Dynamic Evaluation Method of Landscape Ecological Security Based on Multi-Attribute Decision-Making Model

Abstract

The present invention discloses a dynamic evaluation method for landscape ecological security based on a multi-attribute decision-making model, including: adopting the gap degree landscape scale analysis method to obtain the spatial characteristic scales of each land use type, and using them as the characteristic grid scales for landscape pattern analysis; calculating the landscape pattern characteristic indices at each characteristic grid scale in different periods to construct a regional landscape ecological security evaluation model; using the Lagrange interpolation method to obtain the landscape ecological security indices within the continuous time range of each characteristic grid scale, and predicting the landscape ecological security index values of each characteristic grid scale in the future period; adopting the multi-attribute decision-making CoCoSo model to calculate the weighted sum and the sum of the exponential weighted products of landscape ecological security, and using different aggregation functions to obtain the dynamic index of landscape ecological security. Finally, the natural breakpoint method is used to conduct hierarchical and zonal management on the dynamic index of landscape ecological security in the research area. The present invention can provide technical support for the protection and restoration of regional landscape ecosystems.

Description

Technical Field

[0001] The present invention relates to the technical field of regional landscape ecosystem management and planning, and particularly to a dynamic evaluation method for landscape ecological security based on a multi-attribute decision-making model. Background Art

[0002] The development of social productivity and urbanization has increased the intensity of land use, disrupted the balance of land use resources, made ecological landscapes more fragmented, and significantly reduced their internal connectivity, leading to problems such as land desertification and loss of ecosystem functions, endangering regional ecological security and affecting the sustainable development of human social economy. Ecological security refers to the state in which the ecological environment can meet the needs of human sustainable development without destroying the potential of the natural environment. Comprehensive evaluation of ecological security can provide an effective means for ecological security management. The main methods for evaluating ecological security include: mathematical model method, ecological model method, digital ecological security model method, and landscape model method. Among them, landscape ecological security assessment quantifies the landscape pattern characteristic index to characterize the change trend of landscape structure and form, and further reflects the quality of the landscape ecosystem. This method is not only easy to carry out, but also can reflect the evolution process of the quality of large-area landscape ecosystems in the long term.

[0003] Currently, landscape ecological security assessment is mostly limited to constructing models through landscape disturbance and vulnerability indices to calculate landscape ecological sensitivity. However, as a dynamic system of the landscape, the ecosystem has its own adjustment mechanism, that is, the adaptability, which can usually be quantified by landscape diversity and evenness indices. The higher the diversity and evenness of the landscape, the stronger the ability to make adaptive adjustments after making a sensitive response to external disturbances. A single landscape sensitivity index cannot comprehensively reflect the complex impact of land use change on the landscape ecosystem, and the landscape ecological security should be comprehensively evaluated by combining landscape sensitivity and landscape adaptability. In addition, the current dynamic evaluation of regional landscape ecological security still evaluates the landscape ecological security indices of multiple periods separately and compares their spatio-temporal change characteristics. This method is essentially a static evaluation method and cannot describe the evolution process of the regional landscape ecosystem, and also ignores the interaction relationship between the landscape ecological security indices of each period. Summary of the Invention

[0004] To solve the problems existing in the prior art, the purpose of the present invention is to provide a dynamic evaluation method for landscape ecological security based on a multi-attribute decision-making model, which can provide a theoretical basis and technical support for the protection and restoration of regional landscape ecosystems.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is: a dynamic evaluation method for landscape ecological security based on a multi-attribute decision-making model, comprising the following steps:

[0006] Step 1: With the aid of fractal theory, the porosity index of each land use type in the study area at different spatial scales is obtained by using the porosity landscape scale analysis method, and the spatial characteristic scale is obtained and used as the characteristic grid scale for landscape pattern analysis;

[0007] Step 2: Based on the long-term land use type layer of the study area, the landscape pattern characteristic indexes at each characteristic grid scale in different periods are calculated, and a regional landscape ecological security assessment model is constructed from two perspectives of landscape sensitivity and landscape adaptability;

[0008] Step 3: Based on the landscape ecological security indexes at each characteristic grid scale in different periods, the landscape ecological security indexes within the continuous time range of each characteristic grid scale are obtained by using the Lagrange interpolation method, and the landscape ecological security index values of each characteristic grid scale in the future period are predicted by using the LSTM model;

[0009] Step 4: Based on the spatial distribution layer of the landscape ecological security indexes at each characteristic grid scale in different periods, considering the additive and multiplicative synergistic relationships between the landscape ecological security indexes in different periods, the multi-attribute decision-making CoCoSo model is used to calculate the weighted sum and the weighted product sum of landscape ecological security, and different aggregation functions are used to obtain the landscape ecological security dynamic index. Finally, the natural break point method is used to conduct hierarchical and zonal management on the landscape ecological security dynamic index of the study area.

[0010] As a further improvement of the present invention, in Step 2, the landscape pattern characteristic indexes include landscape fragmentation, landscape isolation, landscape richness, and landscape interference degree.

[0011] As a further improvement of the present invention, the calculation method of the landscape fragmentation is specifically as follows:

[0012] ;

[0013] Wherein, C ki is the landscape fragmentation, NP ki is the number of patches of the k th i landscape type in the grid, CA ki is the area of the k th i landscape type in the grid;

[0014] The calculation method of the landscape isolation is specifically as follows:

[0015] ;

[0016] Wherein, S kiis the landscape separation degree, TA k is the grid k the total area of various landscapes within;

[0017] The calculation method of the landscape richness is specifically as follows:

[0018] ;

[0019] Wherein, D ki is the landscape richness, m ki is the total number of patches of the k th type of landscape within the statistical cell i ; M k is the total number of patches within the statistical cell k ; TN k is the total number of patches within the statistical cell k ;

[0020] The calculation method of the landscape interference degree is specifically as follows:

[0021] ;

[0022] Wherein, U ki is the landscape interference degree, a, b, c respectively represent the weights of landscape fragmentation C ki , separation degree S ki , dominance D ki .

[0023] As a further improvement of the present invention, in step 2, the regional landscape ecological security assessment model is constructed from the two perspectives of landscape sensitivity and landscape adaptability as follows:

[0024] The landscape sensitivity index is calculated according to the following formula:

[0025] ;

[0026] In the formula, LSI k represents the landscape sensitivity index in the grid k ; U ki represents the landscape interference degree of the k th land use type in the grid i ; V ki represents the k th in the gridi The landscape vulnerability of a land use type;

[0027] The calculation formula of the landscape adaptability index is as follows:

[0028] ;

[0029] In the formula, PRD k is the patch richness density in the grid k ; SHDI k represents the diversity degree of the landscape in the grid k ; SHEI k represents the Shannon evenness index in the grid k ;

[0030] The calculation formula of the landscape ecological security index that comprehensively considers landscape sensitivity and landscape suitability is as follows:

[0031] ;

[0032] In the formula, LESI k represents LES in grid k; LSI k represents the landscape sensitivity index in the grid k ; A ki represents the area of the k th i land use type in grid A k represents the total area of all land use types in the grid k ; LAI k represents the landscape adaptability index in the grid k ;

[0033] As a further improvement of the present invention, in step 4, the multi-attribute decision-making CoCoSo model is a comprehensive index weighted product based on the simple weighted model, and its calculation formula is specifically as follows:

[0034] ;

[0035] ;

[0036] In the formula, represents the number of different periods, that is, the number of year data used for calculation; is the landscape ecological security index of the rd year of the is the The weight of the annual landscape ecological security index is the weighted sum of the variable values; is the sum of the weighted products of the landscape ecological security indices;

[0037] Apply the aggregation method to and respectively for summarization:

[0038] ;

[0039] ;

[0040] ;

[0041] ;

[0042] In the formula: represents the number of evaluation units, that is, the number of grid cells; represents and the arithmetic mean of the sum; represents and the sum of the quantification of landscape ecological security relative to the best case; is the balance compromise score for and , where the parameter represents the preference for or , that is the dynamic index of landscape ecological security of grid cells, and the spatio-temporal analysis characteristics of the dynamic index of landscape ecological security in the study area can be analyzed according to the value.

[0043] The present invention couples the landscape feature scale analysis - landscape ecological security assessment model construction - landscape ecological security prediction, quantitatively evaluates the landscape ecological security in different periods of the region from the grid scale. On this basis, the multi-attribute decision-making CoCoSo model is used to obtain the dynamic evaluation index of landscape ecological security, and it is classified and partitioned, providing a theoretical basis and technical support for the protection and restoration of the regional landscape ecological system.

[0044] The beneficial effects of the present invention are:

[0045] 1. The present invention constructs a regional landscape ecological security assessment model from two perspectives of landscape sensitivity and landscape adaptability, taking into account the degree of influence of nature and human activities on the landscape, the sensitivity of landscape types to external disturbances, and the adaptation and recovery ability under current external disturbances, which can more comprehensively reflect the impact of land use change on the landscape ecosystem, make up for the shortcoming of traditional landscape ecological security that only considers landscape sensitivity, and provide method support for accurately assessing landscape ecological security;

[0046] 2. Based on the multi-attribute decision-making model, the landscape ecological security dynamic assessment method constructed by the present invention takes into account the additive and multiplicative synergistic relationship between landscape ecological security indices in different periods of long time series, can better reflect the dynamic evolution process of the landscape ecosystem, make up for the shortcoming of traditional static assessment of landscape ecological security, and can provide a more reasonable method support for regional landscape ecological security dynamic assessment;

[0047] 3. The present invention provides a new method for reflecting the spatial distribution of the landscape ecological security dynamic assessment index under specific landscape management, classifies and partitions the comprehensive index, identifies the protected area and restoration area of the regional landscape ecological security dynamic index, and improves the management efficiency of the regional landscape ecosystem. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flowchart of an embodiment of the present invention;

[0049] Figure 2(a) is a porosity index distribution map of agricultural land in the embodiment of the present invention;

[0050] Figure 2(b) is a porosity index distribution map of forest land in the embodiment of the present invention;

[0051] Figure 2(c) is a porosity index distribution map of grassland in the embodiment of the present invention;

[0052] Figure 2(d) is a porosity index distribution map of water body in the embodiment of the present invention;

[0053] Figure 2(e) is a porosity index distribution map of construction land in the embodiment of the present invention;

[0054] Figure 2(f) is a porosity index distribution map of unused land in the embodiment of the present invention;

[0055] Figure 3 is a spatial distribution map of landscape sensitivity index at grid scale in the embodiment of the present invention;

[0056] Figure 4 is a spatial distribution map of landscape adaptability index at grid scale in the embodiment of the present invention;

[0057] Figure 5 This is the spatial distribution map of the landscape ecological security index at the grid scale in the embodiment of the present invention;

[0058] Figure 6 This is the predicted spatial distribution map of the landscape ecological security index at the grid scale in the embodiment of the present invention;

[0059] Figure 7 This is the classification and zoning map of the landscape ecological security dynamic index at the grid scale in the embodiment of the present invention. Detailed implementation manners

[0060] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0061] Embodiment 1

[0062] As Figure 1 shown, a dynamic assessment method for landscape ecological security based on a multi-attribute decision-making model. This method starts from the synergistic relationship between addition and multiplication among landscape ecological security indices in different periods of a long time series, and is a new method for dynamic assessment of regional landscape ecological security combining gap degree landscape scale analysis, landscape ecological security assessment model, Lagrange interpolation, LSTM prediction model, multi-attribute decision-making CoCoSo model and spatial analysis technology. The method includes the following steps:

[0063] S1. With the help of fractal theory, the gap degree index of each land use type in the study area at different spatial scales is obtained by using the gap degree landscape scale analysis method, and the spatial characteristic scale is obtained, that is, at this spatial characteristic scale of the land use type, the spatial distribution tends to be homogeneous, and it is used as the characteristic grid scale for landscape pattern analysis and also the basic calculation unit for landscape ecological security evaluation;

[0064] S2. Based on the long time series land use type layers in the study area, using Fragstats landscape index analysis software, calculate the landscape pattern characteristic indices (including landscape fragmentation degree, landscape separation degree, landscape richness, landscape interference degree) at each grid scale in different periods, and construct a regional landscape ecological security assessment model from two perspectives of landscape sensitivity and landscape adaptability;

[0065] S3. Based on the landscape ecological security indices of each grid in different periods, use the Lagrange interpolation method to obtain the landscape ecological security index within the continuous time range of each grid, and use the LSTM model to predict the landscape ecological security index value of each grid in the future period;

[0066] S4. To reflect the dynamic change process of regional landscape ecological security, based on the spatial distribution layers of landscape ecological security index at different grid scales in different periods, considering the additive and multiplicative synergy relationships between landscape ecological security indices in different periods, the multi-attribute decision-making CoCoSo model is used to calculate the weighted sum and the weighted product sum of landscape ecological security, and different aggregation functions are adopted to obtain the landscape ecological security dynamic index. Finally, the natural breakpoint method is used to conduct hierarchical and zonal management of the landscape ecological security dynamic index in the study area;

[0067] In the calculation of the landscape ecological security index for each grid in step S2, the landscape disturbance degree characterizes the heterogeneity and diversity characteristics of the landscape under natural and human disturbances, indicating the degree of influence of natural and human activities on the landscape, and is related to landscape fragmentation, isolation, and dominance. Landscape fragmentation refers to the degree of fragmentation of landscape types at a specific time and with specific properties. The greater the degree of adverse human disturbance, the more fragmented the landscape type. Landscape isolation reflects the spatio-temporal distribution characteristics of the interference of human-natural factors on the landscape. Landscape dominance is a sensitive indicator for analyzing the changes in diversity and heterogeneity of different landscape types in different periods, and is related to the coverage, density, and frequency of landscape types. In the process of facing natural-human disturbances, the sensitivity and anti-interference ability of various landscape types show differences, that is, the landscape has vulnerability and landscape adaptability. Landscape vulnerability represents the magnitude of the sensitivity of landscape types to external disturbances. The larger the value, the more sensitive to external disturbances. Landscape adaptability is the ability of the landscape to adapt and recover under external disturbances, and is closely related to the structure, function, diversity, and distribution uniformity of the landscape. The larger the value, the stronger the ability to adapt and recover to external disturbances. Finally, a regional landscape ecological security assessment model is constructed from the two perspectives of landscape sensitivity and landscape adaptability.

[0068] The calculation formulas of different landscape pattern characteristic indices and landscape disturbance degree are shown in Table 1:

[0069] Table 1 Related calculation descriptions of landscape pattern characteristic indices and landscape disturbance degree

[0070]

[0071] The landscape sensitivity index can be calculated according to the following formula:

[0072] ;

[0073] In the formula, LSI k represents the landscape sensitivity index in grid k ; U ki represents the landscape disturbance degree of the k th land use type in grid i ; Vki Represents the grid k in the i land use type of the

[0074] landscape vulnerability, as shown in Table 2:

[0075]

[0076] The calculation formula of the landscape adaptability index is as follows:

[0077] ;

[0078] In the formula, where PRD k is the patch richness density in the grid k ; SHDI k represents the diversity degree of the landscape in the grid k ; SHEI k represents the Shannon's Evenness Index in the grid k . A high SHEI k value indicates that there is no obvious dominant landscape class, and each type of landscape is evenly distributed in the grid k .

[0079] The calculation formula of the landscape ecological security index that comprehensively considers landscape sensitivity and landscape suitability is as follows:

[0080] ;

[0081] In the formula, LESI k represents the k in the grid LES ; LSI k represents the landscape sensitivity index in the grid k ; A ki represents the area of the k in the grid i land use type; A k represents the total area of all land use types in the grid k ; LAI k represents the landscape adaptability index in the grid k .

[0082] In the calculation of the landscape ecological security dynamic index in step S4, to reflect the dynamic change process of regional landscape ecological security, the CoCoSo model is used to aggregate the spatial distribution layers of landscape ecological security indices at different grid scales in different periods obtained from the analysis of landscape scale based on porosity - calculation of landscape ecological security index - prediction of future landscape ecological security index, and the landscape ecological security dynamic index of the study area is calculated. CoCoSo synthesizes the exponentially weighting product on the basis of the simple additive weighting model and is a multi - attribute decision - making method that can provide a compromise solution for different aggregation functions. Its calculation formula is: ;

[0083] ;

[0084] In the formula, represents the number of different periods, that is, the number of year data used for calculation; is the landscape ecological security index of the th grid in the th year; is the weight of the landscape ecological security index in the th year. Generally, it is considered that the landscape ecological security indices in each period are of equal importance, so equal weights are assigned to them, that is = 1 / n . is the weighted sum of variable values; is the sum of the exponentially weighted products of the landscape ecological security index. Based on this, aggregation methods are respectively applied to summarize and .

[0085] ;

[0086] ;

[0087] ;

[0088] ;

[0089] In the formula: represents the number of evaluation units, that is, the number of grids; represents and the arithmetic mean of the sum; represents and the sum of the quantification of landscape ecological security relative to the best case; is then about and The balanced compromise score, where the parameter represents the preference for or Generally, there is no preference for and Therefore, is set to 0.5. That is The dynamic index of landscape ecological security of grid cells, which can analyze the spatio-temporal analysis characteristics of the dynamic index of landscape ecological security in the study area according to the value.

[0090] This embodiment proposes to couple the landscape feature scale analysis - landscape ecological security static assessment - landscape ecological security prediction and multi-attribute decision-making CoCoSo model, so that the dynamic index of landscape ecological security under the characteristic calculation grid cells can more accurately reflect the interaction relationship between the landscape ecological security indices in different periods of the long time series, as well as the dynamic evolution process of the landscape ecosystem. By integrating landscape ecology, geoscience, management science and environmental science, it can accurately evaluate the dynamic index of regional landscape ecological security, provide method support for improving the quality and function of the regional landscape ecosystem, and provide data support for realizing the hierarchical and zonal management of the dynamic index of regional landscape ecological security.

[0091] Embodiment 2

[0092] This embodiment will be further described in detail in combination with a certain basin. As Figure 1 shown, a method for dynamically evaluating landscape ecological security based on a multi-attribute decision-making model is a new method for dynamically evaluating landscape ecological security that combines gap degree landscape scale analysis, landscape ecological security assessment model, Lagrangian interpolation, LSTM prediction model, multi-attribute decision-making CoCoSo model and spatial analysis technology. The specific contents are as follows:

[0093] S1. With the help of fractal theory, using the gap degree landscape scale analysis method, obtain the gap degree indices of 6 land use types in the basin (30 m × 30 m in this embodiment) at different spatial scales, as shown in Figures 2(a) - 2(f), and obtain the spatial characteristic scale, that is, at this spatial characteristic scale, the spatial distribution of land use types tends to be homogeneous, and it is used as the characteristic grid scale for landscape pattern analysis and also the basic calculation unit for landscape ecological security evaluation in this area (3 km × 3 km in this embodiment);

[0094] S2. Based on the land use type layers of the basin for long time series (in 1980, 1990, 1995, 2000, 2005, 2008, 2010, 2015, 2018, 2020), using the Fragstats landscape index analysis software, calculate the landscape pattern characteristic indices (including landscape fragmentation degree, landscape separation degree, landscape richness, landscape interference degree) at each grid scale in different previous periods, and construct a regional landscape ecological security assessment model from two perspectives of landscape sensitivity and landscape adaptability. The spatial distribution maps of the landscape sensitivity index and the landscape adaptability index are respectively as Figure 3 and Figure 4 shown;

[0095] S3. Based on the landscape ecological security indices of each grid in the above different periods, use the Lagrange interpolation method to obtain the landscape ecological security indices within the continuous time range for each grid as Figure 5 shown, and use the LSTM model to predict the landscape ecological security index values of each grid in the future periods (set as 2025 and 2030 in this embodiment), as Figure 6 shown;

[0096] S4. To reflect the dynamic change process of the landscape ecological security of the basin, based on the spatial distribution layers of the landscape ecological security indices at each grid scale in different periods, considering the additive and multiplicative synergistic relationship between the landscape ecological security indices in different periods (1980 - 2025, 1980 - 2030), use the multi-attribute decision-making CoCoSo model to calculate the weighted sum and the index weighted product sum of the landscape ecological security, and use different aggregation functions to obtain the landscape ecological security dynamic index, as Figure 7 shown, and finally use the natural breakpoint method to conduct hierarchical and zonal management on the landscape ecological security dynamic index of the basin;

[0097] In the calculation of the landscape ecological security index for each grid in the basin in step S2, the landscape disturbance degree characterizes the heterogeneity and diversity characteristics of the landscape under natural and human disturbances, indicating the degree of influence of natural and human activities on the landscape, and is related to landscape fragmentation, isolation, and dominance. Landscape fragmentation refers to the degree of fragmentation of a landscape type at a specific time and with specific properties. The greater the degree of adverse human disturbance, the more fragmented the landscape type. Landscape isolation reflects the spatio-temporal distribution characteristics of human-natural factors' interference with the landscape. Landscape dominance is a sensitive indicator for analyzing the changes in diversity and heterogeneity of different landscape types at different times, and is related to the coverage, density, and frequency of landscape types. During the natural-human disturbance process, the sensitivity and anti-interference ability of various landscape types show relatively significant differences, that is, the landscape has vulnerability and landscape adaptability. Landscape vulnerability indicates the degree of sensitivity of a landscape type to external disturbances. The larger the value, the more sensitive it is to external disturbances. Landscape adaptability is the ability of a landscape to adapt and recover under external disturbances, and is closely related to the structure, function, diversity, and distribution uniformity of the landscape. The larger the value, the stronger the ability to adapt and recover to external disturbances. Finally, a regional landscape ecological security assessment model is constructed from the two perspectives of landscape sensitivity and landscape adaptability.

[0098] The landscape sensitivity index can be calculated by the following formula:

[0099] ;

[0100] In the formula, LSI k represents the landscape sensitivity index in grid k ; U ki represents the landscape disturbance degree of the k th land use type in grid i ; V ki represents the landscape vulnerability of the k th land use type in grid i ;

[0101] The calculation formula for the landscape adaptability index is as follows:

[0102] ;

[0103] In the formula, where PRD k is the patch richness density in grid k ; SHDI k represents the diversity degree of the landscape in grid k ; SHEI k represents grid kThe Shannon's Evenness Index, a high SHEI k value indicates that there is no obvious dominant landscape class, and each type of landscape is evenly distributed in the grid k .

[0104] The calculation formula of the landscape ecological security index that comprehensively considers landscape sensitivity and landscape suitability is as follows:

[0105] ;

[0106] In the formula, LESI k represents k in the grid LES ; LSI k represents the landscape sensitivity index in the grid k ; A ki represents k in the grid, the i th land use type area; A k represents k in the grid, the total area of all land use types; LAI k represents the landscape adaptability index in the grid k.

[0107] In the calculation of the landscape ecological security dynamic index in step S4, to reflect the dynamic change process of the basin landscape ecological security, the CoCoSo model is used to aggregate the spatial distribution layers of the landscape ecological security index at the grid scale for different periods (1980 - 2025, 1980 - 2030) of the long time series obtained from the porosity landscape scale analysis - landscape ecological security index calculation - future landscape ecological security index prediction, and calculate the basin landscape ecological security dynamic index. CoCoSo combines the Exponentially Weighting Product on the basis of the Simple Additive Weighting model, which is a multi-attribute decision-making method that can provide a compromise solution for different aggregation functions. Its calculation formula is:

[0108] ;

[0109] ;

[0110] In the formula, represents the number of different periods, that is, the number of year data used for calculation; is the th grid's Annual landscape ecological security index; is the weight of the annual landscape ecological security index. Generally, the landscape ecological security indices in each period are considered equally important, so equal weights are assigned to them, that is = 1 / n . is the weighted sum of variable values; is the sum of the exponential weighted products of landscape ecological security. Based on this, the aggregation method is respectively applied to and for summarization.

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] In the formula: represents the number of evaluation units, that is, the number of grids; represents and the arithmetic mean of the sum; represents and the sum of the quantification of landscape ecological security relative to the best case; is the balance compromise score for and , where the parameter represents the preference for or . Generally, there is no preference for and , so is set to 0.5. That is the dynamic index of landscape ecological security of grid cells, and the spatio-temporal analysis characteristics of the dynamic index of landscape ecological security in the study area can be analyzed according to the value.

[0116] The above-described embodiments only express the specific implementation manners of the present invention, and the description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A dynamic assessment method for landscape ecological security based on a multi-attribute decision-making model, characterized in that: The following steps are involved: Step 1: With the help of fractal theory, the gap index of each land use type in the study area at different spatial scales is obtained by using the gap landscape scale analysis method, and the spatial characteristic scale is obtained, which is used as the characteristic grid scale for landscape pattern analysis; Step 2: Based on the long-term land use type layer of the study area, calculate the landscape pattern characteristic index at each characteristic grid scale in different periods, and construct a regional landscape ecological security assessment model from the perspectives of landscape sensitivity and landscape adaptability; In step 2, the regional landscape ecological security assessment model is constructed from the two perspectives of landscape sensitivity and landscape adaptability as follows: The landscape sensitivity index is calculated as follows: LSI k =U ki ×V ki In the formula, LSI k represents the landscape sensitivity index in grid k; U ki V represents the landscape disturbance degree of the i-th land use type in grid k; ki represents the landscape vulnerability of the i-th land use type in grid k; The calculation formula of landscape adaptability index is as follows: LAI k =PRD k ×SHDI k ×SHEI k Where PRD k is the patch richness density in grid k; SHDI k Represents the diversity of the landscape in grid k; SHEI k represents the Shannon uniformity index in grid k; The calculation formula of the landscape ecological security index that integrates landscape sensitivity and landscape adaptability is as follows: In the formula, LESI k represents the LES in grid k; LSI k represents the landscape sensitivity index in grid k; A ki A represents the area of ​​the i-th land use type in grid k; k represents the total area of ​​all land use types in grid k; LAI k represents the landscape suitability index in grid k; Step 3: Based on the landscape ecological security index of each characteristic grid scale at different periods, the Lagrange interpolation method is used to obtain the landscape ecological security index of each characteristic grid scale within a continuous time range, and the LSTM model is used to predict the landscape ecological security index value of each characteristic grid scale in the future period; Step 4: Based on the spatial distribution layers of landscape ecological security index at each characteristic grid scale in different periods, considering the additive and multiplicative synergistic relationship between landscape ecological security indices in different periods, the multi-attribute decision-making CoCoSo model is used to calculate the weighted sum and exponential weighted product sum of landscape ecological security, and different aggregation functions are used to obtain the dynamic index of landscape ecological security. Finally, the natural breakpoint method is used to carry out hierarchical and regional management of the dynamic index of landscape ecological security in the study area.

2. The method for dynamic assessment of landscape ecological security based on a multi-attribute decision-making model according to claim 1 is characterized in that: In step 2, the landscape pattern characteristic index includes landscape fragmentation, landscape separation, landscape richness and landscape disturbance.

3. The method for dynamic assessment of landscape ecological security based on a multi-attribute decision-making model according to claim 2 is characterized in that: The calculation method of the landscape fragmentation is as follows: Among them, C ki is landscape fragmentation, NP ki is the number of patches of the i-th landscape type in grid k, CA ki is the area of ​​the i-th landscape type in grid k; The calculation method of the landscape separation degree is as follows: Among them, S ki is the landscape separation, TA k is the total area of ​​all types of landscapes in grid k; The calculation method of the landscape richness is as follows: Among them, D ki is the landscape richness, m ki is the total number of patches of the i-th type of landscape in the statistical cell k, M k is the total number of plaques in cell k, TN k is the total number of patches in statistical cell k; The calculation method of the landscape disturbance degree is as follows: U ki =aC ki +bS ki +cD ki Among them, U ki is the degree of landscape disturbance, a, b, c represent the degree of landscape fragmentation C ki , Separation S ki , Richness D ki The weight of .

4. The method for dynamic assessment of landscape ecological security based on a multi-attribute decision-making model according to claim 1 is characterized in that: In step 4, the multi-attribute decision-making CoCoSo model is a comprehensive exponential weighted product based on a simple weighted model, and its calculation formula is as follows: In the formula, n represents the number of different periods, that is, the number of years of data used for calculation; r ij is the landscape ecological security index of the i-th grid in the j-th year; w j is the weight of the landscape ecological security index in the jth year, S i is the weighted sum of the variable values; P i It is the sum of weighted products of landscape ecological security indexes; Apply aggregation methods to S i and P i To summarize: Where: m represents the number of evaluated units, that is, the number of grids; k ia Represents S i With P i arithmetic mean of the sum; k ib Represents S i With P i The sum of the quantified landscape ecological security relative to the best case; k ic Then about S i and P i The balanced compromise score of S i or P i Preference, k i That is, the landscape ecological security dynamic index of grid unit i, according to k i The spatiotemporal analysis characteristics of the dynamic index of regional landscape ecological security were studied by value analysis.

Images