A method, medium and system for assessing the ecological environment damage value of Spartina alterniflora

Abstract

The present invention provides a method, medium and system for assessing the ecological environment damage value of Spartina alterniflora, belonging to the technical field of Spartina alterniflora ecological environment damage assessment. The present invention introduces an ecological damage diffusion equation to describe the propagation coupling mechanism of multiple factors, utilizes a spatiotemporal damage analysis model based on a graph convolutional network to process complex factor interactions, establishes a time weight factor correction model to quantify the dynamic changes of cumulative effects, combines a biological inhibitor effect model to analyze the synergistic effects of multiple inhibition mechanisms such as local reed competition, natural enemy insect control, and salt stress, and achieves accurate damage quantification and prevention and control strategy formulation through a regional ecological environment damage value comprehensive assessment model and an inhibitor optimization configuration model, thereby solving the technical problem that the existing technology cannot effectively analyze the interaction mechanism of multiple ecological factors during the invasion of Spartina alterniflora, resulting in insufficient accuracy in ecological damage value assessment.

Description

Technical Field

[0001] The present invention belongs to the technical field of ecological environment damage assessment of Spartina alterniflora, and specifically relates to a method, medium and system for assessing the ecological environment damage value of Spartina alterniflora. Background Art

[0002] The assessment of the ecological and environmental damage value of Spartina alterniflora is an important research area in invasion biology and ecological economics. Traditional assessment methods mainly use single-factor analysis and static assessment models. Field surveys are used to obtain data on the distribution area of ​​Spartina alterniflora and the overall damage value is calculated based on the unit price of ecosystem service functions. These methods are widely used in wetland ecological protection planning, invasive species prevention and control decisions, and ecological restoration project assessments, providing government departments and scientific research institutions with a basis for quantifying damage. However, traditional assessment techniques have significant flaws, mainly manifested in ignoring the dynamic correlation between invasion rate and ecological service function damage, lacking comprehensive consideration of temporal cumulative effects and spatial transmission mechanisms, and failing to effectively integrate the synergistic effects of multiple ecological factors such as biological competition, environmental stress, and niche overlap. In current ecological damage assessment practices, the inability of existing technologies to deeply analyze the complex interactions between multiple factors such as Spartina alterniflora invasion, competition among native species, changes in environmental factors, and ecosystem responses leads to large deviations between assessment results and the actual ecological damage situation, making it difficult to accurately reflect the true impact of the invasion process. Summary of the Invention

[0003] In view of this, the present invention provides a method, medium and system for assessing the ecological environment damage value of Spartina alterniflora, which can solve the technical problem in the prior art that the interaction mechanism of multiple ecological factors during the invasion of Spartina alterniflora cannot be effectively analyzed, resulting in insufficient accuracy in the ecological damage value assessment.

[0004] The present invention is achieved as follows: In a first aspect, the present invention provides a method for assessing the ecological environment damage value of Spartina alterniflora, comprising the following steps: establishing a multi-temporal remote sensing image database of the area invaded by Spartina alterniflora, and extracting time-series change data of the area covered by Spartina alterniflora; calculating the invasion rate parameters of Spartina alterniflora based on the multi-temporal remote sensing image database, and establishing an invasion rate prediction model; constructing an ecosystem service function loss assessment index system, selecting five core ecological service functions, namely, water purification function, biodiversity maintenance function, soil conservation function, climate regulation function, and fishery resource supply function, and calculating the impact of Spartina alterniflora invasion on various ecological service functions using an ecological damage diffusion equation; establishing a propagation mechanism for the invasion of Spartina alterniflora. The damage coefficient matrix of invasion to various ecological service functions is calculated, and the optimal distribution of the damage coefficient matrix is ​​solved based on the multi-constraint optimization problem of minimizing the total cost of ecological damage; a time weight factor correction model is constructed, and the ecological value comprehensive assessment model is used to dynamically integrate the damage value of various ecological service functions; a biological inhibition factor assessment module is introduced, and an inhibition factor effect model is established; a regional ecological environment damage value comprehensive assessment model is established; based on the overall ecological environment damage value and the biological inhibition factor effect model, the ecological damage value per unit area in each period is multiplied by the invasion area in the corresponding period, and combined with the correction value of the biological inhibition factor effect model, the overall ecological environment damage value of Spartina alterniflora invasion in the study area is accumulated to obtain.

[0005] Among them, the steps to establish a multi-temporal remote sensing image database of the area invaded by Spartina alterniflora are to obtain high-resolution satellite images for more than five consecutive years, extract the temporal change data of the area covered by Spartina alterniflora, and construct a basic data set for dynamic monitoring of invasion and spread.

[0006] Among them, the step of calculating the invasion rate parameters of Spartina alterniflora based on the multi-temporal remote sensing image database is to establish the invasion rate time series by analyzing the ratio of the change in the coverage area of ​​Spartina alterniflora in adjacent years to the time interval, and use the least squares method to fit the invasion rate prediction model.

[0007] Among them, the ecological damage diffusion equation is used to describe the propagation and diffusion mechanism of ecological function damage caused by the invasion of Spartina alterniflora in spatial and temporal dimensions. The input includes the invasion source intensity, environmental resistance coefficient, ecosystem sensitivity index, diffusion time parameter, and boundary condition parameter. The output is the distribution of ecological function damage intensity at each spatial location at different times.

[0008] Among them, the multi-constraint optimization problem belongs to the multi-objective linear programming problem. By establishing a mathematical optimization model with minimizing the total ecological damage cost as the objective function, and with the sum constraint of the damage coefficients of various ecological service functions, the upper and lower limit constraints of the damage coefficients of single functions, and the difference constraint of the damage coefficients of adjacent functions as constraints, the simplex method is used to solve the optimal damage coefficient distribution plan for various ecological service functions.

[0009] Among them, before calculating the value of ecological damage per unit area caused by the invasion of Spartina alterniflora in each period based on the invasion rate prediction model and the damage coefficient matrix, it also includes the introduction of a spatiotemporal damage analysis model to achieve accurate identification and quantification of the degree of damage in different regions and periods.

[0010] Among them, the spatiotemporal damage analysis model is an ecological damage spatiotemporal prediction model built based on the graph convolutional network architecture. Its structure is a deep learning network architecture including an input layer, a multi-layer graph convolution layer, a temporal feature extraction layer, a spatial feature fusion layer and an output layer.

[0011] Among them, the biological inhibition factor effect model is used to calculate the inhibitory effect of local reed population density, natural enemy insect community structure of Spartina alterniflora, and soil salt concentration on the invasion of Spartina alterniflora. The inputs include local reed population density, natural enemy insect community structure of Spartina alterniflora, soil salt concentration, environmental load capacity, and niche competition intensity.

[0012] Among them, the graph convolution layer adopts the neighbor sampling mechanism for message transmission, and the number of neighbor sampling is dynamically determined according to three parameters: invasion diffusion range, ecosystem connectivity index and evaluation accuracy requirements.

[0013] Among them, the steps for establishing the training dataset for the spatiotemporal damage analysis model include collecting multi-temporal remote sensing image data and corresponding ecological damage assessment reports of 100 typical cases of Spartina alterniflora invasion worldwide, and constructing a multidimensional data sample containing spatial coordinates, time series, ecological damage intensity and environmental characteristics.

[0014] Among them, the training steps of the spatiotemporal damage analysis model include first using an unsupervised pre-training method to learn the structural features of the graph convolutional network to obtain initialization parameters, and then using a supervised learning method to input training data for end-to-end model training.

[0015] Among them, the sampling weight adjustment function is used to adjust the neighbor sampling parameters of the spatiotemporal damage analysis model, and the sampling adjustment factor value is calculated based on the invasion diffusion rate, ecosystem complexity index, environmental change intensity, and evaluation time scale.

[0016] Among them, the gating weight function is used to adjust the gating mechanism parameters of the spatiotemporal damage analysis model, and the gating adjustment value is calculated based on the intensity of ecological damage transmission, the degree of spatial heterogeneity, the severity of temporal changes, and the confidence of model prediction.

[0017] Among them, the comprehensive ecological value assessment model is used to weightedly calculate the damage value of various ecological service functions to obtain the overall damage value. The inputs include the damage value of water purification function, the damage value of biodiversity maintenance function, the damage value of soil conservation function, the damage value of climate regulation function, and the damage value of fishery resource supply function.

[0018] Among them, the inhibition factor optimization configuration model establishes an optimization problem with the objective function of minimizing the ecological restoration cost and maximizing the damage reduction benefit ratio, and with the constraints of ecosystem stability, biodiversity protection, and technical feasibility, to output the local reed population input density range, the natural enemy insect input density range of Spartina alterniflora, and the soil salt regulation concentration range.

[0019] Among them, the time weight factor refers to the parameter that weights and corrects the cumulative effect of ecological damage according to the duration of invasion, reflecting the irreversible damage caused by long-term invasion to the ecosystem, and is used to correct the damage value weight coefficient at different invasion stages.

[0020] A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions. When the program instructions are run in a computer, the program instructions are used to execute the above-mentioned method for assessing the ecological environment damage value of Spartina alterniflora.

[0021] The third aspect of the present invention provides a system for assessing the ecological environmental damage value of Spartina alterniflora, comprising the above-mentioned computer-readable storage medium. The system is any one of a computer, a server, and a single-chip microcomputer. The computer-readable storage medium is arranged in the system, and the system is provided with a microprocessor for executing program instructions stored in the computer-readable storage medium.

[0022] The present invention, by constructing a multiple ecological factor coupling analysis system, establishes a damage coefficient matrix multi-constraint optimization model and a spatiotemporal damage analysis model, and combines the biological inhibitor effect assessment to achieve an in-depth analysis of the interaction mechanism of multiple ecological factors during the invasion of Spartina alterniflora. This method solves the limitations of single factor analysis in traditional assessment techniques. By introducing an invasion rate prediction model, a time weight factor correction model, and an ecological damage diffusion equation, it can effectively capture the dynamic correlation between invasion diffusion, ecological function loss, and environmental response. By establishing a spatiotemporal damage analysis model based on a graph convolutional network, this method can handle the complex interactions of multiple factors such as species competition, environmental stress, and niche overlap in the ecosystem, and combines the biological inhibitor effect model to quantify the synergistic effects of inhibitory mechanisms such as local reed competition, natural enemy insect control, and salt stress, thereby solving the technical problem that the prior art cannot effectively analyze the interaction mechanism of multiple ecological factors during the invasion of Spartina alterniflora, resulting in insufficient accuracy in ecological damage value assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of the method of the present invention.

[0024] Figure 2It is a structural diagram of the spatiotemporal damage analysis model involved in the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0026] like Figure 1 1 is a flow chart of a method for assessing the ecological environment damage value of Spartina alterniflora provided by the first aspect of the present invention, and the method comprises the following steps:

[0027] S01. Establish a multi-temporal remote sensing image database of the areas invaded by Spartina alterniflora. By acquiring high-resolution satellite images for more than five consecutive years, extract the temporal variation data of the area covered by Spartina alterniflora and construct a basic data set for dynamic monitoring of invasion and spread.

[0028] S02. Calculate the invasion rate parameters of Spartina alterniflora based on a multi-temporal remote sensing image database, establish an invasion rate time series by analyzing the ratio of the change in Spartina alterniflora coverage area to the time interval between adjacent years, and obtain an invasion rate prediction model using the least squares method;

[0029] S03. Construct an indicator system for assessing ecosystem service loss. Select five core ecosystem services: water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource provision. Establish quantitative standards for the per-unit-area service value of each ecosystem service. Use the ecological damage diffusion equation to calculate the impact of Spartina alterniflora invasion on each ecosystem service.

[0030] S04. Establish a damage coefficient matrix for various ecological service functions caused by the invasion of Spartina alterniflora. Through field surveys and experimental analysis, determine the proportion of ecological service functions that decline relative to native vegetation in areas covered by Spartina alterniflora, form quantitative parameters for damage intensity, and solve the optimal allocation of the damage coefficient matrix based on a multi-constrained optimization problem that minimizes the total cost of ecological damage.

[0031] S05. Construct a time-weighted factor correction model. Based on the impact of invasion duration on the cumulative effect of ecological damage, establish a time-weighted function to correct the damage value weight coefficients at different invasion stages. At the same time, use the ecological value comprehensive assessment model to dynamically integrate the damage value of various ecological service functions.

[0032] S06. Based on the invasion rate prediction model and damage coefficient matrix, calculate the ecological damage value per unit area caused by the invasion of Spartina alterniflora at each period. Combined with the time weight factor, a dynamically revised damage value sequence is obtained. By introducing a spatiotemporal damage analysis model, the degree of damage in different regions and periods can be accurately identified and quantified.

[0033] S07. Introduce a biological inhibitor assessment module. By analyzing the inhibitory effects of local reed population density, the community structure of natural enemies of Spartina alterniflora, and soil salinity on the growth of Spartina alterniflora, establish an inhibitor effect model and quantify the impact of each inhibitor on the invasion rate parameters and damage coefficient matrix.

[0034] S08. Establish a comprehensive assessment model for the regional ecological and environmental damage value. Multiply the ecological damage value per unit area in each period by the invasion area in the corresponding period, combine it with the correction value of the biological inhibitor effect model, and accumulate it to obtain the overall ecological and environmental damage value of Spartina alterniflora invasion in the study area.

[0035] Optionally, it also includes S09, establishing an inhibitor optimization configuration model based on the overall ecological environmental damage value and the biological inhibitor effect model, and determining the optimal inhibitor investment plan for areas with different damage degrees through cost-benefit analysis.

[0036] The multi-temporal remote sensing image database refers to a collection of remote sensing image data acquired over the same area at different time points, used to monitor time series data on changes in land cover. The invasion rate parameter refers to the rate of increase in the area covered by Spartina alterniflora per unit time, reflecting the speed of the invasion and spread. Ecosystem services refer to the various direct and indirect benefits provided by ecosystems to humans, including provisioning services, regulating services, supporting services, and cultural services. The damage coefficient matrix is ​​a numerical matrix describing the extent of damage caused by the invasion of Spartina alterniflora to various ecosystem services, with each coefficient representing the relative proportion of function lost. The temporal weighting factor is a parameter that weights the cumulative effects of ecological damage based on the duration of the invasion, reflecting the irreversible damage caused to the ecosystem by long-term invasion. Dynamic correction refers to the process of adjusting assessment results to account for temporal changes, ensuring that the assessment more accurately reflects the dynamic characteristics of the invasion process. Biotic inhibitors refer to biological or environmental factors that inhibit the growth and spread of Spartina alterniflora, including the density of native reed populations, the community structure of Spartina alterniflora's natural enemy insects, and soil salinity.

[0037] The ecological damage diffusion equation is used to describe the propagation and diffusion mechanism of ecological function damage caused by the invasion of Spartina alterniflora in spatial and temporal dimensions. The input includes the invasion source intensity, environmental resistance coefficient, ecosystem sensitivity index, diffusion time parameter, and boundary condition parameter. The output is the distribution of ecological function damage intensity at each spatial location at different times.

[0038] The comprehensive ecological value assessment model is used to perform weighted comprehensive calculation of the damage value of various ecological service functions to obtain the overall damage value. The input includes the damage value of water purification function, the damage value of biodiversity maintenance function, the damage value of soil conservation function, the damage value of climate regulation function, the damage value of fishery resource supply function, the ecological function importance weight coefficient, the regional ecological sensitivity factor, the time attenuation coefficient, and the environmental restoration difficulty coefficient. The output is the comprehensive ecological environment damage value.

[0039] The multi-constraint optimization problem involved in step S04 is a multi-objective linear programming problem. By establishing a mathematical optimization model with minimizing the total ecological damage cost as the objective function, and with the sum constraint of the damage coefficients of various ecological service functions, the upper and lower limit constraints of the damage coefficients of individual functions, and the difference constraint of the damage coefficients of adjacent functions as constraints, the simplex method is used to solve the optimal damage coefficient distribution plan for various ecological service functions.

[0040] The spatiotemporal damage analysis model is an ecological damage spatiotemporal prediction model constructed based on a graph convolutional network architecture. The structure of the spatiotemporal damage analysis model is a deep learning network architecture including an input layer, a multi-layer graph convolution layer, a temporal feature extraction layer, a spatial feature fusion layer and an output layer. The graph convolution layer adopts a neighbor sampling mechanism for message transmission. The number of neighbor sampling is dynamically determined based on three parameters: the invasion spread range, the ecosystem connectivity index and the evaluation accuracy requirements. When the invasion spread range is less than 500m, the neighbor sampling number is set to 8 to 12 nodes. When the invasion spread range is between 500m and 2000m, the neighbor sampling number is adjusted to 15 to 20 nodes. When the invasion spread range exceeds 2000m, the neighbor sampling number is increased to 25 to 30 nodes.

[0041] The steps for establishing a training data set for the spatiotemporal damage analysis model include collecting multi-temporal remote sensing image data and corresponding ecological damage assessment reports of 100 typical cases of Spartina alterniflora invasion worldwide, extracting invasion area change series, ecological service function loss data, and environmental factor change information for each case, constructing a multidimensional data sample containing spatial coordinates, time series, ecological damage intensity, and environmental characteristics, and stratifying the data according to geographical regions and climate types to form training sets, validation sets, and test sets, of which 70% of the training set is used for model parameter learning, 15% of the validation set is used for hyperparameter tuning, and 15% of the test set is used for model performance evaluation.

[0042] The spatiotemporal damage analysis model training steps include first using an unsupervised pre-training method to learn the structural features of the graph convolutional network to obtain initialization parameters, and then using a supervised learning method to input training data for end-to-end model training. During the training process, the mean square error loss function is used to measure the difference between the predicted damage value and the actual damage value. The Adam optimizer is used for gradient descent optimization, the learning rate is set to 0.001 and the cosine annealing strategy is used for dynamic adjustment. The training batch size is set to 32, the total number of training rounds is 200, the model performance is evaluated on the validation set every 10 rounds and the optimal model parameters are saved. After the training is completed, the final performance evaluation is performed on the test set.

[0043] The sampling weight adjustment function is used to adjust the neighbor sampling parameters of the spatiotemporal damage analysis model. The sampling weight adjustment function calculates the sampling adjustment factor value based on the invasion diffusion rate, the ecosystem complexity index, the environmental change intensity, and the evaluation time scale. When the sampling adjustment factor value is less than 0.3, the linear growth weight adjustment function is used to linearly increase the neighbor sampling number by 10% to 15%. When the sampling adjustment factor value is between 0.3 and 0.7, the exponential decay weight adjustment function is used to adjust the neighbor sampling number according to the exponential law. When the sampling adjustment factor value is greater than 0.7, the saturation function weight adjustment function is used to control the neighbor sampling number within the maximum threshold range.

[0044] The gating weight function is used to adjust the gating mechanism parameters of the spatiotemporal damage analysis model. The gating weight function calculates the gating adjustment value based on the ecological damage transmission intensity, spatial heterogeneity, temporal variation intensity, and model prediction confidence. When the gating adjustment value is less than 0.4, the sigmoid activation weight adjustment function is used to enhance the selectivity of the gating. When the gating adjustment value is between 0.4 and 0.8, the tanh hyperbolic tangent weight adjustment function is used to balance the openness of the gating. When the gating adjustment value is greater than 0.8, the ReLU modified linear weight adjustment function is used to ensure the non-negativity and stability of the gating parameters.

[0045] The biological inhibition factor effect model is used to calculate the inhibitory effect of local reed population density, natural enemy insect community structure of Spartina alterniflora, and soil salt concentration on the invasion of Spartina alterniflora. The inputs include local reed population density, natural enemy insect community structure of Spartina alterniflora, soil salt concentration, environmental load capacity, and niche competition intensity. The outputs include reed competition inhibition coefficient, natural enemy insect control coefficient, and salt stress inhibition coefficient. The inhibition coefficient is used to correct the corresponding values ​​in the invasion rate parameter and damage coefficient matrix.

[0046] The inhibition factor optimization configuration model establishes an optimization problem with the objective function of minimizing the ecological restoration cost and maximizing the damage reduction benefit ratio, and with the constraints of ecosystem stability, biodiversity protection, and technical feasibility. According to the overall ecological environmental damage value level of different regions, it outputs the corresponding local reed population input density range, Spartina alterniflora natural enemy insect input density range, and soil salt adjustment concentration range.

[0047] The native reed population density refers to the number of native reed plants per unit area and is used to measure the reed's ability to competitively inhibit Spartina alterniflora. The community structure of natural enemy insects of Spartina alterniflora refers to the species composition and abundance distribution of insects that feed on Spartina alterniflora, primarily including Spartina alterniflora sawflies, Spartina alterniflora borers, and Spartina alterniflora aphids. Soil salinity refers to the content of soluble salts in the soil; high salt concentrations exert a stressful effect on the growth of Spartina alterniflora. The reed competitive inhibition coefficient refers to the intensity coefficient of the inhibition exerted by native reeds on the growth of Spartina alterniflora through competition for light, nutrients, and space. The natural enemy insect control coefficient refers to the intensity coefficient of the control exerted by natural enemy insects of Spartina alterniflora on the Spartina alterniflora population through feeding and oviposition. The salt stress inhibition coefficient refers to the intensity coefficient of the inhibition exerted by a high soil salinity environment on the physiological activities and growth and development of Spartina alterniflora. The invasion spread range refers to the maximum distance over which Spartina alterniflora can spread from the invasion source to the surrounding area. The ecosystem connectivity index is an indicator of the degree of connectivity between the flows of matter, energy, and information among ecosystem components. Assessment accuracy requirements refer to the accuracy and reliability standards that the ecological damage assessment results must meet. The invasion spread rate refers to the rate at which the area covered by Spartina alterniflora increases per unit time.

[0048] The ecosystem complexity index refers to an indicator of the complexity of the internal structural levels, component types, and interrelationships of an ecosystem. The intensity of environmental change refers to the severity of changes in environmental factors within the study area. The assessment time scale refers to the time span involved in the ecological damage assessment. The intensity of ecological damage transmission refers to the intensity of the spatial spread of ecological function damage caused by the invasion of Spartina alterniflora. The degree of spatial heterogeneity refers to the degree of unevenness and variability in the spatial distribution of ecosystems. The intensity of temporal change refers to the rapidity of changes in ecological damage over time. The confidence level of model prediction refers to the level of credibility of the prediction results of the spatiotemporal damage analysis model. The environmental load capacity refers to the maximum bioload that an environmental system can bear.

[0049] Niche competition intensity refers to the intensity of competition between different species occupying the same ecological niche. The inhibitor optimization allocation model refers to a mathematical optimization model that determines the biological inhibitor input plan based on the results of an ecological damage value assessment. The local reed population input density range refers to the reed planting density interval determined by the regional damage level. The natural enemy insect release density range of Spartina alterniflora refers to the natural enemy insect release density interval determined by the regional damage level. The soil salinity regulation concentration range refers to the soil salinity control concentration interval determined by the regional damage level.

[0050] The specific implementation of the above steps is described in detail below.

[0051] The specific embodiment of step S01 is to obtain the high-resolution image data of the continuous time series of the study area by a multi-source satellite remote sensing platform. First, platforms such as Landsat series satellites and Sentinel-2 satellites are utilized to obtain multispectral remote sensing images with a spatial resolution of not less than 10m and a temporal resolution of no more than 16 days, and an image database covering at least 5 years of time span in the target area is established. Then, an atmospheric correction algorithm is adopted to carry out radiation calibration and atmospheric correction processing to the original image, to eliminate the impact of atmospheric scattering and absorption on spectral information. Then, vegetation water index algorithms such as normalized vegetation index NDVI and normalized water index NDWI are used, in conjunction with the support vector machine classifier in the supervised classification algorithm, the Spartina alterniflora covered area in the image is automatically identified and extracted. The effect of this step is to provide accurate and reliable Spartina alterniflora spatiotemporal distribution basic data for subsequent analysis, wherein the image spatial resolution threshold is set at 10m, and the temporal resolution threshold is set at 16 days, and the classification accuracy threshold requirement reaches more than 85%.

[0052] The specific implementation method of step S02 is to calculate the rate parameters of invasion and diffusion based on the extracted time series data of the area covered by Spartina alterniflora. First, the variation of the area covered by Spartina alterniflora between adjacent time nodes is calculated, and the area growth rate of each time period is obtained by the differential calculation method. Then, the least squares fitting algorithm is used to perform trend analysis on the area growth rate time series, and a mathematical prediction model of the invasion rate changing over time is established. Then, the autoregressive moving average model ARIMA in the time series analysis is used to predict the invasion rate, and the validity of the prediction model is verified by model parameter estimation and residual analysis. The purpose of this step is to quantify the dynamic characteristics of the invasion and diffusion of Spartina alterniflora, and to provide a parameter basis of the time dimension for damage assessment, wherein the determination coefficient of the fitting model is 100%. The threshold requirement is no less than 0.8, and the prediction error threshold is controlled within 15%.

[0053] The specific implementation method of step S03 is to construct a loss assessment indicator system covering multiple ecological service functions. First, based on the theory of ecosystem service function classification, five core functions such as water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource supply are selected as assessment objects. Then, environmental economics assessment methods such as market value method, replacement cost method, and conditional value method are used to establish a quantitative standard for the unit area value of each ecological service function. Then, an ecological damage diffusion equation is constructed, and the mathematical principles of the diffusion reaction equation are used to describe the propagation mechanism of damage in the time and space dimensions. The input parameters include the intensity of the invasion source, the environmental resistance coefficient, the ecosystem sensitivity index, etc. The purpose of this step is to establish a scientific and reasonable ecological damage assessment framework, in which the uncertainty threshold of the quantification of the value of each ecological service function is controlled within 20%, and the spatial resolution threshold of the diffusion equation is set to 100m.

[0054] The specific implementation method of step S04 is to determine the quantitative parameters of the degree of damage to various ecological service functions caused by the invasion of Spartina alterniflora through field surveys and comparative experiments. First, sample plots in the Spartina alterniflora invasion area and the native vegetation control area are set, and the measured data of various ecological service functions are collected using a paired sample plot survey method. The decline ratio of the invasion area relative to the various functions in the control area is then calculated to form the initial value of the damage coefficient. Then a multi-objective linear programming optimization model is constructed, with minimizing the total ecological damage cost as the objective function, and constraints such as the sum constraint of the damage coefficient of each function, the upper and lower limit constraints of the single function coefficient are set. Finally, the simplex method is adopted to solve the optimization problem to obtain the optimal damage coefficient allocation scheme for each ecological service function. The purpose of this step is to scientifically determine the damage intensity parameter and provide an accurate coefficient basis for quantitative assessment, wherein the field survey sample plot quantity threshold is no less than 30 pairs, and the damage coefficient value range threshold is set between 0.1 and 0.9.

[0055] The specific implementation method of step S05 is to establish a weight correction model that takes into account the cumulative effect of time. First, based on the irreversibility and cumulative characteristics of ecological damage, a time-weighted function is constructed to reflect the continuous impact of long-term invasion on the ecosystem. Then, an exponential decay function is used to describe the accumulation law of damage over time, and the decay parameter is determined by fitting historical case data. Then, an ecological value comprehensive assessment model is established, and a weighted summation algorithm is used to comprehensively calculate the damage value of various ecological service functions. The weight coefficient is dynamically adjusted according to the functional importance and regional sensitivity. The purpose of this step is to correct the limitations of static assessment and accurately reflect the dynamic damage process. The threshold range of the time weight decay parameter is set to 0.05 to 0.15, and the threshold of the total weight coefficient of the comprehensive assessment model is 1.0.

[0056] The specific implementation method of step S06 is to calculate the dynamic damage value of each period based on the above-mentioned model. First, the output results of the invasion rate prediction model are combined with the damage coefficient matrix to calculate the ecological damage value per unit area caused by the invasion of Spartina alterniflora at each time node. Then, the time weight factor is applied to correct the damage value of different periods to obtain a dynamic damage value sequence that takes into account the cumulative effect. Then, a spatiotemporal damage analysis model is introduced, which is based on a graph convolutional network architecture to accurately identify and quantify the degree of damage in different regions and periods. The purpose of this step is to achieve spatiotemporal dynamic quantification of the damage value, where the calculation accuracy threshold of the damage value per unit area is required to be within 10%, and the prediction accuracy threshold of the spatiotemporal analysis model is not less than 80%.

[0057] The specific implementation method of step S07 is to quantify the effect of biological inhibition factors on the invasion of Spartina alterniflora. First, the current status parameters of key inhibition factors such as local reed population density, Spartina alterniflora natural enemy insect community structure, and soil salinity concentration are measured through field surveys. Then, based on the niche competition theory and the interspecific competition model, the competitive relationship between reed and Spartina alterniflora is analyzed, and the reed competition inhibition coefficient is calculated. Then, the predator-prey dynamics model is adopted to analyze the control effect of natural enemy insects on Spartina alterniflora, and the natural enemy insect control coefficient is calculated. Simultaneously, based on the principle of plant physiological ecology, the stress effect of soil salinity on the growth of Spartina alterniflora is analyzed, and the salt stress inhibition coefficient is calculated. The effect of this step is to quantify the effect of the natural inhibition mechanism, which provides a scientific basis for subsequent optimization configuration. wherein the reed density survey accuracy threshold is 95%, and the inhibition coefficient value range threshold is set at 0.1 to 0.8.

[0058] The specific implementation method of step S08 is to set up a comprehensive assessment model for the overall damage value of the region. First, the ecological damage value per unit area of ​​each period is multiplied by the invasion area of ​​Spartina alterniflora in the corresponding period to obtain the total damage value of each period. Then, in conjunction with the correction value of the biological inhibitor effect model, the damage value is adjusted to reflect the influence of the inhibitory effect. Then, a cumulative summation algorithm is adopted to add up the correction damage value of each period to obtain the overall ecological environment damage value of the invasion of Spartina alterniflora in the study area. The purpose of this step is to realize the comprehensive quantification of the damage value at the regional scale, and provide an overall assessment result for decision-making, wherein the area measurement accuracy threshold requirement reaches within 5%, and the uncertainty threshold of the overall damage value is controlled within 25%.

[0059] Step S09 is an optional step, and its specific implementation method is to establish an optimal configuration model of inhibitory factors based on the principle of cost-benefit analysis. First, the ratio of minimizing the cost of ecological restoration to maximizing the benefit of damage reduction is used as the objective function to construct a multi-constraint optimization problem. Then, constraints such as ecosystem stability constraints, biodiversity protection constraints, and technical feasibility constraints are set to ensure the ecological safety and feasibility of the optimization plan. Then, according to the overall ecological environmental damage value level of different regions, a hierarchical optimization algorithm is used to determine the corresponding inhibitory factor input plan. The purpose of this step is to provide scientific inhibitory measure configuration plans for areas with different degrees of damage and achieve precise governance, wherein the cost-benefit ratio threshold is set to above 1:3, the reed input density range threshold is 50 to 200 plants per square meter, the natural enemy insect release density range threshold is 10 to 50 heads per square meter, and the soil salt adjustment concentration range threshold is 3 to 8 grams per kilogram.

[0060] The spatiotemporal damage analysis model utilizes a deep learning architecture based on graph convolutional networks (GCNs) to predict and analyze the spatiotemporal dynamics of ecological damage caused by the invasion of Spartina alterniflora. The model's detailed structure comprises five main components: an input layer, a multi-layer graph convolutional layer, a temporal feature extraction layer, a spatial feature fusion layer, and an output layer. The input layer receives multidimensional input data, including spatial coordinates, time series, and environmental characteristics, and performs standardized preprocessing to ensure data quality. The multi-layer GCN layer uses a neighbor sampling mechanism for message passing, aggregating information about neighboring nodes to learn spatial dependencies. The number of neighbor samples is dynamically adjusted based on the invasion's spread range: 8 to 12 nodes when the spread range is less than 500 meters, 15 to 20 nodes when the spread range is between 500 and 2000 meters, and 25 to 30 nodes when the spread range exceeds 2000 meters. The temporal feature extraction layer utilizes a long short-term memory (LSTM) architecture to capture long-term dependencies in the time series. A gating mechanism controls the flow of information to effectively extract temporal features. The spatial feature fusion layer uses an attention mechanism to perform a weighted fusion of features from different spatial locations, enhancing the model's focus on key areas. The output layer maps the fused features into predicted ecological damage intensity values ​​through a fully connected network.

[0061] The steps for establishing the model training dataset include four phases: data collection, feature extraction, sample construction, and data partitioning. The data collection phase involved literature research and field surveys to collect multi-temporal remote sensing imagery and corresponding ecological damage assessment reports for 100 representative cases of Spartina alterniflora invasion worldwide, ensuring data representativeness and integrity. The feature extraction phase extracted key features for each case, including a series of changes in the invaded area, data on the loss of various ecological service functions, and information on changes in environmental factors. Principal component analysis was used to reduce feature dimensionality while retaining key information. The sample construction phase organized the extracted feature data into multidimensional data samples containing spatial coordinates, time series, ecological damage intensity, and environmental characteristics. Each sample corresponded to the damage status of a spatiotemporal node. The data partitioning phase involved stratifying the constructed samples by geographic region and climate type to form training, validation, and test sets. The training set accounted for 70% of the model parameter learning, the validation set accounted for 15% of the model hyperparameter tuning and model selection, and the test set accounted for 15% of the final performance evaluation to ensure the model's generalization and predictive accuracy.

[0062] It should be noted that the specific implementation of the present invention comprehensively adopts the following three key technologies: the combination of multi-temporal remote sensing data and invasion rate prediction model, the spatiotemporal damage analysis model based on graph convolutional network, and the integrated application of biological inhibitor effect model.

[0063] First, the combination of multi-temporal remote sensing data and invasion rate prediction models constitutes the core technical basis of this method. Traditional ecological damage assessments often rely on static current status survey data, which makes it difficult to accurately capture the dynamic characteristics of the invasion process and future development trends. The present invention establishes a high-resolution remote sensing image database for many consecutive years, combines the least squares method and ARIMA time series analysis, and constructs a mathematical model that can quantitatively predict the invasion spread rate. The technical advantage of this method is that it can extract the temporal evolution law of invasion from historical data, realize scientific prediction of future invasion trends, and provide a quantitative basis for forward-looking ecological protection decisions. Compared with traditional qualitative descriptions or simple linear extrapolation methods, this technical idea significantly improves the time dimension accuracy of the assessment and the reliability of the prediction.

[0064] Secondly, the spatiotemporal damage analysis model based on graph convolutional networks represents an important innovation in the spatial analysis technology of this method. Existing ecological damage assessment methods usually use simple spatial interpolation or statistical analysis techniques, which are difficult to effectively deal with the complex spatial heterogeneity of ecosystems and the nonlinear characteristics of damage propagation. The present invention introduces the graph convolutional network architecture in deep learning, which can adaptively learn damage propagation patterns at different spatial scales through a dynamic neighbor sampling mechanism and an attention mechanism. The technical advantage of this model is that it can simultaneously capture spatial dependencies and time series characteristics, automatically discover hidden spatiotemporal correlation patterns through end-to-end learning, and achieve accurate identification and quantification of the degree of damage in different regions. This intelligent spatial analysis method has stronger adaptability and prediction accuracy than traditional empirical models.

[0065] Third, the integrated application of the biological inhibitor effect model reflects the depth of the method's understanding of ecological mechanisms. Traditional invasive species damage assessments often ignore the natural regulatory mechanisms inherent in ecosystems, resulting in pessimistic assessment results and a lack of targeted management guidance. The present invention systematically analyzes biological inhibitory factors such as local reed population density, natural enemy insect community structure, and soil environmental factors, establishes a mathematical model for quantifying the inhibitory effect, and integrates it into the damage assessment framework. The advantage of this technical approach is that it can more realistically reflect the complexity and self-regulation ability of ecosystems, providing scientific support for the formulation of precise management strategies based on ecological mechanisms.

[0066] The synergy of these three key technical approaches has resulted in a significant amplification of technical impact. The multi-temporal remote sensing prediction model provides a dynamic foundation in the temporal dimension, the graph convolutional network model enables intelligent analysis in the spatial dimension, and the biological inhibitor model provides in-depth understanding of the ecological mechanism dimension. The organic combination of these three forms a comprehensive assessment system encompassing time, space, and ecological mechanisms. Compared to existing technologies, this system achieves a technological leap from static to dynamic, from extensive to precise, and from single to comprehensive, providing a systematic technical solution for the scientific management of the Spartina alterniflora invasion.

[0067] A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions. When the program instructions are run in a computer, the program instructions are used to execute the above-mentioned method for assessing the ecological environment damage value of Spartina alterniflora.

[0068] The third aspect of the present invention provides a system for assessing the ecological environmental damage value of Spartina alterniflora, comprising the above-mentioned computer-readable storage medium. The system is any one of a computer, a server, and a single-chip microcomputer. The computer-readable storage medium is arranged in the system, and the system is provided with a microprocessor for executing program instructions stored in the computer-readable storage medium.

[0069] Specifically, the principle of the present invention is as follows: the key to the present invention's ability to solve the technical problem of insufficient analysis of the interaction of multiple ecological factors lies in its construction of a systematic multi-factor coupling analysis framework and an accurate interaction modeling mechanism. First, by establishing a multi-temporal remote sensing image database and an invasion rate prediction model, this method can simultaneously capture the invasion dynamics of Spartina alterniflora, environmental change trends, and ecosystem response processes, providing comprehensive data support for multi-factor interaction analysis. Secondly, by constructing a damage assessment index system that includes water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource supply functions, combined with multi-constrained optimization solution of the damage coefficient matrix, this method can quantify the correlation and complementarity between different ecological service functions and accurately reflect the synergistic effects of multiple damage factors. Thirdly, the spatiotemporal damage analysis model based on the graph convolutional network architecture can effectively handle the complex interaction relationships between nodes in the ecosystem through the neighbor sampling mechanism and message passing algorithm, and realize accurate modeling of the interaction mechanism of multiple ecological factors. The ecological damage diffusion equation can describe the coupling law of various factors in the damage propagation process by integrating multi-dimensional parameters such as invasion source intensity, environmental resistance coefficient, and ecosystem sensitivity index. Finally, the biotic inhibitor effect model accurately assesses the combined effects of different inhibitory factor combinations on the invasion process by quantifying the synergistic effects of multiple inhibitory mechanisms, including competition from native reeds, control by natural enemies, and salt stress. The logical rationality of this technical solution lies in its transformation of complex multi-factor interactions into a computable mathematical model. Through multi-level data fusion and intelligent optimization algorithms, it effectively characterizes and accurately quantifies the complexity of the ecosystem.

[0070] A specific embodiment 1 of the present invention is provided below. The specific implementation of each step in this embodiment 1 is described in detail as follows.

[0071] The specific implementation of step S01 is to obtain high-resolution image data of a continuous time series of the study area through a multi-source satellite remote sensing platform. First, vegetation information is extracted from the remote sensing image using the normalized vegetation index calculation formula. The normalized vegetation index calculation formula is as follows:

[0072] ;

[0073] Where, is the normalized difference vegetation index; is the reflectivity in the near-infrared band; is the reflectivity in the red light band.

[0074] The calculation formula of normalized water index is as follows:

[0075] ;

[0076] Where, is the normalized water index; is the reflectivity of green light band; is the reflectivity in the near-infrared band.

[0077] The coverage area of ​​Spartina alterniflora was extracted using the threshold segmentation method, and the calculation formula is as follows:

[0078] ;

[0079] Where, For the Annual Spartina alterniflora coverage area; For the The recognition probability of Spartina alterniflora in pixels; is the area of ​​a single pixel; is the total number of pixels in the study area. Calculated by support vector machine classifier, Determined according to the image spatial resolution, when the spatial resolution is 10m, .

[0080] The specific implementation of step S02 is to calculate the invasion and diffusion rate parameter based on the extracted time series data of the coverage area of ​​Spartina alterniflora. The invasion rate parameter calculation formula is as follows:

[0081] ;

[0082] Where, For the invasion rate during the period; For the the annual coverage of Spartina alterniflora; For the the annual coverage of Spartina alterniflora; is the time interval.

[0083] The invasion rate prediction model was fitted using the least squares method to establish the following linear regression equation:

[0084] ;

[0085] Where, To predict the invasion rate; is the intercept parameter; is the coefficient of the first-order term; is the coefficient of the quadratic term; is the time variable; is the prediction error term, and its value range is 0.05~0.15. The objective function is obtained by solving the problem using the least squares method:

[0086] ;

[0087] Where, is the total number of observation periods.

[0088] The specific implementation of step S03 is to construct a loss assessment indicator system covering multiple ecological service functions. The unit area value of each ecological service function is quantified using the following comprehensive assessment formula:

[0089] ;

[0090] Where, For the Total value of regional ecological service functions; For the The ecological service function is The weight coefficient of the region; For the The unit area value of each ecological service function; For the Area of ​​the region; They represent the functions of water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource supply respectively.

[0091] The ecological damage diffusion equation adopts the principle of diffusion reaction equation, and the specific expression is as follows:

[0092] ;

[0093] Where, is the ecological damage intensity; For time; is the environmental resistance coefficient, ranging from 0.1 to 0.5; is the Laplace operator; is the intrusion source intensity, ranging from 0.2 to 0.8; for ecosystem carrying capacity; It is the ecosystem sensitivity index, ranging from 0.3 to 0.7.

[0094] The specific implementation of step S04 is to determine the quantitative parameters of the degree of damage to various ecological service functions caused by the invasion of Spartina alterniflora through field surveys and comparative experiments. The damage coefficient matrix expression is as follows:

[0095] ;

[0096] Where, is the damage coefficient matrix; For the Region No. The damage coefficient of each ecological service function ranges from 0.1 to 0.9; is the number of study areas.

[0097] The objective function of the multi-objective linear programming optimization model is:

[0098] ;

[0099] Constraints include:

[0100] ;

[0101] ;

[0102] ;

[0103] Where, For the The upper limit of the total regional damage coefficient ranges from 2.0 to 4.0; It is the constraint on the difference of adjacent function damage coefficients and takes the value of 0.2.

[0104] The specific implementation of step S05 is to establish a weight correction model that takes into account the cumulative effect of time. The time weight factor calculation formula is as follows:

[0105] ;

[0106] Where, For the Time weight factor of the period; is the time decay parameter, ranging from 0.05 to 0.15; is the cumulative effect coefficient, ranging from 0.1 to 0.3; The duration of the invasion.

[0107] The calculation formula of the comprehensive evaluation model of ecological value is as follows:

[0108] ;

[0109] Where, The value of comprehensive ecological and environmental damage; For the Importance weight coefficient of each ecological service function; For the The regional ecological sensitivity factor of each function ranges from 0.5 to 1.5; For the The function is in The intensity of damage during the period; For the The area invaded by Spartina alterniflora during the period.

[0110] The specific implementation of step S06 is to calculate the dynamic damage value of each period based on the above model. The calculation formula of the ecological damage value per unit area is as follows:

[0111] ;

[0112] Where, For the The ecological damage value per unit area during the period; For the Period Damage coefficient of each function; For the The unit area value of each function; For the Period The correction factor of the function ranges from 0.8 to 1.2.

[0113] The calculation formula for the damage value sequence after dynamic revision is as follows:

[0114] ;

[0115] Where, For the The period-adjusted value of damage per unit area; It is the correction coefficient of the spatiotemporal damage analysis model, ranging from 0.9 to 1.1.

[0116] The specific implementation of step S07 is to quantify the effect of the biological inhibitor on the invasion of Spartina alterniflora. The formula for calculating the competitive inhibition coefficient of reed is as follows:

[0117] ;

[0118] Where, is the competitive inhibition coefficient of reed; is the population density of local reed, in plants / square meter; is the upper limit of environmental load capacity; is the niche competition intensity parameter, ranging from 0.01 to 0.05.

[0119] The formula for calculating the natural enemy insect control coefficient is as follows:

[0120] ;

[0121] Where, is the natural enemy insect control coefficient; For the Control efficiency of natural enemy insects; For the Population density of natural enemy insects; For the The activity factor of natural enemy insects ranges from 0.6 to 1.0; They represent the Spartina alterniflora sawfly, Spartina alterniflora borer, and Spartina alterniflora aphid respectively.

[0122] The calculation formula of salt stress inhibition coefficient is as follows:

[0123] ;

[0124] Where, is the salt stress inhibition coefficient; is the soil salinity concentration in g / kg; The salt tolerance threshold of Spartina alterniflora is 3 g / kg; It is a parameter of salt stress sensitivity, ranging from 0.1 to 0.3.

[0125] The specific implementation of step S08 is to establish a comprehensive assessment model for the overall damage value of the region. The calculation formula for the overall ecological environmental damage value of the region is as follows:

[0126] ;

[0127] Where, The value of damage to the overall ecological environment of the region; is the comprehensive suppression coefficient, and the calculation formula is:

[0128] .

[0129] The specific implementation of the optional step S09 is to establish an optimization configuration model of the inhibition factor based on the principle of cost-benefit analysis. The optimization objective function is:

[0130] ;

[0131] Where, For the cost of reed planting; the cost of stocking natural enemy insects; Cost of regulating soil salinity; Stocking density for natural enemy insects; To regulate the amount of soil salinity; The value of the damage after the implementation of the restraining measures.

[0132] Constraints include:

[0133] ;

[0134] ;

[0135] ;

[0136] ;

[0137] In the formula, the units of each parameter are plant / square meter, head / square meter, and gram / kilogram respectively.

[0138] It should be noted that in this embodiment, the normalized vegetation index formula is Based on the spectroscopic principle of differences in vegetation reflectance in the near-infrared and red bands, a standardized process eliminates interference from atmospheric influences and changing lighting conditions. Compared to traditional visual interpretation methods, this formula automates and standardizes Spartina alterniflora identification, significantly improving the consistency and comparability of multi-temporal monitoring and laying the technical foundation for establishing an accurate invasion dynamics database.

[0139] Normalized Water Index formula Using the differences in the spectral characteristics of water in the green and near-infrared bands, this method effectively distinguishes the boundaries between water and vegetation in coastal wetlands. This formula, combined with a vegetation index, improves the accuracy of Spartina alterniflora identification in complex coastal environments compared to single spectral index methods, resolving the technical challenge of ambiguous identification at the water-land interface caused by traditional methods.

[0140] Calculation formula for the coverage area of ​​Spartina alterniflora Using the mathematical principle of probability weighted summation, pixel-level recognition probabilities are converted into regional-scale area statistics. Compared to traditional hard classification methods, this formula can handle mixed pixel problems, improves the accuracy of area calculations, and provides more reliable basic data for subsequent invasion rate parameter calculations.

[0141] Intrusion rate calculation formula Using differential calculus, the temporal rate of spread of Spartina alterniflora was quantified. Compared to static area statistics, this formula provides the first quantitative description of the invasion dynamics. This provides key parameters for predictive model construction and temporal analysis of damage assessment, addressing the limitations of existing technologies, which lack dynamic monitoring capabilities.

[0142] Invasion rate prediction model The principle of quadratic polynomial regression is used to capture the nonlinear temporal variation of the invasion rate. Compared to linear prediction methods, this model can reflect the acceleration or deceleration trends during the invasion process, improving the accuracy of medium- and long-term predictions and providing a scientific basis for forward-looking damage assessment and the formulation of preventive measures.

[0143] Comprehensive evaluation formula for the value of ecological service functions Based on multi-objective decision-making theory, different types of ecological services are uniformly quantified into economic values. Compared to single-function assessment methods, this formula comprehensively considers multiple ecosystem services and provides a systematic framework for comprehensively assessing the ecological and economic losses caused by the invasion of Spartina alterniflora.

[0144] Ecological damage diffusion equation The reaction-diffusion theory describes the propagation mechanism of damage in space and time. Compared to static damage assessment methods, this equation introduces the spatial propagation and temporal evolution dynamics of damage for the first time. It can predict the spread and intensity of damage, providing theoretical guidance for the precise spatial configuration of remediation measures.

[0145] Time weight factor formula Combining exponential decay and logarithmic growth mathematical forms reflects the compound time effects of ecological damage. Compared with the equal-weighted accumulation method, this formula more accurately depicts the ecosystem's response to long-term invasions, reflects the irreversibility and cumulative nature of ecological damage, and improves the temporal precision of damage value assessment.

[0146] Formula for competitive inhibition coefficient of reed The inhibitory effect of native vegetation is quantified based on the principles of niche theory and interspecific competition dynamics. Compared to qualitative descriptions, this formula, for the first time, quantitatively expresses the inhibitory effect of organisms. It provides a mathematical tool for optimizing the density of native vegetation and significantly improves the scientific nature and accuracy of ecological restoration measures.

[0147] Formula for natural enemy insect control coefficient Based on the theory of predator-prey dynamics, this formula comprehensively considers the synergistic control effects of multiple natural enemies. Compared to single-enemy assessment methods, this formula systematically quantifies the control capacity of insect communities and provides a theoretical basis for the diversified deployment of biological control strategies.

[0148] Regional total damage value formula This comprehensive calculation combines time-series damage values ​​with spatial area and biological inhibition effects. Compared to traditional static assessment methods, this formula dynamically adjusts damage values ​​and quantitatively deducts inhibition effects. This provides a complete assessment framework for developing differentiated governance investment strategies and cost-benefit analysis, significantly improving the accuracy and practicality of ecological and environmental damage assessments.

[0149] In order to better understand and implement the present invention, Example 2 of a specific application scenario of the present invention is provided below: A research team was commissioned to conduct an assessment of the ecological environmental damage value caused by the invasion of Spartina alterniflora in an important wetland reserve. In recent years, the invasion of Spartina alterniflora has become increasingly serious, and there is an urgent need to scientifically assess the damage value caused by it to the ecological environment to provide technical support for the formulation of precise control measures.

[0150] The research team first used a multi-source satellite remote sensing platform to obtain high-resolution image data for the study area for seven consecutive years from 2018 to 2024. Using satellite data, a total of 152 multispectral remote sensing images with a spatial resolution of 10 meters and a temporal resolution of 16 days were obtained. The original images were radiometrically calibrated and atmospherically corrected using an atmospheric correction algorithm to eliminate the effects of atmospheric scattering and absorption. The Normalized Difference Vegetation Index (NDVI) was used. and normalized water index Extract vegetation and water body information. Combined with a support vector machine classifier, the system automatically identifies areas covered by Spartina alterniflora, achieving a classification accuracy of 87.3%, meeting the technical requirement of over 85%.

[0151] By calculating the formula Extract the coverage area of ​​Spartina alterniflora in each year, including The temporal variation data shown in Table 1 indicate that the invasion of Spartina alterniflora is showing a trend of continuous expansion.

[0152] Table 1 Time series data on the coverage area of ​​Spartina alterniflora

[0153]

[0154] Based on the data in Table 1, the invasion rate parameters are calculated using the formula Calculate the invasion rate in each period. Use the least squares method to fit and establish an invasion rate prediction model. ,in , , , model determination coefficient , meeting the technical requirement of no less than 0.8, and the prediction error is controlled at 12.7%, which meets the accuracy requirement of less than 15%.

[0155] A loss assessment indicator system covering five core ecological service functions—water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource provision—was constructed. The per-unit-area value of each function was determined using market value, replacement cost, and contingent value methods, as shown in Table 2.

[0156] Table 2 Quantitative standards for the value of ecological service functions per unit area

[0157]

[0158] Through field surveys, 30 pairs of Spartina alterniflora invasion plots and native vegetation control plots were set up to collect measured data on various ecological service functions. The decline in various functions in the invasion area relative to the control area was calculated, and a damage coefficient matrix was constructed. A multi-objective linear programming optimization model was used, with minimizing the total ecological damage cost as the objective function and setting constraints. , , The optimal damage coefficient allocation scheme is obtained by using the simplex method, as shown in Table 3.

[0159] Table 3 Ecological service function damage coefficient matrix of Spartina alterniflora invasion

[0160]

[0161] Establish a time weight factor correction model, using the formula , where the time decay parameter , cumulative effect coefficient . Construct a comprehensive ecological value assessment model, the calculation formula is , the sum of the weight coefficients of the comprehensive evaluation model is 1.0, which meets the technical requirements.

[0162] Calculate the ecological damage value per unit area in each period based on the invasion rate prediction model and damage coefficient matrix , where the correction factor The value range is 0.85 to 1.15. A spatiotemporal damage analysis model was introduced. This model, based on a graph convolutional network architecture, dynamically adjusts the number of neighbor samples based on the intrusion spread range. When the spread range is less than 500 meters, the number of nodes is set to 10; when the spread range is between 500 and 2000 meters, the number is adjusted to 18; and when the spread range exceeds 2000 meters, the number is increased to 28. The model was trained using data from 100 typical cases, achieving a training accuracy of 83.2%, meeting the technical requirement of over 80%.

[0163] Quantify the effect of biological inhibitors on the invasion of Spartina alterniflora. The local reed population density was determined to be 135 plants / , using the formula Calculate the competitive inhibition coefficient of reed, where , ,get The survey found three main natural enemy insects, using the formula The natural enemy insect control coefficient was calculated to be 0.28. The soil salt concentration was determined to be 5.6 g / kg, using the formula Calculate the salt stress inhibition coefficient, where g / kg, ,get Comprehensive inhibition coefficient .

[0164] Establish a comprehensive assessment model for regional overall damage value, using the formula After accounting for the effects of biological inhibitors, the cumulative ecological and environmental damage value from 2018 to 2024 was 48.267 million yuan, including 13.521 million yuan in water purification function losses, 11.584 million yuan in biodiversity maintenance function losses, 8.673 million yuan in soil conservation function losses, 10.136 million yuan in climate regulation function losses, and 4.353 million yuan in fishery resource supply function losses, as shown in Table 4.

[0165] Table 4 Results of regional ecological environmental damage value assessment

[0166]

[0167] Based on cost-benefit analysis, an optimal configuration model of inhibition factors was established, with the objective function being to minimize the ecological restoration cost and maximize the damage reduction benefit ratio. The constraint condition was set as reed planting density range of 50 to 200 plants / The density of natural enemy insects is 10 to 50 heads / , the soil salt concentration range is 3-8g / kg, and the comprehensive inhibition coefficient is not less than 0.3. The optimal inhibition factor input plan for different damage levels is determined through a hierarchical optimization algorithm. The recommended reed input density for high damage areas is 180 plants / , the density of natural enemy insects is 45 heads / , soil salinity is adjusted to 7.2g / kg. The recommended reed planting density for moderately damaged areas is 120 plants / , the density of natural enemy insects is 30 heads / , soil salinity is adjusted to 5.8g / kg. The recommended reed planting density in low damage areas is 80 plants / , the density of natural enemy insects is 18 heads / , the soil salinity is adjusted to 4.5g / kg.

[0168] Traditional assessments of Spartina alterniflora ecological damage primarily rely on static assessment methods, estimating damage value based on survey data from a single point in time, lacking a scientific depiction of the dynamic process of invasion. Traditional methods typically employ fixed damage coefficients, ignoring the impact of cumulative effects over time and spatial heterogeneity, while also lacking systematic consideration of biological inhibitors. In terms of assessment accuracy, the uncertainty of traditional methods typically ranges from 35% to 45%, with prediction errors generally exceeding 25%. This present invention achieves significant advancements in multiple aspects compared to traditional methods. First, a dynamic monitoring system based on multi-temporal remote sensing was established. Second, a spatiotemporal damage analysis model was constructed. Third, a biological inhibitor assessment module was introduced, making damage value assessments more comprehensive and accurate, avoiding the problem of traditional methods overestimating damage value by approximately 12% to 20%. Fourth, an inhibitor optimization configuration model was established, providing scientific governance solutions for areas with varying degrees of damage.

[0169] It should be noted that the detailed explanation of the variables involved in the present invention is shown in Table 5.

[0170] Table 5 Variable Explanation Table

[0171]

[0172] The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for assessing the ecological environment damage value of Spartina alterniflora, characterized in that: The following steps are involved: A multi-temporal remote sensing image database of the S. alterniflora invasion area was established, and temporal variation data on the area covered by S. alterniflora were extracted. The invasion rate parameters of S. alterniflora were calculated based on the multi-temporal remote sensing image database, and an invasion rate prediction model was established. An ecosystem service function loss assessment indicator system was constructed, selecting five core ecological service functions: water purification, biodiversity maintenance, soil conservation, climate regulation, and fishery resource provision. The ecological damage diffusion equation was used to calculate the impact propagation mechanism of S. alterniflora invasion on each ecological service function. A damage coefficient matrix for each ecological service function caused by S. alterniflora invasion was established, and the optimal allocation of the damage coefficient matrix was solved based on a multi-constrained optimization problem that minimizes the total ecological damage cost. A time weight factor correction model was constructed, and the ecological value comprehensive assessment model was used to dynamically integrate the damage value of each ecological service function. A biological inhibitor assessment module was introduced, and an inhibitor effect model was established. A comprehensive assessment model for regional ecological environmental damage value was established. Based on the overall ecological environmental damage value and the biological inhibitor effect model, the ecological damage value per unit area in each period was multiplied by the invaded area in the corresponding period. Combined with the correction value of the biological inhibitor effect model, the overall ecological environmental damage value of S. alterniflora invasion in the study area was accumulated.

2. The method according to claim 1, characterized in that The steps to establish a multi-temporal remote sensing image database of the area invaded by Spartina alterniflora are to obtain high-resolution satellite images for more than five consecutive years, extract the temporal change data of the area covered by Spartina alterniflora, and construct a basic dataset for dynamic monitoring of invasion and spread.

3. The method according to claim 2, characterized in that The steps for calculating the invasion rate parameters of Spartina alterniflora based on the multi-temporal remote sensing image database are to establish an invasion rate time series by analyzing the ratio of the change in the coverage area of ​​Spartina alterniflora in adjacent years to the time interval, and to obtain an invasion rate prediction model using the least squares method.

4. The method according to claim 3, characterized in that The ecological damage diffusion equation is used to describe the propagation and diffusion mechanism of ecological function damage caused by the invasion of Spartina alterniflora in spatial and temporal dimensions. The input includes the invasion source intensity, environmental resistance coefficient, ecosystem sensitivity index, diffusion time parameter, and boundary condition parameter. The output is the distribution of ecological function damage intensity at each spatial location at different times.

5. The method according to claim 4, characterized in that The multi-constraint optimization problem belongs to the multi-objective linear programming problem. By establishing a mathematical optimization model with minimizing the total ecological damage cost as the objective function, and with the sum constraint of the damage coefficients of various ecological service functions, the upper and lower limit constraints of the damage coefficients of a single function, and the difference constraint of the damage coefficients of adjacent functions as constraints, the simplex method is used to solve the optimal damage coefficient distribution plan for various ecological service functions.

6. The method according to claim 5, characterized in that Before calculating the ecological damage value per unit area caused by the invasion of Spartina alterniflora in each period based on the invasion rate prediction model and the damage coefficient matrix, it also includes the introduction of a spatiotemporal damage analysis model to accurately identify and quantify the degree of damage in different regions and periods.

7. The method according to claim 6, characterized in that The spatiotemporal damage analysis model is an ecological damage spatiotemporal prediction model built based on a graph convolutional network architecture. Its structure is a deep learning network architecture that includes an input layer, a multi-layer graph convolution layer, a temporal feature extraction layer, a spatial feature fusion layer, and an output layer.

8. The method according to claim 7, characterized in that The biotic inhibitor effect model was used to calculate the inhibitory effects of native reed population density, community structure of natural enemy insects of Spartina alterniflora, and soil salinity on the invasion of Spartina alterniflora. The inputs included native reed population density, community structure of natural enemy insects of Spartina alterniflora, soil salinity, environmental loading capacity, and niche competition intensity.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores program instructions, and when the program instructions are run in a computer, they are used to execute the method for assessing the ecological environment damage value of Spartina alterniflora according to any one of claims 1 to 8.

10. A system for assessing the ecological damage value of Spartina alterniflora, characterized in that: The computer-readable storage medium according to claim 9 is included, the system is any one of a computer, a server, and a single-chip microcomputer, the computer-readable storage medium is arranged in the system, and the system is provided with a microprocessor for executing program instructions stored in the computer-readable storage medium.

Images