-ecological systems by linking human-driven pressures to ecological states and policy responses—this research provides a holistic understanding of the factors influencing urban riparian forests and offers practical insights for policymakers and stakeholders. This approach not only integrates ecological and social complexities but also enhances the applicability of management strategies across different urban contexts, ultimately contributing to the development of sustainable urban environments.

2. Methodology

2.1. Study Areas

This study focused on four major streams in Seoul, South Korea—Jungnang Stream, Tan Stream, Yangjae Stream, and Anyang Stream—as shown in Figure 1, which illustrates the city’s ecological and geographical diversity. Initially, Bulkwang Stream, located in the northwest, was considered but excluded due to insufficient data. Watershed areas and stream lengths were obtained using GIS data, while administrative boundaries and ecological attributes were sourced from official datasets provided by the Seoul Metropolitan Government.

Jungnang Stream, with a watershed area of 2,745,576 m² and a length of 53.8 km, flows through seven districts (Gwangjin-gu, Nowon-gu, Dobong-gu, Dongdaemun-gu, Seongdong-gu, Seongbuk-gu, and Jungnang-gu) and is designated as a protected area for migratory birds and wildlife. Tan Stream, covering 1,467,245 m² and extending 35.5 km, passes through Gangnam-gu and Songpa-gu and is recognized as an ecological landscape conservation area. Yangjae Stream, with a watershed area of 666,257 m² and a length of 18.2 km, is situated in Seocho-gu and has been designated as a future heritage site of Seoul. Anyang Stream, spanning 2,375,698 m² and 36.1 km, flows through five districts (Gangseo-gu, Geumcheon-gu, Guro-gu, Yangcheon-gu, and Yeongdeungpo-gu) and is also recognized as a protected area for migratory birds. These streams were chosen based on their ecological significance, spatial characteristics, and representation of diverse administrative, environmental, and social conditions.

The selection of these streams aligns with Seoul’s “30 Million Tree Planting” initiative, a large-scale afforestation policy designed to expand urban green spaces, enhance waterfront and riparian buffer zones, and support the city’s carbon neutrality goals [9]. This initiative specifically targets waterfront and riparian areas as strategic sites for increasing ecological connectivity, improving air and water quality, and mitigating the urban heat island effect. To ensure accurate ecological observations, this study concentrated on riparian zones located beyond the levees—areas that are less affected by artificial infrastructure and more representative of natural ecological processes. These zones, characterized by native vegetation and limited human disturbance, play a central role in Seoul’s climate change mitigation efforts. They also serve as critical sites for evaluating the ecological functions of urban riparian forests and provide essential ecosystem services, including water purification, flood regulation, carbon sequestration, and wildlife habitat provision.

2.2. Methods

A multi-step approach was employed to develop and validate a causal framework for urban riparian forest management. The process began with formulating causal research questions and establishing theoretical foundations through a systematic literature review of academic journal articles and reports published between January 2010 and June 2024, following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) protocol. In the framework development phase, text-mining techniques were applied to extract key concepts and identify causal relationships among critical variables using SciBERT (Scientific Bidirectional Encoder Representations from Transformers)-based sentence classification and thematic analysis. A DPSIR-based causal model was constructed from the thematic structures derived from 1001 research abstracts and 9652 sentences.

In the data preparation and preprocessing phase, relevant indicators were selected, and data collection and cleaning were performed. Causal inference and relationship analysis were conducted using Pearson correlation and network analysis to identify key interrelations and dominant causal pathways within the DPSIR framework. Finally, the empirical validity of the framework was assessed by comparing the analytical outcomes with existing studies. The overall procedure for developing and validating the causal framework for urban riparian forest management is illustrated in Figure 2.

2.2.1. Conceptualizing and Structuring the DPSIR Framework

To conceptualize causal dynamics in urban riparian ecosystems, this study employed a systematic text mining approach grounded in the DPSIR framework. The analysis was based on a comprehensive dataset comprising academic journal articles and technical reports published between January 2014 and June 2024. The literature selection followed the PRISMA protocol, which includes identification, screening, eligibility, and inclusion phases.

In the identification phase, 3525 documents were retrieved from the Web of Science Core Collection using the keywords “Urban Riparian” (1670 results), “Urban Waterfront” (1066 results), and “Urban River Green Space” (789 results). After screening for title and abstract relevance, 2496 documents were excluded. Three members of the research team reviewed abstracts to minimize bias. During the eligibility phase, 28 duplicates or redundant documents were removed. Ultimately, 1001 articles were selected for analysis.

The selected documents underwent keyword co-occurrence network analysis and SciBERT-based sentence clustering. To improve network clarity, we applied a filtering threshold in which only keywords that appeared at least 15 times across the corpus and co-occurred with at least one other keyword that appeared at least 15 times were retained. Based on this criterion, the top 100 keywords were selected to construct the co-occurrence network, which consisted of 100 nodes and 322 edges, with a network density of 0.065. Modularity analysis using Gephi (v0.10.1) and the ForceAtlas2 layout algorithm revealed six modular clusters, yielding a modularity score of 0.349.

SciBERT-based clustering was applied to 9652 sentences extracted from abstracts, initially resulting in eight clusters. Sentences related to copyright (67), study area descriptions (1488), and methodology (1094) were excluded. The remaining 7003 sentences were categorized into five thematic clusters. Silhouette scores and sensitivity tests guided optimal cluster determination. For each cluster, ten representative keywords and twenty key sentences were selected based on frequency and cosine similarity to centroid vectors. Manual validation by the research team ensured thematic coherence and relevance.

Causal linkages were derived by mapping the identified keywords and sentence clusters to DPSIR components. The DPSIR model was selected due to its systematic structure for capturing socio-ecological feedback mechanisms, particularly suited to urban ecosystem assessments. To ensure validity, three study authors and five external experts with over 20 years of field experience reviewed the classification and linkage structure.

2.2.2. Determining Indicators

Indicators aligned with each DPSIR category were identified to operationalize the causal framework for application in Seoul. Selection was informed by the text mining results and a review of relevant municipal, environmental, and academic data sources. Emphasis was placed on indicators that reflect both ecological and socio-political dimensions of Seoul’s urban riparian zones.

The selection process considered three criteria: policy relevance, data availability, and comparability with prior research. Indicators had to align with Seoul’s environmental goals and be supported by publicly accessible and credible datasets. Cross-referencing with indicators from existing urban riparian studies ensured methodological consistency.

Final indicators included metrics related to land use intensity, impervious surface ratio, water quality parameters (e.g., Biochemical Oxygen Demand (BOD) and Total Phosphorus (T-P)), biodiversity indices, climate impact variables, and governance measures (e.g., public participation rate, budget allocation). These indicators enabled both qualitative and quantitative assessments within the DPSIR structure.

2.2.3. Data Collection and Preprocessing

A mixed-methods data collection strategy was employed. Secondary datasets were sourced from the Seoul Metropolitan Government, the Ministry of Environment, and academic repositories covering 2013–2023. These included statistics on aquatic health, endangered species, water quality, and urban forest cover. Additionally, telephone interviews with public officials (March 2024) provided contextual data on policy implementation, climate education, and disaster response.

A GIS-based spatial analysis was conducted using the Seoul Land Cover Maps (2014, 2019, 2023). Land cover classes—forest, grassland, and wetland—were extracted using ArcGIS Pro’s Spatial Analyst Tool (100 m² pixel resolution). Temperature maps were created using AWS data and Kriging interpolation to compare riparian and urban zones. Cooling effects and carbon storage were quantified using the InVEST Carbon Storage and Sequestration Model with a 1:5000 land cover map and literature-derived coefficients [27].

2.2.4. Causal Inference and Relationship Analysis

Pearson correlation analysis and network visualization were used to establish causal relationships among selected indicators. Only correlations exceeding 0.5 were retained to ensure statistical significance. These relationships were visualized using Gephi, with edge thickness denoting correlation strength and color indicating direction (red = negative, blue = positive). A causal network diagram was constructed with standardized node sizes and directional edges corresponding to DPSIR categories. This visualization enabled identification of dominant drivers, pressures, and responses in Seoul’s urban riparian systems. Comparative analysis with previous literature was conducted to validate key causal assumptions. The structural coherence of the network was further evaluated using modularity analysis with the ForceAtlas2 layout, allowing insight into thematic integration across ecological, policy, and urban factors.

3. Results

3.1. Development of a Causal Framework Based on the DPSIR Framework Through Text Mining

3.1.1. The Key Factors Extraction Under the DPSIR Framework Using Text Mining

A Text mining analysis was conducted to identify and examine relationships among key concepts on urban riparian forests. A keyword co-occurrence network was constructed to visualize term associations and categorize them into thematic clusters. As a result, six primary clusters were identified, each representing a distinct theme relevant to urban riparian forest research (Figure 3). The largest cluster, accounting for 29% of the total network, focused on water quality management, followed by ecosystem services and restoration (16%), water resources and basin management (16%), land use (14%), climate change and land impact (13%), and greenhouse gases (6%).

The water quality cluster (29%) encompassed key terms such as water pollution, stream water, water temperature, and water management. This cluster underscored the importance of managing water quality in urban riparian environments, with frequent co-occurrence of terms related to stream restoration and pollution control. The ecosystem services and restoration cluster (16%) highlighted the ecological functions provided by urban riparian forests, with key terms including vegetation cover and habitat. This cluster emphasized the significance of restoration efforts in enhancing biodiversity and maintaining ecosystem health. The water resources and basin management cluster (16%) incorporated terms such as water supply and basin management, indicating the necessity of strategic planning for the long-term sustainability of water resources in urbanized riparian areas. The land use cluster (14%) addressed the impact of land use changes on riparian ecosystems, including terms such as impervious surface ratio and urbanization. This cluster reflected the adverse effects of land cover changes resulting from urban development, particularly in relation to the increase in impervious surfaces and subsequent water pollution. The climate change and land impact cluster (13%) incorporated terms such as climate change, temperature, and carbon dioxide, illustrating concerns regarding climate-induced changes in urban riparian environments, particularly in terms of temperature fluctuations, altered precipitation patterns, and their effects on water quality and ecosystem services. The air quality and greenhouse gas factors cluster (6%) featured terms like air temperature, carbon dioxide, and greenhouse gases, highlighting the interconnections between urban air pollution and the overall health of riparian ecosystems. The co-occurrence network visualization revealed that water quality and land use functioned as central concepts within the network, linking multiple clusters. The water quality cluster exhibited strong connections to terms such as water body, stream water, and pollution, while the land use cluster was closely associated with urbanization and impervious surfaces, reinforcing the negative impact of urban development on riparian environments.

Land use changes demonstrated strong associations with both the water quality and ecosystem services clusters, suggesting that urbanization and the increase in impervious surfaces directly affect riparian ecosystem conditions. Several significant relationships were identified through the co-occurrence analysis. Water quality was found to be closely linked to ecosystem services, as indicated by the frequent co-occurrence of terms related to stream restoration and ecosystem health. This finding suggests that effective water quality management is crucial for the restoration and maintenance of ecosystem services in urban riparian spaces. The land use cluster was strongly associated with impervious surface ratio and urban development, indicating the detrimental effects of urbanization on riparian environments, particularly in terms of increased runoff and pollution. Climate change related terms were connected to temperature fluctuations and water quality, demonstrating the vulnerability of urban riparian ecosystems to climate-induced stresses such as rising temperatures and changes in precipitation patterns. Additionally, the relationship between greenhouse gases suggests that urban riparian ecosystems are influenced not only by localized water management but also by broader environmental factors such as air pollution and global climate change.

3.1.2. Establishment of Causal Relationships Based on the DPSIR Framework

A summary of all thematic categories and their respective frequencies is presented in

Table 1. Thematic Clustering and Analysis of Urban Riparian Ecosystem Research Using SciBERT: Identification of Key Topics from Abstracts.

Cluster ID	Topic	Top 10 Words	Representative Sentences	Sentence Count
1	Impact of Urbanization on Biodiversity and Ecosystems	water quality (72), land use (38), ecosystem service (20), ecosystem (19), quality (17), area (17), land cover (14), climate change (13), stream (13), water (12)	Understanding the climate growth relationship patterns in floodplains is important for providing insights into the species persistence and longevity in vulnerable riverine ecosystems experiencing human induced hydrology alteration. It is notable that all hydrological changes from the LID implementation, ranging from 0.01 to 0.06 km2 across the study watershed, were modest, which suggests a potentially limited efficacy of LID practices in mixed land cover watersheds. Ecosystem classification presents a tool that allows for quantifying the social and ecological processes of ecosystems so as to shape them into relatively homogeneous management objectives, and provides the potential of offering decision support to urban planners based on urban ecological principles. The results indicate increased daytime and nighttime temperatures and increased urban flooding due to sealed soil, and river rise from regional precipitation under climate change. Based on our results, we conclude that Chinese cities are still in the process of ‘decorating’ rivers, and that the ‘Landscape Garden City’ designation promoted such ‘decorating’ projects, especially ‘linear greening’ projects and ‘public spaces along rivers’.	1283
2	Environmental Factors Affecting Water Quality and Biodiversity in Riparian Zones	water quality (51), land use (49), stream (34), area (24), season (22), wet season (18), CO2 (18), condition (17), cover (16), species (15)	Multiple upstream sites were the major nonpoint sources of nutrient pollution. Nitrate supply and temperature finally decided the spatiotemporal distribution patterns of urban riparian denitrification. Abrupt transitions in a suite of abiotic and biotic variables were observed at the entrances and exits of the culverts whereas some variables showed no response to the culvert presence. The results showed that urban riparian corridors have a positive effect on the spread of alien and invasive plants as well as increasing the native plant diversity. The data suggest that perennial flow in urban streams results from the removal of riparian vegetation and deepening of channels rather than from urban runoff.	1778
3	Climate Change-Driven Challenges in Urban Waterfront Spaces	ecosystem service (18), area (16), environment (13), city (12), water (11), climate change (11), space (10), water quality (9), effect (9), water resource (8)	As many cities face similar challenges, they are looking for novel approaches in urban ecosystem governance. Current waterfront studies focus mainly on a land-based perspective, failing to include the water side. Compact urbanization is the main strategy for sustainable urban development. A severe drought in 2014–2015 led to a major water crisis and highlighted the fragility of the regional water supply system. The challenge that the port cities have to face lies in the disposal of large areas (often located on the waterfronts), in which it is necessary to establish new functions, to overcome the condition of marginal and degraded areas and become an integral space of the cities and of interaction with the element of water.	1164
4	Impact of Land Use and Green Coverage on River Ecosystem	water quality (59), land use (38), area (25), water body (14), river (13), quality (13), air temperature (11), zone (10), buffer zone (9), water (9)	Results from this study revealed urban and rural areas contribute to 79% and 21% of municipal wastewater generation, respectively. Significant differences in water quality were observed. Observations in the field suggested that the damage may be related to distance to water sources. The sedimentation rate derived from radionuclide profiles increased by a factor of five since the 1960s due to the urbanization of the town waterfront. The statistical analysis of the data clearly concluded that water quality of river Ganga at Varanasi was a function of adjacent land use.	1286
5	Human Activities and Their Impact on Stream Ecosystems	water quality (75), land use (46), ecosystem (22), quality (20), zone (19), stream (19), water body (19), stream restoration (17), ecosystem service (17), area (17)	The impact of environmental change on transmission patterns of waterborne enteric diseases is a major public health concern. It is insufficient to reflect actual soil quality based on soil fertility or heavy metal pollution assessment, respectively. Diversion of surface water to support production agriculture in arid and semi-arid regions has degraded ecosystems but also created potential habitat along and in canals specifically designed to transport water. Although the relationships between local environments and soil denitrification are well understood, relatively little is known about the indirect effects of landscape factors (e.g., catchment agriculture) on the soil denitrification of riparian zones. Source controls dominate eutrophication management, whilst biological regulation of nutrients is largely neglected, although aquatic microbial organisms have huge potential to process nutrients.	1492

Notes: The thematic clusters were identified based on the top 10 keywords and five representative sentences extracted for each cluster.

. The largest thematic category, environmental factors affecting water quality and biodiversity in riparian zones, comprised 1778 sentences (25.4%). This cluster emphasized how hydrological and seasonal variations influence riparian water quality and biodiversity. Key terms such as “water quality,” “land use,” “stream,” and “season” frequently appeared, highlighting concerns regarding urban pollution sources and their impact on aquatic ecosystems. Representative sentences discussed nonpoint source pollution, nitrate supply, and how urban riparian corridors influence native and invasive plant species distribution. The second-largest cluster, the impact of human activities on stream ecosystems, contained 1492 sentences (21.3%). This theme focused on how human-induced environmental changes affect waterborne diseases, nutrient cycling, and ecosystem stability. The top terms included “water quality,” “land use,” “ecosystem,” and “stream restoration,” underscoring the need for improved water management strategies. Representative sentences addressed eutrophication control, biological nutrient regulation, and the consequences of surface water diversion on aquatic habitats.

The impact of urbanization on biodiversity and ecosystems ranked third, encompassing 1283 sentences (18.3%). This cluster examined how land use changes, climate variability, and urban planning impact biodiversity and ecosystem services. The most frequently occurring terms—“water quality,” “land use,” “ecosystem services,” and “climate change”—reflected research concerns regarding rising temperatures, urban flooding, and the expansion of impervious surfaces. Representative sentences highlighted the implications of urbanization on species persistence, hydrological changes, and river ecosystem degradation. The impact of land use and green coverage on river ecosystems accounted for 1286 sentences (18.4%). This cluster focused on the effects of land use patterns, buffer zones, and vegetation cover on water bodies. Key terms such as “water quality,” “land use,” “water body,” and “buffer zone” frequently appeared, indicating a strong research emphasis on mitigating urban runoff and improving riparian water retention. Representative sentences discussed wastewater generation contributions from urban versus rural areas, sedimentation trends, and water quality variations in heavily urbanized river systems.

The smallest cluster, climate change driven challenges in urban waterfront spaces, contained 1164 sentences (16.6%). This theme focused on how climate change impacts urban waterfront governance, water resource management, and ecosystem resilience. Frequently occurring terms—“ecosystem services,” “climate change,” “water quality,” and “environment”—highlighted the necessity of integrating climate adaptation measures into urban planning. Representative sentences addressed water crises caused by extreme droughts, governance challenges in port cities, and the need for multifunctional urban waterfront spaces.

The thematic analysis revealed strong interconnections among key research themes, demonstrating that urban riparian ecosystem studies of ten address overlapping concerns. The impact of land use and green coverage on river ecosystems showed a significant overlap with water quality and biodiversity, reinforcing that land development and vegetation cover strongly influence hydrological conditions. Similarly, climate change challenges were closely linked to water quality management, reflecting that urban flooding, rising temperatures, and altered precipitation patterns exacerbate pollution risks in riparian zones. A notable research gap was identified in the integration of biodiversity conservation with climate resilience strategies. While the biodiversity and ecosystem conservation theme addressed species preservation and invasive species control, relatively few studies examined how climate change might affect riparian biodiversity over time. This indicates a need for future research on the role of riparian forests in providing climate refuge for native species.

3.1.3. Urban Riparian Forest Management Framework Development

Text mining techniques were applied to analyze the top 100 terms extracted from 1001 research abstracts, focusing on identifying key factors influencing urban riparian ecosystems. These terms were analyzed for their relationships and then classified into five major categories—Driving Forces, Pressures, State, Impact, and Response—based on the DPSIR framework.

Table 2. DPSIR Framework-Based Classification of Key Factors in Urban Riparian Ecosystem.

DPSIR	Key Factor	Related Terms
Driving Forces	Urbanization	city (37), development (29), urbanization (27), urban impact (18)
	Climate change	climate change (43), CO2 (33), temperature (24)
Pressures	Land use	land use (248), land cover (80), land-use (65), landscape (26), land (22), use change (22)
	Water pollution	water quality (372), effect water (26), water pollution (22)
	Air Temperature	air temperature (29), temperature (24), CH4 (18)
State	Water quality	water quality (372), water (58), river water (35), stream water (34), NO3 (18), water supply (18)
	Water body	water body (72), river (79), basin (19), flow (18), body (18)
	Soil	soil water (19), soil (19), surface water (50)
	Biodiversity	biodiversity (60), species (51), ecosystem (48), species richness (31), habitat (29), macroinvertebrate community (28)
	Vegetation cover	vegetation (32), vegetation cover (20), forest (20), landscape metric (17)
Impact	Ecosystem service	ecosystem service (75), service (27), ecosystem (48), watershed (45), stream temperature (20)
	Health	health (17), community health (15)
	Food	food (18), food web (18), nutrition (15)
Response	Management	management (26), water management (19), management practice (21), planning (20), activity (22)
	Restoration	restoration (20), stream restoration (33), habitat restoration (25)
Excluded		area (138), stream (90), quality (77), lake (75), condition (60), water (58), river basin (56), zone (54), region (52), season (47), scale (46), system (42), environment (39), city (37), year (36), water resource (35), buffer zone (35), use (34), effect (34), stream water (34), cover (33), model (32), space (30), community (29), catchment (27), concentration (27), process (27), time (27), landscape (26), pollution (26), effect water (26), level (26), period (26), agriculture (25), type (25), change (24), impact (24), water temperature (24), nitrogen phosphorus (24), wet season (23), effect land (21), reach (21), china (21), source (21), stream reach (20), community composition (20), forest (20), stream ecosystem (19), event (19), resource (19), impact land (18), stream channel (18), relationship land (17), project (17), use pattern (17), use water (17), community structure (17)

Notes: In the third column, Related terms are the top 100 terms from 1001 abstracts, classified to identify the key factors. The “Excluded” category refers to high-frequency keywords that were intentionally omitted from the DPSIR classification due to a lack of direct conceptual alignment with the framework’s components.

provided a structured overview of the complex interactions shaping urban riparian ecosystem health and management. The driving forces represent the fundamental causes of environmental changes in urban riparian ecosystems. Urbanization emerged as a dominant factor, with city expansion, infrastructure development, and land use changes leading to increased impervious surfaces and disruptions in hydrological processes. These modifications significantly affect water flow, quality, and biodiversity [2,3]. Climate change was also identified as a key driver, as rising greenhouse gas emissions contribute to shifts in temperature and precipitation patterns, impacting urban riparian zones through altered water temperatures and increased hydrological variability [14].

The pressures exerted by these driving forces introduce direct stressors on riparian ecosystems. Land use changes from urban development have resulted in the conversion of natural land cover into impervious surfaces, reducing water infiltration and increasing runoff, leading to heightened pollution levels [7]. Water pollution was identified as another significant pressure, stemming from industrial, agricultural, and domestic sources [2]. Pollutants such as excess nutrients, heavy metals, and toxins compromise water quality and disrupt aquatic habitats. Additionally, air temperature fluctuations due to climate change intensify thermal stress in urban waterways, affecting aquatic species and ecosystem stability [14].

The state of the urban riparian ecosystem reflects the conditions resulting from these pressures. Water quality serves as a primary indicator of ecosystem health, with pollution, nutrient loading, and filtration capacity determining overall conditions [3]. The integrity of water bodies, including rivers, streams, and lakes, is heavily influenced by sedimentation, flow alterations, and habitat degradation [2,3]. Biodiversity within riparian zones is also a critical measure of ecosystem health, as urbanization and pollution have contributed to the decline of native species and the spread of invasive species, leading to significant ecological imbalances [12].

The impacts of changes in the state of urban riparian ecosystems extend to both the environment and human wellbeing [13]. Ecosystem services, such as water filtration, flood regulation, and wildlife habitat provision, are increasingly compromised as urbanization and pollution degrade riparian functions. This has direct consequences for public health, with poor water quality heightening the risks of waterborne diseases and reducing the availability of safe drinking water [13]. The loss of ecosystem functions also threatens food security, as fisheries and agriculture reliant on riparian resources face declines in productivity due to environmental degradation [2].

The responses to these challenges encompass various management and restoration strategies aimed at mitigating negative effects and promoting ecosystem resilience. Management efforts focused on integrated water resource management, land use regulations, and pollution control were identified as essential for reducing pressures on riparian ecosystems [5]. Additionally, restoration initiatives, such as habitat rehabilitation, stream restoration, and biodiversity conservation, play a crucial role in improving water quality, reestablishing ecological balance, and enhancing overall riparian resilience [5,6].

The application of the DPSIR framework provides a systematic method for assessing and addressing the interdependent factors influencing urban riparian ecosystems. By identifying the key drivers, pressures, environmental states, impacts, and necessary responses, this framework enables the development of targeted strategies for riparian conservation and restoration. These findings offer valuable insights for policymakers, urban planners, and environmental managers in designing sustainable interventions that ensure the long-term health and functionality of urban riparian zones.

3.2. Application of the Causal Framework of Urban Riparian Forest Management for Seoul

3.2.1. Selection of Indicators for Urban Riparian in Seoul

This section applied the developed DPSIR-based causal framework to the case of Seoul to evaluate its applicability and to derive context-specific management strategies for urban riparian forests. Indicators were selected based on four criteria: relevance, measurability, data availability, and comparability. These indicators played a crucial role in assessing the ecological characteristics and sustainability of urban riparian forests in Seoul. A detailed list of the indicators and their classification is provided in

Table 3. Selection of Indicators for Urban Riparian Forests in Seoul.

DPSIR	Key Factor	Indicator Candidates	Selected Indicator	Reason for Selection
Driving Forces	Urbanization	Urbanization rate, Population density	Population density	Chosen due to ease of measurement
	Climate change	Greenhouse gas emissions, Temperature anomaly	Not selected	Excluded due to regional specificity, making comparison difficult
Pressures	Land use	Land use change rate, Impervious surface ratio	Impervious surface ratio	Chosen for ease of measurement
	Water pollution	BOD, COD, Total nitrogen (TN), Total phosphorus (TP)	BOD	Chosen for ease of measurement
	Air Temperature	Mean air temperature, Max/Min temperature	Mean air temperature in summer	Due to Korea’s distinct four seasons, summer temperature was chosen over the annual average temperature
State	Water quality	DO, pH, Electrical conductivity (EC), Turbidity, Fish grade	Fish grade	Chosen for ease of measurement
	Water body	Flow rate, Water temperature	Flow rate	Chosen for ease of measurement
	Soil	Organic matter content, Soil moisture	Not selected	Excluded due to high impervious surface cover in Seoul
	Biodiversity	Species diversity index, Habitat occupancy rate, Bird species count	Bird species count	Birds are top predators and a good representative indicator of biodiversity
	Vegetation cover	Greenness ratio, Tree canopy cover	Greenness ratio	Chosen for ease of measurement
Impact	Ecosystem service	Ecosystem service valuation	InVEST model (carbon storage, etc.)	Based on Seoul’s tree planting initiatives near rivers
	Health	Air quality index (AQI), Health metrics (e.g., asthma rates)	Air quality index (AQI)	Chosen for ease of measurement
	Food	Agricultural/Fishing productivity	Not selected	Excluded due to the minimal agricultural/fishing activities in Seoul’s urban riparian areas
Response	Management	Management investment rate, Policy implementation rate	Protected area, Program count	Based on Seoul’s designated protected areas and related management programs
	Restoration	Restored area, Vegetation restoration rate	Number of planting sites	Chosen based on Seoul’s 30 million tree planting initiative

Notes: The indicators were selected based on four criteria: relevance, measurability, data availability, and comparability.

.

3.2.2. Analysis of the Selected Indicators

The correlation analysis provided valuable insights into the interrelations among various ecological, environmental, and urban factors critical to the management of riparian forests in Seoul (Figure 4). A particularly significant finding was the inverse relationship between urbanization and green space availability, as evidenced by the strong negative correlations between Green Area Ratio and Population Density (0.79) and Impervious Surface Ratio and Green Area Ratio (0.77). This suggests that urbanization leads to an increase in impervious surfaces while reducing green spaces, highlighting the urgent need for sustainable urban planning to preserve ecological infrastructure. Additionally, the positive correlation between Afforestation Business and Green Area Ratio (0.63) underscores the effectiveness of afforestation efforts in increasing green space within urban areas. Further analysis revealed a notable association between water quality and biodiversity. The positive correlation between BOD (Biochemical Oxygen Demand) and Bird Species (0.55) suggests that improved water quality (indicated by lower BOD levels) correlates with higher biodiversity, particularly in urban riparian ecosystems. Similarly, the relationship between the Fish Assessment Index and Bird Species (0.48) indicates that healthier aquatic ecosystems, as reflected by fish populations, promote greater biodiversity overall. In terms of climate and air quality, the significant positive correlation between Temperature and Ambient Particulate Matter (PM10) (0.66) suggests that higher temperatures exacerbate air pollution, likely due to the urban heat island effect. Finally, the correlation between Carbon Storage and Sequestration and Green Area Ratio (0.74) emphasizes the role of urban greenery in carbon sequestration, demonstrating that increased green space in cities can enhance carbon absorption capacity.

While these findings are insightful, several relationships require careful interpretation, as correlation does not imply causation. For example, the negative correlation between Temperature and Bird Species (0.57) suggests that elevated temperatures may negatively affect avian biodiversity, but it remains unclear whether this is due to heat stress, habitat degradation, or other factors. Similarly, the positive correlation between Civic Activity Group and Carbon Storage and Sequestration (0.52) indicates that higher civic engagement might correlate with improved carbon sequestration outcomes. However, it is uncertain whether this relationship is causal or if regions with higher environmental awareness and civic participation have better green infrastructure.

One of the most prominent findings from the causal framework, as illustrated in Figure 5, is the inverse relationship between urbanization (represented by the Impervious Surface Ratio) and the availability of green space (represented by the Green Area Ratio). A strong negative correlation (0.77) suggests that as urbanization increases, the proportion of impervious surfaces also rises, which in turn reduces the amount of green space. This finding emphasizes the pressures that urbanization places on the environment, making sustainable urban planning and the preservation of green spaces crucial for mitigating the ecological effects of urban growth. The analysis shows a positive correlation (0.63) between the Afforestation Business and the Green Area Ratio, highlighting the role of afforestation projects in increasing urban greenery. This relationship underscores the effectiveness of afforestation initiatives in expanding green spaces within urban environments, which are essential for maintaining biodiversity and improving carbon sequestration.

The positive correlation between BOD and Bird Species (0.55) indicates that improved water quality, as indicated by lower BOD levels, is associated with greater biodiversity in urban riparian ecosystems. Additionally, the relationship between Fish Assessment Index and Bird Species (0.48) suggests that healthier aquatic ecosystems contribute to overall biodiversity. These findings emphasize the interconnectedness of water quality and biodiversity, suggesting that efforts to improve aquatic health will have positive effects on terrestrial species.

Another significant finding is the positive correlation (0.52) between Civic Activity Group and Carbon Sequestration. This relationship suggests that increased community participation in environmental activities, such as afforestation and sustainability initiatives, is associated with higher levels of carbon sequestration. The Civic Activity Group also shows a negative correlation (0.57) with Ambient Particulate Matter (PM10), suggesting that community engagement may play a role in improving air quality by reducing pollution levels. This highlights the importance of public involvement in addressing environmental challenges like air pollution and carbon emissions.

The causal framework illustrates the complexity of interactions between urbanization, green space management, biodiversity, community engagement, and air quality in Seoul’s riparian ecosystems. It underscores the importance of an integrated approach to urban ecological management, where afforestation efforts, community involvement, and sustainable urban planning are central to preserving the ecological health of the city. By addressing both pressures (e.g., urbanization) and responses (e.g., afforestation, civic engagement), it is possible to foster sustainable urban ecosystems that support biodiversity, mitigate climate change, and enhance public health.

3.2.3. Evaluation of the Causal Framework in Urban Riparian Forest Management

Figure 6 delineates the relationships among these factors, organized into four distinct groups that reflect the key elements of the DPSIR framework. The first group comprises indicators such as Afforestation Business, Maintenance Stream Flow, Impervious Surface Ratio, and Mean Air Temperature. This group demonstrates the tension between urbanization and the effective management of green spaces. While afforestation efforts and stream flow management positively influence the availability of green spaces, urbanization pressures, such as the increase in impervious surfaces and higher temperatures, negatively impact the environment. These findings underscore the necessity of sustainable urban planning that prioritizes ecological infrastructure.

The second group focuses on the interrelationship between water quality and biodiversity. Key indicators such as BOD, Fish Assessment Index, Bird Species, and Population Density are included in this cluster. The positive correlation between BOD and Bird Species suggests that improved water quality supports greater biodiversity, particularly in urban riparian ecosystems. Moreover, the relationship between Fish Assessment Index and Bird Species reflects the interconnectedness of aquatic health and terrestrial biodiversity, indicating that maintaining water quality is essential for supporting urban biodiversity. Population Density influences these factors, pointing to the additional pressures placed on water quality and biodiversity as urban density increases.

In the third group, indicators such as Civic Activity Group, Ambient Particulate Matter (PM10), and Temperature Difference are highlighted. This group underscores the influence of community engagement on urban air quality. The positive correlation between Civic Activity Group and PM 10 indicates that civic participation is associated with better management of air quality. Additionally, the Temperature Difference indicator reveals the significant role of urban heat island effects, which intensify air pollution in cities, further emphasizing the need for urban policies that address both air quality and community involvement.

The final group includes Carbon Sequestration, Green Area Ratio, Endangered Species, and Protected Area. This cluster emphasizes the role of green spaces in promoting carbon sequestration and preserving biodiversity. The positive correlation between Green Area Ratio and Carbon Sequestration highlights the importance of expanding urban greenery to enhance the city’s capacity for carbon storage. Furthermore, the presence of Endangered Species underlines the ecological value of these green areas, stressing the importance of preserving biodiversity through protected areas and sustainable land management.

4. Discussion

4.1. Identification of Key Themes in Urban Riparian Forest Management Through Text Mining and Network Analysis

The text mining and co-occurrence network analysis identified six predominant conceptual clusters related to urban riparian forest management, with water quality emerging as the most critical theme. This observation aligns with previous studies emphasizing the detrimental effects of urbanization on water quality through mechanisms such as increased impervious surfaces and pollutant loads [7,28]. The strong association between water quality and land use highlights how urbanization, particularly the expansion of impervious surfaces, contributes to the degradation of aquatic ecosystems by increasing surface runoff and pollutant inflow [29].

Moreover, the identified relationship between ecosystem services and restoration and water resources and basin management underscores the importance of integrated approaches that address multiple ecosystem functions [23,30]. These findings suggest that effective management strategies must reconcile water quality improvement efforts with land use adaptation policies to achieve sustainable urban riparian forest management [31]. However, while most studies acknowledge the importance of ecological functions, socioeconomic and policy-related factors that may influence these relationships remain unconsidered [32]. This gap calls for a more comprehensive approach that bridges ecological understanding with social and economic dynamics.

Furthermore, the thematic analysis using SciBERT revealed a disproportionate emphasis on environmental factors affecting water quality (25.4%), urbanization and land use impacts (39.6%), and climate change resilience (16.6%), while comparatively less attention was given to biodiversity conservation. This imbalance suggests a critical need for the development of integrated policies that incorporate biodiversity conservation efforts within broader frameworks of urban sustainability and climate resilience [33]. Future studies should address this imbalance by incorporating biodiversity metrics into existing frameworks to enhance the understanding of urban riparian forest functions and their contributions to ecological resilience.

4.2. Validation of the Causal Framework for Urban Riparian Forest Management in Seoul

The application of the causal framework to urban riparian forest management in Seoul revealed several critical insights and limitations, emphasizing the complex interplay between urbanization, climate change, and socioeconomic factors. While the DPSIR framework effectively demonstrated that urbanization and climate change are primary driving forces influencing environmental pressures and ecosystem health [34]. The analysis also exposed several areas where the framework may require modification to enhance its applicability to urban environments.

The population density variable exhibited a weak correlation with impervious surface ratios but a strong correlation with bird species richness. This finding suggests the phenomenon of green gentrification, where improvements in natural environments and urban infrastructure lead to increased property values and declining population density in certain areas [35]. Such trends indicate that urbanization processes do not always result in environmental degradation; rather, they can sometimes facilitate environmental improvements through social and economic mechanisms. This observation highlights the necessity of integrating socioeconomic factors into urban riparian management frameworks, which are currently underrepresented in the DPSIR model.

Additionally, although Seoul’s rivers generally maintain high water quality, making it difficult to detect significant effects of riparian forest management on water quality improvements, positive trends were observed in specific indicators [36]. For instance, improvements in water quality were associated with increased Fish Assessment scores and bird species counts. This suggests that even in highly managed urban environments, riparian forest creation and management can enhance ecosystem functions [26]. However, this effect may be overshadowed by Seoul’s extensive artificial water management systems, which have already established a high baseline of water quality. Therefore, it is crucial to consider how engineered interventions may obscure or interact with the natural benefits of riparian forests.

Moreover, the limited correlation between riparian ecosystems and water quality suggests that the DPSIR framework may inadequately capture the influence of socioeconomic processes on ecosystem health [37]. While the model effectively identifies key drivers and pressures, it fails to address how economic policies, governance structures, and public engagement shape urban riparian ecosystems. Integrating these factors into the framework would provide a more comprehensive understanding of urban riparian forest management.

Furthermore, the limited effectiveness of urban riparian forests in mitigating temperature increases and improving air quality likely results from their small and fragmented nature [38]. This finding emphasizes the need for long-term riparian forest expansion projects aimed at enhancing the connectivity and effectiveness of these green spaces. Additionally, the development of AI-driven models that incorporate social, economic, and environmental variables would allow researchers to predict the outcomes of policy interventions more accurately and develop more robust strategies for enhancing urban resilience [39].

4.3. Interpretation of Findings and Comparison with Previous Studies

The findings of this study largely corroborate previous research highlighting the significant role of urban riparian forests in improving water quality, enhancing biodiversity, and contributing to climate resilience [26,31,37,39]. The negative correlation between impervious surface ratios and water quality indicators, such as BOD and biodiversity, is consistent with earlier studies demonstrating that urbanization adversely affects aquatic ecosystems [7,28,29].

Unlike prior studies that primarily focused on ecological and environmental variables, this study emphasizes the importance of incorporating social and economic dimensions into urban riparian forest management [33]. One notable finding is the positive correlation observed between population density and bird species richness. While this may suggest that denser areas with more green infrastructure support higher biodiversity, we acknowledge that this pattern could be partially influenced by observation bias, such as increased monitoring activities or citizen science participation in more urbanized regions.

However, beyond statistical interpretation, recent empirical trends in Seoul support the relevance of the green gentrification hypothesis. Several riverside districts—despite benefiting from ecological and infrastructural improvements—are experiencing population decline, a trend often associated with green gentrification, where environmental enhancement contributes to changes in the social or demographic composition of neighborhoods. In this light, our interpretation of green gentrification is presented as a contextually grounded and plausible mechanism, though further research with spatially controlled models is necessary to validate this hypothesis.

In addition to these socio-ecological considerations, our study acknowledges an emerging yet underexplored dimension in urban riparian research: the increasing hydrological stochasticity associated with climate change. While our text mining results identified frequent mentions of “climate change,” they did not adequately capture its connection to extreme or episodic hydrological events, such as prolonged low-flow conditions, flash droughts, and temporary rivers. Recent environmental incidents—such as the mass fish mortality in the Oder River caused by elevated salinity levels during a period of exceptionally low discharge—highlight the urgency of addressing these dynamics in both research and practice [40]. Such events are no longer rare anomalies but are becoming recurring stressors in riverine systems across Central Europe and beyond [40,41]. Given the current lack of scientific literature on this subject, incorporating the concept of hydrological variability and stochastic disturbance into future iterations of the DPSIR framework could improve its sensitivity to climate-induced risks.

Furthermore, the findings underscore the limitations of the DPSIR framework, particularly its inability to capture nonlinear interactions and feedback mechanisms between urbanization processes and ecosystem functions [42]. Addressing this limitation through the integration of AI-based modeling techniques would significantly enhance the predictive power and practical applicability of the framework.

4.4. Research Limitations and Future Research Directions

While the DPSIR framework applied in this study provides a useful structure for identifying environmental drivers and pressures, it has limited capacity to capture the complexity of interactions between urbanization, climate change, and socio-economic processes. In particular, factors such as green gentrification, engineered water infrastructure, and urban governance dynamics remain underrepresented in the conventional DPSIR approach. Moreover, the study relied primarily on publicly available datasets collected by administrative and research institutions. Although these datasets offer consistent and city-wide coverage, they are often constrained by fixed monitoring locations, limited temporal resolution, and uneven observation intensity across spatial units. Such characteristics may influence correlation-based findings, particularly in urban areas with varying levels of data density and reporting activity.

However, the case of Seoul provides an important empirical context where these limitations become analytically valuable. For instance, despite improvements in ecological conditions along urban waterways, some riverside districts have experienced population decline—demonstrating a real-world pattern consistent with green gentrification dynamics. This suggests that integrating socio-economic dimensions is not merely theoretical but necessary for understanding contemporary urban ecological transitions.

Future research should therefore aim to develop a multidimensional modeling framework that incorporates social dynamics, economic behavior, and feedback-based mechanisms. AI-driven tools such as system dynamics modeling (SDM), agent-based modeling (ABM), and spatial causal inference approaches may provide more accurate simulations and stronger policy guidance. Furthermore, long-term data on biodiversity, land-use, and civic engagement will be essential for monitoring the effectiveness of riparian forest interventions. Ultimately, moving toward an integrated framework that unites environmental, social, and economic perspectives will better support evidence-based policymaking for resilient, equitable, and ecologically functional urban landscapes [23,25,42].

5. Conclusions

This study developed a DPSIR framework to enhance the sustainable management of urban riparian forests. By conducting a systematic literature review and applying text mining techniques to 1001 research abstracts, this study identified key variables and validated the applicability of the developed framework through an empirical analysis of factor interactions in Seoul. The major findings are as follows.

First of all, water quality emerged as the most critical variable in urban riparian forest management, significantly influenced by urbanization, land use changes, and climate variability. Notably, policy interventions aimed at improving water quality play a pivotal role in integrating ecological restoration with sustainable urban planning. Furthermore, climate change challenges were closely linked to water quality management, with urban flooding, rising temperatures, and altered precipitation patterns exacerbating pollution risks in riparian zones.

Second, this study underscored the necessity of integrating climate adaptation strategies into urban riparian forest management to enhance urban ecosystem resilience. Additionally, a significant research gap was identified regarding the integration of climate resilience strategies and biodiversity conservation, indicating the need for further investigation.

Third, the analysis of Protected Area designation and biodiversity index (Bird Species) revealed a positive trend over time, demonstrating that biodiversity indices increased following the establishment of protected areas. This finding underscores the crucial role of protected area designation and management in enhancing the conservation of urban riparian forests.

Finally, the modular grouping of factors using the DPSIR framework provided critical insights into how urbanization, afforestation, community involvement, and green infrastructure influence the health of riparian ecosystems. This approach contributes to a more comprehensive understanding of how policy interventions impact specific variables within the system.

Despite certain limitations, such as the exclusion of socioeconomic factors and the reliance on correlation-based analysis, this study provides a valuable foundation for developing more robust urban riparian forest management strategies. The proposed framework provides a structured approach to identifying and addressing the complex interactions between environmental variables, thereby supporting evidence-based policymaking. Future research should aim to refine and expand this framework by incorporating socioeconomic factors, employing AI-driven predictive modeling, and establishing long-term monitoring systems to enhance the accuracy and comprehensiveness of policy impact assessments.