A method and related apparatus for assessing coastal vegetation damage and recovery after extreme storms

Abstract

This application discloses a method and related equipment for assessing coastal vegetation damage and recovery after extreme storms, relating to the field of data processing technology. The method includes: constructing a synchronous grid stack; mapping four states—storm, moisture, scattering, and damage—to a set of nodes, and constructing a pixel-level community lag causality graph based on each node; solving for the expected value of water film scattering from the pixel-level community lag causality graph, and using the expected value of water film scattering to determine the net damage sequence; dividing the area into ecological zones of different environments, calculating scattering values ​​in each ecological zone to obtain the natural fluctuation envelope; encapsulating the net damage sequence, the natural fluctuation envelope, and the edge weights and confidence factors of pixel-level moisture and scattering in the pixel-level community lag causality graph as game inputs and inputting them to the parameter player, thereby outputting multiple parameters of vegetation damage and recovery, and determining the peak damage rate, half-recovery time, and full recovery time of the vegetation. This application can improve the accuracy of vegetation damage and recovery assessment.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and related equipment for assessing coastal vegetation damage and recovery after extreme storms. Background Technology

[0002] Current technical approaches for monitoring coastal vegetation damage after typhoon landfall largely rely on the parallel use of optical and microwave remote sensing. A common practice is to first calculate vegetation indices such as NDVI and EVI using multi-temporal Sentinel-2 or Landsat images, then perform differential analysis between pre- and post-disaster images to delineate suspected damaged areas. Next, C-band Sentinel-1 dual-polarization scattering intensity is used to compensate for cloud cover and imaging time differences. Finally, the optical threshold and radar amplitude ratio are superimposed to generate a first-order damage distribution map. Both types of images require geometric correction at the sub-pixel level, and comparable spatial benchmarks are achieved through topographic correction and unified projection.

[0003] Existing technologies generally lack a unified spatiotemporal reference and a traceable causal chain, making it impossible to simultaneously characterize the multi-level hysteresis relationship between storm intensity, surface moisture, radar scattering, and vegetation damage at the pixel level. They also lack an explicit stripping mechanism for water film scattering interference, making it difficult to obtain reliable quantitative indicators of damage-recovery under high humidity conditions. Summary of the Invention

[0004] The main objective of this application is to propose a method and related equipment for assessing coastal vegetation damage and recovery after extreme storms, so as to improve the accuracy of vegetation damage and recovery assessment.

[0005] To achieve the above objectives, one aspect of this application proposes a method for assessing coastal vegetation damage and recovery after extreme storms, the method comprising the following steps:

[0006] By integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain, a synchronous grid stack is constructed.

[0007] The four states of storm, moisture, scattering and damage in the synchronous grid stack are mapped to a set of nodes, and a pixel-level cluster hysteresis causal graph is constructed based on each node in the set of nodes.

[0008] The expected value of water film scattering is obtained by solving the pixel-level community hysteresis causality graph, and the net damage sequence is determined by using the expected value of water film scattering.

[0009] Based on the synchronous grid stack, ecological zones of different environments are obtained, and scattering values ​​are calculated in each ecological zone to obtain the natural wave envelope;

[0010] The net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal graph, and the confidence factor are encapsulated as game inputs and input to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery.

[0011] Based on various parameters of vegetation damage and recovery, the peak damage rate, half-recovery time, and full recovery time of the vegetation are output.

[0012] In some embodiments, the process of integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common cell domain and time domain to construct a synchronous raster stack includes the following steps:

[0013] Acquire recorded data for a first set number of consecutive days before and after the storm's landfall date; wherein, the recorded data includes radar images from dual-polarization synthetic aperture radar, soil moisture grids, typhoon intensity grids, and damage records from field sample plots;

[0014] A uniform daily-scale timescale is appended to the recorded data;

[0015] The radar image is sequentially subjected to orbit refinement, digital elevation model correction, radiometric calibration and slant range projection, and then reprojected to a unified reference frame after conversion.

[0016] Based on the radar pixel resolution, the typhoon intensity grid and the soil moisture grid are jointly resampled using topographic weights and wind field weights; wherein, the weighting factors are automatically allocated under the dual constraints of elevation difference and radial distance of typhoon wind field;

[0017] The time series is completed using the observation dates of the field sample plots as anchor points, and rigid body correction is used to ensure that the geographic coordinates of the same pixel remain consistent across all dates;

[0018] By combining daily tide levels and incident angle thresholds to set quality labels, storm intensity, moisture content, VV polarization scattering, VH polarization scattering, and damage index are integrated in the common pixel domain and time domain to obtain the synchronous grid stack.

[0019] In some embodiments, mapping the four states—storm, moisture, scattering, and damage—in the synchronization grid stack to a set of nodes includes the following steps:

[0020] A community response prior library is generated using historical typhoon cases, and a three-level lag window for the community is obtained based on the community response prior library.

[0021] Based on the principles of pixel consistency, community consistency, and lag matching, the four states of storm, moisture, scattering, and damage in the synchronous grid stack are mapped to the node set.

[0022] The step of constructing a pixel-level community hysteresis causality graph based on each node in the node set includes the following steps:

[0023] The positive edge weights corresponding to storm and moisture, moisture and scattering, and scattering and damage are initialized based on the lag correlation coefficient, normalized, and then truncated according to the prior.

[0024] When the probability of residual tide level or foam coating meets the set conditions, add the negative edge of scattering and damage;

[0025] The pixel-level cluster hysteresis causal graph is obtained by iteratively updating the edge weights through gradient descent along the time axis until the pixel residual variance is lower than the threshold.

[0026] In some embodiments, obtaining the expected value of water film scattering based on the pixel-level community hysteresis causality graph includes the following steps:

[0027] Simultaneously extract the VV polarization scattering, the VH polarization scattering, the water content, and the edge weights of water content and scattering obtained from the pixel-level community hysteresis causality graph for each pixel;

[0028] The expected value of water film scattering is obtained by combining the edge weights of water and scattering with the water and scattering gain function;

[0029] The determination of the net damage sequence using the expected value of water film scattering includes the following steps:

[0030] The structural scattering residual is obtained by weighting the dual-polarization scattering energy and subtracting the expected value of the water film scattering.

[0031] After applying a two-day window first-order Savitzky-Golay filter to the structural scattering residuals, the mean residual value of the second set number of days before the storm landslide is extracted as the baseline and normalized to obtain the net damage sequence.

[0032] In some embodiments, the step of dividing the ecological zones of different environments according to the synchronization grid stack includes the following steps:

[0033] Based on the synchronous raster stack, the elevation, slope aspect, conductivity, and community type raster are read in pixel alignment to form an environment vector;

[0034] The environmental vectors are divided into ecological zones for different environments by progressive similarity splitting clustering based on weighted Mahalanobis distance;

[0035] The process of calculating scattering values ​​in each ecological zone to obtain the natural wave envelope includes the following steps:

[0036] Extract dual-polarized radar images of the same month as the target storm from stormless images of the most recently set number of years to form a stormless comparison set;

[0037] The natural wave envelope is obtained by calculating the 5th and 95th percentiles of the scattering values ​​in each of the ecological zones according to the calendar day based on the storm-free control set.

[0038] Abnormal curves that have been manually verified are removed from the natural fluctuation envelope.

[0039] In some embodiments, the process of using the parameters to output various parameters of vegetation damage and recovery by the player includes the following steps:

[0040] Using these parameters, players generate an initial simulation curve based on damage peak, latency, and recovery rate;

[0041] The weighted residual scores are calculated based on the confidence level and envelope out-of-bounds conditions of the initial simulation curve using residual players.

[0042] Using the parameters, the player corrects the parameters according to the vector level rules until the change value of the weighted residual score is lower than the error threshold or reaches the preset maximum round, and outputs the recovery amplitude, latency, recovery rate and residual index as multiple parameters of the vegetation damage and recovery.

[0043] In some embodiments, the step of outputting the peak damage rate, half-recovery time, and full recovery time of vegetation based on multiple parameters of vegetation damage and recovery includes the following steps:

[0044] The time axis of various parameters related to vegetation damage and recovery was corrected based on on-site water level records;

[0045] The various parameters of vegetation damage and recovery are mapped to piecewise exponential curves, and the peak damage rate is obtained by normalizing the peak daily envelope amplitude.

[0046] The half-recovery time is obtained by locating the date on which net damage decreases to half of the peak value from the peak date.

[0047] The full recovery time is recorded when the net damage is below the upper limit of the envelope for the third consecutive set number of days.

[0048] Output the peak damage rate, the half-recovery time, and the full recovery time.

[0049] To achieve the above objectives, another aspect of this application proposes a device for assessing coastal vegetation damage and recovery after extreme storms, the device comprising:

[0050] The data acquisition unit is used to integrate storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common cell domain and time domain to construct a synchronous raster stack.

[0051] The graph construction unit is used to map the four states of storm, moisture, scattering and damage in the synchronous grid stack into a set of nodes, and to construct a pixel-level community hysteresis causal graph based on each node in the set of nodes.

[0052] The damage determination unit is used to obtain the expected value of water film scattering based on the pixel-level community hysteresis causality graph, and to determine the net damage sequence using the expected value of water film scattering.

[0053] The natural wave envelope determination unit is used to divide the ecological zones of different environments according to the synchronous grid stack, calculate the scattering value in each ecological zone, and then obtain the natural wave envelope.

[0054] The parameter fitting unit is used to encapsulate the net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal map, and the confidence factor as game inputs and input them to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery.

[0055] The damage and recovery assessment unit is used to output the peak damage rate, half recovery time, and full recovery time of the vegetation based on various parameters of the vegetation damage and recovery.

[0056] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0057] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0058] The embodiments of this application include at least the following beneficial effects:

[0059] This application provides a method and related equipment for assessing coastal vegetation damage and recovery after extreme storms. The scheme integrates storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in a common pixel domain and time domain to construct a synchronous grid stack. The four states (storm, moisture, scattering, and damage) in the synchronous grid stack are mapped to a set of nodes. A pixel-level community lag causal graph is constructed based on each node in the node set. The expected value of water film scattering is obtained from the pixel-level community lag causal graph, and the net damage sequence is determined using this expected value. Ecological zones with different environments are divided according to the synchronous grid stack, and scattering values ​​are calculated in each ecological zone to obtain the natural fluctuation envelope. The net damage sequence, natural fluctuation envelope, pixel-level moisture and scattering edge weights, and confidence factors in the pixel-level community lag causal graph are encapsulated as game inputs and input to a parameter player. The parameter player outputs various parameters of vegetation damage and recovery. Based on these parameters, the peak damage rate, half-recovery time, and full recovery time of the vegetation are output. This application employs a storm-moisture-scattering synchronous grid stack, which ensures that typhoon intensity, soil moisture, and dual-polarized scattering maintain the same step size and resolution within the pixel domain. This provides an equivalent spatiotemporal coordinate system for subsequent analysis, avoids convolution errors caused by inconsistent resolution of multi-source field quantities, and thus improves the accuracy of vegetation damage and recovery assessment. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 Comparison diagrams between some embodiments of this application and the prior art are provided for illustration purposes;

[0062] Figure 2 A flowchart illustrating a method for assessing coastal vegetation damage and recovery after an extreme storm, provided as an embodiment of this application;

[0063] Figure 3 An example flowchart of a method for assessing coastal vegetation damage and recovery after an extreme storm, provided in an embodiment of this application;

[0064] Figure 4 A flowchart illustrating the processing of a cause-effect graph baseline is provided in this embodiment of the application.

[0065] Figure 5 A flowchart illustrating the process of water film-structure demixing provided in this application embodiment;

[0066] Figure 6 A flowchart illustrating the construction process of an ecological resilience baseline is provided in this application embodiment;

[0067] Figure 7 A schematic diagram of a device for assessing coastal vegetation damage and recovery after an extreme storm, provided in an embodiment of this application;

[0068] Figure 8 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0070] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0071] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0072] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0073] Before providing a detailed description of the embodiments of this application, some related technologies involved in the embodiments of this application will be described first, as follows:

[0074] 1. Parametric Player: In the game theory framework, the parametric player is responsible for generating and updating the simulated damage-recovery curve. Its core task is to vectorize and adjust the ternary parameters (initial damage magnitude A, latency τ, and recovery rate k) based on the current fitting residual, in order to make the new round of simulation more closely resemble the observed sequence.

[0075] 2. Residual Player: The residual player takes on the role of "nitpicking". By calculating the difference between the current simulated curve and the actual net damage sequence, the maximum continuous residual segment is located using the sliding window accumulation method, and a weighted residual score is formed accordingly, providing feedback on the direction and magnitude of correction to the parameter player.

[0076] 3. Confidence Factor: The confidence factor Cp originates from the pixel-level water scattering edge weight mapping and is a quantitative measure of the reliability of the observation of that pixel. In the residual weighting stage, Cp is used to adjust the weights of the residual score and parameter updates, ensuring that high-confidence pixels converge faster and have a greater impact in the game.

[0077] 4. Sliding Window Accumulation Method: Also known as the sliding window accumulation method, this method involves moving a fixed-length window across the residual sequence, accumulating the absolute values ​​of the residuals within the window, and finding the interval with the maximum sum. This method can highlight periods of concentrated continuous errors, helping residual traders accurately locate the curve segments that most need correction.

[0078] 5. Weighted Residual Score: Based on the identified maximum residual segment, the residual player calculates the weighted residual score Sr using the confidence factor and the out-of-bounds penalty index. Sr reflects both the degree to which the fit deviates from the observation and the priority of this deviation in the overall game.

[0079] 6. Game Iteration and Termination Conditions: "Game" refers to the entire process of parameter player and residual player alternately optimizing on the same pixel. After each round, if the residual score and residual variance are both lower than the preset threshold, or the maximum number of iterations is reached, the pixel is determined to have converged and the iteration is terminated.

[0080] To reduce false detections caused by tidal range and turbid water reflection, some existing technologies introduce the MODIS water color index in the preprocessing stage to remove highly turbid water surfaces, and rely on tidal level forecast models or actual water level records to mask flooding time slices; on the radar side, transient specular scattering pixels are directly set to invalid. Subsequently, all indicators are resampled to a 30m or 10m regular grid, and the effects of slope aspect and shadow are adjusted with the help of digital elevation models to ensure consistent topographic reference across multi-source data.

[0081] Time-series analysis often employs exponential smoothing or BreaksForAdditiveSeasonandTrend models to capture abrupt drops in the annual vegetation index series, treating typhoons as short-cycle negative jumps. Subsequently, the recovery slope is used to fit the ecological resilience curve, and the half-recovery time is calculated. Since this approach assumes a single index can describe damage and recovery, post-regression correction using field quadrat breakage rates is often necessary to improve numerical accuracy.

[0082] On the other hand, some studies, starting from the physical scattering mechanism, use the coherence coefficient of dual-polarized radar to infer the direction of collapse through volumetric scattering and specular scattering; others are based on the changes in coherence of interferometric SAR, interpreting the decrease in coherence as structural damage. These methods require sufficiently short imaging intervals to maintain interferometric quality and are easily affected by rapid moisture fluctuations, so they are often combined with SMAP or ASCAT soil moisture grids for regression denoising.

[0083] In recent years, machine learning approaches have been explored: researchers compile data on wind speed, tide levels, soil moisture, radar scattering, and optical indices to construct pixel-level feature vectors, which are then input into random forests or convolutional neural networks to output damage probabilities or levels. Training relies on historical typhoon samples, while inference utilizes current day's satellite quick-scan imagery for near real-time assessment. This approach offers greater flexibility in high-dimensional feature fusion but has limited interpretability, requiring continuous supplementation of labeled data to maintain generalization ability.

[0084] Problems with existing technology:

[0085] Current pre- and post-disaster differential methods mostly rely on the simple superposition of optical vegetation indices and radar amplitudes. The capture of temporal features usually remains at the empirical level of single-point sudden drops and exponential rebounds. They lack independent characterization of the rapidly changing soil moisture content and leaf free water film in the days following storm landfall, making it difficult to distinguish between the two distinct physical signals of "high humidity-enhanced scattering" and "structural breakage attenuation." The same pixel often exhibits increased scattering during the water film effect stage but correspondingly weakened scattering during the breakage stage. If existing models use a uniform threshold, it is easy to lead to both false positives and false negatives.

[0086] While the idea of ​​separating volume scattering with interferometric coherence or dual-polarization coherence coefficient is closer to the scattering mechanism, it is highly sensitive to imaging intervals, post-rain moisture disturbances, and the stability of the interferometric baseline. Once a long baseline is encountered or the geoelectric constant changes drastically after heavy rain, the coherence drops significantly, making the results lack spatiotemporal continuity.

[0087] Data-driven methods, such as machine learning, demonstrate flexibility in multi-source feature fusion, but they heavily rely on the completeness of the annotations of historical typhoon samples. When training data is insufficient or the ecological community changes, the model performance deteriorates sharply. Moreover, black-box decision-making makes it difficult to incorporate prior knowledge such as typhoon intensity, tide level, and micro-topography, resulting in a lack of interpretable physical coherence in the output.

[0088] Existing technologies generally lack a unified spatiotemporal reference and a traceable causal chain, making it impossible to simultaneously characterize the multi-level hysteresis relationship between storm intensity, surface moisture, radar scattering, and vegetation damage at the pixel level. They also lack an explicit stripping mechanism for water film scattering interference, making it difficult to obtain reliable quantitative indicators of damage-recovery under high humidity conditions.

[0089] The embodiments of this application aim to solve the technical problem that it is difficult to accurately extract the true structural damage of coastal vegetation and its subsequent recovery dynamics under the high humidity and strong tidal disturbance environment after the landfall of an extreme storm. The specific goal is to establish an interpretable, quantifiable storm-water-scattering causal chain with community-specific lag at the pixel scale, and on this basis, strip the water film scattering gain to obtain a temporally continuous and physically consistent vegetation damage-recovery index.

[0090] Exemplary examples are shown in the following embodiments compared with the prior art. Figure 1 As shown.

[0091] This application provides a method and related equipment for assessing coastal vegetation damage and recovery after extreme storms, relating to the field of data processing technology. The method and related equipment provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle-mounted terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing a method for assessing coastal vegetation damage and recovery after extreme storms, but is not limited to the above forms.

[0092] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0093] Reference Figure 2 This application provides a method for assessing coastal vegetation damage and recovery after extreme storms. This method may include, but is not limited to, steps S200 to S250, as follows:

[0094] S200: Synchronous raster stack is constructed by integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering and damage index in the common pixel domain and time domain.

[0095] S210: Map the four states of storm, moisture, scattering and damage in the synchronization grid stack to a set of nodes, and construct a pixel-level community hysteresis causal graph based on each node in the set of nodes;

[0096] S220: Obtain the expected value of water film scattering based on the pixel-level community hysteresis causality graph, and use the expected value of water film scattering to determine the net damage sequence;

[0097] S230: Based on the synchronous grid stack, ecological zones of different environments are obtained, and the scattering value is calculated in each ecological zone to obtain the natural wave envelope;

[0098] S240: The net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal graph, and the confidence factor are encapsulated as game inputs and input to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery.

[0099] S250: Outputs the peak damage rate, half-recovery time, and full recovery time of vegetation based on various parameters of vegetation damage and recovery.

[0100] Optionally, the step of integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain to construct a synchronous raster stack includes the following steps:

[0101] Acquire recorded data for a first set number of consecutive days before and after the storm's landfall date; wherein, the recorded data includes radar images from dual-polarization synthetic aperture radar, soil moisture grids, typhoon intensity grids, and damage records from field sample plots;

[0102] A uniform daily-scale timescale is appended to the recorded data;

[0103] The radar image is sequentially subjected to orbit refinement, digital elevation model correction, radiometric calibration and slant range projection, and then reprojected to a unified reference frame after conversion.

[0104] Based on the radar pixel resolution, the typhoon intensity grid and the soil moisture grid are jointly resampled using topographic weights and wind field weights; wherein, the weighting factors are automatically allocated under the dual constraints of elevation difference and radial distance of typhoon wind field;

[0105] The time series is completed using the observation dates of the field sample plots as anchor points, and rigid body correction is used to ensure that the geographic coordinates of the same pixel remain consistent across all dates;

[0106] By combining daily tide levels and incident angle thresholds to set quality labels, storm intensity, moisture content, VV polarization scattering, VH polarization scattering, and damage index are integrated in the common pixel domain and time domain to obtain the synchronous grid stack.

[0107] Optionally, mapping the four states—storm, moisture, scattering, and damage—in the synchronization grid stack to a set of nodes includes the following steps:

[0108] A community response prior library is generated using historical typhoon cases, and a three-level lag window for the community is obtained based on the community response prior library.

[0109] Based on the principles of pixel consistency, community consistency, and lag matching, the four states of storm, moisture, scattering, and damage in the synchronous grid stack are mapped to the node set.

[0110] The step of constructing a pixel-level community hysteresis causality graph based on each node in the node set includes the following steps:

[0111] The positive edge weights corresponding to storm and moisture, moisture and scattering, and scattering and damage are initialized based on the lag correlation coefficient, normalized, and then truncated according to the prior.

[0112] When the probability of residual tide level or foam coating meets the set conditions, add the negative edge of scattering and damage;

[0113] The pixel-level cluster hysteresis causal graph is obtained by iteratively updating the edge weights through gradient descent along the time axis until the pixel residual variance is lower than the threshold.

[0114] Optionally, obtaining the expected value of water film scattering based on the pixel-level community hysteresis causality graph includes the following steps:

[0115] Simultaneously extract the VV polarization scattering, the VH polarization scattering, the water content, and the edge weights of water content and scattering obtained from the pixel-level community hysteresis causality graph for each pixel;

[0116] The expected value of water film scattering is obtained by combining the edge weights of water and scattering with the water and scattering gain function;

[0117] The determination of the net damage sequence using the expected value of water film scattering includes the following steps:

[0118] The structural scattering residual is obtained by weighting the dual-polarization scattering energy and subtracting the expected value of the water film scattering.

[0119] After applying a two-day window first-order Savitzky-Golay filter to the structural scattering residuals, the mean residual value of the second set number of days before the storm landslide is extracted as the baseline and normalized to obtain the net damage sequence.

[0120] Optionally, the step of dividing the ecological zones into different environments based on the synchronization grid stack includes the following steps:

[0121] Based on the synchronous raster stack, the elevation, slope aspect, conductivity, and community type raster are read in pixel alignment to form an environment vector;

[0122] The environmental vectors are divided into ecological zones for different environments by progressive similarity splitting clustering based on weighted Mahalanobis distance;

[0123] The process of calculating scattering values ​​in each ecological zone to obtain the natural wave envelope includes the following steps:

[0124] Extract dual-polarized radar images of the same month as the target storm from stormless images of the most recently set number of years to form a stormless comparison set;

[0125] The natural wave envelope is obtained by calculating the 5th and 95th percentiles of the scattering values ​​in each of the ecological zones according to the calendar day based on the storm-free control set.

[0126] Abnormal curves that have been manually verified are removed from the natural fluctuation envelope.

[0127] Optionally, the step of using the parameters to output various parameters of vegetation damage and recovery by the player includes the following steps:

[0128] Using these parameters, players generate an initial simulation curve based on damage peak, latency, and recovery rate;

[0129] The weighted residual scores are calculated based on the confidence level and envelope out-of-bounds conditions of the initial simulation curve using residual players.

[0130] Using the parameters, the player corrects the parameters according to the vector level rules until the change value of the weighted residual score is lower than the error threshold or reaches the preset maximum round, and outputs the recovery amplitude, latency, recovery rate and residual index as multiple parameters of the vegetation damage and recovery.

[0131] Optionally, the step of outputting the peak damage rate, half-recovery time, and full recovery time of vegetation based on multiple parameters of vegetation damage and recovery includes the following steps:

[0132] The time axis of various parameters related to vegetation damage and recovery was corrected based on on-site water level records;

[0133] The various parameters of vegetation damage and recovery are mapped to piecewise exponential curves, and the peak damage rate is obtained by normalizing the peak daily envelope amplitude.

[0134] The half-recovery time is obtained by locating the date on which net damage decreases to half of the peak value from the peak date.

[0135] The full recovery time is recorded when the net damage is below the upper limit of the envelope for the third consecutive set number of days.

[0136] Output the peak damage rate, the half-recovery time, and the full recovery time.

[0137] The following sections will provide a detailed description and explanation of some optional embodiments of this application, using specific application examples.

[0138] Reference Figure 3 This embodiment provides a method for assessing coastal vegetation damage and recovery after extreme storms based on scene coupling causal unmixing and enhanced game theory optimization, including the following steps S1~S6:

[0139] S1. Construct a synchronous raster stack for storm-moisture-scattering, specifically: acquire dual-polarization synthetic aperture radar images, passive microwave soil moisture raster, typhoon center intensity log, and field quadrat damage records for 30 consecutive days before and after the storm's landfall date, and add a unified daily-scale timescale to all data; perform orbit refinement, digital elevation model correction, radiometric calibration, and slant range projection conversion on the radar images in sequence, and then reproject them to a unified reference frame; perform topographic weighting and wind field weighting joint resampling on the typhoon intensity raster and soil moisture raster according to the radar pixel resolution, where the weighting factors are automatically allocated under the dual constraints of elevation difference and typhoon wind field radial distance; complete the time series with the quadrat observation date as the anchor point, and use rigid body correction to ensure that the same pixel maintains consistent geographic coordinates on all dates; set quality labels by combining daily tide level and incident angle thresholds, and finally integrate storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain to obtain a synchronous raster stack.

[0140] S2. Construct a pixel-level community lag causal graph, specifically: Use historical typhoon cases to generate a community response prior library to obtain a community-specific three-level lag window; Map the four states—storm, moisture, scattering, and damage—to node sets according to the principles of pixel consistency, community consistency, and lag matching; Initialize the three positive edge weights (storm-moisture, moisture-scattering, and scattering-damage) based on the lag correlation coefficient, normalize them, and then truncate them according to the prior; Add a scattering-damage suppression edge and assign negative weights when the probability of residual tide level or foam cover meets the set conditions; Then, traverse along the time axis using a sliding window, iteratively updating each edge weight through gradient descent until the pixel residual variance is below a threshold. For example, Figure 4 This embodiment provides a flowchart for processing a causal graph baseline.

[0141] S3. Water film-structure scattering demixing: Specifically, for each pixel, VV scattering, VH scattering, surface moisture, and the water-scattering edge weights obtained in step S2 are simultaneously extracted. The edge weights are combined with the water-scattering gain function to obtain the expected value of water film scattering. The dual-polarization scattering energy is weighted and the expected value of water film scattering is subtracted to obtain the structure scattering residual. A two-day window first-order Savitzky-Golay filter is applied to the structure scattering residual, and the mean residual value for the seven days prior to storm landfall is extracted as the baseline and normalized to obtain the net damage sequence. For example, Figure 5 This embodiment provides a process flow diagram for water film-structure demixing.

[0142] S4. Construct an ecological resilience baseline, specifically: Collect raster data of elevation, slope aspect, electrical conductivity, and community type according to pixel alignment to form an environmental vector; use weighted Mahalanobis distance-based progressive similarity splitting clustering to divide ecological zones; extract dual-polarized radar images of the same month as the target storm from storm-free imagery of the past ten years to form a storm-free control set; calculate the 5th and 95th percentiles of scattering values ​​for each ecological zone according to calendar day to obtain the natural fluctuation envelope, and remove manually verified anomalous curves. For example, Figure 6 This embodiment provides a flowchart for constructing an ecological resilience baseline.

[0143] S5. Fitting vegetation damage-recovery curves based on enhanced game theory: The net damage sequence from step S3, the natural fluctuation envelope from step S4, the pixel-level water-scattering edge weights, and the confidence factor are encapsulated as game inputs. The parameter player generates an initial simulation curve based on the damage peak, latency, and recovery rate. The residual player calculates the weighted residual score based on the confidence level and envelope out-of-bounds conditions. The parameter player corrects the parameters according to vector-level rules until the residual score change is lower than the error threshold or reaches the preset maximum round. The recovery amplitude, latency, recovery rate, and residual index are then output.

[0144] S6. Calculate the damage-recovery index, specifically: based on the corrected time axis of the on-site water level record, map the parameters obtained in step S5 to a piecewise exponential curve, and normalize the peak day envelope amplitude to obtain the peak damage rate; locate the date when the net damage drops to half of the peak value from the peak day to obtain the half recovery time; record the full recovery time when the net damage is below the upper limit of the envelope for five consecutive days; output the peak damage rate, half recovery time and full recovery time raster.

[0145] More specifically, this embodiment can be implemented through the following methods:

[0146] Step S1 includes the following steps:

[0147] S11. Obtain dual-polarization synthetic aperture radar imagery for the thirty consecutive days before and after the storm's landfall date. Passive microwave soil moisture grid Typhoon Center Intensity Log and field sample plot damage records Constructing daily-scale time series based on Gregorian calendar dates. A unified time stamp label is attached to the four types of data to achieve initial alignment of multi-source data in the time domain.

[0148] S12, to Each image undergoes orbit refinement, topographic distortion is corrected using a digital elevation model, and radiometric calibration and slant range projection conversion are performed, converting digital quantization values ​​into backscattering coefficients. .

[0149] All images are reprojected onto a unified coastal reference frame to ensure comparability and spatial homogeneity among the images.

[0150] S13, Based on SAR pixel resolution Construct the target mesh, for and Joint resampling of topographic weights and wind field weights was performed. Weighting factors were automatically allocated under the dual constraints of elevation difference and radial distance of the typhoon wind field, resulting in fine-scale moisture content... With typhoon intensity It can express the spatial attenuation characteristics of lateral seepage gradient and wind speed caused by micro-topography.

[0151] S14, Based on the set of observation dates of field quadrats As an anchor point, , and Corresponding to the most recent observation day; a placeholder is set when observations are missing for a certain day. Preserve the integrity of the time series.

[0152] S15. Extract two types of strong scattering artificial targets, namely stable shoreline and harbor basin retaining wall, to form a control point set. The cross-date translation is calculated using phase correlation and sub-pixel interpolation. With rotation angle Rigid body correction is performed on each image to ensure that the same pixel has consistent geographic coordinates on all dates.

[0153] S16, combined with daily tide levels With the angle of incidence threshold ,shield or Pixels; add quality labels to scattering values ​​with a signal-to-noise ratio below 3dB. The remaining pixel labels are set to To eliminate the interference of seawater-covered areas and extreme slant-view areas on backscattering reliability.

[0154] S17, in the common pixel domain With the time domain By integrating storm intensity, moisture content, bipolar scattering, and quadrat damage index, a storm-moisture-scattering synchronous grid stack is constructed:

[0155] ;

[0156] in: Pixel ,date The five-dimensional feature column vector; Typhoon intensity index of the corresponding pixel after spatial interpolation; Soil moisture content after topographic weighting and wind field weighting sampling; : VV polarization backscattering coefficient; : VH polarization backscattering coefficient; : Vegetation damage index corresponding to the field quadrat; A synchronized raster stack with global cell-time dual indexes, used as a unified input for subsequent causal graph modeling; : Meta index; Time index; : Effective pixel set of the coastal zone; : Unified daily-scale time series.

[0157] Step S2 includes the following steps:

[0158] S21. Response Baseline Establishment: In multiple years of typhoon landfall cases, salt marshes, sand dune shrublands, and seagrass beds exhibit drastically different water swelling rates, radar scattering gain, and physical breakage thresholds in response to high-humidity storm environments due to differences in morphology, root systems, and leaf surface wax. To capture this community-element-time three-dimensional response pattern, historical validation plot data were integrated with corresponding moisture, scattering, and damage sequences. Peak times, peak amplitudes, and half-lives were statistically analyzed for each community to generate a priori library. .

[0159] The parameters within the library are centrally recorded using a three-level lag window. , , This baseline is used to characterize the temporal gradients of surface moisture surge, water film-structure scattering peak shift, and structural damage manifestation caused by storms. It inherits the actual physical timeline and provides community-dependent temporal constraints for subsequent node mapping.

[0160] S22. To address the alignment problem of multi-source elements across both spatial and temporal axes, this step transforms the four states—storm, moisture, scattering, and damage—into a set of nodes according to three principles: same pixel, community consistency, and lag matching. A one-to-one correspondence is then established through a community-specific lag window.

[0161] ;

[0162] in: Pixel At any moment The intensity of the typhoon; After a typhoon makes landfall, it experiences a specific lag within its community. Surface moisture; , : Dual-polarized backscattering sequence, including contributions from both water film and structure; : On-site measurement of vegetation damage index; At the reference time and the five-dimensional node vector formed by the lag at each level; Pixel The complete set of nodes along the entire timeline.

[0163] Through the above node mapping, observations originating from different physical processes are synchronized to the same pixel index and community scale, ensuring that subsequent edge weight calculations follow the causal context of same origin, same community, and same hysteresis.

[0164] S23. Directed Path Setting: Considering that storms first change surface humidity, and high humidity then increases radar backscattering through the water film effect, ultimately manifesting as a physical chain of structural scattering attenuation during blade breakage or lodging, this step writes the three forward edges into the adjacency matrix and initializes the edge weights using a calculation method with hysteresis correlation coefficients:

[0165] ;

[0166] ;

[0167] in: Pixel A three-layer directed graph adjacency matrix; :storm The initial weights of the water edges; :Moisture Initial weights of the scattering edge; :scattering Initial weights of the damaged edges; , ; , ; , ; Community The Level lag window; ; : Complete timeline.

[0168] Based on the above calculation method, the edge weights not only consider numerical correlation, but also explicitly incorporate community lag characteristics, so that the phenomena of water film causing scattering advance peak and structural damage causing damage lag are quantified in the initial weight assignment stage.

[0169] S24, to The three positive edge weights are column-normalized to ensure that the sum of the outgoing edge weights is always 1. Simultaneously, the weight range of the corresponding community is read from the prior library. As upper and lower bounds, the normalization result is truncated to the interval to generate the initial causal weight matrix of the pixels. .

[0170] S25. During the storm's high tide phase, some low-lying pixels are submerged by seawater for extended periods, accompanied by foam infill, resulting in high specular scattering and random volumetric scattering. To isolate this type of non-vegetation scattering, this step detects residual tide levels. Probability of foam coating ,like or Then in scattering Add a suppression edge along the damage direction and assign negative weights. Suppressing edges reduces the risk of misjudgment by weakening the transmission of structural scattering signals to damaged nodes.

[0171] S26. In high humidity scenarios, water film thickness, leaf curling, and light intensity all evolve dynamically over time, making it difficult for initial weights to accurately reflect the true causal strength in the long run. This step involves a sliding window traversal along the time axis. Calculate observed damage and based on current The residuals between the defects are extrapolated, and the edge weights are updated using gradient descent. When the variance of the cell residuals decreases to a threshold... When the iteration terminates, the final causal weight tensor is output. .

[0172] The results will proceed to the subsequent water film-structure unmixing step, providing pixel-level, community-specific physical priors for stripping away interference and reconstructing pure structural damage sequences.

[0173] Step S3 includes the following steps:

[0174] S31, Pixel fetching, for causal graph domains Each cell within According to the timeline Simultaneous capture of four types of observations: VV polarization scattering VH polarization scattering Surface moisture and the moisture calculated by S2 Scattering causality weight The retrieval process performs pixel-level alignment based on the same row and column coordinates, and then assigns the cluster labels. With quality label They are bound together to ensure that subsequent unmixing takes into account both the differences in water absorption rates between communities and the confidence level markers of S1–S2.

[0175] S32. Storms cause a simultaneous surge in near-surface relative humidity and leaf surface free water film thickness, resulting in a linear-saturated composite gain for backscattering. This is achieved through pixel-specific... Modulate the water film response intensity and... The expected water film scattering conditions are obtained by averaging the moisture values ​​within the daily high humidity window.

[0176] ;

[0177] in: Pixel ,time Expected value of water film scattering; :Moisture Scattering causal edge weight reflects the gain sensitivity of the water film in that pixel; The experimentally calibrated water-scattering gain function exhibits an initial increasing and then saturating characteristic. : Effective window length for high humidity after a storm (days); Pixel exist Surface moisture observation at any given time; Community The multi-year average water film scattering is used for low humidity background compensation.

[0178] S33. Physical damage such as broken vegetation stems and loosened root systems reduces volume scattering and birefringence intensity. To quantify this attenuation, bipolar scattering is weighted by energy conservation, and the expected water film value calculated in S32 is subtracted to obtain the structural scattering residual:

[0179] ;

[0180] in: Pixel ,time Structural scattering residuals; The polarization energy weighting coefficients of VV and VH satisfy the following conditions: ; Measured values ​​of dual-polarized backscattering; : S32 estimate of water film scattering expectation.

[0181] S34. Timing continuity correction: Impulse noise caused by instantaneous changes in tide level or foam decay. The sequence exhibits high-frequency spikes. To preserve the true low-frequency characteristics of the destruction-buffering-repair ecological process while suppressing meaningless high-frequency disturbances, this step applies a two-day window first-order Savitzky-Golay filter to the residual sequence. The filter smooths the amplitude while preserving the inflection point positions, making the destruction peaks and recovery trends more physically interpretable.

[0182] S35. The difference in the original scattering intensity of different pixels may reach 10–15 dB. If it is not normalized, it will be difficult to unify the learning rate when fitting the game in the future.

[0183] This step extracts the mean structural residuals from the seven days prior to storm landfall. As a steady-state baseline, the normalized net damage was calculated:

[0184] ;

[0185] Normalization result The interval is defined as follows: 0 represents no loss or complete recovery, and 1 corresponds to the peak damage caused by the storm, creating a unified scale for cross-pixel and cross-community comparisons.

[0186] S36. Finally, the normalized net damage... Cell Index by Row Index by column time Assembled into a two-dimensional matrix Matrix elements also carry quality labels. Community tags This will support the next step of enhancing the game theory algorithm to converge first in high-confidence pixels and correct later in low-confidence pixels, thereby achieving a robust fit to the vegetation damage-recovery dynamics after extreme storms.

[0187] Step S4 includes the following steps:

[0188] S41. The wind resistance of coastal vegetation is highly coupled with the environmental background: elevation Directly determines the duration of tidal immersion; slope aspect Influences on solar radiation and evapotranspiration; substrate conductivity Reflects salt infiltration rate; community type This implies differences in root depth and stem strength. This step utilizes these four rasters and aligns them one-to-one with the cell indices to form an environment vector. Simultaneously, the four-dimensional sample variance is calculated, and the covariance matrix used for subsequent weighted Mahalanobis distance calculation is determined. This is pre-calculated at this stage to ensure comparability of different dimensions in distance metrics.

[0189] S42. Ecological zoning: Considering that salt marshes and seagrass beds have similar topography but drastically different community functions; and that the southeastern slope area, which also belongs to dune shrubland, exhibits independent resilience due to stronger evapotranspiration on the windward side, this step uses a progressive similarity splitting algorithm to... Perform hierarchical clustering.

[0190] The algorithm first uses the entire domain as the root node and calculates the weighted Mahalanobis distance for any pair of pixels:

[0191] ;

[0192] in: Pixel The weighted Mahalanobis distance is used to measure the overall niche differences. : A two-pixel environment vector; The diagonal covariance matrix formed by the variances of each dimension of the environment vector; : The overall variance of the four factors.

[0193] Then iteratively select the maximum value within the region. Pixel pairs are divided along their connecting lines until the maximum distance within any sub-region is below a threshold. The process ensures that the zoning retains significant ecological differences while avoiding excessive fragmentation, ultimately outputting three major ecological zones—salt marshes, dune shrublands, and seagrass beds—along with several micro-landscape sub-zones. .

[0194] S43. The resilience baseline must be assessed under the same illumination and temperature conditions; otherwise, natural seasonal fluctuations will be misinterpreted as storm effects. This step involves querying typhoon tracks and wave warning records for the past ten years, removing all periods affected by typhoons; from the remaining years, Sentinel-1 dual-polarization SAR images of the same calendar month as the target storm event are extracted to construct a storm-free control set. During the process, the trajectory should be kept consistent with the imaging geometry to avoid introducing additional scattering noise due to differences in viewing angle.

[0195] S44, in each partition Within, for stormless sequences, according to calendar day Aggregate the scattered pixels from all years, calculate the 5th and 95th percentiles, and form the natural fluctuation envelope for this region:

[0196] ;

[0197] in: Partition On calendar days The lower and upper limits of natural scattering; Percentage operator; VV polarization scattering in storm-free years in history; The time index is mapped to calendar days. The 5%–95% range covers normal fluctuations such as tides and vegetation growth, while also tolerating extreme values, thus setting an objective threshold for damage anomaly detection.

[0198] S45. Land reclamation or emergency dredging in certain years can cause a sudden increase or decrease in scattering; directly recording this to the baseline would mask the true natural state. This step compares each historical curve with the envelope to locate curves exceeding the specified limits. The isolated peaks and valleys were identified; then, land use change records and on-site inspection records were consulted. If human disturbance was confirmed, the corresponding curve was removed from the dataset. The remaining samples were smoothed using LOESS to remove high-frequency noise, making the envelope closer to long-term natural fluctuations.

[0199] S46. Write the upper and lower bounds of the envelope of each partition back to its pixel domain to generate two tough baseline gratings. When fitting the net damage curve for the same pixel in subsequent games... Exceeding This will trigger a penalty term in the player's reward function, forcing the fitted result to revert to a reasonable range for the partition, thus avoiding global distortion caused by a few outlier pixels. Simultaneously, a partition statistics table will be output. This allows regulatory authorities to quickly compare the resilience differences between storm years and natural years, guiding subsequent repair strategies.

[0200] Step S5 includes the following steps:

[0201] S51. Vegetation responses to extreme storms often span weeks to months, and the accuracy of time alignment directly determines the convergence efficiency of the game fitting. This step first uses the net damage sequence calculated by S3. The partitioned toughness envelope generated by S4 According to a unified timeline Complete alignment and remove items marked as quality labels in S1. Invalid observations. Subsequently, pixel-level moisture content was read. Scattering edge weight Using linear mapping functions Generate confidence factors. The confidence level is set by the median of the entire domain to ensure that the confidence level is distributed in the range of the median. Within the range and with sufficient dispersion. Finally, the maximum number of rounds in the game is preset. With convergence error threshold This completes the encapsulation of pixel-level game input.

[0202] S52. In the initial stage of a storm's landfall, plants are rapidly damaged by the impact of breakage, and then gradually initiate physiological repair mechanisms only after entering a latent period. Regarding this two-stage process, the parameter player first... (The sentence is incomplete and requires more context to translate accurately.) Find the damage peak value and set the corresponding time as 0. Based on this, the initial damage range is given. Incubation period The initial recovery rate is then calculated by taking the time difference between the peak damage and the first inflection point of the curve, and using the envelope slope. Driven by these ternary parameters, the expression for generating the initial simulation curve is:

[0203] ;

[0204] in: First simulation of net damage; Initial damage magnitude; Incubation period; Preliminary estimate of recovery rate; Reference time for storm landfall; Global nonlinear acceleration factor, set to 1.2; : Attenuation modulation coefficient, take .

[0205] The initial version of the simulation curve uses the Sigmoid function to capture the latent-initiation transition in the first part, and the exponential term in the second part simulates the decreasing repair rate over time, so that the curve has the continuity of ecological process.

[0206] S53. Residual players nitpick: Under high humidity interference, net damage sequences often contain two types of errors: transient spikes and slow exudation. If the residuals are minimized globally in one go, the curve shape can easily be skewed by the spikes. Residual players first calculate the difference between the two curves. Then, the maximum continuous residual segment is found using the sliding window accumulation method. For this segment of the error signal, the residual analyst constructed a weighted residual score by combining confidence level and baseline out-of-bounds weights:

[0207] ;

[0208] in: : First round residual score; Confidence factor; The penalty coefficients for exceeding the upper and lower bounds are both set to 1.5. Indicator function.

[0209] This design prioritizes exposing ecologically unacceptable biases while reducing the weight of noisy pixels to prevent them from misleading parameter users.

[0210] S54, parameters according to player The magnitude and direction of the deviation are used to perform vector-level correction on the ternary parameters: if the deviation is positive and occurs in the later stage of the repair, then the value is appropriately increased. To accelerate recovery; if the deviation is negative and concentrated in the latency period, then reduce And fine-tuning The correction range strictly follows ,in For learning rate, The direction vector is adaptively determined by the shape of the deviation segment. During this process, the confidence level... pass With pre-amplification, high-confidence pixels will have a larger step size, thus converging earlier.

[0211] S55, The game enters the [missing information] stage. After the round, the residual players are re-evaluated. ,like and If the game continues, it ends and the game is recorded; otherwise, it ends and the game is stopped. .

[0212] Based on this, for high-confidence salt marsh pixels, which have been completely stripped away due to water film scattering, the residuals mainly come from true fracture recovery and usually converge within 5-7 cycles; while seagrass bed pixels are more affected by tidal initiation and often require close convergence. .

[0213] S56. After all pixels have completed the game, write the final parameters and residual indices back to the four grids:

[0214] recovery range Incubation period Recovery rate Residual penalty Simultaneously, it outputs a zonal statistics table and an error distribution histogram, providing ecological authorities with information to assess the priority and critical vulnerable areas of storm recovery projects. Through the aforementioned enhanced game theory, this step, after removing water film scattering interference, successfully achieves pixel-level robust fitting of the dynamics of vegetation structure damage and recovery.

[0215] Step S6 includes the following steps:

[0216] S61. Parameter raster loading: After a typhoon passes, residual tide levels and sea fog backflow can cause timeline misjudgments. Therefore, the first step is to read... After three layers of grids, the storm's landfall reference date is determined using on-site water level records. The curve is then corrected backward to the actual peak date of the high tide to ensure that the starting point of the curve is synchronized with the maximum structural impact. Subsequently, all grid lines are projected onto a unified diurnal time axis. Furthermore, observable markers are set for pixels that are easily obscured by salt marsh water, and missing days are automatically skipped in subsequent calculations. This step provides a data foundation with complete time series and guaranteed quality for curve reconstruction.

[0217] S62. During the latent period, vegetation remains in a state of tissue water saturation and growth stagnation. Even if leaves collapse, they will not immediately recover. Therefore, The section directly maintains the peak value As the tide recedes and sunlight returns, the repair rate gradually becomes dominated by respiration, which can be... The single exponential decay description. The piecewise function can be specifically expressed as:

[0218] ;

[0219] Therefore, it conforms to physiological delay while avoiding the uncertainties of multi-parameter fitting. The numerical values ​​will gradually approach each other. This reflects the maximum natural restoration limit allowed by the ecological niche.

[0220] S63, the instantaneous wind pressure of extreme storms and drifting debris together determine The height. To convert this absolute amplitude into a cross-regional comparable relative indicator, this step uses the peak day... envelope amplitude on that day Normalization is performed to obtain the peak damage rate. .like This indicates that the structural damage of the pixel has exceeded the upper limit of historical natural fluctuations, and it belongs to a severely damaged area that requires urgent intervention.

[0221] S64, Half-recovery time This reflects the time required for vegetation to transition from a critical survival state to a self-healing state, serving as a quantitative indicator of early post-storm repair efficiency. This step involves... Scanning daily from the peak date, once detected... (Allows a 2% tolerance), i.e., the date of recording. .short This often corresponds to areas with sunny slopes or good bottom permeability, verifying the regulatory effect of environmental factors on the recovery rate in S1–S4.

[0222] S65. Even if the index declines well, high humidity or subsequent small storms can cause secondary disturbances. Therefore, this step requires... First fall After five consecutive days of stable condition, the duration of full recovery was recorded. If two transgressions occur within five days, the area will be marked as unrecovered and designated as a priority patrol area in subsequent monitoring plans. This design aligns with the actual situation of frequent intertidal submersions and repeated salt stress on vegetation along the coast.

[0223] S66, Peak Damage Rate, Half-Recovery Time Full recovery time The three indicator grids can be directly overlaid onto the existing causal weight layer for dynamic linkage display by the intelligent monitoring platform. Simultaneously, S4 partition masks are used. Calculate the mean, standard deviation, and 25%–75% quantiles to generate partitioned layers and statistical tables.

[0224] The data results are directly fed back to the management department to quickly identify areas with both severe damage and slow recovery, and to deploy targeted restoration measures such as artificial support, planting grass to stabilize sand, or dredging tidal channels to achieve differentiated ecological governance after extreme storms.

[0225] In summary, the technical solution proposed in this embodiment first constructs a synchronous raster stack of "storm-moisture-scattering" at the data organization level, and achieves strict alignment of multi-source field quantities at the pixel domain and daily scale through joint resampling of topographic weights and wind field weights. The resampling weights are jointly determined by the local elevation difference and the radial distance of the typhoon wind field, which preserves the lateral infiltration gradient controlled by micro-topography and synchronously maps the spatial attenuation of wind speed, so that the typhoon intensity and soil moisture field evolving over time can be coupled to the radar scattering pixels with consistent resolution. On this basis, the solution uses two types of high signal-to-noise artificial targets, stable shoreline and harbor retaining wall, to perform cross-date rigid body correction, ensuring that the same pixel has stable geographic coordinates on the complete time axis. The above processing not only provides a unified and frame-free data cube for subsequent algorithms, but also establishes a traceable reference for the drastically changing high humidity background in the very short time after the storm makes landfall. This is the biggest structural difference from traditional pre-disaster and post-disaster two-scene difference or long baseline interferometry strategies.

[0226] At the causal modeling level, this embodiment, based on the physical resilience characteristics of different vegetation communities, pre-calculates historical typhoon quadrat curves, extracts three sets of parameters—peak amplitude, peak time, and half-life—to form a community response prior library and provides a three-level lag window. Subsequently, in the pixel-level node set, it explicitly distinguishes between "storm intensity nodes," "soil moisture nodes," "VV / VH dual-polarization scattering nodes," and "damage index nodes," and establishes three positive principal edges—storm-moisture, moisture-scattering, and scattering-damage—as well as a negative edge for bubble suppression based on specific lag relationships within the community. The initial edge weights are calculated using lag correlation coefficients and normalized on the column vector, then truncated using the edge weight intervals set by the prior library to ensure that the weights reflect real-time correlation while also satisfying the constraints of the community's physical temporal sequence. To overcome the edge weight drift caused by water film thickness, leaf curling, and light variations under high humidity conditions, the proposed scheme applies a sliding window gradient descent strategy along the complete time axis. The edge weights are adaptively updated until the residual variance converges, using the residual between observed and inferred damage as the objective function. This yields a causal weight tensor with temporal continuity and physical interpretability. This causal graph construction and dynamic calibration mechanism, centered on the community hysteresis window, differs from existing methods that directly establish feature-label mappings using empirical thresholds or black-box learning. It explicitly characterizes the multi-level transmission and driving rate differences of the water film-breakage-recovery chain.

[0227] In the damage extraction and recovery fitting stage, this embodiment uses pixel-level moisture-scattering edge weights as modulation coefficients. Combined with an experimentally calibrated incremental-saturation gain function, the moisture sequence within the high-humidity window is convolved to obtain the expected water film scattering. This expectation is then subtracted using a weighted average based on dual-polarization energy conservation to obtain the pure structural scattering residual. Subsequently, a short-window Savitzky-Golay filter is used to suppress tidal impulse noise, and the average of the seven days prior to storm landfall is introduced to normalize the net damage sequence. To fit the damage-recovery curve, a dual-subject enhanced game of "parameter player - residual player" is designed: the parameter player describes the entire latency-initiation-repair process using a sigmoid-exponential piecewise function, while the residual player weights the coherence error based on confidence level and resilience envelope out-of-bounds conditions, driving the parameter vector to adaptively converge during iteration. After the game terminates, three grids are output: recovery amplitude, latency, and recovery rate. The peak damage rate, half-recovery duration, and full recovery duration are further calculated. This step-by-step unmixing and game-theoretic fitting strategy, by explicitly stripping water film gain and using residual feedback to dynamically correct model parameters, avoids the problem of damage overestimation or recovery distortion in traditional single-exponential or static machine learning models under high humidity disturbances, and constitutes the core algorithm that this application intends to protect.

[0228] Compared to existing schemes based on pre-disaster-post-disaster difference or fixed-window exponential models, this embodiment employs a synchronous grid stack of storm-moisture-scattering and joint resampling of topographic weights and wind field weights. This ensures that typhoon intensity, soil moisture, and dual-polarization scattering maintain the same step size and resolution within the pixel domain. This structure provides an equivalent spatiotemporal coordinate system for subsequent analysis, avoiding convolution errors caused by resolution inconsistencies in multi-source field quantities. Existing methods typically rely on image-level or zone-level thresholds, while this scheme establishes a causal graph containing community-specific hysteresis windows at the pixel level. This explicitly quantifies the multi-level progressive relationship between storm-moisture-scattering-damage and continuously updates the edge weights through sliding window gradient descent, achieving dynamic adaptive extrapolation that matches the physical process.

[0229] In the damage extraction stage, this embodiment modulates the high-humidity window using water-scattering edge weights, first stripping the water film gain before calculating the structural scattering residual. This fundamentally separates the two opposing signals: "increased scattering due to rising water content" and "decreased scattering due to structural breakage." Traditional amplitude ratio or coherence threshold methods cannot provide this resolution in high-humidity environments. Furthermore, an enhanced game mechanism involving parameter players and residual players is introduced, linking ecological process parameters such as latency and recovery rate with confidence and resilience envelope updates. This achieves a balance between morphological continuity and statistical consistency in the damage-recovery curve, rather than relying on a large number of labeled samples to train a black-box model. Through these differentiated designs, this embodiment can maintain stable detection and precise quantification of structural damage under conditions of drastic tidal changes and short-term strong water film interference, demonstrating its targeted adaptation to extreme storm scenarios in coastal zones and its inherent superiority in principle.

[0230] One specific embodiment is as follows:

[0231] Following a super typhoon, a provincial bay area wetland protection center launched an 18-month monitoring project on damage and recovery in salt marshes, sand dune shrublands, and adjacent seagrass beds. Historically, this area has repeatedly used the pre-disaster–post-disaster NDVI differential method for rapid assessment, but the misjudgment rate remained high under high humidity and salt fog conditions, and the half-recovery time was difficult to quantify. This monitoring project was jointly implemented by the Oceanographic Institute's remote sensing team and the Department of Natural Resources, fully incorporating the "storm-moisture-scattering" synchronous raster stack and pixel-level causal game model proposed in this embodiment.

[0232] In the data preparation phase, this embodiment collected 11 Sentinel-1A / BIW dual-polarization SAR images (10m) from August 28, 2022 to October 12, 2022, simultaneously downloaded SMAPL3 daily soil moisture products (9km), and obtained the 6-hour intensity trajectory of the typhoon center from JTWC. In the field, 32 quadrats were set up in both salt marshes and sand dunes, and the Damage Index (DI) was recorded using an inclinometer and a breakage rate scale. All raster data were reprojected to UTM-50N after orbit refinement, SRTM-1DEM correction, radiometric calibration, and slant range projection conversion. Subsequently, the JTWC intensity grid and SMAP moisture grid were resampled to 10m resolution using the ΔH–d_r dual-constraint algorithm and merged into a synchronous stack containing Ty, SM, VV, VH, and DI, fully covering the 30-day diurnal sequence before and after landfall.

[0233] In the causal modeling phase, this embodiment uses sample records from five typhoon cases between 2016 and 2021 to construct a priori library, statistically obtaining the lag windows for salt marshes {1d, 3d, 6d}, sand dune shrublands {2d, 5d, 7d}, and seagrass beds {0d, 2d, 4d}. Following the pixel-community consistency rule, Ty, SM, VV / VH, and DI are mapped to nodes, and three positive principal edge weights are initialized. Negative suppression edges are automatically added to pixels with a 300mm tidal residual and a 0.65 foam probability. When the global residual variance iterates to 0.018, the causal weight tensor converges.

[0234] In the water film-structure demixing process, this embodiment is based on The expected value of water film scattering was calculated, and a structural scattering residual sequence was generated under the constraint of dual-polarization energy conservation. Subsequent 2d Savitzky-Golay filtering effectively weakened the spike pulses caused by the tidal peak reversal. The net damage ND(t) was obtained by normalizing the values ​​to the mean of the 7 days prior to landfall, with peak values ​​of 0.83 for salt marshes, 0.57 for sand dunes, and 0.48 for seagrass beds.

[0235] The ecological resilience baseline was divided into five micro-landscape sub-regions using a weighted Mahalanobis distance splitting algorithm. The 5%-95% natural fluctuation envelope was calculated using storm-free Sentinel-1 images from the same month period between 2012 and 2021. In the augmented game, the parametric players provided an initial curve using a sigmoid-exponential piecewise function. The residual player performs weighted fault-finding based on confidence level and out-of-bounds conditions, ultimately achieving an average convergence time of 6.3 rounds per pixel. The output includes the recovery amplitude A, latency, and recovery rate grid, and further calculates the peak damage rate PD and half-recovery time. With full recovery time.

[0236] During the interim acceptance test on October 18, 2022, this embodiment selected one of the following: The results of comparing the dense quadrat regions with the traditional NDVI difference method, the dual polarization amplitude ratio method, and this embodiment are shown in Table 1:

[0237]

[0238] Table 1

[0239] The results show that this embodiment exhibits smaller numerical deviations in peak damage localization and half-recovery time estimation, and provides better comprehensive detection of damaged pixels. An increase of 13 percentage points. A priority list for remediation was issued in January 2023, prioritizing those with a PD > 0.7 and... The salt marshland was identified as a key area for emergency sand fixation and tidal dredging, and subsequent on-site verification confirmed that the accuracy rate of the inference plan reached 87%.

[0240] This embodiment fully verifies the feasibility of applying the combined use of synchronous grid stack, community hysteresis causal graph and water film-structure unmixing to quantitative monitoring of vegetation damage after extreme storms. It also demonstrates the stability of enhanced game curve fitting in parameter estimation during the recovery process, laying a technical foundation for its subsequent promotion over longer time periods and larger areas.

[0241] Reference Figure 7 This application also provides a device for assessing coastal vegetation damage and recovery after extreme storms, which can realize the above-mentioned method for assessing coastal vegetation damage and recovery after extreme storms. The device includes:

[0242] The data acquisition unit is used to integrate storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common cell domain and time domain to construct a synchronous raster stack.

[0243] The graph construction unit is used to map the four states of storm, moisture, scattering and damage in the synchronous grid stack into a set of nodes, and to construct a pixel-level community hysteresis causal graph based on each node in the set of nodes.

[0244] The damage determination unit is used to obtain the expected value of water film scattering based on the pixel-level community hysteresis causality graph, and to determine the net damage sequence using the expected value of water film scattering.

[0245] The natural wave envelope determination unit is used to divide the ecological zones of different environments according to the synchronous grid stack, calculate the scattering value in each ecological zone, and then obtain the natural wave envelope.

[0246] The parameter fitting unit is used to encapsulate the net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal map, and the confidence factor as game inputs and input them to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery.

[0247] The damage and recovery assessment unit is used to output the peak damage rate, half recovery time, and full recovery time of the vegetation based on various parameters of the vegetation damage and recovery.

[0248] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0249] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method of this application. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0250] It is understood that the content of the above method embodiments is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the methods of this application, and the beneficial effects achieved are the same as those achieved by the methods of this application.

[0251] Please see Figure 8 , Figure 8 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0252] The processor 801 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0253] The memory 802 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802 and is called and executed by the processor 801.

[0254] The 803 input / output interface is used to implement information input and output.

[0255] The communication interface 804 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0256] Bus 805 transmits information between various components of the device (e.g., processor 801, memory 802, input / output interface 803, and communication interface 804);

[0257] The processor 801, memory 802, input / output interface 803, and communication interface 804 are connected to each other within the device via bus 805.

[0258] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of this application.

[0259] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0260] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0261] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0262] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0263] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0264] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0265] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0266] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0267] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0268] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0269] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0270] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0271] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for assessing coastal vegetation damage and recovery after extreme storms, characterized in that, The method includes the following steps: By integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain, a synchronous grid stack is constructed. The four states of storm, moisture, scattering and damage in the synchronous grid stack are mapped to a set of nodes, and a pixel-level cluster hysteresis causal graph is constructed based on each node in the set of nodes. The expected value of water film scattering is obtained by solving the pixel-level community hysteresis causality graph, and the net damage sequence is determined by using the expected value of water film scattering. Based on the synchronous grid stack, ecological zones of different environments are obtained, and scattering values ​​are calculated in each ecological zone to obtain the natural wave envelope; The net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal graph, and the confidence factor are encapsulated as game inputs and input to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery. Based on the various parameters of vegetation damage and recovery, the peak damage rate, half-recovery time, and full recovery time of the vegetation are output. The process of integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain to construct a synchronized raster stack includes the following steps: Acquire recorded data for a first set number of consecutive days before and after the storm's landfall date; wherein, the recorded data includes radar images from dual-polarization synthetic aperture radar, soil moisture grids, typhoon intensity grids, and damage records from field sample plots; A uniform daily-scale timescale is appended to the recorded data; The radar image is sequentially subjected to orbit refinement, digital elevation model correction, radiometric calibration and slant range projection, and then reprojected to a unified reference frame after conversion. Based on the radar pixel resolution, the typhoon intensity grid and the soil moisture grid are jointly resampled using topographic weights and wind field weights; wherein, the weighting factors are automatically allocated under the dual constraints of elevation difference and radial distance of typhoon wind field; The time series is completed using the observation dates of the field sample plots as anchor points, and rigid body correction is used to ensure that the geographic coordinates of the same pixel remain consistent across all dates; By combining daily tide levels and incident angle thresholds to set quality labels, storm intensity, moisture content, VV polarization scattering, VH polarization scattering, and damage index are integrated in the common pixel domain and time domain to obtain the synchronous grid stack.

2. The method for assessing coastal vegetation damage and recovery after an extreme storm, as described in claim 1, is characterized in that... The process of mapping the four states—storm, moisture, scattering, and damage—in the synchronized grid stack to a set of nodes includes the following steps: A community response prior library is generated using historical typhoon cases, and a three-level lag window for the community is obtained based on the community response prior library. Based on the principles of pixel consistency, community consistency, and lag matching, the four states of storm, moisture, scattering, and damage in the synchronous grid stack are mapped to the node set. The step of constructing a pixel-level community hysteresis causality graph based on each node in the node set includes the following steps: The positive edge weights corresponding to storm and moisture, moisture and scattering, and scattering and damage are initialized based on the lag correlation coefficient, normalized, and then truncated according to the prior. When the probability of residual tide level or foam coating meets the set conditions, add the negative edge of scattering and damage; The pixel-level cluster hysteresis causal graph is obtained by iteratively updating the edge weights through gradient descent along the time axis until the pixel residual variance is lower than the threshold.

3. The method for assessing coastal vegetation damage and recovery after an extreme storm, as described in claim 1, is characterized in that... The step of obtaining the expected value of water film scattering based on the pixel-level community hysteresis causality graph includes the following steps: Simultaneously extract the VV polarization scattering, the VH polarization scattering, the water content, and the edge weights of water content and scattering obtained from the pixel-level community hysteresis causality graph for each pixel; The expected value of water film scattering is obtained by combining the edge weights of water and scattering with the water and scattering gain function; The determination of the net damage sequence using the expected value of water film scattering includes the following steps: The structural scattering residual is obtained by weighting the dual-polarization scattering energy and subtracting the expected value of the water film scattering. After applying a two-day window first-order Savitzky-Golay filter to the structural scattering residuals, the mean residual value of the second set number of days before the storm landslide is extracted as the baseline and normalized to obtain the net damage sequence.

4. The method for assessing coastal vegetation damage and recovery after an extreme storm, as described in claim 1, is characterized in that... The process of dividing the ecological zones into different environments based on the synchronized grid stack includes the following steps: Based on the synchronous raster stack, the elevation, slope aspect, conductivity, and community type raster are read in pixel alignment to form an environment vector; The environmental vectors are divided into ecological zones for different environments by progressive similarity splitting clustering based on weighted Mahalanobis distance; The process of calculating scattering values ​​in each ecological zone to obtain the natural wave envelope includes the following steps: Extract dual-polarized radar images of the same month as the target storm from stormless images of the most recently set number of years to form a stormless comparison set; The natural wave envelope is obtained by calculating the 5th and 95th percentiles of the scattering values ​​in each of the ecological zones according to the calendar day based on the storm-free control set. Abnormal curves that have been manually verified are removed from the natural fluctuation envelope.

5. The method for assessing coastal vegetation damage and recovery after an extreme storm, as described in claim 1, is characterized in that... The process of using these parameters to output various parameters related to vegetation damage and recovery includes the following steps: Using these parameters, players generate an initial simulation curve based on damage peak, latency, and recovery rate; The weighted residual scores are calculated based on the confidence level and envelope out-of-bounds conditions of the initial simulation curve using residual players. Using the parameters, the player corrects the parameters according to the vector level rules until the change value of the weighted residual score is lower than the error threshold or reaches the preset maximum round, and outputs the recovery amplitude, latency, recovery rate and residual index as multiple parameters of the vegetation damage and recovery.

6. A method for assessing coastal vegetation damage and recovery after an extreme storm, as described in any one of claims 1 to 5, characterized in that, The method of outputting the peak damage rate, half-recovery time, and full recovery time of vegetation based on various parameters of vegetation damage and recovery includes the following steps: The time axis of various parameters related to vegetation damage and recovery was corrected based on on-site water level records; The various parameters of vegetation damage and recovery are mapped to piecewise exponential curves, and the peak damage rate is obtained by normalizing the peak daily envelope amplitude. The half-recovery time is obtained by locating the date on which net damage decreases to half of the peak value from the peak date. The full recovery time is recorded when the net damage is below the upper limit of the envelope for the third consecutive set number of days. Output the peak damage rate, the half-recovery time, and the full recovery time.

7. A device for assessing coastal vegetation damage and recovery after extreme storms, characterized in that, The device includes: The data acquisition unit is used to integrate storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common cell domain and time domain to construct a synchronous raster stack. The graph construction unit is used to map the four states of storm, moisture, scattering and damage in the synchronous grid stack into a set of nodes, and to construct a pixel-level community hysteresis causal graph based on each node in the set of nodes. The damage determination unit is used to obtain the expected value of water film scattering based on the pixel-level community hysteresis causality graph, and to determine the net damage sequence using the expected value of water film scattering. The natural wave envelope determination unit is used to divide the ecological zones of different environments according to the synchronous grid stack, calculate the scattering value in each ecological zone, and then obtain the natural wave envelope. The parameter fitting unit is used to encapsulate the net damage sequence, the natural fluctuation envelope, the edge weights of pixel-level water and scattering in the pixel-level community hysteresis causal map, and the confidence factor as game inputs and input them to the parameter player. The parameter player then outputs various parameters of vegetation damage and recovery. The damage and recovery assessment unit is used to output the peak damage rate, half recovery time and full recovery time of the vegetation based on various parameters of the vegetation damage and recovery. The process of integrating storm intensity, moisture, VV polarization scattering, VH polarization scattering, and damage index in the common pixel domain and time domain to construct a synchronized raster stack includes the following steps: Acquire recorded data for a first set number of consecutive days before and after the storm's landfall date; wherein, the recorded data includes radar images from dual-polarization synthetic aperture radar, soil moisture grids, typhoon intensity grids, and damage records from field sample plots; A uniform daily-scale timescale is appended to the recorded data; The radar image is sequentially subjected to orbit refinement, digital elevation model correction, radiometric calibration and slant range projection, and then reprojected to a unified reference frame after conversion. Based on the radar pixel resolution, the typhoon intensity grid and the soil moisture grid are jointly resampled using topographic weights and wind field weights; wherein, the weighting factors are automatically allocated under the dual constraints of elevation difference and radial distance of typhoon wind field; The time series is completed using the observation dates of the field sample plots as anchor points, and rigid body correction is used to ensure that the geographic coordinates of the same pixel remain consistent across all dates; By combining daily tide levels and incident angle thresholds to set quality labels, storm intensity, moisture content, VV polarization scattering, VH polarization scattering, and damage index are integrated in the common pixel domain and time domain to obtain the synchronous grid stack.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images