A highly migratory fish cross-border resource correlation degree dynamic calculation method and system

Abstract

This invention discloses a method and system for dynamically calculating the cross-border resource correlation of highly migratory fish species, belonging to the field of marine fisheries resource monitoring and data processing technology. The method acquires basic fish species data, migration route data, sea area boundary data, fishing behavior data, marine environmental data, and resource status data, and performs coordinate, time, and fish species coding standardization processing to construct a spatiotemporal grid analysis system. Based on this, it calculates resource occurrence probability parameters, spatial overlap parameters, and fishing impact parameters, and constructs a correction mechanism combining resource status and data confidence levels. A cross-border resource correlation score is generated through dynamic weight fusion. This system realizes dynamic quantitative analysis of the resource correlation of highly migratory fish species in cross-border sea areas, and can be used in fisheries resource monitoring, fishing supervision, aquatic product traceability, and IUU fishing risk identification, improving the accuracy and interpretability of cross-border resource assessment.

Description

Technical Field

[0001] This invention relates to the field of marine fishery resource monitoring and intelligent data processing technology, and in particular to a method and system for dynamically calculating the cross-border resource correlation of highly migratory fish. Background Technology

[0002] Highly migratory fish species (HMFS) refer to fish resources that migrate across a large spatial range and periodically or in stages between different sea areas throughout their life cycle. Typical species include tuna, skipjack tuna, swordfish, mackerel, herring, and eel. These fish species are typically characterized by wide ranges, long migration distances, significant differences in growth stages, and migration across multiple national jurisdictional waters and high seas. Their resource distribution and population changes are influenced by a variety of factors, including marine environmental conditions, changes in resource status, and fishing activities.

[0003] With increasingly stringent requirements for global marine fisheries resource management, seafood trade regulation, and marine ecological protection, the need for resource monitoring, fishing regulation, seafood traceability verification, and risk identification of illegal, unreported, and unregulated (IUU) fishing for highly migratory fish species is constantly growing. In practical applications, regulatory agencies, resource management agencies, and trade regulatory departments typically need to determine the degree of correlation between specific fishing entities, specific fishing areas, specific transport batches, or specific import batches and the cross-border resources of a particular highly migratory fish species to assist in resource conservation decisions, risk warnings, law enforcement supervision, and trade compliance verification.

[0004] Currently, the monitoring and assessment of highly migratory fish resources typically utilizes a variety of information sources, including fish tagging and release data, electronic tag data, satellite remote sensing data, marine environmental monitoring data, Automatic Identification System (AIS) data, Vessel Monitoring System (VMS) data, catch log data, catch statistics, and resource assessment data released by regional fisheries management organizations. These data can respectively reflect fish migration patterns, changes in the marine environment, the distribution of fishing activities, and changes in resource status.

[0005] However, due to the complexity of the data sources and the significant differences in data structure, collection standards, and update frequency, practical applications commonly encounter problems such as inconsistent coordinate benchmarks, inconsistent time granularity, inconsistent fish species identification, data gaps, and significant differences in data quality. It is often difficult to establish unified data relationships between data from different sources, making it challenging to comprehensively analyze fish migration routes, marine boundary information, fishing behavior information, and resource status information within a unified spatiotemporal framework.

[0006] Meanwhile, some existing resource assessment methods primarily predict the abundance of target fish species based on historical resource survey data, catch statistics, or environmental factor data; some vessel monitoring methods mainly identify suspected fishing activities based on AIS or VMS trajectories; and some geographic information system methods are mainly used to analyze the spatial overlap between sea area boundaries and target areas. Most of these technical solutions are designed for single business scenarios, and their analysis objects are usually limited to one aspect of resource assessment, vessel monitoring, or spatial analysis, lacking a comprehensive calculation mechanism that can simultaneously integrate fish migration routes, cross-border sea areas, fishing activity characteristics, and changes in resource status.

[0007] Especially for highly migratory fish species, their resource distribution exhibits significant temporal and spatial dynamics. The same fish species often displays different activity ranges and migration patterns during the breeding, feeding, overwintering, and migratory periods; the origins of fish populations in the same sea area may also vary considerably across different time windows. Therefore, correlation analysis based solely on fixed sea area boundaries, static resource distribution maps, or single fishing locations is insufficient to accurately reflect the actual cross-border resource flows and resource relationships of the target fish species.

[0008] Furthermore, existing technologies have limitations in analyzing the relationship between fishing activities and target resources. In actual regulatory processes, the impact of fishing activities on target resources is not only related to the fishing location, but also closely related to the fishing time, operational intensity, catch size, operation duration, and the resource distribution status of the target fish species within the corresponding time window. Existing technologies typically lack the technical means to perform spatiotemporal matching analysis between vessel fishing activities and the dynamic distribution of fish resources, thus making it difficult to accurately quantify the actual impact of fishing activities on target transboundary resources.

[0009] Furthermore, in applications such as risk assessment of imported aquatic products, traceability verification of aquatic products, and risk identification of IUU (individually caught fish) fishing, the regulatory targets typically involve multiple stages, including vessels, catches, transshipment, processing, transportation, and trade. Existing technologies often lack dynamic models capable of unified correlation analysis of fish migration route information, marine spatial relationship information, fishing behavior information, and resource status information, making it difficult to accurately determine the degree of correlation between the target fishing entity, target fishing area, or target trade batch and cross-border resources of specific migratory fish species.

[0010] Furthermore, most existing risk assessment or resource association analysis methods employ fixed evaluation weights or fixed grading thresholds. When the target fish species is in its breeding season, resource decline period, or high fishing pressure, fixed evaluation models struggle to fully reflect changes in resource sensitivity. Conversely, when data is missing, conflicting, or has low confidence levels, fixed threshold models are prone to misjudgments or omissions, thus affecting the accuracy and reliability of the association analysis results.

[0011] Therefore, there is an urgent need for a method and system for dynamically calculating the cross-border resource correlation of highly migratory fish species. This system should be able to perform unified and standardized processing and spatiotemporal grid management of fish migration route data, marine environmental data, target sea area boundary data, fishing behavior data, and resource status data. Under a unified spatiotemporal framework, it should calculate resource occurrence probability parameters, spatial overlap parameters, and fishing impact parameters. Furthermore, by combining resource status information and data quality information, it should dynamically generate cross-border resource correlation scores and correlation levels, thereby improving the accuracy, dynamism, and traceability of identifying cross-border resource correlations of highly migratory fish species. This would provide reliable technical support for fisheries resource monitoring, distant-water fishing supervision, aquatic product traceability verification, IUU fishing risk identification, and cross-border resource protection. Summary of the Invention

[0012] The purpose of this invention is to provide a method and system for dynamically calculating the cross-border resource correlation of highly migratory fish, in order to solve the technical problems in the prior art such as the lack of unified standards for multi-source fishery data, the difficulty in accurately quantifying the cross-border resource correlation of fish, the difficulty in effectively identifying the relationship between fishing behavior and cross-border resource impact, and the difficulty in dynamically evaluating the correlation level.

[0013] To achieve the above objectives, the present invention provides the following technical solution.

[0014] In one possible implementation, a method for dynamically calculating the cross-border resource correlation of highly migratory fish species is provided, comprising the following steps:

[0015] Acquire basic data on fish species, migration routes, target sea area boundaries, fishing behavior, marine environment, and resource status for fish species migrating at the target altitude;

[0016] The basic data of fish species, migration path data, target sea area boundary data, fishing behavior data, marine environment data, and resource status data are uniformly processed by coordinate reference, time granularity, and fish species coding to obtain standardized multi-source data;

[0017] According to the preset spatial resolution and preset time window, the target analysis sea area is divided into multiple spatiotemporal grid units with spatial and temporal attributes, and the standardized multi-source data is mapped to the corresponding spatiotemporal grid units.

[0018] Based on the migration route data, marine environment data, and resource status data, the resource occurrence probability parameter of migratory fish at the target altitude in each of the spatiotemporal grid cells is calculated, wherein the resource occurrence probability parameter is determined according to the migration distance attenuation value, the migration stage correction coefficient, and the environmental suitability correction coefficient;

[0019] Based on the target sea area boundary data, the spatial overlap parameters between each of the spatiotemporal grid cells and at least two cross-border associated sea areas are calculated, wherein the spatial overlap parameters are determined according to the overlap area ratio and cross-border connectivity parameters.

[0020] Based on the fishing behavior data, suspected fishing operation segments are identified, the suspected fishing operation segments are matched with the spatiotemporal grid cells, and the impact parameters of the target fishing entity or target fishing area on the target highly migratory fish are calculated based on the matching results.

[0021] Based on the resource status data and data quality information, determine the resource status correction parameters and the comprehensive confidence level parameters;

[0022] Based on the migration stage of the target height migratory fish, changes in resource status, and comprehensive confidence parameters, dynamic weights are determined. The resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters are then fused according to the dynamic weights to generate a cross-border resource correlation score for the target height migratory fish.

[0023] The correlation degree grading threshold is dynamically determined based on the resource status data and comprehensive confidence parameters. The cross-border resource correlation score is compared with the correlation degree grading threshold, and the cross-border resource correlation level of the target highly migratory fish is output.

[0024] In one possible implementation, the unified processing of the fish species basic data, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data in terms of coordinate reference, time granularity, and fish species coding includes:

[0025] Transform marine boundary data, vessel tracks, fishing locations, fish migration routes, and marine environmental data from different sources into the same geographic coordinate system;

[0026] The data on vessel trajectories, fishing locations, migration routes, and marine environments are sliced ​​according to a preset time window.

[0027] Map the same fish species identifier to fish species names, Latin names, commercial names, and codes from different data sources;

[0028] Data confidence weights are set for data records that are missing, conflicting, or abnormal, and the data mapped to the spatiotemporal grid cells are corrected based on the data confidence weights.

[0029] In one possible implementation, dividing the target analysis sea area into multiple spatiotemporal grid units with spatial and temporal attributes includes:

[0030] The target analysis area is divided into multiple spatial grids according to the preset spatial resolution;

[0031] Each spatial grid is expanded into a spatiotemporal grid unit according to a preset time window;

[0032] Configure each spatiotemporal grid cell with a grid number, time window, sea area affiliation attribute, migration stage attribute, initial value of resource occurrence probability, and data confidence label;

[0033] The sea area attribution includes at least one or more of the following: national jurisdiction waters, high seas, and cross-border migration routes.

[0034] In one possible implementation, the calculation of the resource occurrence probability parameter includes:

[0035] Based on the migration path data, determine the migration centerline, migration buffer zone, and migration direction of migratory fish at the target altitude within different time windows;

[0036] Calculate the distance between each of the aforementioned spatiotemporal grid cells and the migration centerline or migration buffer zone to obtain the migration distance attenuation value;

[0037] Based on the characteristics of breeding, foraging, overwintering, or migratory fish at different migration stages at the target altitude, a correction coefficient for the migration stage is determined.

[0038] Determine the environmental suitability correction factor based on at least one of the following: water temperature, salinity, water depth, ocean current, chlorophyll concentration, or habitat suitability data.

[0039] Based on the migration distance attenuation value, migration stage correction coefficient, and environmental suitability correction coefficient, the resource occurrence probability parameters in each of the spatiotemporal grid cells are obtained.

[0040] In one possible implementation, the resource occurrence probability parameter is determined as follows:

[0041] The migration distance attenuation value, migration stage correction coefficient, and environmental suitability correction coefficient are multiplied and fused or weighted and fused to obtain the initial resource occurrence probability;

[0042] The initial resource occurrence probability is corrected based on the resource abundance index, resource decline index, or historical resource distribution records to obtain the resource occurrence probability parameter.

[0043] In one possible implementation, the calculation of the spatial overlap parameter includes:

[0044] Determine multiple spatiotemporal grid cells that migratory fish at the target altitude will pass through within a preset time window;

[0045] Calculate the percentage of overlapping area between the multiple spatiotemporal grid cells and different national jurisdictional waters, high seas areas, or cross-border migration routes;

[0046] Calculate cross-border connectivity parameters based on the connectivity relationships between adjacent spatiotemporal grid cells in different cross-border associated sea areas;

[0047] The spatial overlap parameters are generated based on the overlap area ratio and cross-border connectivity parameters.

[0048] In one possible implementation, the calculation of the fishing impact parameters includes:

[0049] Obtain at least one of the following information about the target fishing entity: vessel trajectory, operation time, operation speed, changes in course, duration of stay, type of fishing gear, fishing log, catch amount, or transshipment record;

[0050] Suspected fishing operation segments were identified based on changes in speed, course, and duration of stay in the vessel's trajectory.

[0051] The suspected fishing operation segments are matched with the spatiotemporal grid cells within the corresponding time window to obtain a fishing operation grid set;

[0052] The fishing impact parameters are calculated based on the fishing operation intensity, catch percentage, temporal proximity, spatial proximity, and confidence level of abnormal operations in the fishing operation grid set.

[0053] In one possible implementation, the cross-border resource correlation score is determined as follows:

[0054] The resource occurrence probability parameter, spatial overlap parameter, fishing impact parameter, and resource status correction parameter are used as positive correlation parameters;

[0055] The comprehensive confidence parameter is used as the scoring correction parameter;

[0056] The dynamic weights of each positive correlation parameter are updated based on the migration stage of the target migratory fish, changes in resource status, and overall confidence parameters.

[0057] The positive correlation parameters are weighted and summed, and the weighted summation result is corrected using the comprehensive confidence parameter to obtain the cross-border resource correlation score.

[0058] In one possible implementation, the step of dynamically determining the correlation degree classification threshold based on the resource status data and the comprehensive confidence parameter includes:

[0059] When the target highly migratory fish species are in their breeding season, resource decline period, or high fishing pressure period, increase the dynamic weights corresponding to the resource occurrence probability parameter and the fishing impact parameter, and lower the scoring threshold corresponding to the high correlation level.

[0060] When the target high-altitude migratory fish are in a non-critical migration phase, increase the dynamic weight corresponding to the spatial overlap parameter;

[0061] When the overall confidence level parameter is lower than the preset confidence threshold, the data review weight is increased and the correlation level to be reviewed is output.

[0062] The cross-border resource association level includes at least three or four of the following: high association level, medium association level, low association level, and level pending review.

[0063] In one possible implementation, a dynamic measurement system for the cross-border resource correlation of highly migratory fish species is provided, comprising:

[0064] The data acquisition module is used to acquire basic data on target migratory fish species, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data.

[0065] The data standardization module is used to uniformly process the fish species basic data, migration path data, target sea area boundary data, fishing behavior data, marine environment data, and resource status data in terms of coordinate reference, time granularity, and fish species coding to obtain standardized multi-source data.

[0066] The spatiotemporal grid construction module is used to divide the target analysis sea area into multiple spatiotemporal grid units with spatial and temporal attributes according to a preset spatial resolution and a preset time window, and to map the standardized multi-source data to the corresponding spatiotemporal grid units.

[0067] The resource occurrence probability calculation module is used to calculate the resource occurrence probability parameters of migratory fish at the target altitude in each of the spatiotemporal grid cells based on the migration path data, marine environment data, and resource status data.

[0068] The spatial overlap calculation module is used to calculate the spatial overlap parameters between each of the spatiotemporal grid cells and at least two cross-border associated sea areas based on the target sea area boundary data.

[0069] The fishing impact calculation module is used to identify suspected fishing operation segments based on the fishing behavior data, match the suspected fishing operation segments with the spatiotemporal grid cells, and calculate the fishing impact parameters of the target fishing entity or target fishing area on the target highly migratory fish based on the matching results.

[0070] The status and confidence calculation module is used to determine the resource status correction parameters and the comprehensive confidence parameters based on the resource status data and data quality information.

[0071] The correlation scoring module is used to determine dynamic weights based on the migration stage of the target height migratory fish, changes in resource status, and comprehensive confidence parameters. The module then integrates the resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters according to the dynamic weights to generate a cross-border resource correlation score for the target height migratory fish.

[0072] The dynamic grading output module is used to dynamically determine the correlation grading threshold based on the resource status data and comprehensive confidence parameters, compare the cross-border resource correlation score with the correlation grading threshold, and output the cross-border resource correlation level of the target highly migratory fish.

[0073] Based on the above technical solutions, the present invention provides a dynamic calculation method and system for the cross-border resource correlation of highly migratory fish species. By acquiring basic data on fish species, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data, the system performs unified and standardized processing on multi-source data from different sources, formats, and time scales. It also constructs spatiotemporal grid units with temporal and spatial attributes to achieve correlation analysis of fish migration activities, sea area spatial relationships, fishing behavior, and resource status information under a unified spatiotemporal benchmark. This solves the problems in the prior art where it is difficult to unify and integrate multi-source fishery data and accurately quantify the cross-border resource correlation of fish.

[0074] Furthermore, this invention incorporates fish migration routes, migration stage characteristics, and marine environmental suitability information into the resource occurrence probability calculation process. Based on the migration distance attenuation value, migration stage correction coefficient, and environmental suitability correction coefficient, the resource occurrence probability parameters in each spatiotemporal grid cell are determined. This enables the resource distribution results to not only reflect the historical activity patterns of fish but also to combine the dynamic changes under different growth stages and environmental conditions, thereby improving the ability to characterize the actual resource distribution status of highly migratory fish.

[0075] Furthermore, this invention constructs a cross-border sea area spatial overlap analysis mechanism, utilizing the overlap area ratio and cross-border connectivity parameters between spatiotemporal grid units and national jurisdictional waters, high seas areas, and cross-border migration channel areas to generate spatial overlap parameters, thereby achieving a quantitative description of the spatial correlation of resources during the cross-border migration of fish.

[0076] Furthermore, this invention identifies suspected fishing operation segments and matches them with spatiotemporal grid cells. Based on operation intensity, catch proportion, temporal proximity, spatial proximity, and confidence level of abnormal operations, it calculates fishing impact parameters, thereby achieving a quantitative analysis of the correlation between fishing activities and target fish species resources and improving the accuracy of assessing the actual resource impact of fishing behavior.

[0077] Furthermore, this invention combines resource status data and data quality information to determine resource status correction parameters and comprehensive confidence parameters, and dynamically adjusts the weights of each evaluation index according to the migration stage of the target migratory fish, changes in resource status, and comprehensive confidence parameters. This enables dynamic fusion calculation of resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters, thereby improving the adaptability of cross-border resource correlation scoring results to different fish species, different seasons, and different resource statuses.

[0078] Furthermore, this invention dynamically determines the correlation degree grading threshold based on resource status data and comprehensive confidence parameters. When the target highly migratory fish is in its breeding season, resource decline period, or high fishing pressure period, it can automatically increase the sensitivity to resource protection. When the data integrity is insufficient or the data confidence is low, it can automatically output the correlation level to be reviewed, thereby reducing the risk of misjudgment caused by missing data, data anomalies, or data uncertainty, and improving the reliability and interpretability of the correlation level output results.

[0079] Compared with existing technologies, this invention achieves dynamic correlation analysis between fish resource distribution, cross-border migration paths, and the impact of fishing activities by constructing a collaborative calculation mechanism of "spatiotemporal grid mapping—resource occurrence probability calculation—cross-border spatial overlap analysis—fishing impact identification—dynamic weight fusion—dynamic threshold grading." Compared to existing technologies that use fixed sea areas, static resource distribution models, or single fishing locations to determine resource correlation, this invention can adjust the correlation calculation results and correlation level output results in real time according to changes in fish migration stages, environmental suitability, resource status, and data integrity. This improves the accuracy, dynamic adaptability, and risk identification capabilities of cross-border resource correlation identification, providing reliable technical support for fisheries resource monitoring, distant-water fishing supervision, aquatic product traceability verification, imported aquatic product risk assessment, illegal fishing risk identification, cross-border resource protection, and marine ecological resource management. Attached Figure Description

[0080] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0081] Figure 1 A flowchart illustrating a method for dynamically calculating the cross-border resource correlation of highly migratory fish species, provided in an embodiment of the present invention;

[0082] Figure 2 This is a structural block diagram of a dynamic measurement system for the cross-border resource correlation of highly migratory fish species, provided in an embodiment of the present invention.

[0083] The steps are as follows: 101. Data acquisition; 102. Data standardization; 103. Spatiotemporal grid construction; 104. Resource occurrence probability calculation; 105. Spatial overlap parameter calculation; 106. Fishing impact parameter calculation; 107. Resource status and comprehensive confidence calculation; 108. Cross-border resource correlation score generation; 109. Dynamic hierarchical output of correlation.

[0084] 201. Data Acquisition Module; 202. Data Standardization Module; 203. Spatiotemporal Grid Construction Module; 204. Resource Occurrence Probability Calculation Module; 205. Spatial Overlap Calculation Module; 206. Fishing Impact Calculation Module; 207. State and Confidence Calculation Module; 208. Correlation Scoring Module; 209. Dynamic Hierarchical Output Module. Detailed Implementation

[0085] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0086] I. Overall Technical Solution of the Invention

[0087] The following detailed description, in conjunction with the accompanying drawings, of the method and system for dynamically calculating the cross-border resource correlation of highly migratory fish species provided by this invention.

[0088] It should be noted that the following embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Various modifications, substitutions, and improvements made by those skilled in the art without departing from the technical concept of the present invention should fall within the scope of protection of the present invention.

[0089] like Figure 1 As shown in this embodiment, a method for dynamically calculating the cross-border resource correlation of highly migratory fish species is provided. This method targets highly migratory fish resources such as tuna, skipjack tuna, swordfish, mackerel, herring, and eel. It achieves dynamic quantitative analysis of the cross-border resource correlation of target fish species by uniformly processing fish migration routes, marine spatial boundaries, fishing activity information, marine environmental information, and resource status information.

[0090] In this embodiment, the method includes a data acquisition step 101, a data standardization processing step 102, a spatiotemporal grid construction step 103, a resource occurrence probability parameter calculation step 104, a spatial overlap parameter calculation step 105, a fishing impact parameter calculation step 106, a resource status and comprehensive confidence calculation step 107, a cross-border resource correlation score calculation step 108, and a correlation level dynamic output step 109.

[0091] In step 101, the system acquires multi-source data related to the target migratory fish species. This multi-source data includes basic fish species data, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data. This data can originate from satellite remote sensing systems, AIS (Automatic Identification System), VMS (Vessel Monitoring System), electronic tagging systems, marine environmental monitoring platforms, catch statistics databases, and resource assessment data published by regional fisheries management organizations.

[0092] In step 102, the system performs unified standardization processing on data from different sources. Since different data sources often have inconsistent coordinate bases, different time granularities, different fish species encoding methods, and significant differences in data quality, it is necessary to perform unified transformation and preprocessing on all data to establish a unified foundation for data analysis.

[0093] In step 103, the system constructs a spatiotemporal grid system based on a preset spatial resolution and a preset time window. By dividing the target analysis sea area into multiple spatiotemporal grid units with spatial and temporal attributes, data from different sources can be mapped to a unified spatiotemporal coordinate system, thereby enabling correlation analysis between fish migration activities, fishing activities, and changes in resource status.

[0094] In step 104, the system calculates the resource occurrence probability parameters of the target migratory fish species in each spatiotemporal grid cell based on fish migration route information, marine environmental information, and resource status information. These resource occurrence probability parameters reflect the likelihood of the target fish species' resources existing in a specific time window and spatial region, providing a basis for subsequent correlation calculations.

[0095] In step 105, the system analyzes the spatial relationships between the spatiotemporal grid cells traversed by the target fish species and the waters under the jurisdiction of different countries, the high seas, and cross-border migration channels based on the target sea area boundary information, and calculates spatial overlap parameters. These spatial overlap parameters characterize the degree of spatial association between the target fish species resources and different cross-border associated sea areas.

[0096] In step 106, the system identifies suspected fishing activities based on vessel trajectory information, operational behavior information, and catch record information. It then matches the identified fishing activities with corresponding spatiotemporal grid cells to calculate fishing impact parameters. These parameters reflect the actual impact of the target fishing entity or target fishing area on the target migratory fish resources.

[0097] In step 107, the system determines resource status correction parameters based on resource abundance index, resource decline index, closed fishing season status, breeding season status, and historical fishing pressure information; simultaneously, it determines comprehensive confidence parameters based on data source reliability, data integrity, time freshness, and spatial accuracy. These resource status correction parameters and comprehensive confidence parameters are used to dynamically correct subsequent correlation calculation results.

[0098] In step 108, the system performs a fusion calculation on resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters to generate a cross-border resource correlation score. Unlike traditional static evaluation models, this invention dynamically adjusts the weights of each evaluation parameter based on the target fish species' migration stage, changes in resource status, and changes in data quality, thereby improving the accuracy and adaptability of the correlation calculation results.

[0099] In step 109, the system dynamically determines the correlation level classification threshold based on resource status information and comprehensive confidence information, compares the cross-border resource correlation score with the corresponding threshold, and outputs correlation results such as high correlation level, medium correlation level, low correlation level, or level pending review. Through the dynamic threshold adjustment mechanism, the correlation level determination results can better reflect the dynamic changes in fish resources and actual regulatory needs.

[0100] like Figure 2 As shown in the figure, this embodiment also provides a dynamic measurement system for the cross-border resource correlation of highly migratory fish. The system includes a data acquisition module 201, a data standardization module 202, a spatiotemporal grid construction module 203, a resource occurrence probability calculation module 204, a spatial overlap calculation module 205, a fishing impact calculation module 206, a state and confidence calculation module 207, a correlation scoring module 208, and a dynamic hierarchical output module 209.

[0101] The modules interact with each other to form a complete dynamic calculation link for cross-border resource correlation. Specifically, the data acquisition module 201 acquires raw data; the data standardization module 202 completes unified processing; the spatiotemporal grid construction module 203 establishes a unified analysis framework; the resource occurrence probability calculation module 204, the spatial overlap calculation module 205, and the fishing impact calculation module 206 calculate evaluation parameters of different dimensions respectively; the status and confidence calculation module 207 completes resource status and data quality assessment; the correlation scoring module 208 generates a cross-border resource correlation score; and the dynamic hierarchical output module 209 outputs the final correlation level result.

[0102] Through the above technical solutions, this invention constructs a dynamic calculation system that integrates fish migration patterns, cross-border marine spatial relationships, fishing behavior characteristics, resource status changes, and data quality evaluation. It realizes dynamic quantitative analysis of the cross-border resource correlation of highly migratory fish, providing reliable technical support for fishery resource monitoring, deep-sea fishing supervision, aquatic product traceability and verification, illegal fishing risk identification, and cross-border resource protection.

[0103] II. Explanation of the Correlation Calculation Principle of this Invention

[0104] To facilitate understanding of the technical solution of the present invention, before providing a detailed explanation of steps 101 to 109, the overall principle of the dynamic calculation of cross-border resource correlation of the present invention will be explained first.

[0105] The core objective of this invention is to address the difficulty in accurately quantifying the cross-border resource correlations of highly migratory fish species. Because highly migratory fish typically traverse multiple national jurisdictional waters, high seas, and cross-border migration routes during their life cycle, their resource distribution is influenced by various factors, including fish migration patterns, marine environmental conditions, changes in resource status, and fishing activities. Therefore, relying solely on single indicators such as fishing location, sea area affiliation, or historical catch volume is insufficient to accurately reflect the true correlation between the target species and cross-border resources.

[0106] To this end, this invention first acquires basic fish species data, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data, and establishes a standardized data system through unified coordinate reference, unified time reference, and unified fish species identification. Subsequently, using a spatiotemporal grid as a unified analysis carrier, fish resource distribution information, sea area spatial information, fishing activity information, and resource status information are mapped to the same spatiotemporal framework for management and calculation.

[0107] Based on this, the present invention constructs an evaluation parameter system from the dimensions of resource distribution, spatial correlation, and fishing impact. Among them, the resource occurrence probability parameter is used to characterize the possibility of the target fish species existing in the corresponding spatiotemporal grid; the spatial overlap parameter is used to characterize the degree of spatial correlation between the target resource and different transboundary sea areas; and the fishing impact parameter is used to characterize the potential impact of the target fishing behavior on the target resource.

[0108] Furthermore, this invention constructs resource status correction parameters by combining information such as resource abundance, resource decline degree, breeding period status, and historical fishing pressure, and constructs comprehensive confidence parameters by combining information such as data source reliability, data integrity, time freshness, and spatial accuracy. Thus, the correlation measurement results can not only reflect the resource change pattern, but also reflect the credibility of the evaluation results.

[0109] Ultimately, this invention employs a dynamic weighted fusion mechanism to comprehensively calculate resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, and resource status correction parameters. It then uses a comprehensive confidence level parameter to correct the evaluation results, obtaining a cross-border resource correlation score. Subsequently, the correlation level classification threshold is dynamically adjusted based on the target fish species' resource status and data quality, outputting the corresponding cross-border resource correlation level.

[0110] Through the above technical approach, this invention constructs a complete measurement system from multi-source data acquisition, spatiotemporal grid modeling, resource correlation analysis, dynamic scoring to grade output, realizing dynamic quantitative analysis of cross-border resource correlation of highly migratory fish.

[0111] III. Step 101: Data Acquisition

[0112] In this embodiment, as Figure 1 As shown, step 101 is used to obtain multi-source basic data required for dynamic calculation of the cross-border resource correlation of the target highly migratory fish species. Since highly migratory fish species typically cross multiple national jurisdictional waters, high seas areas, and cross-border migration channels during their life cycle, their resource distribution is not only affected by the fish's own migration patterns, but also by a combination of factors such as changes in the marine environment, changes in resource status, and fishing activities. Therefore, to ensure the accuracy of the subsequent correlation calculation results, this embodiment uses a multi-source data fusion method to construct a resource correlation analysis dataset for the target fish species.

[0113] Specifically, the acquired data includes basic fish species data, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data. Basic fish species data describes the fundamental biological attributes and resource characteristics of the target fish species, and may include species name, species code, Latin name, growth stage, reproductive cycle, feeding cycle, overwintering cycle, and historical resource distribution information. Migration route data reflects the migration trajectory and activity range of the target fish species within different time windows, and may be derived from electronic tag monitoring systems, satellite tagging systems, release and recapture data, and historical resource survey data. Target sea area boundary data characterizes national exclusive economic zones, high seas areas, marine protected areas, fishing ban areas, and cross-border migration routes. Spatial boundary information; fishing behavior data reflects the activity status of the target fishing entity within the corresponding time window, which may include vessel trajectory, operation time, heading information, speed information, dwell time, fishing gear type, fishing log, and catch record; marine environmental data reflects the characteristics of the fish's living environment, which may include indicators such as sea surface temperature, salinity, ocean current, water depth, chlorophyll concentration, and dissolved oxygen concentration; resource status data reflects the resource change trend of the target fish species, which may include information such as resource abundance index, resource decline index, historical fishing pressure index, and resource recovery status.

[0114] In one implementation, the aforementioned data can originate from data sources such as the AIS Automatic Identification System, VMS Vessel Monitoring System, marine environmental monitoring platform, regional fisheries management organization database, fisheries resource survey database, and government regulatory database. Since data from different sources typically differ in data format, coordinate reference, time granularity, and fish species coding methods, it is necessary to standardize and unify the data in subsequent step 102 to establish a unified data analysis foundation.

[0115] Furthermore, to improve the reliability of subsequent resource correlation calculation results, this embodiment also simultaneously acquires data quality information corresponding to each data source. This data quality information may include evaluation indicators such as data source reliability, data integrity, time freshness, spatial accuracy, and the proportion of abnormal data, and is used for subsequent calculation of comprehensive confidence parameters and correction of correlation scores.

[0116] Through the above methods, this embodiment can construct a comprehensive data system covering multiple dimensions such as fish resource distribution, cross-border spatial relationships, impact of fishing activities, and changes in resource status, providing a unified data foundation for subsequent spatiotemporal grid construction, resource occurrence probability calculation, spatial overlap analysis, fishing impact assessment, and cross-border resource correlation measurement.

[0117] IV. Step 102: Data Standardization Processing

[0118] After acquiring basic data on fish species, migration routes, target sea area boundaries, fishing behavior, marine environment, and resource status, since the above data come from different data platforms and business systems, their data formats, coordinate references, time scales, and fish species identification methods usually vary greatly. Therefore, it is necessary to first standardize all types of data to establish a unified data analysis foundation.

[0119] In this embodiment, data standardization mainly includes processes such as coordinate benchmark unification, time granularity unification, fish species identification unification, and data quality correction. Through these processes, data from different sources can be mapped to the same spatiotemporal analysis framework, thereby providing reliable data support for subsequent spatiotemporal grid construction and correlation calculation.

[0120] First, all spatial data undergo coordinate benchmark unification processing. Since ship trajectory data, sea area boundary data, fish migration path data, and marine environmental monitoring data may be stored in different coordinate systems, such as the WGS84 coordinate system, the CGCS2000 coordinate system, or other regional geographic coordinate systems, it is necessary to uniformly convert the above data to the preset target coordinate system.

[0121] In this embodiment, the WGS84 geographic coordinate system is used as a unified spatial reference datum. For any data record, its spatial location is represented by longitude (Lon) and latitude (Lat), which are then transformed to form a unified spatial coordinate expression:

[0122] S=(Lon,Lat)

[0123] Where S represents the standardized spatial coordinates.

[0124] After unifying the coordinates, the time data was further standardized.

[0125] Since fish electronic tag data is usually recorded at minute-level or hour-level intervals, AIS trajectory data may be updated at second-level intervals, while resource survey data, catch statistics data and resource status data are usually released in units of weeks, months or quarters, so there are obvious differences in time scales between different data.

[0126] To improve the consistency of subsequent spatio-temporal correlation analysis, this embodiment adopts a unified time window mechanism to perform time standardization processing on all data. The preset time window is set as ΔT, and in practical applications, ΔT can be set to 1 day, 3 days, 7 days, 15 days or 30 days according to business requirements.

[0127] For any data record, its timestamp is marked as T. When the data time falls within the same time window range, it is mapped to the corresponding time window identifier:

[0128] TW_k=[T_k,T_(k+1)]

[0129] where TW_k represents the k-th time window.

[0130] Through the above method, data with different time scales can be uniformly mapped to the same time analysis unit, thereby realizing subsequent cross-data source correlation analysis.

[0131] Further, since different data sources may adopt different naming methods for the same fish resource, it is necessary to perform unified processing on fish species identifiers.

[0132] For example, for bluefin tuna, different databases may use the Chinese name "蓝鳍金枪鱼", the English name "Bluefin Tuna", the Latin scientific name "Thunnus thynnus" or the corresponding commodity code to represent it respectively. If data fusion is performed directly, it is easy to cause the same fish species to be identified as multiple different objects.

[0133] Therefore, this embodiment establishes a unified fish species mapping relation table, which maps fish species names, Latin scientific names, commodity names, resource codes and trade codes to the unified fish species identifier Fish_ID. For any data record, Fish_ID is used as the unique fish species identifier in the subsequent calculation process, so as to ensure accurate correlation between data from different sources.

[0134] In addition, since multi-source data may have problems such as missing data, abnormal values, conflicting records and duplicate records during collection, transmission and storage, this embodiment further introduces a data quality correction mechanism.

[0135] For each data record, the system calculates the corresponding data quality evaluation result based on factors such as data source reliability, field completeness, time validity, and spatial accuracy, and generates data confidence weights. Data source reliability is used to evaluate the credibility of the data source; field completeness is used to evaluate the absence of key fields; time validity is used to evaluate the difference between the data update time and the target analysis time; and spatial accuracy is used to evaluate the level of positioning error.

[0136] In one implementation, the above evaluation results can be normalized to form data confidence weights, which can then be used as correction factors in subsequent resource occurrence probability calculations, fishing impact analysis, and correlation score calculations.

[0137] After the above-mentioned processes of unifying coordinate references, time granularity, fish species identification, and data quality correction, the system generates a standardized multi-source dataset. All data in the standardized multi-source dataset have a unified spatial reference, a unified time reference, and a unified fish species identification, and are accompanied by corresponding data quality evaluation information.

[0138] The above processing not only solves the problem of difficulty in directly linking data from different sources, but also improves the accuracy and stability of subsequent resource occurrence probability calculations, spatial overlap analysis, fishing impact assessments, and cross-border resource correlation measurement results, providing a unified data input basis for the spatiotemporal grid construction in step 103.

[0139] V. Step 103: Spatiotemporal grid construction

[0140] After completing the data standardization process in step 102, although the data from different sources already have a unified coordinate reference, a unified time reference, and a unified fish species identifier, since fish migration activities, fishing behavior, and changes in the marine environment all have both spatial and temporal attributes, it is still necessary to establish a unified spatiotemporal analysis framework to achieve correlation calculations between multi-source data.

[0141] Therefore, this embodiment adopts a spatiotemporal gridded modeling approach to discretize the target sea area, converting continuous ocean space and continuous time processes into computable spatiotemporal grid units, thereby providing a unified data carrier for subsequent resource occurrence probability calculation, spatial overlap analysis, fishing impact assessment, and correlation score calculation.

[0142] Specifically, a spatial grid system is first established based on the target analysis area.

[0143] Let the sea area corresponding to the target analysis region be A_total, and divide the target analysis region into grids according to the preset spatial resolution ΔS.

[0144] Where ΔS represents the side length of the spatial grid, and the corresponding unit can be kilometers (km) or latitude and longitude resolution (°).

[0145] In one implementation, the spatial resolution can be set to 5km, 10km, 20km, 50km, or other spatial scales suitable for the migratory characteristics of the target fish species.

[0146] After partitioning, the target analysis region is divided into multiple spatial grid cells:

[0147] G_s(i,j)

[0148] in:

[0149] i represents the grid number in the longitude direction;

[0150] j represents the grid number in the latitudinal direction;

[0151] G_s(i,j) represents the spatial grid in the i-th row and j-th column.

[0152] Subsequently, the time dimension is discretized according to the preset time window ΔT.

[0153] ΔT can be set to 1 day, 3 days, 7 days, 15 days, or 30 days.

[0154] Let the target analysis period be: [T_start, T_end]

[0155] The target analysis period can then be divided into multiple time windows: TW(k)

[0156] Where k represents the time window number.

[0157] Furthermore, spatial grids are combined with temporal windows to construct three-dimensional spatiotemporal grid units:

[0158] G(i,j,k)

[0159] Where: i represents the longitude grid index; j represents the latitude grid index; k represents the time window index.

[0160] Therefore, G(i,j,k) represents the spatiotemporal grid cell corresponding to the spatial location within the k-th time window.

[0161] In this embodiment, each spatiotemporal grid cell serves as the smallest computational unit for subsequent analysis and is used to carry fish resource information, marine area attribute information, fishing activity information, and marine environmental information within the corresponding time window.

[0162] Specifically, each spatiotemporal grid cell includes at least the following attributes:

[0163] Grid ID; Time Window; Region Type;

[0164] Migration stage attribute; resource occurrence probability parameter P_resource; spatial overlap parameter P_overlap; fishing impact parameter P_fishing; resource state parameter R_state; overall confidence parameter C_conf.

[0165] Among them, Region_Type is used to identify the type of sea area to which the corresponding grid belongs, which can be national jurisdiction waters, high seas, marine protected areas, fishing ban areas, or cross-border migration channels.

[0166] Migration_Stage is used to identify the migration stage of the target fish species within the current time window, including the breeding season, feeding season, overwintering season, and return migration season.

[0167] To ensure that all subsequent evaluation indicators can be calculated in a unified manner, this embodiment further adopts a normalization mechanism.

[0168] Specifically, for data with different physical dimensions such as distance, area, catch, resource abundance and environmental indicators involved in subsequent calculations, they are first converted into corresponding evaluation indicators, and then normalized to the interval [0,1] to obtain dimensionless parameters.

[0169] Therefore, the resource occurrence probability parameter P_resource, spatial overlap parameter P_overlap, fishing impact parameter P_fishing, resource state parameter R_state, and comprehensive confidence parameter C_conf calculated in subsequent steps 104 to 109 are all dimensionless parameters, and their value ranges satisfy the following:

[0170] 0≤P_resource≤1; 0≤P_overlap≤1; 0≤P_fishing≤1; 0≤R_state≤1; 0≤C_conf≤1.

[0171] Through the above spatiotemporal grid construction process, this embodiment maps fish migration information, marine environment information, sea area boundary information, and fishing behavior information, which were originally scattered in different data sources, into the same spatiotemporal grid unit, thereby establishing a unified spatiotemporal analysis framework for cross-border resource correlation measurement, providing basic support for subsequent calculation of resource occurrence probability parameters.

[0172] VI. Step 104: Calculation of resource occurrence probability parameters

[0173] After completing the spatiotemporal grid construction in step 103, data from different sources have been uniformly mapped to the corresponding spatiotemporal grid cells G(i,j,k). To reflect the resource distribution status of migratory fish at the target altitude in different time windows and different spatial locations, this embodiment further calculates the resource occurrence probability parameter P_resource in each spatiotemporal grid cell.

[0174] The resource occurrence probability parameter is used to characterize the possibility of the target fish species having resources in the corresponding spatiotemporal grid cell. Its value range is: 0≤P_resource≤1, where the closer P_resource is to 1, the higher the probability of the target fish species having resources in the corresponding spatiotemporal grid cell; the closer P_resource is to 0, the lower the probability of the resource occurrence.

[0175] Since highly migratory fish have obvious migration patterns, their spatial distribution is not only affected by historical migration routes, but also by the migration stage and changes in marine environmental conditions. Therefore, this embodiment uses migration route factors, migration stage factors, and environmental suitability factors to jointly construct a resource occurrence probability model.

[0176] First, the migration centerline of the target fish species within the current time window is determined based on the migration path data obtained in step 101.

[0177] Specifically, for the set of fish trajectory points within the same time window TW(k), the main activity area of ​​the target fish species is determined by trajectory clustering analysis or kernel density estimation method, and the corresponding migration center line L_m is extracted.

[0178] Subsequently, the distance between the center of the spatiotemporal grid cell and the migration centerline is calculated.

[0179] Let d(i,j,k) represent the shortest distance between the center point of the spatiotemporal grid cell G(i,j,k) and the migration centerline L_m, in km.

[0180] To eliminate the influence of distance dimensions, this embodiment uses a distance attenuation function to convert the distance into a dimensionless distance attenuation factor F_d:

[0181] F_d=exp(-d(i,j,k) / D_ref)

[0182] Where: D_ref represents the reference migration distance; the unit is km; its value can be determined based on the historical migration characteristics of the target fish species.

[0183] For example: tuna takes 500km; skipjack tuna takes 300km; eel takes 100km.

[0184] Since d and D_ref have the same dimension, the exponent term remains dimensionless.

[0185] It can be guaranteed that: 0 < F_d ≤ 1, and the closer the distance is, the larger F_d is.

[0186] Further, since different migration stages correspond to different resource aggregation degrees, this embodiment introduces a migration stage correction factor F_m.

[0187] In one embodiment, a corresponding correction coefficient is assigned according to the current migration stage of the target fish species:

[0188] Breeding period: F_m=1.00, feeding period: F_m=0.85, migration period: F_m=0.75, overwintering period: F_m=0.65.

[0189] It should be noted that the above values are only examples, and can be adjusted according to the ecological characteristics of the target fish species in practical applications.

[0190] Further, in order to reflect the influence of the marine environment on fish distribution, this embodiment establishes an environmental suitability index F_e.

[0191] Specifically, the corresponding environmental suitability score is calculated based on environmental indicators such as water temperature, salinity, ocean current velocity, water depth, and chlorophyll concentration.

[0192] Suppose:

[0193] T_n represents the normalized water temperature suitability; S_n represents the normalized salinity suitability; C_n represents the normalized ocean current suitability; D_n represents the normalized water depth suitability; Ch_n represents the normalized chlorophyll suitability.

[0194] All the above indicators are mapped to the interval using the min-max normalization method: [0,1]

[0195] Then the environmental suitability index F_e can be expressed as:

[0196] F_e=w_tT_n+w_sS_n+w_cC_n+w_dD_n+w_chCh_n

[0197] wherein w_t+w_s+w_c+w_d+w_ch=1, and each weight is determined according to the ecological preference of the target fish species.

[0198] Thus it is obtained that: 0≤F_e≤1.

[0199] After obtaining the distance attenuation factor F_d, the migration stage correction factor F_m and the environmental suitability index F_e, this embodiment further constructs a resource occurrence probability parameter.

[0200] Specifically, the resource occurrence probability parameter P_resource is defined as:

[0201] P_resource=(F_d×F_m×F_e) / P_max

[0202] Where P_max represents the theoretical maximum value correction coefficient; it is used to ensure that 0≤P_resource≤1.

[0203] In one implementation, P_max can be set to max(F_d×F_m×F_e), or a normalization process can be used to obtain the final resource occurrence probability parameter.

[0204] Through the above calculation process, this embodiment can comprehensively consider the characteristics of fish migration paths, the characteristics of different migration stages, and the suitability characteristics of the marine environment, thereby obtaining resource occurrence probability parameters that are more consistent with the actual resource distribution patterns.

[0205] Compared with traditional methods that estimate resources based on fixed distribution areas or historical catch statistics, this embodiment can dynamically reflect the changes in resource distribution of target fish species in different time windows and different spatial regions, thereby improving the accuracy and reliability of subsequent cross-border resource correlation calculation results.

[0206] After the above processing, each spatiotemporal grid cell G(i,j,k) corresponds to a resource occurrence probability parameter P_resource. This parameter will serve as an important input parameter for the calculation of spatial overlap parameters in step 105 and the calculation of cross-border resource correlation score in step 108.

[0207] VII. Step 105: Calculation of Spatial Overlap Parameters

[0208] After obtaining the resource occurrence probability parameters in step 104, in order to further reflect the spatial correlation between target height migratory fish resources and different cross-border associated sea areas, this embodiment performs spatial overlap analysis on each spatiotemporal grid cell and calculates the spatial overlap parameter P_overlap.

[0209] The spatial overlap parameter is used to characterize the degree of spatial correlation between target fish species resources in different national jurisdictional waters, high seas areas, and transboundary migration routes. Its value range is: 0 ≤ P_overlap ≤ 1.

[0210] The closer P_overlap is to 1, the higher the spatial correlation between the target fish species and multiple cross-border related sea areas; the closer P_overlap is to 0, the lower the spatial correlation between the target fish species and cross-border related sea areas.

[0211] Since highly migratory fish typically traverse multiple national jurisdictional waters and high seas areas during their life cycle, simply using the area of ​​overlapping sea areas as an evaluation criterion is insufficient to fully reflect the cross-border flow characteristics of fish resources. Therefore, this embodiment divides spatial overlap analysis into two parts: overlap area analysis and cross-border connectivity analysis, and integrates these two parts to form the final spatial overlap parameters.

[0212] Specifically, for any spatiotemporal grid cell G(i,j,k), the set of cross-border sea areas associated with the target fish species resources within its corresponding time window is first identified.

[0213] Let R = {R1, R2, ..., R} n} represents the set of cross-border sea areas involved in the target fish species within the current time window.

[0214] in:

[0215] R1, R2...R n These respectively represent waters under the jurisdiction of different countries, high seas areas, or cross-border migration routes.

[0216] The overlap ratio between the spatiotemporal grid cells and each cross-border sea area was then calculated.

[0217] Let A_grid represent the spatial grid area corresponding to the target spatiotemporal grid cell; the unit is km. 2 A_overlap,m represents the overlap area between this spatiotemporal grid cell and the m-th cross-border sea area; the unit is also km. 2 .

[0218] The overlapping area ratio of the corresponding cross-border sea areas can then be expressed as:

[0219] O_m=A_overlap,m / A_grid

[0220] Since the numerator and denominator have the same area dimension, O_m is a dimensionless parameter.

[0221] And satisfy 0≤O_m≤1;

[0222] When the spatiotemporal grid cells are entirely located within the corresponding transboundary sea area: O_m=1

[0223] When there is no spatial overlap: O_m=0

[0224] For cases involving multiple cross-border sea areas, this embodiment further calculates the overall overlap area ratio:

[0225] O_total=∑(P_resource,m×O_m)

[0226] Wherein, P_resource,m represents the probability parameter of the occurrence of the target fish species in the corresponding cross-border sea area; O_total represents the comprehensive overlap area index.

[0227] To ensure consistency of subsequent parameters, O_total is normalized to obtain F_overlap.

[0228] Where 0 ≤ F_overlap ≤ 1.

[0229] Furthermore, considering only spatial overlap is still insufficient to reflect the characteristics of cross-border migration of fish.

[0230] For example, even if two sea areas are in spatial contact, if the target fish species do not have a migration path connection within the corresponding time window, the actual degree of cross-border resource association is still low.

[0231] Therefore, this embodiment further introduces cross-border connectivity analysis.

[0232] Specifically, based on the migration centerline obtained in step 104 and the migration trajectory of the target fish species within a continuous time window, the spatiotemporal grid sequence traversed during the resource migration process is identified: G1→G2→…→G n

[0233] When adjacent spatiotemporal grids belong to different cross-border sea areas, it is considered that there is a cross-border migration connection.

[0234] Let N_cross represent the number of cross-border migration connections that occur within the target time window;

[0235] N_total represents the total number of migration connections within the target time window.

[0236] The cross-border connectivity parameter is defined as: F_connect = N_cross / N_total

[0237] Where 0≤F_connect≤1

[0238] Since both the numerator and denominator represent the number of connections, F_connect is also a dimensionless parameter.

[0239] The larger the F_connect value, the more frequent the migration activities of the target fish species between different cross-border sea areas, and the more obvious its cross-border resource flow characteristics.

[0240] After obtaining the comprehensive overlap area index F_overlap and the cross-border connectivity parameter F_connect, this embodiment constructs the spatial overlap parameter:

[0241] P_overlap=w_oF_overlap+w_cF_connect

[0242] Where: w_o represents the overlap area weight; w_c represents the connectivity weight; and the following condition is satisfied: w_o + w_c = 1

[0243] In one implementation, w_o = 0.6 and w_c = 0.4 can be chosen to enhance the ability to represent the actual sea area ownership relationship.

[0244] Of course, in other implementations, the weights can be adjusted according to the ecological characteristics of the target fish species or business needs.

[0245] Through the above calculation process, this embodiment can not only reflect the spatial overlap between the target fish species resources and different cross-border sea areas, but also reflect the migration and connectivity of resources between multiple cross-border sea areas, thereby establishing a spatial association model that is more in line with the ecological laws of highly migratory fish.

[0246] Compared with the traditional method that relies solely on the overlapping area of ​​sea boundaries, this embodiment introduces cross-border connectivity parameters, enabling spatial overlap analysis to simultaneously consider the spatial distribution and cross-border flow of resources, thereby improving the accuracy and interpretability of subsequent cross-border resource correlation calculation results.

[0247] After the above processing, each spatiotemporal grid cell G(i,j,k) corresponds to a spatial overlap parameter P_overlap, which will serve as an important input parameter for calculating the fishing impact parameter in step 106 and the cross-border resource correlation score in step 108.

[0248] 8. Step 106: Calculation of parameters affecting fishing

[0249] After obtaining the resource occurrence probability parameter P_resource in step 104 and the spatial overlap parameter P_overlap in step 105, in order to further evaluate the actual impact of the target fishing entity or target fishing area on the target highly migratory fish resources, this embodiment analyzes the fishing behavior data and calculates the fishing impact parameter P_fishing.

[0250] The fishing impact parameter is used to characterize the correlation strength between fishing activities and target fish species resources, and its value range is: 0≤P_fishing≤1

[0251] Wherein, the closer P_fishing is to 1, the greater the potential impact of the target fishing behavior on the target fish species resources; the closer P_fishing is to 0, the smaller the impact.

[0252] In actual marine fishing activities, judging the impact on fishing based solely on the vessel's current location is often insufficient to accurately reflect the actual situation. For example, different vessels in the same sea area may have different operating methods, different operating durations, and different catch compositions; furthermore, even if a vessel is near the distribution area of ​​the target fish species, whether it is actually conducting fishing activities and the duration of those activities can vary. Therefore, this embodiment comprehensively considers factors such as operational intensity, catch contribution, temporal proximity, spatial proximity, and abnormal operational characteristics to quantitatively assess the impact on fishing.

[0253] Specifically, suspected fishing operation segments are first identified based on AIS data, VMS data, fishing logs, and catch records.

[0254] For any ship trajectory sequence, if a ship simultaneously meets the following conditions within a preset time window: its speed is lower than a preset speed threshold, its course changes frequently, and its dwell time exceeds a preset dwell time threshold, the corresponding trajectory segment can be determined to be a suspected fishing operation segment.

[0255] Subsequently, suspected fishing operation segments are mapped to corresponding spatiotemporal grid units G(i,j,k) to form a fishing operation grid set.

[0256] Furthermore, to reflect the duration of fishing activities, this embodiment defines an operation intensity factor F_intensity.

[0257] Let T_fishing represent the duration of the fishing operation identified within the target time window; the unit is hours (h).

[0258] T_window represents the total duration of the corresponding time window; the unit is also hours (h).

[0259] The job intensity factor is defined as: F_intensity = T_fishing / T_window

[0260] Where: 0 ≤ F_intensity ≤ 1

[0261] Since the numerator and denominator have the same time dimension, F_intensity is a dimensionless parameter.

[0262] Furthermore, to reflect the contribution of the target fish species to the total catch, this embodiment defines a catch contribution factor F_catch.

[0263] Let C_target represent the catch of the target fish species, in tons (t); and C_total represent the total catch within the corresponding time window, also in tons (t).

[0264] The catch contribution factor is then expressed as: F_catch = C_target / C_total

[0265] Where: 0 ≤ F_catch ≤ 1

[0266] Since the numerator and denominator have the same mass dimension, F_catch is also a dimensionless parameter.

[0267] Furthermore, to reflect the degree of correlation between fishing activities and the time of resource appearance, this embodiment defines a time proximity factor F_time.

[0268] Let Δt represent the time difference between the time when fishing occurs and the time when resources are most likely to appear; the unit is hours (h).

[0269] T_ref represents the reference time scale; the unit is also hours (h).

[0270] The time proximity factor is then expressed as: F_time = exp(-Δt / T_ref)

[0271] Among them: 0 <F_time≤1

[0272] The closer F_time is to 1, the closer the fishing activity occurs is to the time when the resource is most likely to appear.

[0273] Furthermore, to reflect the degree of spatial correlation between fishing activities and areas with a high probability of resource distribution, this embodiment defines a spatial proximity factor F_space.

[0274] Let Δs represent the distance between the fishing location and the center of the area where resources are likely to appear; the unit is km.

[0275] S_ref represents the reference spatial scale; the unit is also km.

[0276] The spatial proximity factor is then expressed as: F_space = exp(-Δs / S_ref)

[0277] Where: 0 <F_space≤1

[0278] The closer the fishing location is to the area where resources are likely to be distributed, the closer F_space is to 1.

[0279] In addition, to improve the ability to identify abnormal fishing activities, this embodiment further introduces an abnormal operation confidence factor F_abnormal.

[0280] The abnormal operation confidence factor is used to reflect the degree of deviation between the target fishing behavior and the historical normal operation mode, and can be calculated in combination with the abnormal trajectory identification model, the abnormal transshipment identification model, or the abnormal catch declaration identification model.

[0281] In one implementation, normalization can be performed based on the number and intensity of abnormal features to obtain: 0 ≤ F_abnormal ≤ 1

[0282] The larger the value of F_abnormal, the more obvious the abnormal characteristics of the target fishing behavior.

[0283] After obtaining the above-mentioned influencing factors, this embodiment constructs the fishing influence parameter P_fishing:

[0284] P_fishing=α1F_intensity+α2F_catch+α3F_time+α4F_space+α5F_abnormal

[0285] Where α1, α2, α3, α4 and α5 represent the weight coefficients of the corresponding influencing factors, and satisfy: α1+α2+α3+α4+α5=1.

[0286] In one implementation, the weights can be dynamically adjusted based on the characteristics of the target fish species or regulatory requirements. For example, in a resource protection scenario, the weights of F_time and F_space can be increased; in an IUU fishing risk identification scenario, the weight of F_abnormal can be increased.

[0287] Since all the above factors have been normalized, the final fishing impact parameters satisfy:

[0288] 0≤P_fishing≤1

[0289] Through the above calculation process, this embodiment can not only identify whether the target fishing entity is carrying out fishing activities, but also quantify the actual correlation between fishing activities and target fish species resources, realizing the technical extension from "fishing behavior identification" to "fishing resource impact assessment".

[0290] After the above processing, each spatiotemporal grid cell G(i,j,k) corresponds to a fishing influence parameter P_fishing. This parameter, together with the resource occurrence probability parameter P_resource and the spatial overlap parameter P_overlap, will serve as important input parameters for the cross-border resource correlation calculation in subsequent steps 107 and 108.

[0291] IX. Step 107: Calculation of Resource Status and Comprehensive Confidence Parameters

[0292] After obtaining the resource occurrence probability parameter P_resource in step 104, the spatial overlap parameter P_overlap in step 105, and the fishing impact parameter P_fishing in step 106, in order to further improve the adaptability of the cross-border resource correlation measurement results to changes in resource status and data quality, this embodiment further calculates the resource status correction parameter R_state and the comprehensive confidence parameter C_conf.

[0293] Among them, the resource status correction parameter R_state is used to reflect the degree of influence of the current resource status of the target fish species on the resource correlation; the comprehensive confidence parameter C_conf is used to reflect the credibility and completeness of the data involved in the calculation.

[0294] Both serve as important input parameters for subsequent calculations of cross-border resource correlation scores.

[0295] (a) Calculation of resource status correction parameters

[0296] For highly migratory fish species, fishing activities of the same intensity under different resource conditions may have different degrees of impact on the resources.

[0297] For example, when the target fish species is in the resource recovery period or at a high level of resource abundance, the resource system usually has a strong recovery capacity; while when the target fish species is in the resource decline period, breeding period, or period of continuous high fishing pressure, even low-intensity fishing activities may have a significant impact on resource recovery.

[0298] Therefore, this embodiment introduces the resource state correction parameter R_state.

[0299] Specifically, resource status evaluation indicators are extracted based on the resource status data obtained in step 101.

[0300] The resource status evaluation indicators include at least the following:

[0301] Resource abundance index; resource decline index; historical fishing pressure index; breeding season status; fishing ban status.

[0302] To facilitate standardized calculations, the above indicators are first normalized. Let:

[0303] A_n represents the normalized resource abundance index; D_n represents the normalized resource decline index; F_n represents the normalized historical fishing pressure index; S_n represents the breeding season status index; C_n represents the closed season status index.

[0304] Where: 0≤A_n≤1; 0≤D_n≤1; 0≤F_n≤1; 0≤S_n≤1; 0≤C_n≤1.

[0305] Furthermore, construct resource status correction parameters:

[0306] R_state=β1A_n+β2D_n+β3F_n+β4S_n+β5C_n

[0307] Where: β1, β2, β3, β4 and β5 represent the corresponding index weights; satisfying: β1+β2+β3+β4+β5=1.

[0308] Therefore, we have: 0 ≤ R_state ≤ 1.

[0309] The closer R_state is to 1, the more sensitive the current resource status is or the more resource protection is needed.

[0310] (II) Calculation of comprehensive confidence parameters

[0311] Since this invention uses multi-source heterogeneous data for correlation analysis, different data sources may have different collection frequencies, update times, spatial precision, and degrees of data missing.

[0312] If all data are used directly in the calculation, the correlation results may be affected by abnormal or low-quality data.

[0313] Therefore, this embodiment further establishes a comprehensive confidence evaluation mechanism.

[0314] Specifically, for the data used in the correlation coefficient calculation, the following quality evaluation indicators are extracted:

[0315] Data source reliability; data integrity; time freshness; spatial accuracy.

[0316] in:

[0317] Data source reliability is used to reflect the trustworthiness of the data provider or the data source platform;

[0318] Data integrity is used to reflect the absence of key fields;

[0319] Time freshness is used to reflect the proximity between the data update time and the target analysis time;

[0320] Spatial accuracy is used to reflect the level of positioning error.

[0321] Let R_source represent the reliability index of the source of normalized data; R_complete represent the integrity index of normalized data; R_time represent the freshness index of normalized time; and R_space represent the spatial accuracy index of normalized data.

[0322] All the above indicators have been normalized and mapped to the interval [0,1].

[0323] Further construct comprehensive confidence parameters:

[0324] C_conf=γ1R_source+γ2R_complete+γ3R_time+γ4R_space

[0325] in:

[0326] γ1, γ2, γ3 and γ4 represent the weights of the corresponding quality indicators; satisfying: γ1+γ2+γ3+γ4=1.

[0327] Therefore, we have: 0 ≤ C_conf ≤ 1.

[0328] When C_conf is close to 1, it indicates that the current data has a high degree of reliability;

[0329] When C_conf is close to 0, it indicates that the data quality is low or there is a large degree of uncertainty.

[0330] (III) Joint Correction of Resource Status and Data Quality

[0331] In one implementation, the resource state correction parameter R_state and the comprehensive confidence parameter C_conf can be used in subsequent correlation score calculations.

[0332] R_state is used to enhance the evaluation capability in resource-sensitive scenarios; C_conf is used to correct the impact of data quality on the evaluation results.

[0333] For example, when the target fish species is in its breeding season and the resource decline index is high, R_state will be increased accordingly, thereby enhancing the responsiveness of the subsequent correlation score to the needs of resource protection.

[0334] When AIS trajectory data is severely missing, catch records are incomplete, or marine environmental data is outdated, C_conf will be reduced accordingly, thereby reducing the risk of low-quality data misleading the correlation results.

[0335] Through the aforementioned resource status correction mechanism and comprehensive confidence evaluation mechanism, this embodiment can not only reflect the changes in the resource status of the target fish species, but also reflect the reliability of the data involved in the calculation, providing a more stable and reliable evaluation basis for the cross-border resource correlation score calculation in step 108 and the dynamic output of the correlation level in step 109.

[0336] Step 108: Calculation of Cross-border Resource Relevance Score

[0337] After obtaining the resource occurrence probability parameter P_resource in step 104, the spatial overlap parameter P_overlap in step 105, the fishing impact parameter P_fishing in step 106, and the resource state correction parameter R_state and comprehensive confidence parameter C_conf in step 107, this embodiment further performs fusion calculation on the above parameters to generate the cross-border resource correlation score Score corresponding to the target highly migratory fish species.

[0338] The cross-border resource correlation score is used to comprehensively reflect the degree of cross-border correlation of target fish species resources in the current spatiotemporal grid unit. Its evaluation results can reflect not only the spatial distribution of resources, but also cross-border migration relationships, the degree of impact of fishing activities, and the sensitivity of resource protection.

[0339] Unlike traditional methods that use fixed-weight scoring models, this embodiment introduces a dynamic weighting mechanism, which enables each evaluation indicator to automatically adjust its contribution under different resource states, different migration stages, and different data quality conditions, thereby improving the adaptability of the correlation measurement results to actual resource changes.

[0340] Specifically, in this embodiment, the evaluation parameter set is first constructed as follows:

[0341] X={P_resource,P_overlap,P_fishing,R_state}

[0342] Where P_resource represents the resource occurrence probability parameter; P_overlap represents the spatial overlap parameter;

[0343] P_fishing represents the fishing impact parameter; R_state represents the resource status correction parameter.

[0344] The above parameters have all been normalized, therefore they satisfy:

[0345] 0≤P_resource≤1; 0≤P_overlap≤1; 0≤P_fishing≤1; 0≤R_state≤1.

[0346] Furthermore, construct the corresponding dynamic weight vector: W={w1,w2,w3,w4}

[0347] Among them, w1 corresponds to the resource occurrence probability parameter weight; w2 corresponds to the spatial overlap parameter weight; w3 corresponds to the fishing impact parameter weight; w4 corresponds to the resource status correction parameter weight, and satisfies w1+w2+w3+w4=1.

[0348] In one implementation, the initial values ​​of each weight can be set as follows:

[0349] w1=0.35; w2=0.25; w3=0.25; w4=0.15.

[0350] Subsequently, the weights are dynamically adjusted based on the current migration stage of the target fish species, changes in resource status, and comprehensive confidence parameters.

[0351] For example, when the target fish species is in its breeding season, the need for resource conservation is significantly enhanced, so the weights of the resource occurrence probability parameter and the fishing impact parameter should be increased.

[0352] When the target fish species is in the feeding period or resource dispersal stage, the weight of the spatial overlap parameter can be appropriately increased.

[0353] When the resource decline index continues to rise, the corresponding weight of the resource status correction parameter can be increased.

[0354] In one implementation, a dynamic adjustment factor can be constructed: K_m

[0355] Where K_m represents the adjustment coefficient corresponding to the current migration stage.

[0356] For example:

[0357] Breeding period: K_m=1.20; Foraging period: K_m=1.00; Migratory period: K_m=0.90; Overwintering period: K_m=0.80.

[0358] Simultaneously, a resource sensitivity adjustment factor, K_r, is constructed based on resource status parameters.

[0359] And a data confidence adjustment factor, K_c, is constructed based on the comprehensive confidence parameter.

[0360] Furthermore, the initial weights are adjusted:

[0361]

[0362] in, This represents the dynamic adjustment coefficient for the corresponding indicator.

[0363] Then, normalization was performed:

[0364]

[0365] Thus, the final dynamic weights are obtained:

[0366]

[0367] And satisfy: .

[0368] After obtaining the dynamic weights, this embodiment calculates the basic correlation score:

[0369]

[0370] Since all parameters are dimensionless and their values ​​range from [0,1], therefore .

[0371] Furthermore, to reflect the impact of data quality on the reliability of the evaluation results, this embodiment uses the comprehensive confidence parameter C_conf obtained in step 107 to correct the basic correlation score.

[0372] Specifically, the final cross-border resource correlation score is defined as: Score = Score_base × C_conf

[0373] Where: 0≤C_conf≤1.

[0374] Therefore: 0 ≤ Score ≤ 1.

[0375] In one implementation, to improve the readability of subsequent rating classifications, the relevance score can be further mapped to a percentage range: Score_final = 100 × Score

[0376] Therefore, we have: 0 ≤ Score_final ≤ 100.

[0377] The higher the Score_final, the greater the correlation between the target fishing entity, target fishing area, or target resource object and the target highly migratory fish transboundary resource; the lower the Score_final, the lower the correlation.

[0378] Through the aforementioned dynamic weight fusion mechanism, this embodiment can not only comprehensively reflect the resource availability status, cross-border spatial relationships, and the impact of fishing activities, but also automatically adjust the evaluation model according to changes in resource status and data quality, so that the correlation score results have better dynamic adaptability and practical interpretability.

[0379] Compared with the traditional fixed-weight scoring model, this embodiment can more accurately reflect the dynamic changes in the cross-border resource associations of highly migratory fish, providing a reliable data foundation for the dynamic output of association levels in the subsequent step 109.

[0380] XI. Step 109: Dynamic Output of Association Level

[0381] After obtaining the cross-border resource association score Score_final in step 108, this embodiment further determines the association level classification threshold based on the resource status of the target fish species and the data quality, and outputs the corresponding cross-border resource association level.

[0382] The association level is used to characterize the degree of association between the target fishing entity, target fishing area, target catch batch or target trade batch and the cross-border resources of highly migratory fish species, thereby providing a basis for decision-making in fishery resource supervision, aquatic product traceability and verification, resource protection and identification of illegal fishing risks.

[0383] Unlike traditional methods that use fixed scoring thresholds for grading, this embodiment employs a dynamic threshold mechanism. The grading criteria are dynamically adjusted based on the current resource status, migration stage, and data quality of the target fish species, enabling the associated grading output to more accurately reflect resource conservation needs and the credibility of the evaluation results.

[0384] Specifically, in one implementation, basic level classification rules are first established.

[0385] Let the correlation score obtained in step 108 be: Score_final

[0386] Its value range is: 0≤Score_final≤100.

[0387] Correspondingly, a basic level threshold can be set:

[0388] High correlation level: Score_final ≥ 70;

[0389] Association level: 40 ≤ Score_final < 70;

[0390] Low correlation level: Score_final < 40.

[0391] The above thresholds are for illustrative purposes only and can be adjusted according to business needs in actual applications.

[0392] Furthermore, since the actual resource risks represented by the same score differ under different resource states, this embodiment dynamically adjusts the level threshold based on the resource state correction parameter R_state.

[0393] For example, when the target fish species is in its breeding season, the resources are more sensitive to external fishing activities. At this time, even if the correlation score does not reach the usual high correlation threshold, there may still be a high risk of resource impact.

[0394] Therefore, when the target fish species is in its breeding season, the threshold corresponding to a high association level can be lowered.

[0395] For example:

[0396] High_Threshold=70-δ1

[0397] Where: δ1 represents the adjustment amount during the breeding season.

[0398] In one implementation: δ1 can be taken as 5 to 15 points.

[0399] Similarly, when the target fish species is in a period of resource decline or high fishing pressure, the threshold for high correlation level can be lowered.

[0400] Let R_state be the resource state correction parameter.

[0401] The high correlation level threshold can then be expressed as:

[0402] High_Threshold=70-ηR_state

[0403] Where η represents the resource status adjustment coefficient.

[0404] This makes it easier to trigger a high correlation level when the resource is more sensitive, thereby improving the resource protection capability.

[0405] Furthermore, when the target fish species is in the foraging period, resource recovery period, or a stage of high resource abundance, the threshold corresponding to the high correlation level can be appropriately increased to avoid excessive warnings.

[0406] In addition to resource status factors, this embodiment also considers the impact of data quality factors on the graded output results.

[0407] Since the correlation score is based on multi-source data, when there is significant data gaps or uncertainty, even if a high score is obtained, it is not advisable to directly output a deterministic level result.

[0408] Therefore, this embodiment introduces a review level mechanism.

[0409] Specifically, when the comprehensive confidence parameter calculated in step 107 satisfies:

[0410] C_conf < C_threshold

[0411] At that time, the system does not directly output the high correlation level, medium correlation level, or low correlation level, but instead outputs the level to be reviewed.

[0412] Where C_threshold represents the preset confidence threshold.

[0413] In one implementation:

[0414] C_threshold can be set to 0.50, 0.60, or 0.70.

[0415] For example, when there are large gaps in AIS trajectory data, incomplete catch records, or when key marine environmental data cannot be obtained, the overall confidence parameter will decrease.

[0416] Even if the correlation score is high, there may still be significant evaluation uncertainty.

[0417] By displaying the level of review pending, regulators can be prompted to supplement data or conduct manual verification, thereby reducing the risk of misjudgment.

[0418] Furthermore, in one implementation, the system can also generate association level explanation information.

[0419] The association level explanation information includes:

[0420] Contribution of resource occurrence probability; contribution of spatial overlap; contribution of fishing impact; contribution of resource status; and overall confidence level evaluation results.

[0421] The above explanation clarifies the reasons for the formation of association levels, improving the interpretability and traceability of association degree calculation results.

[0422] For example, for a specific target fishing species, the system can output:

[0423] The resource's probability of occurrence contributes 35%;

[0424] Spatial overlap contributed 22%;

[0425] The impact of fishing accounted for 28%;

[0426] The resource status contribution is 15%;

[0427] The overall confidence level is 0.91;

[0428] The final association level is high association level.

[0429] Through the aforementioned dynamic threshold adjustment mechanism and the pending review level output mechanism, this embodiment can not only dynamically adjust the risk sensitivity according to the resource status of the target fish species, but also dynamically control the credibility of the evaluation results according to the data quality, thereby improving the accuracy, stability and interpretability of the cross-border resource association level output results.

[0430] Finally, the system outputs the cross-border resource association level and association score results corresponding to the target height migratory fish, completing the dynamic calculation process of cross-border resource association.

[0431] 12. Example 1 – Case Study on Cross-border Resource Correlation Measurement of Bluefin Tuna

[0432] To further illustrate the specific application process of the technical solution of this invention, the following explanation uses the cross-border resource correlation calculation of bluefin tuna (Thunnusthynnus) as an example.

[0433] This embodiment selects June 1, 2024 to June 7, 2024 as the target analysis time window, with a window length of 7 days. Based on the electronic tag trajectory data, AIS vessel trajectory data, target sea area boundary data, marine environmental data, and resource status data obtained in step 101, it was found that bluefin tuna migrated from the exclusive economic zone of country A to the high seas within this time window, and continued to move towards the exclusive economic zone of country B, forming a typical cross-border migration path. Subsequently, according to steps 102 and 103, various types of data were standardized and subjected to spatiotemporal grid mapping, and corresponding spatiotemporal grid units were established.

[0434] In calculating the probability of resource occurrence, the main migration centerline of bluefin tuna within the current time window is first determined based on electronic tag trajectory data and historical migration path data. Calculations show that the shortest distance between the center of the target spatiotemporal grid cell and the migration centerline is approximately 120 km. Combining this with historical bluefin tuna migration patterns, a reference migration distance of 500 km is set, and the distance attenuation factor is calculated using the distance attenuation model from step 104. The distance attenuation factor, calculated using the formula F_d=exp(-d / D_ref), is approximately 0.787.

[0435] Meanwhile, based on resource survey results released by regional fisheries management organizations, the current time window corresponds to the breeding stage of bluefin tuna, therefore the migration stage correction factor is set to 1.00. Further combining marine environmental monitoring data, environmental indicators such as sea surface temperature, salinity, ocean current velocity, water depth, and chlorophyll concentration are normalized, and the environmental suitability index is obtained as approximately 0.843 according to the environmental suitability calculation model in step 104. Subsequently, the resource occurrence probability parameter is calculated using the distance attenuation factor, the migration stage correction factor, and the environmental suitability index. Finally, the resource occurrence probability parameter P_resource corresponding to the target spatiotemporal grid unit is approximately 0.663, indicating that the area has a high probability of bluefin tuna resource occurrence within the current time window.

[0436] During the spatial overlap analysis, based on the maritime boundary analysis results in step 105, it was found that the target spatiotemporal grid has spatial relationships with the exclusive economic zones (EEZs) of country A, the high seas, and country B. Specifically, the overlap area between the target grid and country A's EEZ accounts for approximately 35% of the total grid area, with the high seas approximately 40%, and with country B's EEZ approximately 25%. Through overlap area analysis and normalization, the comprehensive overlap area index is approximately 0.73.

[0437] Further analysis of the bluefin tuna's migration trajectory within a continuous time window revealed 10 migration connections within the statistical period, 8 of which occurred between different sea areas. Therefore, based on the cross-border connectivity model in step 105, the cross-border connectivity parameter was calculated to be approximately 0.80. Subsequently, the spatial overlap parameter P_overlap was calculated by combining the overlap area index and the cross-border connectivity parameter, ultimately obtaining a P_overlap of approximately 0.758, indicating a strong spatial correlation between the target resource and multiple cross-border associated sea areas.

[0438] During the fishing impact analysis, AIS trajectory data and catch log data identified that the target vessel was engaged in continuous fishing activities during the analysis period. Statistical results showed that the vessel's cumulative operating time was approximately 84 hours, accounting for 50% of the entire time window. Simultaneously, the catch of the target fish species was approximately 32 tons, representing 80% of the total catch of 40 tons. Combining the temporal proximity between fishing time and the time of high-probability resource occurrence, and the spatial proximity between fishing location and the area of ​​high-probability resource distribution, temporal proximity factors and spatial proximity factors were obtained. Further combining the abnormal operation analysis results, and calculating according to the fishing impact model in step 106, the final fishing impact parameter P_fishing was approximately 0.727.

[0439] Subsequently, the resource status was analyzed based on the resource assessment data. After normalization, indicators such as resource abundance index, resource decline index, historical fishing pressure index, and breeding season status were input into the resource status model, ultimately yielding a resource status correction parameter R_state of approximately 0.840. Simultaneously, a comprehensive confidence parameter was calculated based on quality evaluation indicators such as data source reliability, data integrity, time freshness, and spatial accuracy, resulting in a comprehensive confidence parameter C_conf of approximately 0.928, indicating that the data used in the calculation has high reliability.

[0440] After calculating the above parameters, the resource occurrence probability parameter, spatial overlap parameter, fishing impact parameter, and resource status correction parameter are fused according to the dynamic weight fusion mechanism in step 108. In this embodiment, the resource occurrence probability parameter has a weight of 0.35, the spatial overlap parameter has a weight of 0.25, the fishing impact parameter has a weight of 0.25, and the resource status correction parameter has a weight of 0.15. After weighted fusion, the basic correlation score is approximately 0.729. After further correction using the comprehensive confidence parameter, the final correlation score is approximately 0.676, corresponding to a percentage score of approximately 67.6.

[0441] According to the basic grading rules, 67.6 points corresponds to a medium correlation level. However, based on the resource status analysis, the bluefin tuna are currently in their breeding season, and the resource decline index is at a high level, thus triggering the dynamic threshold adjustment mechanism in step 109. The system automatically lowers the threshold corresponding to the high correlation level based on the resource status correction parameters, adjusting the high correlation level threshold from the original 70 points to 61.6 points. Since the final correlation score of 67.6 points is higher than the adjusted threshold, the system ultimately outputs a high correlation level.

[0442] It can be seen that this invention can not only accurately identify the cross-border correlation of target resources by utilizing fish migration routes, marine boundary information and fishing behavior information, but also dynamically adjust the evaluation results by combining changes in resource status and data quality, so that the correlation measurement results are more in line with the actual resource change patterns, thereby improving the accuracy, reliability and interpretability of cross-border resource correlation analysis.

[0443] XIII. Example 2 – A Case Study on Cross-border Resource Linkage Tracing of Imported Frozen Tuna Batches

[0444] To further illustrate the application of this invention in the scenarios of aquatic product traceability and import risk supervision, the following uses a batch of imported frozen bluefin tuna as an example to explain the dynamic calculation process of cross-border resource correlation of this invention.

[0445] In this example, an importing company declared a batch of frozen bluefin tuna products to the regulatory authorities. The declaration documents show that the batch of products originated from the deep-sea fishing vessel V001, was caught in mid-May 2024, and after being transshipped at sea, was transported to an overseas processing plant for primary processing before entering the import distribution chain as frozen fish products. The regulatory authorities wish to determine the degree of correlation between this batch of products and the cross-border resources of highly migratory fish species, and assess whether it involves sensitive resource areas or potential IUU (intellectually unripe) fishing risks.

[0446] First, based on product traceability information, we obtain the fishing vessel identification, transshipment vessel identification, fishing time, fishing area, and catch record information corresponding to this batch. Then, based on the vessel identification, we obtain AIS trajectory data, VMS monitoring data, electronic fishing logs, and transshipment record data. Simultaneously, we obtain bluefin tuna electronic tag data, marine environmental monitoring data, resource status assessment data, and relevant sea area boundary data for the corresponding time window.

[0447] Subsequently, the data was standardized according to step 102. The system uniformly converted data from different sources to the WGS84 coordinate system and processed it into time slices according to a 7-day time window. At the same time, the species code, trade code, and commodity name of bluefin tuna were uniformly mapped to establish the association between the target species and the target batch.

[0448] Further, a spatiotemporal grid system was constructed according to step 103. The system constructed spatiotemporal grid units with a spatial resolution of 10km×10km and a time window of 7 days, and mapped fishing trajectories, transshipment trajectories, fish migration trajectories, and resource status data to the corresponding spatiotemporal grids. Analysis revealed that the target vessel mainly operated in the area where the exclusive economic zone of country A and the adjacent high seas met during the fishing period, and moved eastward along the main migration route of bluefin tuna in the following days.

[0449] In step 104, the system calculates the resource occurrence probability parameter based on the bluefin tuna electronic tag trajectory and marine environmental data. Analysis results show that the target fishing area is located near the main activity area of ​​bluefin tuna during their breeding season, and the sea surface temperature, ocean current conditions, and chlorophyll concentration are all within suitable ranges. Therefore, the resource occurrence probability parameter in the corresponding spatiotemporal grid cell is relatively high. The calculated resource occurrence probability parameter for the target batch's corresponding fishing area reaches 0.79, indicating a high probability of bluefin tuna resource distribution in this area.

[0450] In step 105, the system performs spatial overlap analysis on the target fishing area. The analysis results show that the target spatiotemporal grid overlaps with the exclusive economic zone, high seas, and cross-border migration channels of country A. Furthermore, based on fish migration trajectories within continuous time windows, it was found that the target fish population continuously traversed multiple national jurisdictional waters and high seas before and after fishing, exhibiting clear cross-border migration characteristics. Combining the overlap area ratio and cross-border connectivity analysis results, the spatial overlap parameter is approximately 0.82.

[0451] In step 106, the system analyzes the fishing behavior. Based on AIS trajectory and electronic fishing log data, it was found that the target vessel continuously conducted fishing operations in areas with a high probability of resource occurrence, and there was a high proportion of bluefin tuna catches. Furthermore, the fishing time highly overlapped with the target fish population's breeding season, and the fishing location showed a high spatial proximity to areas with a high probability of resource distribution. Further analysis of transshipment records revealed a complete traceability chain between the catch in the target batch and the target vessel's fishing records; therefore, the final fishing impact parameter was determined to be approximately 0.76.

[0452] Subsequently, in step 107, the system calculates resource status correction parameters based on resource assessment reports released by regional fisheries management organizations. Analysis results show that the target fish species is currently in its breeding season, and the resource decline index is at a high level; therefore, the resource status correction parameter reaches 0.88. Furthermore, since the AIS track, VMS data, catch records, and transshipment records are all relatively complete and the data updates are recent, the overall confidence parameter reaches 0.93.

[0453] In step 108, the system dynamically weights and fuses the resource occurrence probability parameter, spatial overlap parameter, fishing impact parameter, and resource status correction parameter, and then corrects them using a comprehensive confidence parameter. The final cross-border resource correlation score for this import batch is 74.8 points.

[0454] Furthermore, in step 109, the system determines the level based on resource status and a dynamic threshold mechanism. Since the bluefin tuna are in their breeding season and the resource decline index is high, the threshold corresponding to a high correlation level is appropriately lowered. The final correlation score is higher than the adjusted high correlation level threshold, therefore the system outputs a high correlation level result.

[0455] Simultaneously, the system generates correlation explanation information, displaying the contribution of resource occurrence probability, spatial overlap, fishing impact, resource status, and overall confidence level evaluation results. Among these, the high proportions of resource occurrence probability and spatial overlap indicate that this batch of products mainly originates from areas with active cross-border migration of bluefin tuna; the high contribution of resource status indicates a high sensitivity to current resource conservation.

[0456] Based on the above analysis, regulatory authorities can further include this batch in the key inspection targets and conduct subsequent verification work by combining fishing permit information, transshipment permit information, and data from international fisheries management organizations. Compared with the traditional method of judging solely based on fishing location or trade documents, this invention can comprehensively utilize information on fish migration patterns, cross-border maritime relationships, fishing behavior characteristics, and changes in resource status to dynamically and quantitatively analyze the correlation between the target imported batch and highly migratory fish cross-border resources, thereby improving the accuracy and reliability of aquatic product traceability verification and risk identification.

[0457] As can be seen from the above embodiments, this invention establishes a dynamic measurement mechanism for the cross-border resource correlation of highly migratory fish species through multi-source fisheries data standardization processing, spatiotemporal gridding mapping, resource occurrence probability calculation, cross-border spatial overlap analysis, fishing impact parameter calculation, resource status and comprehensive confidence correction, and dynamic threshold hierarchical output. This mechanism can adapt to the resource correlation analysis needs of different fish species, different sea areas, different time windows, and different data quality conditions, thereby improving the accuracy and reliability of cross-border fisheries resource monitoring, fishing activity supervision, aquatic product traceability verification, and illegal fishing risk identification.

[0458] It should be understood that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. For those skilled in the art, various modifications, substitutions, or combinations can be made to the data sources, time windows, spatial resolutions, parameter weights, threshold settings, calculation models, and module structures in the above embodiments without departing from the concept and essence of the present invention, and all such modifications, substitutions, or combinations should fall within the scope of protection of the present invention.

[0459] The specific numerical values, weighting coefficients, threshold ranges, and example scenarios described in this specification are only for illustrating the implementation of the technical solution of this invention and do not constitute a limitation on the scope of protection of this invention. In practical applications, the relevant parameters can be adaptively adjusted according to the characteristics of the target fish species, regulatory requirements, data quality, and changes in resource status.

[0460] It should be noted that, where there is no conflict, the various embodiments of the present invention and the technical features thereof can be combined with each other. All equivalent substitutions, equivalent transformations, or obvious improvements made based on the specification and claims of this invention should be included within the scope of protection of this invention. The scope of protection of this invention shall be determined by the scope defined in the claims.

Claims

1. A method for dynamically calculating the cross-border resource correlation of highly migratory fish species, characterized in that, Includes the following steps: Acquire basic data on fish species, migration routes, target sea area boundaries, fishing behavior, marine environment, and resource status for fish species migrating at the target altitude; The basic data of fish species, migration path data, target sea area boundary data, fishing behavior data, marine environment data, and resource status data are uniformly processed by coordinate reference, time granularity, and fish species coding to obtain standardized multi-source data; According to the preset spatial resolution and preset time window, the target analysis sea area is divided into multiple spatiotemporal grid units with spatial and temporal attributes, and the standardized multi-source data is mapped to the corresponding spatiotemporal grid units. Based on the migration route data, marine environment data, and resource status data, the resource occurrence probability parameter of migratory fish at the target altitude in each of the spatiotemporal grid cells is calculated, wherein the resource occurrence probability parameter is determined according to the migration distance attenuation value, the migration stage correction coefficient, and the environmental suitability correction coefficient; Based on the target sea area boundary data, the spatial overlap parameters between each of the spatiotemporal grid cells and at least two cross-border associated sea areas are calculated, wherein the spatial overlap parameters are determined according to the overlap area ratio and cross-border connectivity parameters. Based on the fishing behavior data, suspected fishing operation segments are identified, the suspected fishing operation segments are matched with the spatiotemporal grid cells, and the impact parameters of the target fishing entity or target fishing area on the target highly migratory fish are calculated based on the matching results. Based on the resource status data and data quality information, determine the resource status correction parameters and the comprehensive confidence level parameters; Based on the migration stage of the target height migratory fish, changes in resource status, and comprehensive confidence parameters, dynamic weights are determined. The resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters are then fused according to the dynamic weights to generate a cross-border resource correlation score for the target height migratory fish. The correlation degree grading threshold is dynamically determined based on the resource status data and comprehensive confidence parameters. The cross-border resource correlation score is compared with the correlation degree grading threshold, and the cross-border resource correlation level of the target highly migratory fish is output.

2. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The process of unifying the coordinate reference, time granularity, and fish species coding of the basic fish species data, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data includes: Transform marine boundary data, vessel tracks, fishing locations, fish migration routes, and marine environmental data from different sources into the same geographic coordinate system; The data on vessel trajectories, fishing locations, migration routes, and marine environments are sliced ​​over time according to a preset time window. Map the same fish species identifier to fish species names, Latin names, commercial names, and codes from different data sources; Data confidence weights are set for data records that are missing, conflicting, or abnormal, and the data mapped to the spatiotemporal grid cells are corrected based on the data confidence weights.

3. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The process of dividing the target sea area into multiple spatiotemporal grid units with spatial and temporal attributes includes: The target analysis area is divided into multiple spatial grids according to the preset spatial resolution; Each spatial grid is expanded into a spatiotemporal grid unit according to a preset time window; Configure each spatiotemporal grid cell with a grid number, time window, sea area affiliation attribute, migration stage attribute, initial value of resource occurrence probability, and data confidence label; The sea area attribution includes at least one or more of the following: national jurisdiction waters, high seas, and cross-border migration routes.

4. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The calculation of the resource occurrence probability parameter includes: Based on the migration path data, determine the migration centerline, migration buffer zone, and migration direction of migratory fish at the target altitude within different time windows; Calculate the distance between each of the aforementioned spatiotemporal grid cells and the migration centerline or migration buffer zone to obtain the migration distance attenuation value; Based on the characteristics of breeding, foraging, overwintering, or migratory fish at different migration stages at the target altitude, a correction coefficient for the migration stage is determined. Determine the environmental suitability correction factor based on at least one of the following: water temperature, salinity, water depth, ocean current, chlorophyll concentration, or habitat suitability data. Based on the migration distance attenuation value, migration stage correction coefficient, and environmental suitability correction coefficient, the resource occurrence probability parameters in each of the spatiotemporal grid cells are obtained.

5. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 4, characterized in that, The resource occurrence probability parameter is determined in the following way: The initial resource occurrence probability is obtained by multiplying or weighting the migration distance attenuation value, migration stage correction coefficient, and environmental suitability correction coefficient. The initial resource occurrence probability is corrected based on the resource abundance index, resource decline index, or historical resource distribution records to obtain the resource occurrence probability parameter.

6. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The calculation of the spatial overlap parameter includes: Determine multiple spatiotemporal grid cells that migratory fish at the target altitude will pass through within a preset time window; Calculate the percentage of overlapping area between the multiple spatiotemporal grid cells and different national jurisdictional waters, high seas areas, or cross-border migration routes; Calculate cross-border connectivity parameters based on the connectivity relationships between adjacent spatiotemporal grid cells in different cross-border associated sea areas; The spatial overlap parameters are generated based on the overlap area ratio and cross-border connectivity parameters.

7. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The calculation of the fishing impact parameters includes: Obtain at least one of the following information about the target fishing entity: vessel trajectory, operation time, operation speed, changes in course, duration of stay, type of fishing gear, fishing log, catch amount, or transshipment record; Suspected fishing operation segments were identified based on changes in speed, course, and duration of stay in the vessel's trajectory. The suspected fishing operation segments are matched with the spatiotemporal grid cells within the corresponding time window to obtain a fishing operation grid set; The fishing impact parameters are calculated based on the fishing operation intensity, catch percentage, temporal proximity, spatial proximity, and abnormal operation confidence in the fishing operation grid set.

8. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The cross-border resource correlation score is determined in the following way: The resource occurrence probability parameter, spatial overlap parameter, fishing impact parameter, and resource status correction parameter are used as positive correlation parameters; The comprehensive confidence parameter is used as the scoring correction parameter; The dynamic weights of each positive correlation parameter are updated based on the migration stage of the target migratory fish, changes in resource status, and overall confidence parameters. The positive correlation parameters are weighted and summed, and the weighted summation result is corrected using the comprehensive confidence parameter to obtain the cross-border resource correlation score.

9. The method for dynamically calculating the cross-border resource correlation of highly migratory fish species according to claim 1, characterized in that, The step of dynamically determining the correlation level threshold based on the resource status data and comprehensive confidence parameters includes: When the target highly migratory fish species are in their breeding season, resource decline period, or high fishing pressure period, increase the dynamic weights corresponding to the resource occurrence probability parameter and the fishing impact parameter, and lower the scoring threshold corresponding to the high correlation level. When the target high-altitude migratory fish are in a non-critical migration phase, increase the dynamic weight corresponding to the spatial overlap parameter; When the overall confidence level parameter is lower than the preset confidence threshold, the data review weight is increased and the correlation level to be reviewed is output. The cross-border resource association level includes at least three or four of the following: high association level, medium association level, low association level, and level pending review.

10. A dynamic measurement system for the cross-border resource correlation of highly migratory fish species, characterized in that, include: The data acquisition module is used to acquire basic data on target migratory fish species, migration route data, target sea area boundary data, fishing behavior data, marine environmental data, and resource status data. The data standardization module is used to uniformly process the fish species basic data, migration path data, target sea area boundary data, fishing behavior data, marine environment data, and resource status data in terms of coordinate reference, time granularity, and fish species coding to obtain standardized multi-source data. The spatiotemporal grid construction module is used to divide the target analysis sea area into multiple spatiotemporal grid units with spatial and temporal attributes according to a preset spatial resolution and a preset time window, and to map the standardized multi-source data to the corresponding spatiotemporal grid units. The resource occurrence probability calculation module is used to calculate the resource occurrence probability parameters of migratory fish at the target altitude in each of the spatiotemporal grid cells based on the migration path data, marine environment data, and resource status data. The spatial overlap calculation module is used to calculate the spatial overlap parameters between each of the spatiotemporal grid cells and at least two cross-border associated sea areas based on the target sea area boundary data. The fishing impact calculation module is used to identify suspected fishing operation segments based on the fishing behavior data, match the suspected fishing operation segments with the spatiotemporal grid cells, and calculate the fishing impact parameters of the target fishing entity or target fishing area on the target highly migratory fish based on the matching results. The status and confidence calculation module is used to determine the resource status correction parameters and the comprehensive confidence parameters based on the resource status data and data quality information. The correlation scoring module is used to determine dynamic weights based on the migration stage of the target height migratory fish, changes in resource status, and comprehensive confidence parameters. The module then integrates the resource occurrence probability parameters, spatial overlap parameters, fishing impact parameters, resource status correction parameters, and comprehensive confidence parameters according to the dynamic weights to generate a cross-border resource correlation score for the target height migratory fish. The dynamic grading output module is used to dynamically determine the correlation grading threshold based on the resource status data and comprehensive confidence parameters, compare the cross-border resource correlation score with the correlation grading threshold, and output the cross-border resource correlation level of the target highly migratory fish.

Images