Method and system for evaluating influence of extreme climate on ecological toughness of wetland

Abstract

The invention discloses a method and system for evaluating the influence of extreme climate on wetland ecological toughness, and the method comprises the steps: obtaining multi-source extreme climate data, carrying out the scale alignment, inputting a Bayesian hierarchical fusion model, and obtaining extreme climate fusion data, the method comprises the following steps: calculating comprehensive pressure indexes of different time and space, carrying out wetland state dynamic simulation to update a wetland state vector, calculating a time-varying toughness index, carrying out graded ecological early warning according to the time-varying toughness index, calculating the correlation between the time-varying toughness index and the wetland state index, and carrying out causal correlation. And determining ecological regulation and control measures according to the causal association result and the graded ecological early warning result. The method not only can improve the efficiency and accuracy of evaluating the influence of the extreme climate on the wetland ecological toughness, but also has better interpretability, and can be directly applied to a system for evaluating the influence of the extreme climate on the wetland ecological toughness.

Description

Technical Field

[0001] This invention relates to the field of ecological monitoring and disaster early warning technology, and in particular to an evaluation method and system for the impact of extreme climate on wetland ecological resilience. Background Technology

[0002] Wetland ecosystems are crucial ecological barriers on Earth, playing an irreplaceable role in regulating climate, conserving water resources, and protecting biodiversity. In recent years, against the backdrop of global climate change, extreme weather events have become more frequent and intense, posing a serious threat to the stability and resilience (i.e., ecological resilience) of wetland ecosystems. Therefore, scientifically and accurately assessing the impact of extreme climate events on wetland ecological resilience and achieving forward-looking early warning and regulation has become a key scientific issue that urgently needs to be addressed in ecological protection and adaptive management.

[0003] Current wetland ecological resilience assessment technologies suffer from three main shortcomings: First, the assessment paradigm is static, failing to characterize the nonlinear response of ecological variables to extreme pressures, critical threshold effects, and recovery lags. Second, data fusion is disconnected from process simulation, failing to consider wetland surface heterogeneity and ecological state feedback as constraints, resulting in a lack of causal correlation between fusion results and ecological process mechanisms. Finally, the decision support system is outdated, lacking a real-time causal inference and measure effectiveness prediction mechanism for "pressure-state-resilience," making it difficult to achieve a closed loop of "monitoring-early warning-regulation." Therefore, this invention proposes an evaluation method and system for the impact of extreme climate on wetland ecological resilience. By achieving alignment and bidirectional constraints between climate and ecological spatiotemporal scales through Bayesian hierarchical fusion, a set of stochastic differential equations with pressure-driven heteroscedasticity and multi-scale lag is constructed. A three-dimensional time-varying resilience index of hydrology, biology, and function is designed, and intelligent regulation and matching are carried out based on information entropy causal correlation. This shifts response measures from empirical selection to probabilistic optimal decision-making. This technological breakthrough upgrades wetland ecological resilience assessment from static ex-post evaluation to dynamic ex-ante early warning, providing core technical support for improving the adaptability of wetlands to climate change and the level of precise management. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for evaluating the impact of extreme climate on wetland ecological resilience.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution: This invention includes the following steps: Multi-source extreme climate data are acquired, scale-aligned, and then input into a Bayesian hierarchical fusion model to perform spatiotemporal fusion with wetland status to obtain extreme climate fused data. The comprehensive pressure index at different times and spaces is calculated by using the three-dimensional features of extreme climate fusion data. The comprehensive pressure index at different times and spaces is then used to dynamically simulate the wetland state and update the wetland state vector. The dynamic simulation of the wetland state is carried out through a set of stochastic differential equations for stochastic dynamic evolution. The time-varying resilience index is calculated based on the updated wetland state vector, and a graded ecological early warning is carried out based on the time-varying resilience index. The time-varying resilience index reflects the sensitivity of wetland ecology to extreme climate, including the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index. The correlation between the time-varying resilience index and wetland status indicators is calculated to establish a causal relationship. Based on the causal relationship results and the results of hierarchical ecological early warning, ecological regulation measures are determined.

[0006] Furthermore, the method for obtaining extreme climate fusion data includes: Multi-source extreme climate data is acquired and scale-aligned; the multi-source extreme climate data includes global climate model data, satellite remote sensing data, ground-based fixed-point monitoring data, and social perception data. The specific steps for scale alignment are as follows: drive a high-resolution regional climate model to increase the resolution of global climate model data to be consistent with satellite remote sensing data; use multivariate statistical relationships and geographic weighted regression models to downscale global climate model data and satellite remote sensing data to be consistent with ground-based fixed-point monitoring data; and resample all data to the same spatiotemporal grid framework. Scale-aligned multi-source extreme climate data and wetland prior knowledge are input into a Bayesian hierarchical fusion model for spatiotemporal fusion to obtain extreme climate fused data; the wetland prior knowledge includes initial estimates of wetland state and uncertainties. The Bayesian hierarchical fusion model comprises an input layer, a fusion layer, and an output layer. The fusion layer quantifies the uncertainty of each data source based on Bayesian statistical theory, dynamically weights and fuses them to generate the optimal estimate of the true wetland state, and includes an observation model unit, a process model unit, and a parameter model unit. The observation model unit describes the relationship between the observed values ​​of each data source and the true state through a joint likelihood function. The process model unit uses a Gaussian process to construct the prior distribution of the wetland state. The parameter model unit sets the prior distribution for the parameters of each layer. The joint likelihood function expression is: ; ; ; ; in Real wetland conditions All observations The joint likelihood function, For location index, For time indexing, for Extreme climate observations from data sources For the number of data sources, This is a set of spatiotemporal monitoring points for the region to be merged. , The local expected value and estimation uncertainty of extreme climate observations are labeled with differences. for Dynamic weights of the data source for The inherent credibility parameters of the data source, , This is the smoothing coefficient.

[0007] Furthermore, the method for calculating the comprehensive pressure index in different times and spaces includes: By analyzing the three-dimensional characteristics of statistically fused extreme climate data, the comprehensive pressure index at different times and spaces is calculated, expressed as: ; in for Moment Wetland The overall pressure index at the location. The number of extreme event categories, Weights for extreme event types, As the attenuation factor, The number of events within the time window. This is the cumulative effect coefficient. , , For the first The real-time intensity, frequency, and duration of extreme events. , The average of the frequency and duration of the baseline period. For the first Extreme event-like wetland spatial heterogeneous weight field The term is a random disturbance that follows a spatiotemporal heteroscedastic distribution.

[0008] Furthermore, the dynamic simulation of the wetland state is performed through a set of stochastic differential equations to achieve stochastic dynamic evolution; The stochastic differential equations are described by the Iton process of the following form, which describes the stochastic dynamics of the wetland state vector: ; in This is the wetland state vector. For time indexing, These are deterministic ecological process functions, belonging to the category of ordinary differential equations. This is the intensity function of the pressure-driven random perturbation, belonging to the diffusion term of the stochastic differential equation. For the spatial heterogeneous weight field of wetlands, It is a delayed response function, belonging to the time-delay differential equation. As a composite pressure vector, For time delay parameters; The wetland state vector Including water level, vegetation cover, soil organic carbon, and biodiversity index; the comprehensive pressure vector Including flood pressure, drought pressure, and high temperature pressure; The deterministic ecological process function is based on a logistic growth term, combined with a pressure inhibition term in the form of a Hill function and an ecological coupling feedback term to update the wetland state. The expression is as follows: ; in The intrinsic growth rate As an indicator of wetland status, It is a set of wetland status indicators. For environmental carrying capacity, The stress sensitivity coefficient, This is the conversion factor. This is the pressure tolerance vector. Wetland status indicators and The coupling feedback term; The expression for the pressure-driven random disturbance intensity function is: ; in Based on the intensity of fluctuations, it characterizes the inherent randomness within an ecosystem in the absence of extreme weather conditions. This is the pressure amplification factor. This is the pressure reference value. It is a state-dependent index; The expression for the hysteresis response function is: ; in For maximum memory duration, To fix the lag time, This is the memory decay time constant.

[0009] Furthermore, the method for conducting tiered ecological early warning includes: The time-varying toughness index is calculated based on the updated wetland state vector, and the expression is: ; in The time-varying toughness index, Let be the expected vector of wetland state variables. Let be the standard deviation of the wetland state variables. For comprehensive pressure With wetland status The correlation coefficient, The threshold for the critical correlation coefficient; The time-varying resilience index includes the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index; the hydrological time-varying resilience index is determined by water level updates and vegetation cover updates; the biological time-varying resilience index is determined by biodiversity index updates; and the system function time-varying resilience index is determined by vegetation cover updates, soil organic carbon updates, and biodiversity index updates. Determine the primary and secondary resilience indices, and conduct graded ecological early warning based on the time-varying resilience indices.

[0010] Furthermore, the method for determining ecological regulation measures includes: The mutual information entropy between each time-varying resilience index and the corresponding wetland state index is calculated as the data correlation degree. The two sets of wetland state indices with the highest data correlation degree are selected as the influencing factors of the corresponding time-varying resilience index. Ecological regulation measures are determined based on the results of hierarchical ecological early warning and the influencing factors.

[0011] Secondly, an evaluation system for the impact of extreme climate on wetland ecological resilience includes: Cache module: used to extract raw mass spectrometry spectrum data and corresponding multi-source mass spectrometry metadata, perform layered processing on the multi-source mass spectrometry metadata, load the same timestamp and instrument unique identifier for each layer of mass spectrometry metadata and the corresponding spectrum data, and package and partition them to be stored in the metadata cache pool. Preliminary fusion module: This module is used to input the mass spectrometry metadata of each layer into the dual-channel adaptive fusion module to obtain adaptive fusion weights, and then weight and fuse the mass spectrometry metadata of each layer to obtain preliminary fusion metadata. Spatiotemporal verification module: used to construct a Markov random field undirected graph based on the initial fused metadata node set, iteratively calculate the global consistency score through the belief propagation algorithm, and perform spatiotemporal consistency verification on the initial fused metadata; Quality assessment module: used to calculate the quality transfer entropy of the preliminary fusion metadata that has passed the spatiotemporal consistency check, and guide the fusion strategy based on the quality transfer entropy to re-layer the corresponding mass spectrometry metadata and output reliable metadata; Storage module: Used to spatially align and compress the trusted metadata and the corresponding original mass spectrometry data using the tensor ring decomposition operator, bind them to generate a unified data package and store it in the fusion database.

[0012] The beneficial effects of this invention are: This invention provides a method and system for evaluating the impact of extreme climates on wetland ecological resilience. Compared with existing technologies, this invention has the following technical advantages: This invention significantly improves the spatiotemporal accuracy and ecological relevance of extreme climate pressure fields by spatiotemporally aligning multi-source extreme climate data and embedding wetland state feedback constraints into a Bayesian hierarchical fusion model, effectively solving the problem of pressure underestimation caused by scale mismatch. This invention employs a set of stochastic differential equations coupled with four-dimensional state variables: water level, vegetation, soil carbon, and biodiversity. It introduces a pressure-driven heteroscedastic term and a multi-timescale lag feedback function, which can characterize the critical slowdown and state transition phenomena under extreme climates. It can realize nonlinear stochastic dynamic simulation and threshold early warning of wetland ecological state. This invention establishes a three-dimensional time-varying resilience index system of hydrological resilience (stability-resilience), biological resilience (structure-adaptability), and functional resilience (productivity-sustainability). Combined with a probabilistic three-level early warning mechanism, it can identify the bottleneck effect and synergistic degradation patterns of different ecological components, and realize the process-oriented assessment of ecological resilience. This invention automatically identifies key influencing factors and determines control strategies by calculating the mutual information entropy between the time-varying resilience index and state indicators. It solves the key technical problems that have long existed in wetland ecological resilience assessment, such as insufficient accuracy, lack of dynamism, and lagging control, and provides an integrated technical solution that is precise, forward-looking, and operable for wetland ecosystems to cope with extreme climate risks. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating the steps of an evaluation method for the impact of extreme climate on wetland ecological resilience according to the present invention. Detailed Implementation

[0014] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0015] The present invention provides a method and system for evaluating the impact of extreme climate on wetland ecological resilience, comprising the following steps: like Figure 1 As shown, this embodiment includes the following steps: Multi-source extreme climate data are acquired, scale-aligned, and then input into a Bayesian hierarchical fusion model to perform spatiotemporal fusion with wetland status to obtain extreme climate fused data. The comprehensive pressure index at different times and spaces is calculated by using the three-dimensional features of extreme climate fusion data. The comprehensive pressure index at different times and spaces is then used to dynamically simulate the wetland state and update the wetland state vector. The dynamic simulation of the wetland state is carried out through a set of stochastic differential equations for stochastic dynamic evolution. The time-varying resilience index is calculated based on the updated wetland state vector, and a graded ecological early warning is carried out based on the time-varying resilience index. The time-varying resilience index reflects the sensitivity of wetland ecology to extreme climate, including the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index. The correlation between the time-varying resilience index and wetland status indicators is calculated to establish a causal relationship. Based on the causal relationship results and the results of hierarchical ecological early warning, ecological regulation measures are determined.

[0016] In this embodiment, the method for obtaining extreme climate fusion data includes: Multi-source extreme climate data is acquired and scale-aligned; the multi-source extreme climate data includes global climate model data, satellite remote sensing data, ground-based fixed-point monitoring data, and social perception data. The specific steps for scale alignment are as follows: drive a high-resolution regional climate model to increase the resolution of global climate model data to be consistent with satellite remote sensing data; use multivariate statistical relationships and geographic weighted regression models to downscale global climate model data and satellite remote sensing data to be consistent with ground-based fixed-point monitoring data; and resample all data to the same spatiotemporal grid framework. Scale-aligned multi-source extreme climate data and wetland prior knowledge are input into a Bayesian hierarchical fusion model for spatiotemporal fusion to obtain extreme climate fused data; the wetland prior knowledge includes initial estimates of wetland state and uncertainties. The Bayesian hierarchical fusion model comprises an input layer, a fusion layer, and an output layer. The fusion layer quantifies the uncertainty of each data source based on Bayesian statistical theory, dynamically weights and fuses them to generate the optimal estimate of the true wetland state, and includes an observation model unit, a process model unit, and a parameter model unit. The observation model unit describes the relationship between the observed values ​​of each data source and the true state through a joint likelihood function. The process model unit uses a Gaussian process to construct the prior distribution of the wetland state. The parameter model unit sets the prior distribution for the parameters of each layer. The joint likelihood function expression is: ; ; ; ; in Real wetland conditions All observations The joint likelihood function, For location index, For time indexing, for Extreme climate observations from data sources For the number of data sources, This is a set of spatiotemporal monitoring points for the region to be merged. , The local expected value and estimation uncertainty of extreme climate observations are labeled with differences. for Dynamic weights of the data source for The inherent credibility parameters of the data source, , For smoothing coefficients; In practical assessments, multi-source extreme climate data includes global climate model data, satellite remote sensing data, ground-based fixed-point monitoring data, and social perception data. Global climate model data is acquired through the CMIP6 global climate model, including future scenarios for temperature, precipitation, and extreme indices (such as the number of consecutive drought days). Satellite remote sensing data is acquired through Sentinel-1 SAR, including surface temperature and flooding frequency. Ground-based fixed-point monitoring data is acquired through meteorological stations / hydrological stations / ecological monitoring stations, including measured temperature, precipitation, wind speed, groundwater level, and water quality. Social perception data is acquired through social media, news reports, and historical disaster records, including the spatiotemporal distribution and impact reports of extreme climate events. The spatial nature of social perception data is ambiguous, and its temporal nature is driven by the time of extreme climate events, exhibiting a lag. It is primarily used for coarse verification of global climate model data, satellite remote sensing data, and ground-based fixed-point monitoring data within the corresponding spatiotemporal region (to determine if they conform to the corresponding spatiotemporal distribution and impact reports). During scale alignment, a high-resolution regional climate model (RCM) is driven to upscale the resolution of the global climate model data (GCM) (100 km) to match the resolution of satellite remote sensing data (1-10 km). This step mainly addresses the interaction between large-scale circulation and regional topography and land-sea distribution. Multivariate statistical relationships and geographically weighted regression models are used to downscale the resolution of the global climate model data and satellite remote sensing data (1-10 km) to match the resolution of ground-based fixed-point monitoring data (10 m). All data are uniformly resampled to the same spatiotemporal grid framework (10-meter spatial resolution, ten-day time step). For irregular data, spatiotemporal kriging interpolation is used. The expression for the multivariate statistical relationship and geographically weighted regression model is as follows:

[0017] in Location of target wetland patch In time Downscaled climate elements It is a nonlinear mapping function that can be obtained by training through machine learning algorithms (such as random forests and neural networks) (establishing the relationship between large-scale climate fields and local geographical features). To correspond to climate elements at low-resolution (1-10 km) grid points over time, For position High-resolution geographic feature vectors, including digital elevation models, slope and aspect, land use type, and distance from water bodies (these are key factors affecting local climate). This is the residual term (representing random variation not explained by the model); The output of the Bayesian hierarchical fusion model is extreme climate fusion data and the corresponding fusion uncertainty. Taking extreme precipitation data as an example, global climate model data, satellite remote sensing data, and ground-based fixed-point monitoring data all contain specific precipitation data. After spatiotemporal fusion by the Bayesian hierarchical fusion model, the three sets of precipitation data are simplified into one set of precipitation data. The wetland status of the corresponding spatiotemporal points is embedded into the precipitation data and verified by social perception data (e.g., a sudden increase in rainfall reported in a certain place, traffic congestion / visible water accumulation reflected on social media, etc., to verify whether the set of precipitation data conforms to this pattern). Other extreme climate data are fused in the same way, and finally, extreme climate fusion data is obtained.

[0018] In this embodiment, the method for calculating the comprehensive pressure index in different times and spaces includes: By analyzing the three-dimensional characteristics of statistically fused extreme climate data, the comprehensive pressure index at different times and spaces is calculated, expressed as: ; in for Moment Wetland The overall pressure index at the location. The number of extreme event categories, Weights for extreme event types, As the attenuation factor, The number of events within the time window. This is the cumulative effect coefficient. , , For the first The real-time intensity, frequency, and duration of extreme events. , The average of the frequency and duration of the baseline period. For the first Extreme event-like wetland spatial heterogeneous weight field The term represents a random disturbance that follows a spatiotemporal heteroscedastic distribution. In actual assessments, extreme event categories include floods, droughts, high temperatures, and typhoons, and the weights of extreme event types are determined through inversion of historical disaster data. The average frequency and duration of the 30-year climatological baseline period are taken; The spatial heterogeneous weighted field of wetlands characterizes the two-dimensional spatial distribution of the differences in the response of wetland interior space to extreme climate pressures, quantifying the coordinates. The deviation of the actual impact intensity of a unit intensity extreme climate event on the wetland ecosystem from the regional average is taken as an example at (11,201) in the low-lying area / dense vegetation area / peat marshland area. The average water accumulation in the wetland system caused by 1 mm / h rainfall is 0.4 mm, and the standard weight is 1. The water accumulation at this location is 0.8 mm, indicating that this location is highly sensitive to extreme rainfall climate. The corresponding spatial heterogeneity weight is 1.5. Similarly, the average decrease in water content in the wetland system caused by 1℃ temperature rise is 0.5%, and the standard weight is 1. The water content at this location decreases by 0.2%, indicating that this location is not highly sensitive to extreme high temperature climate. The corresponding spatial heterogeneity weight is 0.75.

[0019] In this embodiment, the dynamic simulation of wetland state is performed through a set of stochastic differential equations to achieve stochastic dynamic evolution; The stochastic differential equations are described by the Iton process of the following form, which describes the stochastic dynamics of the wetland state vector: ; in This is the wetland state vector. For time indexing, These are deterministic ecological process functions, belonging to the category of ordinary differential equations. This is the intensity function of the pressure-driven random perturbation, belonging to the diffusion term of the stochastic differential equation. For the spatial heterogeneous weight field of wetlands, It is a delayed response function, belonging to the time-delay differential equation. As a composite pressure vector, For time delay parameters; The wetland state vector Including water level, vegetation cover, soil organic carbon, and biodiversity index; the comprehensive pressure vector Including flood pressure, drought pressure, and high temperature pressure; The deterministic ecological process function is based on a logistic growth term, combined with a pressure inhibition term in the form of a Hill function and an ecological coupling feedback term to update the wetland state. The expression is as follows: ; in The intrinsic growth rate As an indicator of wetland status, It is a set of wetland status indicators. For environmental carrying capacity, The stress sensitivity coefficient, This is the conversion factor. This is the pressure tolerance vector. Wetland status indicators and The coupling feedback term; The expression for the pressure-driven random disturbance intensity function is: ; in Based on the intensity of fluctuations, it characterizes the inherent randomness within an ecosystem in the absence of extreme weather conditions. This is the pressure amplification factor. This is the pressure reference value. It is a state-dependent index; The expression for the hysteresis response function is: ; in For maximum memory duration, To fix the lag time, This is the memory decay time constant; In actual assessment, a comprehensive pressure field is constructed based on the comprehensive pressure index at different times and spaces, and a three-dimensional pressure vector is extracted at a specific wetland patch to obtain the comprehensive pressure vector. Among them, the spatially continuous downscaling result of the coarse-scale climate grid of the comprehensive pressure index (such as GCM 100km×100km) at the wetland patch scale (10m×10m) is a scalar field, and the comprehensive pressure vector is a three-dimensional vector obtained by decoupling the comprehensive pressure index according to the extreme event type at a specific wetland monitoring point or representative patch, which is the real-time input of the state evolution model. Among them, flood stress characterizes the comprehensive effects of flooding stress on wetland plant root hypoxia, soil anaerobicness, and microbial metabolic inhibition; drought stress characterizes the effects of water deficit on plant stomatal closure, photosynthesis inhibition, soil respiration reduction, and biological habitat shrinkage; and high temperature stress characterizes the effects of heat stress on protein denaturation, enzyme activity inhibition, net photosynthetic rate reduction, and accelerated soil carbon decomposition. Wetland status indicators and The coupling feedback relationships include: (1) positive regulation of vegetation cover by water level, inhibition and decomposition of soil organic carbon, and habitat filtration for biodiversity; (2) regulation feedback of vegetation cover to water level, carbon source input for soil organic carbon, and microhabitat creation for biodiversity; (3) fertility feedback of soil organic carbon to vegetation cover and resource support for biodiversity; (4) engineering promotion of vegetation cover by biodiversity and carbon regulation of soil organic carbon. In the stress-driven random disturbance intensity function, the baseline fluctuation intensity characterizes the inherent randomness within the ecosystem when there are no extreme climates (such as daily evapotranspiration fluctuations and biological activities), the state dependence index reflects the buffering capacity of large populations / high state variables to random disturbances (the "dilution effect" in ecology), and the stress baseline value is usually taken as the 30-year climatological average. In the lag response function, the fixed lag time (lag of major physiological processes) is set to 15 days according to the plant physiological memory (short-term memory window), the maximum memory duration (the longest time the system can remember history) is set to 180 days according to the interannual dynamics of the seed bank (long-term memory window), and the memory decay time constant is set to 30 days (reflecting the half-life of plant stress memory). The lag response terms include vegetation lag recovery, water level lag, soil carbon lag, and biodiversity lag related to wetland status indicators.

[0020] In this embodiment, the method for conducting tiered ecological early warning includes: The time-varying toughness index is calculated based on the updated wetland state vector, and the expression is: ; in The time-varying toughness index, Let be the expected vector of wetland state variables. Let be the standard deviation of the wetland state variables. For comprehensive pressure With wetland status The correlation coefficient, The threshold for the critical correlation coefficient; The time-varying resilience index includes the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index; the hydrological time-varying resilience index is determined by water level updates and vegetation cover updates; the biological time-varying resilience index is determined by biodiversity index updates; and the system function time-varying resilience index is determined by vegetation cover updates, soil organic carbon updates, and biodiversity index updates. Determine the primary and secondary resilience indices, and conduct graded ecological early warning based on the time-varying resilience indices; In practical assessments, the expected vector of wetland state variables in the expression for the time-varying resilience index is... The standard deviation of wetland state variables reflects the health level of wetlands. Reflects the stability of wetlands; The wetland ecosystem is categorized and warned according to the time-varying resilience index. A yellow warning is issued when the time-varying resilience index is less than the first-level resilience index (the 25th percentile of the historical time-varying resilience index), indicating that extreme weather has a slight impact on the wetland's ecological resilience, and the wetland ecosystem can recover on its own. An orange warning is issued when the time-varying resilience index is greater than the second-level resilience index (the 75th percentile of the historical time-varying resilience index), indicating that extreme weather has a significant impact on the wetland's ecological resilience, and the wetland ecosystem needs to be restored through artificial means. A red warning is issued when the time-varying resilience index is greater than 1.5 times the second-level resilience index, indicating that the wetland's ecological resilience is weak, it is extremely vulnerable to collapse due to extreme weather, and recovery is difficult.

[0021] In this embodiment, the method for determining ecological regulation measures includes: The mutual information entropy between each time-varying resilience index and the corresponding wetland state index is calculated as the data correlation degree. The two sets of wetland state indices with the highest data correlation degree are selected as the influencing factors of the corresponding time-varying resilience index. Ecological regulation measures are determined based on the results of hierarchical ecological early warning and the influencing factors. In the actual assessment, the Banghu sub-lake system of the Poyang Lake National Nature Reserve (with an area of ​​12.7 km²) was selected. 2 (A typical saucer-shaped depression wetland, dominated by reed and sedge communities). Based on the continuous heavy rain event from July 15 to 18 (total rainfall of 320 mm), the hydrological time-varying resilience index, biological time-varying resilience index, and system function time-varying resilience index were calculated according to the above steps and were 2.7, 2.44, and 0.66, respectively. Taking the first-level resilience index and the second-level resilience index as 1.2 and 2.2, the hydrological time-varying resilience index, biological time-varying resilience index, and system function time-varying resilience index correspond to red warning, orange warning, and yellow warning, respectively. The mutual information entropy between the hydrological time-varying resilience index 2.7 and wetland state indicators was calculated, and the water level (1.85), vegetation cover (1.38), and biodiversity (1.2), which had the highest data correlation, were selected as the influencing factors of the hydrological time-varying resilience index. The mutual information entropy between the biological time-varying resilience index 2.44 and wetland state indicators was calculated, and the vegetation cover (1.72), soil organic carbon (1.51), and flood pressure (1.45), which had the highest data correlation, were selected as the influencing factors of the hydrological time-varying resilience index. The mutual information entropy between the system function time-varying resilience index 0.66 and wetland state indicators was calculated, and the biodiversity (1.43), vegetation cover (1.35), and water level (1.28), which had the highest data correlation, were selected as the influencing factors of the hydrological time-varying resilience index. The cumulative impact values ​​of water level (9.38), vegetation cover (8.81), biodiversity (4.18), and soil organic carbon (3.68) were calculated by multiplying the correlation between the time-varying resilience index and each wetland state index. Based on these cumulative impact values, ecological regulation was implemented: lowering the water level to a safe threshold to reduce inundation stress; air-dropping reed rhizomes (density 2000 plants / hectare) by drone to increase vegetation cover in shallow water areas after water recedes (H<1.0 m); and constructing ecological islands and reefs in the receding water area to provide shelter for organisms.

[0022] Secondly, an evaluation system for the impact of extreme climate on wetland ecological resilience includes: Fusion module: Used to scale-align multi-source extreme climate data, input Bayesian hierarchical fusion model and wetland status for spatiotemporal fusion to obtain extreme climate fused data; State update module: used to calculate the comprehensive pressure index in different times and spaces through the three-dimensional features of extreme climate fusion data, and update the wetland state vector by dynamically simulating the comprehensive pressure index in different times and spaces; the wetland state dynamic simulation is carried out through stochastic dynamic evolution by a set of stochastic differential equations. The graded early warning module is used to calculate the time-varying resilience index based on the updated wetland state vector and to conduct graded ecological early warning based on the time-varying resilience index. The time-varying resilience index reflects the sensitivity of wetland ecology to extreme climate and includes the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index. Regulation module: Calculates the correlation between time-varying resilience index and wetland status indicators to establish causal relationships, and determines ecological regulation measures based on the causal relationship results and hierarchical ecological early warning results.

[0023] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the impact of extreme climate on wetland ecological resilience, characterized in that, Includes the following steps: S1. Acquire multi-source extreme climate data, perform scale alignment, input the Bayesian hierarchical fusion model and wetland status for spatiotemporal fusion to obtain extreme climate fusion data; S2. Calculate the comprehensive pressure index in different times and spaces using the three-dimensional features of extreme climate fusion data, and update the wetland state vector by dynamically simulating the comprehensive pressure index in different times and spaces; the dynamic simulation of wetland state is carried out through a set of stochastic differential equations for stochastic dynamic evolution. S3. Calculate the time-varying resilience index based on the updated wetland state vector, and conduct graded ecological early warning based on the time-varying resilience index; the time-varying resilience index reflects the sensitivity of wetland ecology to extreme climate, including hydrological time-varying resilience index, biological time-varying resilience index and system function time-varying resilience index. S4. Calculate the correlation between the time-varying resilience index and wetland status indicators to establish a causal relationship, and determine ecological regulation measures based on the causal relationship results and the results of hierarchical ecological early warning.

2. The method for evaluating the impact of extreme climate on wetland ecological resilience according to claim 1, characterized in that, The method for obtaining extreme climate fusion data includes: Multi-source extreme climate data is acquired and scale-aligned; the multi-source extreme climate data includes global climate model data, satellite remote sensing data, ground-based fixed-point monitoring data, and social perception data. The specific steps for scale alignment are as follows: drive a high-resolution regional climate model to increase the resolution of global climate model data to be consistent with satellite remote sensing data; use multivariate statistical relationships and geographic weighted regression models to downscale global climate model data and satellite remote sensing data to be consistent with ground-based fixed-point monitoring data; and resample all data to the same spatiotemporal grid framework. Scale-aligned multi-source extreme climate data and wetland prior knowledge are input into a Bayesian hierarchical fusion model for spatiotemporal fusion to obtain extreme climate fused data; the wetland prior knowledge includes initial estimates of wetland state and uncertainties. The Bayesian hierarchical fusion model comprises an input layer, a fusion layer, and an output layer. The fusion layer quantifies the uncertainty of each data source based on Bayesian statistical theory, dynamically weights and fuses them to generate the optimal estimate of the true wetland state, and includes an observation model unit, a process model unit, and a parameter model unit. The observation model unit describes the relationship between the observed values ​​of each data source and the true state through a joint likelihood function. The process model unit uses a Gaussian process to construct the prior distribution of the wetland state. The parameter model unit sets the prior distribution for the parameters of each layer. The joint likelihood function expression is: ； ； ； ； in Real wetland conditions All observations The joint likelihood function, For location index, For time indexing, for Extreme climate observations from data sources For the number of data sources, This is a set of spatiotemporal monitoring points for the region to be merged. , The local expected value and estimation uncertainty of extreme climate observations are labeled with differences. for Dynamic weights of the data source for The inherent credibility parameters of the data source, , This is the smoothing coefficient.

3. The method for evaluating the impact of extreme climate on wetland ecological resilience according to claim 1, characterized in that, The method for calculating the comprehensive pressure index in different times and spaces includes: By analyzing the three-dimensional characteristics of statistically fused extreme climate data, the comprehensive pressure index at different times and spaces is calculated, expressed as: ； in for Moment Wetland The overall pressure index at the location. The number of extreme event categories, Weights for extreme event types, As the attenuation factor, The number of events within the time window. This is the cumulative effect coefficient. , , For the first The real-time intensity, frequency, and duration of extreme events. , The average of the frequency and duration of the baseline period. For the first Extreme event-like wetland spatial heterogeneous weight field The term is a random disturbance that follows a spatiotemporal heteroscedastic distribution.

4. The method for evaluating the impact of extreme climate on wetland ecological resilience according to claim 1, characterized in that, The dynamic simulation of wetland status is performed through a set of stochastic differential equations to simulate stochastic dynamic evolution. The stochastic differential equations are described by the Iton process of the following form, which describes the stochastic dynamics of the wetland state vector: ； in This is the wetland state vector. For time indexing, These are deterministic ecological process functions, belonging to the category of ordinary differential equations. This is the intensity function of the pressure-driven random perturbation, belonging to the diffusion term of the stochastic differential equation. For the spatial heterogeneous weight field of wetlands, It is a delayed response function, belonging to the time-delay differential equation. As a composite pressure vector, For time delay parameters; The wetland state vector Including water level, vegetation cover, soil organic carbon, and biodiversity index; the comprehensive pressure vector Including flood pressure, drought pressure, and high temperature pressure; The deterministic ecological process function is based on a logistic growth term, combined with a pressure inhibition term in the form of a Hill function and an ecological coupling feedback term to update the wetland state. The expression is as follows: ； in The intrinsic growth rate As an indicator of wetland status, It is a set of wetland status indicators. For environmental carrying capacity, The stress sensitivity coefficient, This is the conversion factor. This is the pressure tolerance vector. Wetland status indicators and The coupling feedback term; The expression for the pressure-driven random disturbance intensity function is: ； in Based on the intensity of fluctuations, it characterizes the inherent randomness within an ecosystem in the absence of extreme weather conditions. This is the pressure amplification factor. This is the pressure reference value. It is a state-dependent index; The expression for the hysteresis response function is: ； in For maximum memory duration, To fix the lag time, This is the memory decay time constant.

5. The method for evaluating the impact of extreme climate on wetland ecological resilience according to claim 1, characterized in that, The method for conducting tiered ecological early warning includes: The time-varying toughness index is calculated based on the updated wetland state vector, and the expression is: ； in The time-varying toughness index, Let be the expected vector of wetland state variables. Let be the standard deviation of the wetland state variables. For comprehensive pressure With wetland status The correlation coefficient, The threshold for the critical correlation coefficient; The time-varying resilience index includes the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index; the hydrological time-varying resilience index is determined by water level updates and vegetation cover updates; the biological time-varying resilience index is determined by biodiversity index updates; and the system function time-varying resilience index is determined by vegetation cover updates, soil organic carbon updates, and biodiversity index updates. Determine the primary and secondary resilience indices, and conduct graded ecological early warning based on the time-varying resilience indices.

6. The method for evaluating the impact of extreme climate on wetland ecological resilience according to claim 1, characterized in that, The method for determining ecological regulation measures includes: The mutual information entropy between each time-varying resilience index and the corresponding wetland state index is calculated as the data correlation degree. The two sets of wetland state indices with the highest data correlation degree are selected as the influencing factors of the corresponding time-varying resilience index. Ecological regulation measures are determined based on the results of hierarchical ecological early warning and the influencing factors.

7. An evaluation system for the impact of extreme climate on wetland ecological resilience, used to implement the method according to any one of claims 1-6, characterized in that, include: Fusion module: Used to scale-align multi-source extreme climate data, input Bayesian hierarchical fusion model and wetland status for spatiotemporal fusion to obtain extreme climate fused data; State update module: used to calculate the comprehensive pressure index in different times and spaces through the three-dimensional features of extreme climate fusion data, and update the wetland state vector by dynamically simulating the comprehensive pressure index in different times and spaces; the wetland state dynamic simulation is carried out through stochastic dynamic evolution by a set of stochastic differential equations. The graded early warning module is used to calculate the time-varying resilience index based on the updated wetland state vector and to conduct graded ecological early warning based on the time-varying resilience index. The time-varying resilience index reflects the sensitivity of wetland ecology to extreme climate and includes the hydrological time-varying resilience index, the biological time-varying resilience index, and the system function time-varying resilience index. Regulation module: Calculates the correlation between time-varying resilience index and wetland status indicators to establish causal relationships, and determines ecological regulation measures based on the causal relationship results and hierarchical ecological early warning results.

Images