Land transfer and planning in suburban villages of mountain valley area based on scenario analysis

Abstract

Influenced by terrain constraints and urban negative externality in mountain valley areas, the suburban village construction lands are not only occupied by cities, but also face hollow-out and decline. Various complex situations of construction land use should be considered in the rural land transfer and planning. In this study, scenario analysis method is used to propose a research framework of “Scenario construction–Benefit evaluation–Dynamic planning”. In the case of Huazhaizi and Ershilipu villages in Bali, Lanzhou, the construction land use demands are identified initially, and 27 possible development scenarios are designed. Next, thresholds are set to screen out 9 land transfer scenarios, and their respective social and economic benefits are evaluated. Finally, the optimal scenario is obtained based on the benefits, and dynamic planning is carried out with the application of artificial intelligence (AI) technology. The results show that the construction of Ershilipu and Huazhaizi villages should be coordinated to achieve the target scenario of Culture-oriented travel, which needs to go through three phases: Family farm renovation, Ecological tourism construction and Cultural tourism construction. From 2025 to 2030, the abandoned residential lands in rural settlements will be consolidated, and 24.89 ha of construction lands near urban areas will be transferred to construct high-rise residential buildings, vegetable processing workshops and family farms. From 2030 to 2035, 14.95 ha of construction lands on both river sides and around settlements should be transferred to build parks and scenic spots, with a view to improving the rural commercial and ecological environments. From 2035 to 2040, 9.93 ha of low-value construction lands inside settlements needs to be transferred to construct cultural tourism projects and tourism service facilities. It is estimated that by the end of planning period, the total revenue will be approximately 3.20 billion yuan. Our research results can provide decision-making reference for the transfer, development direction and use arrangement of suburban village construction lands in mountain valley areas.

Introduction

Mountain valley areas are mainly located in northwest China’s Loess Plateau, accounting for approximately 33% of China’s total territory. These areas are densely covered with ravines, which have fragmented topography and fragile ecological environment¹. Development of local suburban villages is not only limited by the valley topography, but also faces the urban negative externality. The production lands outside settlements are occupied by cities, while the hollow-out inside settlements is serious. These further lead to the shortage of available construction and production lands in rural areas, which seriously affects their sustainable development². In developed countries, the rural population in some suburban areas may increase as a result of overspill³. However, in the economically backward mountain valley areas, the construction of suburban villages is faced with a more complex scenario. On the one hand, cities are located in the narrow floodplains of the lower reaches of river valleys, where construction lands are in shortage. It is economical for the cities to develop towards the upstream valleys, forming multiple residential areas and industrial zones. As a result, suburban villages lose lands, agricultural order is disrupted, and rural settlements decline⁴. On the other hand, higher income makes most of villagers flood into the cities, who can enjoy high-quality public services while working in the cities, leading to the separation of suburban villagers from farmlands and of their residence from workplace^5,6.This further leads to the central through road congestion, backward public services and serious hollow-out⁷. During the transfer and planning of rural construction lands, it is necessary to consider the joint impact of topography and cities, thereby enhancing the use efficiency of rural lands. Rural construction land transfer refers to allowing rural collective lands to be used for other purposes after transfer, lease and investment on the premise of meeting urban planning and use regulations, so that more flexible and diversified construction demands are achievable⁸. However, compared to the villages in plain areas, the construction land transfer in valley suburban villages faces more uncertainty. Firstly, valleys have narrow landforms and agricultural lands are scarce. Rural buildings are dense, encroaching heavily on the river banks. Since roads connecting the cities are single and narrow, it is necessary to meet the multiple demands of future rural lands for agricultural production, ecological environment and commuter traffic. Next, urban–rural connections are frequent in these areas, population travel purposes are diverse and spatial characteristics are complex. Rural land transfer not only should meet the demands of multiple types of public services, but also can match the uncertain spatial changes⁹. Finally, these areas are located at the urban fringes, and the rural land transfer has higher returns, which will attract more stakeholders to participate in the project construction. The demands of multiple stakeholders will lead to the uncertainty of land transfer location, area and function¹⁰. To coordinately solve the above problems, it is not only necessary to predict various uncertain scenarios in the rural construction land transfer, but also to calculate the comprehensive income from possible scenarios. Scenario analysis can not only make a series of plausible, recognizable, challenging and internal consistently assumptions about future trends to provide various development scenarios after the land transfer, but can also make a further description of possible rural development trend in combination with other computational methods, such as statistical analysis technology and AI-based solution generation technology, under a certain scenario framework^11,12.Thus, compared to conventional quantitative analysis, inductive-deductive analysis, scenario-based analysis can comprehensively consider the uncertainty of land use changes and intuitively display the land use transfer scenarios and planning scheme, which can better solve the complex problems encountered by the valley suburban villages.

Huazhaizi and Ershilipu villages, the research objects of this study, are located in Bali Town of Lanzhou City. They are only 7 km away from Lanzhou urban area, where the urban-rural population flow is frequent. Most of the villagers go to cities for work, consumption and house purchase, leading to substantial vacant rural houses and a shortage of production lands to result in regional poverty and decline. How to guide the shrewd and efficient transfer of rural lands, create higher social and economic benefits, and form a better living environment is the core of sustainable development of valley suburban villages. The research objectives of this study are as follows: (1) To predict various scenarios regarding land use changes in the valley suburban villages through the application of scenario analysis methods. (2) To quantitative evaluate the social and economic benefits of multiple land transfer scenarios for obtaining the optimal development scenario. (3) To develop spatial optimization strategy for valley suburban villages by scenario description based AI generation technology. The contributions of this study are as follows: (1) Through collaborative scenario analysis, quantitative analysis and AI technology, eliminate the uncertainty in the construction processes of valley suburban villages, and propose optimal solutions. (2) To provide methodologies for prediction, evaluation indicators, and planning generation related to land transfer and changes in suburban villages.

Literature review

Transfer of rural construction lands stems from China’s rapid urbanization, especially in mountain valley areas, where urban land resources are becoming increasingly valuable with the increasing demands for urban industrial and residential lands¹³. The per capita construction land in suburban villages is four times that in cities. Therefore, in a free land market, the situation would resolve itself as developers purchase the land needed from the rural owners at a price that suits both parties¹⁴. Relevant research has focused on rural politics and economics. In the political field, most studies have concerned the correlation of construction land with other factors in rural regions, such as the relationship between rural construction land and urbanization rate, or the increase and decrease relationship between rural and urban construction lands^15,16. In the economic field, focus has been on the land market research of construction lands, covering land value evaluation, land market openness and land resource information management^17,18,19. However, most of the above studies form an evaluation framework or put forward policy recommendations, which cannot guide the transfer of rural lands based on spatial environment. There are three major types of methods for the spatial analysis and transfer of rural construction lands. The first type encompasses the classified comparison and induction-deduction methods, which can compare the feasibility and economical efficiency of land transfer model, promotion area and implementation process. Under the representative sustainable development framework for land consolidation proposed by Zang, classified comparison is made among factors like economic activities, ecological resources, social capital and human capital²⁰. Relevant studies are reflected in the conversion of farmland in the Hudson River Valley, the Land abandonment projections in Fukushima Prefecture and the Land consolidation potential of hollow village in China^21,22,23. However, cases in China are mostly from developed and coastal plain areas, while rural governments in mountain valley areas have strained finance and weak collective economy, which differ largely from the developed areas in policies, indicators and programs based on classified comparison. The second type of methods highlights public participation, which includes land transfer methods based on stakeholders’ appeal²⁴. Decisions are usually made by statistically summarizing the needs of governments, local villagers and migrant workers. The data comes from a wide range of sources and has been applied in the aspects of regionally integrated spatial layout and rural land acquisition^25,26. Since this type of methods can reflect the demands of multiple parties, the decisions can be better trusted and supported. They can provide a reference for this study regarding transfer intention and land use demand. The third type includes statistical analysis-based research methods, which usually make comprehensive statistics on economic, ecological and social security values, and guide the transfer of rural construction lands by making rational value judgment^27,28. However, under the influence of cities, suburban villages in mountain valley areas face hardly predictable permanent population, complicated construction conditions and continuous emergence of new changes. These factors can lead to errors in value judgment. Currently, there is a notable deficiency in the scenario description of various spatial changes occurring in the valley suburban villages following land transfer, as well as a lack of evaluation indicators for the systematic functional changes that ensue. Furthermore, most of them are divorced from the actual construction projects. Consequently, it is challenging to eliminate the uncertainties associated with land transfer and development construction in the valley suburban villages.

Scenario analysis is a method of future uncertainty research²⁹, where a “scenario” is a construct that describes the future developmental prospect of a system. As a flexible means of research, it not only can describe both the complex future prospect of space and the new changes in the development process, but also can make further value judgments on the social and economic benefits of results in combination with quantitative analysis measures³⁰. Relevant research has been widely applied in the field of urban and rural planning, and favourable effects have been achieved regarding urban development strategy, low-carbon urban planning, land use change and vulnerability of suburban agricultural lands^21,31,32,33. The research process covers factor identification, future event prediction, causal analysis, development path prediction, evaluation and decision making. It can describe specific spatial changes and eliminate uncertainties³⁴. Recently, AI technology has been able to preferably realize the text generation scenarios. These advantages make it possible to apply scenario analysis to the land transfer research in the valley suburban villages. Hence, by drawing on the scenario analysis-based planning model proposed by Li et al. (2012), this study statistically summarizes the demands of multiple parties within the research range, describes various complex scenarios and assesses the benefits of each scenario, thereby completing the spatial layout in stages to intuitively guide the transfer and planning of rural construction lands³⁵. Taking Ershilipu and Huazhaizi villages in Bali Town of Lanzhou City as examples, the problems of idle construction land and low utilization efficiency are solved, and a relatively ideal layout scheme of village construction lands is proposed.

Materials and methods

Research scope and data sources

The research objects, Ershilipu and Huazhaizi villages, are located in the Leitan River Valley in the southeast of Lanzhou urban area (Fig. 1). The valley is about 15 km long and 500–1000 m wide, which is flanked by continuous mountains. Since the two administrative villages have adjacent boundaries and some village buildings are continuous, unified implementation of land transfer and planning is required (Fig. 2). The construction land area of Ershilipu is 30.81 ha, representing 14.04% of its total land area, where the per capita arable land is only 98 m². The construction land area of Huazhaizi is 18.96 ha, accounting for 11.59% of its total land area, where the per capita arable land is only 42 m². Influenced by topography, the total area of rural production and construction lands is in serious shortage. However, the per capita construction land within the villages is 15% higher than the local standard, and there exists a housing vacancy rate of about 25.3%.

The research data includes boundary vector, satellite image, questionnaire interview and land use data of villages, as well as the benchmark land price of Lanzhou. Among them, the satellite image data comes from the National Platform for Common GeoSpatial Information Services (https://www.tianditu.gov.cn/), the village boundary and land use data are from the Third National Land Survey conducted by the Ministry of Natural Resources and Planning, the Lanzhou benchmark land price data is from the open government information (https://zrzyj.lanzhou.gov.cn), and the questionnaire interview data comes from field investigation. According to the Guidelines for Rural Planning in Gansu Province 2022, the number of questionnaires must exceed 30% of the total number of households. The distribution of these questionnaires has been entrusted to village officials, who disseminated them to villagers through door-to-door visits, telephone calls, and online channels throughout August 2023, with a distribution period lasting one month. A total of 720 questionnaires were distributed, and 714 were recovered, of which 636 were valid. The questionnaire covers 21 questions in four categories: village development scale, land use layout, environmental governance and industrial development. Additionally, the interview records of officials at various levels are used as the basis for determining the influencing factors of land transfer.

Fig. 1

Research range of this study.

Fig. 2

Settlement characteristics within the research range.

Research methods

This study adopts scenario analysis to form a research framework of “Scenario construction–Benefit evaluation–Dynamic planning”. Firstly, the construction land use needs are identified and the scenario conditions are designed. Next, by setting the thresholds and considering the key factors, the possible land transfer scenarios are preliminarily screened and respective development benefits are evaluated. Finally, the optimal scenario is screened out, and the ideal scene is described. It generated by AI technology to guide the development planning (Fig. 3).

Fig. 3

Technical route of this study.

Scenario construction

Scenario construction, which is based on a full understanding of the background and current situation of rural development, takes the predictable influencing factors of rural development into consideration as the constant variables or constraints³⁶. The predicted influencing factors can be classified into two categories: stable and variable. By drawing on Zhang’s research, this study puts forward five types of key factors influencing the construction land transfer, which cover the requirements of the master planning, the current situation of construction, the economic development, the government support and the villagers’ cooperation³⁷ (Table 1).

Table 1 Key influencing factors of rural construction land transfer.

In Table 1, the requirements of the master planning and the construction situation constitute the stable influencing factors. Among them, the master planning period can be extended to 2040, during which its guidance requirements for village construction should be strictly implemented. The current construction situation will not change before the land transfer, and the building style and function can be continued in the future to reduce the reconstruction cost. Hence, the requirements of the master planning and the construction situation are taken as the basis of scenario construction, based on which the land transfer zoning is determined (Fig. 4). Specifically, the northern part is zoned as high-density construction area (Zone A) relying on the existing high-rise residential community. The central part is zoned as rural factory agglomeration area (Zone B) relying on the existing agricultural and sideline product processing factories along with some rural settlements. The southern part is zoned as scattered settlement area (Zone C) relying on the height, quality and style of buildings.

Fig. 4

Current situation and zoning of rural construction lands.

In Table 1, three indicators, including economic development, government support and villagers’ cooperation, are closely related to the regional economic changes, managers’ administrative decisions and villagers’ transfer willingness. Hence, these three indicators are likely to mutate in the future, which are thus arranged and combined as variable factors to construct scenarios. Accordingly, a maximum of 3*3*3, or 27 possible land transfer scenarios can be formed. The specific spatial changes can be derived through scenario analysis. This approach is based on conjecture and aims to mitigate negative outcomes while enhancing positive effects by integrating several related individual forecast sets into a comprehensive overall forecast set¹¹. Firstly, in conjunction with the specific requirements for master planning, new project needs, settlement relocation plans, villagers’ willingness to work, and infrastructure construction, three land transfer models are developed: High-strength exploitation, Optimization and upgrading, and Renovation and improvement. Secondly, based on these models, the aforementioned projects will be implemented sequentially in space, leading to the formulation of a possible and relatively ideal blueprint. In this process, scenarios with similar functions are summarized to establish planning objectives and positioning. Finally, 27 development scenarios are generated, including Ecological livable, Upgrade industrie, Culture-oriented travel, Ecological tourism, etc. Additionally, more likely scenarios are selected by setting thresholds and progressively deliberating key factors (Fig. 5).

Fig. 5

Possible future land transfer models and development scenarios.

Benefit evaluation

According to Wang et al. (2021)’s study on the relationship between agricultural land transfers and household livelihood capitals, the benefits of rural construction land transfer primarily comprise economic and social benefits³⁸. Through the benefit evaluation, we can rank the possible scenario values, and describe the future ideal scenario.

Economic benefit estimation. The economic benefits are estimated using full costing, which means that net income is calculated as total revenue minus total costs. The relevant computational formula is presented in Eq. (1). The set Q represents the net revenue generated from the development of construction land, where Q={Q₁,Q₂,Q₃}. Here, Q₁, Q₂, and Q₃ denote the net income derived from the transfer of residential, commercial, and industrial land, respectively. Additionally, the sets A={A₁,A₂,A₃}, B={B₁,B₂,B₃} and C={C₁,C₂,C₃} are the sets of total construction revenue, construction cost and land cost of these three types of land respectively. Besides, their full development and full sale are assumed.

$$Q=A-B-C$$

(1)

Among them, the revenue from residential land transfer is denoted by Q₁. The net revenue from residential land development A₁ is the product of the sold residential floor area and the housing price E, with the residential floor area being the product of the floor area ration (FAR) and the transferred land area N. Construction cost B₁ is the sum of building demolition cost H, resettlement cost D and residential project construction cost G. Land cost C₁ is the product of the local benchmark residential land price F₁ and the transferred land area N. Accordingly, the computational formula for Q₁ is shown in Eq. (2).

$${{\text{Q}}_{\text{1}}}\,=\,{\text{E}}*{\text{N}}*{\text{FAR}} - ({\text{H}}\,+\,{\text{D}}\,+\,{\text{G}}) - {{\text{F}}_{{\text{1}}*}}{\text{N}}$$

(2)

Revenue from commercial land transfer. It is denoted by Q₂. The total revenue of commercial land development A₂, which mainly comes from rental and value-added profit, is highly uncertain. Thus, referring to the annual return on investment of local commercial and service industry 6.3%, and the number of years is n. Construction cost B₂ is the total development cost (including demolition, construction and operation) of commercial service projects. Land cost C₂ is the product of the local benchmark commercial land price F₂ and the transferred land area N. The resettlement cost during project development is D. Hence, Q₂ is comprehensively calculated by Eq. (3).

$${{\text{Q}}_{\text{2}}}=\left( {{\text{n}}*{\text{6}}.{\text{3}}\% +{\text{1}}} \right)*({{\text{B}}_{\text{2}}}\,+\,{{\text{F}}_{\text{2}}}*{\text{N}}) - {\text{D}}$$

(3)

Revenue from industrial land transfer. It is denoted by Q₃. In rural areas, the primary construction pattern of industrial lands is private development and construction, with features including simple construction and high return rate. The total revenue of industrial land development A₃ mainly comes from factory rental. Thus, referring to the local annual return on investment 6%, and the number of years is n. Construction cost B₃ is the total development cost (including demolition, construction and operation) of industrial projects. Land cost C₃ is the product of the local benchmark industrial land price F₃ and the transferred land area N. The resettlement cost during project development is D. Hence, Q₃ is comprehensively calculated by Eq. (4).

$${{\text{Q}}_{\text{3}}}=\left( {{\text{n}}*{\text{6}}0.0\% +{\text{1}}} \right)*({{\text{B}}_{\text{3}}}\,+\,{{\text{F}}_{\text{3}}}*{\text{N}}) - {\text{D}}$$

(4)

Social benefit estimation. According to He’s research, the social benefits of construction land transfer are reflected in the total pension insurance of local residents³⁹.By dividing the total insurance premium by the construction land area, the social security value of unit construction land transfer is obtained, and relevant computational formula is shown in Eq. (4). The total social benefit after land transfer is the product of the social security value of unit construction land transfer and the transferred land area N.

$$V = \frac{{Y*P}}{S}$$

(5)

where V signifies the social security value of unit construction land, Y denotes the one-time insurance payment by individuals based on average age, P represents the total insured population, and S is the area of construction lands. Besides, the additional social benefits of different scenarios should be qualitatively considered.

Dynamic planning

On the basis of ideal scenario, the prospect prediction is made and the planning scheme is put forward. Since both the scenario construction and benefit evaluation are based on possible assumptions, their results are relatively ideal. However, in the actual rural land transfer, factors like early investment and return period influence the project progress, thereby affecting the land transfer benefits. Thus, it is necessary to predict the phased variations by using the scenario analysis in 2040 (the end of planning period) as the final result. With reference to the recent construction plan, this study takes 5 years as one transfer cycle. Initially, based on the development process from the current phase to the final scene, the land transfer scheme should be considered by four factors: requirements of rural planning, possible development areas, the scale of land transfer and the difficulty of project construction. Among these factors, rural construction planning and phased development areas serve as constraints for the spatial elements change in dynamic planning. They can reflect the forecast results at various stages. The scale of land transfer and the challenges associated with project construction indicate the pressures on villagers’ livelihoods and resource allocation, which are the two most core indicators influencing land transfer⁴⁰. Next, the final scenario theme is defined, the design vocabulary of the theme is proposed, and each scene is described in depth. Further, the ideal scene is generated using Stable Diffusion based on the AI-generated logic. Finally, the theme is dynamically improved and the planning scheme is developed in phases.

The software Stable Diffusion can be downloaded from the website (https://www.stablediffusion-cn.com). For visualization and operational convenience, data in this study is processed through an open-source launcher (https://pan.quark.cn/s/2c832199b09b). Models and lora are loaded from the website (https://civitai.com), with the UrbanDesign_V7 model and UrbanAerial_V2 lora used in this study. Key elements of the current situation and ideal scenario are extracted. Each scene can be described in 40–50 words, refined from the scenario description. The ‘#test to image’ tool is utilized to generate the ideal scene, followed by iterative generation of scenes using software for multiple iterations and local modifications.

Results

The results cover those of scenario construction, benefit evaluation and dynamic planning.

Scenario construction results

Influenced by three variable factors (economic development, government support and villagers’ cooperation), up to 27 possible scenarios for the future rural development are formed. However, according to the development rate survey of Lanzhou rural economy and the interviews with relevant officials, a 2% economic development rate does not conform to the “National Economic and Social Development Plan of Lanzhou”, nor can it reach the development goals set by the superior government. Thus, all scenarios involving “Low speed development” are eliminated. Meanwhile, according to the national “Regulations on the Transfer of Rural Land Management Rights”, the construction land transfer should only be carried out with the consent of two-thirds of the villagers. If the current support rate of villagers is less than 40%, it will be difficult to meet the minimum (two-thirds) villagers’ cooperation requirement through lobbying. Thus, all scenarios involving “Refusal to cooperate” are eliminated. After the first round of scenario screening, 12 possible development scenarios remain.

In the 12 scenarios, the survey found that the rapid economic development necessitates high cooperation of villagers, which is only achievable by fully completing the land transfer. Thus, scenarios involving “Rapid development” and “Generally cooperate” should also be eliminated. Meanwhile, through questionnaire survey and interviews with villagers, it is difficult to lobby for the “Generally cooperate” scenario with the “Conditional support” of government, and even more difficult to attain the goal of “Medium-speed development”. After eliminating these 3 scenarios, 9 possible development scenarios remain.

According to the master planning of Lanzhou, Ershilipu and Huazhaizi will be included in the scope of urban construction lands in the future to become urban villages. Thus, under the land transfer model of high-strength exploitation, 100% of rural construction land transferred for urbanization construction is necessary, including commercial centers, industrial parks, residential areas and parks, with the goal of modern urban construction. Under the optimization and upgrading model, it is necessary to use 80% of rural construction lands for transfer by taking into account the livelihood guarantee and inherent lifestyle of farmers apart from the urbanization needs, so as to achieve balanced urban-rural development. Under the renovation and improvement model, 50% of transferred rural construction lands should be used for project construction to a limited extent, so that the villages can be renovated and upgraded to ensure basic agricultural production. Finally, we can derive 9 rural development scenarios and land transfer needs that may appear in the future (Table 2). Due to text limitations, table does not present a complete description for each scenario (Approximately 600 words).

Table 2 Rural development scenarios and land transfer factors.

Benefit evaluation results

Based on the land transfer needs under the 9 scenarios in Table 2, each scenario is evaluated for economic and social benefits.

Economic value assessment

Revenue from residential land transfer: If the area of transferred rural land is N m², according to the field investigation, the average price E of surrounding community is 5,627 yuan/m² and the average FAR of residential community is 2. Since the demolition cost is 34 yuan/m² based on average price in Lanzhou and the FAR of village buildings is 0.8, the cost of building demolition H is 27.2 N yuan. The resettlement cost is set following the demolition compensation standard of Bali Town. Since the resettlement compensation is 720,000 yuan per household, and the construction land area per household is 345.9 m², the resettlement cost D is 2081.53 N yuan. If the average price of residential construction cost in Lanzhou is 2,000 yuan/m², the construction cost of residential project G is 2,000 N yuan. Given that the benchmark land price of Lanzhou residential communities F₁ is 2,125 yuan/m², the estimated revenue from residential land transfer Q₁ is 5,020.27 N yuan.

Revenue from commercial land transfer: Excluding land transfer and project construction time, the number of years n of income from commercial and service facilities is 10 years. Since the average investment price of relevant commercial construction projects in Lanzhou is 4,000 yuan /m², the construction cost B₂ is 4,000 N yuan. Considering that the benchmark land price of Lanzhou commercial and service industry F₂ is 2,030 yuan/m² and the resettlement cost D is 2,081.53 N yuan, the estimated revenue from commercial land transfer Q₂ is 7,747.37 N yuan.

Revenue from industrial land transfer: Excluding land transfer and project construction time, the number of years n of income from industrial projects is 10 years. Since the average investment price of relevant industrial construction projects in Lanzhou is 1,500 yuan /m², the construction cost B₃ is 1,500 N yuan. Given that the benchmark land price of Lanzhou industrial estate F₃ is 484 yuan/m² and the resettlement cost D is 2,081.53 N yuan, the estimated revenue from industrial land transfer Q₃ is 1,092.87 N yuan.

Social value assessment

According to the questionnaire and interview, the average age of local villagers is 42, and the number of insured individuals represents 82.3% of the whole village population, so Y is 23,000 yuan. Through a combination of dependent coefficient and average growth rate methods, the number of people P under different scenarios is calculated to be 3,200–4,200 and S = 497,700 m². Using Eq. (4), V is calculated to be 0.046P yuan/m², so the social value is 0.046P*N yuan. Meanwhile, among the 9 scenarios, Ecological livable, Upgrade industrie, Culture-oriented travel, Ecological tourism and Development of characteristic agriculture not only elevate villagers’ income and rural living environment, but also improve ecological environment, increase employment rate, extend historical context and enhance villagers’ happiness. Thus, the above scenarios also have additional social benefits.

Comprehensive evaluation results

Table 3 lists the results of comprehensive benefit evaluation. With the villagers’ high degree of cooperation, a greater extent of rural construction land transfer indicates greater economic and social benefits. Through benefit evaluation, it is concluded that the optimal development scenario is “Culture-oriented travel” (Fig. 6). In this scenario, totally 49.8 ha of land is transferred, of which 53.8% is residential land and 46.2% is commercial land, including 12.2 ha of High-density construction area (Zone A), 19.8 ha of Rural factory agglomeration area (Zone B), and 17.8 ha of Scattered settlement area (Zone C).

Fig. 6

Land transfer and rural planning under the Culture-oriented travel scenario.

Table 3 Comprehensive benefit evaluations of various scenarios.

Dynamic planning results

To realize the final scenario, the land transfer and planning will be completed in three phases: Family farm renovation, Ecological tourism construction and Cultural tourism construction.

The 2025–2030 period will be the family farm renovation phase, where the quality of rural living environment will be improved. Initially, the residential lands in urban vicinities will be consolidated, and the ecological lands around the villages will be coordinated to build parks and squares. Next, a commercial complex called Shuimo Small Town will be constructed to create a commercial atmosphere. On its northern side, new high-rise residential will be built to meet the living needs of villagers and citizens. Afterwards, two new fruit-vegetable processing workshops will be built in the rural factory agglomeration area. Finally, five family farms will be built in the scattered settlement area to preliminarily realize the leisure tourism function. In this phase, 24.89 ha of construction lands will be transferred. The construction lands will be transferred in large scale during this period, and the investment volume will be large. The investment return and villagers’ resettlement are expected to be completed within 10 years.

The 2030–2035 period will be the ecological tourism construction phase, during which the quality of rural ecological environment will be improved and the sightseeing function will be realized. Initially, the riverways will be cleared to restore natural water flow, and greenways and continuous open spaces will be constructed in the riverfront area. Next, a new through-traffic highway will be built along the mountain in the village peripheries to address traffic needs. The original through road will be rebuilt into a living main road, with greenery and pedestrian spaces arranged on both sides. Afterwards, a landmark building will be constructed in the rural factory agglomeration area for reception and commodity sales. Finally, the quantity and quality of family farms and agritainment in the scattered settlement area will be increased. In this phase, 14.95 ha of construction lands will be transferred. The lands will be transferred in small scale during this period to relieve the project investment pressure from the first phase.

The 2035–2040 period will be the construction phase of cultural tourism, during which a tourism brand will be built based on farming culture, and a comprehensive tourist area for consumption, residence, sightseeing, experience and entertainment will be formed. Initially, in the riverfront area of Shuimo Small Town project, the traditional watermill facilities will be restored, the waterways and reservoirs will be rehabilitated, and a watermill culture exhibition hall will be built. Next, villagers on both sides of main road will be encouraged to rebuild their houses for providing homestay, catering and shopping functions. Besides, characteristic commercial blocks will be constructed. Finally, tourism service facilities such as toilets, management stations, medical and health care and sports venues will be built in each scenic spot. In this phase, 9.93 h of construction lands will be transferred, including 7.92 ha for cultural tourism projects and 2.01 ha for public service facilities. Through land use planning and projects construction over the three phases, land transfer is expected to be 100% completed in 2040, generating economic and social benefits of 3.20 billion yuan (Fig. 7).

Fig. 7

Three phases of land transfer and planning.

Figure 8 displays the future planning of rural construction projects. Firstly, by inheriting the characteristic local watermill culture, a Shuimo tourist area will be built in Ershilipu, which will cover retro watermill facilities (5), watermill culture exhibition hall (9) and tourist center (7). Totally 6,000 high-rise residential units will be built on the north side of tourist area, including the commercial houses for external sales (1) to solve the urban housing shortage problem; and the villagers’ resettlement houses (2) to solve the living problem of land-lost villagers. Secondly, the original polluting enterprises in the rural factory agglomeration area will be removed, and fruit-vegetable processing plant (25) will be constructed relying on the orchards on both sides for the processing, packaging and distribution of vegetables and fruits. In the scattered settlement area, family farm tourism projects (19, 20, 21, 22) will be developed relying on the orchards and a small amount of farmland, and villagers’ self-built houses will be renovated to construct homestay inns with local characteristics (23, 24). Some commercial land will be used to provide accommodation, catering and recreational facilities for tourists. Thirdly, a new through-traffic highway will be opened on the east side of villages to address the urban-rural connection needs. The original through road will be rebuilt into a living main road, with greenery and pedestrian spaces arranged on both sides (6). Relying on the natural river landscape on the east side, park and recreational greenway (9, 10, 11) will be built to provide outdoor activities such as fishing and rafting. Fourthly, the abandoned railway on the east side of villages will be reused to construct a train theme park running through the settlements (8, 13, 12). Hiking and cycling paths, mountaineering trails and observation deck will be established along the orchards on both sides of villages (14, 16, 17). Finally, three major travel routes will be formed in the region, namely the mountain fitness route (12-18-17-14-9-3), the agricultural experience route (23-13-19-22) and the hydrophilic health care route (8-10-11-12-13-20-21). There are about 40 subject words of AI describing each scene, and relevant examples are shown in Table 4.

Table 4 Ideal scenes and examples of their AI subject words.

Fig. 8

AI-generated intentional design diagram based on scene description.

Discussion and conclusion

This chapter discusses the advantages of this study in suburban village planning, its contributions to land transfer research along with the suggestions for spatial layout, and puts forward conclusions.

Discussion

Advantages of this study in suburban village planning

This study is based on our team’s rural planning of Huazhaizi and Ershilipu in 2021. Initially, we completed the planning by the method of conventional inductive-deductive. However, due to the epidemic, the planning could hardly be implemented. On the one hand, given the lack of autonomy in the suburban village development, the projects construction are completely affected by urban demands, which lacks resilience to external changes. On the other hand, most of the villagers work in the city, who are away from the traditional, fixed way of production and living. Thus, the land transfer willingness of villagers and government is difficult to predict, and may change in the short term. It is necessary to consider the multiple possible future development scenarios of the villages and make a value ranking. Although the “Culture-oriented travel” is chosen as the ideal scenario in this study, we can achieve faster transformation of planning objectives and scheme when facing abrupt changes. Moreover, the maximum period of rural planning is 20 years. During the progressive realization of planning objectives, the project construction will inevitably change, and the phased dismantling of scenarios and possible changes will be conducive to the planning implementation^41,42. Accordingly, the advantage of this study is its possibility to explore multiple scenarios and dynamic changes. Besides, with the rise of AI technologies like Chatgpt and Stable Diffusion in recent years, they have been gradually applied to the urban and rural planning, such as crowd behavior analysis, index calculation and scheme deduction^43,44. Despite the long history of scenario analysis, the scenario prediction, description and logic generation based on it is highly compatible with the current “Chat Generative Pre-trained Transformer”. Hence, this study preliminarily realizes the collaborative research of scenario analysis and AI scene generation in suburban village planning, which can offer new research ideas for the AI-based urban and rural planning.

Contributions of this study to land transfer research

Transfer and intensive use of land are not only related to the optimization of industrial development model and economic structure, but also have a great impact on the technological progress and national competitiveness⁴⁵. Existing studies primarily focus on the transfer of land and its associated benefits, often neglecting the potential outcomes of land transfer itself. The former belongs to the fields of economics and management, while the latter belongs to the fields of urban-rural planning and human geography. As a result, most of the existing studies lack a complete logic for problem solving. For instance, Jin and Deininger⁴⁶ explored the relationship between the area and value of land transfer through various factors such as transaction costs, risk assessment, and labor transfer; however, he did not clarify the specific location and function of land transfer. In the context of urban and rural planning, land value is intricately linked to land location and its systematic function following transfer. This study aims to provide an initial systematic discussion on the value of land transfer, incorporating a multidisciplinary perspective. Moreover, this result-oriented study presets the ideal scenario and completes the land transfer and planning. Notably, mainstream research on predicting future land changes predominantly utilizes the PLUS and CA-Markov models, resulting in overly macro-level insights⁴⁷. In contrast, this study focuses on the aspirations of villagers, governmental needs, and actual projects to predict the development trends of individual villages and address the specific challenges encountered in rural construction. To address the scientific problem of eliminating uncertainty in the development and construction of villages, this study enhances Li et al.’s scenario analysis model by expanding the 3 proposed analysis scenarios to 27. With the assistance of computational tools, the number of uncertain scenarios that can theoretically be analyzed and eliminated in this study approaches infinity³⁵. During the testing and assessment, quantitative approach is adopted to make the evaluation more scientific. Besides, AI analysis technology is applied in scenario description and definition, which can present research results more intuitively. Currently, the study on land transfer model, scheme and feasibility demonstration is relatively mature, and the policy suggestions are mostly put forward by senior scholars. Nonetheless, land transfer is implemented by the local government, and the target is villagers with low educational level. During the land transfer, it is necessary to encourage villagers for participation and active cooperation through lobbying and publicity. Thus, in the empirical research, proposing an ideal scene for acceptance from more villagers is necessary. To achieve this, generation of scene images is somewhat effective. If the AI model can be deeply applied to present the scene realization via video or VR, it will be more conducive to the public participation and the implementation of planning scheme. Consequently, this study demonstrates improved applicability and operability at the rural spatial level. Currently, the methods for benefit evaluation are overly simplistic, necessitating a more comprehensive approach that incorporates professional evaluation models from management science and engineering cost major.

Suggestions on the suburban village Spatial layout in mountain Valley area

Influenced by terrain constraints and urban negative externality, the valley suburban villages display spatial form features such as gathering along the river system, extension along the roadway, diverse landscape and rapid changes. This study corroborates the findings of Chen et al.⁴⁸. and Li and Jiang⁴⁹ regarding the distribution of rural settlements in river valleys. However, the current research remains overly macroscopic in its approach. While the authors argue that rural layout should depend on road or water systems at the watershed level, they fail to offer specific layout recommendations at the rural level. Through scenario analysis, the following aspects should be considered in the spatial layout of local villages: (1) Due to the shortage of urban construction land in the valley area and the convenient urban-rural connection in the suburban villages, the demand and income of land transfer are extremely high. The villages should be constructed as new urban districts, and the standards for residential communities and commercial, public service facilities should be higher. Despite the construction land shortage, it is still necessary to reserve a large area of greenbelts and open spaces, and complete the water management and riverfront construction to create a high-quality environment. (2) The current congestion of through traffic has seriously affected the environmental quality of suburban villages and the value of transferred lands. It is recommended to open a new through-traffic highway along the mountain in the village peripheries to address traffic needs. The original through road should be rebuilt into a living main road, with greenery and pedestrian spaces arranged on both sides. Multiple pedestrian crossings extending to the river system should be built from the main road, which connect various commercial districts, scenic spots and living quarters together with the riverfront tour lines. (3) The large-scale factory layout in the valley villages has low returns and affects environmental quality. However, small factories that produce no impact on the valley environment, such as agricultural and sideline, fruit processing and handicraft workshops, can be chosen as ancillary spaces for the development of characteristic agriculture and tourism. (4) A higher land transfer ratio should be encouraged, although it remains necessary to reserve a small area of construction land for coping with possible changes in spatial demand.

Conclusion

This study innovatively adopts scenario analysis to explore the construction land transfer and planning in the valley suburban villages, which solves the scientific problem of eliminating uncertainties during the suburban village development and construction. Taking Ershilipu and Huazhaizi villages in Lanzhou as examples, all possible scenarios are preset with variable factors, and uncertainties are progressively eliminated by setting thresholds, discussing key elements and evaluating benefits, so as to predict the possible future development trends, describe the scenarios based on AI technology and complete dynamic planning. The results show that the construction of Ershilipu and Huazhaizi should be coordinated to realize the development scenario of Culture-oriented travel. The land transfer should be completed in three phases: Family farm renovation, Ecological tourism construction and Cultural tourism construction. From 2025 to 2030, the abandoned residential lands in rural settlements will be consolidated, and 24.89 ha of construction lands near urban areas will be transferred to construct high-rise residential buildings, vegetable processing workshops and family farms. During the 2030–2035 period, 14.95 ha of construction lands on both river sides and around settlements should be transferred to build parks and scenic spots, with a view to improving the rural commercial and ecological environments. During the 2035–2040 period, 9.93 ha of low-value construction lands inside settlements needs to be transferred to construct cultural tourism projects and tourism service facilities. It is estimated that by the end of planning period, the total revenue will be approximately 3.20 billion yuan. Our research results can provide decision-making reference for the transfer, development direction and use arrangement of suburban village construction lands in mountain valley areas. The advantage of this study is its possibility to explore multiple scenarios and dynamic changes, which can identify the key elements affecting rural future development and how they are reflected in the spatial layout. Given the extremely complex problems faced by suburban villages, this study only constructs scenarios based on the key elements to simplify the analytical process. Since the evaluation of transfer benefits is rather simple, which only involves 6 types of land use indicators, it is necessary to further optimize the interpretation of scenario analysis on various land use changes. In this study, 27 scenarios are constructed, from which the ideal scheme is deduced progressively. In the future research, we will focus on more complex scenarios and more types of land use. The research logic of scenario analysis resembles the scenario generation procedure of current AI technologies such as Chatgpt and Stable Diffusion. In particular, the emergence of Sora makes the text-to-video research possible. Hence, the AI-assisted dynamic prediction of spatial scenarios will be our future research direction.