Sea area phytoplankton monitoring system and method based on artificial intelligence

Abstract

The invention relates to the technical field of sea area phytoplankton monitoring, and particularly discloses a sea area phytoplankton monitoring system and method based on artificial intelligence, and the system comprises a data collection module which is used for obtaining multi-source three-dimensional monitoring data of a target sea area, and the multi-source three-dimensional monitoring data comprises a surface layer optical remote sensing image; the vertical profile data of the water body comprises optical parameters and seawater physical and chemical parameters which are acquired by a profile buoy, a towed sensor, an underwater unmanned vehicle or a shipborne profile measurement system, have at least two different depths and are related to phytoplankton; and the data preprocessing and fusion module is in communication connection with the data acquisition module. According to the sea area phytoplankton monitoring system and method based on artificial intelligence, comprehensive technology upgrading of phytoplankton from a two-dimensional plane to a three-dimensional space, from representation to mechanism and from post-event statistics to prospective early warning is achieved, and the dimension limitation of traditional monitoring is broken through.

Description

Technical Field

[0001] This invention relates to the field of marine phytoplankton monitoring technology, specifically to an artificial intelligence-based marine phytoplankton monitoring system and method. Background Technology

[0002] Currently, the monitoring of marine phytoplankton mainly relies on three methods: satellite remote sensing, field observation, and numerical models. Although satellite remote sensing can achieve large-scale coverage, it can only acquire surface information and has limited detection capabilities for ecological processes occurring in the subsurface or mid-water layers (such as the subsurface chlorophyll-maximum layer). Field observation (such as ship sampling and profiling buoys) can obtain high-precision vertical data, but it suffers from bottlenecks such as sparse spatial coverage, high cost, and poor timeliness. The accuracy of numerical models is highly dependent on the assimilation quality of observation data and the model parameterization scheme. Existing technologies as a whole are characterized by "two-dimensionality, surface-level focus, and data-model disconnect," making it impossible to economically and efficiently acquire and analyze the three-dimensional distribution structure of phytoplankton in water bodies. In particular, it is difficult to quantify their vertical stratification characteristics and niche separation, resulting in widespread "lag" and "blind spots" in early warning of ecological events such as harmful algal blooms.

[0003] Although artificial intelligence technology has been introduced to improve data processing capabilities, its application is mostly limited to optimizing two-dimensional remote sensing inversion, and has not yet fundamentally built an innovative system that integrates multi-source three-dimensional observation to achieve three-dimensional intelligent monitoring and vertical mechanism early warning. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides the following technical solution: a marine phytoplankton monitoring system and method based on artificial intelligence, comprising:

[0005] The data acquisition module is used to acquire multi-source three-dimensional monitoring data of the target sea area, wherein the multi-source three-dimensional monitoring data includes:

[0006] Surface optical remote sensing images;

[0007] Vertical profile data of water bodies, including optical parameters and seawater physicochemical parameters related to phytoplankton at at least two different depths, collected by profile buoys, towed sensors, underwater unmanned vehicles or shipborne profile measurement systems.

[0008] The data preprocessing and fusion module is communicatively connected to the data acquisition module and is used to preprocess and spatiotemporally register the surface optical remote sensing images and the water body vertical profile data, and to construct a three-dimensional standardized spatiotemporal grid dataset.

[0009] The feature extraction and selection module is communicatively connected to the data preprocessing and fusion module, and is used to extract a multidimensional feature set related to phytoplankton biomass, species composition and vertical distribution structure from the three-dimensional standardized spatiotemporal grid dataset;

[0010] An artificial intelligence analysis engine, communicatively connected to the feature extraction and selection module, includes:

[0011] The trained phytoplankton three-dimensional monitoring model is used to output the three-dimensional distribution information of phytoplankton in the target sea area based on the multi-dimensional feature set. The three-dimensional distribution information includes: the three-dimensional distribution field of chlorophyll a concentration, the depth and thickness of the vertical dominance layer of the main phytoplankton functional groups, and the vertical range of the harmful algal bloom area.

[0012] The dynamic early warning and visualization module is connected to the artificial intelligence analysis engine. It is used to generate early warning information based on the three-dimensional distribution information of the phytoplankton and combined with a preset vertical structure threshold, and to display the three-dimensional distribution information and early warning information in three dimensions.

[0013] Preferably, the optical parameters related to phytoplankton included in the vertical profile data of the water body include, but are not limited to, chlorophyll fluorescence values ​​at different depths, beam attenuation coefficients, or chlorophyll a concentration profiles obtained by inverting the above parameters.

[0014] Preferably, the data preprocessing and fusion module, when constructing a three-dimensional standardized spatiotemporal grid dataset, is specifically used for:

[0015] The surface optical remote sensing image is used as a horizontal plane constraint.

[0016] The discrete vertical profile data of the water body are fused into a three-dimensional mesh constrained by the horizontal plane through three-dimensional kriging interpolation, three-dimensional variational assimilation or machine learning interpolation methods to form a continuous three-dimensional data volume.

[0017] Preferably, the features extracted by the feature extraction and selection module that are related to the vertical distribution structure include, but are not limited to: the maximum chlorophyll layer depth, the vertical heterogeneity index, the gradient changes of feature parameters at different depths, and the combination of features characterizing the vertical stratification phenomenon.

[0018] Preferably, the phytoplankton three-dimensional monitoring model in the artificial intelligence analysis engine is trained by fusing surface remote sensing features and vertical profile features, and its architecture includes a network structure for extracting vertical contextual information, such as a one-dimensional convolutional neural network or a recurrent neural network layer.

[0019] Preferably, the dynamic early warning and visualization module is used for:

[0020] Identify whether there are abnormal subsurface chlorophyll maximum values ​​in the three-dimensional distribution field of chlorophyll a concentration, and evaluate their depth, intensity and duration;

[0021] Based on the vertical extent of the harmful algal bloom area, determine whether it has expanded from the surface to a depth that may affect water mixing or the underlying ecosystem;

[0022] Generate a comprehensive ecological risk assessment report that includes vertical dimension information.

[0023] An artificial intelligence-based method for monitoring marine phytoplankton includes the following steps:

[0024] S1. Multi-source stereo data acquisition steps: Simultaneously acquire surface remote sensing image data of the target sea area and profile observation data covering the vertical direction of the water column;

[0025] S2. Three-dimensional data fusion and reconstruction steps: The surface remote sensing image data and profile observation data are preprocessed and spatiotemporally matched. The continuous information on the horizontal plane provided by the surface remote sensing image is used as a guide. Through three-dimensional spatial interpolation or data assimilation technology, the discrete profile observation data is reconstructed into a continuous three-dimensional ecological environment data field.

[0026] S3. Three-dimensional feature engineering steps: Extract a three-dimensional feature set from the three-dimensional ecological environment data field, which contains information in both horizontal and vertical dimensions. The three-dimensional feature set is used to characterize phytoplankton biomass, taxonomic composition and its vertical stratification structure.

[0027] S4. AI Intelligent Stereoscopic Inversion Step: Input the stereoscopic feature set into the pre-trained phytoplankton stereoscopic monitoring artificial intelligence model; the model is trained to simultaneously learn horizontal distribution patterns and vertical distribution patterns, and output stereoscopic monitoring products including the three-dimensional distribution of chlorophyll a concentration, the vertical functional layer structure of phytoplankton, and the three-dimensional occurrence range of harmful algal blooms.

[0028] S5. Vertical Structure Analysis and Early Warning Steps: Analyze the vertical structure information in the three-dimensional monitoring product, including but not limited to: the depth change of the maximum chlorophyll layer, the degree of vertical separation of different groups, and the vertical migration trend of harmful algal blooms. When the vertical structure parameters deviate from the normal range or indicate ecological risks, generate a special early warning containing the vertical dimension.

[0029] S6. Three-dimensional visualization and output steps: The three-dimensional monitoring products and special early warnings are visualized and output in various forms such as three-dimensional rendering, vertical cross-sectional slices, and depth-time change curves.

[0030] Preferably, the features extracted in step S3 to characterize the vertical layered structure include: calculating the statistics (mean, standard deviation) of the feature parameters in different depth intervals, and calculating the vertical gradient, integral depth and centroid depth in the entire water column or a specific layer.

[0031] Preferably, the "vertical structure analysis and early warning step" in step S5 specifically includes:

[0032] S5a: Extract vertical profile data of chlorophyll a concentration and vertical distribution data of phytoplankton groups from the three-dimensional monitoring product;

[0033] S5b: Based on the vertical profile data, calculate a set of predefined vertical structure quantification parameters, which include at least: maximum chlorophyll layer depth, thickness and intensity, vertical heterogeneity index, vertical dominance depth and separation of phytoplankton groups, and vertical range of harmful algal blooms.

[0034] S5c: Compare the calculated real-time vertical structure parameters with historical climatological baselines and analyze their short-term trends;

[0035] S5d: Based on preset multi-rule logic and parameter anomalies or trends, trigger different types of vertical special warnings. The warning types include subsurface algal bloom risk warning, vertical structure drastic change warning, and community vertical structure anomaly warning.

[0036] S5e: Generates a vertical structure-specific analysis report that includes quantification parameters, comparison results, and early warning conclusions.

[0037] It has the following beneficial effects:

[0038] This invention, by constructing an integrated three-dimensional monitoring and intelligent analysis system encompassing air, space, and sea, achieves a comprehensive technological upgrade in phytoplankton monitoring, moving from two-dimensional to three-dimensional analysis, from surface phenomena to underlying mechanisms, and from post-monitoring to proactive early warning. It overcomes the dimensional limitations of traditional monitoring, and by integrating multi-source profile data with satellite remote sensing, employing physically constrained three-dimensional reconstruction technology, it has for the first time operationally generated a continuous and dynamic three-dimensional phytoplankton biomass distribution field. This significantly enhances the detection capability of concealed phenomena such as subsurface algal blooms, reducing monitoring blind spots by more than 70%, and enabling in-depth insight and precise early warning of ecological processes. The system can not only quantitatively monitor biomass but also analyze phytoplankton... The system analyzes the vertical distribution and separation of phytophyte functional groups and identifies subsurface algal bloom risks 24-72 hours in advance based on abnormal changes in structural parameters such as the maximum chlorophyll layer depth and vertical range expansion. The warning information is upgraded from "potential" to "forming at a specific depth." The system achieves full-chain automation and intelligence, requiring no human intervention from data fusion and AI 3D inversion to warning generation. The model also has online incremental learning capabilities, can adapt to environmental evolution, and provides an immersive 3D visualization platform to display the 3D structure of algal blooms. It outputs structured reports containing quantitative parameters and risk assessments, directly serving fisheries management, disaster prevention and mitigation, and ecological protection. Detailed Implementation

[0039] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] In a first embodiment, the present invention provides a technical solution: a method for monitoring marine phytoplankton based on artificial intelligence, comprising the following steps:

[0041] System hardware and data source configuration:

[0042] Satellite remote sensing data source: Deploy data interfaces to receive and download multispectral images from the Sentinel-3 OLCI sensor and sea surface temperature data from MODISAqua in real time. These data provide large-area, periodic surface information.

[0043] On-site three-dimensional monitoring network:

[0044] Profiling buoy array: Several Argo biogeochemical buoys (BGC-Argo) are deployed in key monitoring areas. Each buoy completes a profile measurement from the sea surface to a depth of 2,000 meters every 10 days and uploads vertical profile data including chlorophyll fluorescence, particulate backscattering, downlink irradiance, as well as temperature, salinity, and nitrate concentration.

[0045] Unmanned vessel patrol: During critical periods, unmanned vessels equipped with vertical profile pumping systems and CTD multi-parameter water quality meters are dispatched to conduct high-resolution (horizontal resolution 1 km, vertical resolution 1 m) mobile profile observations at specific cross sections.

[0046] Validation data: In collaboration with marine research vessels, stratified water samples are regularly collected in the warning area. Precise vertical distribution data of phytoplankton species composition and pigment concentration are obtained through laboratory microscopy and high-performance liquid chromatography (HPLC) for model training and validation.

[0047] 3D Data Fusion and Reconstruction:

[0048] The system data preprocessing and fusion module performs the following operations:

[0049] Atmospheric correction and Rayleigh scattering correction were performed on OLCI images to generate normalized water ionization radiance data, and spectral features related to phytoplankton were extracted.

[0050] Quality control was performed on the profile data of BGC-Argo and unmanned ships to remove outliers and calibrate the system bias between different sensors;

[0051] Core 3D Reconstruction: Using geocoded OLCI imagery as the horizontal reference field, co-kriging interpolation is employed for each 3D grid cell. Specifically, surface chlorophyll retrieved from satellites is used as a secondary variable, while chlorophyll at discrete depth points measured by BGC-Argo is used as the primary variable. During interpolation, the thermocline depth defined by temperature and salinity profiles is introduced as a spatially relevant structural function constraint to ensure that interpolation is performed within a physically reasonable layered structure. Ultimately, a spatiotemporally continuous 3D data field of chlorophyll a concentration and 3D data field of environmental parameters covering the target sea area are generated.

[0052] 3D Feature Engineering:

[0053] The feature extraction module extracts a 3D feature vector containing vertical information for each horizontal grid point from the above 3D data field, for example:

[0054] Vertical distribution morphology characteristics: Calculate the chlorophyll profile of the 0-100m water column at this point, and extract its centroid depth, maximum concentration layer depth (SCM depth), SCM layer thickness, and the average concentration ratio of the 0-20m and 20-50m layers.

[0055] Stratified environment characteristics: Extract the depth of the mixed layer, the intensity of the thermocline, and the depth of the nitrate thermocline at this location;

[0056] Surface remote sensing derived features: satellite reflectance band values, fluorescence line height (FLH), and sea surface temperature corresponding to this location;

[0057] Spatiotemporal context features: the SCM depth change trend of this point in the previous week, and the current average centroid depth of the 8 adjacent grid points (horizontal spatial context).

[0058] AI-powered 3D inversion model training and operation:

[0059] Model Architecture: A multi-task deep learning model is adopted, the core of which is a shared encoder used to extract deep features from the input stereo feature vector. The decoder is divided into three branches:

[0060] Regression branch: Output chlorophyll a concentration for every 5 meters from 0 to 100 meters, and reconstruct the vertical profile;

[0061] Classification branches: Output the relative dominance probability of major functional groups such as diatoms, dinoflagellates, and cyanobacteria in the surface layer, SCM layer, and subsurface layer;

[0062] Anomaly detection branch: Outputs the confidence level of "high-value algal bloom in the subsurface" or "vertical structural anomaly" for each grid point;

[0063] Model training: The training set was constructed using historical BGC-Argo data, geostationary satellite data and shipborne verification data from the East China Sea region over the past 5 years. The loss function was a weighted sum of the losses of the three branches, with the regression branch loss focusing on optimizing the inversion accuracy of the SCM layer.

[0064] Model execution: Input the processed stereo feature vector at the current time step into the trained model;

[0065] Vertical structure depth analysis and early warning generation:

[0066] The system dynamic evaluation module performs secondary analysis on the AI ​​output:

[0067] Automatic parameter extraction: The system traverses the three-dimensional grid, automatically identifies and calculates the core SCM parameters (depth, intensity, thickness) and vertical uniformity index of each effective water column;

[0068] Climatological Comparison and Trend Detection: The calculated SCM depth is compared with the "East China Sea Summer SCM Depth Climatological Distribution Map" generated based on historical data. At the same time, the moving average rate of change of SCM depth in the region over the past 72 hours is calculated.

[0069] Warning rule triggered:

[0070] Scenario A: In a certain sea area, SCM intensity > 5 mg / m³ is identified, and the SCM depth is between 20-40 meters (subsurface). At the same time, the surface temperature is high and the wind and waves are calm. The system compares the climate state and finds that the SCM in this area is usually shallow (<15 meters). Rule trigger: Generate "Subsurface High Biomass Algal Bloom Risk Warning (SSHB)" with a warning level of "yellow". It indicates that there may be a high concentration of algae that is not easily detected by satellite, which poses a potential risk to oxygen consumption in the middle water and the bottom ecosystem.

[0071] Scenario B: In another sea area, the AI ​​model identified a high probability of cyanobacteria aggregation on the surface, and the vertical distribution showed that it hardly descended and was concentrated in the 0-10 meter range. At the same time, satellites showed that the sea surface in this area had a typical serrated texture. The rule was triggered: a "Surface Cyanobacterial Bloom Warning" was generated, with a warning level of "orange", indicating the risk of direct impact on coastal tourism and mariculture.

[0072] Report generation: Automatically generates monitoring reports that include the latitude and longitude of the warning area, core depth, impact range, major algae, environmental driving factor analysis, and recommended measures for attention;

[0073] Model iterative optimization:

[0074] Once the vessel has completed verification sampling in the warning area, the system will import the actual vertical stratified sample data.

[0075] Accuracy assessment: Calculate the root mean square error (RMSE) between the chlorophyll profile retrieved by the model at this location and the measured profile, as well as the accuracy of taxonomic identification.

[0076] Incremental learning: If new data reaches a certain batch size and the overall evaluation shows that the model's performance is declining in a specific season or region, the system automatically starts the model fine-tuning process. It uses only the new data to train the last few layers of the model or specific task branches for a small number of epochs, quickly adapting to the new ecological conditions without retraining the entire large model.

[0077] Through the above specific implementation methods, the present invention realizes full-process automation and intelligence from multi-source three-dimensional data acquisition, intelligent three-dimensional inversion, to vertical structure depth analysis and early warning, which significantly improves the monitoring and early warning capabilities for marine phytoplankton ecological processes, especially their key vertical distribution dynamics.

[0078] Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art and related fields based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described and explained in the present invention, unless otherwise specified or limited, shall be implemented according to conventional means in the art.

Claims

1. A marine phytoplankton monitoring system based on artificial intelligence, characterized in that, include: The data acquisition module is used to acquire multi-source three-dimensional monitoring data of the target sea area, wherein the multi-source three-dimensional monitoring data includes: Surface optical remote sensing images; Vertical profile data of water bodies, including optical parameters and seawater physicochemical parameters related to phytoplankton at at least two different depths, collected by profile buoys, towed sensors, underwater unmanned vehicles or shipborne profile measurement systems. The data preprocessing and fusion module is communicatively connected to the data acquisition module and is used to preprocess and spatiotemporally register the surface optical remote sensing images and the water body vertical profile data, and to construct a three-dimensional standardized spatiotemporal grid dataset. The feature extraction and selection module is communicatively connected to the data preprocessing and fusion module, and is used to extract a multidimensional feature set related to phytoplankton biomass, species composition and vertical distribution structure from the three-dimensional standardized spatiotemporal grid dataset; An artificial intelligence analysis engine, communicatively connected to the feature extraction and selection module, includes: The trained phytoplankton three-dimensional monitoring model is used to output the three-dimensional distribution information of phytoplankton in the target sea area based on the multi-dimensional feature set. The three-dimensional distribution information includes: the three-dimensional distribution field of chlorophyll a concentration, the depth and thickness of the vertical dominance layer of the main phytoplankton functional groups, and the vertical range of the harmful algal bloom area. The dynamic early warning and visualization module is connected to the artificial intelligence analysis engine. It is used to generate early warning information based on the three-dimensional distribution information of the phytoplankton and combined with a preset vertical structure threshold, and to display the three-dimensional distribution information and early warning information in three dimensions.

2. The marine phytoplankton monitoring system based on artificial intelligence according to claim 1, characterized in that: The optical parameters related to phytoplankton included in the vertical profile data of the water body include, but are not limited to, chlorophyll fluorescence values ​​at different depths, beam attenuation coefficients, or chlorophyll a concentration profiles obtained by inverting the above parameters.

3. The marine phytoplankton monitoring system based on artificial intelligence according to claim 1, characterized in that: The data preprocessing and fusion module is specifically used in constructing a three-dimensional standardized spatiotemporal grid dataset for: The surface optical remote sensing image is used as a horizontal plane constraint. The discrete vertical profile data of the water body are fused into a three-dimensional mesh constrained by the horizontal plane through three-dimensional kriging interpolation, three-dimensional variational assimilation or machine learning interpolation methods to form a continuous three-dimensional data volume.

4. The marine phytoplankton monitoring system based on artificial intelligence according to claim 1, characterized in that: The features extracted by the feature extraction and selection module that are related to the vertical distribution structure include, but are not limited to: the maximum chlorophyll layer depth, the vertical heterogeneity index, the gradient changes of feature parameters at different depths, and the combination of features characterizing vertical stratification.

5. The marine phytoplankton monitoring system based on artificial intelligence according to claim 1, characterized in that: The phytoplankton three-dimensional monitoring model in the artificial intelligence analysis engine is trained by fusing surface remote sensing features and vertical profile features. Its architecture includes network structures for extracting vertical contextual information, such as one-dimensional convolutional neural network or recurrent neural network layers.

6. The marine phytoplankton monitoring system based on artificial intelligence according to claim 1, characterized in that: The dynamic early warning and visualization module is used for: Identify whether there are abnormal subsurface chlorophyll maximum values ​​in the three-dimensional distribution field of chlorophyll a concentration, and evaluate their depth, intensity and duration; Based on the vertical extent of the harmful algal bloom area, determine whether it has expanded from the surface to a depth that may affect water mixing or the underlying ecosystem; Generate a comprehensive ecological risk assessment report that includes vertical dimension information.

7. A method for monitoring marine phytoplankton based on artificial intelligence, characterized in that, Includes the following steps: S1. Multi-source stereo data acquisition steps: Simultaneously acquire surface remote sensing image data of the target sea area and profile observation data covering the vertical direction of the water column; S2. Three-dimensional data fusion and reconstruction steps: The surface remote sensing image data and profile observation data are preprocessed and spatiotemporally matched. The continuous information on the horizontal plane provided by the surface remote sensing image is used as a guide. Through three-dimensional spatial interpolation or data assimilation technology, the discrete profile observation data is reconstructed into a continuous three-dimensional ecological environment data field. S3. Three-dimensional feature engineering steps: Extract a three-dimensional feature set from the three-dimensional ecological environment data field, which contains information in both horizontal and vertical dimensions. The three-dimensional feature set is used to characterize phytoplankton biomass, taxonomic composition and its vertical stratification structure. S4. AI Intelligent Stereoscopic Inversion Step: Input the stereoscopic feature set into the pre-trained phytoplankton stereoscopic monitoring artificial intelligence model; the model is trained to simultaneously learn horizontal distribution patterns and vertical distribution patterns, and output stereoscopic monitoring products including the three-dimensional distribution of chlorophyll a concentration, the vertical functional layer structure of phytoplankton, and the three-dimensional occurrence range of harmful algal blooms. S5. Vertical Structure Analysis and Early Warning Steps: Analyze the vertical structure information in the three-dimensional monitoring product, including but not limited to: the depth change of the maximum chlorophyll layer, the degree of vertical separation of different groups, and the vertical migration trend of harmful algal blooms. When the vertical structure parameters deviate from the normal range or indicate ecological risks, generate a special early warning containing the vertical dimension. S6. Three-dimensional visualization and output steps: The three-dimensional monitoring products and special early warnings are visualized and output in various forms such as three-dimensional rendering, vertical cross-sectional slices, and depth-time change curves.

8. The method for monitoring marine phytoplankton based on artificial intelligence according to claim 7, characterized in that: The features extracted in step S3 to characterize the vertical layered structure include: calculating the statistics (mean, standard deviation) of the feature parameters in different depth intervals, and calculating the vertical gradient, integral depth and centroid depth in the entire water column or a specific layer.

9. A method for monitoring marine phytoplankton based on artificial intelligence according to claim 7, characterized in that: The "vertical structure analysis and early warning step" in step S5 specifically includes: S5a: Extract vertical profile data of chlorophyll a concentration and vertical distribution data of phytoplankton groups from the three-dimensional monitoring product; S5b: Based on the vertical profile data, calculate a set of predefined vertical structure quantification parameters, which include at least: maximum chlorophyll layer depth, thickness and intensity, vertical heterogeneity index, vertical dominance depth and separation of phytoplankton groups, and vertical range of harmful algal blooms. S5c: Compare the calculated real-time vertical structure parameters with historical climatological baselines and analyze their short-term trends; S5d: Based on preset multi-rule logic and parameter anomalies or trends, trigger different types of vertical special warnings. The warning types include subsurface algal bloom risk warning, vertical structure drastic change warning, and community vertical structure anomaly warning. S5e: Generates a vertical structure-specific analysis report that includes quantification parameters, comparison results, and early warning conclusions.