Method and system for identifying range and type of multi-scale geographic transition area, and medium

Abstract

The invention discloses a method and a system for identifying the range and the type of a multi-scale geographic transition area, and a medium, and belongs to the technical field of geographic information science. The method comprises the following steps: S1, constructing a humanity component and natural component index framework for identifying a transition region; s2, improving the coupling model, and measuring the interaction direction and strength between the social economic system and the natural ecosystem; s3, identifying transition areas of the grid units and performing type division; s4, using a gradient mutation detection method to identify a transition area in a typical direction, and combining a spatial mutation detection method to identify coupling index mutation on a regional scale so as to determine the boundary and type of the geographic transition area; and S5, training and predicting a transition type of a target level based on a random forest intelligent classification method, and further downscaling the geographic transition region to a target administrative unit, thereby obtaining different types of coupling system partitions. Through combination of the administrative unit and the grid unit, accurate matching of data is realized, and the efficiency of transition area identification is improved.

Description

Technical Field

[0001] The present invention relates to the field of geographic information science and technology, and in particular to a method, system and medium for identifying the scope and type of multi-scale geographic transition zones. Background Art

[0002] The obvious environmental gradients in mountainous areas have shaped interconnected social-ecological systems with diverse landscapes, complex system structures and multifunctional lands. The huge sustainability challenges faced by social-ecological systems are closely intertwined in time and space. Under the situation of global change, the evolution of social-ecological systems has become more dynamic and complex. Understanding the interaction between humans and the environment in social-ecological systems is a core issue to reduce system pressure and ease the tension between internal elements. As a nonlinear dynamic system, social-ecological systems present a transition state from one stable equilibrium to another. Quantitative measurement of this transition state or property can effectively reveal the internal relationship of social-ecological systems.

[0003] Environmental gradients are a proxy for transition and are increasingly being used to support the understanding of rapidly changing and increasingly complex socioecological systems, particularly in mountainous areas with significant regional heterogeneity. The environmental gradient paradigm suggests that environmental change is spatially ordered and drives the distribution of land structure and functional components. The gradient perspective implies the identification of transition zones for specific land use / environmental factors, such as agroforestry and agropastoral transition zones. These transition zones have specific and critical ecological processes and are representative of complex landscapes, where unstable conditions involve both internal landscape configurations and relationships with the surrounding environment. The urban-rural natural gradient exists in a spatial sequence of different scales and directions, and it can be used to illustrate the role of a series of spatial movements such as rapid urbanization, rural transformation and reconstruction, and the construction of nature reserves in enhancing regional spatial complexity, diversity, and dynamics. The gradient consists of a series of adjacent and interacting landscapes, forming a continuous landscape sequence, among which transitional landscapes that are connected to the main landscape and promote landscape functional interactions have received increasing attention. Transitional landscapes play an indispensable role in environmental gradients and are a means of spatial connection and communication between landscapes in different functional areas. Therefore, transitional landscapes have complex spatial interfaces that transmit physical resource and energy information from different functional spaces on both sides of the interface.

[0004] In addition to the urban-rural transition zone and the agro-pastoral transition zone, many other ecological transition zones or geographical transition zones are concentrations of extensive transitional landscapes. However, research on urban-rural transition zones often emphasizes the definition and analysis of transitional geographical spaces and gradient landscapes within the socio-economic gradient. The concept of the agro-pastoral transition zone focuses more on the natural environment shaping along landscape gradients and transitional geographical spaces rather than human factors. Recent studies have focused on incorporating the impact of human activities and environmental changes into the assessment and demarcation of landscape gradients and transitional zones, such as demarcating and investigating landscape gradients based solely on urban and rural natural land use. These studies show that despite considering multiple elements of social ecology, scholars have a certain tendency towards socio-economic systems or natural ecological systems. In order to explore the social and ecological coupling processes in complex regional systems, it is necessary to conduct a comprehensive study of multi-scale social-ecological systems from the perspective of transition.

[0005] In the prior art, such as the authorized CN115661670B, a method and device for identifying transition zones on the gradient of urban and rural natural landscapes, which extracts transition curves in each county by constructing a raster pixel value reordering technology, the identified transition curves are more accurate, and the accuracy of identifying transition zones on the gradient of urban and rural natural landscapes is improved. However, it does not combine administrative units and grid units to accurately match data, the workload is large, the computing resources are more occupied, the efficiency of transition zone identification is low, and the identification accuracy needs to be improved.

[0006] In existing studies, geographical transition zones are identified and analyzed based on single-scale research. In urban-rural transition zones, agricultural-pastoral transition zones, and ecological transition zones, quantitative methods using single data are used to identify and analyze or extract gradient mutation characteristics and breakpoints of multiple factors. These mainly include density threshold, static linear global weighted algorithm analysis, gradient threshold mutation points of multiple factors, etc. The identification technology of transition zones considering a single indicator or significant feature can only be used in specific transition areas. For example, based on the proportion of construction land in land use or urbanization rate, only urban-rural transition zones can be identified, but ecological transition zones such as oasis-desert transition zones and agricultural-pastoral transition zones cannot be identified. Vegetation coverage, climate and other indices applicable to ecological transition zones are less sensitive to urban-rural transition zones. In areas containing multiple types of transition zones, especially when there is a strong game between social and economic development and ecological protection within the system, a single indicator cannot fully characterize its transition characteristics. Although static linear weighting can superimpose multi-factor characteristics, it will cause a lot of information loss, especially the insufficient measurement of the interaction between systems. In addition, when multiple transition zones are complexly intertwined, traditional methods cannot clearly distinguish the types and numbers of transition zones contained within them, as well as the precise scope of stripping off a certain type of transition zone. However, many concepts such as ecological transition zones, agricultural-pastoral transition zones, urban-rural transition zones, oasis-desert transition zones, etc. lack a systematic and comprehensive overview of their connotations and scope definition standards. Overall, the identification and classification of transition zones within complex and intertwined social-ecological systems face great limitations and high uncertainty. The following problems also exist in the prior art: 1. The definition and analysis of socioeconomic gradient landscapes within specific transition zones, such as urban-rural transition zones or agricultural-pastoral transition zones, require specific research methods and frameworks, and the research efficiency is low. 2. Traditional gradient identification based on environmental factors or comprehensive integration and identification using static linear global weighted algorithms results in a large amount of information loss. 3. Transition zones identified at a single grid scale or administrative unit scale have great limitations on the specific processes of regional planning and land management. Summary of the invention

[0007] The purpose of the present invention is to overcome the deficiencies of the prior art and to provide a method, system and medium for identifying the scope and type of multi-scale geographic transition zones.

[0008] The object of the present invention is achieved through the following technical solutions: The first aspect of the present invention provides: a method for identifying the scope and type of a multi-scale geographic transition zone, comprising the following steps: S1: Select the characterization factors of socio-economic attributes and natural ecological attributes, and construct the indicator framework for identifying the human and natural components of the transition zone; S2: Improve the coupling model, perform nonlinear superposition of human and natural components, and measure the direction and intensity of interaction between socio-economic systems and natural ecosystems; S3: Based on the positive and negative values ​​of the coupling index, the proportion of human and natural components in the target area is characterized, the transition zone of the grid unit is identified and classified; S4: Use the gradient mutation detection method to identify transition zones in typical directions, and combine it with the spatial mutation detection method to identify coupling index mutations at the regional scale, thereby determining the boundaries and types of geographical transition zones; S5: Based on the random forest intelligent classification method, the transition type at the target level is trained and predicted, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

[0009] Preferably, the S1 further comprises the following steps: Characterization factors of socio-economic attributes and natural ecological attributes are selected, settlement patch density and agricultural production functions are spatially expressed, and a grid-scale representation data set is obtained, thereby constructing an indicator framework for identifying the human and natural components of the transition zone.

[0010] Preferably, the S2 further comprises the following steps: The initial coupling model is obtained by integrating the physical coupling model and the management trade-off model. CSI , , ; ; , ; in, CP is the coupling degree, indicating n The coupling relationship between the systems; CD is the degree of coordination, which indicates the overall coordination level between systems; TO yes n The trade-offs between systems; b i It is a system i relative benefits; It is a system i expected benefits; It is a system i The importance of Improved coupling model obtained by derivation simplification ; , ;in, x for the relative benefits of the socio-economic system; y for the relative benefits of natural ecosystems; Using coupled models The human and natural components are further superimposed nonlinearly to measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem.

[0011] Preferably, the S3 further comprises the following steps: Based on the positive and negative values ​​and the size of the coupling index, the proportion of human and natural components in the target area is characterized, the coupling effect between the socio-economic system and the natural ecosystem is represented, the geographical transition zones of the grid units are identified and divided into types, and all traditional transition zones are obtained by adjusting the classification criteria of the coupling index.

[0012] Preferably, the S4 further comprises the following steps: Use the gradient mutation detection method to identify transition regions in typical directions. ,in t is the gradient sequence in mutation detection; t 1 is the horizontal coordinate of the starting point of the gradient; t 2 is the abscissa of the end point of the gradient; y 1 is the starting ordinate of the gradient; y 2 is the ordinate of the mutation point; y 3 is the ordinate of the end point of the gradient; T 1 is the abscissa of the mutation point; f is a piecewise function representing the gradient partition; region 1 is the function before the mutation point; region 2 is the function after the mutation point; if Function means if t <= T 1 Return if the condition is met region 1, if not true, return region 2; The spatial slope function is introduced to identify the coupling index mutation on the regional scale using the spatial mutation detection method. ;in, C _ S ij is the spatial slope of factor i at pixel j, and a larger value indicates a greater uncertainty in the environmental factor; dz / dx is the rate of change of the factor in the first-order neighborhood in the horizontal direction; dz / dy is the rate of change of the factor in the vertical direction; p m and p n denote the factor values ​​in the first-order neighborhood in the north-south and east-west directions, respectively; w s and w eRepresents the horizontal weighted counts of valid pixels in the north-south and east-west directions, respectively; p size is the size of a pixel; Then, typical directional gradient county units are used to verify the accuracy of mutation detection in the entire space, thereby determining the boundaries and types of geographical transition zones.

[0013] Preferably, the S5 further comprises the following steps: Based on the random forest intelligent classification method, the environmental factors divided by the transition zone type and impact type of the county unit and the indicators for measuring the information amount of the environmental factors are used as training samples to predict the transition type of the target level, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

[0014] Preferably, the target level is the township level, and the target administrative unit is the township administrative unit.

[0015] The second aspect of the present invention provides: a system for identifying the scope and type of a multi-scale geographic transition zone, which is used to implement any of the above methods for identifying the scope and type of a multi-scale geographic transition zone, comprising: The framework building module is used to select the characterization factors of socio-economic attributes and natural ecological attributes, and to construct the indicator framework for identifying the human and natural components of the transition zone; The measurement module is used to improve the coupling model, nonlinearly superimpose the human and natural components, and measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem; The type classification module is used to characterize the proportion of human and natural components in the target area based on the positive and negative values ​​and numerical values ​​of the coupling index, identify the transition area of ​​the grid unit and perform type classification; The mutation detection module is used to identify the transition zone in the typical direction using the gradient mutation detection method, and to identify the coupling index mutation at the regional scale by combining the spatial mutation detection method, so as to determine the boundary and type of the geographical transition zone; The zoning module is used to train and predict the transition type at the target level based on the random forest intelligent classification method, and further downscale the geographical transition zone to the target administrative unit to obtain different types of coupled system zoning.

[0016] The third aspect of the present invention provides: a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are loaded and executed by a processor, any of the above-mentioned methods for identifying the scope and type of multi-scale geographic transition zones is implemented.

[0017] The beneficial effects of the present invention are: 1) The improved coupling model has the ability to simultaneously detect the direction and intensity of interaction between systems, and can more accurately detect the interaction between systems in the region; combining the improved coupling model, mutation detection, and intelligent classification technology based on random forests can more accurately identify the scope and type of the transition zone.

[0018] 2) By combining administrative units and grid units and using multi-source geospatial data to characterize the coupling of mountain social-ecological systems, accurate data matching was achieved, which greatly reduced the workload, reduced the use of computing resources, and improved the efficiency of transition zone identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Flowchart of the method for identifying the extent and type of multi-scale geographic transition zones; Figure 2 Schematic diagram for determining the degree of coupling, integration level and interaction direction between systems for the improved coupling model; Figure 3 Schematic diagram of the spatial pattern of the social-ecological system coupling index in mountainous areas; Figure 4 Schematic diagram of various types of transition zones determined for different classifications of coupling index; Figure 5 It is a schematic diagram of the cross-sectional gradient of the coupling index in eight typical directions in the study area in a specific embodiment; Figure 6 This is a schematic diagram of the spatial mutation index of counties in the Sichuan Basin; Figure 7 This is a schematic diagram of the transition types of towns in the Sichuan Basin. DETAILED DESCRIPTION

[0020] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0021] This paper combines administrative units and grid units, uses multi-source geospatial data to characterize the coupling of mountain social-ecological systems, and identifies multi-scale geographic transition zones. It can fully capture the complexity and dynamics of geographic transition zones, improve the accuracy and reliability of transition zone identification, and overcome the limitations of single-scale research.

[0022] The coupling coordination model and the trade-off model are integrated and improved, and multi-source geospatial data are nonlinearly superimposed to measure the interaction and coupling relationship between socio-economic systems and natural ecosystems, so as to accurately identify and classify transition areas and avoid information loss caused by linear superposition.

[0023] By analyzing the coupling direction and strength between social-ecological systems, the characteristics of the transition zone and the internal mechanism of its formation are portrayed, so that the characteristics of the transition zone such as urban-rural, ecological, topographic, and climatic are systematically characterized through unified means, the scope of the region is identified, and the geographical transition zone is systematically characterized.

[0024] See also Figure 1-Figure 4 The first aspect of the present invention provides: a method for identifying the scope and type of a multi-scale geographic transition zone, comprising the following steps: S1: Select the characterization factors of socio-economic attributes and natural ecological attributes, and construct the indicator framework for identifying the human and natural components of the transition zone; S2: Improve the coupling model, perform nonlinear superposition of human and natural components, and measure the direction and intensity of interaction between socio-economic systems and natural ecosystems; S3: Based on the positive and negative values ​​of the coupling index, the proportion of human and natural components in the target area is characterized, the transition zone of the grid unit is identified and classified; S4: Use the gradient mutation detection method to identify transition zones in typical directions, and combine it with the spatial mutation detection method to identify coupling index mutations at the regional scale, thereby determining the boundaries and types of geographical transition zones; S5: Based on the random forest intelligent classification method, the transition type at the target level is trained and predicted, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

[0025] In this embodiment, administrative units and grid units are combined to determine the multi-scale geographical transition zone: through the improved coupling model, the interaction between the socio-economic system and the natural ecosystem, as well as their degree of coupling at different scales are analyzed; mutation detection technology is applied, a linear piecewise function is used to identify transitions in typical directions, and spatial mutation technology is combined to identify mutations in environmental factors at the regional scale, thereby determining the boundaries of the geographical transition zone; intelligent classification based on random forests: a random forest model is constructed to train and predict the transition type at the township level, and the geographical transition zone is further subdivided into different types of coupled system areas; the spatial balance of the geographical transition zone is analyzed, and its heterogeneity and spatial development balance are evaluated, which helps to understand the structural complexity and functional diversity within the transition zone.

[0026] In some embodiments, the S1 further comprises the following steps: Characterization factors of socio-economic attributes and natural ecological attributes are selected, settlement patch density and agricultural production functions are spatially expressed, and a grid-scale representation data set is obtained, thereby constructing an indicator framework for identifying the human and natural components of the transition zone.

[0027] In this embodiment, information such as topography, climate, soil, vegetation, population, land use, economy and location is used to reflect the regional natural and human systems. Among them, the topographic location index, humidity index, cumulative temperature (≥10°C), soil erosion and normalized vegetation index are used to describe the natural system. Settlement patch density, land function index, GDP density and traffic accessibility are used to symbolize the human system. The spatial (quantitative) expression process of the characterization factors of socio-economic attributes and natural ecological attributes is shown in Table 1.

[0028] Table 1 In some embodiments, the S2 further comprises the following steps: The initial coupling model is obtained by integrating the physical coupling model and the management trade-off model. CSI , , ; ; , ; in, CP is the coupling degree, indicating n The coupling relationship between the systems; CD is the degree of coordination, which indicates the overall coordination level between systems; TO yes n The trade-offs between systems; b i It is a system i relative benefits; It is a system i expected benefits; It is a system i The importance of Improved coupling model obtained by derivation simplification ; , ;in, x for the relative benefits of the socio-economic system; y for the relative benefits of natural ecosystems; Using coupled models The human and natural components are further superimposed nonlinearly to measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem.

[0029] In this embodiment, the coupling degree ( CP ) originated from physics. n system or within a system n The relationship between the elements is expressed. Considering the importance of each part ( w i ), coordination degree ( CD ) reflects the overall coordination level between systems. n The trade-off between the goals TO ) can be calculated by the root mean square error. CP , CD , TO Fusion to get the initial coupling model CSI The trade-off between two systems is defined as the distance from the zero trade-off line, and a larger distance means a larger trade-off between the two. By improving the trade-off model and the coordination model, the present invention obtains a new model. The new model can determine the coupling degree, integration level and interaction direction between systems, such as Figure 2 As shown in the figure, it is an illustration of the coupling, coordination, trade-off and improved coupling between two systems: the order of points a, b and c in the figure is |a|>|b|=|c|. b=–c can only be obtained in the process of trade-off and improved coupling. The trade-off model is opposite to the coupling model. The smaller its value, the better the system state. Therefore, it is necessary to negatively normalize part of the new model and finally obtain the coupling model , in the coupled model Human systems (X) and natural systems (Y) are of equal importance.

[0030] In some embodiments, the S3 further comprises the following steps: Based on the positive and negative values ​​and the size of the coupling index, the proportion of human and natural components in the target area is characterized, the coupling effect between the socio-economic system and the natural ecosystem is represented, the geographical transition zones of the grid units are identified and divided into types, and all traditional transition zones are obtained by adjusting the classification criteria of the coupling index.

[0031] In some embodiments, the S4 further comprises the following steps: Use the gradient mutation detection method to identify transition regions in typical directions. ,in t is the gradient sequence in mutation detection; t 1 is the horizontal coordinate of the starting point of the gradient; t 2 is the abscissa of the end point of the gradient; y 1 is the starting ordinate of the gradient; y 2 is the ordinate of the mutation point; y 3 is the ordinate of the end point of the gradient; T1 is the abscissa of the mutation point; f is a piecewise function representing the gradient partition; region 1 is the function before the mutation point; region 2 is the function after the mutation point; if Function means if t <= T 1 Return if the condition is met region 1, if not true, return region 2; The spatial slope function is introduced to identify the coupling index mutation on the regional scale using the spatial mutation detection method. ;in, C _ S ij is the spatial slope of factor i at pixel j, and a larger value indicates a greater uncertainty in the environmental factor; dz / dx is the rate of change of the factor in the first-order neighborhood in the horizontal direction; dz / dy is the rate of change of the factor in the vertical direction; p m and p n denote the factor values ​​in the first-order neighborhood in the north-south and east-west directions, respectively; w s and w e Represents the horizontal weighted counts of valid pixels in the north-south and east-west directions, respectively; p size is the size of a pixel; Then, typical directional gradient county units are used to verify the accuracy of mutation detection in the entire space, thereby determining the boundaries and types of geographical transition zones.

[0032] In the present embodiment, mutation detection is based on superimposed environmental factors. The coupling index is an index covering comprehensive environmental factors and can be directly used for mutation detection. By constructing the gradient of the cross section, a linear piecewise function is used to identify the transition in the typical direction. The difference between the rate of change before and after the mutation (mutation intensity) reflects the transition intensity. The method of selecting the cross section is only applicable to the change of geographical features along a certain direction in the region. Although the transition characteristics of the typical region can be accurately identified, the situation of the entire study area cannot be obtained, especially in areas where the gradient characteristics are unclear or non-monotonic. In the present invention, a spatial mutation method is further used to detect mutations and classify the transition types in the Sichuan Basin. Clearly distinguishing land use water bodies (especially rivers) and construction land with a spatial resolution of 500 meters is not conducive to detecting sudden changes in environmental factors in the region. Therefore, the raster data of the entire region is smoothed. The spatial slope of the entire region was calculated, and the average spatial slope (spatial mutation) of each county-level administrative unit was statistically analyzed.

[0033] In some embodiments, the S5 further comprises the following steps: Based on the random forest intelligent classification method, the environmental factors divided by the transition zone type and impact type of the county unit and the indicators for measuring the information amount of the environmental factors are used as training samples to predict the transition type of the target level, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

[0034] In this embodiment, a random forest model is constructed to train, simulate and predict the transition types of towns in the Sichuan Basin. In this model, natural and human factors affecting regional geographical transformation are used as independent variables, and the transformation types of 170 county units in the Sichuan Basin are used as dependent variables. Strong transitions (positive samples) and weak transitions (negative samples) in the training and test samples in the study are jointly considered to prevent model bias. All samples are divided into two parts, 70% are allocated to training samples, and the remaining 30% are allocated to test samples. For training, a random sampling method with a constant sample size is adopted, and the out-of-bag scoring parameter of the random forest model is used.

[0035] In some embodiments, the target level is the township level, and the target administrative unit is the township administrative unit.

[0036] A specific embodiment is given below: (1) Geographical transition division of grid units The spatial pattern of the coupling index of social-ecological systems in mountainous areas is shown in Figure 3As shown in the figure, (a), (b) and (c) represent the coupling index of the Sichuan Basin, the proportion of each group of coupling index and the ten counties with the largest proportion of EST types. WDN = nature-dominated (weak transition dominated by natural ecological processes), SNE = strong humanities and weak nature (strong transition with natural ecological characteristics stronger than socio-economic characteristics), EST = semi-humanities and semi-nature (extremely strong transition of complex coupling of humanities and nature), SSE = strong nature and weak humanities (strong transition with socio-economic characteristics stronger than natural ecological characteristics), WDS = humanities-dominated (weak transition dominated by socio-economic processes). The higher the coupling index, the more obvious the socio-economic characteristics of the patch; the lower the index, the stronger the natural ecological characteristics of the patch. When the coupling index is close to 0, both socio-economic and natural ecological characteristics of the patch exist, but neither is prominent. Based on the coupling index from low to high, the geographical transition of the Sichuan Basin is divided into five groups according to the coupling intensity between systems. The natural ecological attributes are stronger than the socio-economic attributes, which are dominated by natural processes (WDN) and strong nature and weak humanity (SNE). The natural ecological attributes are weaker than the socio-economic attributes, which are dominated by human processes (WDS) and strong humanity and weak nature (SSE). The complex coupling of natural ecosystems and socio-economic systems is semi-human and semi-natural (EST). There are more WDN-type areas in the western part of the study area than in the eastern part, and the central part of the basin is dominated by WDS-type areas. The EST area is distributed in a circular shape between the WDN and WDS-type areas.

[0037] WDS covers an area of ​​96,500 square kilometers, accounting for 35.2% of the Sichuan Basin. WDN has the smallest area, 36,200 square kilometers, accounting for only 13.2% of the region ( Figure 3 (b) in the text). Figure 3 Part (c) of the report lists the top ten counties / districts with the highest EST ratio in the Sichuan Basin. The EST area of ​​Gulin County is 26,700 square kilometers, accounting for 84.6% of the total area of ​​the county. Nine of the top ten counties / districts with the highest EST ratio in the Sichuan Basin were previously national-level poverty-stricken counties.

[0038] Specifically, the division of coupling strength is a division of the overall socio-ecological system type of the region. Based on different research and application purposes, different levels of coupling index can determine the transition zone of each type ( Figure 4 When the coupling index is > 0.4, that is, the humanities component exceeds 90%, the region can be determined as an urban densely populated area; when the coupling index is around 0.2~0.4, that is, the humanities component accounts for 70%~90%, the region can be determined as an urban-rural transition zone ( Figure 4 (b) and (c)); When the coupling index is around -0.15~0.15, the ecological component and the human component account for almost half each, between 40% and 60%. At this time, the area can be determined as a mountain-plain transition zone ( Figure 4Similarly, the ecological attributes of the hilly plain transition zone are lower than those of the mountainous plain transition zone, which may be between -0.1 and 0.15, and can be adjusted according to actual conditions. When the coupling index is around -0.4 to -0.3, the overall ecological attributes of the region are strong, and the ecological components are about 80% to 90%. At this time, the region can be determined as an agricultural-pastoral transition zone ( Figure 4 (e) and (f) in the figure).

[0039] (2) Division of county-level administrative units 1. Typical county transition type: The coupling index of the Sichuan Basin shows a clear upward trend from the edge of the basin to the interior. The geographical transition of administrative units can be identified by analyzing the changes in the coupling index.

[0040] The results show that the cross-sectional gradients of the coupling index in eight typical directions within the study area are as follows: Figure 5 As shown, the change rate is expressed as CR, and its value is the change of coupling index per 1000 km. According to the mutation intensity from strong to weak, the transition of each county can be divided into the following three categories: sharp transition (coupling index mutation is greater than 6 per 1000 km), fast transition (coupling coefficient mutation is 2-3 per 1000 km) and slow transition (coupling index mutation is less than 1.5 per 1000 km) (Table 2). The transition of the northwest gradient is the strongest and is the only sharp transition type among the eight transects. Shifang is a typical county with a coupling index change of 6.65 per 1000 km. This area represents the transition zone between the Qinghai-Tibet Plateau and the Chengdu Plain. The southeast is characterized by a slow transition from the Sichuan hilly basin to the Dalou Mountain, with the smallest change in coupling index of 0.77 / 1000 km, and the typical representative county is Jiangjin District.

[0041] Table 2 2. Spatial mutation and type of each county: The spatial mutation index (spatial variability) of each county in the Sichuan Basin is as follows: Figure 6 As shown. Low spatial mutation (0-0.12) means that there are many WDN or WDS areas in this type of area, areas with a spatial mutation range of 0.12 to 0.2 present a slow transition, areas with a spatial mutation amplitude of 0.2 to 0.34 correspond to a rapid transition, and areas with a spatial mutation greater than 0.34 represent a sharp transition. Among all the transition counties obtained using the spatial mutation method, only the transition type of Lizhou District, Guangyuan City is inconsistent with the transition type calculated by the cross-sectional gradient (Table 3). Cross-sectional mutation detection requires a path that is as long as possible across the county in order to detect a larger amount of information, while the transect in Lizhou District crosses the county horizontally. Spatial mutation has a strong applicability in detecting mutations and identifying transition types.

[0042] Table 3 More than 54% of the counties in the Sichuan Basin belong to the weak transition type. Among them, the counties distributed in the central part of the basin belong to the WDS type, with a total of 87 counties and a high level of socioeconomic development. Six counties in the western and southwestern parts of the basin—Maoxian, Luding County, Leibo County, Meigu County, Yuexi County, and Baoxing County—belong to the WDN type (Table 4). These areas have high altitudes, steep slopes, and low intensity of human activities. There are 38 slowly transitioning counties, accounting for 22.4% of the Sichuan Basin. Slowly transitioning counties are widely distributed in the north and south of the Sichuan Basin. There are 23 rapidly transitioning counties; these are distributed on the edge of the basin and are rare in the south. Sharply transitioning counties are mainly located in the western part of the study area, that is, the transition zone between the Chengdu Plain and the Qinghai-Tibet Plateau. There are 16 sharply transitioning counties in the Sichuan Basin, accounting for 9.4% of the Sichuan Basin.

[0043] Table 4 (3) Downscaling results based on random forest The spatial pattern of the township transition type is basically the same as that of the county, e.g. Figure 7 As shown in Table 5, the weak transition type dominated by socio-economic and natural ecological processes is represented by WDS and WDN, respectively. The number of WDS-type towns (sub-districts) is the highest in the entire region, with 2,571 towns covering an area of ​​1,072,800 square kilometers, accounting for 39.14% of the total area (Table 5). There are 1,357 slowly transitioning towns covering an area of ​​901,700 square kilometers, accounting for 32.9% of the total area of ​​the region. Rapidly transitioning towns tend to be distributed on the northern and eastern edges of the Sichuan Basin, while slowly transitioning towns are more common in the west. There are 438 such towns covering an area of ​​49,960 square kilometers (accounting for 18.23%). The number of sharply transitioning towns is relatively small. Due to the small area of ​​towns, it is difficult for environmental factors in the region to undergo sudden changes. There are only 24 sharply transitioning towns covering an area of ​​187,200 square kilometers. The geographical transition of WDN-type towns tends to be stable, with low intensity of human activities, and is mainly composed of high natural ecological components. WDN-type towns are distributed at the westernmost end of the basin, including 107 towns and 248,000 square kilometers, accounting for 9.05%, which is significantly higher than the proportion of counties of the same type.

[0044] Table 5 The model accuracy was 96.4%, and the confusion matrix of the training set is shown in Table 6. The weak transition type (WDS) dominated by socioeconomic processes had the highest accuracy, with only 2 of the 87 samples being incorrectly classified as the slow transition type. All six samples of the weak transition type (WDN) dominated by natural ecological processes were correct. Two samples each of the slow transition and the sharp transition were incorrectly classified as the fast transition type.

[0045] Table 6 The results show that the number of sharp transition types is small, accounting for less than 1% of the total. Townships are almost the smallest administrative units for evaluating geographic transitions. If the area of ​​the evaluation unit is too small, it is impossible to effectively detect mutation information.

[0046] The second aspect of the present invention provides: a system for identifying the scope and type of a multi-scale geographic transition zone, which is used to implement any of the above methods for identifying the scope and type of a multi-scale geographic transition zone, comprising: The framework building module is used to select the characterization factors of socio-economic attributes and natural ecological attributes, and to construct the indicator framework for identifying the human and natural components of the transition zone; The measurement module is used to improve the coupling model, nonlinearly superimpose the human and natural components, and measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem; The type classification module is used to characterize the proportion of human and natural components in the target area based on the positive and negative values ​​and numerical values ​​of the coupling index, identify the transition area of ​​the grid unit and perform type classification; The mutation detection module is used to identify the transition zone in the typical direction using the gradient mutation detection method, and to identify the coupling index mutation at the regional scale by combining the spatial mutation detection method, so as to determine the boundary and type of the geographical transition zone; The zoning module is used to train and predict the transition type at the target level based on the random forest intelligent classification method, and further downscale the geographical transition zone to the target administrative unit to obtain different types of coupled system zoning.

[0047] The third aspect of the present invention provides: a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are loaded and executed by a processor, any of the above-mentioned methods for identifying the scope and type of multi-scale geographic transition zones is implemented.

[0048] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art shall not deviate from the spirit and scope of the present invention, and shall be within the scope of protection of the claims attached to the present invention.

Claims

1. A method for identifying the scope and type of multi-scale geographic transition zones, characterized by: The following steps are involved: S1: Select the characterization factors of socio-economic attributes and natural ecological attributes, and construct the indicator framework for identifying the human and natural components of the transition zone; S2: Improve the coupling model, perform nonlinear superposition of human and natural components, and measure the direction and intensity of interaction between socio-economic systems and natural ecosystems; S3: Based on the positive and negative values ​​of the coupling index, the proportion of human and natural components in the target area is characterized, the transition zone of the grid unit is identified and classified; S4: Use the gradient mutation detection method to identify transition zones in typical directions, and combine it with the spatial mutation detection method to identify coupling index mutations at the regional scale, thereby determining the boundaries and types of geographical transition zones; S5: Based on the random forest intelligent classification method, the transition type at the target level is trained and predicted, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

2. The method for identifying the scope and type of multi-scale geographic transition zones according to claim 1, characterized in that: The S1 further comprises the following steps: Characterization factors of socio-economic attributes and natural ecological attributes are selected, settlement patch density and agricultural production functions are spatially expressed, and a grid-scale representation data set is obtained, thereby constructing an indicator framework for identifying the human and natural components of the transition zone.

3. The method for identifying the scope and type of multi-scale geographic transition zones according to claim 1, characterized in that: The S2 further comprises the following steps: The initial coupling model is obtained by integrating the physical coupling model and the management trade-off model. CSI , ， ； ； ， ； in, CP is the coupling degree, indicating n The coupling relationship between the systems; CD is the degree of coordination, which indicates the overall coordination level between systems; TO yes n The trade-offs between systems; b i It is a system i relative benefits; It is a system i expected benefits; It is a system i The importance of Improved coupling model obtained by derivation simplification ; , ;in, x for the relative benefits of the socio-economic system; y Relative benefits to natural ecosystems; using coupled models The human and natural components are further superimposed nonlinearly to measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem.

4. The method for identifying the scope and type of multi-scale geographic transition zones according to claim 1, characterized in that: The S3 further comprises the following steps: Based on the positive and negative values ​​and the size of the coupling index, the proportion of human and natural components in the target area is characterized, the coupling effect between the socio-economic system and the natural ecosystem is represented, the geographical transition zones of the grid units are identified and divided into types, and all traditional transition zones are obtained by adjusting the classification criteria of the coupling index.

5. The method for identifying the scope and type of multi-scale geographic transition zones according to claim 1, characterized in that: The S4 further comprises the following steps: Use the gradient mutation detection method to identify transition regions in typical directions. ,in t is the gradient sequence in mutation detection; t 1 is the horizontal coordinate of the starting point of the gradient; t 2 is the abscissa of the end point of the gradient; y 1 is the starting ordinate of the gradient; y 2 is the ordinate of the mutation point; y 3 is the ordinate of the end point of the gradient; T 1 is the abscissa of the mutation point; f is a piecewise function representing the gradient partition; region 1 is the function before the mutation point; region 2 is the function after the mutation point; if Function means if t <= T 1 Return if the condition is met region 1, if not established, return region 2; The spatial slope function is introduced to identify the coupling index mutation on the regional scale using the spatial mutation detection method. ;in, C _ S ij is the spatial slope of factor i at pixel j, and a larger value indicates a greater uncertainty in the environmental factor; dz / dx is the rate of change of the factor in the first-order neighborhood in the horizontal direction; dz / dy is the rate of change of the factor in the vertical direction; p m and p n denote the factor values ​​in the first-order neighborhood in the north-south and east-west directions, respectively; w s and w e Represents the horizontal weighted counts of valid pixels in the north-south and east-west directions, respectively; p size is the size of a pixel; Then, typical directional gradient county units are used to verify the accuracy of mutation detection in the entire space, thereby determining the boundaries and types of geographical transition zones.

6. The method for identifying the scope and type of multi-scale geographic transition zones according to claim 1, characterized in that: The S5 further comprises the following steps: Based on the random forest intelligent classification method, the environmental factors divided by the transition zone type and impact type of the county unit and the indicators for measuring the information amount of the environmental factors are used as training samples to predict the transition type of the target level, and the geographical transition zone is further downscaled to the target administrative unit to obtain different types of coupled system partitions.

7. The method for identifying the scope and type of multi-scale geographic transition zones according to any one of claims 1 to 6, characterized in that: The target level is the township level, and the target administrative unit is the township administrative unit.

8. A system for identifying the extent and type of multi-scale geographic transition zones, characterized by: The method for implementing the method of identifying the scope and type of a multi-scale geographic transition zone as described in any one of claims 1 to 7 comprises: The framework building module is used to select the characterization factors of socio-economic attributes and natural ecological attributes, and to construct the indicator framework for identifying the human and natural components of the transition zone; The measurement module is used to improve the coupling model, nonlinearly superimpose the human and natural components, and measure the direction and intensity of the interaction between the socio-economic system and the natural ecosystem; The type classification module is used to characterize the proportion of human and natural components in the target area based on the positive and negative values ​​and numerical values ​​of the coupling index, identify the transition area of ​​the grid unit and perform type classification; The mutation detection module is used to identify the transition zone in the typical direction using the gradient mutation detection method, and to identify the coupling index mutation at the regional scale by combining the spatial mutation detection method, so as to determine the boundary and type of the geographical transition zone; The zoning module is used to train and predict the transition type at the target level based on the random forest intelligent classification method, and further downscale the geographical transition zone to the target administrative unit to obtain different types of coupled system zoning.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are loaded and executed by the processor, the method for identifying the scope and type of multi-scale geographic transition zones as described in any one of claims 1-7 is implemented.

Images