Method and system for early warning of water body cyanobacterial bloom

Abstract

The application discloses a water body cyanobacteria bloom early warning monitoring method and system, and relates to the technical field of early warning monitoring. Dynamic timing data of physicochemical parameters of a water body monitoring area is collected, and variable speed analysis is performed on the dynamic timing data of the physicochemical parameters to obtain water body physicochemical parameter change information of the water body monitoring area. Multidimensional physiological timing data of algae is synchronously monitored and collected, and event window feature extraction processing is performed to obtain a sequence of multidimensional feature vectors of algae response data. Bloom risk analysis is performed on the sequence of multidimensional feature vectors of algae response data of the water body monitoring area to obtain a cyanobacteria bloom risk assessment result of the water body monitoring area. According to the cyanobacteria bloom risk assessment result of the water body monitoring area, deduction analysis is performed to obtain cyanobacteria bloom risk early warning information of the water body monitoring area, and cyanobacteria bloom risk early warning is performed according to the cyanobacteria bloom risk early warning information. The potential cyanobacteria bloom risk can be identified in advance, and the accuracy and timeliness of cyanobacteria bloom early warning are improved.

Description

Technical Field

[0001] This invention relates to the field of early warning and monitoring technology, specifically to a method and system for early warning and monitoring of cyanobacterial blooms in water bodies. Background Technology

[0002] In the process of water body monitoring and management, early warning and monitoring of cyanobacterial blooms are key tasks to ensure water quality safety. Especially when facing drastic environmental changes, such as extreme weather or pollution events, the drastic fluctuations in the physical and chemical conditions of water bodies, such as rapid changes in water temperature, often trigger rapid changes in algal populations, especially the outbreak of harmful cyanobacteria.

[0003] Current technologies primarily rely on biomass indicators, such as chlorophyll a concentration and spectral characteristics, to assess cyanobacterial abundance and potential bloom risks in water bodies. To achieve precise monitoring and early warning of water health, systems need to analyze the physicochemical conditions of the water body and the physiological responses of algae in real time to identify potential bloom risks. However, existing monitoring methods are limited by their reliance on single biomass or pigment concentration indicators, failing to effectively capture the dynamic changes in algal populations caused by drastic changes in water physicochemical conditions. When water physicochemical conditions, such as water temperature, change rapidly, some algal populations may exhibit temporary dormancy, sedimentation, or other metabolic responses. These changes are not immediately reflected in traditional biomass or pigment concentrations. This means that if environmental conditions change drastically, existing methods cannot detect the physiological responses of algal populations during these changes, leading to the early warning system's inability to promptly detect potential risks. The monitoring system may misjudge the water body as safe, fail to accurately assess the risk of blooms, miss the optimal warning window, or even only discover the problem after a bloom has occurred, resulting in severe ecological damage. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for early warning and monitoring of cyanobacterial blooms in water bodies, which can effectively solve the problems mentioned in the background technology.

[0005] To achieve the above objectives, the first aspect of the present invention is implemented through the following technical solution: a method for early warning and monitoring of cyanobacterial blooms in water bodies, comprising collecting dynamic time-series data of physicochemical parameters in a water body monitoring area, performing variable-speed analysis on the dynamic time-series data of physicochemical parameters, and obtaining information on the changes in physicochemical parameters of the water body in the water body monitoring area.

[0006] The system receives information on changes in the physicochemical parameters of the water body in the monitoring area, simultaneously monitors and collects multidimensional physiological time-series data of algae in the monitoring area, and obtains a multidimensional feature vector sequence of algal response data in the monitoring area through event window feature extraction processing.

[0007] Algal bloom risk analysis was performed on the multidimensional feature vector sequence of algal response data in the water monitoring area to obtain the risk assessment results of cyanobacterial blooms in the water monitoring area.

[0008] Based on the risk assessment results of cyanobacterial blooms in the water monitoring area, extrapolation and analysis are conducted to obtain early warning information on the risk of cyanobacterial blooms in the water monitoring area, and early warning of cyanobacterial bloom risks is issued based on the early warning information.

[0009] Furthermore, the method for performing variable-speed analysis on the dynamic time-series data of physicochemical parameters is as follows: the dynamic time-series data of physicochemical parameters in the water monitoring area includes physicochemical parameters, the collection timestamps of the physicochemical parameters, and spatial location identifiers.

[0010] Based on the physicochemical parameters and their acquisition timestamps in the dynamic time series data of physicochemical parameters, the acceleration sequence of the changes in physicochemical parameters is obtained through analysis.

[0011] Based on the acceleration sequence of changes in physicochemical parameters, the changes in physicochemical parameters of the water body in the monitoring area can be determined.

[0012] Information on changes in the physicochemical parameters of the water body in the water monitoring area includes: event identifiers, event start times, event duration ranges, spatial location identifiers, key change parameter types, and acceleration sequences of key change parameters.

[0013] Furthermore, the method for monitoring and collecting multidimensional physiological time-series data of algae in the water monitoring area is as follows: receiving information on changes in the physicochemical parameters of the water in the water monitoring area, and analyzing the event start time, event duration range, and spatial location identifier of each physicochemical condition change event according to the event identifier.

[0014] When the acceleration sequence of changes in physicochemical parameters is determined to meet the conditions for physicochemical changes in water bodies, algal response data of the water body monitoring area are obtained.

[0015] The algal response data of the water body monitoring area are associated with the corresponding collection timestamps and spatial location identifiers, and arranged continuously in the order of collection timestamps to form the multidimensional physiological time-series data of algae in the water body monitoring area. The multidimensional physiological time-series data of algae includes the original measurement value sequence of algal response data, collection timestamps and spatial location identifiers.

[0016] Furthermore, the method for obtaining the multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing is as follows: receiving information on changes in the physicochemical parameters of the water body and multidimensional physiological time-series data of algae in the water monitoring area; locating and extracting the event-related algal physiological observation sequence corresponding to the event-related algal physiological observation sequence in the multidimensional physiological time-series data of algae based on the event start time and spatial location identifier of the physicochemical condition change event; and recording this as the event-related algal physiological observation sequence of the water monitoring area.

[0017] The event-related algal physiological observation sequence for the water body monitoring area includes the original measurement sequence of algal response data in the water body monitoring area within the preset physiological response observation window, as well as the collection timestamp and spatial location identifier.

[0018] Based on the event-related algal physiological observation sequence of the water body monitoring area, the response amplitude feature vector and time scale feature vector of algal response data in the event-related algal physiological observation sequence are extracted.

[0019] Based on the original measurement sequence of algal response data in the water monitoring area within the preset physiological response observation window, and the response amplitude feature vector and time scale feature vector of algal response data in the event-related algal physiological observation sequence, a multidimensional feature vector sequence of algal response data in the water monitoring area is constructed.

[0020] Furthermore, the method for conducting algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area is as follows: the multidimensional feature vector sequence of algal response data includes the original measurement value sequence of algal response data, the response amplitude feature vector sequence, and the time scale feature vector sequence.

[0021] The system receives multidimensional feature vector sequences of algal response data from the water monitoring area, performs a three-level risk analysis on these sequences, and obtains the physiological damage level, prototype of algal community physiological response, and risk score of cyanobacterial bloom from the algal response data in the water monitoring area. These data are then packaged into a risk assessment result for cyanobacterial bloom in the water monitoring area.

[0022] Furthermore, the method for performing three-level risk analysis on the multidimensional feature vector sequence of algal response data is as follows: the first level is physiological damage analysis, which extracts all the original measurement values ​​of algal response data from the original measurement value sequence of algal response data and analyzes to obtain the physiological damage level of algal response data in the water monitoring area.

[0023] The second level is physiological response prototype matching. Quantitative features are obtained from the original measurement sequence of algal response data. The obtained quantitative features are compared one by one with the feature conditions of the algal community physiological response prototype in the predefined physiological response prototype rule base to obtain the algal community physiological response prototype of the water body monitoring area.

[0024] The third level is dynamic feature verification and risk integral synthesis. Based on the physiological damage level of algal response data output from the first level and the algal community physiological response prototype obtained by matching with the second level, dynamic feature verification is performed by combining the response amplitude feature vector sequence and time scale feature vector sequence of algal response data in the multidimensional feature vector sequence of algal response data to generate the risk integral value of cyanobacterial bloom in the water monitoring area.

[0025] Furthermore, the method for obtaining early warning information on the risk of cyanobacterial blooms in the water monitoring area through extrapolation and analysis is as follows:

[0026] The system receives the risk assessment results of cyanobacterial blooms, compares the risk score of cyanobacterial blooms with the preset low-risk and medium-risk thresholds, and obtains the risk level of cyanobacterial blooms in the water monitoring area.

[0027] For the prototype of the physiological response of algal communities in the water monitoring area, combined with the inference knowledge rule set associated with the prototype of the physiological response of algal communities, the inference logic analysis of the prototype of the physiological response of algal communities is performed to obtain the inference output conclusions corresponding to the prototype of the physiological response of algal communities in the water monitoring area.

[0028] Based on the inferences and conclusions derived from the risk levels of cyanobacterial blooms and the prototypes of algal community physiological responses in the water monitoring area, the potential dominant physiological types in the water monitoring area were analyzed.

[0029] The risk levels and potential dominant physiological types of cyanobacterial blooms in the water monitoring area are integrated and packaged to generate early warning information on the risk of cyanobacterial blooms in the water monitoring area.

[0030] Furthermore, the method for obtaining the cyanobacterial bloom risk level of the water monitoring area is as follows: the cyanobacterial bloom risk score is compared with preset low-risk and medium-risk thresholds. If the cyanobacterial bloom risk score is lower than the low-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as low-risk; if the cyanobacterial bloom risk score is higher than or equal to the low-risk threshold and lower than the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as medium-risk; if the cyanobacterial bloom risk score is higher than or equal to the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as high-risk.

[0031] Furthermore, the method for issuing cyanobacterial bloom risk warnings based on cyanobacterial bloom risk warning information is as follows: receiving cyanobacterial bloom risk warning information for the water monitoring area; when the cyanobacterial bloom risk level is low, triggering the monitoring frequency increase warning mode; when the cyanobacterial bloom risk level is medium, sending risk attention reminder information to local management personnel and initiating the manual review and inspection process; when the cyanobacterial bloom risk level is high, sending an emergency warning notice containing specific spatial location identifiers and potential dominant physiological groups to multi-level joint management departments.

[0032] The second aspect of the present invention provides a water body cyanobacterial bloom early warning and monitoring system, comprising: a data variable speed analysis module, used to collect dynamic time series data of physicochemical parameters in the water body monitoring area, perform variable speed analysis on the dynamic time series data of physicochemical parameters, and obtain information on the changes in physicochemical parameters of the water body in the water body monitoring area.

[0033] The time-series data fitting and decomposition module is used to receive information on changes in the physicochemical parameters of the water body in the water monitoring area, synchronously monitor and collect multidimensional physiological time-series data of algae in the water monitoring area, and obtain a multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing.

[0034] The algal bloom risk analysis module is used to perform algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area, and obtain the cyanobacterial bloom risk assessment results of the water monitoring area.

[0035] The risk warning module is used to extrapolate and analyze the risk assessment results of cyanobacterial blooms in the water monitoring area to obtain cyanobacterial bloom risk warning information for the water monitoring area, and to issue cyanobacterial bloom risk warnings based on the cyanobacterial bloom risk warning information.

[0036] The present invention has the following beneficial effects:

[0037] This invention overcomes the technical bottleneck of traditional cyanobacterial bloom monitoring methods that rely excessively on static indicators such as chlorophyll a concentration or biomass. By introducing accelerated analysis of physicochemical parameter changes and an event-level triggering mechanism, it can quickly and reliably identify real-world environmental upheavals when key physicochemical conditions such as water temperature, pH, and dissolved oxygen undergo "accelerated changes," and simultaneously trigger monitoring of algal physiological responses. This method enables the system to capture the physiological stress and functional changes of algal communities before the algal population increases significantly, thereby identifying potential bloom risks in advance. It reduces the problem that existing technologies can only issue warnings after a bloom is approaching or has already formed, thus improving the timeliness of warnings.

[0038] This invention systematically transforms algal fluorescence kinetic physiological response data into response amplitude characteristics, time-scale characteristics, and their temporal evolution sequences. Based on this, it introduces a three-level risk analysis mechanism: physiological damage degree analysis, physiological response prototype matching, and dynamic feature verification. This mechanism can not only determine whether algae are in an abnormal state, but also clarify the abnormal physiological type, response mechanism, and the rationality of its evolution. Through prototype rule base and dynamic consistency verification, the system can effectively distinguish between short-term recoverable photoprotective responses and tolerance adaptation processes with algal bloom potential, avoiding misjudgments caused by transient noise or similar surface features, and improving the scientificity, stability, and biological interpretability of the risk assessment results.

[0039] Based on risk score assessment, this invention further introduces physiological response prototype deduction and population dynamics simulation mechanisms, combining the currently monitored physiological state of algae with future succession trends. This enables the identification of potential dominant physiological types before algal blooms occur and the prediction of their competitive advantage evolution direction. By structurally encapsulating risk levels, potential dominant physiological types, and spatial information and linking them with hierarchical early warning strategies, the invention enhances the guiding value of early warning results for management decisions, realizing the transformation from passive response-based monitoring to proactive prediction-based early warning. This comprehensively improves the intelligence level and practical application value of the cyanobacterial bloom control system in water bodies. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0041] Figure 2 This is a schematic diagram of the system module connections of the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0043] Please see Figure 1 As shown, the first aspect of the present invention provides a technical solution: a method for early warning and monitoring of cyanobacterial blooms in water bodies, including collecting dynamic time-series data of physicochemical parameters in a water body monitoring area, performing variable-speed analysis on the dynamic time-series data of physicochemical parameters, and obtaining information on the changes in physicochemical parameters of the water body in the water body monitoring area.

[0044] The method for collecting dynamic time-series data of physicochemical parameters in the water monitoring area is as follows:

[0045] Water quality monitoring units are deployed in the water monitoring area to collect physicochemical parameters of the water monitoring area. These physicochemical parameters include, but are not limited to, water temperature, pH value, and dissolved oxygen concentration. The water quality monitoring units transmit these physicochemical parameters to a data preprocessing unit. The data preprocessing unit receives the physicochemical parameters of the water monitoring area and labels all parameters with corresponding collection timestamps and spatial location identifiers to form dynamic time-series data of the physicochemical parameters that are continuous in time. The dynamic time-series data of the physicochemical parameters of the water monitoring area includes the physicochemical parameters, their collection timestamps, and spatial location identifiers.

[0046] This implementation plan establishes a foundation for continuous data acquisition with multiple parameters and high temporal resolution, providing raw data streams for subsequent capture of transient and drastic environmental changes, and reducing the limitations of single-parameter monitoring in continuously analyzing dynamic environmental processes.

[0047] It should be added that a water quality monitoring unit refers to a device used to collect physicochemical parameters such as water temperature, pH value and dissolved oxygen concentration in a water body monitoring area in real time, and transmit the collected data to a data preprocessing unit. The data preprocessing unit is a device that receives the raw data stream from the water quality monitoring unit, performs data cleaning and data alignment processing, and converts the raw data into dynamic time-series data of physicochemical parameters that are continuous in time and have clear spatial location identification for subsequent analysis.

[0048] The method for variable-speed analysis of dynamic time-series data of physicochemical parameters is as follows:

[0049] Based on the acquisition timestamp order of the physicochemical parameters in the dynamic time-series data, the ratio of the absolute difference between the current physicochemical parameter and the adjacent previous physicochemical parameter to the absolute difference of the acquisition timestamp is recorded as the instantaneous change rate of the corresponding physicochemical parameter. Based on the instantaneous change rate of the physicochemical parameters within a preset historical data rolling window, the central difference method is used to calculate the change acceleration of the corresponding physicochemical parameter. The change acceleration of the physicochemical parameters within the historical data rolling window is continuously statistically analyzed using a sliding update method, forming a time-evolving sequence of change acceleration of the physicochemical parameters. This sequence includes, but is not limited to, the change acceleration sequences of water temperature, pH, and dissolved oxygen concentration. The length of the preset historical data rolling window is determined based on the normal fluctuation period and sampling frequency of the target water body's physicochemical parameters. When the sampling frequency is on the minute level, the preset historical data rolling window covers data for at least one diurnal fluctuation period. When the sampling frequency is on the second level, the second-level data is first consolidated into minute-level statistical values ​​according to a preset statistical interval before constructing the preset historical data rolling window. The preset historical data scrolling window updates as new data is collected and deletes the data corresponding to the earliest collection time in the window.

[0050] In this implementation plan, acceleration is introduced as the core analytical indicator, which can sensitively identify the turning point of drastic change in environmental parameters that is not only changing but also accelerating. Traditional threshold alarms or rate of change monitoring are insufficient in distinguishing between gradual and drastic changes, while calculating acceleration can effectively amplify the signal characteristics of drastic events. Acceleration is calculated using the central difference method, and the statistical benchmark is dynamically updated based on a sliding historical window, so that the judgment criteria can adapt to the daily rhythm and long-term trend of the water body, thereby distinguishing true abnormal drastic changes from normal periodic fluctuations, such as diurnal temperature range.

[0051] The method for determining the changes in physicochemical parameters of the water body in the monitoring area based on the acceleration sequence of changes in physicochemical parameters is as follows:

[0052] The system sequentially scans the acceleration of each physicochemical parameter in the acceleration sequence. When the absolute value of the acceleration of a certain physicochemical parameter is higher than the corresponding dynamic threshold for acceleration, the corresponding physicochemical parameter is identified as an anomaly candidate and a continuous counting state is initiated. If, in the acceleration sequence of the corresponding physicochemical parameter, the absolute value of the acceleration of multiple consecutive physicochemical parameters following the anomaly candidate is consistently higher than the corresponding dynamic threshold for acceleration, and the total number of consecutive anomaly candidate points is higher than the threshold for consecutive exceedances, then the acceleration sequence of the corresponding physicochemical parameter is determined to meet the conditions for physicochemical changes in the water body and is recorded as a physicochemical condition change event in the water monitoring area. Otherwise, the acceleration sequence of the corresponding physicochemical parameter is determined to not meet the conditions for physicochemical changes in the water body.

[0053] This implementation scheme achieves robust and low-false-alarm intelligent identification of drastic events. By replacing fixed thresholds with dynamic judgment thresholds, the system can adapt to the differences in background fluctuations in different seasons and water areas. The introduction of a threshold for the number of consecutive exceedances requires that the abnormal signal must last for a period of time. This utilizes the characteristics of real environmental drastic events, such as the passage of cold fronts and the arrival of pollution clouds, which are continuous, while equipment noise or transient interference is mostly isolated. This spatiotemporal persistence verification logic improves the reliability of event judgment.

[0054] It should be explained that the specific method for setting the dynamic judgment threshold for the acceleration change of physicochemical parameters and the threshold for the number of consecutive exceedances is as follows: In the sequence of acceleration changes of physicochemical parameters, the average value and standard deviation of the acceleration changes for the corresponding physicochemical parameter are statistically analyzed. The sum of the average value and standard deviation of the acceleration changes is recorded as the dynamic judgment threshold for the acceleration change of the corresponding physicochemical parameter. The threshold for the number of consecutive exceedances is determined by those skilled in the art through analysis of the sequence characteristics of historical false alarm events and valid events. Its value is intended to ensure the robustness of event judgment and avoid false triggering due to single-point noise or instantaneous interference. For example, in a typical embodiment, those skilled in the art have found through retrospective analysis that real events such as changes in the thermocline of water bodies usually cause the acceleration change to be abnormally high more than 3 times consecutively, while random noise rarely lasts more than 2 times. Therefore, the threshold for the number of consecutive exceedances is set to 3. This threshold for the number of consecutive exceedances can be adaptively modified according to the actual sampling frequency and the tolerance for the risk of missed alarms.

[0055] If the acceleration sequence of the corresponding physicochemical parameter is determined to meet the conditions for physicochemical change in water, the corresponding physicochemical parameter is selected as the key change parameter type of the physicochemical condition change event, and the start time, spatial location identifier, and acceleration sequence of the key change parameter of the physicochemical condition change event are recorded; after completing the scanning of the acceleration sequence of all physicochemical parameters, the physicochemical parameter change information of the water body in the water monitoring area is recorded.

[0056] In one specific embodiment, water temperature is selected as the key parameter type for the physicochemical condition change event. The start time of the physicochemical condition change event is recorded as 10:30, the spatial location is identified as station 1, and the acceleration of the water temperature change at the start time of the event is +0.18℃ / s. 2 .

[0057] Information on changes in the physicochemical parameters of the water body in the water monitoring area includes: event identifiers, event start times, event duration ranges, spatial location identifiers, key change parameter types, and acceleration sequences of key change parameters.

[0058] In a specific embodiment, the event identifier of a certain physicochemical condition change event included in the water body physicochemical parameter change information of the water body monitoring area is an automatically generated unique ID. The event start time is 10:30 AM, the event duration is from 10:30 AM to 10:45 AM, the spatial location identifier is station 1, the key change parameter type is water temperature value, and the acceleration sequence of the water temperature value change is [+0.18℃ / s², +0.22℃ / s², +0.25℃ / s², +0.20℃ / s²]. This water body physicochemical parameter change information will be used to trigger subsequent targeted algal physiological response monitoring.

[0059] If the acceleration sequence of the corresponding physicochemical parameter change is determined to not meet the conditions for physicochemical change in water, then the physicochemical parameter change information of the water body monitoring area will be marked as no change.

[0060] Event identifiers are used to achieve event-level indexing and association during the subsequent synchronous triggering and fusion analysis of algal multidimensional physiological time-series data. The event start time and event duration range are used to limit the intercept window of the algal multidimensional dynamic response characteristic curve. Spatial location identifiers are used to ensure the consistency of water body physicochemical parameter change events and algal multidimensional physiological time-series data in the spatial dimension. Key change parameter types and corresponding change acceleration sequences are used to describe the key driving characteristics of this physicochemical condition change event.

[0061] In this implementation plan, the process from signal recognition to encapsulation of structured event information is completed, generating an accurate event diagnosis report, rather than just an alarm signal. It contains all the key information of the event, providing accurate spatiotemporal coordinates and analytical focus for subsequent steps. This enables the system to accurately trigger algal physiological monitoring and ensures that subsequent physiological response data can be rigorously correlated with this specific physicochemical disturbance event.

[0062] The system receives information on changes in the physicochemical parameters of the water body in the monitoring area, simultaneously monitors and collects multidimensional physiological time-series data of algae in the monitoring area, and obtains a multidimensional feature vector sequence of algal response data in the monitoring area through event window feature extraction processing.

[0063] This implementation plan enables high spatiotemporal resolution monitoring of the immediate physiological responses of algal communities driven by changes in physicochemical conditions. It can capture rapid physiological responses that are difficult to reflect by traditional biomass indicators, which helps to identify potential risks of cyanobacterial blooms in advance. Through event-driven synchronous triggering, the monitoring system can focus on the critical time window after drastic physicochemical changes, thereby improving monitoring efficiency and sensitivity.

[0064] The method for synchronously monitoring and collecting multidimensional physiological time-series data of algae in the water monitoring area is as follows:

[0065] The system receives information on changes in the physicochemical parameters of the water body in the monitoring area. It analyzes the start time, duration range, and spatial location of each physicochemical change event according to the event identifier. When the acceleration sequence of the physicochemical parameter changes is determined to meet the conditions for physicochemical change in the water body, a synchronization trigger command is sent to the fluorescence dynamics monitoring unit deployed in the corresponding water body monitoring area and consistent with the corresponding spatial location identifier. This helps to identify the immediate physiological response of algae to sudden environmental changes and effectively reduces the problem that traditional timed sampling may miss the critical response window.

[0066] It should be added that the fluorescence dynamics monitoring unit is a device used to immediately apply a specific photoexcitation sequence to the water body after receiving a synchronization trigger command, and simultaneously and at high speed collect algal response data generated by the algal community.

[0067] The fluorescence dynamics monitoring unit starts immediately upon receiving the synchronization trigger command and acquires algal response data of the water monitoring area. The algal response data includes maximum quantum efficiency, actual quantum yield value, non-photochemical quenching coefficient, initial slope of the fast photoresponse curve, and half-saturated light intensity value.

[0068] It should be explained that the maximum quantum efficiency refers to the maximum potential efficiency of the algal community in converting absorbed light energy into chemical energy. It is calculated by measuring the minimum and maximum fluorescence intensity recorded when saturated flash is applied after dark adaptation using a fluorescence dynamics monitoring unit, and dividing the difference between the maximum and minimum values ​​by the maximum value. The actual quantum yield value refers to the actual light energy conversion efficiency of the algal community under illumination conditions. It is calculated by measuring the steady-state fluorescence intensity recorded during continuous photochemical light irradiation and the maximum fluorescence intensity recorded when saturated flash is applied using a fluorescence dynamics monitoring unit, and dividing the difference between the maximum fluorescence intensity value and the steady-state fluorescence intensity value by the maximum fluorescence intensity value.

[0069] The non-photochemical quenching coefficient represents the loss of photosynthetic efficiency in algal communities under illumination due to photodamage or photoadaptation. It is calculated by measuring the maximum fluorescence intensity after dark adaptation and under illumination using a fluorescence kinetics monitoring unit. The ratio of the maximum fluorescence intensity after dark adaptation to the maximum fluorescence intensity under illumination is calculated, and the logarithm of this ratio (base 10) is obtained to obtain the non-photochemical quenching coefficient. The initial slope of the fast light response curve refers to the initial response rate of photosynthetic efficiency of algal communities under low light conditions as light intensity increases. The fluorescence kinetics monitoring unit simultaneously records the corresponding actual quantum yield values ​​when a series of photochemical lights with increasing intensity are applied. A linear fit is performed between the actual quantum yield values ​​in the low light intensity stage and the corresponding photochemical light intensity data; the slope of the fitted line is the initial slope of the fast light response curve. The half-saturation light intensity value refers to the light intensity required for the photosynthetic rate of the algal community to reach half of its maximum potential. The half-saturation light intensity value is calculated by dividing the maximum quantum efficiency by the initial slope of the fast light response curve obtained from the fluorescence kinetics monitoring unit.

[0070] The algal response data of the water body monitoring area are associated with the corresponding collection timestamps and spatial location identifiers, and arranged continuously in the order of collection timestamps to form the multidimensional physiological time-series data of algae in the water body monitoring area. This data is used to characterize the real-time physiological response state and dynamic evolution characteristics of algal communities under the drive of physicochemical condition changes. The multidimensional physiological time-series data of algae includes the original measurement value sequence of algal response data, collection timestamps and spatial location identifiers.

[0071] The method for obtaining the multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing is as follows:

[0072] It receives information on changes in the physicochemical parameters of the water body in the water monitoring area and multidimensional physiological time-series data of algae. Based on the event start time and spatial location identifier of the physicochemical condition change event, it locates and extracts all multidimensional physiological time-series data of algae in the corresponding spatial location and whose collection timestamp is later than the event start time from the multidimensional physiological time-series data of algae.

[0073] Using the event start time as the zero point, the extracted algal multidimensional physiological time-series data are re-aligned according to the collection timestamp. Then, all algal multidimensional physiological time-series data within the preset physiological response observation window, starting from the zero point, are extracted to form an event-related algal physiological observation sequence corresponding to the physicochemical condition change event. This sequence is denoted as the event-related algal physiological observation sequence of the water body monitoring area. The event-related algal physiological observation sequence of the water body monitoring area includes the original measurement value sequence of algal response data in the water body monitoring area within the preset physiological response observation window, the collection timestamp, and the spatial location identifier. This step transforms the dynamic physiological response into a standardized observation sequence, enabling direct comparison of data from different events and monitoring stations, facilitating statistical analysis and feature extraction.

[0074] Based on the event-related algal physiological observation sequence of the water body monitoring area, the response amplitude feature vector and time scale feature vector of algal response data in the event-related algal physiological observation sequence are extracted.

[0075] The method for extracting the response amplitude feature vector and time scale feature vector from the algal response data in the event-related algal physiological observation sequence is as follows:

[0076] The system receives event-related algal physiological observation sequences from the water monitoring area. For each algal response data point in the event-related algal physiological observation sequence, segmented statistical processing is performed within a preset physiological response observation window. The preset physiological response observation window is divided into three continuous and non-overlapping statistical sub-windows of equal duration according to time sequence, labeled as the first time sub-window, the second time sub-window, and the third time sub-window, respectively, to ensure the comparability of each time sub-window on a time scale. The arithmetic mean of all original measurements of the algal response data within the first, second, and third time sub-windows is calculated. The three arithmetic mean values ​​are then arranged and combined according to the time sequence of the first, second, and third time sub-windows to construct the response amplitude feature vector of the corresponding algal response data. Each dimension of the response amplitude feature vector represents the average response intensity level of the algal community in the early, middle, and late stages of the physiological response observation window, respectively.

[0077] Based on the response change relationship between adjacent statistical sub-windows, time scale features are extracted; the absolute value of the difference between the arithmetic mean of the algal response data in the first time sub-window and the corresponding arithmetic mean in the second time sub-window is calculated, and the absolute value of the difference is divided by the time span of the corresponding time sub-window to obtain the first rate of change; further, the absolute value of the difference between the arithmetic mean of the algal response data in the second time sub-window and the corresponding arithmetic mean in the third time sub-window is calculated, and the absolute value of the difference is divided by the time span of the corresponding time sub-window to obtain the second rate of change.

[0078] The first and second rates of change are arranged and combined in chronological order to construct a time-scale feature vector corresponding to the algal response data. Each dimension of the time-scale feature vector is used to characterize the average rate of change of the algal community between adjacent physiological response stages, thereby reflecting the speed of algal physiological response evolution over time.

[0079] The original measurement values, response amplitude feature vectors, and time-scale feature vectors corresponding to each algal response data in the event-related algal physiological observation sequence are aligned and combined according to a unified time reference to construct a multidimensional feature vector sequence of algal response data in the water monitoring area that evolves over time. The multidimensional feature vector sequence of algal response data includes the original measurement value sequence, response amplitude feature vector sequence, and time-scale feature vector sequence of algal response data. This provides a complete and operable data foundation for subsequent risk integral calculation, prototype matching, and dynamic evolution analysis, thereby improving the accuracy and reliability of algal bloom risk early warning.

[0080] The multidimensional feature vector sequence of algal response data is used to characterize the intensity, rate and temporal evolution of the multidimensional physiological response of algal communities under the drive of physicochemical condition change events, and to identify and associate the corresponding physicochemical condition change events.

[0081] This implementation plan clarifies the core role of multidimensional feature vector sequences, enabling a deep binding between physicochemical changes and algal physiological responses. This provides a complete data foundation for subsequent risk analysis, overcoming the limitations of traditional monitoring that only focuses on biomass and cannot establish a correlation between drivers and responses. This allows for the accurate identification of latent algal population dynamics caused by physicochemical changes, avoiding delayed or misjudged warnings. Through event identification association, each set of feature vector sequences can be traced back to the corresponding physicochemical change event. The response intensity, rate, and other features contained within can directly quantify the algal adaptation status to the event, providing a basis for determining whether the population is in a dormant, metabolically adjusted, or other risky state, facilitating subsequent traceability analysis.

[0082] Algal bloom risk analysis was performed on the multidimensional feature vector sequence of algal response data in the water monitoring area to obtain the risk assessment results of cyanobacterial blooms in the water monitoring area.

[0083] Cyanobacteria are a type of photosynthetic bacteria widely distributed in water bodies. They possess blue-green pigments and are capable of photosynthesis, using sunlight to synthesize organic matter. A cyanobacterial bloom refers to the massive proliferation of cyanobacteria in a water body, forming a suspended algal community. This phenomenon typically occurs in eutrophic water bodies, especially when the water is rich in nutrients such as nitrogen and phosphorus. Cyanobacterial blooms not only affect water quality but can also pose threats to aquatic ecosystems and public health.

[0084] The method for conducting algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area is as follows:

[0085] The system synchronously receives multidimensional feature vector sequences of algal response data from the water monitoring area, performs three-level risk analysis on these sequences, and improves the robustness and interpretability of risk assessment under complex environmental disturbance conditions.

[0086] The method for performing a three-level risk analysis on the multidimensional feature vector sequence of algal response data is as follows:

[0087] The first level is physiological damage analysis. All original measurements of the algal response data are extracted from the raw measurement sequence, normalized, and then averaged to obtain the mean value of the original measurements. A baseline fluctuation range is set. If the mean value is within the baseline fluctuation range, the corresponding algal response data physiological damage level in the water monitoring area is marked as no physiological damage. If the mean value exceeds the baseline fluctuation range, but the deviation from the historical average value is less than a preset threshold, the algal response data physiological damage level is marked as moderate physiological damage. If the mean value deviates from the historical average value by a factor greater than or equal to the preset threshold, the algal response data physiological damage level is marked as high physiological damage. Using the statistical baseline formed during historical calm periods, it is quickly determined whether the current algal population has deviated from the normal physiological fluctuation range. This enables rapid screening of the immediate physiological stress of algae caused by sudden changes in physicochemical conditions, allowing early damage or stress states of the algal population to be identified before algal blooms form, thus compensating for the inadequacy of relying solely on biomass indicators to reflect functional damage.

[0088] It should be added that the method for setting the baseline fluctuation range is as follows: Those skilled in the art conduct normality tests and distribution characteristic analyses on a large amount of historical calm period data. Based on the mean and standard deviation of the historical calm period data, a fluctuation range is constructed. In one specific implementation, based on statistical analysis experience and referring to the cumulative probability characteristics of the standard normal distribution, the mean of the historical calm period data plus 1.5 times the standard deviation is used as the upper limit of the baseline fluctuation range, and the mean of the historical calm period data minus 1.5 times the standard deviation is used as the lower limit of the baseline fluctuation range. Setting the multiplier to 1.5 can balance monitoring sensitivity and false alarm rate. The original measured values ​​of the corresponding algal response data and the historical calm period data are calculated. The absolute difference between the average values ​​is then divided by the standard deviation of the historical calm period data. The resulting ratio is the magnitude of the deviation from the average value of the historical calm period data. The magnitude of the deviation from the average value of the historical calm period data is used to quantitatively characterize the degree of deviation of the current observation value from the historical normal fluctuation range. The magnitude threshold is a boundary value set based on the correlation analysis of the degree of deviation of physiological parameters and the final ecological risk consequences in historical physiological response event cases. In one specific implementation, those skilled in the art, by analyzing historical data, set the case where the deviation reaches 2.0 times the standard deviation of the historical calm period data as the magnitude threshold for distinguishing between moderate and high physiological damage.

[0089] The second level is physiological response prototype matching. Quantitative features are analyzed from the raw measurement sequence of algal response data. These features include: a first quantitative feature and a second quantitative feature. The raw measurement sequence of the non-photochemical quenching coefficient is extracted, and a linear fitting curve for the non-photochemical quenching coefficient is obtained by fitting the raw measurement value of the non-photochemical quenching coefficient with time as the independent variable and the raw measurement value of the non-photochemical quenching coefficient as the dependent variable. The slope of the linear fitting curve for the non-photochemical quenching coefficient is denoted as the first quantitative feature. The raw measurement sequence of the maximum quantum efficiency is extracted, and the arithmetic mean of the raw measurement value of the maximum quantum efficiency within the first and second preset analysis windows is calculated to obtain the first and second arithmetic mean of the maximum quantum efficiency. The difference between the second and first arithmetic mean of quantum efficiency is divided by the first arithmetic mean of maximum quantum efficiency to obtain the recovery ratio of maximum quantum efficiency. This recovery ratio of maximum quantum efficiency is denoted as the second quantitative feature. The quantitative feature obtained from the analysis is compared with the feature conditions of the algal community physiological response prototype in the predefined physiological response prototype rule base to obtain the algal community physiological response prototype of the water body monitoring area. This maps the continuously changing physiological response process into response types with clear ecological meaning, and can distinguish between short-term photosynthetic protection response, continuous photosynthetic inhibition state, and tolerant energy redistribution state, thereby avoiding misjudging algal responses with recovery potential as high-risk precursors of algal blooms.

[0090] It should be added that the method for constructing the physiological response prototype rule base is as follows: In a laboratory environment, representative cyanobacterial species are cultured individually using a standard algal culture device. Predefined types and intensities of standard stresses are applied to each culture. These standard stresses include at least sudden temperature rise, sudden temperature drop, sudden pH rise, sudden pH drop, and sudden dissolved oxygen concentration drop. At the start of each standard stress application, a fluorescence kinetic monitoring device is activated to record the algal response data of the tested algal population throughout the preset observation period. All data points with extracted features are then analyzed based on the known ecological attributes of their source algal species and the conclusions of stress response literature. The patterns exhibiting a rapid and continuous increase in the non-photochemical quenching coefficient and a significant subsequent recovery of the maximum quantum efficiency are classified as energy redistribution-tolerant prototype candidate cases. Patterns exhibiting a continuous and synchronous decrease in both maximum quantum efficiency and actual quantum yield are classified as photosynthetic inhibition-prototype candidate cases. Then, for each type of prototype candidate case set, the prototype name, complete feature condition logic (including the conditions that must be met simultaneously, clear comparison relationships, and specific thresholds) are stored as a complete rule entry in the physiological response prototype rule base, thereby completing the construction of the physiological response prototype rule base.

[0091] The algal community physiological response prototypes include, but are not limited to, photoprotection-initiated prototypes, photosynthetic inhibition-persistent prototypes, and energy redistribution-tolerant prototypes. Taking the energy redistribution-tolerant prototype as an example, the characteristic conditions of the energy redistribution-tolerant prototype are: the first quantitative characteristic, namely, the slope of the linear fitting curve of the non-photochemical quenching coefficient is positive and greater than the slope threshold; the second quantitative characteristic, namely, the recovery ratio of the maximum quantum efficiency is higher than the ratio threshold. When the quantitative characteristics obtained from the analysis of the original measurement value sequence of algal response data simultaneously meet the characteristic conditions of the corresponding energy redistribution-tolerant prototype, it is determined that the algal community physiological response prototype of the water monitoring area matches the corresponding energy redistribution-tolerant prototype in the physiological response prototype rule base.

[0092] It should be explained that the slope threshold and proportion threshold involved in the characteristic conditions of the energy redistribution tolerance prototype are set after analyzing the statistical distribution of the first and second quantitative characteristics of all cases in the historically verified energy redistribution tolerance event case library. In one specific implementation, those skilled in the art set the slope threshold as the median of the positive value sequence of the slope of the linear fitting curve of the non-photochemical quenching coefficient in the first quantitative characteristic of the historical cases, and set the proportion threshold as the median of the maximum quantum efficiency recovery proportion sequence in the second quantitative characteristic of the historical cases.

[0093] The third level is dynamic feature verification and risk integral synthesis. Based on the physiological damage level of algal response data output from the first level and the algal community physiological response prototype obtained by matching with the second level, dynamic feature verification is performed by combining the response amplitude feature vector sequence and time scale feature vector sequence of algal response data in the multidimensional feature vector sequence of algal response data. This generates the risk integral value of cyanobacterial bloom, which dynamically verifies the evolutionary continuity and internal consistency of algal physiological response. It can distinguish between the evolution of real ecological processes and spurious changes caused by monitoring noise and local disturbances, further reducing the probability of misjudgment.

[0094] Based on the physiological damage level of algal response data and the prototype of algal community physiological response, the corresponding basic risk score is obtained from the pre-set basic risk score mapping table. The basic risk score is used to characterize the initial intensity level of cyanobacterial bloom risk without considering dynamic evolution consistency.

[0095] It should be added that the essence of the basic risk score mapping table is to map the damage level of physiological state and response prototype to a quantitative initial risk value. The mapping relationship is based on the consensus of algal physiological ecology. For example, high physiological damage means that the algal community has been stressed and its own stability has decreased; the energy redistribution tolerance prototype indicates that the algal community has initiated a strong photoprotection and repair mechanism and may have a stronger competitive advantage in the later stage of stress. The combination of the two indicates a high-risk population replacement potential. The construction of the basic risk score mapping table relies on historical monitoring data. Based on the above principles, domain experts set the initial risk score value for the combination of damage level and prototype.

[0096] The basic risk score mapping table includes at least a physiological damage level field, an algal community physiological response prototype field, and a basic risk score value field. When called, the physiological damage level of the algal response data output from the first level is used as the first search key, and the algal community physiological response prototype output from the second level is used as the second search key to retrieve a unique corresponding basic risk score value in the basic risk score mapping table. When multiple candidate basic risk score values ​​are retrieved, the basic risk score value that matches the current key change parameter type is selected as the output.

[0097] The method for performing dynamic feature verification is as follows:

[0098] Dynamic feature verification includes two parallel verification rules: feature sequence stability evaluation rule and prototype consistency verification rule.

[0099] In the characteristic sequence stability evaluation rule, for the response amplitude characteristic vector sequence of algal response data, the Euclidean distance between time-adjacent response amplitude characteristic vectors is calculated sequentially, and the standard deviation of all Euclidean distances is calculated. The standard deviation of all Euclidean distances is used as the adjacent variation dispersion index of the response amplitude characteristic vector sequence. For the time-scale characteristic vector sequence, the same processing method as the response amplitude characteristic vector sequence is adopted to obtain the adjacent variation dispersion index of the time-scale characteristic vector sequence. The adjacent variation dispersion index of the response amplitude feature vector sequence and the adjacent variation dispersion index of the time scale feature vector sequence are compared with their corresponding dispersion judgment thresholds. When both adjacent variation dispersion indices are lower than their corresponding dispersion judgment thresholds, it is determined that the physiological dynamic response of the algal community exhibits continuous, stable, and abrupt evolutionary characteristics without obvious mutations in the time dimension, and the confidence support flag is set to high. When any adjacent variation dispersion index is higher than or equal to its corresponding dispersion judgment threshold, it is determined that the physiological dynamic response of the algal community exhibits stage-based disorder or sudden jump characteristics, and the confidence support flag is set to low. By calculating the Euclidean distance between temporally adjacent feature vectors and statistically analyzing their dispersion, an adjacent variation dispersion index is constructed to quantify the continuity of multidimensional features over time. This adjacent variation dispersion index can reflect whether the evolution of physiological response conforms to biological regulatory laws from both the overall variation amplitude and stability aspects.

[0100] It should be explained that the dispersion threshold is determined by those skilled in the art through statistical analysis of adjacent variation dispersion indices of corresponding characteristic sequences in a large number of historical stationary evolutionary event cases. The dispersion threshold is typically set as the average of adjacent variation dispersion indices in historical cases plus N times the standard deviation, where N is a coefficient chosen based on the requirements for data continuity and stability. For example, in embodiments aimed at identifying stable physiological processes, those skilled in the art, by analyzing historical data, set N to 1, thereby setting this dispersion threshold as the historical average plus one standard deviation. This dispersion threshold can be adaptively modified according to the physiological fluctuation characteristics of different algal communities or the data quality of different monitoring periods.

[0101] In the prototype consistency verification rule, based on the algal community physiological response prototype obtained from the second-level matching, key dynamic features corresponding to the algal community physiological response prototype are extracted from the time-scale feature vector sequence and response amplitude feature vector sequence of the algal response data. Consistency comparison is performed according to predefined logical conditions. By jointly comparing the key dynamic features in the time and intensity dimensions, it is verified whether the actual evolution process conforms to the internal logic of the matched prototype, thus preventing mismatches caused by static feature similarity alone.

[0102] Taking the energy redistribution tolerant prototype as an example, in the time-scale feature vector sequence of algal response data, the corresponding dimension used to characterize the rate of change in physiological adaptation in the later stage is selected as the feature index of the long-term adaptation process. The arithmetic mean of this time-scale feature in the early stage and the later stage of the physiological response observation window is calculated respectively. When the average value in the later stage time window is strictly higher than the average value in the early stage time window, it is determined that the long-term adaptation process shows an increasing trend with the evolution of the event, indicating that the algal community gradually establishes a stable adaptive regulation mechanism under the continuous physicochemical disturbance.

[0103] In the response amplitude feature vector sequence of algal response data, the corresponding dimension used to characterize the mid-term response intensity level is selected as the energy redistribution intensity characterization feature, and the arithmetic mean of this response amplitude feature is calculated over the entire physiological response observation window. When this arithmetic mean is higher than the arithmetic mean of the corresponding dimension used to characterize the early transient response intensity and higher than the arithmetic mean of the corresponding dimension used to characterize the late steady-state response intensity, it is determined that the energy flow redistribution process within the algal community dominates the entire physiological response process. When the above comparison relationship is not valid, it is determined that the energy flow redistribution process within the algal community does not dominate. When all the predefined consistency judgment conditions for the energy redistribution tolerance prototype are met, it is determined that the dynamic feature supports the corresponding energy redistribution tolerance prototype, and the prototype consistency flag is set to high. When any consistency judgment condition is not met, the prototype consistency flag is set to low.

[0104] After completing the above two verification rules, based on the combination of the credibility support flag and the prototype consistency flag, the corresponding dynamic verification coefficient is selected from the preset rule and verification coefficient mapping table. When both the credibility support flag and the prototype consistency flag are high, the dynamic verification coefficient used to enhance the risk weight is selected; when both flags are low, the dynamic verification coefficient used to weaken the risk weight is selected; when only one flag is high, the dynamic verification coefficient used to maintain risk neutrality adjustment is selected.

[0105] The rule-verification coefficient mapping table includes at least a credibility support flag field, a prototype consistency flag field, and a dynamic verification coefficient field. When both the credibility support flag and the prototype consistency flag are high, the dynamic verification coefficient is greater than 1. When both the credibility support flag and the prototype consistency flag are low, the dynamic verification coefficient is a weakening coefficient less than 1 and greater than 0. When only one of the flags is high, the dynamic verification coefficient is 1. This allows the basic risk score to be enhanced, weakened, or maintained based on the dynamic verification results.

[0106] After determining the dynamic verification coefficient, the risk score synthesis calculation process is executed. The product of the basic risk score and the dynamic verification coefficient is recorded as the cyanobacterial bloom risk score of the water monitoring area at the current assessment time. The physiological damage level of algae response data, the prototype of algae community physiological response, and the cyanobacterial bloom risk score of the water monitoring area are packaged into the cyanobacterial bloom risk assessment result of the water monitoring area.

[0107] In this implementation plan, a multidimensional feature vector sequence of algal response data is introduced as the analysis object. Instead of relying solely on algal biomass or pigment concentration, the plan analyzes the ecological status of the water body from the physiological aspects such as the photosynthetic function status of the algal community, energy allocation mechanism, and response rate. This allows for the early identification of potential algal bloom risks before the algal population increases significantly, thus solving the problem of delayed early warning in the early stages of drastic changes in physicochemical conditions in existing technologies.

[0108] Based on the risk assessment results of cyanobacterial blooms in the water monitoring area, extrapolation and analysis are conducted to obtain early warning information on the risk of cyanobacterial blooms in the water monitoring area, and early warning of cyanobacterial bloom risks is issued based on the early warning information.

[0109] The method for obtaining early warning information on the risk of cyanobacterial blooms in the water monitoring area through extrapolation and analysis is as follows:

[0110] The system receives the risk assessment results of cyanobacterial blooms and compares the cyanobacterial bloom risk score with preset low-risk and medium-risk thresholds. If the cyanobacterial bloom risk score is lower than the low-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as low-risk. If the cyanobacterial bloom risk score is higher than or equal to the low-risk threshold but lower than the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as medium-risk. If the cyanobacterial bloom risk score is higher than or equal to the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as high-risk. The risk score comprehensively reflects the degree of physiological damage to algae, the type of physiological response, and its dynamic stability. When the score crosses the statistical threshold, it means that the overall physiological state of the algal community has changed from normal fluctuation to an abnormal state with bloom potential.

[0111] It should be noted that the low-risk and medium-risk thresholds are jointly determined by those skilled in the art based on the statistical distribution characteristics of cyanobacterial bloom risk scores in long-term historical monitoring data and actual risk management needs. The low-risk threshold is usually set at a lower percentile of the historical risk score sequence, such as the 50th percentile, i.e., the median, to distinguish between normal fluctuations and potential risks that need attention. The medium-risk threshold is usually set at a higher percentile, such as the 80th percentile, to identify significant risk states that require proactive intervention. These thresholds can be seasonally revised or regionally adjusted according to the risk tolerance and management strategies of different water areas.

[0112] For the physiological response prototypes of algal communities in the water monitoring area, combined with the inference knowledge rule set associated with the physiological response prototypes of algal communities, the inference logic analysis of the physiological response prototypes of algal communities is performed to obtain the inference output conclusions corresponding to the physiological response prototypes of algal communities in the water monitoring area. Different physiological response prototypes of algae correspond to different adaptation strategies and competition paths. Through rule-based inference, their evolutionary trend under subsequent environmental conditions can be predicted.

[0113] It should be added that the inductive knowledge rule set of the prototype association of algal community physiological response is used to explain how the algal community may evolve in the future under a certain physiological state. Based on well-known ecological principles, those skilled in the art have summarized and analyzed historical cases to obtain the inductive knowledge rule set of the prototype association of algal community physiological response.

[0114] In a specific embodiment, if the algal population physiological response prototype is an energy redistribution tolerance prototype, then a set of inference rules associated with the energy redistribution tolerance prototype is loaded. This set of inference rules defines that when matching the energy redistribution tolerance prototype, a joint judgment must be made based on the physiological damage level of the parsed algal response data. Specifically, if the algal response data shows high physiological damage, the inference output conclusion indicates that high physiological damage limits long-term competitive advantage, but tolerance may delay population decline; if the algal response data shows moderate or no physiological damage, the inference output conclusion indicates significant tolerance characteristics and a clear potential for population competitive advantage in the later stages of drastic change. By jointly inferring the physiological response prototype and the physiological damage level, a one-sided judgment of algal competitive ability based solely on a single physiological characteristic is avoided, thus solving the problem of difficulty in distinguishing between short-term physiological inhibition and long-term adaptive potential of algae under drastic environmental changes.

[0115] Based on the inference output conclusions corresponding to the risk level of cyanobacterial blooms and the prototype of algal community physiological responses in the water monitoring area, a pre-set population dynamics simulator is invoked to analyze and obtain the potential dominant physiological types in the water monitoring area. Using the multidimensional physiological time-series data of algae in the water monitoring area, the inference output conclusions corresponding to the risk level of cyanobacterial blooms and the prototype of algal community physiological responses as initial conditions and boundary constraints, the succession process of algal populations within a preset time period is simulated. The physiological types of algae that exceed the preset dominant species threshold at the end of the preset time period are identified and determined. These physiological types of algae are marked as potential dominant physiological types in the water monitoring area. The algal succession process is quantitatively inferred through the population dynamics simulator, enabling the system to identify algal physiological types that may occupy ecological niches before the bloom occurs. This solves the problem that traditional monitoring methods can only identify dominant species after the bloom has formed, achieving true early warning.

[0116] The population dynamics simulator can be implemented using existing population dynamics simulation methods and is configured to take the initial biomass percentage of each algal physiological type, the risk level of cyanobacterial blooms, the inferred output conclusions corresponding to the prototype of the algal community's physiological response, and a preset time period as inputs, and iteratively update the biomass percentage of each algal physiological type according to a discrete time step. The initial biomass percentage of each algal physiological type is obtained statistically from historical algal community monitoring data or current sampling and detection data in the water monitoring area. Within each discrete time step, the biomass percentage is updated according to the competitive advantage coefficient corresponding to the prototype of the algal community's physiological response. The data includes damage inhibition coefficients corresponding to the physiological damage levels of algal response data and environmental disturbance correction coefficients corresponding to the types of key change parameters. Growth corrections for each algal physiological type are calculated, and the biomass proportion of each algal physiological type is normalized and updated based on these corrections. The competitive advantage coefficient, damage inhibition coefficient, and environmental disturbance correction coefficient are determined statistically from biomass change records of each algal physiological type during historical cyanobacterial bloom events. At the end of a preset time period, algal physiological types with a biomass proportion exceeding a preset dominant species threshold are identified as potential dominant physiological types in the water monitoring area. The population dynamics simulator does not directly predict whether an algal bloom will occur, but rather simulates the changes in the biomass proportion of different algal physiological types over a future period, and identifies physiological types with a proportion exceeding a preset dominant species threshold as potential dominant physiological types.

[0117] It should be added that the preset dominant species threshold is set by those skilled in the art based on the general laws of population competition and dominance replacement in aquatic ecology, and is used to determine whether a physiological type has acquired a structural advantage sufficient to trigger an algal bloom in simulated succession. In one specific implementation, referring to the conventions of defining dominant species in common phytoplankton community succession studies, those skilled in the art set the preset dominant species threshold to 40% of the biomass percentage. This threshold can be adaptively modified based on historical algal bloom dominant species composition data of the target water body.

[0118] Ultimately, by integrating the risk levels and potential dominant physiological types of cyanobacterial blooms in the water monitoring areas, a structured early warning information on the risk of cyanobacterial blooms in the water monitoring areas is generated.

[0119] The method for issuing cyanobacterial bloom risk warnings based on cyanobacterial bloom risk warning information is as follows:

[0120] Upon receiving early warning information on the risk of cyanobacterial blooms in the water monitoring area, when the risk level of cyanobacterial blooms is low, an enhanced monitoring mode with increased monitoring frequency is triggered, which increases the monitoring frequency to twice the normal frequency; when the risk level of cyanobacterial blooms is medium, a risk attention reminder is sent to local management personnel and a manual review and inspection process is initiated; when the risk level of cyanobacterial blooms is high, an emergency early warning notice containing specific spatial location identifiers and potential dominant physiological groups is sent to multi-level joint management departments.

[0121] Please see Figure 2 As shown, the second aspect of the present invention provides a water body cyanobacterial bloom early warning and monitoring system, including: a data variable speed analysis module, used to collect dynamic time series data of physicochemical parameters in the water body monitoring area, perform variable speed analysis on the dynamic time series data of physicochemical parameters, and obtain information on the changes in physicochemical parameters of the water body in the water body monitoring area.

[0122] The time-series data fitting and decomposition module is used to receive information on changes in the physicochemical parameters of the water body in the water monitoring area, synchronously monitor and collect multidimensional physiological time-series data of algae in the water monitoring area, and obtain a multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing.

[0123] The algal bloom risk analysis module is used to perform algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area, and obtain the cyanobacterial bloom risk assessment results of the water monitoring area.

[0124] The risk warning module is used to extrapolate and analyze the risk assessment results of cyanobacterial blooms in the water monitoring area to obtain cyanobacterial bloom risk warning information for the water monitoring area, and to issue cyanobacterial bloom risk warnings based on the cyanobacterial bloom risk warning information.

[0125] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0126] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A method for early warning and monitoring of cyanobacterial blooms in water bodies, characterized in that, include: Collect dynamic time-series data of physicochemical parameters in the water body monitoring area, perform variable speed analysis on the dynamic time-series data of physicochemical parameters, and obtain information on the changes of physicochemical parameters in the water body monitoring area. The system receives information on changes in the physicochemical parameters of the water body in the water monitoring area, simultaneously monitors and collects multidimensional physiological time-series data of algae in the water monitoring area, and obtains a multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing. Algal bloom risk analysis was performed on the multidimensional feature vector sequence of algal response data in the water monitoring area to obtain the risk assessment results of cyanobacterial blooms in the water monitoring area. Based on the risk assessment results of cyanobacterial blooms in the water monitoring area, extrapolation and analysis are conducted to obtain early warning information on the risk of cyanobacterial blooms in the water monitoring area, and early warning of cyanobacterial bloom risks is issued based on the early warning information.

2. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 1, characterized in that, The method for performing variable-speed analysis on dynamic time-series data of physicochemical parameters is as follows: The dynamic time-series data of the physicochemical parameters in the water monitoring area includes the physicochemical parameters, the timestamps of the collection of the physicochemical parameters, and the spatial location identifiers. Based on the physicochemical parameters and their acquisition timestamps in the dynamic time series data of physicochemical parameters, the acceleration sequence of the changes in physicochemical parameters is obtained through analysis; Based on the acceleration sequence of changes in physicochemical parameters, the information on changes in physicochemical parameters of the water body in the monitoring area can be determined; Information on changes in the physicochemical parameters of the water body in the water monitoring area includes: event identifiers, event start times, event duration ranges, spatial location identifiers, key change parameter types, and acceleration sequences of key change parameters.

3. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 2, characterized in that, The method for monitoring and collecting multidimensional physiological time-series data of algae in the water monitoring area is as follows: Receive information on changes in the physicochemical parameters of the water body in the water monitoring area, and analyze the start time, duration range, and spatial location of each physicochemical condition change event according to the event identifier; When the acceleration sequence of changes in physicochemical parameters is determined to meet the conditions for changes in the physicochemical properties of the water body, algal response data of the water body monitoring area is obtained. The algal response data of the water body monitoring area are associated with the corresponding collection timestamps and spatial location identifiers, and arranged continuously in the order of collection timestamps to form the multidimensional physiological time-series data of algae in the water body monitoring area. The multidimensional physiological time-series data of algae includes the original measurement value sequence of algal response data, collection timestamps and spatial location identifiers.

4. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 1, characterized in that, The method for obtaining the multidimensional feature vector sequence of algal response data in the water monitoring area through event window feature extraction processing is as follows: Receive information on changes in the physicochemical parameters of the water body in the water monitoring area and multidimensional physiological time-series data of algae. Based on the event start time and spatial location identifier of the physicochemical condition change event, locate and extract the event-related algal physiological observation sequence corresponding to the physicochemical condition change event from the multidimensional physiological time-series data of algae. This sequence is recorded as the event-related algal physiological observation sequence of the water monitoring area. The event-related algal physiological observation sequence for the water body monitoring area includes the original measurement sequence of algal response data in the water body monitoring area within the preset physiological response observation window, as well as the collection timestamp and spatial location identifier; Based on the event-related algal physiological observation sequence of the water body monitoring area, the response amplitude feature vector and time scale feature vector of algal response data in the event-related algal physiological observation sequence are extracted; Based on the original measurement sequence of algal response data in the water monitoring area within the preset physiological response observation window, and the response amplitude feature vector and time scale feature vector of algal response data in the event-related algal physiological observation sequence, a multidimensional feature vector sequence of algal response data in the water monitoring area is constructed.

5. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 4, characterized in that, The method for performing algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area is as follows: The multidimensional feature vector sequence of algal response data includes the original measurement value sequence, response amplitude feature vector sequence, and time scale feature vector sequence of algal response data; The system receives multidimensional feature vector sequences of algal response data from the water monitoring area, performs a three-level risk analysis on these sequences, and obtains the physiological damage level, prototype of algal community physiological response, and risk score of cyanobacterial bloom from the algal response data in the water monitoring area. These data are then packaged into a risk assessment result for cyanobacterial bloom in the water monitoring area.

6. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 5, characterized in that, The method for performing three-level risk analysis on the multidimensional feature vector sequence of algal response data is as follows: The first level is physiological damage analysis, which extracts all the original measurement values ​​of algal response data from the original measurement value sequence of algal response data and analyzes them to obtain the physiological damage level of algal response data in the water monitoring area; The second level is physiological response prototype matching. Quantitative features are obtained from the original measurement value sequence of algal response data. The obtained quantitative features are compared with the feature conditions of algal community physiological response prototypes in the predefined physiological response prototype rule base to obtain the algal community physiological response prototype of the water body monitoring area. The third level is dynamic feature verification and risk integral synthesis. Based on the physiological damage level of algal response data output from the first level and the algal community physiological response prototype obtained by matching with the second level, dynamic feature verification is performed by combining the response amplitude feature vector sequence and time scale feature vector sequence of algal response data in the multidimensional feature vector sequence of algal response data to generate the risk integral value of cyanobacterial bloom in the water monitoring area.

7. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 1, characterized in that, The method for obtaining early warning information on the risk of cyanobacterial blooms in the water monitoring area through extrapolation and analysis is as follows: Receive the risk assessment results of cyanobacterial blooms, compare the risk score of cyanobacterial blooms with the preset low-risk level threshold and medium-risk level threshold, and obtain the risk level of cyanobacterial blooms in the water monitoring area. For the prototype of the physiological response of algal community in the water monitoring area, combined with the inference knowledge rule set associated with the prototype of the physiological response of algal community, the inference logic analysis of the prototype of the physiological response of algal community is performed to obtain the inference output conclusion corresponding to the prototype of the physiological response of algal community in the water monitoring area. Based on the inferences and outputs corresponding to the risk level of cyanobacterial blooms and the prototype of the physiological response of algal communities in the water monitoring area, the potential dominant physiological types in the water monitoring area were analyzed and obtained. The risk levels and potential dominant physiological types of cyanobacterial blooms in the water monitoring area are integrated and packaged to generate early warning information on the risk of cyanobacterial blooms in the water monitoring area.

8. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 7, characterized in that, The method for obtaining the risk level of cyanobacterial blooms in the water monitoring area is as follows: The risk score of cyanobacterial bloom is compared with preset low-risk and medium-risk thresholds. If the risk score is lower than the low-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as low-risk; if the risk score is higher than or equal to the low-risk threshold but lower than the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as medium-risk; if the risk score is higher than or equal to the medium-risk threshold, the cyanobacterial bloom risk level of the water monitoring area is marked as high-risk.

9. The method for early warning and monitoring of cyanobacterial blooms in water bodies according to claim 7, characterized in that, The method for issuing cyanobacterial bloom risk warnings based on cyanobacterial bloom risk warning information is as follows: Receive early warning information on the risk of cyanobacterial blooms in the water monitoring area. When the risk level of cyanobacterial blooms is low, trigger the monitoring frequency increase early warning mode. When the risk level of cyanobacterial bloom is medium, a risk awareness message is sent to the local management personnel and a manual review and inspection process is initiated. When the risk level of cyanobacterial blooms is high, an emergency early warning notice containing specific spatial location identifiers and potential dominant physiological types is sent to multi-level joint management departments.

10. A water body cyanobacterial bloom early warning and monitoring system, used to implement the water body cyanobacterial bloom early warning and monitoring method according to any one of claims 1-9, characterized in that, include: The variable speed analysis module is used to collect dynamic time-series data of physicochemical parameters in the water monitoring area, perform variable speed analysis on the dynamic time-series data of physicochemical parameters, and obtain information on the changes of physicochemical parameters in the water monitoring area. The time-series data fitting and decomposition module is used to receive information on changes in the physicochemical parameters of the water body in the water body monitoring area, synchronously monitor and collect multidimensional physiological time-series data of algae in the water body monitoring area, and obtain a multidimensional feature vector sequence of algal response data in the water body monitoring area through event window feature extraction processing. The algal bloom risk analysis module is used to perform algal bloom risk analysis on the multidimensional feature vector sequence of algal response data in the water monitoring area, and obtain the cyanobacterial bloom risk assessment results of the water monitoring area. The risk warning module is used to extrapolate and analyze the risk assessment results of cyanobacterial blooms in the water monitoring area to obtain cyanobacterial bloom risk warning information for the water monitoring area, and to issue cyanobacterial bloom risk warnings based on the cyanobacterial bloom risk warning information.

Images