A marine ecological environment dynamic evaluation and zoning management system

Abstract

The application discloses a kind of marine ecological environment dynamic evaluation and zoned management system, it is related to marine ecological environment monitoring management technical field.The data acquisition module obtains long time sequence remote sensing image of target sea area and extracts chlorophyll concentration distribution characteristics, sea surface temperature gradient variation characteristics and suspended matter sedimentation diffusion trajectory characteristics;Evaluation modeling module, construct multi-scale nested evaluation model for gridding evaluation, generate each grid unit primary productivity grade;Source analysis module, spatial coupling analysis is carried out to suspended matter sedimentation diffusion trajectory characteristics and primary productivity grade, and source influence area and endogenous influence area are identified;Dynamic correction module, establish dynamic zoned boundary correction function, according to real-time tidal phase change data and river into sea flux data, the spatial distribution boundary is iteratively corrected, and real-time dynamic zoning map is generated;Zoned control module, according to the corresponding ecological control threshold of each zoned type configuration and executes zoned management operation.

Description

Technical Field

[0001] This invention relates to the field of marine ecological environment monitoring and management technology, specifically a dynamic assessment and zoning management system for marine ecological environment. Background Technology

[0002] Marine ecological environment management requires zoning assessment and regulation of sea areas. Existing technologies often employ indicator extraction and single-scale assessment methods based on remote sensing imagery. Chlorophyll concentration inversion or sea surface temperature analysis can determine the primary productivity level of certain sea areas, and this, combined with suspended solids concentration distribution, delineates affected areas, thus completing basic management zoning. However, in this process, assessment models typically operate on a single spatial scale, treating the entire target sea area as a homogeneous spatial unit, or simply dividing it into fixed grids for indicator statistics. This fails to account for both large-scale background trends and small-scale local anomalies, resulting in inaccurate determination of primary productivity levels in complex nearshore areas. Furthermore, the boundaries of pollution-affected areas or ecologically sensitive areas delineated by existing technologies are often static or based on historical averages, lacking a response mechanism to real-time environmental driving force changes. The zoning results cannot reflect the dynamic migration of boundaries caused by factors such as tidal phase alternation and fluctuations in river flux into the sea, leading to significant lags in zoning management. In the apportionment of suspended solids sources, existing methods either rely solely on the geometric features of diffusion trajectories to distinguish between terrestrial and internal resuspension sources, or depend solely on productivity anomalies at the cell level for inference. Both lack deep coupling analysis within the same spatial framework, easily confusing the impact of terrestrial inputs with the impact of sediment resuspension. They fail to accurately distinguish between productivity changes driven by exogenous nutrients and productivity anomalies caused by endogenous release, thus hindering the generation of targeted, differentiated regulation strategies. During the iterative correction of zonal boundaries, conventional methods generally only consider the simple translational effect of single physical quantities such as tidal currents, without comprehensively quantifying the relationship between tidal direction and buffer zone normal, or the synergistic effect of freshwater flux into the sea and suspended solids flux. Boundary correction lacks precise calculation basis for direction and amplitude, and the correction results cannot accurately characterize the dynamic behavior of boundaries under different combinations of driving forces. Summary of the Invention

[0003] This paper provides a dynamic assessment and zoning management system for the marine ecological environment, which can accurately assess the primary productivity level of the sea area under a multi-scale nested grid, accurately identify the spatial range of the external and internal influence zones, and enable the zoning boundaries to be dynamically corrected based on real-time tidal and river inflow fluctuation data, thereby implementing differentiated ecological regulation for different zones.

[0004] To achieve the above objectives, this invention provides the following technical solution: This invention provides a dynamic assessment and zoning management system for the marine ecological environment. This system, through deep fusion of multi-source remote sensing data and real-time monitoring data, achieves refined dynamic assessment and zoning regulation of the marine ecological state. The system includes a data acquisition module, an assessment and modeling module, a source analysis module, a dynamic correction module, and a zoning regulation module. These modules work collaboratively to depict the evolution of the marine ecological environment from macro to micro levels and from static to dynamic perspectives, and accordingly execute differentiated management operations.

[0005] In the technical solution of this invention, the data acquisition module is used to acquire long-term remote sensing image data of the target sea area and extract chlorophyll concentration distribution characteristics, sea surface temperature gradient change characteristics, and suspended matter sedimentation and diffusion trajectory characteristics from them. Preferably, when extracting the above features, the data acquisition module selects remote sensing images with cloud coverage rates lower than a preset cloud threshold from the remote sensing database according to a preset time sampling interval to form a long-term series, so as to ensure the effectiveness and continuity of the data. For each image, the chlorophyll concentration band value is obtained through band operation, and the difference between adjacent sampling times is calculated. The chlorophyll concentration change rate is used as the chlorophyll concentration distribution feature, which can sensitively reflect the dynamic response of primary producers. The sea surface temperature value is retrieved by inverting the thermal infrared band, and the maximum absolute value of the temperature difference between each pixel and its neighboring pixels is calculated as the sea surface temperature gradient value. These are summarized to form the sea surface temperature gradient change feature, thereby accurately capturing key temperature interfaces such as ocean fronts that affect the distribution of productivity. For the characteristics of suspended matter sedimentation and diffusion trajectory, the optical flow method is used to track the movement trajectory of the peak position based on the suspended matter concentration band values ​​in three or more consecutive remote sensing images. This is used as the characteristics of suspended matter sedimentation and diffusion trajectory, which intuitively reveals the transport path and collection area of ​​particulate matter in water.

[0006] The assessment modeling module constructs a multi-scale nested assessment model to characterize the spatiotemporal evolution of primary productivity in a marine area based on chlorophyll concentration distribution characteristics and sea surface temperature gradient variation characteristics. This model is then used to perform a gridded assessment of the target marine area, generating the primary productivity level for each grid cell. As a preferred embodiment of this invention, the multi-scale nested assessment model employs a nested structure combining a first-level coarse grid and a second-level fine grid. The first-level coarse grid captures large-scale background trends in primary productivity, while the second-level fine grid captures anomalous fluctuations in primary productivity at local scales. In specific implementation, the module first divides the target marine area into two levels of grids with progressively increasing spatial resolution. Using the first-level coarse grid as a unit, the mean chlorophyll concentration change rate and the mean sea surface temperature gradient value within it are calculated and input into a pre-trained primary productivity regression model to obtain a basic primary productivity estimate. This effectively ensures the robustness of the background trend assessment. Using a second-level fine grid as the unit, the differences between the rate of change of chlorophyll concentration within the grid and the mean of the corresponding first-level coarse grid, as well as the differences between the sea surface temperature gradient and the mean of the corresponding first-level coarse grid, are calculated to obtain residuals. These residuals are then input into the residual correction branch of the regression model to obtain the primary productivity correction value. This step significantly enhances the ability to analyze local small-scale anomaly signals. The basic estimated value and the correction value are added together to obtain the final primary productivity assessment value, which is then mapped to a productivity level. The resulting assessment result conforms to regional macro-level patterns while preserving important local spatial details.

[0007] The source apportionment module performs spatial coupling analysis on the characteristics of suspended solids sedimentation and diffusion trajectories with the primary productivity levels of each grid cell to identify the exogenous influence zone dominated by terrestrial input and the endogenous influence zone dominated by sediment resuspension. Preferably, during the spatial coupling analysis, the module places the suspended solids diffusion trajectory lines and the primary productivity level distribution layer in the same spatial coordinate system. For each diffusion trajectory line, if its starting point is located within the buffer zone of a river estuary, it is marked as an exogenous trajectory line; if the change in suspended solids concentration band value at its endpoint exceeds a preset sedimentation threshold, it is marked as a sedimentation trajectory line. Further, grid cells covered by continuous trajectory lines near the estuary are selected from the exogenous trajectory lines to form the initial exogenous influence zone; grid cells with high trajectory line density and no overlap with the exogenous influence zone are extracted from the sedimentation trajectory lines to form the initial endogenous influence zone. To improve the reliability of the zoning, the deviation of the primary productivity level in each initial region from the average level of the sea area is also calculated, and confirmation is given when the deviation exceeds a preset threshold. As a better option, the confirmation of the external source influence zone will also include additional judgment conditions, namely, the spatial distance between the starting point of the suspended matter sedimentation and diffusion trajectory in the initial external source influence zone and the river estuary must be less than the source-sink distance threshold, and there is a time lag correlation between the time-series peak value of the chlorophyll concentration change rate in the initial external source influence zone and the time-series peak value of the river runoff. This condition enhances the causal logic of the external source zone determination from the perspective of the source and ecological response mechanism, and effectively avoids misjudgment.

[0008] The dynamic correction module establishes a dynamic zoning boundary correction function based on the spatial distribution boundaries of the external and internal influence zones. It iteratively corrects the spatial distribution boundaries using real-time acquired tidal phase change data and river flux data, generating a real-time dynamic zoning map. The generation process involves extracting the external and internal boundary lines, marking the remaining areas as background zones, and establishing buffer zones along the normals of each boundary line. The tidal rise and fall direction vector and tidal velocity scalar calculated from the real-time tidal phase data, along with the freshwater flux scalar and suspended solids flux scalar calculated from the real-time river flux data, are input into the dynamic zoning boundary correction function. For external source boundaries, the function operation logic uses the dot product of the tidal direction and the buffer zone normal as the first driving force component, and combines it with the main driving strength obtained by dividing the product of tidal velocity and freshwater flux by suspended solids flux to generate an external source offset vector for each boundary point. For internal source boundaries, the dot product of the opposite tidal direction and the buffer zone normal is used as the second driving force component, and the product of the square of tidal velocity and the sediment resuspension coefficient is used as the main driving strength to generate an internal source offset vector. After each boundary point moves along the corresponding offset vector, the corrected external and internal source boundary lines can be obtained, which are finally combined with the background zone boundary to form a real-time dynamic zoning map. This correction mechanism enables the zoning boundaries to dynamically adjust with the instantaneous changes in hydrodynamic conditions, realistically reproducing the real-time expansion and contraction process of the external and internal source influence ranges.

[0009] The zoning control module configures corresponding ecological control thresholds for each zoning type in the real-time dynamic zoning map and performs zoning management operations based on these thresholds. Specifically, the configured ecological control thresholds include upper limits for exogenous pollutant concentration and nutrient flux in the exogenous influence zone, upper limits for sediment disturbance intensity and dissolved oxygen in the bottom layer in the endogenous influence zone, and upper limits for chlorophyll concentration fluctuation and transparency in the background zone. The module analyzes the real-time dynamic zoning map and sets differentiated monitoring modes and control strategies for the exogenous influence zone, endogenous influence zone, and background zone. In the exogenous influence zone, a high-frequency acquisition mode is configured to collect pollutant concentration and nutrient flux in real time. When any indicator exceeds the limit, an exogenous reduction instruction or a marine discharge restriction instruction containing the reduction measure level and execution parameters is generated, achieving rapid response and precise control of land-based inputs. In the endogenous influence zone, the focus is on the bottom layer, with monitoring of the bottom water and sediment interface layer. Shear stress values ​​are obtained using a sediment disturbance monitor as the disturbance intensity. When this value exceeds the limit, a sediment solidification or sediment cover command is generated based on the liquefaction potential index calculated from pore water pressure. This allows for tiered management of endogenous release risks, effectively curbing secondary release of sediment nutrients. In the background zone, a low-frequency acquisition mode is configured. When actual chlorophyll concentration fluctuations or transparency deviate from their respective thresholds, the system calculates the regional fluctuation dispersion to determine the abnormal coverage area, generating a key inspection command or a full-area patrol command carrying the coordinates of hotspot anomalies. This saves on routine monitoring costs while ensuring alertness to sudden ecological anomalies. Through a closed-loop process of "assessment-source tracing-dynamic zoning-differential control," the system significantly improves the temporal and spatial targeting and adaptability of marine ecological environment management.

[0010] The technical effects and advantages provided by the present invention in the above technical solution are as follows: By constructing a multi-scale nested assessment model, the target sea area is divided into a first-level coarse grid and a second-level fine grid. Within the first-level coarse grid, the background trend of primary productivity is captured, and the basic primary productivity estimate is calculated using the chlorophyll concentration change rate and sea surface temperature gradient. Within the second-level fine grid, the chlorophyll residual value and temperature gradient residual value are input into the residual correction branch of the regression model to obtain the primary productivity correction value at the local scale. These two values ​​are then superimposed to obtain the final assessment value. This nested structure allows for the separation of background trends from local anomalous fluctuations, avoiding the problem of productivity level determination being skewed by local extremes or misjudging large-scale gradual changes as local anomalies under a single-scale grid. In areas with strong spatial gradients, such as nearshore estuaries and tidal convergence zones, the spatial resolution of primary productivity levels is maintained, while the overall productivity distribution trend characteristics of the sea area are not lost, improving the precision and accuracy of productivity assessment in complex sea areas.

[0011] When establishing a dynamic partition boundary correction function based on the spatial distribution boundaries of the external and internal influence zones, the tidal rise and fall direction vector is multiplied by the normal vector of each point within the external buffer zone to obtain the first driving force component of the tide on the external boundary. Simultaneously, the tidal velocity scalar is multiplied by the freshwater flux scalar and then divided by the suspended solids flux scalar to obtain the main driving intensity of the external boundary offset. The external offset vector is determined by multiplying the first driving force component by the main driving intensity. For the internal boundary, the opposite direction of the tidal rise and fall direction vector is multiplied by the normal vector of each point within the internal buffer zone to obtain the second driving force component. The tidal velocity scalar is squared and then multiplied by the bottom sediment resuspension coefficient to obtain the main driving intensity of the internal boundary offset, thereby determining the internal offset vector. This correction function incorporates tidal vectors, freshwater flux, suspended solids flux, and sediment resuspension characteristics into the same calculation framework. This ensures that the magnitude and direction of external boundary offsets are not only affected by the direction of tidal rise and fall, but also by the synergistic regulation of inflowing freshwater flux and suspended solids flux. Internal boundary offsets, on the other hand, independently respond to the sediment resuspension driving force enhanced by the tidal square effect. This avoids the inaccuracy of boundary correction caused by using only a single tidal current vector for translation processing. The zoning boundaries can achieve dynamic migration with both direction and magnitude accurately quantified during tidal phase transitions and river inflow pulse events. This allows the real-time dynamic zoning map to truly reflect the spatial redistribution of marine ecologically sensitive areas as they change with multiple driving forces. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0013] Figure 1 This is a schematic diagram of the structure of the marine ecological environment dynamic assessment and zoning management system; Figure 2 This is a flowchart of a primary productivity multi-level grid assessment method based on remote sensing imagery; Figure 3 This is a flowchart of the suspended matter source analysis process; Figure 4 It is a time-series curve of the rate of change of chlorophyll concentration versus sea surface temperature gradient; Figure 5 This is a time-series diagram showing the relationship between external boundary offset, tidal velocity, and freshwater flux into the sea; Figure 6 It is a real-time monitoring and early warning curve of chemical oxygen demand concentration in the area affected by external sources; Figure 7 This is a map showing the fluctuation of chlorophyll concentration and the distribution of abnormal hotspots in the background area. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] See Figure 1 This invention provides a dynamic assessment and zoning management system for the marine ecological environment, including a data acquisition module, an assessment modeling module, a source apportionment module, a dynamic correction module, and a zoning control module. The data acquisition module acquires long-term remote sensing image data of the target sea area and extracts chlorophyll concentration distribution characteristics, sea surface temperature gradient change characteristics, and suspended matter sedimentation and diffusion trajectory characteristics from these data. The assessment modeling module constructs a multi-scale nested assessment model to characterize the spatiotemporal evolution of primary productivity in the sea area based on the chlorophyll concentration distribution characteristics and sea surface temperature gradient change characteristics. This multi-scale nested assessment model adopts a nested structure combining a first-level coarse grid and a second-level fine grid. The first-level coarse grid is used to capture large-scale background trends in primary productivity, while the second-level fine grid is used to capture abnormal fluctuations in primary productivity at local scales. The assessment modeling module performs a gridded assessment of the target sea area using this multi-scale nested assessment model, generating the primary productivity level for each grid unit. The source apportionment module performs spatial coupling analysis between the suspended matter sedimentation and diffusion trajectory characteristics and the primary productivity level of each grid unit, identifying the exogenous influence zone dominated by land-based input and the endogenous influence zone dominated by sediment resuspension. The dynamic correction module establishes a dynamic zoning boundary correction function based on the spatial distribution boundaries of the external and internal influence zones. It iteratively corrects the spatial distribution boundaries using real-time acquired tidal phase change data and river flux data, generating a real-time dynamic zoning map. The zoning control module configures corresponding ecological control thresholds for each zoning type in the real-time dynamic zoning map. These ecological control thresholds include upper limits for external pollutant concentrations and external nutrient fluxes in the external influence zone, upper limits for sediment disturbance intensity and lower limits for dissolved oxygen in the bottom layer in the internal influence zone, and upper limits for chlorophyll concentration fluctuations and lower limits for transparency in the background zone. The zoning control module then performs zoning management operations based on these ecological control thresholds.

[0016] Example 1:

[0017] In one implementation of the data acquisition module, see [reference]. Figure 2The preset time sampling interval is determined based on the tidal cycle and satellite revisit cycle of the target sea area. The preset time sampling interval ranges from 1 hour to 24 hours, and is specifically set to 6 hours. Remote sensing images with cloud coverage below a preset cloud threshold are selected from the remote sensing database to form long-term remote sensing image data. The preset cloud threshold is set to 10%, meaning that remote sensing images with cloud coverage below 10% are retained.

[0018] For each remote sensing image in a long-term remote sensing image series, chlorophyll concentration band values ​​are extracted through band operations. The OC3M algorithm is used for chlorophyll concentration band value extraction, which performs a logarithmic operation on the ratio of blue band reflectance to green band reflectance of the remote sensing image to obtain the chlorophyll concentration band value. The difference between the chlorophyll concentration band values ​​at two adjacent sampling times is calculated: the chlorophyll concentration band value at the later sampling time is subtracted from the chlorophyll concentration band value at the earlier sampling time, and the difference is taken as the chlorophyll concentration change rate. The chlorophyll concentration change rates at all times constitute the chlorophyll concentration distribution characteristics.

[0019] For each remote sensing image in a long-term remote sensing image series, sea surface temperature (SST) values ​​are retrieved using the thermal infrared band. The SST retrieval employs a nonlinear sea surface temperature algorithm, calculating the SST value based on the brightness temperature value of the thermal infrared band and satellite zenith angle parameters using a piecewise linear regression model. For each pixel, the absolute value of the temperature difference between that pixel and its eight neighboring pixels is calculated. The eight neighboring pixels are defined as the eight adjacent pixels in a 3x3 matrix centered on that pixel, excluding the center pixel. The maximum absolute value of these eight temperature differences is taken as the SST gradient value for that pixel. The SST gradient values ​​of all pixels are then aggregated to form the SST gradient variation characteristics.

[0020] For three or more consecutive remote sensing images in a long-term remote sensing image series, the suspended particulate matter (SPM) concentration band values ​​are extracted from each image. The extraction of SPM concentration band values ​​employs a single-band inversion method. Near-infrared reflectance is selected, atmospherically corrected, and then converted into SPM concentration band values ​​using empirical formulas. The Lucas-Kanade optical flow method is used to track the movement trajectory of the peak position of the SPM concentration band values ​​at different times. Based on the assumption of constant brightness at the peak position in the image sequence and the constraint of neighborhood optical flow smoothing, the optical flow method solves for the displacement vectors of the peak position in the horizontal and vertical directions. The path formed by connecting the peak positions in three or more consecutive remote sensing images is used as the SPM sedimentation and diffusion trajectory feature.

[0021] In one implementation of the evaluation modeling module, the target sea area is divided into a first-level coarse grid, with each coarse grid cell having a spatial resolution of 0.1° × 0.1°. Each coarse grid cell is further divided into a second-level fine grid, with a spatial resolution of 0.01° × 0.01°, higher than that of the first-level coarse grid. For each coarse grid cell, the arithmetic mean of the chlorophyll concentration change rate of all pixels within that cell is calculated to obtain the first-level chlorophyll mean. Simultaneously, the arithmetic mean of the sea surface temperature gradient values ​​of all pixels within that cell is calculated to obtain the first-level temperature gradient mean. The first-level chlorophyll mean and the first-level temperature gradient mean are then input into a pre-trained primary productivity regression model. This pre-trained primary productivity regression model employs a deep neural network structure, containing one input layer, three fully connected hidden layers, and one output layer. The input layer receives two neurons, corresponding to the mean chlorophyll level and the mean temperature gradient level, respectively. The three fully connected hidden layers have 64, 128, and 64 neurons respectively, and the ReLU activation function is used. The output layer contains one neuron, outputting the basic primary productivity estimate. The training process of the pre-trained primary productivity regression model is as follows: historical synchronous observation data are collected, including measured primary productivity values, corresponding remotely sensed mean chlorophyll level and mean temperature gradient level, to construct a training sample set; the mean squared error loss function and Adam optimizer are used, with the learning rate set to 0.001 and the batch size set to 32, and iterative training is performed until the loss function converges to obtain the pre-trained primary productivity regression model.

[0022] For each secondary fine-grid cell, the difference between the chlorophyll concentration change rate of all pixels within that cell and the mean primary chlorophyll value of the corresponding primary coarse-grid cell is calculated to obtain the secondary chlorophyll residual value. Simultaneously, the difference between the sea surface temperature gradient value of all pixels within that cell and the mean primary temperature gradient of the corresponding primary coarse-grid cell is calculated to obtain the secondary temperature gradient residual value. These secondary chlorophyll and temperature gradient residual values ​​are then input into the residual correction branch of the pre-trained primary productivity regression model. The residual correction branch is a sub-network parallel to the backbone of the pre-trained primary productivity regression model, containing one input layer, two fully connected hidden layers, and one output layer. The input layer receives two neurons, corresponding to the secondary chlorophyll residual value and the secondary temperature gradient residual value, respectively. Both fully connected hidden layers have 32 neurons each, using the ReLU activation function. The output layer contains one neuron, outputting the primary productivity correction value. The training of the residual correction branch is carried out simultaneously with the backbone network, using the same training data and loss function, but the input data is the second-order residual values, and the target value is the residual between the measured primary productivity value and the basic primary productivity estimate.

[0023] The initial primary productivity estimate is added to the revised primary productivity value to obtain the final primary productivity assessment value for the second-level fine grid cell. Based on a preset productivity grading threshold, the final primary productivity assessment value is mapped to a primary productivity level. The preset productivity grading threshold is determined using the quantile method, with the 20th, 40th, 60th, and 80th percentiles of the historical primary productivity assessment values ​​for the target sea area used as dividing points to classify primary productivity into five levels: low productivity, low-to-medium productivity, medium productivity, medium-to-high productivity, and high productivity. If the final primary productivity assessment value falls within the corresponding range, it is assigned the corresponding primary productivity level.

[0024] See Figure 4 In the figure, the horizontal axis represents the sampling time number, and the vertical axis represents the numerical change, with the vertical axis value ranging from approximately -1 to 1.2. The solid curve represents the rate of change of chlorophyll concentration, and the dashed curve represents the sea surface temperature gradient value. The chlorophyll concentration change rate curve shows obvious periodic fluctuations, with a period of approximately 15 to 20 sampling units, and the value fluctuates between -0.8 and 1.1, showing multiple alternating peaks and troughs, reflecting the dynamic change characteristics of chlorophyll concentration at different sampling times. The sea surface temperature gradient value curve is generally stable between 0.7 and 1.1, with small fluctuations and no obvious periodicity, showing relatively gentle and high-frequency small fluctuations.

[0025] This figure illustrates how the data acquisition module in Example 1 extracts and calculates chlorophyll concentration band values ​​and sea surface temperature gradient values ​​from long-term remote sensing image data, forming chlorophyll concentration distribution characteristics and sea surface temperature gradient variation characteristics. The periodic trend of the chlorophyll concentration change rate curve corresponds to the short-term variation characteristics in the spatiotemporal evolution of marine primary productivity, while the stability of the sea surface temperature gradient curve reflects the relatively uniform distribution of sea surface temperature in the spatial gradient.

[0026] The fluctuation trends of the two curves in the figure show a certain synchronicity with the sampling time sequence, indicating a correlation between the chlorophyll concentration change rate and the sea surface temperature gradient value. This aligns with the characterization requirements of the multi-scale nested assessment model for the spatiotemporal evolution of primary productivity. The significant positive and negative fluctuations in the chlorophyll concentration change rate indicate substantial spatiotemporal anomalies in chlorophyll concentration within the sea area, providing a valid input data basis for correcting the primary productivity residuals of the first-level coarse grid and the second-level fine grid. Overall, this figure illustrates the key environmental parameters obtained through remote sensing data processing and feature extraction, providing data support for the subsequent input and dynamic assessment of the primary productivity regression model.

[0027] Example 2

[0028] In specific implementation, please refer to Figure 3The source resolution module plots the suspended matter settling and diffusion trajectory features on a grid base map of the target sea area. The grid base map of the target sea area uses the same grid division scheme as the first-level coarse grid in the evaluation modeling module, with a spatial resolution of 0.1° × 0.1°. Each movement trajectory in the suspended matter settling and diffusion trajectory features is superimposed on the grid base map as a vector line feature. The direction of the vector line is determined by the temporal sequence of the peak positions in the trajectory. The grid cells traversed by the vector line are recorded through spatial intersection operations, forming a suspended matter diffusion trajectory line layer. Under the same spatial coordinate system, the primary productivity level of each grid cell is labeled as a polygon feature. The color or grayscale of each polygon feature is distinguished according to the five-level classification of primary productivity level: low productivity level corresponds to the first grayscale value, low-medium productivity level corresponds to the second grayscale value, medium productivity level corresponds to the third grayscale value, medium-high productivity level corresponds to the fourth grayscale value, and high productivity level corresponds to the fifth grayscale value, forming a primary productivity level distribution layer.

[0029] For each diffusion trajectory line in the suspended matter diffusion trajectory line layer, the source resolution module identifies the starting position of the diffusion trajectory line using a linear reference method. The starting position is defined as the coordinates of the first vertex of the vector feature of the diffusion trajectory line. A river estuary buffer zone is pre-set. This buffer zone is generated by using the location points of each river estuary along the target sea area as the center, with a preset buffer radius as the parameter. The preset buffer radius ranges from 500 meters to 5000 meters, and is set to 2000 meters based on the estuary topography of the target sea area. If the starting position of a diffusion trajectory line falls within the circular area of ​​any river estuary buffer zone, then the diffusion trajectory line is marked as an external source trajectory line. A Boolean field with a true value is added to the attribute information of the external source trajectory line.

[0030] For each dispersion trajectory line in the suspended matter dispersion trajectory line layer, the source resolution module identifies the endpoint position of the dispersion trajectory line. The endpoint position is defined as the coordinates of the last vertex of the vector feature of the dispersion trajectory line. The suspended matter concentration band values ​​at two adjacent sampling times at the endpoint position of the dispersion trajectory line are obtained. The two adjacent sampling times refer to the last sampling time corresponding to the dispersion trajectory line and the sampling time preceding that last sampling time. The suspended matter concentration band values ​​at these two times are then subtracted from the suspended matter concentration band value at the last sampling time. When the difference exceeds a preset settling threshold, the dispersion trajectory line is marked as a settling trajectory line. The preset settling threshold is based on the statistical value of the daily average variation of suspended matter concentration observations in the target sea area, and the upper quartile value of the historical daily average variation data of suspended matter concentration in the target sea area is selected as the preset settling threshold.

[0031] The source resolution module extracts spatial distribution features from all diffusion trajectories marked as external source trajectory lines. For each external source trajectory line, the spatial distance between the trajectory line and the boundaries of each river estuary buffer zone is calculated. The spatial distance is defined as the minimum of the shortest Euclidean distances from each vertex on the trajectory line to the river estuary buffer zone boundary. All external source trajectory lines are sorted in ascending order of the spatial distance. Three or more consecutive external source trajectory lines with the smallest spatial distance are selected. The grid cells covered by the selected external source trajectory lines in the suspended matter diffusion trajectory line layer are extracted, and these grid cells are marked as the initial external source influence area. The initial external source influence area is stored in the form of polygon features.

[0032] The source resolution module extracts spatial distribution features from all diffusion trajectories marked as settlement trajectories. Using grid cells as units, the number of settlement trajectories within each grid cell is counted, and the ratio of the number of settlement trajectories to the area of ​​that grid cell is used as the settlement trajectory density. Grid cells covered by settlement trajectories are selected that do not spatially overlap with the initial external source influence zone. Spatial overlap is determined through surface feature intersection operations; if the surface features of a grid cell do not intersect with the surface features of the initial external source influence zone, it is considered to have no spatial overlap. Among the non-overlapping grid cells, grid cells with settlement trajectory density exceeding a preset density threshold are further selected and marked as the initial internal source influence zone. The preset density threshold is determined by calculating the arithmetic mean of the settlement trajectory density of all grid cells in the target sea area, multiplying the arithmetic mean by an adjustment factor of 1.5.

[0033] The source resolution module calculates the first deviation between the primary productivity level of each grid cell within the initial external influence zone and the average primary productivity level of the target sea area. The average primary productivity level of the target sea area is calculated as follows: the arithmetic mean of the primary productivity levels of all grid cells within the target sea area is taken. The primary productivity levels are: low productivity level corresponds to value 1, low-to-medium productivity level corresponds to value 2, medium productivity level corresponds to value 3, medium-to-high productivity level corresponds to value 4, and high productivity level corresponds to value 5. The formula for calculating the first deviation is:

[0034] in, Indicates the first degree of deviation. This represents the total number of grid cells contained within the initial external influence zone. Indicates the first [unit] in the initial external influence zone. The level value corresponding to the initial productivity level of each grid cell. The value ranges from 1 to integers, This represents the grade value corresponding to the average primary productivity level of the target sea area. If the first deviation exceeds the preset external deviation threshold, additional condition judgments are made regarding the external influence zone. The preset external deviation threshold is set to 1.2, which is based on twice the overall standard deviation of the primary productivity level values ​​of the target sea area.

[0035] In the additional condition judgment, the source apportionment module calculates the spatial distance between the starting point of the suspended matter sedimentation and diffusion trajectory within the initial external source influence area and the river estuary. The specific calculation method for the spatial distance is as follows: extract the starting coordinates of all external source trajectory lines within the initial external source influence area, calculate the Euclidean distance from each starting coordinate to the nearest river estuary, and take the minimum value among all Euclidean distances. If this minimum value is less than a preset source-sink distance threshold, then the first condition of the additional condition is satisfied. The preset source-sink distance threshold is determined based on the diffusion range of the river plume in the target sea area and is set to 5000 meters.

[0036] The source analysis module simultaneously determines whether there is a time-lag correlation between the temporal peak of chlorophyll concentration change rate and the temporal peak of river runoff within the initial external source influence area. The temporal peak of chlorophyll concentration change rate is extracted as follows: within the initial external source influence area, the spatial average of the chlorophyll concentration change rate for all grid cells at each sampling time is calculated, forming a time series curve. The local maxima on this time series curve are identified as the temporal peak of chlorophyll concentration change rate. The temporal peak of river runoff is extracted as follows: daily runoff observation data of major rivers along the target sea area are obtained within the same time period, forming a river runoff time series curve. The local maxima on this time series curve are identified as the temporal peak of river runoff. The time-lag correlation is calculated using a sliding time window cross-correlation function. The temporal peak sequence of chlorophyll concentration change rate and the temporal peak sequence of river runoff are time-shifted and matched. Pearson correlation coefficients are calculated for different time lag values, and the time lag value corresponding to the largest Pearson correlation coefficient is taken. If a time lag value exists such that the Pearson correlation coefficient is greater than 0.6, then a time-lag correlation is determined to exist, and the second condition of the additional condition is met. If both the first and second conditions of the additional condition are met, the source analysis module confirms the initial external influence region as an external influence region.

[0037] The source resolution module calculates the second degree of deviation between the primary productivity level of each grid cell within the initial internal influence zone and the average primary productivity level of the target sea area. The formula for calculating the second degree of deviation is:

[0038] in, Indicates the second degree of deviation. This represents the total number of mesh cells contained within the initial intrinsic influence zone. Indicates the first term within the initial endogenous influence zone. The level value corresponding to the initial productivity level of each grid cell. The value ranges from 1 to integers, This represents the grade value corresponding to the average primary productivity level of the target sea area. If the second deviation exceeds the preset internal deviation threshold, the initial internal influence zone is confirmed as an internal influence zone. The preset internal deviation threshold is set to 0.8, which is based on 1.5 times the overall standard deviation of the primary productivity level values ​​of the target sea area.

[0039] Example 3:

[0040] In practice, the dynamic correction module extracts the boundary lines of the surface features corresponding to the external influence zone, and uses these extracted boundary lines as the external boundary lines. The extraction of the external boundary lines employs an edge tracing algorithm. Specifically, starting from any boundary grid cell of the surface features in the external influence zone, it traces the adjacent external grid cells that belong to the boundary of the external influence zone in a clockwise direction, recording the coordinates of the center points of these boundary grid cells sequentially, and connecting them to form a closed external boundary line. Similarly, the dynamic correction module extracts the boundary lines of the surface features corresponding to the internal influence zone, and uses these extracted boundary lines as the internal boundary lines. All grid cells in the target sea area that do not belong to either the external or internal influence zone are marked as background areas, and the boundary lines of these background areas are also extracted using the surface feature edge tracing algorithm.

[0041] Centered on each boundary point on the extracted external boundary line, the dynamic correction module establishes an external buffer zone perpendicular to the tangent direction at that boundary point. The tangent direction at the boundary point is determined as follows: Take the preceding and following boundary points on the external boundary line adjacent to the current boundary point; subtract the coordinates of the preceding boundary point from the coordinates of the following boundary point to obtain the tangent vector; normalize the tangent vector to obtain the unit tangent vector. The normal vector perpendicular to the tangent direction is obtained by rotating the unit tangent vector counterclockwise by 90 degrees; the direction of the normal vector points outward from the external influence zone. The external buffer zone is set as a line segment extending a preset width along both the positive and negative directions of the normal vector, centered on the boundary point. The preset width is determined by multiplying the square root of the area of ​​the external influence zone by a scaling factor, which is set to 0.05. The internal buffer zone is established in the same way as the external buffer zone: Centered on each boundary point on the internal boundary line, an internal buffer zone perpendicular to the tangent direction at that boundary point is established; the normal vector within the internal buffer zone points outward from the internal influence zone.

[0042] The dynamic correction module acquires real-time tidal phase data. This data is obtained using a global tidal model. By inputting the current timestamp and the latitude and longitude coordinates of the boundary point, the model outputs the tidal level height and tidal ebb / flow status indicators at that point. When the tidal level height increases over time, the indicator is high tide; when it decreases, it is low tide. The tidal ebb / flow direction vector is calculated based on the tidal phase data. The direction of this vector is determined as follows: during high tide, the vector points towards the inward normal to the coastline; during low tide, it points towards the outward normal to the coastline. The magnitude of the tidal ebb / flow direction vector is set to 1 unit. The tidal current velocity scalar is calculated using the empirical relationship between current velocity and tidal amplitude. The tidal current velocity scalar is equal to the tidal amplitude multiplied by an empirical coefficient. The empirical coefficient is determined based on the linear regression slope of historical measured tidal amplitude and current velocity data for the target sea area. The empirical coefficient is between 0.4 and 0.8, and the empirical coefficient for the target sea area is set to 0.6.

[0043] The dynamic correction module acquires real-time river flux data into the sea. This data is collected from flow monitoring stations deployed along the coast of the target sea area at the river's inflow sections. Each station reports its flow rate every 15 minutes. The freshwater flux scalar is calculated as the sum of the flow rates reported by each major river monitoring station at the current moment. The suspended solids flux scalar is calculated as the sum of the products of the current flow rates of each major river and the corresponding suspended solids concentration. The suspended solids concentration of each river is obtained from turbidity sensors located at the same cross-section of the flow monitoring station through a turbidity-suspended solids concentration calibration curve.

[0044] The tidal rise and fall direction vector, tidal velocity scalar, freshwater flux into the sea scalar, and suspended solids flux scalar are input into the dynamic partition boundary correction function. The internal calculations of the dynamic partition boundary correction function are first performed on the external source boundary.

[0045] The first driving force component of the tide on the external boundary is obtained by performing a dot product operation between the tidal rise and fall direction vector and the normal vector of each point within the external buffer zone. Each point within the external buffer zone is a sequence of points sampled along the normal vector direction centered on the boundary point at a preset step size, which is one-tenth of the preset width. The normal vector at each sampling point is the same as the normal vector used when establishing the external buffer zone, because the normal vector direction within the buffer zone is constant at the same boundary point. The dot product operation is the sum of the products of the corresponding components of the two vectors, and the result is a scalar. If the dot product result is positive, it indicates that the angle between the tidal rise and fall direction and the buffer zone normal vector direction is less than 90 degrees, and the tide exerts an outward thrust on the external boundary; if the dot product result is negative, it indicates that the tide exerts an inward thrust on the external boundary. The arithmetic mean of the dot product results of all sampling points is taken to obtain the first driving force component.

[0046] Calculate the main driving force of the external boundary offset. The formula for calculating the main driving force of the external boundary offset is:

[0047] in, Indicates the main driving force of external boundary offset. Represents the scalar value of tidal current velocity. This represents a scalar quantity of freshwater flux into the sea. This represents the scalar quantity of suspended matter flux. To avoid singularities caused by a zero denominator when the suspended matter flux scalar quantity approaches zero, a small constant is added to the denominator, set to... . The larger the value, the stronger the driving force for the outward expansion of the external boundary.

[0048] Multiplying the first driving force component by the main driving intensity of the external boundary offset yields the direction adjustment coefficient and amplitude adjustment coefficient of the external offset vector at each boundary point. The direction adjustment coefficient is the sign of the product; a positive value indicates that the boundary point moves in the positive direction of the external buffer zone normal vector, and a negative value indicates that it moves in the opposite direction. The amplitude adjustment coefficient is equal to the absolute value of the product. The direction of the external offset vector is parallel to the direction of the external buffer zone normal vector. The magnitude of the external offset vector is equal to the amplitude adjustment coefficient multiplied by a length conversion factor. The length conversion factor converts the dimensionless driving intensity into actual spatial distance, and is set to one-third of the first-level coarse grid resolution of the target sea area, i.e., 0.033°.

[0049] The internal operation logic of the dynamic partition boundary correction function on the internal source boundary is as follows: The reverse direction of the tidal fluctuation direction vector is multiplied by the normal vector of each point within the internal source buffer zone. The reverse direction of the tidal fluctuation direction vector is obtained by multiplying the tidal fluctuation direction vector by -1. The calculation method for the dot product is the same as that for the external source part, yielding the second driving force component of the tide on the internal source boundary.

[0050] The primary driving force of the internal boundary migration is calculated. The primary driving force of the internal boundary migration is calculated by squaring the tidal velocity scalar and multiplying it by the sediment resuspension coefficient. The sediment resuspension coefficient is obtained by analyzing sediment samples from the target sea area to obtain the median grain size and critical initiation shear stress of the sediment. According to the non-cohesive sediment resuspension formula, the sediment resuspension coefficient is expressed as a function of the ratio of the square of the tidal velocity scalar to the critical initiation shear stress. Specifically, when the median grain size of the sediment is in the range of 0.02 mm to 0.1 mm, the sediment resuspension coefficient is set to 0.01. When the median grain size of the sediment exceeds this range, the sediment resuspension coefficient is adjusted by linear interpolation based on the reciprocal of the grain size.

[0051] Multiplying the second driving force component by the main driving intensity of the internal boundary offset yields the direction adjustment coefficient and amplitude adjustment coefficient of the internal offset vector at each boundary point of the internal boundary. The direction of the internal offset vector is parallel to the normal vector direction of the internal buffer zone, and the magnitude of the internal offset vector is equal to the amplitude adjustment coefficient multiplied by the same length conversion factor.

[0052] The dynamic correction module moves each boundary point on the external boundary line along its corresponding external offset vector. The moved coordinates are the original boundary point coordinates plus the offset of the external offset vector. Connecting all moved boundary points forms a closed, corrected external boundary line. Similarly, the dynamic correction module moves each boundary point on the internal boundary line along its corresponding internal offset vector, forming a closed, corrected internal boundary line. The background boundary line remains unchanged. The corrected external boundary line, the corrected internal boundary line, and the background boundary line are combined in the same spatial coordinate system. The polygon features of the three regions are regenerated based on their respective boundary lines, ensuring seamless connection between adjacent region boundaries and generating a real-time dynamic zoning map. The real-time dynamic zoning map is stored in vector polygon feature format. Each polygon feature's attributes include a zoning type field, with values ​​for external influence area, internal influence area, and background area, respectively.

[0053] See Figure 5 The horizontal axis in the figure represents time in hours, covering a continuous period of approximately 10 days. The left vertical axis represents the magnitude of the external boundary offset vector in degrees (°), and the right vertical axis represents the tidal current velocity scalar (in meters per second, m / s) and the freshwater flux into the sea scalar (scaled and labeled as freshwater flux into the sea scalar / 100, in meters per second, m / s).

[0054] The solid curve represents the trend of the magnitude of the external boundary offset vector over time. The curve fluctuates relatively little in the first 50 hours, remaining between -0.1° and 0.15°. After approximately 50 hours, the magnitude of the external boundary offset begins to fluctuate significantly, especially between 100 and 170 hours, with multiple peaks appearing. The maximum value exceeds 0.4°, and the minimum value is close to -0.6°, indicating that the external boundary underwent significant dynamic correction during this period, with both the offset direction and magnitude changing significantly. After 170 hours, the offset magnitude gradually decreases, the fluctuation amplitude diminishes, and it tends to stabilize close to zero.

[0055] The black dashed line represents the tidal velocity scalar, exhibiting typical tidal periodic fluctuations with a period of approximately 12.4 hours and an amplitude of about 0.5 m / s, maintaining a value between 0 and 0.5 m / s. The periodic variation of the tidal velocity provides the fundamental dynamic support for the periodic driving force of the external boundary offset.

[0056] The black dashed line represents the scalar flux of freshwater flowing into the sea, scaled 100 times and displayed on the right vertical axis. The overall trend is one of initial increase followed by a gradual decrease. In the first 80 hours, the freshwater flux gradually rises to near its peak (approximately 3 m / s), then begins to slowly decline, reaching a low of nearly 1 m / s by the 240-hour mark. The changes in the freshwater flux correspond to the fluctuations in the external boundary offset modulus, particularly during the high-flux phase, where the fluctuations in the external boundary offset modulus are significantly enhanced.

[0057] Overall, this figure reflects the temporal characteristics of the external boundary offset in the dynamic correction module, indicating that the magnitude of the external boundary offset vector is affected by two key driving factors: tidal current velocity and freshwater flux into the sea. Furthermore, the changing trends of these two factors are significantly correlated with the intensity of the boundary offset. The periodicity of the tidal current velocity provides continuous momentum for the boundary offset, while the variation in the freshwater flux into the sea modulates the intensity and amplitude of the boundary offset, reflecting the combined effect of the driving force components calculated within the dynamic partition boundary correction function and the main driving force intensity. The data in this figure conforms to the description of the external boundary offset mechanism and real-time data driving force of the dynamic correction module in Example 3.

[0058] Example 4:

[0059] In practical implementation, the zoning control module loads a real-time dynamic zoning map. The real-time dynamic zoning map is stored in vector polygon feature format. Each polygon feature contains a zoning type field, which has three enumerated values ​​corresponding to external influence areas, internal influence areas, and background areas. The zoning control module iterates through all polygon features in the real-time dynamic zoning map, reads the zoning type field value for each polygon feature, and classifies the polygon features into the corresponding zoning type set based on the field value, thus completing the identification of the three zoning types: external influence areas, internal influence areas, and background areas.

[0060] For the identified external source impact areas, the zoning control module configures upper limit thresholds for external pollutant concentrations and external nutrient fluxes. The upper limit threshold for external pollutant concentrations is determined as follows: the spatial range of all areal features corresponding to the external source impact area is extracted, and the pollutant concentration observations within this spatial range over the past five years are retrieved from the historical water quality monitoring database of the target sea area. Chemical oxygen demand (COD) concentration is selected as the pollutant concentration indicator, and the 90th percentile of the COD concentration over the past five years is used as the upper limit threshold for external pollutant concentration. The upper limit threshold for external nutrient fluxes is determined as follows: the monthly average total nitrogen flux and total phosphorus flux at the river estuary section within the external source impact area over the past five years are extracted, and the 85th percentile of the total nitrogen flux and the 85th percentile of the total phosphorus flux are calculated respectively. These two percentiles are used as the upper limit thresholds for total nitrogen flux and total phosphorus flux, respectively, and together they constitute the upper limit threshold for external nutrient fluxes. The zone control module sets the external source monitoring frequency parameter to a high-frequency acquisition mode for the external source influence zone. The sampling interval corresponding to the high-frequency acquisition mode is set to 30 minutes. This sampling interval is set based on one-sixth of the time scale of water quality parameter changes within the tidal cycle.

[0061] For the identified endogenous influence zones, the zonal control module configures upper thresholds for sediment disturbance intensity and lower thresholds for bottom dissolved oxygen. The upper threshold for sediment disturbance intensity is determined as follows: representative stations within the endogenous influence zone are selected for sediment sampling. The critical initiation shear stress of the sediment is determined through laboratory annular flume experiments. The critical initiation shear stress is multiplied by a safety factor to obtain the upper threshold for sediment disturbance intensity. The safety factor is set to 0.8, based on statistical analysis of the repeatability error of the sediment initiation experiment. The coefficient of variation for the repeatability experiment is approximately 20%, and the safety factor of 0.8 is obtained by reducing the coefficient of variation accordingly. The lower threshold for bottom dissolved oxygen is determined as follows: based on the ecological threshold of dissolved oxygen concentration tolerated by benthic organisms in the target sea area, the lower threshold for bottom dissolved oxygen is set to 2.0 mg / L. This setting is based on the average value of the critical dissolved oxygen concentration at which significant structural changes occur in the benthic animal community. The zonal control module sets the internal source monitoring depth parameter for the internal source influence zone as the bottom water body and sediment interface layer. The specific depth range of the bottom water body and sediment interface layer is 1 meter above the sediment-water interface to 0.1 meters below the sediment-water interface.

[0062] For the identified background area, the zoning control module configures upper and lower thresholds for chlorophyll concentration fluctuations. The upper threshold for chlorophyll concentration fluctuations is determined by extracting daily chlorophyll concentration time-series data from the past three years within the corresponding spatial range of the background area, calculating the absolute value of the difference in chlorophyll concentration between adjacent days to obtain a diurnal chlorophyll concentration fluctuation sequence, and taking the 95th percentile of this sequence as the upper threshold for chlorophyll concentration fluctuations. The lower threshold for transparency is determined by setting the lower threshold to the 10th percentile of the historical transparency observations of the background area, based on the optical characteristics of the target sea area's water. This setting is based on the minimum transparency level required to prevent significant degradation of the background area's water optical characteristics. The zoning control module sets the background monitoring frequency parameter to a low-frequency acquisition mode for the background area, with a sampling interval of 24 hours. This sampling interval is set based on complete coverage of the daily variation cycle of water quality parameters in the background area.

[0063] After threshold configuration and monitoring parameter settings are completed, the zone control module initiates a real-time data acquisition and threshold comparison process. For areas affected by external sources, the zone control module uses online chemical oxygen demand (COD) analyzers deployed at fixed monitoring stations within these areas to collect actual pollutant concentrations in real time. The COD analyzers automatically collect water samples every 30 minutes and output the COD concentration value as the actual pollutant concentration. The real-time collected actual pollutant concentration value is compared with the upper limit threshold for external pollutant concentration. When the actual pollutant concentration value exceeds the upper limit threshold, the zone control module generates an external source reduction command. This command includes the name of the pollutant exceeding the standard, the multiple of exceedance, and a trigger timestamp.

[0064] Simultaneously, the zoned control module collects real-time nutrient flux values ​​through nutrient flux monitoring devices deployed at river estuaries. These devices consist of a flow meter and an online nutrient analyzer. The flow meter collects the flow rate every 30 minutes, while the online analyzer simultaneously collects total nitrogen and total phosphorus concentrations. The actual total nitrogen flux value and actual total phosphorus flux value are obtained by multiplying the flow rate value by the total nitrogen and total phosphorus concentrations, respectively. These two values ​​together constitute the actual nutrient flux value. The actual total nitrogen flux value is compared with the upper limit threshold for total nitrogen flux, and the actual total phosphorus flux value is also compared with the upper limit threshold for total phosphorus flux. When either the actual total nitrogen flux value or the actual total phosphorus flux value exceeds the upper limit threshold for total nitrogen flux, the zoned control module generates a discharge restriction command for the sea. This command includes the name of the exceeding flux indicator and the exceeding percentage.

[0065] For the endogenous influence zone, the zonal control module collects the actual sediment disturbance intensity value in real time using sediment disturbance monitoring instruments deployed at the sediment interface within the endogenous influence zone. These instruments include shear force sensors, which continuously collect the shear stress value experienced by the sediment surface at a 1-minute interval as the actual sediment disturbance intensity value. Each collected shear stress value is compared with a sediment disturbance intensity upper limit threshold. When the shear stress value exceeds the threshold, the zonal control module generates a sediment solidification or covering command, which includes the disturbance intensity exceeding the limit value and the trigger location coordinates.

[0066] Simultaneously, the zone control module collects real-time dissolved oxygen values ​​in the bottom water of the endogenous influence zone using dissolved oxygen sensors deployed 0.5 meters above the sediment-water interface. The sensors collect dissolved oxygen concentration values ​​every 5 minutes as the actual bottom dissolved oxygen value. The actual bottom dissolved oxygen value is compared with a lower limit threshold. When the actual bottom dissolved oxygen value falls below the lower limit threshold, the zone control module generates a bottom oxygenation command, which includes the current dissolved oxygen concentration and the target dissolved oxygen concentration increase value.

[0067] For the background area, the zoning control module collects real-time fluctuations in actual chlorophyll concentration using fluorescence chlorophyll sensors deployed at a preset background monitoring grid. The fluorescence chlorophyll sensors collect raw chlorophyll concentration values ​​at 24-hour intervals. The actual chlorophyll concentration fluctuation is calculated by subtracting the raw chlorophyll concentration value collected at the previous moment from the current raw chlorophyll concentration value collected by the fluorescence chlorophyll sensor, and taking the absolute value of the difference as the current actual chlorophyll concentration fluctuation value. The actual chlorophyll concentration fluctuation value is compared with a chlorophyll concentration fluctuation upper limit threshold. When the actual chlorophyll concentration fluctuation value exceeds the upper limit threshold, the zoning control module generates a background area key inspection command, which includes the coordinates of the location where the fluctuation exceeds the limit and the fluctuation amplitude.

[0068] Simultaneously, the zone control module collects real-time actual transparency values ​​using online transparency monitoring devices deployed within the background zone. These devices utilize the beam attenuation method and automatically measure water transparency every 24 hours as the actual transparency value. The actual transparency value is then compared to a lower transparency threshold. When the actual transparency value falls below this threshold, the zone control module generates a suspended matter tracing command, which includes the degree of transparency reduction and the coordinates of the trigger location.

[0069] See Figure 6The figure shows the real-time monitoring data change curve of the actual chemical oxygen demand (COD) concentration in the external source influence area by the zone control module in Example 4. The horizontal axis represents the monitoring time in hours, ranging from 0 to 240 hours, covering a continuous 10-day monitoring cycle; the vertical axis represents the COD concentration in milligrams per liter (mg / L), with a range set from 0 to 10 mg / L. The solid curve depicts the temporal trend of the actual COD concentration in the external source influence area, while the dashed line represents the pre-configured upper limit threshold of the external pollutant concentration, with a value of approximately 4.8 mg / L, serving as the standard for judging whether the pollutant concentration exceeds the limit.

[0070] The graph shows three distinct concentration peaks, located at approximately the 55th, 175th, and 217th hours, with peak concentrations of approximately 9.5 mg / L, 6.3 mg / L, and 7.5 mg / L, respectively, all significantly exceeding the set threshold of 4.8 mg / L. Each peak region is marked with a shaded line, indicating that the actual concentration exceeded the threshold during that time period, falling within the range of exogenous exceedances. The curves show rapid upward and downward trends before and after the exceedance peaks, demonstrating the sudden and short-duration peak characteristics of pollutant concentrations.

[0071] This figure illustrates the specific application of the real-time data acquisition and threshold comparison functions of the zone control module, which can effectively capture chemical oxygen demand (COD) exceedance events within the external influence zone. Based on the threshold configuration and monitoring parameter settings in Example 4, the online COD analyzer automatically samples every 30 minutes, dynamically monitors pollutant concentrations, and triggers external source reduction commands when exceedances occur. The occurrence of exceedance intervals and their peak amplitudes in the figure provide a clear data signal basis for the zone control module to generate reduction commands, ensuring timely response and effective implementation of management measures.

[0072] Example 5:

[0073] In practical implementation, the zonal control module refines the generation process of external source reduction instructions. In-situ water quality monitoring probes are deployed at the river estuary sections within the external source influence zone. These probes employ ultraviolet-visible absorption spectroscopy sensors to continuously detect the chemical oxygen demand (COD) concentration in the water. The probes operate in a high-frequency acquisition mode, set according to the external source monitoring frequency parameters. The first sampling interval in this high-frequency acquisition mode is set to 30 minutes, based on one-quarter of the characteristic time of water quality changes within the tidal semi-diurnal cycle. The probes collect and analyze water samples every 30 minutes, and the output COD concentration value is used as the actual pollutant concentration value.

[0074] After each data collection, the zone control module calculates the difference between the actual pollutant concentration value collected and the pre-configured upper limit threshold for external pollutant concentration. The difference between the actual pollutant concentration value and the upper limit threshold is taken as the external pollution exceedance. The module then checks if the external pollution exceedance is greater than zero. If it is, a timer is started to record the duration of the exceedance. The timer is reset when the exceedance drops to zero or is less than zero. If the duration of the continuously greater than zero exceedance exceeds the preset external pollution response delay time, the module proceeds to the external pollution reduction measure decision-making stage. The preset external pollution response delay time is set to 3 hours, based on twice the average water transport time between the river estuary section and a representative monitoring point within the external pollution impact zone. The average water transport time is obtained from historical tracer experimental data.

[0075] In the decision-making process for external source mitigation measures, the zonal control module queries the pre-stored external source mitigation measure database for the corresponding mitigation measure level based on the magnitude of the external source exceedance. The external source mitigation measure database is a structured data table stored in the zonal control module's memory. The data table contains two fields: an exceedance range field and a mitigation measure level field. The exceedance range field defines five consecutive numerical ranges: the first range is where the exceedance is greater than 0 and less than or equal to 20% of the upper limit threshold for external pollutant concentration; the second range is where the exceedance is greater than 20% and less than or equal to 50% of the upper limit threshold; the third range is where the exceedance is greater than 50% and less than or equal to 100%; the fourth range is where the exceedance is greater than 100% and less than or equal to 200%; and the fifth range is where the exceedance is greater than 200%. The mitigation measure level field assigns a level of 1, 2, 3, 4, and 5 to each range, with higher levels indicating stronger mitigation measures required. The zonal control module maps the current external source exceedance to the corresponding exceedance range and retrieves the corresponding mitigation measure level.

[0076] Based on the queried level of external pollution reduction measures, the zone control module generates an external pollution reduction instruction containing specific execution parameters. The external pollution reduction instruction is a structured data message; the message body includes the activation identifier and operating power parameters of the reduction measure execution device. The reduction measure execution device refers to the control interface of land-based pollution reduction facilities deployed near the river's estuary, such as wastewater treatment plants or bypass treatment systems. The activation identifier is set to a Boolean true value, indicating that the facility has entered the reduction operation state. The operating power parameters are determined according to the reduction measure level: Level 1 corresponds to 60% of the facility's rated power, Level 2 to 80%, Level 3 to 100%, Level 4 to 100% with the activation of standby treatment units, and Level 5 to 100% with the activation of all standby treatment units and the emergency chemical treatment module. The external pollution reduction instruction is transmitted to the reduction measure execution device via a wireless communication link.

[0077] During the generation of sediment solidification or cover commands, the zone control module deploys sediment disturbance monitoring instruments at the sediment interface in the endogenous influence zone. The sediment disturbance monitoring instrument is an integrated seabed observation platform, fixedly installed at representative stations in the endogenous influence zone, at water depths between 5 and 20 meters. The instrument includes a shear force sensor and a pore water pressure sensor. The shear force sensor uses a thermal pulse shear force measurement probe, inserted 0.05 meters below the sediment-water interface, continuously collecting shear stress values ​​on the sediment surface at 1-minute intervals, using each collected shear stress value as the actual sediment disturbance intensity. The pore water pressure sensor uses a fiber optic grating pressure sensor, buried 0.2 meters below the sediment-water interface, and is in standby mode.

[0078] The zone control module compares the shear stress value collected each time with the pre-configured upper limit threshold for sediment disturbance intensity. When the shear stress value exceeds the upper limit threshold for sediment disturbance intensity, the zone control module immediately activates the pore water pressure sensor. The pore water pressure sensor continuously collects the pore water pressure value of the current sediment at a period of 1 second for 3 minutes, and takes the arithmetic mean of the pore water pressure values ​​within 3 minutes as the representative value of the current sediment pore water pressure.

[0079] Based on the collected representative values ​​of pore water pressure, the zonal control module calculates the liquefaction potential index of the sediment. The liquefaction potential index is calculated as the ratio of the representative value of pore water pressure to the effective overlying stress at the sediment location. The effective overlying stress at the sediment location is calculated by multiplying the sediment bulk density by the burial depth of the pore water pressure sensor. The sediment bulk density is obtained through on-site sampling and is measured to be between 1.6 g / cm³ and 2.0 g / cm³. The liquefaction potential index is greater than or equal to 0. When the liquefaction potential index is greater than or equal to 1.0, it indicates that the sediment is in a fully liquefied state. The module then determines whether the liquefaction potential index exceeds a preset liquefaction threshold, which is set to 0.8. This threshold is based on an analysis of the relationship between the measured liquefaction potential index and sediment erosion flux during historical sediment resuspension events in the target sea area. A sharp increase in sediment erosion flux occurs when the liquefaction potential index reaches 0.8.

[0080] If the liquefaction potential index exceeds the liquefaction threshold, the zone control module generates a sediment solidification command. This command includes the coordinates of the solidifier placement location and the solidifier dosage parameters. The solidifier placement location coordinates are set to the latitude and longitude coordinates of the sediment disturbance monitoring instrument's installation location, and a circular placement area with a radius of 50 meters is generated centered on these coordinates. The solidifier dosage parameters are calculated as follows: multiply the placement area by the required sediment reinforcement thickness of 0.5 meters to obtain the required solidification volume, then multiply by the required solidifier admixture ratio of 15% per unit volume of sediment to obtain the total solidifier mass. Cement-based solidification materials are selected as the solidifier.

[0081] If the liquefaction potential index does not exceed the liquefaction threshold, the zone control module generates a sediment cover instruction. The sediment cover instruction includes parameters for the cover material type and the cover layer thickness. The cover material type parameter is determined as follows: when the proportion of shear stress values ​​exceeding the upper limit threshold for sediment disturbance intensity is less than 50%, the cover material type parameter is set to sand; when the proportion of shear stress values ​​exceeding the upper limit threshold for sediment disturbance intensity is greater than or equal to 50% but less than 100%, the cover material type parameter is set to gravel; when the proportion of shear stress values ​​exceeding the upper limit threshold for sediment disturbance intensity is greater than or equal to 100%, the cover material type parameter is set to a geotextile plus gravel composite layer. The cover layer thickness parameter is determined based on the proportion of shear stress values ​​exceeding the threshold, and the cover layer thickness is set to 0.3 meters plus the excess proportion multiplied by 0.7 meters, with an upper limit of 1.0 meter.

[0082] During the generation of the background area key inspection instruction, the zoning control module deploys multiple fluorescent chlorophyll sensors within the background area according to a preset background monitoring grid. The preset background monitoring grid uses an equidistant grid layout with a grid spacing of 1 kilometer, covering the entire background area surface. Each grid point deploys one fluorescent chlorophyll sensor, employing a dual-band fluorescence analysis method with an excitation wavelength of 470 nm and a detection wavelength of 685 nm. Each fluorescent chlorophyll sensor operates in a low-frequency acquisition mode set according to the background monitoring frequency parameters. In this low-frequency acquisition mode, the second sampling interval is set to 24 hours, based on complete sampling of the daily variation cycle of chlorophyll concentration in the background area. The fluorescent chlorophyll sensor automatically collects a water sample once daily at 0:00, outputting the raw chlorophyll concentration value.

[0083] For each fluorescence-based chlorophyll sensor, the zonal control module maintains a first-in-first-out queue of length 2 in memory, storing the raw chlorophyll concentration values ​​collected at the current and previous times. The difference between the raw chlorophyll concentration values ​​collected at the current and previous times is calculated, and the absolute value of the difference is taken to obtain the single-point chlorophyll concentration fluctuation value at the sensor's location.

[0084] The arithmetic mean of the chlorophyll concentration fluctuation values ​​at each individual point corresponding to all fluorescence-based chlorophyll sensors within the background area is calculated to obtain the overall average chlorophyll concentration fluctuation value of the background area. Simultaneously, the standard deviation of all individual point chlorophyll concentration fluctuation values ​​is calculated to obtain the background area fluctuation dispersion. The formula for calculating the background area fluctuation dispersion is:

[0085] in, Indicates the dispersion of fluctuations in the background region. This indicates the total number of fluorescence chlorophyll sensors in the background area. Indicates the first The single-point chlorophyll concentration fluctuation value at the location of the fluorescence chlorophyll sensor. The value ranges from 1 to integers, This represents the average fluctuation of chlorophyll concentration in the overall background area.

[0086] The zone control module compares the average fluctuation of the overall chlorophyll concentration in the background zone with a pre-configured upper limit threshold for chlorophyll concentration fluctuation. If the average fluctuation of the overall chlorophyll concentration in the background zone exceeds the upper limit threshold, it further determines whether the fluctuation dispersion of the background zone exceeds a preset dispersion threshold. The preset dispersion threshold is set to the 90th percentile of the historical fluctuation dispersion data sequence of the background zone, which consists of the daily fluctuation dispersion values ​​of the background zone calculated over the past three years.

[0087] If the background area fluctuation dispersion exceeds the dispersion threshold, the zoning control module identifies the locations of fluorescence chlorophyll sensors where the single-point chlorophyll concentration fluctuation value exceeds the upper limit threshold of chlorophyll concentration fluctuation, and marks these locations as hotspot anomalies. The zoning control module generates a background area key inspection instruction, which is a structured data message containing a list of hotspot anomaly coordinates, where each element is a pair of latitude and longitude coordinates of the hotspot anomaly. The background area key inspection instruction is sent to the patrol vessel's onboard terminal or the UAV control platform via the communication network, instructing the patrol equipment to proceed to the site for verification according to the hotspot anomaly coordinates.

[0088] If the background area fluctuation dispersion does not exceed the dispersion threshold, the zoning control module generates a full-area patrol command for the background area. This command includes a patrol line covering the entire background area's surface features. The patrol line originates from the geometric center point of the background area's surface features and extends outwards in an Archimedean spiral pattern, sequentially traversing all grid points. The full-area patrol command is also sent to the patrol equipment control platform.

[0089] See Figure 7 In the figure, the horizontal axis represents the numbers of the chlorophyll fluorescence sensors deployed in the background area, totaling 500 sensors; the vertical axis represents the chlorophyll concentration fluctuation value at each sensor location, in micrograms per liter (μg / L). Points marked with circles "○" represent normal grid points, whose chlorophyll concentration fluctuation values ​​are all below the preset upper limit threshold for chlorophyll concentration fluctuation (represented by the black dashed line, approximately 2.3 μg / L), reflecting the normal range of chlorophyll concentration variation in the background area. Points marked with stars "★" represent hotspot anomaly points, whose chlorophyll concentration fluctuation values ​​are significantly higher than the upper limit threshold for chlorophyll concentration fluctuation, reaching a maximum of approximately 8 μg / L, clearly deviating from the normal range, indicating abnormal fluctuations in chlorophyll concentration in the area monitored by these sensors.

[0090] The black dashed line in the figure represents the upper limit threshold for chlorophyll concentration fluctuation, which is a key threshold determined based on the 95th percentile of the daily chlorophyll concentration difference in the background area over the past three years. The black dotted line represents the average chlorophyll concentration fluctuation of the overall background area, which is around 2 μg / L, lower than the upper limit threshold, indicating that the overall chlorophyll concentration fluctuation of the background area is within a reasonable range.

[0091] The comparison shows that the chlorophyll concentration fluctuations of the vast majority of sensors are stable and within limits. Only some sensor numbers correspond to grid points that exhibit abnormal fluctuations, forming hotspot anomalies. This distribution characteristic conforms to the technical solution for generating background area key inspection instructions in Example 5. The zoning control module identifies anomalies where the chlorophyll concentration fluctuation exceeds the threshold by comparing the average chlorophyll concentration fluctuation of the overall background area with the threshold and by analyzing the fluctuation dispersion. These anomalies are marked as hotspot anomalies, and key inspection instructions are sent to the inspection equipment accordingly.

[0092] This figure effectively demonstrates the spatial distribution of chlorophyll concentration fluctuations and the results of anomaly identification within the background area, providing data support for the key inspection of the background area in Example 5. It also reflects the real-time monitoring and anomaly warning function of the zoning control module for chlorophyll concentration fluctuations in the background area.

[0093] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A dynamic assessment and zoning management system for marine ecological environment, characterized in that, include: The data acquisition module acquires long-term remote sensing image data of the target sea area and extracts the chlorophyll concentration distribution characteristics, sea surface temperature gradient change characteristics, and suspended matter sedimentation and diffusion trajectory characteristics from each remote sensing image. The assessment modeling module, based on the chlorophyll concentration distribution characteristics and sea surface temperature gradient variation characteristics, constructs a multi-scale nested assessment model to characterize the spatiotemporal evolution of primary productivity in the sea area. This multi-scale nested assessment model is then used to perform a gridded assessment of the target sea area, generating the primary productivity level for each grid cell, including: The target sea area is divided into a first-level coarse grid, and each first-level coarse grid cell is further divided into a second-level fine grid. The spatial resolution of the second-level fine grid is higher than that of the first-level coarse grid. Taking each first-level coarse grid cell as a unit, calculate the average rate of change of chlorophyll concentration of all pixels in the first-level coarse grid cell to obtain the first-level chlorophyll average value. At the same time, calculate the average sea surface temperature gradient value of all pixels in the first-level coarse grid cell to obtain the first-level temperature gradient average value. The mean value of first-level chlorophyll and the mean value of first-level temperature gradient are input into the pre-trained primary productivity regression model, and the basic primary productivity estimate of the first-level coarse grid cell is output. Using each secondary fine grid cell as a unit, the difference between the chlorophyll concentration change rate of all pixels in the secondary fine grid cell and the mean value of the primary chlorophyll in the primary coarse grid cell is calculated to obtain the secondary chlorophyll residual value. At the same time, the difference between the sea surface temperature gradient value of all pixels in the secondary fine grid cell and the mean value of the primary temperature gradient in the primary coarse grid cell is calculated to obtain the secondary temperature gradient residual value. The secondary chlorophyll residual value and the secondary temperature gradient residual value are input into the residual correction branch of the pre-trained primary productivity regression model to obtain the primary productivity correction value of the secondary fine grid cell. The final primary productivity assessment value of the second-level fine grid cell is obtained by adding the basic primary productivity estimate to the primary productivity correction value. Based on a preset productivity grading threshold, the final primary productivity assessment value is mapped to the primary productivity level; The source analysis module performs spatial coupling analysis on the characteristics of the suspended matter sedimentation and diffusion trajectory and the primary productivity level of each grid cell to identify the exogenous influence zone dominated by terrestrial input and the endogenous influence zone dominated by sediment resuspension. The dynamic correction module establishes a dynamic partition boundary correction function based on the spatial distribution boundaries of the external and internal influence zones. It iteratively corrects the spatial distribution boundaries according to the real-time acquired tidal phase change data and river inflow flux data to generate a real-time dynamic partition map. The zoning control module configures corresponding ecological control thresholds for each zoning type in the real-time dynamic zoning map, and performs zoning management operations based on the ecological control thresholds.

2. The marine ecological environment dynamic assessment and zoning management system according to claim 1, characterized in that, The acquisition of long-term remote sensing image data of the target sea area, and the extraction of chlorophyll concentration distribution characteristics, sea surface temperature gradient change characteristics, and suspended matter sedimentation and diffusion trajectory characteristics from each remote sensing image, includes: According to a preset time sampling interval, remote sensing images with cloud coverage rates lower than a preset cloud threshold are selected from the remote sensing database to form the long-term remote sensing image data. For each remote sensing image in the long-term remote sensing image data, chlorophyll concentration band values ​​are extracted through band operations, and the difference between the chlorophyll concentration band values ​​at two adjacent sampling times is calculated to obtain the chlorophyll concentration change rate as the chlorophyll concentration distribution feature. For each remote sensing image in the long-term remote sensing image data, the sea surface temperature value is retrieved by thermal infrared band, and the absolute value of the temperature difference between each pixel and its eight neighboring pixels is calculated. The maximum value of the absolute value of the temperature difference is taken as the sea surface temperature gradient value of that pixel. The sea surface temperature gradient values ​​of all pixels are summarized to form the sea surface temperature gradient change feature. For three or more consecutive remote sensing images in the long-term remote sensing image data, the suspended matter concentration band values ​​in each remote sensing image are extracted. The optical flow method is used to track the movement trajectory of the peak position of the suspended matter concentration band at different times, and the movement trajectory is used as the feature of the suspended matter sedimentation and diffusion trajectory.

3. The marine ecological environment dynamic assessment and zoning management system according to claim 2, characterized in that, The spatial coupling analysis of the suspended matter sedimentation and diffusion trajectory characteristics with the primary productivity level of each grid cell identifies the exogenous influence zone dominated by terrestrial input and the endogenous influence zone dominated by sediment resuspension, including: The suspended matter sedimentation and diffusion trajectory features are plotted on a grid base map of the target sea area to obtain a suspended matter diffusion trajectory line layer. The primary productivity level of each grid cell is marked in the same spatial coordinate system to obtain a primary productivity level distribution layer. For each diffusion trajectory line in the suspended matter diffusion trajectory line layer, the starting position of the diffusion trajectory line is identified. If the starting position falls within the buffer zone of the river estuary in the target sea area, the diffusion trajectory line is marked as an external trajectory line. For each diffusion trajectory line in the suspended matter diffusion trajectory line layer, the endpoint position of the diffusion trajectory line is identified. If the difference between the suspended matter concentration band values ​​at two adjacent sampling times at the endpoint position exceeds a preset sedimentation threshold, the diffusion trajectory line is marked as a sedimentation trajectory line. Extract the grid cells covered by three or more consecutive external trajectory lines closest to the river estuary buffer zone from the external trajectory lines, and mark these grid cells as the initial external influence zone; Extract grid cells from the settlement trajectory lines that do not spatially overlap with the initial external influence zone and whose settlement trajectory line density exceeds a preset density threshold, and mark these grid cells as the initial internal influence zone; Calculate the first deviation between the primary productivity level of each grid cell in the initial external influence zone and the average primary productivity level of the target sea area. If the first deviation exceeds the preset external deviation threshold, then the initial external influence zone is confirmed as the external influence zone. Calculate the second degree of deviation between the primary productivity level of each grid cell in the initial internal influence zone and the average primary productivity level of the target sea area. If the second degree of deviation exceeds a preset internal deviation threshold, then the initial internal influence zone is confirmed as the internal influence zone.

4. The marine ecological environment dynamic assessment and zoning management system according to claim 3, characterized in that, The conditions for confirming the external source influence zone also include: the spatial distance between the starting point of the suspended matter sedimentation and diffusion trajectory in the initial external source influence zone and the river estuary is less than a preset source-sink distance threshold, and there is a time lag correlation between the time-series peak value of the chlorophyll concentration change rate in the initial external source influence zone and the time-series peak value of the river runoff.

5. The marine ecological environment dynamic assessment and zoning management system according to claim 4, characterized in that, Based on the spatial distribution boundaries of the external and internal influence zones, a dynamic zoning boundary correction function is established. Using real-time acquired tidal phase change data and river flux data, the spatial distribution boundaries are iteratively corrected to generate a real-time dynamic zoning map, including: The spatial distribution boundary line of the external source influence area is extracted as the external source boundary line, and the spatial distribution boundary line of the internal source influence area is extracted as the internal source boundary line. The areas in the target sea area that do not belong to the external source influence area and the internal source influence area are marked as background areas. With each boundary point on the external source boundary line as the center, an external source buffer zone is established perpendicular to the tangent direction at that boundary point; with each boundary point on the internal source boundary line as the center, an internal source buffer zone is established perpendicular to the tangent direction at that boundary point. Acquire real-time tidal phase data, and calculate the tidal rise and fall direction vector and tidal velocity scalar at the current moment based on the tidal phase data; Obtain real-time river flux data into the sea, and calculate the current freshwater flux scalar and suspended solids flux scalar based on the river flux data into the sea. The tidal rise and fall direction vector, tidal velocity scalar, freshwater flux into the sea scalar, and suspended solids flux scalar are input into the dynamic partition boundary correction function, which outputs the external offset vector of each boundary point on the external boundary line and the internal offset vector of each boundary point on the internal boundary line. Each boundary point on the external source boundary line is moved along the corresponding external source offset vector to obtain the corrected external source boundary line. Each boundary point on the internal source boundary line is moved along the corresponding internal source offset vector to obtain the corrected internal source boundary line. The modified external boundary line, the modified internal boundary line, and the background boundary line are combined to generate the real-time dynamic partition map.

6. The marine ecological environment dynamic assessment and zoning management system according to claim 5, characterized in that, The internal operation logic of the dynamic partition boundary correction function is as follows: The first driving force component of the tide on the external boundary is obtained by performing a dot product operation between the tidal rise and fall direction vector and the normal vector of each point in the external buffer zone. Multiply the tidal current velocity scalar by the freshwater flux into the sea scalar and divide by the suspended solids flux scalar to obtain the main driving force of the external boundary offset. Multiply the first driving force component by the main driving intensity to obtain the direction adjustment coefficient and amplitude adjustment coefficient of the external offset vector, thereby determining the external offset vector of each boundary point; The second driving force component of the tide on the internal boundary is obtained by performing a dot product operation between the opposite direction of the tidal rise and fall direction vector and the normal vector of each point in the internal buffer zone. The main driving force of the internal boundary offset is obtained by squaring the tidal velocity scalar and multiplying it by the bottom sediment resuspension coefficient. Multiply the second driving force component by the main driving intensity of the internal boundary offset to obtain the direction adjustment coefficient and amplitude adjustment coefficient of the internal offset vector, thereby determining the internal offset vector of each boundary point.

7. The marine ecological environment dynamic assessment and zoning management system according to claim 6, characterized in that, Based on the partition types in the real-time dynamic partitioning diagram, corresponding ecological regulation thresholds are configured, and partition management operations are performed according to the ecological regulation thresholds, including: The real-time dynamic partition map is analyzed to identify three partition types: external influence area, internal influence area, and background area. For the identified external source impact areas, configure upper limit thresholds for external pollutant concentrations and external nutrient fluxes, and set the external source monitoring frequency parameter to high-frequency acquisition mode; For the identified endogenous influence zone, configure the upper limit threshold of sediment disturbance intensity and the lower limit threshold of dissolved oxygen in the bottom layer, and set the endogenous monitoring depth parameter as the interface layer between the bottom water body and the sediment. For the identified background area, configure the upper limit threshold for chlorophyll concentration fluctuation and the lower limit threshold for transparency, and set the background monitoring frequency parameter to low frequency acquisition mode; The actual pollutant concentration value in the external source influence area is collected in real time, and the actual pollutant concentration value is compared with the upper limit threshold of the external source pollutant concentration. When the actual pollutant concentration value exceeds the upper limit threshold of the external source pollutant concentration, an external source reduction command is generated. The actual nutrient flux value in the external influence area is collected in real time, and the actual nutrient flux value is compared with the upper limit threshold of the external nutrient flux. When the actual nutrient flux value exceeds the upper limit threshold of the external nutrient flux, a discharge restriction order is generated. The actual sediment disturbance intensity value in the internal influence zone is collected in real time. The actual sediment disturbance intensity value is compared with the upper limit threshold of sediment disturbance intensity. When the actual sediment disturbance intensity value exceeds the upper limit threshold of sediment disturbance intensity, a sediment solidification or covering instruction is generated. The actual dissolved oxygen value of the bottom layer in the endogenous influence zone is collected in real time, and the actual dissolved oxygen value of the bottom layer is compared with the lower limit threshold of the bottom layer dissolved oxygen. When the actual dissolved oxygen value of the bottom layer is lower than the lower limit threshold of the bottom layer dissolved oxygen, a bottom layer oxygenation command is generated. The actual chlorophyll concentration fluctuation value in the background area is collected in real time, and the actual chlorophyll concentration fluctuation value is compared with the upper limit threshold of chlorophyll concentration fluctuation. When the actual chlorophyll concentration fluctuation value exceeds the upper limit threshold of chlorophyll concentration fluctuation, a key inspection instruction for the background area is generated. The actual transparency value in the background area is collected in real time and compared with the lower limit threshold of transparency. When the actual transparency value is lower than the lower limit threshold of transparency, a suspended matter tracing instruction is generated.

8. The marine ecological environment dynamic assessment and zoning management system according to claim 1, characterized in that, The multi-scale nested evaluation model adopts a nested structure combining a first-level coarse grid and a second-level fine grid. The first-level coarse grid is used to capture the background trend of primary productivity at a large scale, while the second-level fine grid is used to capture abnormal fluctuations in primary productivity at a local scale.

9. The marine ecological environment dynamic assessment and zoning management system according to claim 1, characterized in that, The ecological regulation thresholds include the upper limit thresholds for the concentration of exogenous pollutants and the upper limit thresholds for the flux of exogenous nutrients in the exogenous influence zone, the upper limit thresholds for the intensity of sediment disturbance and the lower limit thresholds for dissolved oxygen in the bottom layer in the endogenous influence zone, and the upper limit thresholds for the fluctuation of chlorophyll concentration and the lower limit thresholds for transparency in the background zone.

Images