A marine ecosystem monitoring system based on multi-source data fusion

Abstract

The application relates to the technical field of marine environment monitoring, in particular to a marine ecosystem monitoring system based on multi-source data fusion, which comprises a multi-source data acquisition module, a parameter extraction module, an ecological driving analysis module and an ecosystem anomaly monitoring module; wherein: the multi-source data acquisition module is used for acquiring multidimensional data; the parameter extraction module generates unified grid data through space-time alignment processing, and extracts dissolved inorganic phosphorus concentration time series data and diatom community spectral feature data sets; the ecological driving analysis module calculates the fluctuation rate and generates diatom community aggregation degree; the ecosystem anomaly monitoring module is used for identifying the abnormal state of the ecosystem and outputting corresponding early warning signals. Through multi-source data fusion and dynamic trend analysis, the application realizes accurate identification and timely early warning of marine ecological abnormal events, and improves the timeliness and discrimination accuracy of ecological monitoring.

Description

Technical Field

[0001] This invention relates to the field of marine environmental monitoring technology, and in particular to a marine ecosystem monitoring system based on multi-source data fusion. Background Technology

[0002] In recent years, marine ecosystems have been affected by multiple disturbances, including global climate change, land-based pollution input, and overfishing, exhibiting highly complex and dynamic evolutionary trends. Among these, fluctuations in dissolved nutrient concentrations are closely related to changes in phytoplankton community structure and are widely considered to be the core driving factors for red tides, ecological imbalances, and biodiversity degradation. Traditional marine ecological monitoring methods rely on single data sources for on-site observations, which have limited spatiotemporal resolution and cannot fully reflect the changes in the state of regional ecosystems, especially in the prediction and trend analysis of abnormal events, where there is a significant lag.

[0003] Current technologies for monitoring ecological anomalies still face several key challenges: First, the lack of an effective mechanism for fusing multi-source heterogeneous observation data makes it difficult to uniformly express ecological parameters. Second, the identification of driving factors generally neglects the temporal coupling relationship between the dynamic fluctuations of nutrient concentration and changes in community structure. Third, in risk event assessment, static threshold discrimination methods are often used, failing to reflect the process characteristics of trend changes in the community. Therefore, a marine ecosystem monitoring system based on multi-source data fusion is urgently needed to address these issues. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides a marine ecosystem monitoring system based on multi-source data fusion.

[0005] A marine ecosystem monitoring system based on multi-source data fusion includes a multi-source data acquisition module, a parameter extraction module, an ecological driving analysis module, and an ecosystem anomaly monitoring module; wherein: Multi-source data acquisition module: used to acquire satellite multispectral imagery, buoy sensor data, and underwater in-situ spectral data; Parameter extraction module: It receives data from the multi-source data acquisition module, generates unified grid data through spatiotemporal alignment processing, and extracts time-series data of dissolved inorganic phosphorus concentration and spectral feature dataset of diatom communities; Ecological-driven analysis module: This module receives time-series data on soluble inorganic phosphorus concentration output by the parameter extraction module, calculates its fluctuation rate, and generates diatom community aggregation degree based on diatom community spectral characteristic data. Ecosystem Anomaly Monitoring Module: Based on the combined characteristics of fluctuation rate and diatom community aggregation, it identifies abnormal ecosystem states and outputs corresponding early warning signals.

[0006] Optionally, the multi-source data acquisition module includes a satellite image receiving unit, a buoy sensing acquisition unit, and an underwater spectral acquisition unit; wherein: Satellite image receiving unit: used to receive multispectral image data transmitted from satellites in a preset orbit at regular intervals, and to perform geometric correction and radiometric correction on it to generate a standardized reflectance image sequence that matches the target monitoring sea area; Buoy Sensing and Acquisition Unit: Used to collect real-time data on sea surface temperature, salinity, turbidity, and nutrient concentration through buoy sensors deployed in the monitored sea area; Underwater spectral acquisition unit: Used to acquire in-situ underwater spectral data within the target depth range using a fixed or towed underwater spectrometer, along with corresponding acquisition time and three-dimensional position information.

[0007] Optionally, the parameter extraction module includes a spatiotemporal alignment processing unit, a grid generation unit, a phosphorus concentration extraction unit, and a spectral feature extraction unit; wherein: Spatiotemporal alignment processing unit: used to receive multispectral image data, buoy sensor data and underwater spectral data output by multi-source data acquisition module, unify the timestamp format of various data and perform spatial resampling based on geographic coordinate system to achieve synchronous registration in time and space dimensions. Grid generation unit: Based on the set latitude and longitude grid resolution, a unified spatial grid structure is constructed within the target monitoring area, and the aligned multi-source data is mapped to the corresponding grid unit to form a unified cross-source dataset; Phosphorus concentration extraction unit: used to extract the soluble inorganic phosphorus concentration values ​​within the corresponding time series from the gridded buoy sensor data, and summarize them according to the grid index to form time series data of soluble inorganic phosphorus concentration; Spectral feature extraction unit: Based on underwater in-situ spectral data, it identifies the main absorption band and reflection peak characteristics of diatoms and extracts feature values ​​to form a spectral feature dataset of diatom communities.

[0008] Optionally, the phosphorus concentration extraction unit includes: Data filtering subunit: Used to filter raw observation records containing soluble inorganic phosphorus concentrations and remove outliers, missing values ​​or duplicate records; Time series sub-cell construction: Based on the timestamp of each grid cell, the observed values ​​of soluble inorganic phosphorus concentrations after screening are arranged in chronological order to construct a preliminary concentration series within the grid; Grid aggregation sub-cell: Used to perform weighted averaging of observations from multiple buoys within each grid cell to generate standardized time-series data on soluble inorganic phosphorus concentration.

[0009] Optionally, the spectral feature extraction unit includes: Band selection subunit: used to receive underwater in-situ spectral data and, based on a preset diatom characteristic response range, extract reflectance curves in the wavelength range of 400nm to 750nm. Feature recognition subunit: Used to identify the typical main absorption band and reflection peak positions of diatoms in the selected band, detect local extrema using the second derivative method, and locate the wavelength of the main absorption band. and the wavelength of the main reflection peak ; Feature extraction subunit: used to extract the reflectance value corresponding to the identified feature wavelength, denoted as . and And calculate the characteristic reflectance index of diatoms. Finally, the extracted and The spectral features of diatom communities are used as feature vectors to form individual records, and then summarized by time and spatial indexes to form a dataset of diatom community spectral features.

[0010] Optionally, the ecological driving analysis module includes a phosphorus fluctuation analysis unit and a diatom aggregation generation unit; wherein: Phosphorus fluctuation analysis unit: Based on time series data of soluble inorganic phosphorus concentration, extract the concentration change values ​​at adjacent time points, and combine them with continuous time windows to calculate the fluctuation rate characterizing the degree of local change; Diatom Aggregation Generation Unit: Based on the spectral feature dataset of diatom communities, the characteristic distribution of sampling points within the same grid is analyzed, and their spatial concentration is calculated to generate the diatom community aggregation degree.

[0011] Optionally, the phosphorus fluctuation analysis unit includes: The difference calculation sub-cell is used to perform pairwise differences on the soluble inorganic phosphorus concentration at consecutive time points within the same grid cell, extracting the concentration change values ​​between adjacent time points and constructing a difference sequence. ; Rate estimation subunit: used to set the length as A sliding time window is used to calculate the mean of the absolute values ​​of the difference sequence within the window, and the fluctuation rate at the current moment is output. .

[0012] Optionally, the diatom aggregation generation unit includes: Feature vector construction sub-unit: used to extract feature values ​​of each sampling point from the diatom community spectral feature dataset, including the wavelength of the main peak absorption band, the wavelength of the main reflection peak, the corresponding reflectance value and the diatom spectral index, and combine them to form a feature vector; Spatial consistency calculation sub-cell: Calculates the spatial consistency mean based on the Euclidean distance between all feature vectors within the same grid. ; Clustering degree generates sub-units: based on spatial consistency value Calculate the clustering index Its expression is: ,in, Represents grid cells diatom community aggregation degree; This is the preset maximum spatial consistency reference value.

[0013] Optionally, the ecosystem anomaly monitoring module includes a threshold determination unit, a trend determination unit, a risk assessment unit, and a signal generation unit; wherein: Threshold determination unit: used to compare the fluctuation rate with a preset threshold to determine whether it is in a high fluctuation state or a low fluctuation state; Trend determination unit: used to determine whether there is a synchronous upward or continuous downward trend based on the changes in the aggregation degree of diatom communities over a continuous period of time; Risk assessment unit: used to jointly analyze the fluctuation rate status and the aggregation trend. When the fluctuation rate is higher than the threshold and the aggregation increases simultaneously, it is identified as a red tide risk event; when the fluctuation rate is lower than the threshold and the aggregation continues to decline, it is identified as a biodiversity degradation event. Signal generation unit: Used to generate corresponding red tide risk level signals or biodiversity degradation early warning signals based on the judgment results of the risk assessment unit.

[0014] Optionally, the trend determination unit includes: Clustering sequence constructs sub-cells: used to extract specified grid cells in a continuous sequence. The aggregation degree values ​​of diatom communities at each time point were used to construct an aggregation degree sequence; The rate of change calculation subunit is used to perform linear fitting on the clustering sequence to obtain the rate of change of clustering with respect to time. ; Trend classification and determination subunit: used to classify the calculated rate of change. Compared with a preset threshold, when When it is determined to be a synchronous upward trend; when The timeframe indicates a continuous downward trend; among them, and These are the thresholds for the rate of increase and the rate of decrease, respectively.

[0015] The beneficial effects of this invention are: This invention constructs a multi-source data acquisition module to achieve the collaborative acquisition of satellite multispectral imagery, buoy sensor data, and underwater in-situ spectral data. It also completes spatiotemporal alignment and grid fusion processing through a parameter extraction module, and uniformly extracts time-series data of dissolved inorganic phosphorus concentration and spectral characteristic data of diatom communities. This effectively solves the heterogeneity problem of multi-source data at spatiotemporal scales and improves the consistency and comparability of ecological monitoring data.

[0016] This invention, through an ecological driving analysis module and an ecosystem anomaly monitoring module, achieves joint trend identification of nutrient fluctuations and community structure changes. Based on the dynamic fluctuation rate and aggregation degree evolution trend, it accurately judges red tide risk events and biodiversity degradation status, breaking through the limitations of traditional static threshold judgment and enhancing the sensitivity and foresight of marine ecosystem anomaly monitoring. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a marine ecosystem monitoring system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the ecosystem anomaly monitoring module according to an embodiment of the present invention. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0020] like Figures 1-2 As shown, a marine ecosystem monitoring system based on multi-source data fusion includes a multi-source data acquisition module, a parameter extraction module, an ecological driving analysis module, and an ecosystem anomaly monitoring module; wherein: Multi-source data acquisition module: used to acquire satellite multispectral imagery, buoy sensor data, and underwater in-situ spectral data; The multi-source data acquisition module includes a satellite image receiving unit, a buoy sensing acquisition unit, and an underwater spectral acquisition unit; among which: Satellite image receiving unit: used to receive multispectral image data transmitted from satellites in a preset orbit at regular intervals, and to perform geometric correction and radiometric correction on it to generate a standardized reflectance image sequence that matches the target monitoring sea area; Buoy Sensing and Acquisition Unit: Used to collect real-time data on sea surface temperature, salinity, turbidity, and nutrient concentration through buoy sensors deployed in the monitored sea area, and transmit the data to the system processing center via an embedded wireless communication module; Underwater spectral acquisition unit: Used to acquire in-situ underwater spectral data within the target depth range using a fixed or towed underwater spectrometer, along with corresponding acquisition time and three-dimensional location information; through the design of this multi-source data acquisition module, it can comprehensively cover the upper layer remote sensing observation of marine ecosystems, surface environment perception, and in-situ detection of the water body, thereby improving the spatiotemporal resolution of the ecological status of the monitored marine area.

[0021] Parameter extraction module: It receives data from the multi-source data acquisition module, generates unified grid data through spatiotemporal alignment processing, and extracts time-series data of dissolved inorganic phosphorus concentration and spectral feature dataset of diatom communities; The parameter extraction module includes a spatiotemporal alignment processing unit, a mesh generation unit, a phosphorus concentration extraction unit, and a spectral feature extraction unit; wherein: Spatiotemporal alignment processing unit: used to receive multispectral image data, buoy sensor data and underwater spectral data output by multi-source data acquisition module, unify the timestamp format of various data and perform spatial resampling based on geographic coordinate system to achieve synchronous registration in time and space dimensions. Grid generation unit: Based on the set latitude and longitude grid resolution, a unified spatial grid structure is constructed within the target monitoring area, and the aligned multi-source data is mapped to the corresponding grid unit to form a unified cross-source dataset; Phosphorus concentration extraction unit: used to extract the soluble inorganic phosphorus concentration values ​​within the corresponding time series from the gridded buoy sensor data, and summarize them according to the grid index to form time series data of soluble inorganic phosphorus concentration; Spectral feature extraction unit: Based on underwater in-situ spectral data, it identifies the main absorption band and reflection peak characteristics of diatoms and extracts feature values ​​to form a spectral feature dataset of diatom communities. Through the setting of the above parameter extraction module, it is possible to achieve unified alignment and grid fusion of heterogeneous multi-source data, and accurately extract the time series and spectral feature data corresponding to key ecological factors, providing standardized input for subsequent ecological driving analysis.

[0022] The phosphorus concentration extraction unit includes: Data filtering subunit: Used to filter raw observation records containing soluble inorganic phosphorus (DIP) concentrations and remove outliers, missing values ​​or duplicate records to ensure data quality; Time series sub-cell construction: Based on the timestamp of each grid cell, the selected soluble inorganic phosphorus concentration observations are arranged in chronological order to construct a preliminary concentration series within the grid. ,in Represents a grid In time The concentration value of soluble inorganic phosphorus; Grid aggregation sub-cell: Used to perform weighted averaging of observations from multiple buoys within each grid cell, generating standardized time-series data on soluble inorganic phosphorus concentrations. The calculation formula is as follows: ,in, For grid cells In time Summary of soluble inorganic phosphorus concentrations; For the first A buoy in the grid cell Observed values ​​of soluble inorganic phosphorus concentration; For the first The weighting factor for each buoy; For grid cells The total number of buoys within.

[0023] The spectral feature extraction unit includes: Band selection subunit: used to receive underwater in-situ spectral data and, based on the preset diatom characteristic response range, extract the reflectance curve in the wavelength range of 400nm to 750nm as the data basis for subsequent feature identification; Feature recognition subunit: Used to identify the typical main absorption band and reflection peak positions of diatoms in the selected band, detect local extrema using the second derivative method, and locate the wavelength of the main absorption band. and the wavelength of the main reflection peak The basis for this judgment is: ; ,in, It is the reflectance spectrum; Indicates the center wavelength of the absorption band; Indicates the center wavelength of the reflection peak; This represents the second derivative of reflectivity with respect to wavelength, i.e., curvature. The basis for this judgment is explained as follows: In underwater in-situ spectroscopy, different phytoplankton groups exhibit specific spectral absorption and reflection characteristics. Diatoms, due to their rich chlorophyll a and carotenoid cell structure, typically show distinct absorption troughs and reflection peaks in specific wavelength ranges (such as around 670 nm). To accurately identify these locations, the second derivative analysis of reflectance curves is a commonly used method for spectral feature localization: when... When this occurs, it indicates that the point is a local minimum, corresponding to the center wavelength of the absorption band (i.e., );when When this occurs, it indicates that the point is a local maximum, corresponding to the center wavelength of the reflection peak (i.e., This method avoids the misjudgments that may be caused by noise or plateau effect in the simple first derivative method, and more stably identifies the key feature positions in the spectral structure, providing a highly reliable input for the subsequent construction of spectral feature vectors. This strategy is derived from the curvature peak identification principle in spectral analysis and is often used in plant classification, water color monitoring and chlorophyll estimation. In this invention, it is used to accurately extract the main feature sites of diatom communities, and has both theoretical basis and practical adaptability.

[0024] Feature extraction subunit: used to extract the reflectance value corresponding to the identified feature wavelength, denoted as . and And calculate the characteristic reflectance index of diatoms. Finally, the extracted and As feature vectors, individual records are composed and summarized by time and spatial indexes to form a dataset of diatom community spectral features. The aforementioned reflectance index The calculation formula is as follows: By setting up the above-mentioned spectral feature extraction unit, the spectral response characteristics of diatoms in specific bands can be accurately captured, and structured feature data can be formed, providing a reliable basis for the identification and modeling of community change trends in ecological driving analysis.

[0025] Ecological-driven analysis module: This module receives time-series data on soluble inorganic phosphorus concentration output by the parameter extraction module, calculates its fluctuation rate, and generates diatom community aggregation degree based on diatom community spectral characteristic data. The ecological driving analysis module includes a phosphorus fluctuation analysis unit and a diatom aggregation generation unit; wherein: Phosphorus fluctuation analysis unit: Based on time series data of soluble inorganic phosphorus concentration, extract the concentration change values ​​at adjacent time points, and combine them with continuous time windows to calculate the fluctuation rate characterizing the degree of local change; Diatom Aggregation Generation Unit: Based on the spectral feature dataset of diatom communities, the characteristic distribution of sampling points within the same grid is analyzed, and their spatial concentration is calculated to generate the diatom community aggregation degree. Through the setting of the above-mentioned ecological driving analysis module, the dynamic fluctuation of nutrients can be evaluated from the time dimension, and the community aggregation characteristics can be quantified from the spatial dimension, providing key driving indicators for subsequent anomaly monitoring.

[0026] The phosphorus fluctuation analysis unit includes: The difference calculation sub-cell is used to perform pairwise differences on the soluble inorganic phosphorus concentration at consecutive time points within the same grid cell, extracting the concentration change values ​​between adjacent time points and constructing a difference sequence. The calculation formula is as follows: ,in, Represents grid cells In time The change in phosphorus concentration at time t; Represents a grid In time The concentration value of soluble inorganic phosphorus; This indicates that the unit was at the previous moment. The concentration value of soluble inorganic phosphorus; Rate estimation subunit: used to set the length as A sliding time window is used to calculate the mean of the absolute values ​​of the difference sequence within the window, and the fluctuation rate at the current moment is output. The calculation formula is as follows: ,in, Represents grid cells In time The rate of fluctuation of phosphorus concentration; Indicates the length of the sliding time window (in time steps); Indicates time The change in phosphorus concentration in that grid cell at any given time; This indicates the operation of taking the absolute value; by setting up the phosphorus fluctuation analysis unit mentioned above, it is possible to extract and quantify the degree of fluctuation of phosphorus concentration in local time series, providing a precise and controllable analytical means for identifying dynamic disturbances of nutrients.

[0027] Diatom aggregation generating units include: Feature vector construction sub-unit: used to extract feature values ​​of each sampling point from the diatom community spectral feature dataset, including the wavelength of the main peak absorption band, the wavelength of the main reflection peak, the corresponding reflectance value and the diatom spectral index, and combine them to form a feature vector; Spatial consistency calculation sub-cell: Calculates the spatial consistency mean based on the Euclidean distance between all feature vectors within the same grid. The calculation formula is as follows: ,in, This represents the number of sampling points within the grid. Represents the Euclidean distance between vectors; and They represent the first The and the first Diatom spectral feature vectors of each sampling point; Clustering degree generates sub-units: based on spatial consistency value Calculate the clustering index Its expression is: ,in, Represents grid cells diatom community aggregation degree; The preset maximum spatial consistency reference value is used for normalization. By setting the above diatom aggregation degree generation unit, the degree of concentration of diatom communities in the spectral feature space within the same region can be quantitatively characterized, thereby providing a key spatial structure criterion for anomaly detection.

[0028] Ecosystem Anomaly Monitoring Module: Based on the combined characteristics of fluctuation rate and diatom community aggregation degree, it identifies abnormal ecosystem states and outputs corresponding early warning signals; The ecosystem anomaly monitoring module includes a threshold determination unit, a trend determination unit, a risk assessment unit, and a signal generation unit; among which: Threshold determination unit: used to compare the fluctuation rate with a preset threshold to determine whether it is in a high fluctuation state or a low fluctuation state; Trend determination unit: used to determine whether there is a synchronous upward or continuous downward trend based on the changes in the aggregation degree of diatom communities over a continuous period of time; Risk assessment unit: used to jointly analyze the fluctuation rate status and the aggregation trend. When the fluctuation rate is higher than the threshold and the aggregation increases simultaneously, it is identified as a red tide risk event; when the fluctuation rate is lower than the threshold and the aggregation continues to decline, it is identified as a biodiversity degradation event. Signal generation unit: Used to generate corresponding red tide risk level signals or biodiversity degradation early warning signals based on the judgment results of the risk assessment unit; through the setting of the above-mentioned ecosystem anomaly monitoring module, it is possible to automatically identify typical ecological anomalies based on the joint judgment of dynamic factors, and output risk signals with clear type identification, thereby improving the response accuracy and intelligence level of the marine ecological early warning system.

[0029] The trend determination unit includes: Clustering sequence constructs sub-cells: used to extract specified grid cells in a continuous sequence. The aggregation degree values ​​of diatom communities at each time point were used to construct an aggregation degree sequence; The rate of change calculation subunit is used to perform linear fitting on the clustering sequence to obtain the rate of change of clustering with respect to time. The calculation formula is as follows: ,in, Represents grid cells The rate of change of the degree of aggregation; For the first The aggregation value at each moment; The mean of the sequence; Indicates the first The time value corresponding to each moment; This is the average value across all time points; Trend classification and determination subunit: used to classify the calculated rate of change. Compared with a preset threshold, when When it is determined to be a synchronous upward trend; when The timeframe indicates a continuous downward trend; among them, and These are the threshold values ​​for the rate of increase and the rate of decrease, respectively, and the set values ​​are... and By setting up the aforementioned trend determination unit, trend change characteristics can be extracted from the aggregation evolution trajectory over a continuous period of time, accurately identifying whether diatom communities exhibit a dynamic process of enhancement or decline, and providing trend support for the intelligent identification of ecological anomalies.

[0030] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0031] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A marine ecosystem monitoring system based on multi-source data fusion, characterized in that, It includes a multi-source data acquisition module, a parameter extraction module, an ecosystem-driven analysis module, and an ecosystem anomaly monitoring module; among which: Multi-source data acquisition module: used to acquire satellite multispectral imagery, buoy sensor data, and underwater in-situ spectral data; Parameter extraction module: It receives data from the multi-source data acquisition module, generates unified grid data through spatiotemporal alignment processing, and extracts time-series data of dissolved inorganic phosphorus concentration and spectral feature dataset of diatom communities; Ecological-driven analysis module: This module receives time-series data on soluble inorganic phosphorus concentration output by the parameter extraction module, calculates its fluctuation rate, and generates diatom community aggregation degree based on diatom community spectral characteristic data. Ecosystem Anomaly Monitoring Module: Based on the combined characteristics of fluctuation rate and diatom community aggregation, it identifies abnormal ecosystem states and outputs corresponding early warning signals.

2. The marine ecosystem monitoring system based on multi-source data fusion according to claim 1, characterized in that, The multi-source data acquisition module includes a satellite image receiving unit, a buoy sensing acquisition unit, and an underwater spectral acquisition unit; wherein: Satellite image receiving unit: used to receive multispectral image data transmitted from satellites in a preset orbit at regular intervals, and to perform geometric correction and radiometric correction on it to generate a standardized reflectance image sequence that matches the target monitoring sea area; Buoy Sensing and Acquisition Unit: Used to collect real-time data on sea surface temperature, salinity, turbidity, and nutrient concentration through buoy sensors deployed in the monitored sea area; Underwater spectral acquisition unit: Used to acquire in-situ underwater spectral data within the target depth range using a fixed or towed underwater spectrometer, along with corresponding acquisition time and three-dimensional position information.

3. A marine ecosystem monitoring system based on multi-source data fusion according to claim 1, characterized in that, The parameter extraction module includes a spatiotemporal alignment processing unit, a grid generation unit, a phosphorus concentration extraction unit, and a spectral feature extraction unit; wherein: Spatiotemporal alignment processing unit: used to receive multispectral image data, buoy sensor data and underwater spectral data output by multi-source data acquisition module, unify the timestamp format of various data and perform spatial resampling based on geographic coordinate system to achieve synchronous registration in time and space dimensions. Grid generation unit: Based on the set latitude and longitude grid resolution, a unified spatial grid structure is constructed within the target monitoring area, and the aligned multi-source data is mapped to the corresponding grid unit to form a unified cross-source dataset; Phosphorus concentration extraction unit: used to extract the soluble inorganic phosphorus concentration values ​​within the corresponding time series from the gridded buoy sensor data, and summarize them according to the grid index to form time series data of soluble inorganic phosphorus concentration; Spectral feature extraction unit: Based on underwater in-situ spectral data, it identifies the main absorption band and reflection peak characteristics of diatoms and extracts feature values ​​to form a spectral feature dataset of diatom communities.

4. A marine ecosystem monitoring system based on multi-source data fusion according to claim 3, characterized in that, The phosphorus concentration extraction unit includes: Data filtering subunit: Used to filter raw observation records containing soluble inorganic phosphorus concentrations and remove outliers, missing values ​​or duplicate records; Time series sub-cell construction: Based on the timestamp of each grid cell, the observed values ​​of soluble inorganic phosphorus concentrations after screening are arranged in chronological order to construct a preliminary concentration series within the grid; Grid aggregation sub-cell: Used to perform weighted averaging of observations from multiple buoys within each grid cell to generate standardized time-series data on soluble inorganic phosphorus concentration.

5. A marine ecosystem monitoring system based on multi-source data fusion according to claim 3, characterized in that, The spectral feature extraction unit includes: Band selection subunit: used to receive underwater in-situ spectral data and, based on a preset diatom characteristic response range, extract reflectance curves in the wavelength range of 400nm to 750nm. Feature recognition subunit: Used to identify the typical main absorption band and reflection peak positions of diatoms in the selected band, detect local extrema using the second derivative method, and locate the wavelength of the main absorption band. and the wavelength of the main reflection peak ; Feature extraction subunit: used to extract the reflectance value corresponding to the identified feature wavelength, denoted as . and And calculate the characteristic reflectance index of diatoms. Finally, the extracted and The spectral features of diatom communities are used as feature vectors to form individual records, and then summarized by time and spatial indexes to form a dataset of diatom community spectral features.

6. A marine ecosystem monitoring system based on multi-source data fusion according to claim 1, characterized in that, The ecological driving analysis module includes a phosphorus fluctuation analysis unit and a diatom aggregation generation unit; wherein: Phosphorus fluctuation analysis unit: Based on time series data of soluble inorganic phosphorus concentration, extract the concentration change values ​​at adjacent time points, and combine them with continuous time windows to calculate the fluctuation rate characterizing the degree of local change; Diatom Aggregation Generation Unit: Based on the spectral feature dataset of diatom communities, the characteristic distribution of sampling points within the same grid is analyzed, and their spatial concentration is calculated to generate the diatom community aggregation degree.

7. A marine ecosystem monitoring system based on multi-source data fusion according to claim 6, characterized in that, The phosphorus fluctuation analysis unit includes: The difference calculation sub-cell is used to perform pairwise differences on the soluble inorganic phosphorus concentration at consecutive time points within the same grid cell, extracting the concentration change values ​​between adjacent time points and constructing a difference sequence. ; Rate estimation subunit: used to set the length as A sliding time window is used to calculate the mean of the absolute values ​​of the difference sequence within the window, and the fluctuation rate at the current moment is output. .

8. A marine ecosystem monitoring system based on multi-source data fusion according to claim 6, characterized in that, The diatom aggregation generation unit includes: Feature vector construction sub-unit: used to extract feature values ​​of each sampling point from the diatom community spectral feature dataset, including the wavelength of the main peak absorption band, the wavelength of the main reflection peak, the corresponding reflectance value and the diatom spectral index, and combine them to form a feature vector; Spatial consistency calculation sub-cell: Calculates the spatial consistency mean based on the Euclidean distance between all feature vectors within the same grid. ; Clustering degree generates sub-units: based on spatial consistency value Calculate the clustering index Its expression is: ,in, Represents grid cells diatom community aggregation degree; This is the preset maximum spatial consistency reference value.

9. A marine ecosystem monitoring system based on multi-source data fusion according to claim 1, characterized in that, The ecosystem anomaly monitoring module includes a threshold determination unit, a trend determination unit, a risk assessment unit, and a signal generation unit; wherein: Threshold determination unit: used to compare the fluctuation rate with a preset threshold to determine whether it is in a high fluctuation state or a low fluctuation state; Trend determination unit: used to determine whether there is a synchronous upward or continuous downward trend based on the changes in the aggregation degree of diatom communities over a continuous period of time; Risk assessment unit: used to jointly analyze the fluctuation rate status and the aggregation trend. When the fluctuation rate is higher than the threshold and the aggregation increases simultaneously, it is identified as a red tide risk event; when the fluctuation rate is lower than the threshold and the aggregation continues to decline, it is identified as a biodiversity degradation event. Signal generation unit: Used to generate corresponding red tide risk level signals or biodiversity degradation early warning signals based on the judgment results of the risk assessment unit.

10. A marine ecosystem monitoring system based on multi-source data fusion according to claim 9, characterized in that, The trend determination unit includes: Clustering sequence constructs sub-cells: used to extract specified grid cells in a continuous sequence. The aggregation degree values ​​of diatom communities at each time point were used to construct an aggregation degree sequence; The rate of change calculation subunit is used to perform linear fitting on the clustering sequence to obtain the rate of change of clustering with respect to time. ; Trend classification and determination subunit: used to classify the calculated rate of change. Compared with a preset threshold, when When it is determined to be a synchronous upward trend; when The timeframe indicates a continuous downward trend; among them, and These are the thresholds for the rate of increase and the rate of decrease, respectively.

Images