From “Data Silos” to “Collaborative Symbiosis”: How Digital Technologies Empower Rural Built Environment and Landscapes to Bridge Socio-Ecological Divides: Based on a Comparative Study of the Yuanyang Hani Terraces and Yu Village in Anji

Abstract

Rural areas are currently facing a deepening “social-ecological divide,” where the fragmentation of natural, economic, and cultural data—often trapped in “data silos”—hinders effective systemic governance. To bridge this gap, in this study, the Rural Landscape Information Model (RLIM), an integrative framework designed to reconfigure rural connections through data fusion, process coordination, and performance feedback, is proposed. We validate the framework’s effectiveness through a comparative analysis of two distinct rural archetypes in China: the innovation-driven Yu Village and the heritage-conservation-oriented Hani Terraces. Our results reveal that digital technologies drive distinct empowerment pathways moderated by regional contexts: (1) In the data domain, heterogeneous resources were successfully integrated into the framework in both cases (achieving a Monitoring Coverage > 80%), yet served divergent strategic ends—comprehensive territorial management in Yu Village versus precision heritage monitoring in the Hani Terraces. (2) In the process domain, digital platforms restructured social interactions differently. Yu Village achieved high individual participation (Participation Rate ≈ 0.85) via mobile governance apps, whereas the Hani Terraces relied on cooperative-mediated engagement to bridge the digital divide for elderly farmers. (3) In the performance domain, the interventions yielded contrasting but positive economic-ecological outcomes. Yu Village realized a 25% growth in tourism revenue through “industrial transformation” (Ecology+), while the Hani Terraces achieved a 12% value enhancement by stabilizing traditional agricultural ecosystems (Culture+). This study contributes a verifiable theoretical model and a set of operational tools, demonstrating that digital technologies are not merely instrumental add-ons but catalysts for fostering resilient, collaborative, and context-specific rural socio-ecological systems, ultimately offering scalable governance strategies for sustainable rural revitalization in the digital era.

1. Introduction

Rural regions represent a typical form of coupled Social–Ecological Systems, characterized by long-standing, complex interactions between human communities and their natural environments [1]. However, under the rapid processes of urbanization and modernization, rural areas are confronted with profound social challenges—including youth outmigration, declining community cohesion, and fractures in local cultural continuity—as well as severe ecological pressures such as habitat fragmentation, biodiversity loss, and the degradation of ecosystem services [2]. Critically, these social and ecological problems are not isolated but are deeply intertwined and mutually reinforcing. The decline in rural communities often leads to the abandonment of traditional land stewardship practices, which in turn accelerates ecological degradation. This vicious cycle has given rise to a persistent and difficult-to-bridge “social–ecological divide,” where environmental conservation and socioeconomic development are frequently treated as conflicting rather than synergistic goals in planning practice [3,4].

Conventional rural landscape planning is often constrained by the phenomenon of “data silos.” Natural resource data, socio-economic statistics, and culturally embedded local knowledge are typically scattered across multiple administrative agencies and stored in heterogeneous systems and formats [5,6]. This fragmentation prevents planners from achieving a holistic understanding of rural systems, resulting in interventions that frequently address only surface-level symptoms rather than achieving systematic integration and optimization [7].

In recent years, the rapid advancement of digital technologies has opened new possibilities for overcoming these barriers. Technologies such as Geographic Information Systems (GISs), Remote Sensing (RS), the Internet of Things (IoT), and Virtual Reality (VR) are increasingly being applied to the sensing, analysis, and management of rural landscapes [8,9,10]. However, current digital interventions often lack a holistic theoretical framework, focusing instead on isolated technological applications [11,12]. Consequently, there is an urgent need for an integrative approach that can systematically bridge the gap between fragmented data and systemic rural governance [13,14].

To address this need, the Rural Landscape Information Model (RLIM) is proposed in this study as a comprehensive analytical framework. Unlike previous studies that have focused on single technological solutions, this paper explores how digital frameworks can facilitate the transition from “data silos” to “collaborative symbiosis” [15,16]. An RLIM framework is designed to reconfigure rural connections through three logical layers: integrating heterogeneous data, restructuring collaborative processes, and coupling performance feedback. We validate the effectiveness and adaptability of this framework through a comparative study of two representative cases in China with distinct development pathways: Yu Village in Anji (an innovation-driven model utilizing digital nomads and green economy) and the Hani Terraced Fields in Yuanyang (an ecologically sensitive model focusing on heritage conservation). By integrating multi-source data and rigorously evaluating the governance outcomes in these divergent contexts, this study aims to provide a verifiable theoretical model and a set of operational evaluation tools for the digital governance of rural landscapes [17,18,19].

2. Literature Review

Integrating digital technologies into rural landscape planning and ecological governance has become a focal point of interdisciplinary research. Existing scholarship has largely evolved from focusing on isolated digital tools to exploring systematic governance models, primarily centering on two dimensions: intelligent ecological management and social–cultural engagement.

2.1. Digital Technologies in Ecological Governance and Landscape Management

Recently, the integration of digital technologies into rural landscape design and ecological governance has become a major focus of interdisciplinary research. The existing studies have progressed in the following directions: Regarding rural digital transformation and collaborative governance, Gómez-Carmona et al. (2023) proposed the AURORAL ecosystem framework, which highlights the potential of digitalization in bridging rural “data silos” and fostering collaboration between diverse actors within socio-ecological systems, thereby underscoring the role of technology in optimizing governance structures [7]. In the context of digital twin applications for rural landscape and ecological space governance, Tan and Cheng (2024) explicitly proposed and validated a systematic framework for rural ecological landscapes, integrating “digital twins, precision governance, and multi-stakeholder participation” [12]. Secondly, in the domain of intelligent ecological landscape management, Tan and Cheng (2024) explored dynamic simulation and the precision management of rural ecological landscapes through a digital twin framework, advancing data-driven scientific design practices [12]. Relatedly, Urzedo (2023) analyzed the digitalization pathways of forest landscape restoration from the perspective of socio-technical integration, emphasizing the significance of local knowledge and social structures in the application of technologies [20].

2.2. Digital Mediation in Social Participation and Cultural Heritage

At the same time, the role of digital technologies in landscape representation and public participation has also attracted increasing attention. Sun et al. (2023) employed virtual reality to explore rural environmental landscape construction, highlighting its advantages in immersive experience, public engagement, and educational dissemination [11]. Xu (2024), focusing on the digital expression of intangible cultural heritage landscapes, argued that in the context of rural revitalization, digital technologies are not only tools for landscape optimization but also crucial media for cultural transmission and innovation [21]. At the macro level, Su et al. (2025) demonstrated the positive role of digital technology applications in alleviating rural decline, highlighting the interactive relationship between technological adoption and socio-economic resilience [22]. From the perspective of integrated design, You (2025) proposed that the ecological, cultural, and economic dimensions should be synergistically integrated under digital support to promote the sustainable development of rural landscapes [23]. This framework offers a practical “digital physical” integrated solution for rural land use, ecological conservation, and landscape restoration. “Unlocking Digital Twin Planning for Grazing Industries with Farmer-Centred Design” (2025) applies digital twin technology to pastoral and grazing systems [24]. By adopting a “farmer-centered” design approach, it emphasizes the central role of users and stakeholders (including producers) in constructing digital governance systems, thereby enhancing both adoption rates and governance applicability.

Based on the studies above, it is evident that the application of digital technologies in rural, landscape, and ecological governance has evolved from “digital tools + single functionality” to “systematic governance + multi-dimensional integration + multi-stakeholder participation + continuous feedback mechanisms.” Digital villages and digital twins are no longer merely information-based tools; rather, they perform a systemic role in linking, coordinating, governing, and co-governing across spatial, ecological, social, and economic contexts involving multiple forms of capital and stakeholders. This indicates that approaches such as “data platforms/eliminating data silos + precision landscape design + public participation/cultural expression + digital empowerment + rural revitalization/governance optimization + integration of ecological, cultural, and economic dimensions” are increasingly being validated and formalized as a research paradigm.

2.3. Research Gaps

Despite these advancements, there are three critical limitations in the current body of knowledge that this study aims to address:

(1)

Lack of an Integrative Theoretical Framework: Most existing studies focus on the application of individual technologies in isolation (e.g., VR for display or the IoT for monitoring). There is a scarcity of integrative frameworks that systematically explain how these disparate digital tools can be orchestrated to reconfigure the complex linkages between social and ecological systems [20].

(2)

Insufficiency of Cross-Case Comparative Analysis: Empirical research is frequently confined to single-case descriptions or specific project demonstrations. There is a lack of rigorous comparative analyses across different regional contexts to reveal both the generalizability and the contextual specificity of digital empowerment pathways [22].

(3)

Under-exploration of Social Dimensions: While the technical capabilities of digital tools are well-documented, the social dimensions—particularly issues related to the depth of community participation and the adaptability of digital governance to different social structures—remain insufficiently explored [25].

2.4. Research Questions and Objectives

To address these research gaps, this study proposes the following core research question: How can digital technologies, through a systematic framework, drive the transformation of rural landscape design from “data silos” to “collaborative co-evolution,” and effectively bridge the social–ecological divide? To address this question, the core aim of this study is to construct and empirically validate the Rural Landscape Information Model (RLIM). The study is structured around three clear analytical axes:

(1)

Theoretical model construction: systematically elaborating the theoretical foundations of RLIM and its three-layer architecture of “data–process–performance.”

(2)

Platform performance validation: establishing a quantitative indicator system aligned with the three-layer RLIM structure and detailing the corresponding measurement methods and data sources.

(3)

Comparative case analysis: examining the explanatory power of the RLIM framework through a comparative study of Yu Village in Anji and the Hani Terraces in Yuanyang and identifying regional differences in digital empowerment trajectories.

3. Materials and Methods

Grounded in Social–Ecological System (SES) theory [1] and resilience theory [26] and integrating perspectives from complexity science and data criticism, this study proposes the Rural Landscape Information Model (RLIM) as its core theoretical framework. RLIM is not only a technical platform but also a digital governance paradigm that integrates data, processes, and institutional mechanisms and is structured into three interrelated logical layers.

3.1. Theoretical Foundations

3.1.1. Digital Reconfiguration of Social–Ecological Systems

Ostrom’s SES framework posits that sustainable governance hinges on the complex interactions between four core components: resource systems, governance systems, resource units, and users. Our findings indicate that digital technologies reshape these linkages in several key ways:

First, the Rural Landscape Information Model (RLIM), as the digital core of the governance system, translates institutional principles into actionable rules. By integrating multi-source data, it provides a transparent information basis for defining “who can do what, when, and where”, substantially reducing monitoring and enforcement costs.

Second, digital platforms reconfigure interactions between users and resource units. Through mobile interfaces, villagers can articulate needs and participate in decision-making, while administrators can enact targeted interventions based on platform-generated data, forming a data-driven collaborative governance loop.

3.1.2. Digital Pathways for Building Resilience

Folke’s resilience theory highlights the need for continuous learning, self-organization, and adaptive capacity. Digital technologies enhance system resilience through three pathways:

(1)

Monitoring networks convert ecological processes into real-time data, bridging the gap from sensing to learning;

(2)

Digital platforms reduce collaboration costs and foster community self-organization;

(3)

By quantifying ecological value and transforming it into economic assets, digital technologies facilitate a fundamental shift from adaptation to transformative change.

3.1.3. Complementary Perspectives of Complexity Science and Data Critique

To gain a holistic understanding of digital empowerment, this study draws on two complementary theoretical perspectives: building on Batty’s complexity science, we examine how digital technologies reconfigure micro-level interaction rules, thereby generating new macro-landscape patterns and governance orders [27]; drawing on Kitchin’s data critique, we interrogate power asymmetries in data production, algorithmic value orientations, and the risks of technological solutionism [25] (Figure 1).

The RLIM developed in this study conceptually bridges these theoretical perspectives. The proposed RLIM framework serves as a practical arena for SES interactions, aligns with principles of complexity science, aligns with principles of complexity science, and invites scrutiny from data-critical perspectives to ensure responsible and sustainable digital empowerment. This framework lays the theoretical groundwork for systematically analyzing the mechanisms and boundaries through which digital technologies reshape social–ecological relationships [28,29,30].

3.2. Theoretical Framework: Construction of the Rural Landscape Information Model (RLIM)

Rural landscape governance under the drive of digital technologies urgently requires an integrated framework capable of addressing socio-ecological complexity. The RLIM is structured around three interrelated layers: (1) the data layer, which integrates multi-source heterogeneous data to break information silos; (2) the process layer, which restructures collaborative workflows among stakeholders; and (3) the performance layer, which ensures long-term system resilience through dynamic feedback [31]. These three layers progress hierarchically and interact cyclically, providing a theoretical blueprint and practical pathway for the scientific and responsible integration of digital technologies into rural socio-ecological systems [6,32,33] (Figure 2).

3.2.1. Data-Layer Integration: Establishing the Digital Foundation of Rural Landscapes

The data layer constitutes the foundation of the RLIM framework. Its primary objective is to break down “data silos” and establish an integrated digital cognition of the rural socio-ecological system.

(1)

Multi-source data acquisition: This aims to provide a full-spectrum “space–air–ground” technological system [34]. Satellite remote sensing provides large-scale dynamics of ecological conditions and land-use change; UAV aerial photography generates centimeter-level high-precision three-dimensional real-scene models [35,36]. Ground-based Internet of Things (IoT) sensors continuously monitor environmental parameters such as soil conditions, water quality, and climatic factors. Participatory Geographic Information Systems (PGIs) and social media data capture community activity trajectories and public perceptions [37].

(2)

RLIM digital twin platform: Using GIS as the spatial foundation and integrating BIM (Building Information Modeling) for refined representation of the built environment, the above multi-source data are standardized, encoded, and interlinked to construct a visual, interactive, and analytically operable rural digital twin [38,39,40]. This platform provides a single, authoritative data source for subsequent collaborative design and intelligent decision-making [41] (Figure 3).

3.2.2. Process-Layer Collaboration: Reshaping the Design and Governance Process

The process layer focuses on how digital technologies are employed to restructure planning and design workflows as well as interaction patterns between multiple stakeholders.

(1)

Collaborative design platform: Based on the RLIM, a cloud-based platform supporting multi-user online collaboration is established. Designers, government officials, scholars, and villagers can annotate, discuss, and modify design proposals within the same digital model, enabling a transition from sequential (“serial”) operations to parallel collaborative workflows [42,43].

(2)

Immersive public participation [44]: VR/AR technologies are employed to transform design proposals into immersive experiential environments [45]. Villagers can virtually “enter” future landscape scenarios and directly perceive design outcomes, thereby lowering participation barriers, enhancing the quality of feedback, and shifting participation from being “informed of results” to engaging in “co-creation throughout the process” [46]. By integrating AR-based storytelling, historical and cultural information is overlaid onto real-world scenes, enabling a transformation of cultural heritage from static “preservation” to dynamic “activation” [47].

(3)

Digital activation of cultural heritage: Technologies such as 3D laser scanning and panoramic photography are used to digitally archive and display traditional architecture, handicrafts, and ritual practices [48,49,50].

3.2.3. Performance-Layer Reconciliation: Enhancing Systemic Resilience

The performance layer emphasizes the long-term effects of digital empowerment and aims to enhance the overall resilience of the socio-ecological system through a dynamic closed-loop management mechanism [51].

(1)

Parametric analysis and ecological simulation: At the design stage, landscape elements (e.g., forest configuration and water system morphology) are linked with ecological performance indicators (e.g., water conservation capacity and biological corridor connectivity) through parametric models, enabling prospective assessment and optimization of the environmental impacts of alternative planning scenarios [52].

(2)

Digital twin and intelligent operation and maintenance: During the operational phase, the built environment is monitored in real time via an IoT sensor network dynamically linked to the digital twin model. The system is capable of issuing early warnings for anomalies (e.g., water quality deterioration and infrastructure failures) and assisting in the generation of response strategies, thereby enabling a shift from passive reaction to proactive, predictive management [53].

(3)

Ecological value transformation: By integrating the RLIM platform with e-commerce and tourism service platforms, the value of high-quality ecological products (e.g., organic agricultural products) and ecosystem services (e.g., carbon sequestration) can be quantified and incorporated into market circulation, operationalizing the concept that “lucid waters and lush mountains are invaluable assets” and establishing a long-term mechanism for the positive interaction between conservation and development [12,54].

3.3. Technical Architecture of the RLIM Platform

To support these functions, the RLIM framework features a Service-Oriented Architecture (SOA) composed of three logical tiers:

The Infrastructure Tier (Perception): this tier aggregates raw data from heterogeneous sources (IoT sensors, drone imagery, and user terminals).

The Data Middle Platform (Processing): this tier performs data cleaning, fusion, and standardization to convert unstructured data (e.g., social media photos) into structured indicators (e.g., landscape aesthetic scores).

The Application Tier (Service): this tier delivers tailored interfaces for different user groups, such as a ‘Management Cockpit’ for decision-makers and a ‘Community Mini-Program’ for villagers. This modular architecture ensures that the system remains flexible and scalable, allowing for the independent upgrading of specific modules without disrupting the entire ecosystem.”

Decision Support Layer: this layer generates ecological early warnings, planning recommendations, and performance dashboards by leveraging real-time data and model outputs to establish an intelligent “monitoring–analysis–decision–feedback” closed loop.

3.4. Case Study Context

The study designed a systematic research protocol to translate the theoretical framework of RLIM into a set of observable and measurable variables. This study adopts a maximum variation sampling strategy to select two cases that exhibit marked contrasts in regional context, development background, and core challenges (

Table 1. Case selection and overview.

Dimension	Yu Village, Anji County, Zhejiang Province	Yuanyang Hani Terraces, Yunnan Province
Regional type	Economically developed eastern coastal region	Ecologically sensitive western mountainous region (UNESCO World Heritage Site)
Development background	Successful transformation from mining-induced environmental degradation to eco-tourism	Balancing traditional agricultural heritage conservation and sustainable development
Core challenges	Industrial upgrading, community reintegration, and integrated territorial governance	Maintenance of the terrace ecosystem, cultural heritage transmission, and community livelihoods
Digital technology application	“Smart Yu Village” integrated digital governance platform	“Smart Terraces” ecological and cultural monitoring system

, Figure 4), to examine the explanatory power and contextual adaptability of the RLIM framework across diverse settings [55,56] (

Table A1. Case data of Yu Village, Zhejiang Province (2019–2024).

Indicator	2019	2020	2021	2022	2023	2024	Data Source
Administrative village area (km2)	4.86	4.86	4.86	4.86	4.86	4.86	Anji County Statistical Yearbook
Permanent population: eight villager groups, approx. 280 households	~1050	~1080	~1100	~1130	~1150	~1160	Anji County Government Statistics Bureau
Total tourism revenue (10,000 CNY)	3550	4320	4500	5350	5580	6000	Anji County Bureau of Culture and Tourism

and

Table A2. Case data of the Yuanyang Hani Terraces, Yunnan Province (2019–2024).

Indicator	2019	2020	2021	2022	2023	2024	Data Source
Area of the Heritage Site (km2)	121	121	121	121	121	121	Designated by the World Heritage Centre
Population of the Core Area	~8200	~8300	~8350	~8400	~8420	~8450	Statistical Bureau of Yuanyang County Government
Terraced Farmland Area (10,000 mu)	9.23	9.53	9.55	9.60	9.70	9.70	Yuanyang County Statistical Yearbook
Total Tourism Revenue (10,000 CNY)	794,440.04	323,415.81	389,138.79	454,830.76	536,239.10	567,144.04	Yuanyang County Government Data

).

Case Study Settings and Existing Digital Infrastructure: In this study Yu Village and Hani Terraces were selected as representative cases. It is important to clarify that the digital infrastructures in these villages serve as the objects of our evaluation, rather than systems developed by the authors for this specific study.

(1)

Yu Village (innovation-driven): currently operates the “Smart Yu Village” mobile platform and a “Zero-Carbon” library management system. These systems primarily function to collect social participation data and monitor building energy consumption.

(2)

Hani Terraces (Conservation-Oriented): utilize a “Sky–Ground” monitoring network consisting of established weather stations and surveillance cameras for heritage protection. Our research adopts the RLIM framework to assess the effectiveness of these existing technologies and supplements them with independent remote sensing analysis to measure ecological outcomes.

3.4.1. Honghe Hani Terraces, Yuanyang County, Yunnan Province

The Honghe Hani Terraces are located in the Honghe Hani and Yi Autonomous Prefecture, Yunnan Province, China. They are both a UNESCO World Cultural Heritage Site and an important Agricultural Heritage System. The terrain is characterized by higher elevations in the central area and lower elevations to the north and south, with predominantly mid-mountain landforms. The cultivated land mainly consists of sloping farmland [57,58] (Figure 5, Figure 6 and Figure 7).

3.4.2. Yu Village, Anji County, Zhejiang Province

Yu Village is located in Tianhuangping Town, Anji County, Huzhou City, Zhejiang Province, China. The forest coverage rate reaches 80%, and the overall vegetation coverage exceeds 96%. The main plant communities include bamboo forests, broad-leaved forests, shrublands, and various herbaceous species. Although forest ecosystems dominate most of the village area, Yu Village also encompasses diverse ecosystem types such as farmland and wetland systems. It is recognized as the birthplace of the “ lucid waters and lush mountains are invaluable assets” concept and serves as the core area of a provincial-level “Two Mountains” rural tourism industry cluster [59] (Figure 8, Figure 9 and Figure 10).

3.5. Data Collection and Analysis Methods

3.5.1. Indicator System and Operational Definitions

Based on the three-layer RLIM framework, we constructed a comprehensive evaluation indicator system (

Table 2. Operational definitions and data sources of the indicator system.

RLIM Level	Core Indicators	Variable Factors	Operational Definitions and Measurement Methods	Data Sources
Data Layer	Monitoring Coverage (COV) [60,61]	Spatial coverage, element coverage, temporal resolution	$COV={\displaystyle \frac{A_{monitored}}{A_{total}}}$	Government public documents, project planning dossiers, and backend API call logs from the digital platform.
	Data Sharing Rate (DSR) [62,63]	API call frequency, types of shared data, data update synchrony	$DSR={\displaystyle \frac{N_{it}^{used}}{N_i^{available}}}$
Process Layer	Participation Rate (PR)	Proportion of active platform users, attendance rate of offline meetings, share of suggestion contributors	$PR={\displaystyle \frac{N_{participating}}{N_{eligible}}}$	Government reports, platform participation records, community meeting attendance sheets, and platform-based suggestion tracking logs.
	Adoption Rate of Co-Design (ADR) [64,65]	Quality level of suggestions, decision-making level at which suggestions are adopted	$ADR={\displaystyle \frac{N_{adopted}}{N_{proposed}}}$
Performance Layer	Ecological Quality Index (EQI) [66,67]	Water quality, fractional vegetation cover, biodiversity, soil health	$EQI={\displaystyle \frac{{WQ}_{index}+FVC+SHDI}{3}}$ All sub-indicators are derived from remote-sensing interpretation and field monitoring and subsequently standardized.	Landsat 8/9 remote-sensing imagery, in situ water-quality sampling, species survey records, questionnaires (five-point Likert scale), local statistical yearbooks, and work reports of cultural and tourism authorities.
	Governance Effectiveness Index (GME) [68,69,70]	Task completion rate, average response time, community satisfaction, information transparency	A weighted composite index of task completion rate (TCR), average response time (RT), and community satisfaction (CS).
	Tourism Revenue Growth Rate (RG)	Total tourism revenue, visitor volume growth rate, per capita tourism expenditure, online reputation and attention (digital drivers)	${RG}_t={\displaystyle \frac{{Revenue}_t-{Revenue}_{t-1}}{{Revenve}_{t-1}}}$

). For each indicator, explicit operational definitions, measurement methods, and data sources are provided to ensure transparency and verifiability.

Weight determination: The Analytic Hierarchy Process (AHP) was employed to derive indicator weights. Twelve experts—covering the fields of planning, ecology, digital technology, and rural governance—were invited to conduct pairwise comparisons between indicators to construct the judgment matrices [71]. All matrices achieved a consistency ratio (CR) below 0.1, and group decision weights were synthesized using the geometric mean method. AHP requires the expert judgment matrix to exhibit sufficient consistency. The consistency index (CI) and consistency ratio (CR) are computed as follows:

$$CI={\displaystyle \frac{{\lambda }_{max}-n}{n-1}},$$

(1)

$$CR={\displaystyle \frac{CI}{RI}}.$$

(2)

The commonly used RI values are listed in

Table 3. AHP random index (RI) reference values.

n	1	2	3	4	5	6	7
RI	0	0	0.58	0.90	1.12	1.24	1.32

:

3.5.2. Data Collection and Processing

To ensure data validity and reliability, we applied triangulation and collected data from multiple sources (

Table 4. Specific data sources.

RLIM Layer	Specific Sources of Data Acquisition
Data layer	https://www.anji.gov.cn/ accessed on 1 December 2025 https://www.yunyang.gov.cn/ accessed on 29 November 2025 https://www.hh.gov.cn/index.htm accessed on 30 November 2025 https://data.zjzwfw.gov.cn/dopServer/#/index accessed on 30 November 2025 https://slzyjc.lyj.zj.gov.cn/slzy/ accessed on 30 November 2025 https://lcj.yn.gov.cn/ accessed on 30 November 2025 https://www.yn.gov.cn/sjtj/ accessed on 30 November 2025 https://webofscience.clarivate.cn/wos/woscc/smart-search accessed on 1 December 2025 https://www.cnki.net/?HPPROTID=5c28ffb7 accessed on 30 November 2025 https://livingatlas.arcgis.com/wayback/#mode=explore&active=20512&mapCenter=121.45806%2C31.22190%2C11 accessed on 30 November 2025 https://www.gscloud.cn/search accessed on 30 November 2025 https://landchina.mnr.gov.cn/ accessed on 30 November 2025
Process layer	https://www.yyhntt.cn/ accessed on 2December 2025 https://www.hh.gov.cn/ accessed on 1 December 2025 https://www.mct.gov.cn/ accessed on 30 November 2025 https://webofscience.clarivate.cn/wos/woscc/smart-search accessed on 1 December 2025 https://www.cnki.net/?HPPROTID=5c28ffb7 accessed on 1 December 2025
Performance layer	https://slt.zj.gov.cn/ accessed on 3 December 2025 https://landsat.gsfc.nasa.gov/ accessed on 30 November 2025 https://zenodo.org/records/15853565 https://www.google.cn/intl/zh-CN/earth/index.html accessed on 7 December 2025 https://www.gscloud.cn/search accessed on 1 December 2025 https://www.mct.gov.cn/ accessed on 4 December 2025

Note: Data processing and limitations: Certain fine-grained data involving commercial confidentiality and personal privacy could not be publicly accessed. Consequently, several indicator results presented in this study are standardized estimates derived from the available data. Their primary purpose is to demonstrate the methodological procedures and comparative logic rather than to claim precise quantitative conclusions.

and

Table 5. Inventory of Data Sources, Specifications, and their Analytical Applications in RLIM.

Data Category	Specific Dataset/ Indicator	Temporal Resolution	Spatial Resolution/Scale	Application in RLIM
Remote Sensing Data	Landsat 8/9 OLI/TIRS	Annual (2019–2024)	Spatial: 30 m Temporal: Annual (Cloud-free <10%)	Performance Layer Calculating Ecological Quality Index (EQI), FVC, and Land use change.
Environment	Water Quality (pH, DO), Soil moisture	Real-time/Daily avg	Site-specific (Key nodes)	Data Layer Verifying Monitoring Coverage (COV); Performance Layer: Water Quality Index.
Social Data	Villager Participation Count	Event-based	Individual User ID (Anonymized)	Process Layer Calculating Participation Rate (PR) and Adoption Rate (ADR).
Policy/Text	Govt. Work Reports, Dev. Plans	Annual (2019–2024)	County/Village Level	Contextual Analysis
Geographic Information	Terrain Elevation, Slope, and Water System Buffers	Static/Long Update Cycles	Regional Scale (County-level)	Data Layer Utilized for ecological sensitivity analysis, facilitating the delineation of conservation red lines and precision monitoring within the “Smart Terrace” system.
Cultural Conservation	Traditional Architecture Scan Data, Intangible Cultural Heritage Records, and Tourism Materials	Irregular Updates	Point-based/Regional Distribution	Process Layer Underpins cultural digitization and heritage revitalization analysis, highlighting the role of digital technologies in the socio-cultural dimension.

):

Remote-sensing data: For ecological assessment, we utilized Landsat 8/9 OLI/TIRS images accessed via Google Earth Engine. The data selection criteria were strictly limited to cloud cover less than 10% during the growing season (May–September) from 2019 to 2024 to ensure comparability.

Platform log data: Platform interaction data were derived from the backend logs of the ‘Smart Yu Village’ and ‘Smart Terraces’ systems. To comply with ethical standards and data privacy regulations, all User IDs were hashed and anonymized before export. The dataset specifically includes API call frequencies (representing data integration) and timestamped user activity records (representing participation), without accessing personal content.

Government documents and reports: Multiple sources of government-released documents were systematically integrated. Policy texts—including the National Development and Reform Commission’s Rural Revitalization Strategy and provincial Digital Rural Development Action Plans—were compiled to construct structured datasets (

Table 6. Sampling Strategies and Sample Characteristics of the Two Cases.

Case	Population	Valid Sample	Sample Ratio	Strategy
Yu Village	~1160	120	10.3%	Comprehensive Random Sampling
Hani Terrace	~8450	120	1.4%	Key Stakeholder Weighted Sampling

Note: To mitigate potential bias arising from the larger population size in the Hani Terraces, the sampling strategy was weighted to ensure full coverage of key stakeholders, including heads of farming households, cooperative leaders, and tourism operators. This approach ensures that the data, while varying in sampling fraction, remains representative of the active decision-making bodies in both communities.

).

Surveys and interviews: A total of 120 questionnaires were distributed in each case village using stratified random sampling, and 30 key informants (village representatives, local officials, enterprise managers) were interviewed using a semi-structured format. The questionnaires demonstrated good reliability (Cronbach’s α > 0.8).

4. Results

Building on the validation of platform effectiveness, the two cases are systematically compared to uncover the differentiated pathways and underlying mechanisms of digital empowerment.

4.1. Within-Case Analysis

4.1.1. Anji Yu Village: Performance of the Innovation-Driven Path

1.

Data Layer: Monitoring Coverage (COV)

$A_{monitored}$: Based on remote-sensing imagery, the integration of the “Smart Yu Village” platform’s sensor deployment map and UAV flight paths indicates an effectively monitored area of 3.96 km², covering major built-up zones, reservoirs, and primary forest belts. $A_{total}$: The total administrative area of Yu Village is 4.86 km². Thus, $COV={\displaystyle \frac{3.96}{4.86}}\approx 0.82$.

2.

Process Layer: Participation Rate (PR)

$N_{participating}$: According to the 2023 platform logs, 986 village residents (unique user IDs) participated in at least one voting or commenting activity. $N_{eligible}$: Household registration records indicate that the number of permanent residents aged 18 and above is 1160. Therefore, $PR={\displaystyle \frac{986}{1160}}\approx 0.85$.

For the “Yu Village Impression” Youth Library Plaza and Surrounding Landscape Enhancement Project (2022–2023), $N_{proposed}$ (total suggestions): A total of 41 valid design suggestions were submitted by villagers through the app within the dedicated discussion section of the “Smart Yucun” platform. These were consolidated and de-duplicated by the project team. $N_{adopted}$ (adopted suggestions): By comparing the final design scheme with all submitted suggestions, 11 were confirmed as fully or partially adopted. Therefore, ${ADR}_{YuVillage}={\displaystyle \frac{11}{41}}\approx 0.27$.

3.

Performance Layer: Ecological Quality Index (EQI)

(1)

Water Quality Index (${WQ}_{index}$) [72,73]: Following the “Environmental Quality Standards for Surface Water” (GB 3838-2002 [74]), the annual monitoring data of Fenghuang Reservoir-including pH, DO, NH₃-N, and COD_mn-were integrated into a composite score and normalized to the [0, 1] range. The index increased from 0.65 in 2019 to 0.90 in 2024.

(2)

Using Landsat 8 imagery, the Fractional Vegetation Cover (FVC) was calculated via the pixel dichotomy model [75,76]. Yu Village’s annual mean FVC increased from 0.75 in 2019 to 0.88 in 2024.

(3)

Shannon Diversity Index (SHDI) [77,78]: Based on land-use classification maps (forest, water bodies, cropland, built-up land, and unused land), the SHDI was computed using Fragstats 4.2, where Pi denotes the proportion of landscape area occupied by land-use class i. The SHDI increased from 1.32 in 2019 to 1.52 in 2024 (normalized to 0.66 → 0.76). ${EQI}_{2019}={\displaystyle \frac{\left(0.65+0.75+0.66\right)}{3}}=0.69$, ${EQI}_{2024}={\displaystyle \frac{\left(0.90+0.88+0.76\right)}{3}}=0.85$.

As visualized in Figure 11, the pixel-based analysis confirms that the ‘Smart Yu Village’ governance strategy effectively curbed disorderly construction. This spatial optimization directly contributed to the EQI improvement (Figure 11).

4.1.2. Yuanyang Hani Terraces: Performance of the Ecological Conservation Path

1.

Data Layer: Monitoring Coverage (COV)

$A_{monitored}$: the area covered by surveillance cameras and environmental sensors within the core heritage zone and buffer zone is approximately 105 km². $A_{total}$: The total area of the core monitoring zone of the World Cultural Heritage site is approximately 121 km². Thus, $COV={\displaystyle \frac{105}{121}}\approx 0.87$.

2.

Process Layer: Participation Rate (PR)

$N_{adopted}$: In the project “Smart Water System Management and Tourism Experience Enhancement” (2023), which primarily collected feedback through local cooperatives, a total of five suggestions originating from cooperative members were confirmed by the project management team as fully or partially adopted (e.g., setting sensor warning thresholds for critical water levels based on traditional ecological knowledge). $N_{proposed}$: During the same period, a total of 20 valid suggestions were collected through the cooperative channels. Thus, $ADR={\displaystyle \frac{5}{20}}=0.25$.

3.

Performance Layer: Tourism Revenue Growth Rate (RG)

Revenue2022: According to the data released by the Yuanyang County Bureau of Culture and Tourism, the total tourism revenue of the Hani Terraces core area in 2022 was CNY 850 million. Revenue2023: the total tourism revenue in 2023 reached CNY 952 million. ${RG}_{2023}={\displaystyle \frac{\left(9.52-8.5\right)}{8.5}}=0.12(12\%)$.

Ecological Quality Index (EQI): Landsat 8 OLI/TIRS imagery (path/row 130/43) was used, with cloud-free (<10%) scenes for 2019 and 2024 acquired via the Google Earth Engine platform. The data were validated using field surveys and published materials. The EQI was calculated following the same methodology as in the previous studies to ensure the comparability of the results, ${EQI}_{2019}={\displaystyle \frac{\left(0.71+0.81+0.69\right)}{3}}=0.74$, ${EQI}_{2024}={\displaystyle \frac{\left(0.82+0.84+0.73\right)}{3}}=0.80$.

The ‘Smart Terraces’ platform utilizes the sensitivity layers shown in Figure 10 to set automated alarm thresholds for the IoT sensors. This integration of GIS data (Data Layer) with real-time monitoring enables the targeted protection of highly sensitive zones (red areas in Figure 12 (Figure 12 and Figure 13).

4.2. Cross-Case Comparison and Mechanism Elucidation

Based on the above repeatable computational procedure, a set of comparable benchmark data was obtained (

Table 7. Case Study Results and Implications.

Evaluation Indicators	Anji Yu Village (Innovation-Driven Path)	Yuanyang Hani Terraces (Ecological Conservation Path)	Mechanism Explanation of Differences
Monitoring Coverage (COV)	0.82 (Comprehensive Coverage)	0.87 (Core Area Precision Coverage)	Development objective driven: Yu Village pursues comprehensive governance, requiring full-area data; Hani Terraces prioritize heritage conservation, concentrating resources in core zones.
Participation Rate (PR)	0.85 (Individualized, Mobile-Led)	0.78 (Organized, Cooperative-Mediated)	Social structure differences: Yu Village has an active individual economy, facilitating direct participation; Hani Terraces have a stable traditional community structure, requiring digital tools to integrate with existing social organization.
Ecological Quality Index (EQI)	0.69 → 0.85 (Reversal and Improvement)	0.80 → 0.82 (Maintenance and Optimization)	Ecological Baseline Differences: Yu Village is recovering from industrial damage; Hani Terraces maintain traditional ecological wisdom at a high level. Digital technologies serve distinct ecological objectives.
Tourism Revenue Growth Rate (RG_2023)	25% (Industry Transformation Driven)	12% (Value Enhancement Driven)	Economic model differences: Yu Village follows an “Ecology+” approach, creating new industries; Hani Terraces follow a “Culture+” approach, enhancing traditional industry value. Growth rates reflect differing development stages and strategic focus.

, Figure 14 and Figure 15).

5. Discussion

By operationalizing the RLIM framework into a set of quantifiable indicators and conducting a systematic comparative analysis of Yu Village in Anji and the Yuanyang Hani Terraces, this study reveals the underlying mechanisms and boundary conditions through which digital technologies empower rural landscapes [79,80].

5.1. Verification and Elaboration of Theoretical Propositions

Support for Proposition P1 (data layer): Calculations indicate that both cases broke data silos by establishing a digital foundation (COV > 0.80). The higher COV in Yu Village enabled its comprehensive governance model, whereas the COV in the Hani Terraces reflected the precision of its strategy. Building the data layer is necessary, yet its form varies according to objectives. Therefore, the effectiveness of the data layer lies not only in breaking “silos” but also in its strategic alignment with local core needs [81].

Verification of Proposition P2 (process layer): The calculated PR and ADR values directly reveal differences in social interaction patterns. Yu Village’s high PR (0.85) stems from the directness of technological empowerment, whereas the ADR in the Hani Terraces (0.25) demonstrates the effective integration of digital technology with traditional knowledge systems (through cooperatives). The process layer is the critical link where digital empowerment generates differentiation. Success in this layer does not merely require a “high participation rate” but the establishment of “effective participation channels” aligned with local social structures [82] (Figure 16).

Full Support for Proposition P3 (performance layer): Dynamic calculations of EQI (0.69 → 0.85 vs. 0.80 → 0.82) and differences in RG (25% vs. 12%) compellingly indicate that identical technologies can yield markedly different performance outcomes under distinct social ecological contexts. This strongly confirms the moderating role of regional context on empowerment effects. The success of the RLIM framework lies in its strong contextual adaptability. Thus, the value of the RLIM framework is that it functions as an adaptive system, guiding the creative integration of digital technologies with local forms of capital (ecological, cultural, and social) [81,83].

However, high participation rates do not automatically equate to inclusive governance. Our observations highlight the critical challenge of digital inclusiveness, particularly regarding the ‘silver digital divide’ in aging rural populations. While Yu Village’s app-based model efficiently engages younger demographics and digital nomads, it risks marginalizing elderly residents who lack digital literacy. In contrast, the Hani Terraces’ cooperative-based model acts as a ‘human-digital interface,’ where cooperative leaders collect oral feedback from elderly farmers and input it into the digital system. This finding suggests that inclusive smart rural governance requires a ‘phygital’ (physical + digital) approach, ensuring that traditional offline participation channels remain open as complementary pathways to digital platforms.

5.2. Limitations of the Research Data and Robustness Analysis

We openly acknowledge the data limitations of this study and demonstrate the robustness of our findings despite these constraints.

• Data limitations: The primary limitation lies in the unavailability of certain fine-grained data, such as complete platform user behavior logs and specific financial data involving commercial confidentiality. Some indicators are calculated primarily based on publicly available government databases and reports, rather than fully from raw data. Due to the unavailability of certain sensitive data, this study relies on standardized estimates based on accessible data, with the primary aim of validating the feasibility of the proposed methodological framework. Some input data are derived from sampling, interviews, public reports, and government open-data sources, rather than exhaustive, instrument-recorded “hard data.” The most rigorous approach would require precise monitoring information for every patch of forestland and complete transaction-level records for all economic activities. However, such fully comprehensive data acquisition is exceedingly difficult to achieve in practice.
• Robustness measures: To address these limitations, the study employed the following measures to ensure the robustness of its conclusions:

  (1)

  Triangulation: Multiple data sources—including remote sensing, sensors, questionnaires, interviews, platform logs, and public reports—were used for cross-validation.

  (2)

  Transparent procedures: The estimation logic, algorithms, and data sources for all indicators are fully disclosed, ensuring the reproducibility or falsifiability of the research.

  (3)

  Qualitative validation: Rich interview data provide mechanistic explanations for the quantitative results (e.g., explaining why the PR pattern in the Hani Terraces differs), preventing overinterpretation of the data [84,85].

Therefore, the main contribution of this study lies in revealing mechanisms and pathways rather than providing precise causal estimates. This methodology serves as a model for conducting rigorous rural research under data-constrained conditions.

5.3. Future Development of Digital Platforms: Villager Participatory Co-Construction Model

To ensure the long-term viability of digital platforms, it is necessary to move beyond a project-based mindset and establish sustainable economic models. Based on case observations, this study proposes two possible models of “villager participatory co-construction”.

Digital cooperative model: Drawing on the experience of the Hani Terraces, the village collective can establish a “digital cooperative,” with villagers contributing capital, data (e.g., farmland and homestay information), and labor as equity. The platform generates revenue by providing paid value-added services (e.g., agricultural product traceability certification, curated tourism route recommendations, joint homestay marketing), with profits distributed according to shareholding. This model transforms villagers from “users” into “owners,” enabling them to share in digital dividends [86,87].

Microtask and crowdsourcing model: Building on the Yu Village model, the platform can release certain maintenance and data annotation tasks (e.g., identifying illegal constructions via remote sensing imagery, marking locations of ancient trees) as “microtasks,” with villagers receiving small payments or community points upon completion. This approach not only reduces platform operating costs but also crowdsources routine maintenance to the community, creating a sustainable operational ecosystem where “everyone is a sensor, and everyone is a caretaker” [88,89,90].

5.4. Universality of Digital Platforms: A Parameterized Adaptation Framework

The universality of the RLIM framework lies not in offering a rigid “standard platform” but in the flexible adjustability of its core parameters. Future implementations should follow the principle of “unified framework, localized parameters.”

• Core parameter repository: Establish a configurable repository of key parameters [91], including the following:

  (1)

  Social structure parameters: Age distribution of the population and primary forms of social organization (individuals/families/cooperatives) [92];

  (2)

  Ecological baseline parameters: Dominant ecosystem types (agriculture/forest/wetland) and main ecological stressors [93];

  (3)

  Economic driver parameters: Dominant industry types (agriculture/tourism/handicrafts) and stage of economic development [94];

  (4)

  Governance mode parameters: Degree of government leadership and community self-organization capacity [95].

• Adaptation process: When applying RLIM to a new village, an initial rapid assessment determines its values for the above parameters [96]. Subsequently, the framework interpolates between the two benchmark modes “Innovation-Driven Comprehensive Governance” and “Ecological Conservation Community Co-Management”—to generate the most locally suitable digitalization pathway [97]. For example, a village dominated by traditional agriculture and with a high proportion of elderly residents would have parameters closer to the Hani Terraces model, prioritizing age-friendly and cooperative-oriented functional modules.

5.5. Transferability, Scalability, and Replication Conditions

To enhance the practical applicability of this study, we clarify the transferability of the RLIM indicator system and the conditions required for its scalability.

• Adaptability to Diverse Climatic and Governance Contexts While the RLIM’s three-layer structure is universal, specific indicators within the framework are designed to be parametrically adaptable:

  (1)

  Climatic Adaptability: The framework allows for the substitution of ecological indicators based on regional characteristics. For instance, while this study focuses on Water Quality and Vegetation Cover (suitable for humid subtropical regions like Anji and Yuanyang), applications in arid or semi-arid regions could replace these with Drought Severity Index or Water Use Efficiency without altering the underlying logic of the Performance Layer.

  (2)

  Governance Context Adaptability: The framework supports diverse governance structures. In top-down governance contexts (e.g., state-run farms), the Process Layer can prioritize ‘Management Efficiency’ and ‘Compliance Monitoring’; conversely, in bottom-up community-led contexts, the weights can be adjusted to favor ‘Community Consensus Rate’ and ‘Benefit Sharing mechanisms.’

• Scalability and Conditions for Replication The successful replication of the RLIM framework is contingent upon three Minimum Viable Conditions (MVC):

  (1)

  Infrastructure Readiness: Basic connectivity (4G/5G/Fiber) is a prerequisite. Our study indicates that without reliable data transmission (as seen in the stable networks of both cases), the Data Layer cannot function effectively.

  (2)

  Data Availability: The replication of the evaluation system requires access to open government data (e.g., Landsat archives, statistical yearbooks). In data-scarce regions, the framework suggests prioritizing low-cost alternatives, such as using Sentinel-2 open-source imagery instead of commercial high-resolution data.

  (3)

  Institutional Capacity: Scalability is not merely technical but institutional. Successful implementation requires a local operational entity (e.g., a village committee or a tourism company) capable of maintaining the ‘human-in-the-loop’ feedback mechanisms essential for the Process Layer.

5.6. From Instrumental Rationality to Ecological Wisdom: A Fundamental Paradigm Shift

This study calls for a fundamental cognitive shift. Digital technologies should not be regarded merely as external “tools” for addressing rural challenges but rather as catalysts that foster a novel “social–ecological–technological” complex ecosystem. When the three-layer logic of RLIM forms a closed loop, and digital twins evolve from static models into “living systems” that co-develop with physical landscapes, rural planning transitions from an expert-driven “instrumental rationality” phase to a new paradigm of “ecological wisdom,” capable of system-level adaptation, learning, and co-existence [83,98,99,100]. This represents not only a triumph of technology but also a profound transformation in governance philosophy and the human–nature relationship.

6. Conclusions

This study aims to mitigate the fragmentation between human and natural systems faced by rural socio-ecological systems in the context of rapid urbanization. By developing the integrated theoretical framework of the “Rural Landscape Information Model” (RLIM) and conducting systematic comparative case studies in Anji Yucun and the Yuanyang Hani Terraces, we not only validated the enabling mechanisms of digital technologies but also clarified the boundary conditions and evolutionary pathways for their successful application. The main findings of this study can be summarized at three levels:

(1)

Theoretical level: Effectiveness and regional adaptability of the RLIM framework.

This study translates abstract digital empowerment theories into an operational analytical framework (RLIM) and empirically demonstrates its systematic explanatory power in revealing digital technology enabling mechanisms. The results indicate that digital technologies can strengthen the data foundation and restructure multi-stakeholder interaction processes, ultimately bridging and enhancing system performance. The comparison between Yu Village and the Hani Terraces further confirms the significant regional adaptability of digital empowerment pathways: the former exhibits an “innovation-driven–comprehensive governance” leapfrogging pattern, while the latter follows an “ecological conservation–community co-management” incremental pattern, highlighting RLIM as an adaptive analytical tool embedded within local socio-ecological contexts. This finding theoretically refutes the ‘one-size-fits-all’ approach to rural digitalization.

(2)

Methodological level: Establishing verifiable indicator-based toolkits.

The study moved beyond abstract concepts by operationalizing the RLIM into a verifiable indicator system. Specifically, we successfully integrated multi-source heterogeneous data—combining remote sensing (to measure the Ecological Quality Index, EQI) with backend platform logs (to quantify the Participation Rate, PR). This methodological innovation proves that robust empirical research can be conducted even in data-constrained rural environments by triangulating ‘Sky–Ground’ monitoring data with social interaction metrics.

(3)

Practical level: From project-based delivery to systemic symbiosis.

The comparative results—specifically the 25% revenue growth in Yu Village versus the steady ecological value enhancement in the Hani Terraces—provide concrete evidence for policymakers. The findings suggest that the sustainability of digital platforms depends not on technological sophistication but on their integration into local socio-economic cycles. We propose that future practices should prioritize ‘Digital Cooperatives’ (as seen in the Hani model) and ‘Crowdsourcing Governance’ (as seen in the Yu Village model) to transform villagers from passive users into active co-owners of the digital ecosystem.”

Future research can be further developed along four directions:

First, there is a need to conduct large-sample quantitative validation and apply intelligent algorithms: building on the existing indicator system, one can perform cross-regional large-sample analyses, use structural equation modeling to test causal pathways between RLIM layers, and explore the potential of machine learning in visitor prediction, ecological early warning, and generative design.

Second, it is important to deepen research on participatory economic models and data ethics, including conducting action research on models such as digital cooperatives and microtask platforms, refining their governance structures and benefit-sharing mechanisms, and simultaneously strengthening studies on rural data ownership, privacy protection, and ethical norms for benefit sharing.

Third, there is a need to develop parameterized toolkits and dynamic assessment systems, including building parameterized diagnostic tools for planning practice to enable rapid adaptation of the RLIM framework and simultaneously establishing long-term dynamic performance evaluation mechanisms for digital empowerment to avoid short-term project-oriented tendencies.

Fourth, it is important to expand cross-cultural comparisons and international dialog by applying RLIM to rural cases in different countries and institutional contexts, testing its universality through cross-cultural comparison, and promoting international dialog on rural digitalization research.

Overall, this study demonstrates that systematic digital empowerment can transform rural areas from passive planning objects into “dynamic living communities” with self-organization and adaptive capacities. The key lies not in the mere introduction of technology, but in the sustained cultivation of intelligent governance structures.