With the acceleration of urbanization since the 1980s, China has witnessed dramatic growth in urban construction, and this phenomenon has been observed in other developing countries worldwide [1
]. Until now, urban expansion has been the most common change in land-use patterns, and this growth is expected to continue for decades in China [3
]. In addition to promoting social and economic development, urbanization has also resulted in the reduction of the surrounding agricultural and ecological lands. This change has served as the impetus for the declines in regional ecosystem services and environmental quality [4
]. Additionally, the adverse ecological and environmental effects of the unreasonable utilization of land resources have hampered regional and urban sustainable development. Therefore, to achieve sustainable urban development and minimize its negative effects, decision makers must identify the appropriate areas for different development purposes, including land preservation and urban development.
Land-use suitability analysis aims at specifying the most appropriate spatial pattern for future land use according to specific requirements, preferences, or predictors of several activities [8
]. Since the 1980s, computer-aided technologies have been gradually adopted in the field of land-use suitability analysis. GIS-based land-use suitability analysis is undoubtedly the most broadly applied approach [11
]. In agriculture, forestry, and ecological protection, researchers have extended the evaluation of land utilization from specific crops [13
], tree species [14
], livestock and poultry [15
], or protected species [17
] to more general types of land use for multiple species at different scales [18
]. Another important application is site selection for urban [12
] and rural construction [25
], as well as human-made facilities for other purposes [26
]. Other applications include land suitability evaluation for environmental impact analysis [29
], landscape planning [30
], and assessments of geological favorability [31
]. In this study, we focus on the suitability evaluation of urban construction land. The increasing contradiction between the protection of cultivated lands and forests and the unreasonable expansion of urban areas has revealed failures in policies, legislation, and the execution of laws regarding land-use planning and management. Nevertheless, improving the ecological basis for these decisions is also important. The suitability evaluation of land for urban construction based on GIS is quantitative, normalized, and informational; therefore, this system can meet all the relevant demands. Improved analysis methods translate to better results for policy formulation, better land-use decisions, and increased sustainability.
McHarg’s work is generally considered to be of pioneering significance in regard to this approach. His team overlaid individual transparent maps containing natural and anthropogenic attributes with serial shading and constructed overall suitability maps for various utilizations through superimposing individual maps [9
]. Currently, the most widely applied approach involves the integration of GIS and Multiple Criteria Decision Making (MCDM). This approach is believed to be the most advanced method of configuring conventional overlaid maps [10
]. In this process, planners integrate spatial and non-spatial elements and generate predictable results via multi-index synthetic estimation [32
]. MCDM can be classified into multi-objective and multi-attribute decision-making methods [33
]. Because calculating the former is comparatively complex and difficult to achieve in GIS, the latter is much more widely used among researchers and planners. Such approaches include Weighted Linear Combination (WLC) [34
], Analytic Hierarchy Process (AHP) [13
], Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) [37
], and Ordered Weighted Averaging (OWA, commonly treated as an improvement of WLC) [38
However, classical maps overlay techniques based on Boolean operations and weighted analysis oversimplify the complexity of the evaluation process by neglecting certain indefinable factors of influence in the GIS environment. Some of these factors include the development potential of certain land-use components, and others are related to limitations. Unified scoring criteria and vertical units consider only quantitative differences and ignore qualitative differences, making it difficult for qualitative factors to be appropriately addressed [32
]. What is more, from a critical perspective, traditional approaches of land use suitability evaluation considered only the vertical processes of land units and neglected horizontal processes. Horizontal processes, such as the spatial movement of species, the sprawl of urban construction, and the spread of disasters, occur among different landscape units and are related to flows and interactions. These types of situations are unlikely to be expressed in McHarg’s overlay techniques [40
]. Therefore, we consider horizontal processes to describe the flows or interactions of factors among land units at the landscape level.
Numerous models describing horizontal processes have been developed and applied, such as fire ecosystem models [41
], gravity modeling and graph theory [21
], habitat suitability index models [43
], and the graphing of theoretic models [44
]. However, the complex computational processes and stringent data requirements of these models have limited their application [37
]. Knaapen proposed the minimum cumulative resistance (MCR) model to measure habitat isolation [45
]. This model utilizes a simple and clear data structure and fast algorithms, with highly visual results [46
]. Furthermore, it considers both the distance and the effects of adjacent landscape characteristics, which have rarely been addressed in previous studies. Therefore, it has been frequently used for studies in urban expansion [46
]. The main idea of the MCR model can be summarized as follows: in nature, any movement must face resistance. For a horizontal movement of an object (expansion, fluxion, etc.), the less resistance the object encounters, the easier the movement happens. Besides, the results of vertical suitability analysis for urban construction land with calculated resistance values can be directly input into the MCR model, which was designed for horizontal suitability evaluation. Therefore, this makes the vertical and horizontal suitability evaluations easier and more practicable than previous individual case studies [49
]. This combined method may offer the opportunity to overcome the deficiencies of traditional methods of land use suitability evaluation.
This paper aims to provide a scientific basis for selecting land for urban construction at the development scale while informing decisions about the optimal direction for urban growth. This paper also presents a new discussion and theoretical exploration regarding suitability evaluations of urban construction land. This study explores a new method of evaluating the suitability of urban construction land based on the comprehensive analysis of both vertical and horizontal processes. The suitability evaluation of vertical processes in this paper refers to the traditional land suitability evaluation method, that is the classical overlay techniques. ”Vertical” is a picturesque phrase. Here, we describe the process of overlay techniques of different natural or humanistic attributes as ”vertical” instead of processes that actually happen in a vertical direction. Meanwhile, MCR model simulates the urban expansion on horizontal processes and therefore can be described as ”horizontal”, which cannot be expressed or calculated by overlay maps of vertical processes. Our study attempt to combine these two processes, and therefore we call them ”vertical” and ”horizontal” graphically. The Beihu New District of Jining City, Shandong Province, China, was taken as a case study, and it was organized into following sections. First, we simultaneously considered the potential and the resistance of urban construction and development. Then, we established a potential-resistance model following the suitability analysis of urban construction land concerning the vertical processes. Next, based on the vertical suitability analysis results, we regarded the urban built-up areas as “sources” and simulated urban spatial sprawl using the MCR model following the suitability regionalization of urban construction land, which concerns the horizontal processes of the study area. Based on the threshold values of resistance, we divided the study area into the priority, suitable, restricted, and prohibited areas.
Using Beihu New District as a case study, we explored a potential-resistance model for urban development and construction based on previous studies and classical suitability evaluation methods of land use. This model comprehensively considered the potential and ecological resistance of urban construction to analyze the suitability of vertical processes. Subsequently, we applied the MCR model to simulate and predict urban spatial expansion and evaluate the suitability of horizontal processes. Finally, we divided the study area into priority, suitable, restricted, and prohibited areas and defined the proper scales and spatial location for urban construction and development. We also defined the ecological borders of future urban growth in the study area.
Compared to traditional suitability evaluations of land use, which typically consider only the vertical processes, the method proposed in this study integrated both the vertical and horizontal processes that affect urban development and construction. Our new method improves upon traditional evaluation methods and not only enriches the approaches used in land suitability evaluations but also makes the evaluation results more feasible and reasonable. Notably, the method improves both the objectivity of urban spatial decision making and the scientific validity of land-use planning. Moreover, it may have considerable value in guiding the compact and multi-center development of local urbanization systems, determining the industrial functions and residential demands of the central city, and decreasing the unnecessary waste of ecological and cultivated land resources. Nevertheless, in certain regions under different conditions, further explorations are still needed, including selecting, classifying, assigning, and conducting weighted calculations involving evaluation factors and introducing political factors from a spatial perspective. Furthermore, based on our new method of land suitability evaluation, multi-scene urban spatial expansion simulation, and land-use layout optimization, combining the existing urban land-use change prediction models with rational scenario design will be the focus of future research.