A simulation prediction method for the spread of a sedge marsh tussock community

Abstract

This invention relates to a method for simulating and predicting the spread of *Carex chinensis* mound communities in marsh wetlands. The invention belongs to the field of marsh wetland plant community spread simulation technology, specifically a method for simulating and predicting the spread of *Carex chinensis* mound communities. This invention addresses the problem of not being able to simulate and predict the spread paths and rates of *Carex chinensis* mound communities in marsh wetlands using existing ecological and environmental data. The method involves: acquiring and interpreting remote sensing images; acquiring community characteristic data and ecological and environmental data; processing the acquired remote sensing images and ecological and environmental data to form a database; creating an MCE suitability atlas; simulating the spread of *Carex chinensis* mound communities; verifying the accuracy of the simulation results; and then making predictions. This invention can accurately predict the spread paths and rates of *Carex chinensis* mound communities in marsh wetlands and can be applied to wetland restoration and management.

Description

Technical Field

[0001] This invention belongs to the field of swamp wetland plant community diffusion simulation technology, and relates to a method for simulating and predicting the diffusion of *Carex chinensis* mound communities in swamp wetlands under different future ecological scenarios based on remote sensing images and field surveys. Background Technology

[0002] Caragana is an important wetland vegetation type, playing an increasingly important role in wetland carbon sequestration, biodiversity maintenance, and providing habitats for waterbirds. Caragana has strong reproductive capacity and environmental adaptability, and is often widely used in the restoration of degraded marsh vegetation in Northeast China.

[0003] Under natural restoration conditions, *Carex spp.* mounds primarily reproduce sexually through seeds via wind, hydrology, and bird transport, but can also reproduce asexually. In the early stages of artificial transplantation and restoration, *Carex spp.* mounds mainly spread through tillering, producing new offshoots with the same genetic structure, thus continuously dispersing the plant in its habitat. The dispersal of *Carex spp.* mounds in marsh wetlands is a complex ecological process; the dispersal pathways and rates are difficult to accurately simulate using only mathematical methods such as Markov chains. Current research on the dispersal of *Carex spp.* mound communities in marsh wetlands is limited, and how to utilize technological means to simulate and predict the dispersal process of *Carex spp.* mound communities in marsh wetlands remains an unsolved problem. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problem that it is impossible to simulate and predict the diffusion path and rate of *Carex spp.* community in marsh wetlands using existing ecological and environmental data, and to provide a method for simulating and predicting the diffusion of *Carex spp.* community in marsh wetlands.

[0005] A simulation and prediction method for the spread of Caragana korshinskii community in marsh wetlands is carried out according to the following steps:

[0006] Step 1: Obtain remote sensing images of the distribution area of ​​*Carex chinensis* grass mound community in marsh wetlands at different historical periods and interpret the remote sensing image data;

[0007] Step 2: Conduct field investigations and monitoring to obtain data on the characteristics of the *Carex chinensis* grass mound community in marsh wetlands and the ecological environment data of the *Carex chinensis* grass mound community distribution area;

[0008] Step 3: Process the data obtained in Step 1 and Step 2 to form a database of the characteristics and ecological environment of the *Carex chinensis* grass mound community;

[0009] Redundancy analysis was performed on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the distribution area of ​​the *Carex chinensis* mound community collected in step two to screen out the environmental factors that affect the spread of the *Carex chinensis* mound community.

[0010] Spatial interpolation was performed using GIS technology on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the *Carex chinensis* mound community distribution area collected in step two to clarify the image threshold range.

[0011] Step 4: Overlay the acquired image data using the MCE multi-criteria evaluation method to form an MCE suitability atlas;

[0012] Step 5: Use the CA-Markov model to simulate and predict the spread of the *Carex chinensis* mound community, and verify the simulation accuracy. After the simulation results pass the accuracy verification, predict the spread of the *Carex chinensis* mound community under different ecological scenarios in the future.

[0013] The remote sensing images from different historical periods mentioned in Step 1 are high-resolution images with a spatial resolution better than 2 meters and an image time interval of 5 years. In the remote sensing interpretation process, object-oriented interpretation is adopted, supplemented by field surveys to verify point data.

[0014] Step two involves obtaining characteristic data of the *Carex spp.* community in marsh wetlands and ecological environment data of the distribution area of ​​the *Carex spp.* community, including characteristic data of the *Carex spp.* community, hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, landscape pattern data, and dispersal barrier data.

[0015] The data on the characteristics of the *Caragana korshinskii* mound community were collected during the growing season of the year using the quadrat survey method, mainly collecting data on the growth status of *Caragana korshinskii*, the morphological characteristics of the mounds, and the species diversity of the mound community.

[0016] The hydrological data includes water depth data;

[0017] The soil data includes soil moisture content, soil bulk density, and soil pH value.

[0018] The terrain data includes DEM and slope data;

[0019] The temperature data includes soil temperature;

[0020] The seed bank data includes seed dispersal methods and dispersal paths;

[0021] The interspecific competition data includes the competition or invasion of other species in the distribution area of ​​*Caragana korshinskii*.

[0022] The landscape pattern data and diffusion barrier data are obtained from the remote sensing image data interpreted in step one;

[0023] The seed diffusion path and diffusion barrier data are normalized, and the distance from each pixel to the image factor data is calculated.

[0024] Step three involves processing the data acquired in step one, specifically unifying the remote sensing images from different historical periods interpreted in step one into image data of the same size, projection, and resolution, performing Markov operations to form a Markov data chain, and determining the transition probability P. ij The transferred area file is obtained, and the formula is as follows:

[0025] (Formula 1)

[0026] (Formula 2)

[0027] In the formula, S(t) and S(t+1) represent the diffusion state of the *Caragana korshinskii* community at times t and t+1, respectively.

[0028] In step four, the MCE multi-criteria evaluation method is used to perform layer standardization processing on the acquired hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, and dispersal barrier data using fuzzy geometric functions. After standardization, the layer can distribute the factor conditions in a continuous range of 0-255, where the closer to 0 is to unsuitable conditions, and the closer to 255 is to optimal conditions.

[0029] Step five uses the CA-Markov model to simulate and predict the spread of the *Caragana korshinskii* community, simulating the spread process of the *Caragana korshinskii* community. The mathematical expression is:

[0030] (Formula 3)

[0031] In the formula, i and j are the row and column numbers of the cell, respectively, and d represents the dimension of the cell space. The diffusion of the *Caragana korshinskii* community refers to its diffusion from the ground to the surrounding area, which is a two-dimensional diffusion. , Let i and j represent the states of cells (i, j) in the *Caragana korshinskii* community at times t+1 and t, respectively. Represents the cell transformation rule, Let n be the state function of the neighboring cell (i, j) at time t, specifically representing the density of *Caragana korshinskii* community cells within an n×n neighborhood centered on cell (i, j), where n is the density of the neighboring cell (i, j). 2 Indicates the number of cells in the neighborhood. The formula for representing is as follows:

[0032] (Formula 4)

[0033] In the formula, con(S) ij =Carex) is a conditional expression indicating that the cell is affected by its neighborhood;

[0034] The CA-Markov model combines the quantitative simulation capabilities of the Markov model with the spatial simulation capabilities of the CA model. The transformation rules in the CA-Markov model... The formula for representing is as follows:

[0035] (Formula 5)

[0036] In the formula The transformation rule for each cell is composed of the number simulation of the Markov chain, the CA space simulation, and the fitness and restriction factors in the MCE.

[0037] A 5×5 filter was selected, and the iteration time y was equal to the image year interval for forming the Markov data chain, which can predict the spread of the *Carex chinensis* mound community in the next period.

[0038] By adjusting the Markov chain transition probability and the fitness and restriction factors in the MCE to set different ecological scenarios, the future spread of *Caragana korshinskii* mound communities under different ecological scenarios was simulated and predicted.

[0039] The accuracy of the simulation is verified by evaluating the simulation results using a pixel-by-pixel comparison method. The evaluation includes a comparison between the Kappa and FOM methods, as shown in the following formula:

[0040] (Formula 6)

[0041] Kappa measures the agreement between simulation results and actual results, where P0 is the proportion of land use distribution that is correctly simulated; P c To simulate the correct proportions under random conditions; P p To simulate a perfectly accurate scale under ideal conditions;

[0042] (Formula 7)

[0043] FOM measures the ability of the CA-Markov model to capture the dispersal of *Caragana korshinskii* mound communities, where UR... hit This indicates the diffusion cells of the *Caragana korshinskii* community correctly captured by the CA-Markov model, UR miss UR represents the diffusion cells of the *Caragana korshinskii* community that were missed by the CA-Markov model. false This represents the *Caragana korshinskii* community cells that were incorrectly captured by the CA-Markov model.

[0044] The CA-Markov model is a simulation method that, based on mathematical simulation, can represent the complex spatial changes of a system. It is suitable for studying the dynamic changes of plant communities. The main advantage of the CA-Markov model is that it is a spatial diffusion model based on diffusion mechanisms, capable of dynamically simulating the diffusion process of *Carex spp.* mound communities and predicting spatial distribution patterns. In the process of using *Carex spp.* for the restoration of marsh wetlands, it can accurately simulate the spatial dynamic diffusion process of *Carex spp.* mound communities based on the ecological environment. This not only provides data support for the study of *Carex spp.* mound community diffusion patterns but also provides a scientific reference for wetland restoration management.

[0045] The advantages of this invention are:

[0046] This invention combines ecological environment data from field surveys with CA-Markov model predictions to overcome the spatial limitations of purely mathematical simulations. It also considers practical factors such as seed bank dispersal and vegetation spread barriers, enabling a more accurate and objective simulation and prediction of the spatial dynamics of *Carex spp.* community dispersal in marsh wetlands. Furthermore, it utilizes the CA-Markov model to establish simulation methods for *Carex spp.* community dispersal under different ecological scenarios, providing a scientific approach for simulating *Carex spp.* community dispersal in marsh wetlands and serving as a reference for marsh wetland restoration and management. Attached Figure Description

[0047] Figure 1 This is a schematic flowchart of the method of the present invention;

[0048] Figure 2 This is a technical roadmap of the method of the present invention;

[0049] Figure 3 This is a 2013 remote sensing image of the *Caragana korshinskii* restoration area in Sun Island Scenic Area, Harbin, from Experiment 1.

[0050] Figure 4 This is a land cover classification map after interpreting the 2013 remote sensing image of the sedge restoration area in Sun Island Scenic Area, Harbin City, in Experiment 1.

[0051] Figure 5 This is a 2018 remote sensing image of the *Caragana korshinskii* restoration area in Sun Island Scenic Area, Harbin, from Experiment 1.

[0052] Figure 6 This is a land cover classification map after interpreting the 2018 remote sensing image of the sedge restoration area in Sun Island Scenic Area, Harbin City, in Experiment 1.

[0053] Figure 7 This is an ecological environment database atlas of the *Caragana korshinskii* restoration area in Sun Island Scenic Area, Harbin City, from Experiment 1;

[0054] Figure 8This is an atlas of MCE suitability maps for the diffusion of *Carex* community in the restoration area of ​​Sun Island Scenic Area, Harbin City, in Experiment 1.

[0055] Figure 9 This is a simulation diagram of the natural diffusion of the *Carex* community in the restoration area of ​​Sun Island Scenic Area, Harbin City in 2023, from Experiment 1. Detailed Implementation

[0056] The technical solution of the present invention is not limited to the specific embodiments listed below, but also includes any combination of the specific embodiments.

[0057] Specific Implementation Method 1: This implementation method for simulating and predicting the spread of *Carex spp.* grass mound communities in marsh wetlands follows these steps:

[0058] Step 1: Obtain remote sensing images of the distribution area of ​​*Carex chinensis* grass mound community in marsh wetlands at different historical periods and interpret the remote sensing image data;

[0059] Step 2: Conduct field investigations and monitoring to obtain data on the characteristics of the *Carex chinensis* grass mound community in marsh wetlands and the ecological environment data of the *Carex chinensis* grass mound community distribution area;

[0060] Step 3: Process the data obtained in Step 1 and Step 2 to form a database of the characteristics and ecological environment of the *Carex chinensis* grass mound community;

[0061] Redundancy analysis was performed on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the distribution area of ​​the *Carex chinensis* mound community collected in step two to screen out the environmental factors that affect the spread of the *Carex chinensis* mound community.

[0062] Spatial interpolation was performed using GIS technology on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the *Carex chinensis* mound community distribution area collected in step two to clarify the image threshold range.

[0063] Step 4: Overlay the acquired image data using the MCE multi-criteria evaluation method to form an MCE suitability atlas;

[0064] Step 5: Use the CA-Markov model to simulate and predict the spread of the *Carex chinensis* mound community, and verify the simulation accuracy. After the simulation results pass the accuracy verification, predict the spread of the *Carex chinensis* mound community under different ecological scenarios in the future.

[0065] Specific Implementation Method Two: This implementation method differs from Specific Implementation Method One in that the remote sensing images from different historical periods mentioned in step one are high-resolution images with a spatial resolution better than 2 meters and an image time interval of 5 years. Object-oriented interpretation is adopted during remote sensing interpretation, supplemented by on-site investigation to verify point data. Everything else is the same as in Specific Implementation Method One.

[0066] Specific Implementation Method Three: This implementation method differs from Specific Implementation Method One or Two in that step two involves obtaining characteristic data of the *Carex spp.* community in marsh wetlands and ecological environment data of the *Carex spp.* community distribution area, including *Carex spp.* community characteristic data, hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, landscape pattern data, and dispersal barrier data. Everything else is the same as in Specific Implementation Method One or Two.

[0067] Specific Implementation Method Four: This implementation method differs from Specific Implementation Methods One to Three in that the data collection of the characteristics of the *Carex spp.* grass mound community is carried out during the growing season of the year using the quadrat survey method, mainly collecting data on the growth status of *Carex spp.*, the morphological characteristics of the grass mounds, and the species diversity of the grass mound community.

[0068] The hydrological data includes water depth data;

[0069] The soil data includes soil moisture content, soil bulk density, and soil pH value.

[0070] The terrain data includes DEM and slope data;

[0071] The temperature data includes soil temperature;

[0072] The seed bank data includes seed dispersal methods and dispersal paths;

[0073] The interspecific competition data includes the competition or invasion of other species in the distribution area of ​​*Caragana korshinskii*.

[0074] The landscape pattern data and diffusion barrier data are obtained from the remote sensing image data interpreted in step one;

[0075] The seed diffusion path and diffusion barrier data need to be normalized, and the distance from each pixel to the image factor data needs to be calculated. Other aspects are the same as in any of the specific implementation methods one to three.

[0076] Specific Implementation Method Five: This implementation method differs from Specific Implementation Methods One to Four in that step three involves processing the data acquired in step one. Specifically, the remote sensing images from different historical periods interpreted in step one are unified into image data of the same size, projection, and resolution. Markov operations are then performed to form a Markov data chain, and the transition probability P is determined. ij The transferred area file is obtained, and the formula is as follows:

[0077] (Formula 1)

[0078] (Formula 2)

[0079] In the formula, S(t) and S(t+1) represent the diffusion state of the *Caragana korshinskii* community at times t and t+1, respectively. Other aspects are the same as in any of the specific implementation methods one to four.

[0080] Specific Implementation Method Six: This implementation method differs from Specific Implementation Methods One to Five in that, in step four, the MCE multi-criteria evaluation method is used to perform layer standardization processing on the acquired hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, and dispersal barrier data using fuzzy geometric functions. After standardization, the layer distributes factor conditions within a continuous range of 0-255, where values ​​closer to 0 represent unsuitable conditions, and values ​​closer to 255 represent optimal conditions. Everything else is the same as in Specific Implementation Methods One to Five.

[0081] Specific Implementation Method Seven: This implementation method differs from Specific Implementation Methods One through Six in that step five utilizes the CA-Markov model to simulate and predict the diffusion status of the *Caragana korshinskii* mound community. The mathematical expression for simulating the diffusion process of the *Caragana korshinskii* mound community is as follows:

[0082] (Formula 3)

[0083] In the formula, i and j are the row and column numbers of the cell, respectively, and d represents the dimension of the cell space. The diffusion of the *Caragana korshinskii* community refers to its diffusion from the ground to the surrounding area, which is a two-dimensional diffusion. , Let i and j represent the states of cells (i, j) in the *Caragana korshinskii* community at times t+1 and t, respectively. Represents the cell transformation rule, Let n be the state function of the neighboring cell (i, j) at time t, specifically representing the density of *Caragana korshinskii* community cells within an n×n neighborhood centered on cell (i, j), where n is the density of the neighboring cell (i, j). 2 Indicates the number of cells in the neighborhood. The formula for representing is as follows:

[0084] (Formula 4)

[0085] In the formula, con(S) ij =Carex) is a conditional expression indicating that the cell is affected by its neighborhood;

[0086] The CA-Markov model combines the quantitative simulation capabilities of the Markov model with the spatial simulation capabilities of the CA model. The transformation rules in the CA-Markov model... The formula for representing is as follows:

[0087] (Formula 5)

[0088] In the formula The transformation rule for each cell is composed of the numerical simulation of the Markov chain, the CA space simulation, and the fitness and restriction factors in the MCE. Other aspects are the same as in any of the specific implementation methods one through six.

[0089] Specific Implementation Method Eight: This implementation method differs from Specific Implementation Methods One through Seven in that the filter is selected as 5×5, and the iteration time y is equal to the image year interval for forming the Markov data chain, which can predict the spread of the *Caragana korshinskii* community in the next period. Everything else is the same as in Specific Implementation Methods One through Seven.

[0090] Specific Implementation Method Nine: This implementation method differs from Specific Implementation Methods One through Eight in that it sets different ecological scenarios by adjusting the Markov chain transition probability and the fitness and restriction factors in the MCE, and simulates and predicts the future spread of *Caragana korshinskii* mound communities under different ecological scenarios. Everything else is the same as in Specific Implementation Methods One through Eight.

[0091] Specific Implementation Method Ten: This implementation method differs from Specific Implementation Methods One through Nine in that the simulation accuracy verification is performed based on a pixel-by-pixel comparison method to evaluate the accuracy of the simulation results. The evaluation includes a comparison between the Kappa and FOM methods, as shown in the following formula:

[0092] (Formula 6)

[0093] Kappa measures the agreement between simulation results and actual results, where P0 is the proportion of land use distribution that is correctly simulated; P c To simulate the correct proportions under random conditions; P p To simulate a perfectly accurate scale under ideal conditions;

[0094] (Formula 7)

[0095] FOM measures the ability of the CA-Markov model to capture the dispersal of *Caragana korshinskii* mound communities, where UR... hit This indicates the diffusion cells of the *Caragana korshinskii* community correctly captured by the CA-Markov model, UR miss UR represents the diffusion cells of the *Caragana korshinskii* community that were missed by the CA-Markov model. false This represents the *Caragana korshinskii* community cells that were erroneously captured by the CA-Markov model. Everything else is the same as in any of the specific implementations one through nine.

[0096] Experiment 1:

[0097] The wetland sedge restoration area of ​​Sun Island Scenic Area in Harbin City, Heilongjiang Province was selected as the implementation area, combined with... Figure 1-6 As shown, Figure 1 The steps of the simulation and prediction method for the spread of Caragana korshinskii community in the implementation area are explained:

[0098] Step 1: Acquire WorldView 2 remote sensing imagery data for the implementation area from 2013 and 2018, with a resolution of 0.5m × 0.5m. Object-oriented interpretation was performed using the Ecoognition platform, and accuracy was verified using field survey points, ensuring that the classification accuracy of both sets of remote sensing images was above 90% (e.g., ...). Figures 4-6 ).

[0099] Step 2: In 2018, field investigations and monitoring were conducted to obtain data on the characteristics of the *Caragana korshinskii* grass mound community, hydrological data, soil data, temperature data, seed bank data, topographic data, interspecific competition data, and dispersal barrier data.

[0100] Step 3: First, process the data acquired in Step 1. This involves unifying the remote sensing images from different historical periods that were interpreted in Step 1 into image data of the same size, projection, and resolution. Then, perform Markov operations to form a Markov data chain and determine the transition probability P. ij The transferred area file is obtained, and the formula is as follows:

[0101] (Formula 1)

[0102] (Formula 2)

[0103] In the formula, S(t) and S(t+1) represent the diffusion state of the *Caragana korshinskii* community at times t and t+1, respectively, and the transfer area file is obtained.

[0104] Redundancy analysis was performed on the *Carex chinensis* mound community characteristic data and ecological environment data of the *Carex chinensis* mound community distribution area collected in step two. The main environmental factors affecting the dispersal of the *Carex chinensis* mound community were identified as: water level depth, seed dispersal mode, soil moisture content, DEM (Density Image Processing Model), and dispersal barrier data. Spatial interpolation of the *Carex chinensis* mound community characteristic data and the main influencing factors of *Carex chinensis* mound dispersal was performed using GIS technology to clarify the image threshold range, forming a database of *Carex chinensis* mound community characteristics and ecological environment (e.g., ...). Figure 7 ).

[0105] Step 4: Using the MCE multi-criteria evaluation method, the acquired hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, and dispersal barrier data are processed using fuzzy geometric functions for layer standardization. Dispersal barriers are used as the limiting factor, and other data are used as adaptation factors to create corresponding dispersal suitability images for each land cover category. The final MCE suitability atlas is thus formed. Experiment 1 had five land cover categories, but it was assumed that buildings did not undergo dispersal changes; therefore, this suitability atlas includes four suitability images (e.g., ...). Figure 8 ).

[0106] Step 5: The CA-Markov model is used to simulate and predict the spread of the *Carex spp.* community. Using the 2013 *Carex spp.* distribution area as the base map, the Markov transition matrix obtained in Step 3 as the transition area file, and the MCE suitability atlas obtained in Step 4 as a supplement, the pixel iteration count is 5 years, and the pixel filter is selected as 5×5 to obtain the simulated image for 2018. The accuracy of the 2018 simulated image is verified by comparing it with the actual classification image for 2018. The Kappa coefficient is greater than 0.85, and the FOM index also meets the standard level, achieving a good simulation effect. The simulation effect has a certain degree of accuracy and scientific validity. Then, using the 2018 *Carex spp.* distribution area as the base map, the Markov transition matrix obtained in Step 3 as the transition area file, and the MCE suitability atlas obtained in Step 4 as a supplement, the pixel iteration count is 5 years, and the pixel filter is selected as 5×5 to obtain the predicted image under the natural spread ecological scenario in 2023 (e.g., ...). Figure 9 This will provide valuable reference for the protection and restoration of the Caragana korshinskii wetland in 2023.

[0107] Step five uses the CA-Markov model to simulate and predict the spread of the *Caragana korshinskii* community, simulating the spread process of the *Caragana korshinskii* community. The mathematical expression is:

[0108] (Formula 3)

[0109] In the formula, i and j are the row and column numbers of the cell, respectively, and d represents the dimension of the cell space. The diffusion of the *Caragana korshinskii* community refers to its diffusion from the ground to the surrounding area, which is a two-dimensional diffusion. , Let i and j represent the states of cells (i, j) in the *Caragana korshinskii* community at times t+1 and t, respectively. Represents the cell transformation rule, Let n be the state function of the neighboring cell (i, j) at time t, specifically representing the density of *Caragana korshinskii* community cells within an n×n neighborhood centered on cell (i, j), where n is the density of the neighboring cell (i, j). 2 Indicates the number of cells in the neighborhood. The formula for representing is as follows:

[0110] (Formula 4)

[0111] In the formula, con(S) ij =Carex) is a conditional expression indicating that the cell is affected by its neighborhood;

[0112] The CA-Markov model combines the quantitative simulation capabilities of the Markov model with the spatial simulation capabilities of the CA model. The transformation rules in the CA-Markov model... The formula for representing is as follows:

[0113] (Formula 5)

[0114] In the formula The transformation rule for each cell is composed of the number simulation of the Markov chain, the CA space simulation, and the fitness and restriction factors in the MCE.

[0115] A 5×5 filter was selected, and the iteration time y was equal to the image year interval for forming the Markov data chain, which can predict the spread of the *Carex chinensis* mound community in the next period.

[0116] By adjusting the Markov chain transition probability and the fitness and restriction factors in the MCE to set different ecological scenarios, the future spread of *Caragana korshinskii* mound communities under different ecological scenarios was simulated and predicted.

[0117] The accuracy of the simulation is verified by evaluating the simulation results using a pixel-by-pixel comparison method. The evaluation includes a comparison between the Kappa and FOM methods, as shown in the following formula:

[0118] (Formula 6)

[0119] Kappa measures the agreement between simulation results and actual results, where P0 is the proportion of land use distribution that is correctly simulated; P c To simulate the correct proportions under random conditions; P p To simulate a perfectly accurate scale under ideal conditions;

[0120] (Formula 7)

[0121] FOM measures the ability of the CA-Markov model to capture the dispersal of *Caragana korshinskii* mound communities, where UR... hit This indicates the diffusion cells of the *Caragana korshinskii* community correctly captured by the CA-Markov model, UR miss UR represents the diffusion cells of the *Caragana korshinskii* community that were missed by the CA-Markov model. false This represents the *Caragana korshinskii* community cells that were incorrectly captured by the CA-Markov model.

Claims

1. A method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands, characterized in that... The simulation and prediction method for the diffusion of *Carex chinensis* grass mound communities in marsh wetlands is carried out according to the following steps: Step 1: Obtain remote sensing images of the distribution area of ​​*Carex chinensis* grass mound community in marsh wetlands at different historical periods and interpret the remote sensing image data; Step 2: Conduct field investigations and monitoring to obtain data on the characteristics of the *Carex chinensis* grass mound community in marsh wetlands and the ecological environment data of the *Carex chinensis* grass mound community distribution area; Step 3: Process the data obtained in Step 1 and Step 2 to form a database of the characteristics and ecological environment of the *Carex chinensis* grass mound community; Redundancy analysis was performed on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the distribution area of ​​the *Carex chinensis* mound community collected in step two to screen out the environmental factors that affect the spread of the *Carex chinensis* mound community. Spatial interpolation was performed using GIS technology on the characteristic data of the *Carex chinensis* mound community and the ecological environment data of the *Carex chinensis* mound community distribution area collected in step two to clarify the image threshold range. Step 4: Overlay the acquired image data using the MCE multi-criteria evaluation method to form an MCE suitability atlas; Step 5: Use the CA-Markov model to simulate and predict the spread of the *Carex chinensis* mound community, and verify the simulation accuracy. After the simulation results pass the accuracy verification, predict the spread of the *Carex chinensis* mound community under different ecological scenarios in the future.

2. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... The remote sensing images from different historical periods mentioned in Step 1 are high-resolution images with a spatial resolution better than 2 meters and an image time interval of 5 years. In the remote sensing interpretation process, object-oriented interpretation is adopted, supplemented by field surveys to verify point data.

3. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... Step two involves obtaining characteristic data of the *Carex spp.* community in marsh wetlands and ecological environment data of the distribution area of ​​the *Carex spp.* community, including characteristic data of the *Carex spp.* community, hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, landscape pattern data, and dispersal barrier data.

4. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 3, characterized in that... The data on the characteristics of the *Caragana korshinskii* mound community were collected during the growing season of the year using the quadrat survey method, mainly collecting data on the growth status of *Caragana korshinskii*, the morphological characteristics of the mounds, and the species diversity of the mound community. The hydrological data includes water depth data; The soil data includes soil moisture content, soil bulk density, and soil pH value. The terrain data includes DEM and slope data; The temperature data includes soil temperature; The seed bank data includes seed dispersal methods and dispersal paths; The interspecific competition data includes the competition or invasion of other species in the distribution area of ​​*Caragana korshinskii*. The landscape pattern data and diffusion barrier data are obtained from the remote sensing image data interpreted in step one of claim 1; The seed diffusion path and diffusion barrier data need to be normalized, and the distance from each pixel to the image factor data needs to be calculated.

5. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... The data obtained in step one is processed in step three, that is, the remote sensing images of different historical periods interpreted in step one are unified into image data of the same range size, the same projection and the same resolution, and Markov operation is performed to form a Markov data chain and determine the transition probability P ij and obtain a transition area file, and the formula is as follows: S (t+1) = P ij × S t (Equation 2) In the formula, S(t) and S(t+1) represent the diffusion state of the *Caragana korshinskii* community at times t and t+1, respectively.

6. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... In step four, the MCE multi-criteria evaluation method is used to perform layer standardization processing on the acquired hydrological data, soil data, topographic data, temperature data, seed bank data, interspecific competition data, and dispersal barrier data using fuzzy geometric functions. After standardization, the layer will distribute the factor conditions in a continuous range of 0-255, where the closer to 0 is to unsuitable conditions, and the closer to 255 is to optimal conditions.

7. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... Step five uses the CA-Markov model to simulate and predict the spread of the *Caragana korshinskii* community, simulating the spread process of the *Caragana korshinskii* community. The mathematical expression is: In the formula, i and j are the row and column numbers of the cell, respectively, and d represents the dimension of the cell space. The diffusion of the *Caragana korshinskii* community refers to its diffusion from the ground to the surrounding area, which is a two-dimensional diffusion. Let f represent the states of cell (i, j) in the *Caragana korshinskii* community at times t+1 and t, respectively, and let f represent the cell transition rule. Let n be the state function of the neighboring cell (i, j) at time t, specifically representing the density of *Caragana korshinskii* community cells within an n×n neighborhood centered on cell (i, j), where n is the density of the neighboring cell (i, j). 2 Indicates the number of cells in the neighborhood. The formula for representing is as follows: In the formula, con(S) ij =Carex) is a conditional expression indicating that the cell is affected by its neighborhood; The CA-Markov model combines the quantitative simulation capabilities of the Markov model with the spatial simulation capabilities of the CA model. The transformation rule f in the CA-Markov model is expressed as follows: f = (markov, CA, MCE) (Formula 5) In the formula, f represents the transformation rule of each cell, which is composed of the quantitative simulation of the Markov chain, the CA space simulation, and the fitness factor and restriction factor in MCE.

8. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 7, characterized in that... A 5×5 filter was selected, and the iteration time y was equal to the image year interval for forming the Markov data chain, which can predict the spread of the *Carex chinensis* mound community in the next period.

9. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 7, characterized in that... By adjusting the Markov chain transition probability and the fitness and restriction factors in the MCE to set different ecological scenarios, the future spread of *Caragana korshinskii* mound communities under different ecological scenarios was simulated and predicted.

10. The method for simulating and predicting the spread of *Carex chinensis* grass mound communities in marsh wetlands according to claim 1, characterized in that... The accuracy of the simulation is verified by evaluating the simulation results using a pixel-by-pixel comparison method. The evaluation includes a comparison between the Kappa and FOM methods, as shown in the following formula: Kappa measures the agreement between simulation results and actual results, where P0 is the proportion of land use distribution that is correctly simulated; P c To simulate the correct proportions under random conditions; P p To simulate a perfectly accurate scale under ideal conditions; FOM measures the ability of the CA-Markov model to capture the dispersal of *Caragana korshinskii* mound communities, where UR... hit This indicates the diffusion cells of the *Caragana korshinskii* community correctly captured by the CA-Markov model, UR miss UR represents the diffusion cells of the *Caragana korshinskii* community that were missed by the CA-Markov model. false This represents the *Caragana korshinskii* community cells that were incorrectly captured by the CA-Markov model.

Images