A method and system for differentiating typhoon impacts based on seawater CO2 partial pressure data

Abstract

This invention discloses a method and system for differentiating typhoon impacts based on seawater CO2 partial pressure data. The method includes: acquiring marine situation monitoring data during typhoon passage in the target sea area through a multi-source observation platform, constructing a standardized dataset and a climatological background field; using the DBSCAN clustering algorithm to segment water masses, and combining the Man-Whitney U test and correlation analysis to determine the distribution of characteristic typhoon water masses; calculating the seawater CO2 partial pressure anomaly field based on the climatological background field, constructing a typhoon impact index time series, identifying the core impact area, and generating typhoon impact trace paths; subsequently reconstructing seawater CO2 partial pressure grid points, and constructing a spatial map of typhoon impacts; finally, analyzing the spatiotemporal changes in air-sea CO2 flux, assessing carbon sink stability, and generating a typhoon-carbon sink sensitivity zoning map. Using seawater CO2 partial pressure as the core indicator, this method overcomes the limitations of traditional methods and provides reliable technical support for assessing the marine environmental impact of typhoons.

Description

Technical Field

[0001] This invention relates to the field of typhoon and ocean response analysis technology, and in particular to a method and system for distinguishing typhoon impacts based on seawater CO2 partial pressure data. Background Technology

[0002] Typhoons are among the most destructive natural disasters, and one of the natural disasters with the greatest impact on people's property safety and daily life. At the same time, typhoons are also an important dynamic process affecting marine ecosystems. Related studies show that typhoon passage causes strong surface cooling, increases phytoplankton growth, enhances the output of biological organic carbon in the euphotic zone, increases phytoplankton populations and marine fisheries, and also makes a significant contribution to the annual and seasonal air-sea CO2 exchange in regional and global oceans.

[0003] Currently, the analysis of the impact of typhoon processes on air-sea CO2 exchange is mainly based on in-situ sampling data. However, since most in-situ sampling occurs after the typhoon has passed, and typhoons are often accompanied by various dynamic processes such as vortices, upwelling, and horizontal mixing, the analysis of the impact of typhoon processes should first spatially differentiate the relevant data. Therefore, a scientific data segmentation method is particularly important.

[0004] Therefore, this invention proposes a method and system for distinguishing typhoon impacts based on seawater CO2 partial pressure data, which enables accurate and dynamic differentiation of typhoon-affected areas and ultimately serves the accurate assessment of marine carbon sinks and climate change research, thereby improving the accuracy of monitoring the marine ecological effects of typhoons. Summary of the Invention

[0005] This invention overcomes the shortcomings of existing technologies and provides a method and system for distinguishing typhoon impacts based on seawater CO2 partial pressure data. Its main purpose is to achieve accurate and dynamic differentiation of typhoon-affected areas and ultimately serve the accurate assessment of marine carbon sinks and climate change research, thereby improving the accuracy of monitoring the marine ecological effects of typhoons.

[0006] To achieve the above objectives, the first aspect of this invention provides a method for differentiating typhoon impacts based on seawater CO2 partial pressure data, comprising:

[0007] Acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climatological background field based on historical marine situation data of the target sea area.

[0008] Based on the aforementioned marine situation monitoring dataset, a clustering algorithm was used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and conduct correlation analysis and significance tests on each water mass to obtain the distribution information of typhoon characteristic water masses.

[0009] The anomaly field of seawater carbon dioxide partial pressure during the passage of the typhoon is calculated based on the climatological background field. The time series sequence of typhoon impact index is constructed through the anomaly field of seawater carbon dioxide partial pressure and the typhoon impact trace path is generated.

[0010] The grid points of seawater carbon dioxide partial pressure in the target sea area are reconstructed using the sea area situation monitoring dataset. A spatial map of typhoon impact is constructed by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information.

[0011] Based on the spatial map of typhoon impact, the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the passage of typhoons are analyzed, and the carbon sink stability of the target sea area after the passage of typhoons is analyzed, generating a map of typhoon-carbon sink sensitive areas.

[0012] In this scheme, the acquisition of marine situation monitoring data during typhoon passage within the target sea area, the preprocessing to generate a marine situation monitoring dataset, and the construction of a climatological background field based on historical marine situation data of the target sea area specifically include:

[0013] The data obtained from the multi-source observation platform include sea surface wind field data, rainfall data, sea surface temperature data (SST), sea surface chlorophyll data, and sea surface height data retrieved from satellite remote sensing, as well as seawater carbon dioxide partial pressure, salinity, and dissolved oxygen data collected by the on-site observation platforms of buoys and monitoring vessels.

[0014] The acquired marine situation monitoring data is preprocessed, and a threshold screening method is used to remove outliers caused by instrument errors. The spatiotemporal consistency test method is used to identify and correct missing or distorted data.

[0015] Subsequently, all data were resampled to a preset standard spatial grid coordinate system, spatially normalized using a bilinear interpolation algorithm, and averaged daily according to Coordinated Universal Time to generate a marine situation monitoring dataset.

[0016] By introducing a big data network, historical sea area situation data of the target sea area during the same period of typhoon occurrence is obtained through the big data network to generate a historical sea area situation grid map. At each identical spatial grid point, the arithmetic mean of historical sea area situation data for the same month over many years is calculated as the monthly climatological mean, and the standard deviation is calculated as a measure of climate variability to generate a climatological background field, including the mean field and the standard deviation field.

[0017] In this scheme, a clustering algorithm is used to divide water masses based on the marine situation monitoring dataset, identifying water masses of different properties formed under the influence of typhoons, and performing correlation analysis and significance testing on each water mass to obtain the distribution information of typhoon characteristic water masses, specifically including:

[0018] Obtain a marine situation monitoring dataset, generate a marine situation grid map of the target sea area during the passage of a typhoon based on the marine situation monitoring dataset, extract the marine situation features corresponding to each grid in the marine situation grid map, and generate a grid feature matrix after standardization processing.

[0019] The grid feature matrix is ​​used as the input of principal component analysis. The covariance matrix between each feature parameter is calculated. The principal component direction is obtained through eigenvalue decomposition. The top k principal components whose cumulative contribution rate reaches the preset contribution rate threshold are selected. The original high-dimensional feature vector is projected onto the principal component space to generate the multimodal marine situation feature vector of each grid unit.

[0020] The multimodal marine situation feature vectors corresponding to each grid cell are abstracted into individual grid nodes. The DBSCAN clustering algorithm is introduced to classify the nodes. The neighborhood radius ε and the minimum number of points MinPts of the DBSCAN clustering algorithm are set. For each grid node, each grid node is marked as unvisited to complete the initialization. A grid node is randomly selected for visit. The grid node that enters the visited state is marked as grid node p for cluster expansion.

[0021] For a grid node p, calculate its Euclidean distance to all other nodes in the feature space, and find all neighboring nodes whose distance is less than a preset distance threshold. If the number of neighboring nodes is not less than MinPts, then mark node p as a core point and create a new cluster, adding the node and all its neighboring nodes to the cluster. If the number of neighboring nodes is less than MinPts, then mark node p as a noise point.

[0022] For each neighboring node added to the cluster, recursively perform neighborhood query and expansion operations. If the node is a core node, continue to add the unvisited nodes in its neighborhood to the current cluster. Repeat this process until no new nodes are added to the current cluster, completing the division of a water cluster category. Traverse all unvisited nodes and repeat the above process until all nodes are visited, obtaining the final set of clusters.

[0023] In this scheme, the water mass division based on the marine situation monitoring dataset using a clustering algorithm is performed to identify water masses of different properties formed under the influence of typhoons. Correlation analysis and significance testing are then conducted on each water mass to obtain the distribution information of typhoon-characteristic water masses. The scheme also includes:

[0024] Based on the final cluster set, several water masses are generated. The mean, standard deviation, skewness and kurtosis of the situation characteristics of each sea area within each water mass are calculated to generate a water mass feature statistics table. Combined with the preset water mass category labels, the water mass categories are divided to generate initial water mass category classification information.

[0025] Based on the initial water mass category classification information, the Mann-Whitney U test method was used to test the significance of parameter differences between different water mass categories, calculate the significance probability value of each pair of water masses on each parameter, and screen out water mass pairs with statistically significant differences.

[0026] Meanwhile, within each water mass, the Pearson correlation coefficient between seawater carbon dioxide partial pressure and other parameters is calculated to assess the strength of the linear correlation between parameters. Water masses that do not have a linear correlation are separated, and finally, water mass category classification information is obtained.

[0027] Using the water mass category classification information, the Delaunay triangulation algorithm is used to construct a triangular mesh for the spatial sample points contained in each water mass category. Then, the convex hull boundary of each water mass point set is calculated to generate the spatial distribution polygon of each water mass. The boundary polygon is smoothed to eliminate irregular jagged edges, and a spatial distribution map of the water mass is obtained.

[0028] The spatial characteristics of the identified typhoon-affected water masses are calculated using the aforementioned water mass spatial distribution map. The distribution information of typhoon-characteristic water masses is then generated by combining the corresponding sea area situation characteristics, significance test results, and correlation analysis results.

[0029] In this scheme, the step of calculating the seawater carbon dioxide partial pressure anomaly field during the typhoon's passage based on the climatological background field, constructing a typhoon impact index time series using the seawater carbon dioxide partial pressure anomaly field, and generating a typhoon impact trace path specifically includes:

[0030] Obtain the climatological background field, extract the monthly climatological mean field from the climatological background field as the reference data, obtain the marine situation monitoring dataset and extract the daily seawater carbon dioxide partial pressure grid data during the typhoon's passage.

[0031] The climate anomaly algorithm is used to calculate the difference between the daily observation value and the monthly climatological mean value of each spatial grid point, thereby generating the daily seawater carbon dioxide partial pressure anomaly value of each grid point, which covers the entire target sea area during the typhoon.

[0032] The distribution information of characteristic water masses of typhoons is obtained, and the spatial distribution range of the core water mass affected by typhoons is extracted as an analysis mask. Within this spatial mask, multi-dimensional features are extracted from the daily seawater carbon dioxide partial pressure anomaly field. The arithmetic mean of the outlier values ​​of all grid points is used as the anomaly intensity feature, the standard deviation of the outlier values ​​of all grid points is used as the spatial heterogeneity feature, and the proportion of grid points whose absolute outlier values ​​exceed the preset threshold is used as the significant influence range feature.

[0033] The extracted multi-dimensional features are normalized, and then the weight coefficients of each feature are determined by a subjective and objective weighting method. The objective weights are calculated by the entropy weight method based on the degree of variation of each feature value, and the subjective weights are determined by the analytic hierarchy process combined with expert scoring. Finally, the subjective and objective weights are weighted and fused to obtain the comprehensive weight. The normalized multi-dimensional features are weighted and summed to calculate the daily typhoon impact index value, and a continuous time series of typhoon impact indexes throughout the entire life cycle of a typhoon is constructed.

[0034] Based on the anomaly field of carbon dioxide partial pressure in seawater, an image segmentation algorithm based on Otsu's maximum inter-class variance method is used to identify and segment the anomaly region. The anomaly field is binarized into significant anomaly regions and non-significant regions. Morphological operations are used to optimize the segmentation results, remove small noise regions and fill holes, and obtain the spatial distribution map of the typhoon core impact area.

[0035] Based on the spatial distribution map of the typhoon's core impact area, the coordinates of the geometric center point of the polygon in the impact area are calculated. The centroid algorithm is used to obtain the center point position by weighted averaging of the polygon vertex coordinates. The center points of each day during the typhoon's passage are connected by spline curves in chronological order to generate the typhoon impact trace path. At the same time, the date corresponding to each center point and the typhoon impact index value at that time are recorded.

[0036] In this scheme, the reconstruction of seawater carbon dioxide partial pressure grid points in the target sea area using the marine situation monitoring dataset, combined with the typhoon impact index time series, typhoon impact trace paths, and typhoon characteristic water mass distribution information, to construct a typhoon impact spatial map, specifically includes:

[0037] A marine situation monitoring dataset is obtained. Based on the marine situation monitoring dataset, marine situation monitoring features are extracted as feature variables. Seawater carbon dioxide partial pressure data observed in the field are used as target variables. A seawater carbon dioxide partial pressure reconstruction model is constructed using a random forest regression algorithm.

[0038] The target sea area is divided into several spatial grids. The corresponding environmental feature vectors are extracted from each grid point using the sea area situation monitoring dataset. The vectors are then input into the trained seawater carbon dioxide partial pressure reconstruction model to calculate the estimated seawater carbon dioxide partial pressure at the grid points, thereby generating a seawater carbon dioxide partial pressure reconstruction field covering the entire target sea area.

[0039] The typhoon impact index time series and typhoon impact trace path are obtained. After smoothing the typhoon impact index time series, the key feature points of the typhoon impact intensity are identified by the local extremum detection algorithm, including the starting point, peak point and attenuation point. The morphological features of the sequence are extracted by the time series segmentation aggregation approximation method. Combined with the time axis information of the typhoon passage, a time series feature map reflecting the evolution process of the typhoon impact intensity is generated.

[0040] Based on the latitude and longitude coordinates of the center point of the daily impact area and the corresponding typhoon impact index value extracted from the typhoon impact trace path, the inverse distance weighted interpolation algorithm is used for spatial interpolation to transform the discrete point-like impact intensity data into a continuous spatial distribution field, thus obtaining the spatial distribution field of typhoon impact intensity.

[0041] Spatial registration was performed on the seawater carbon dioxide partial pressure reconstruction field, the spatial distribution field of typhoon impact intensity, and the temporal characteristic map. Spatial correlation between different sea area situation characteristics was established through raster calculation and vector overlay to generate an initial typhoon impact spatial map.

[0042] The distribution information of characteristic water masses of typhoons is obtained and superimposed on the initial typhoon impact spatial map. The boundary attributes of water masses are correlated with the partial pressure field of carbon dioxide in the seawater. The degree of abnormal change caused by the typhoon is obtained by difference calculation using the climatological background field as a reference benchmark. The carbon dioxide partial pressure distribution is rendered with gradient colors using visualization technology, the influence intensity is represented by contour lines, and the water mass boundary is marked with preset rule symbols. Finally, the typhoon impact spatial map is obtained.

[0043] This scheme analyzes the spatiotemporal dynamics of carbon flux and the stability of carbon sinks in target sea areas after typhoon passage, generating a typhoon-carbon sink sensitivity zone delineation map, specifically including:

[0044] Obtain the spatial map of typhoon impact, extract the seawater carbon dioxide partial pressure reconstruction field and sea surface wind field data from the typhoon impact spatial map, obtain sea surface water temperature and salinity data from the marine situation monitoring dataset to calculate the solubility of carbon dioxide in seawater, and obtain the daily spatial distribution field of sea-air interface carbon dioxide flux during the typhoon's passage by combining the sea-air carbon dioxide flux calculation rules.

[0045] Based on the calculated spatial distribution field of carbon dioxide exchange flux at the air-sea interface, the average carbon flux values ​​in the target sea area and various characteristic water masses are statistically analyzed in three periods: before, during, and after the passage of the typhoon. The amplitude and duration of flux changes caused by the typhoon are quantified through difference analysis, and the areas where carbon sinks are converted into carbon sources due to the typhoon are identified, generating information on carbon flux changes.

[0046] The distribution map of typhoon characteristic water masses is extracted from the spatial map of typhoon impact. The carbon flux change information is associated with the carbon flux change information and the characteristic parameters of carbon flux change in different water mass regions are extracted, including the carbon flux change amplitude, change duration and recovery rate. The carbon flux change of each water mass during the typhoon is calculated and the differentiated response of carbon sink function of water masses with different properties during the typhoon is analyzed to generate water mass carbon sink response analysis information.

[0047] The climatological background field was obtained, and the recovery status of seawater carbon dioxide partial pressure in the target sea area after the typhoon was analyzed. The degree of recovery of carbon sink function was assessed by comparing the difference between seawater carbon dioxide partial pressure value after the typhoon and the climatological background value. The recovery speed and recovery period of seawater carbon dioxide partial pressure value were calculated by time series trend and used as carbon sink stability assessment indicators.

[0048] Combining information on carbon flux changes, water mass carbon sink response analysis, and carbon sink stability assessment indicators, the K-means clustering algorithm was used to divide the target sea area into sensitivity zones. The carbon sink stability assessment indicators were used as clustering features. Based on the response intensity and recovery capacity of each region to the impact of typhoons, the sea area was divided into three levels: high-sensitivity zone, medium-sensitivity zone, and low-sensitivity zone, generating a typhoon-carbon sink sensitivity zone division map to characterize the impact of typhoons on the target sea area.

[0049] A second aspect of the present invention provides a typhoon impact differentiation system based on seawater CO2 partial pressure data. The system includes a memory, a processor, and a communication interface. The memory contains a typhoon impact differentiation method program based on seawater CO2 partial pressure data. When executed by the processor, the typhoon impact differentiation method program based on seawater CO2 partial pressure data performs the following steps:

[0050] Acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climatological background field based on historical marine situation data of the target sea area.

[0051] Based on the aforementioned marine situation monitoring dataset, a clustering algorithm was used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and conduct correlation analysis and significance tests on each water mass to obtain the distribution information of typhoon characteristic water masses.

[0052] The anomaly field of seawater carbon dioxide partial pressure during the passage of the typhoon is calculated based on the climatological background field. The time series sequence of typhoon impact index is constructed through the anomaly field of seawater carbon dioxide partial pressure and the typhoon impact trace path is generated.

[0053] The grid points of seawater carbon dioxide partial pressure in the target sea area are reconstructed using the sea area situation monitoring dataset. A spatial map of typhoon impact is constructed by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information.

[0054] Based on the spatial map of typhoon impact, the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the passage of typhoons are analyzed, and the carbon sink stability of the target sea area after the passage of typhoons is analyzed, generating a map of typhoon-carbon sink sensitive areas.

[0055] A third aspect of the present invention provides a computer-readable storage medium comprising a typhoon impact differentiation method program based on seawater CO2 partial pressure data. When the typhoon impact differentiation method program based on seawater CO2 partial pressure data is executed by a processor, it implements the steps of the typhoon impact differentiation method based on seawater CO2 partial pressure data as described in any of the preceding claims. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments or examples of the present invention, the drawings used in the embodiments or examples will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained according to these drawings without creative effort.

[0057] Figure 1 This is a flowchart of the first method for differentiating typhoon impacts based on seawater CO2 partial pressure data, provided in an embodiment of the present invention.

[0058] Figure 2 This is a flowchart of a second method for differentiating typhoon impacts based on seawater CO2 partial pressure data, provided in an embodiment of the present invention.

[0059] Figure 3 A block diagram of a typhoon impact differentiation system based on seawater CO2 partial pressure data is provided as an embodiment of the present invention.

[0060] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0061] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0063] Figure 1 This is a flowchart of the first method for differentiating typhoon impacts based on seawater CO2 partial pressure data, provided in an embodiment of the present invention.

[0064] like Figure 1As shown, the present invention provides a first method flowchart for a typhoon impact differentiation method based on seawater CO2 partial pressure data, including:

[0065] S102, acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climate background field based on the historical marine situation data of the target sea area.

[0066] S104. Based on the marine situation monitoring dataset, a clustering algorithm is used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and perform correlation analysis and significance test on each water mass to obtain the distribution information of typhoon characteristic water masses.

[0067] S106, Calculate the seawater carbon dioxide partial pressure anomaly field during the typhoon's passage based on the climatological background field, construct the typhoon impact index time series through the seawater carbon dioxide partial pressure anomaly field, and generate the typhoon impact trace path.

[0068] S108, Reconstruct the grid points of seawater carbon dioxide partial pressure in the target sea area using the sea area situation monitoring dataset, and construct a spatial map of typhoon impact by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information.

[0069] S110, based on the typhoon impact spatial map, analyze the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the typhoon's passage, and analyze the carbon sink stability of the target sea area after the typhoon's passage, generating a typhoon-carbon sink sensitive area delineation map.

[0070] Furthermore, in a preferred embodiment of the present invention, the step of acquiring marine situation monitoring data during the passage of a typhoon in the target sea area, preprocessing it to generate a marine situation monitoring dataset, and constructing a climatological background field based on historical marine situation data of the target sea area specifically includes:

[0071] The data obtained from the multi-source observation platform include sea surface wind field data, rainfall data, sea surface temperature data (SST), sea surface chlorophyll data, and sea surface height data retrieved from satellite remote sensing, as well as seawater carbon dioxide partial pressure, salinity, and dissolved oxygen data collected by the on-site observation platforms of buoys and monitoring vessels.

[0072] The acquired marine situation monitoring data is preprocessed, and a threshold screening method is used to remove outliers caused by instrument errors. The spatiotemporal consistency test method is used to identify and correct missing or distorted data.

[0073] Subsequently, all data were resampled to a preset standard spatial grid coordinate system, spatially normalized using a bilinear interpolation algorithm, and averaged daily according to Coordinated Universal Time to generate a marine situation monitoring dataset.

[0074] By introducing a big data network, historical sea area situation data of the target sea area during the same period of typhoon occurrence is obtained through the big data network to generate a historical sea area situation grid map. At each identical spatial grid point, the arithmetic mean of historical sea area situation data for the same month over many years is calculated as the monthly climatological mean, and the standard deviation is calculated as a measure of climate variability to generate a climatological background field, including the mean field and the standard deviation field.

[0075] It should be noted that the comprehensive marine situation monitoring data acquired from the multi-source observation platform system during the typhoon's passage mainly includes two categories: large-scale macroscopic data retrieved from satellite remote sensing and localized, detailed data collected by on-site observation platforms. Satellite data covers parameters such as sea surface wind field, rainfall, sea surface temperature (SST), chlorophyll concentration, and sea surface height, providing large-scale, continuous spatial coverage. On-site platforms such as buoys and monitoring vessels directly measure key biogeochemical parameters such as seawater carbon dioxide partial pressure, salinity, and dissolved oxygen, providing in-situ verification and high-precision vertical profile information. After acquiring the raw data, a dynamic threshold filtering method is used. Based on the characteristics of different sensors and the historical distribution range of parameters, thresholds are set to automatically identify and remove abnormal observations caused by instrument errors or transmission interference. Subsequently, a spatiotemporal consistency verification algorithm is used to identify data anomalies or missing data caused by cloud cover, signal interruption, etc., by comparing the data values ​​and trends of adjacent spatiotemporal grid points. Interpolation methods are then used to repair and fill in the data, thereby ensuring the reliability and accuracy of the data. Subsequently, a standard spatial grid coordinate system is established, typically using an equal latitude and longitude grid system. After completing data quality control, all source data are uniformly resampled into this standard grid coordinate system using a bilinear interpolation algorithm, achieving spatial normalization of data from different sources and at different resolutions. Simultaneously, all observation data are time-standardized to Coordinated Universal Time (UTC), and an arithmetic mean or weighted average method is used to generate a marine situation monitoring dataset. A climatological background field is further constructed to provide an analytical benchmark. Through international marine data exchange networks and domestic marine observation databases, long-sequence marine situation monitoring data for the target sea area during historical typhoon seasons (same season or same month) are obtained. These historical data undergo the same quality control, outlier removal, and spatiotemporal resampling processing methods as the real-time data to ensure data standard consistency. At each spatial grid point, the arithmetic mean of the historical data of each parameter over many years is calculated to generate a monthly climatological mean field; at the same time, its standard deviation is calculated as a measure of climate variability, and finally a climatological background field containing the mean field and the standard deviation field is formed, which characterizes the normal state range of the sea area under the influence of extreme weather events such as typhoons, and provides a scientific benchmark for accurately identifying abnormal changes caused by typhoons.

[0076] Furthermore, in a preferred embodiment of the present invention, the water mass segmentation based on the marine situation monitoring dataset using a clustering algorithm is performed to identify water masses of different properties formed under the influence of typhoons, and correlation analysis and significance testing are conducted on each water mass to obtain typhoon characteristic water mass distribution information, specifically including:

[0077] Obtain a marine situation monitoring dataset, generate a marine situation grid map of the target sea area during the passage of a typhoon based on the marine situation monitoring dataset, extract the marine situation features corresponding to each grid in the marine situation grid map, and generate a grid feature matrix after standardization processing.

[0078] The grid feature matrix is ​​used as the input of principal component analysis. The covariance matrix between each feature parameter is calculated. The principal component direction is obtained through eigenvalue decomposition. The top k principal components whose cumulative contribution rate reaches the preset contribution rate threshold are selected. The original high-dimensional feature vector is projected onto the principal component space to generate the multimodal marine situation feature vector of each grid unit.

[0079] The multimodal marine situation feature vectors corresponding to each grid cell are abstracted into individual grid nodes. The DBSCAN clustering algorithm is introduced to classify the nodes. The neighborhood radius ε and the minimum number of points MinPts of the DBSCAN clustering algorithm are set. For each grid node, each grid node is marked as unvisited to complete the initialization. A grid node is randomly selected for visit. The grid node that enters the visited state is marked as grid node p for cluster expansion.

[0080] For a grid node p, calculate its Euclidean distance to all other nodes in the feature space, and find all neighboring nodes whose distance is less than a preset distance threshold. If the number of neighboring nodes is not less than MinPts, then mark node p as a core point and create a new cluster, adding the node and all its neighboring nodes to the cluster. If the number of neighboring nodes is less than MinPts, then mark node p as a noise point.

[0081] For each neighboring node added to the cluster, recursively perform neighborhood query and expansion operations. If the node is a core node, continue to add the unvisited nodes in its neighborhood to the current cluster. Repeat this process until no new nodes are added to the current cluster, completing the division of a water cluster category. Traverse all unvisited nodes and repeat the above process until all nodes are visited, obtaining the final set of clusters.

[0082] It should be noted that a marine situation grid map of the target sea area during the typhoon's passage was generated based on a marine situation monitoring dataset. This grid map is based on a standardized spatial grid, with each grid cell containing multi-dimensional environmental parameters, such as synchronously observed data on sea surface temperature, salinity, chlorophyll concentration, seawater carbon dioxide partial pressure, and dissolved oxygen. By extracting the multi-parameter feature vector corresponding to each grid cell, the Z-score normalization method is used to eliminate the dimensional differences between different parameters, ensuring that all feature parameters are of the same order of magnitude, ultimately forming a standardized grid feature matrix.

[0083] The standardized grid feature matrix is ​​used as input data for principal component analysis (PCA). The covariance matrix between each feature parameter is calculated, and eigenvalues ​​and corresponding eigenvectors are obtained through eigenvalue decomposition to determine the principal component directions. Based on a preset cumulative contribution rate threshold (usually above 85%), the top k principal components are selected, and the original high-dimensional feature vectors are projected onto the principal component space to achieve data dimensionality reduction. Most of the variation information of the original data is retained, redundancy between features is eliminated, and low-dimensional multimodal marine situation feature vectors for each grid cell are generated. The dimensionality-reduced multimodal marine situation feature vectors are abstracted as grid nodes in the feature space, and the density-based DBSCAN clustering algorithm is introduced for water mass classification. Two key parameters of the algorithm are set based on the data distribution characteristics: neighborhood radius ε and minimum number of points MinPts. In the initialization phase, all grid nodes are marked as unvisited, and a starting node p is randomly selected for visit. The Euclidean distance between p and all other nodes in the feature space is calculated, and the set of neighboring nodes with a distance less than ε is identified. If the number of neighboring nodes reaches or exceeds MinPts, then node p is marked as a core point and a new cluster is created. At the same time, the node and all its neighboring nodes are added to the current cluster. If the number of neighboring nodes is insufficient, then it is marked as a noise point.

[0084] For each neighboring node already added to a cluster, recursively perform neighborhood queries and expansion operations. For each newly added node, if it is also a core node, its unvisited neighboring nodes are added to the current cluster. Through this density-connected expansion method, new nodes are continuously absorbed into the cluster until no more new nodes are added, completing the classification of a water mass category. The above cluster expansion process is repeated for all unvisited nodes until all nodes are visited, ultimately resulting in a set of clusters containing multiple water mass categories. Each cluster represents a water mass type with specific environmental characteristics, enabling automatic identification and classification of different water masses under typhoon influence. This provides a reliable technical means for in-depth research on the impact mechanisms of typhoons on the marine environment, clearly reflects the spatial distribution characteristics of water masses with different properties formed after a typhoon passes, and provides a scientific basis for marine environmental monitoring and disaster assessment.

[0085] Furthermore, in a preferred embodiment of the present invention, the step of using a clustering algorithm to divide water masses based on the marine situation monitoring dataset, identifying water masses of different properties formed under the influence of typhoons, and performing correlation analysis and significance tests on each water mass to obtain typhoon characteristic water mass distribution information, further includes:

[0086] Based on the final cluster set, several water masses are generated. The mean, standard deviation, skewness and kurtosis of the situation characteristics of each sea area within each water mass are calculated to generate a water mass feature statistics table. Combined with the preset water mass category labels, the water mass categories are divided to generate initial water mass category classification information.

[0087] Based on the initial water mass category classification information, the Mann-Whitney U test method was used to test the significance of parameter differences between different water mass categories, calculate the significance probability value of each pair of water masses on each parameter, and screen out water mass pairs with statistically significant differences.

[0088] Meanwhile, within each water mass, the Pearson correlation coefficient between seawater carbon dioxide partial pressure and other parameters is calculated to assess the strength of the linear correlation between parameters. Water masses that do not have a linear correlation are separated, and finally, water mass category classification information is obtained.

[0089] Using the water mass category classification information, the Delaunay triangulation algorithm is used to construct a triangular mesh for the spatial sample points contained in each water mass category. Then, the convex hull boundary of each water mass point set is calculated to generate the spatial distribution polygon of each water mass. The boundary polygon is smoothed to eliminate irregular jagged edges, and a spatial distribution map of the water mass is obtained.

[0090] The spatial characteristics of the identified typhoon-affected water masses are calculated using the aforementioned water mass spatial distribution map. The distribution information of typhoon-characteristic water masses is then generated by combining the corresponding sea area situation characteristics, significance test results, and correlation analysis results.

[0091] It should be noted that several water masses are generated based on the final cluster set, and characteristic statistical analysis is performed on the water masses represented by each cluster. The mean, standard deviation, skewness, and kurtosis of the marine situational characteristic parameters (including sea surface temperature, salinity, chlorophyll concentration, seawater carbon dioxide partial pressure, dissolved oxygen, etc.) within each water mass are calculated. These statistics provide a comprehensive understanding of the central tendency, dispersion, and distribution characteristics of each water mass. These statistics are compiled into a water mass characteristic statistical table, and combined with pre-defined water mass category labels (such as water mass classification standards based on temperature-salinity characteristics), preliminary category divisions are performed for each water mass, generating initial water mass category classification information. Based on the initial water mass category division, the Mann-Whitney U test is used to test the significance of parameter differences between different water mass categories. This non-parametric statistical method is suitable for non-normally distributed data and can effectively assess the significance of differences between pairs of water masses in various environmental parameters. By calculating the significance probability values ​​for each parameter, water mass pairs with statistically significant differences are selected, thereby verifying the rationality of the initial classification. Simultaneously, Pearson correlation coefficients were calculated between seawater carbon dioxide partial pressure and other environmental parameters within each water mass. By assessing the strength of linear correlations between parameters, the consistency of characteristics within the water mass was further verified. Water masses lacking significant linear correlations between their internal parameters were separated or reclassified, ultimately yielding water mass category information.

[0092] Based on the determined water mass category classification information, the Delaunay triangulation algorithm is used to construct triangular meshes for the spatial sample points contained in each water mass category, forming a triangular mesh by connecting adjacent sample points. On this basis, the convex hull boundary of each water mass point set is calculated, generating spatial distribution polygons for each water mass. Smoothing methods such as spline curve interpolation are applied to the initially obtained boundary polygons to eliminate irregular jagged boundaries, resulting in a smooth and geographically significant spatial distribution map of water masses. Using the obtained spatial distribution map of water masses, spatial characteristics of the identified typhoon-affected water masses are calculated, including morphological indicators such as area, perimeter, and spatial compactness of each water mass. Combining the corresponding sea area situation characteristics, significance test results, and correlation analysis results, the characteristic differences in the response of different water masses to typhoons are comprehensively analyzed. Finally, typhoon characteristic water mass distribution information containing multi-dimensional information such as spatial distribution, characteristic statistics, and significant differences is generated.

[0093] Furthermore, in a preferred embodiment of the present invention, the step of calculating the seawater carbon dioxide partial pressure anomaly field during the typhoon's passage based on the climatological background field, constructing a typhoon impact index time series using the seawater carbon dioxide partial pressure anomaly field, and generating a typhoon impact trace path specifically includes:

[0094] Obtain the climatological background field, extract the monthly climatological mean field from the climatological background field as the reference data, obtain the marine situation monitoring dataset and extract the daily seawater carbon dioxide partial pressure grid data during the typhoon's passage.

[0095] The climate anomaly algorithm is used to calculate the difference between the daily observation value and the monthly climatological mean value of each spatial grid point, thereby generating the daily seawater carbon dioxide partial pressure anomaly value of each grid point, which covers the entire target sea area during the typhoon.

[0096] The distribution information of characteristic water masses of typhoons is obtained, and the spatial distribution range of the core water mass affected by typhoons is extracted as an analysis mask. Within this spatial mask, multi-dimensional features are extracted from the daily seawater carbon dioxide partial pressure anomaly field. The arithmetic mean of the outlier values ​​of all grid points is used as the anomaly intensity feature, the standard deviation of the outlier values ​​of all grid points is used as the spatial heterogeneity feature, and the proportion of grid points whose absolute outlier values ​​exceed the preset threshold is used as the significant influence range feature.

[0097] The extracted multi-dimensional features are normalized, and then the weight coefficients of each feature are determined by a subjective and objective weighting method. The objective weights are calculated by the entropy weight method based on the degree of variation of each feature value, and the subjective weights are determined by the analytic hierarchy process combined with expert scoring. Finally, the subjective and objective weights are weighted and fused to obtain the comprehensive weight. The normalized multi-dimensional features are weighted and summed to calculate the daily typhoon impact index value, and a continuous time series of typhoon impact indexes throughout the entire life cycle of a typhoon is constructed.

[0098] Based on the anomaly field of carbon dioxide partial pressure in seawater, an image segmentation algorithm based on Otsu's maximum inter-class variance method is used to identify and segment the anomaly region. The anomaly field is binarized into significant anomaly regions and non-significant regions. Morphological operations are used to optimize the segmentation results, remove small noise regions and fill holes, and obtain the spatial distribution map of the typhoon core impact area.

[0099] Based on the spatial distribution map of the typhoon's core impact area, the coordinates of the geometric center point of the polygon in the impact area are calculated. The centroid algorithm is used to obtain the center point position by weighted averaging of the polygon vertex coordinates. The center points of each day during the typhoon's passage are connected by spline curves in chronological order to generate the typhoon impact trace path. At the same time, the date corresponding to each center point and the typhoon impact index value at that time are recorded.

[0100] It should be noted that, based on the climatological background field, a monthly climatological mean field was extracted as the baseline data. Simultaneously, daily seawater carbon dioxide partial pressure gridded data during the typhoon's passage were obtained from the marine situation monitoring dataset. Using a climatological anomaly algorithm, the daily observation value at each spatial grid point was interpolated with the corresponding monthly climatological mean to calculate the daily seawater carbon dioxide partial pressure anomaly value for each grid point. This generated a daily seawater carbon dioxide partial pressure anomaly field data cube covering the entire target sea area during the typhoon's passage. This data cube comprehensively records the spatiotemporal variation characteristics of seawater carbon dioxide partial pressure relative to the climatological mean under the influence of the typhoon. Based on the anomaly field data cube, and combined with the distribution information of characteristic water masses affected by the typhoon, the spatial distribution range of the core water masses affected by the typhoon was extracted as an analysis mask. Within the spatial mask area, multi-dimensional features are extracted from the daily seawater carbon dioxide partial pressure anomaly field. The arithmetic mean of the anomaly values ​​of all grid points is calculated as the anomaly intensity feature, representing the overall level of the anomaly on that day. The standard deviation of the anomaly values ​​of all grid points is calculated as the spatial heterogeneity feature, representing the degree of uniformity of the spatial distribution of the anomaly. The proportion of grid points whose absolute anomaly values ​​exceed a preset threshold is statistically analyzed as the significant influence range feature, quantifying the spatial range of the anomaly's influence, thereby characterizing the spatiotemporal characteristics of the anomaly field.

[0101] After normalizing the extracted multi-dimensional features, a combined subjective and objective weighting method was used to determine the weight coefficients of each feature. Objective weights were calculated using the entropy weighting method, automatically determining the weight based on the degree of variation of each feature value; features with greater variation were assigned higher weights. Subjective weights were determined using the analytic hierarchy process (AHP) combined with expert scoring. The subjective and objective weights were weighted and fused to obtain a comprehensive weight. The normalized multi-dimensional features were then weighted and summed to calculate the daily typhoon impact index, thus constructing a continuous time-series sequence of typhoon impact indices throughout the typhoon's lifecycle. Subsequently, based on the seawater carbon dioxide partial pressure anomaly field, an image segmentation algorithm based on Otsu's maximum inter-class variance method was used for anomaly region identification and segmentation. This algorithm automatically determined the optimal segmentation threshold, binarizing the anomaly field into significant and non-significant anomaly regions. Morphological opening and closing operations were used to optimize the segmentation results, removing small noise areas and filling holes to obtain a continuous and complete spatial distribution map of the typhoon core impact area. Based on the spatial distribution map of the typhoon core impact area, the coordinates of the geometric center point of the polygon in the impact area were calculated. The centroid algorithm was used to obtain the center point position by weighted averaging of the polygon vertex coordinates. Connect the daily center points during the typhoon's passage in chronological order using spline curves to generate a smooth typhoon impact trace path. Simultaneously, record the date corresponding to each center point and the typhoon impact index value at that time to form a complete record of the typhoon's impact trajectory.

[0102] Furthermore, in a preferred embodiment of the present invention, the step of reconstructing the grid points of seawater carbon dioxide partial pressure in the target sea area using the sea area situation monitoring dataset, and constructing a spatial map of typhoon impact by combining the typhoon impact index time series, typhoon impact trace paths, and typhoon characteristic water mass distribution information, specifically includes:

[0103] A marine situation monitoring dataset is obtained. Based on the marine situation monitoring dataset, marine situation monitoring features are extracted as feature variables. Seawater carbon dioxide partial pressure data observed in the field are used as target variables. A seawater carbon dioxide partial pressure reconstruction model is constructed using a random forest regression algorithm.

[0104] The target sea area is divided into several spatial grids. The corresponding environmental feature vectors are extracted from each grid point using the sea area situation monitoring dataset. The vectors are then input into the trained seawater carbon dioxide partial pressure reconstruction model to calculate the estimated seawater carbon dioxide partial pressure at the grid points, thereby generating a seawater carbon dioxide partial pressure reconstruction field covering the entire target sea area.

[0105] The typhoon impact index time series and typhoon impact trace path are obtained. After smoothing the typhoon impact index time series, the key feature points of the typhoon impact intensity are identified by the local extremum detection algorithm, including the starting point, peak point and attenuation point. The morphological features of the sequence are extracted by the time series segmentation aggregation approximation method. Combined with the time axis information of the typhoon passage, a time series feature map reflecting the evolution process of the typhoon impact intensity is generated.

[0106] Based on the latitude and longitude coordinates of the center point of the daily impact area and the corresponding typhoon impact index value extracted from the typhoon impact trace path, the inverse distance weighted interpolation algorithm is used for spatial interpolation to transform the discrete point-like impact intensity data into a continuous spatial distribution field, thus obtaining the spatial distribution field of typhoon impact intensity.

[0107] Spatial registration was performed on the seawater carbon dioxide partial pressure reconstruction field, the spatial distribution field of typhoon impact intensity, and the temporal characteristic map. Spatial correlation between different sea area situation characteristics was established through raster calculation and vector overlay to generate an initial typhoon impact spatial map.

[0108] The distribution information of characteristic water masses of typhoons is obtained and superimposed on the initial typhoon impact spatial map. The boundary attributes of water masses are correlated with the partial pressure field of carbon dioxide in the seawater. The degree of abnormal change caused by the typhoon is obtained by difference calculation using the climatological background field as a reference benchmark. The carbon dioxide partial pressure distribution is rendered with gradient colors using visualization technology, the influence intensity is represented by contour lines, and the water mass boundary is marked with preset rule symbols. Finally, the typhoon impact spatial map is obtained.

[0109] It should be noted that multi-dimensional marine situation monitoring features, including sea surface temperature, salinity, chlorophyll concentration, and sea surface height, were extracted from the marine situation monitoring dataset and used as input variables for the machine learning model. Simultaneously, field-measured seawater carbon dioxide partial pressure data were used as the target variable. A random forest regression algorithm was employed to construct a seawater carbon dioxide partial pressure reconstruction model. This model, through an ensemble learning strategy that integrates the prediction results of multiple decision trees, effectively captures complex nonlinear relationships and ranks the importance of features, thereby establishing a reliable mapping relationship from easily observed parameters to the difficult-to-obtain carbon dioxide partial pressure, providing a technical foundation for subsequent large-scale marine carbon distribution research. Subsequently, the target marine area was divided into several spatial grids. Environmental feature vectors were extracted from each grid point using the marine situation monitoring dataset and input into the well-trained seawater carbon dioxide partial pressure reconstruction model. The estimated seawater carbon dioxide partial pressure at each grid point was calculated, generating a continuous spatially distributed seawater carbon dioxide partial pressure reconstruction field covering the entire target marine area. This reconstruction field retains the accuracy of field observation data while possessing the advantage of large-scale coverage from remote sensing data, achieving a spatialized representation of seawater carbon parameters combining point and surface data.

[0110] Furthermore, the time series of typhoon impact index and typhoon impact trace paths are obtained. The time series of typhoon impact index is processed by moving average filtering to eliminate high-frequency fluctuations. A local extremum detection algorithm is used to identify key feature points in the evolution of typhoon impact intensity, including the impact initiation point, intensity peak point, and attenuation inflection point. A time series segmentation and aggregation approximation method is used to extract the global morphological features of the sequence. Combined with the time axis information of typhoon passage, a time series feature map reflecting the dynamic evolution of typhoon impact intensity is generated, providing a quantitative basis for the stage division of the typhoon life cycle. Based on the typhoon impact trace paths, the precise latitude and longitude coordinates of the daily impact area center points and the corresponding typhoon impact index values ​​are extracted. An inverse distance weighted interpolation algorithm is used for spatial interpolation. This algorithm assigns different weights according to spatial location relationships, with closer points receiving greater weights. This transforms discrete point-like impact intensity data into a continuous spatial distribution field, resulting in a high-resolution spatial distribution field of typhoon impact intensity, intuitively displaying the spatial gradient characteristics of typhoon impact.

[0111] Subsequently, the reconstructed seawater carbon dioxide partial pressure field, the spatial distribution field of typhoon impact intensity, and the temporal characteristic map were precisely spatially registered. Spatial correlations between different sea area characteristics were established through raster calculations and vector overlay operations. Map algebra methods were used to fuse and analyze multi-source data, generating an initial spatial map of typhoon impact and preliminarily integrating multi-dimensional characteristic information of typhoon impact. Finally, the distribution information of characteristic water masses caused by typhoons was obtained, and the water mass boundary attributes were spatially correlated with the seawater carbon dioxide partial pressure field. Using the climatological background field as a reference, the degree of anomalous changes caused by typhoons was obtained through difference calculations. Advanced visualization techniques were employed, using gradient colors to render the carbon dioxide partial pressure distribution, isolines to represent the impact intensity gradient, and standard symbols according to preset rules to mark water mass boundaries, ultimately generating a spatial map of typhoon impact. This visually demonstrates the spatial pattern of typhoon impact and provides important analytical tools and decision-making basis for in-depth research on the impact mechanism of typhoons on the marine carbon cycle.

[0112] Figure 2 This is a flowchart of a second method for differentiating typhoon impacts based on seawater CO2 partial pressure data, provided in an embodiment of the present invention.

[0113] like Figure 2 As shown, the present invention provides a second method flowchart for differentiating typhoon impacts based on seawater CO2 partial pressure data, including:

[0114] S202, obtain the spatial map of typhoon impact, extract the seawater carbon dioxide partial pressure reconstruction field and sea surface wind field data through the typhoon impact spatial map, obtain sea surface water temperature and salinity data from the marine situation monitoring dataset to calculate the solubility of carbon dioxide in seawater, and obtain the daily spatial distribution field of sea-air interface carbon dioxide flux during the typhoon's passage by combining the sea-air carbon dioxide flux calculation rules.

[0115] S204, based on the calculated spatial distribution field of carbon dioxide exchange flux at the air-sea interface, statistically analyzes the average carbon flux values ​​in the target sea area and various characteristic water masses during three periods: before, during, and after the passage of a typhoon; quantifies the magnitude and duration of flux changes caused by the typhoon through difference analysis, identifies areas where carbon sinks are converted into carbon sources due to the typhoon, and generates information on carbon flux changes.

[0116] S206, extract the distribution map of typhoon characteristic water masses from the typhoon impact spatial map, associate it with the carbon flux change information and extract the carbon flux change characteristic parameters in different water mass regions, including the carbon flux change amplitude, change duration and recovery rate, calculate the carbon flux change of each water mass during the typhoon and analyze the differentiated response of the carbon sink function of water masses with different properties during the typhoon, and generate water mass carbon sink response analysis information;

[0117] S208, extract the distribution map of typhoon characteristic water masses from the typhoon impact spatial map, associate it with the carbon flux change information and extract the carbon flux change characteristic parameters in different water mass regions, including the carbon flux change amplitude, change duration and recovery rate, calculate the carbon flux change of each water mass during the typhoon and analyze the differentiated response of carbon sink function of water masses with different properties during the typhoon, and generate water mass carbon sink response analysis information;

[0118] S210 combines information on carbon flux changes, water mass carbon sink response analysis, and carbon sink stability assessment indicators. It uses the K-means clustering algorithm to divide the target sea area into sensitive zones, uses the carbon sink stability assessment indicators as clustering features, and divides the sea area into three levels: high-sensitivity zone, medium-sensitivity zone, and low-sensitivity zone according to the response intensity and recovery capacity of each region to the impact of typhoons. It generates a typhoon-carbon sink sensitivity zone division map to characterize the impact of typhoons on the target sea area.

[0119] It should be noted that the reconstruction field of seawater carbon dioxide partial pressure and sea surface wind field data were extracted from the spatial map of typhoon impacts, while sea surface water temperature and salinity data were obtained from the marine situation monitoring dataset. Using the sea surface water temperature and salinity data, the solubility coefficient of carbon dioxide in seawater was calculated using thermodynamic equations, and the gas transport rate was calculated by combining the sea surface wind field data. Combining the solubility coefficient, gas transport rate, and sea-air carbon dioxide partial pressure difference, and based on the sea-air exchange flux calculation formula, the daily sea-air interface carbon dioxide flux during the typhoon's passage was calculated grid-by-grid, generating a complete flux spatial distribution field. Based on the calculated spatial distribution field of sea-air interface carbon dioxide exchange flux, the average carbon flux values ​​for the target sea area as a whole and within each characteristic water mass region were statistically analyzed during three typical periods: before, during, and after the typhoon's passage. By comparing and analyzing the flux data at different times, the magnitude and duration of flux changes were calculated, identifying areas where the carbon sink function was significantly weakened or even transformed from a carbon sink to a carbon source due to the typhoon's impact, quantifying the actual impact of the typhoon on the marine carbon sink function, and generating information on carbon flux changes.

[0120] Subsequently, the distribution information of characteristic water masses affected by typhoons was extracted from the spatial map of typhoon impacts and spatially correlated with information on carbon flux changes. For different water mass regions, key characteristic parameters of carbon flux changes were extracted, including the magnitude of change, duration of change, and recovery rate. By comparing and analyzing the carbon flux response characteristics of water masses with different properties during typhoon passage, the differentiated response mechanisms of water bodies with varying physical properties to typhoon impacts were revealed, generating water mass carbon sink response analysis information and providing an analytical basis for understanding typhoon-ocean interactions. Using the climatological background field as a benchmark, the recovery status of seawater carbon dioxide partial pressure in the target sea area after typhoon passage was analyzed. By comparing the differences between seawater carbon dioxide partial pressure values ​​and climatological background values ​​at different times after the typhoon, the degree of recovery of carbon sink function was assessed. Simultaneously, the recovery rate and recovery period of seawater carbon dioxide partial pressure values ​​were calculated through time series analysis, establishing a carbon sink stability assessment index system to provide a quantitative basis for comprehensively evaluating the persistent effects of typhoon impacts. Finally, by integrating information on carbon flux changes, water mass carbon sink response analysis, and carbon sink stability assessment indicators, the K-means clustering algorithm was used to partition the target sea area into sensitive zones. Stability indicators such as response intensity and resilience of each region were used as clustering features, and the optimal cluster centers were determined through iterative calculations. Based on the clustering results, the sea area was divided into three levels: high-sensitivity, medium-sensitivity, and low-sensitivity zones, generating a typhoon-carbon sink sensitivity zone partition map. This clearly demonstrates the differences in vulnerability to typhoon impacts across different areas of the target sea area, providing a scientific basis for marine carbon sink protection and management, and offering important technical support for in-depth research on the impact mechanisms of typhoons on the marine carbon cycle.

[0121] It is worth noting that sea-air CO2 flux (F) is the prerequisite and foundation for the ocean's regulation of atmospheric CO2 levels. Sea-air CO2 flux refers to the net amount of CO2 exchanged per unit area per unit time at the atmosphere-ocean interface, representing the sea-air CO2 exchange rate. Based on the liquid film diffusion model, the formula for estimating sea-air CO2 flux is:

[0122] ,

[0123] In the formula, F represents the CO2 flux at the air-sea interface. A negative F indicates that the seawater absorbs CO2 from the atmosphere, while a positive F indicates that the seawater releases CO2 into the atmosphere, in which case the seawater acts as a carbon sink (absorbing carbon from the atmosphere). A negative F indicates that the atmosphere emits CO2 into the seawater, in which case the seawater acts as a carbon source. K represents the CO2 transport coefficient, which is the product of the CO2 transport rate k and its solubility as in seawater, i.e., K = k × as. ΔpCO2 represents the CO2 partial pressure difference at the air-sea interface (i.e., pCO2,sw – pCO2, air).

[0124] Furthermore, the typhoon impact differentiation method based on seawater CO2 partial pressure data provided by this invention also includes the following steps:

[0125] After the typhoon passes through the target sea area, the marine situation recovery monitoring data of the target sea area during the recovery period is obtained. The DBSCAN clustering algorithm is used to identify the water mass pattern in the sea area at this time and to perform correlation analysis and significance test to generate characteristic water mass distribution information after the typhoon.

[0126] The distribution information of characteristic water masses during typhoons is obtained and compared with the distribution information of characteristic water masses after typhoons in time and space. The duration, spatial range change rate and decay rate of the core hydration parameters of the corresponding water masses are calculated. The duration and evolution law of water mass disturbance caused by typhoons are quantified, and a water mass disturbance evolution feature dataset is generated.

[0127] Based on the data set of water mass disturbance evolution characteristics, the changing trend of seawater carbon dioxide in each water mass recovering to the climatological background value over time was analyzed. The recovery trajectory was fitted by linear regression and the recovery half-life and final stable value were calculated to generate recovery process analysis information.

[0128] By correlating recovery process analysis information with changes in air-sea CO2 flux, the duration of the impact of typhoon events on regional carbon sink intensity and the time scale required for carbon sink function to recover to baseline levels are calculated, generating a carbon sink disturbance persistence assessment report.

[0129] By using big data retrieval, historical typhoon events in the target sea area are obtained, and a dataset of water mass disturbance evolution characteristics and carbon sink disturbance persistence assessment report for each historical typhoon event are generated to construct a typhoon-carbon sink response case library.

[0130] A typhoon-carbon sink impact prediction model is constructed based on an LSTM network. The model is trained using a training dataset built from the typhoon-carbon sink response case library. Typhoon characteristics and the characteristics of the initial water mass are used as inputs, and the magnitude and duration of carbon sink disturbances are used as outputs to establish a quantitative relationship between typhoon attributes and carbon cycle climate effects.

[0131] By repeatedly training until a typhoon-carbon sink impact prediction model that meets expectations is output, the cumulative impact of the regional marine carbon sink function during the next typhoon is simulated using the trained typhoon-carbon sink impact prediction model, and a regional impact prediction report is generated for management assistance.

[0132] Figure 3 A typhoon impact differentiation system 3 based on seawater CO2 partial pressure data is provided as an embodiment of the present invention. The system includes: a memory 301, a processor 302, and a communication interface 303. The memory 301 contains a typhoon impact differentiation method program based on seawater CO2 partial pressure data. When the processor 302 executes the typhoon impact differentiation method program based on seawater CO2 partial pressure data, it performs the following steps:

[0133] Acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climatological background field based on historical marine situation data of the target sea area.

[0134] Based on the aforementioned marine situation monitoring dataset, a clustering algorithm was used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and conduct correlation analysis and significance tests on each water mass to obtain the distribution information of typhoon characteristic water masses.

[0135] The anomaly field of seawater carbon dioxide partial pressure during the passage of the typhoon is calculated based on the climatological background field. The time series sequence of typhoon impact index is constructed through the anomaly field of seawater carbon dioxide partial pressure and the typhoon impact trace path is generated.

[0136] The grid points of seawater carbon dioxide partial pressure in the target sea area are reconstructed using the sea area situation monitoring dataset. A spatial map of typhoon impact is constructed by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information.

[0137] Based on the spatial map of typhoon impact, the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the passage of typhoons are analyzed, and the carbon sink stability of the target sea area after the passage of typhoons is analyzed, generating a map of typhoon-carbon sink sensitive areas.

[0138] In another aspect, the present invention provides a computer-readable storage medium including a typhoon impact differentiation method program based on seawater CO2 partial pressure data. When the typhoon impact differentiation method program based on seawater CO2 partial pressure data is executed by a processor, it implements the steps of the typhoon impact differentiation method based on seawater CO2 partial pressure data as described in any of the preceding claims.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0140] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0141] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0142] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0143] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0144] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A typhoon influence area division method based on seawater CO2 partial pressure data, characterized in that, include: Acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climatological background field based on historical marine situation data of the target sea area. Based on the aforementioned marine situation monitoring dataset, a clustering algorithm was used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and conduct correlation analysis and significance tests on each water mass to obtain the distribution information of typhoon characteristic water masses. The anomaly field of seawater carbon dioxide partial pressure during the passage of the typhoon is calculated based on the climatological background field. The time series sequence of typhoon impact index is constructed through the anomaly field of seawater carbon dioxide partial pressure and the typhoon impact trace path is generated. The grid points of seawater carbon dioxide partial pressure in the target sea area are reconstructed using the sea area situation monitoring dataset. A spatial map of typhoon impact is constructed by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information. Based on the spatial map of typhoon impact, the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the passage of typhoons are analyzed, and the carbon sink stability of the target sea area after the passage of typhoons is analyzed, generating a map of typhoon-carbon sink sensitive areas.

2. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 1, characterized in that, The process of acquiring marine situation monitoring data during typhoon passage within the target sea area, preprocessing it to generate a marine situation monitoring dataset, and constructing a climatological background field based on historical marine situation data of the target sea area specifically includes: The data obtained from the multi-source observation platform include sea surface wind field data, rainfall data, sea surface temperature data (SST), sea surface chlorophyll data, and sea surface height data retrieved from satellite remote sensing, as well as seawater carbon dioxide partial pressure, salinity, and dissolved oxygen data collected by the on-site observation platforms of buoys and monitoring vessels. The acquired marine situation monitoring data is preprocessed, and a threshold screening method is used to remove outliers caused by instrument errors. The spatiotemporal consistency test method is used to identify and correct missing or distorted data. Subsequently, all data were resampled to a preset standard spatial grid coordinate system, spatially normalized using a bilinear interpolation algorithm, and averaged daily according to Coordinated Universal Time to generate a marine situation monitoring dataset. By introducing a big data network, historical sea area situation data of the target sea area during the same period of typhoon occurrence is obtained through the big data network to generate a historical sea area situation grid map. At each identical spatial grid point, the arithmetic mean of historical sea area situation data for the same month over many years is calculated as the monthly climatological mean, and the standard deviation is calculated as a measure of climate variability to generate a climatological background field, including the mean field and the standard deviation field.

3. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 1, characterized in that, The water mass segmentation based on the marine situation monitoring dataset is performed using a clustering algorithm to identify water masses of different properties formed under the influence of typhoons. Correlation analysis and significance testing are then conducted on each water mass to obtain the distribution information of typhoon-characteristic water masses, specifically including: Obtain a marine situation monitoring dataset, generate a marine situation grid map of the target sea area during the passage of a typhoon based on the marine situation monitoring dataset, extract the marine situation features corresponding to each grid in the marine situation grid map, and generate a grid feature matrix after standardization processing. The grid feature matrix is ​​used as the input of principal component analysis. The covariance matrix between each feature parameter is calculated. The principal component direction is obtained through eigenvalue decomposition. The top k principal components whose cumulative contribution rate reaches the preset contribution rate threshold are selected. The original high-dimensional feature vector is projected onto the principal component space to generate the multimodal marine situation feature vector of each grid unit. The multimodal marine situation feature vectors corresponding to each grid cell are abstracted into individual grid nodes. The DBSCAN clustering algorithm is introduced to classify the nodes. The neighborhood radius ε and the minimum number of points MinPts of the DBSCAN clustering algorithm are set. For each grid node, each grid node is marked as unvisited to complete the initialization. A grid node is randomly selected for visit. The grid node that enters the visited state is marked as grid node p for cluster expansion. For a grid node p, calculate its Euclidean distance to all other nodes in the feature space, and find all neighboring nodes whose distance is less than a preset distance threshold. If the number of neighboring nodes is not less than MinPts, then mark node p as a core point and create a new cluster, adding the node and all its neighboring nodes to the cluster. If the number of neighboring nodes is less than MinPts, then mark node p as a noise point. For each neighboring node added to the cluster, recursively perform neighborhood query and expansion operations. If the node is a core node, continue to add the unvisited nodes in its neighborhood to the current cluster. Repeat this process until no new nodes are added to the current cluster, completing the division of a water cluster category. Traverse all unvisited nodes and repeat the above process until all nodes are visited, obtaining the final set of clusters.

4. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 3, characterized in that, The method involves using a clustering algorithm to segment water masses based on the aforementioned marine situation monitoring dataset, identifying water masses of different properties formed under the influence of typhoons, and performing correlation analysis and significance tests on each water mass to obtain typhoon characteristic water mass distribution information. This also includes: Based on the final cluster set, several water masses are generated. The mean, standard deviation, skewness and kurtosis of the situation characteristics of each sea area within each water mass are calculated to generate a water mass feature statistics table. Combined with the preset water mass category labels, the water mass categories are divided to generate initial water mass category classification information. Based on the initial water mass category classification information, the Mann-Whitney U test method was used to test the significance of parameter differences between different water mass categories, calculate the significance probability value of each pair of water masses on each parameter, and screen out water mass pairs with statistically significant differences. Meanwhile, within each water mass, the Pearson correlation coefficient between seawater carbon dioxide partial pressure and other parameters is calculated to assess the strength of the linear correlation between parameters. Water masses that do not have a linear correlation are separated, and finally, water mass category classification information is obtained. Using the water mass category classification information, the Delaunay triangulation algorithm is used to construct a triangular mesh for the spatial sample points contained in each water mass category. Then, the convex hull boundary of each water mass point set is calculated to generate the spatial distribution polygon of each water mass. The boundary polygon is smoothed to eliminate irregular jagged edges, and a spatial distribution map of the water mass is obtained. The spatial characteristics of the identified typhoon-affected water masses are calculated using the aforementioned water mass spatial distribution map. The distribution information of typhoon-characteristic water masses is then generated by combining the corresponding sea area situation characteristics, significance test results, and correlation analysis results.

5. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 1, characterized in that, The process of calculating the seawater carbon dioxide partial pressure anomaly field during the typhoon's passage based on the climatological background field, constructing a typhoon impact index time series using the seawater carbon dioxide partial pressure anomaly field, and generating typhoon impact trace paths specifically includes: Obtain the climatological background field, extract the monthly climatological mean field from the climatological background field as the reference data, obtain the marine situation monitoring dataset and extract the daily seawater carbon dioxide partial pressure grid data during the typhoon's passage. The climate anomaly algorithm is used to calculate the difference between the daily observation value and the monthly climatological mean value of each spatial grid point, thereby generating the daily seawater carbon dioxide partial pressure anomaly value of each grid point, which covers the entire target sea area during the typhoon. The distribution information of characteristic water masses of typhoons is obtained, and the spatial distribution range of the core water mass affected by typhoons is extracted as an analysis mask. Within this spatial mask, multi-dimensional features are extracted from the daily seawater carbon dioxide partial pressure anomaly field. The arithmetic mean of the outlier values ​​of all grid points is used as the anomaly intensity feature, the standard deviation of the outlier values ​​of all grid points is used as the spatial heterogeneity feature, and the proportion of grid points whose absolute outlier values ​​exceed the preset threshold is used as the significant influence range feature. The extracted multi-dimensional features are normalized, and then the weight coefficients of each feature are determined by a subjective and objective weighting method. The objective weights are calculated by the entropy weight method based on the degree of variation of each feature value, and the subjective weights are determined by the analytic hierarchy process combined with expert scoring. Finally, the subjective and objective weights are weighted and fused to obtain the comprehensive weight. The normalized multi-dimensional features are weighted and summed to calculate the daily typhoon impact index value, and a continuous time series of typhoon impact indexes throughout the entire life cycle of a typhoon is constructed. Based on the anomaly field of carbon dioxide partial pressure in seawater, an image segmentation algorithm based on Otsu's maximum inter-class variance method is used to identify and segment the anomaly region. The anomaly field is binarized into significant anomaly regions and non-significant regions. Morphological operations are used to optimize the segmentation results, remove small noise regions and fill holes, and obtain the spatial distribution map of the typhoon core impact area. Based on the spatial distribution map of the typhoon's core impact area, the coordinates of the geometric center point of the polygon in the impact area are calculated. The centroid algorithm is used to obtain the center point position by weighted averaging of the polygon vertex coordinates. The center points of each day during the typhoon's passage are connected by spline curves in chronological order to generate the typhoon impact trace path. At the same time, the date corresponding to each center point and the typhoon impact index value at that time are recorded.

6. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 1, characterized in that, The process of reconstructing the carbon dioxide partial pressure grid of the target sea area using the aforementioned marine situation monitoring dataset, and constructing a spatial map of typhoon impact by combining the typhoon impact index time series, typhoon impact trace paths, and typhoon characteristic water mass distribution information, specifically includes: A marine situation monitoring dataset is obtained. Based on the marine situation monitoring dataset, marine situation monitoring features are extracted as feature variables. Seawater carbon dioxide partial pressure data observed in the field are used as target variables. A seawater carbon dioxide partial pressure reconstruction model is constructed using a random forest regression algorithm. The target sea area is divided into several spatial grids. The corresponding environmental feature vectors are extracted from each grid point using the sea area situation monitoring dataset. The vectors are then input into the trained seawater carbon dioxide partial pressure reconstruction model to calculate the estimated seawater carbon dioxide partial pressure at the grid points, thereby generating a seawater carbon dioxide partial pressure reconstruction field covering the entire target sea area. The typhoon impact index time series and typhoon impact trace path are obtained. After smoothing the typhoon impact index time series, the key feature points of the typhoon impact intensity are identified by the local extremum detection algorithm, including the starting point, peak point and attenuation point. The morphological features of the sequence are extracted by the time series segmentation aggregation approximation method. Combined with the time axis information of the typhoon passage, a time series feature map reflecting the evolution process of the typhoon impact intensity is generated. Based on the latitude and longitude coordinates of the center point of the daily impact area and the corresponding typhoon impact index value extracted from the typhoon impact trace path, the inverse distance weighted interpolation algorithm is used for spatial interpolation to transform the discrete point-like impact intensity data into a continuous spatial distribution field, thus obtaining the spatial distribution field of typhoon impact intensity. Spatial registration was performed on the seawater carbon dioxide partial pressure reconstruction field, the spatial distribution field of typhoon impact intensity, and the temporal characteristic map. Spatial correlation between different sea area situation characteristics was established through raster calculation and vector overlay to generate an initial typhoon impact spatial map. The distribution information of characteristic water masses of typhoons is obtained and superimposed on the initial typhoon impact spatial map. The boundary attributes of water masses are correlated with the partial pressure field of carbon dioxide in the seawater. The degree of abnormal change caused by the typhoon is obtained by difference calculation using the climatological background field as a reference benchmark. The carbon dioxide partial pressure distribution is rendered with gradient colors using visualization technology, the influence intensity is represented by contour lines, and the water mass boundary is marked with preset rule symbols. Finally, the typhoon impact spatial map is obtained.

7. The method for differentiating typhoon impacts based on seawater CO2 partial pressure data according to claim 1, characterized in that, The analysis of the spatiotemporal dynamics of carbon dioxide flux at the air-sea interface during typhoon passage, based on the aforementioned typhoon impact spatial map, and the analysis of carbon sink stability in the target sea area after typhoon passage, generates a typhoon-carbon sink sensitivity region delineation map, specifically including: Obtain the spatial map of typhoon impact, extract the seawater carbon dioxide partial pressure reconstruction field and sea surface wind field data from the typhoon impact spatial map, obtain sea surface water temperature and salinity data from the marine situation monitoring dataset to calculate the solubility of carbon dioxide in seawater, and obtain the daily spatial distribution field of sea-air interface carbon dioxide flux during the typhoon's passage by combining the sea-air carbon dioxide flux calculation rules. Based on the calculated spatial distribution field of carbon dioxide exchange flux at the air-sea interface, the average carbon flux values ​​in the target sea area and various characteristic water masses are statistically analyzed in three periods: before, during, and after the passage of the typhoon. The amplitude and duration of flux changes caused by the typhoon are quantified through difference analysis, and the areas where carbon sinks are converted into carbon sources due to the typhoon are identified, generating information on carbon flux changes. The distribution map of typhoon characteristic water masses is extracted from the spatial map of typhoon impact. The carbon flux change information is associated with the carbon flux change information and the characteristic parameters of carbon flux change in different water mass regions are extracted, including the carbon flux change amplitude, change duration and recovery rate. The carbon flux change of each water mass during the typhoon is calculated and the differentiated response of carbon sink function of water masses with different properties during the typhoon is analyzed to generate water mass carbon sink response analysis information. The climatological background field was obtained, and the recovery status of seawater carbon dioxide partial pressure in the target sea area after the typhoon was analyzed. The degree of recovery of carbon sink function was assessed by comparing the difference between seawater carbon dioxide partial pressure value after the typhoon and the climatological background value. The recovery speed and recovery period of seawater carbon dioxide partial pressure value were calculated by time series trend and used as carbon sink stability assessment indicators. Combining information on carbon flux changes, water mass carbon sink response analysis, and carbon sink stability assessment indicators, the K-means clustering algorithm was used to divide the target sea area into sensitivity zones. The carbon sink stability assessment indicators were used as clustering features. Based on the response intensity and recovery capacity of each region to the impact of typhoons, the sea area was divided into three levels: high-sensitivity zone, medium-sensitivity zone, and low-sensitivity zone, generating a typhoon-carbon sink sensitivity zone division map to characterize the impact of typhoons on the target sea area.

8. A typhoon impact differentiation system based on seawater CO2 partial pressure data, characterized in that, The system includes: a memory, a processor, and a communication interface. The memory contains a program for differentiating typhoon impacts based on seawater CO2 partial pressure data. When the processor executes the program for differentiating typhoon impacts based on seawater CO2 partial pressure data, it performs the following steps: Acquire marine situation monitoring data during the passage of typhoons in the target sea area, preprocess the data to generate a marine situation monitoring dataset, and construct a climatological background field based on historical marine situation data of the target sea area. Based on the aforementioned marine situation monitoring dataset, a clustering algorithm was used to divide water masses, identify water masses of different properties formed under the influence of typhoons, and conduct correlation analysis and significance tests on each water mass to obtain the distribution information of typhoon characteristic water masses. The anomaly field of seawater carbon dioxide partial pressure during the passage of the typhoon is calculated based on the climatological background field. The time series sequence of typhoon impact index is constructed through the anomaly field of seawater carbon dioxide partial pressure and the typhoon impact trace path is generated. The grid points of seawater carbon dioxide partial pressure in the target sea area are reconstructed using the sea area situation monitoring dataset. A spatial map of typhoon impact is constructed by combining the typhoon impact index time series, typhoon impact trace path and typhoon characteristic water mass distribution information. Based on the spatial map of typhoon impact, the spatiotemporal dynamic changes of carbon dioxide flux at the air-sea interface during the passage of typhoons are analyzed, and the carbon sink stability of the target sea area after the passage of typhoons is analyzed, generating a map of typhoon-carbon sink sensitive areas.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a typhoon impact differentiation method program based on seawater CO2 partial pressure data. When the typhoon impact differentiation method program based on seawater CO2 partial pressure data is executed by a processor, it implements the steps of the typhoon impact differentiation method based on seawater CO2 partial pressure data as described in any one of claims 1 to 7.

Images