An Early Warning System for Red Tide in Subsurface Water Based on LiDAR

Abstract

This invention relates to the field of marine ecological environment monitoring and discloses an early warning system for red tides in subsurface water based on lidar. The system includes a surface support platform, a lidar detection module, a water state monitoring module, and a data processing module. The lidar detection module acquires elastic backscattering signals, chlorophyll fluorescence signals, Raman scattering signals, and polarized echo signals of the target water body. The water state monitoring module collects temperature profiles, salinity profiles, flow velocity data, and flow direction data. The data processing module generates normalized chlorophyll fluorescence profiles, particulate matter backscattering profiles, and depolarization profiles, and compares them with background profiles to extract candidate subsurface enrichment layers below the surface. Then, based on the depolarization profiles, it eliminates disturbances from inorganic suspended particles and performs water mass displacement matching to determine the precursor enrichment layer of red tides, thereby outputting the red tide warning level. This application enables early identification of the process of red tides transitioning from subsurface enrichment to surface manifestation.

Description

Technical Field

[0001] This invention relates to the field of marine ecological environment monitoring, specifically to an early warning system for red tides in subsurface waters based on lidar. Background Technology

[0002] Early warning of red tides is mainly used for ecological risk monitoring in nearshore waters, harbors, aquaculture areas, and river estuaries. Current monitoring methods typically employ buoy water quality sensors, manual sampling, satellite remote sensing, or water optical detection equipment to acquire data such as chlorophyll, water temperature, salinity, dissolved oxygen, turbidity, and water color. The risk of red tides is then assessed based on changes in the monitoring data.

[0003] In real-world waters, abnormal algal proliferation doesn't always manifest as surface water color anomalies first. Influenced by sunlight, temperature and salinity stratification, tidal transport, and nutrient distribution, algae may first form a continuous enriched layer below the surface, at which point surface chlorophyll or water color changes are not obvious. Conventional buoy monitoring primarily reflects the local state at the sensor's depth, while satellite remote sensing mainly reflects surface optical information, making it difficult to continuously obtain the depth, thickness, and changes in the subsurface enriched layer.

[0004] Meanwhile, the resuspension of sediment and increased inorganic suspended particles in nearshore waters, as well as the horizontal movement of water masses, can also cause enhanced scattering or local anomalies. If the risk of red tide is judged solely based on a single chlorophyll threshold, surface water color changes, or optical parameters of the water body at a certain moment, it is easy to misjudge particle disturbance as algal enrichment, or fail to identify the changing trend of the subsurface enrichment layer before it develops to the surface, resulting in delays and misjudgments in early warning of red tides. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an early warning system for red tides in subsurface waters based on lidar, thereby solving the technical problems existing in the prior art.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution: A red tide early warning system for subsurface water based on lidar includes a water surface support platform, a lidar detection module, a water condition monitoring module, and a data processing module. The lidar detection module is mounted on a water surface support platform and is used to emit pulsed lasers toward the target water body and receive elastic backscattered signals, chlorophyll fluorescence signals, Raman scattering signals and polarized echo signals returned by the target water body. The water condition monitoring module is used to collect temperature profiles, salinity profiles, flow velocity data, and flow direction data of the target water body; The data processing module is used to generate the current monitoring profile based on the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal and polarization echo signal. The current monitoring profile includes the normalized chlorophyll fluorescence profile, particulate matter backscattering profile and depolarization profile. The data processing module is also used to call the background profile corresponding to the current monitoring conditions, compare the current monitoring profile with the background profile, extract the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region located below the surface, and determine the overlapping area of ​​the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region in the depth direction as the candidate subsurface enrichment layer. The data processing module is also used to eliminate inorganic suspended particle disturbances in the candidate subsurface enrichment layer based on the depolarization profile, and to match water mass displacements in the candidate subsurface enrichment layer during the continuous monitoring period based on flow velocity data and flow direction data. The data processing module is also used to identify candidate subsurface enrichment layers that meet the water mass displacement matching relationship as red tide precursor enrichment layers, and output red tide warning levels based on the upper boundary depth, layer thickness, normalized chlorophyll fluorescence mean, duration, and normalized chlorophyll fluorescence changes of the surface water layer of the red tide precursor enrichment layer.

[0007] Preferably, the lidar detection module includes a pulsed laser, a transmitting optical component, a receiving telescope, a beam splitting unit, an elastic backscattering receiving channel, a chlorophyll fluorescence receiving channel, a Raman scattering receiving channel, a parallel polarization receiving channel, a vertical polarization receiving channel, and a synchronous acquisition unit; The pulsed laser emits pulsed laser light toward the target water body via a transmitting optical component; The receiving telescope is used to receive the echo light returned by the target water body; The beam splitting unit is set in the output light path of the receiving telescope and introduces the echo light into the elastic backscattering receiving channel, the chlorophyll fluorescence receiving channel, the Raman scattering receiving channel, the parallel polarization receiving channel and the vertical polarization receiving channel, respectively. The synchronous acquisition unit acquires the signals output by each receiving channel using the same sampling clock.

[0008] Preferably, the data processing module includes a profile inversion unit; The profile inversion unit is used to take the time corresponding to the peak echo of the gas-water interface as the zero point of the echo time, convert the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal, parallel polarization receiving channel signal and vertical polarization receiving channel signal into the same depth sampling sequence, and deduct the dark level signal collected by each channel before laser emission. The profile inversion unit is also used to register the channel signals after dark level subtraction according to the same depth sampling sequence to form the current monitoring profile.

[0009] Preferably, the normalized chlorophyll fluorescence profile is generated by the ratio of the chlorophyll fluorescence signal to the Raman scattering signal at the same depth sampling point; The particulate backscattering profile is generated from the elastic backscattering signal after dark level subtraction; The depolarization profile is generated by the ratio of the vertically polarized receiving channel signal to the parallelly polarized receiving channel signal at the same depth sampling point.

[0010] Preferably, the data processing module includes a background profile update unit; The background profile update unit is used to remove monitoring profiles marked as red tide status by on-site verification records from historical monitoring profiles, and to group the removed historical monitoring profiles according to monitoring station location, tidal stage, sunshine period and seasonal window. The background profile update unit is also used to calculate the background mean and background standard deviation corresponding to the depth sampling points for the normalized chlorophyll fluorescence profile, particulate matter backscattering profile and depolarization profile within the same group, and to use the background mean and background standard deviation as background profile parameters under the corresponding monitoring conditions.

[0011] Preferably, the data processing module includes a candidate layer identification unit; The candidate layer identification unit is used to mark depth sampling points whose current normalized chlorophyll fluorescence value is greater than the sum of the background mean and twice the background standard deviation of the corresponding depth sampling point as chlorophyll fluorescence abnormal sampling points. The candidate layer identification unit is also used to mark depth sampling points whose current particulate backscattering value is greater than the sum of the background mean and twice the background standard deviation of the corresponding depth sampling point as particulate backscattering anomalous sampling points. The candidate layer identification unit is also used to merge adjacent chlorophyll fluorescence anomalous sampling points into chlorophyll fluorescence anomalous intervals, merge adjacent particulate matter backscattering anomalous sampling points into particulate matter backscattering anomalous intervals, and determine the overlapping area of ​​the chlorophyll fluorescence anomalous interval and the particulate matter backscattering anomalous interval in the depth direction as a candidate subsurface enrichment layer. The upper boundary of the candidate subsurface enrichment layer is located below the surface water layer, which is the first two depth sampling layers after the gas-water interface echo.

[0012] Preferably, the data processing module includes a depolarization interference elimination unit; The depolarization interference elimination unit is used to calculate the normalized fluorescence enhancement of the normalized chlorophyll fluorescence profile, the normalized scattering enhancement of the particulate matter backscattering profile, and the normalized depolarization enhancement of the depolarization profile in the candidate subsurface enrichment layer, respectively, using the background standard deviation of the corresponding depth sampling point as the denominator. When the normalized depolarization enhancement in the candidate subsurface enriched layer is greater than the normalized fluorescence enhancement, and the normalized scattering enhancement is greater than the normalized fluorescence enhancement, the depolarization interference elimination unit marks the candidate subsurface enriched layer as an inorganic suspended particle disturbance layer and deletes it from the candidate results.

[0013] Preferably, the data processing module includes a current flow displacement correction unit; The tidal current displacement correction unit is used to calculate the horizontal displacement of the water mass within adjacent monitoring cycles based on the time interval, velocity data, and direction data between adjacent monitoring cycles. The current displacement correction unit is also used to project the candidate subsurface enrichment layer of the previous monitoring cycle onto the scanning profile position corresponding to the current monitoring cycle according to the horizontal displacement. If the projected candidate subsurface enrichment layer and the candidate subsurface enrichment layer of the current monitoring period satisfy the condition that the difference in center depth is no greater than two depth sampling intervals and the difference in layer thickness is no greater than one-third of the layer thickness of the previous monitoring period, the tidal displacement correction unit will determine the two as candidate subsurface enrichment layers in the same water mass.

[0014] Preferably, the data processing module includes a spatial continuity determination unit; The spatial continuity discrimination unit is used to compare candidate subsurface enrichment layers in adjacent scanning profiles or adjacent monitoring stations; When the candidate subsurface enrichment layers in three consecutive adjacent profiles all satisfy the condition that the center depth difference is no greater than two depth sampling intervals, and the overlap length of adjacent candidate subsurface enrichment layers in the depth direction is no less than one depth sampling interval, the spatial continuity discrimination unit determines the corresponding candidate subsurface enrichment layer as a red tide precursor enrichment layer.

[0015] Preferably, the data processing module includes a trend analysis unit and an early warning judgment unit; The trend analysis unit is used to extract the center depth, upper boundary depth, layer thickness, normalized average chlorophyll fluorescence value, and subsurface to surface fluorescence ratio of the red tide precursor enrichment layer in the same water mass, using four consecutive monitoring periods as a sliding window. The trend analysis unit is also used to calculate the upward movement speed of the upper boundary based on the slope of the change of the upper boundary depth within the sliding window over time, to calculate the change in layer thickness based on the slope of the change of layer thickness over time, and to calculate the normalized chlorophyll fluorescence enhancement rate based on the slope of the change of the normalized chlorophyll fluorescence mean over time. The early warning discrimination unit outputs a subsurface aggregation prompt when there are four consecutive monitoring cycles in the red tide precursor enrichment layer, and the upper boundary depth of the later monitoring cycle is not always less than the upper boundary depth of the previous monitoring cycle. When there are four consecutive monitoring periods in the red tide precursor enrichment layer, and the upper boundary depth of the subsequent monitoring period is less than that of the previous monitoring period, the normalized chlorophyll fluorescence mean of the subsequent monitoring period is greater than that of the previous monitoring period, and the layer thickness of the subsequent monitoring period is not less than that of the previous monitoring period, an early development warning is output. When the upper boundary depth of the red tide precursor enrichment layer is less than or equal to the sum of the lower boundary depth of the surface water layer and a depth sampling interval, and the normalized mean chlorophyll fluorescence value of the surface water layer is greater than the sum of the corresponding surface background mean and twice the background standard deviation, an outbreak warning for the adjacent surface layer is output.

[0016] In summary, the present invention has the following main beneficial effects: This application acquires elastic backscattering, chlorophyll fluorescence, Raman scattering, and polarization echo signals of the target water body using a lidar detection module. A data processing module then generates normalized chlorophyll fluorescence, particulate matter backscattering, and depolarization profiles. The current monitoring profile is compared with the background profile under corresponding monitoring conditions to extract the overlapping areas of abnormal chlorophyll fluorescence and particulate matter backscattering regions below the surface. This achieves the effect of identifying precursors of algal enrichment in subsurface water before obvious anomalies in surface water color. Compared to early warning methods that rely solely on surface water quality parameters, surface water color, or single-point chlorophyll concentration, this application shifts the early warning target from surface anomalies to the subsurface enrichment process, providing a more definitive water profile basis for early red tide warnings.

[0017] This application utilizes depolarization profiles to eliminate inorganic suspended particle disturbances in candidate subsurface enrichment layers, and combines velocity and direction data to match water mass displacement within the candidate subsurface enrichment layers during continuous monitoring periods. This achieves the effect of reducing interference from sediment resuspension, inorganic particle disturbances, and horizontal water mass movement on the judgment of red tide precursors. In other words, this application does not simply use lidar inversion results for alarms, but further determines whether the candidate subsurface enrichment layer, after its formation, belongs to a water layer with enhanced algal fluorescence characteristics and can be continuously tracked within the same water mass, thereby improving the reliability of early warning in complex nearshore waters.

[0018] This application continuously analyzes the upper boundary depth, layer thickness, normalized chlorophyll fluorescence mean, duration, and normalized chlorophyll fluorescence changes in the surface water layer of the red tide precursor enrichment layer. It then outputs subsurface accumulation indications, early development warnings, or near-surface outbreak warnings, achieving a phased identification of the red tide development process from subsurface enrichment to surface manifestation. This processing method eliminates reliance on a single threshold for instantaneous judgment of warning results. Instead, it comprehensively discriminates based on the persistence of the enrichment layer, its movement towards the surface, changes in layer thickness, and surface response, enabling a clearer distinction between subsurface transient anomalies, continuous development precursors, and near-surface outbreak states. Attached Figure Description

[0019] Figure 1 This is a structural block diagram of the early warning system for red tides in subsurface water based on lidar according to the present invention. Detailed Implementation

[0020] 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.

[0021] Example 1 refer to Figure 1 A red tide early warning system for subsurface water based on lidar includes a water surface support platform, a lidar detection module, a water condition monitoring module, and a data processing module. The lidar detection module is mounted on a water surface support platform and is used to emit pulsed lasers toward the target water body and receive elastic backscattered signals, chlorophyll fluorescence signals, Raman scattering signals and polarized echo signals returned by the target water body. The water condition monitoring module is used to collect temperature profiles, salinity profiles, flow velocity data, and flow direction data of the target water body; The data processing module is used to generate the current monitoring profile based on the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal and polarization echo signal. The current monitoring profile includes the normalized chlorophyll fluorescence profile, particulate matter backscattering profile and depolarization profile. The data processing module is also used to call the background profile corresponding to the current monitoring conditions, compare the current monitoring profile with the background profile, extract the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region located below the surface, and determine the overlapping area of ​​the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region in the depth direction as the candidate subsurface enrichment layer. The data processing module is also used to eliminate inorganic suspended particle disturbances in the candidate subsurface enrichment layer based on the depolarization profile, and to match water mass displacements in the candidate subsurface enrichment layer during the continuous monitoring period based on flow velocity data and flow direction data. The data processing module is also used to identify candidate subsurface enrichment layers that meet the water mass displacement matching relationship as red tide precursor enrichment layers, and output red tide warning levels based on the upper boundary depth, layer thickness, normalized chlorophyll fluorescence mean, duration, and normalized chlorophyll fluorescence changes of the surface water layer of the red tide precursor enrichment layer.

[0022] This application applies to nearshore waters, harbors, aquaculture areas, river estuaries, lake and reservoir exchange zones, and other waters requiring red tide monitoring. This system utilizes lidar to acquire optical profiles of the target water body in the depth direction. It first identifies candidate subsurface enrichment layers below the surface, then determines the precursor red tide enrichment layer through depolarization interference elimination, tidal displacement correction, and spatial continuity discrimination. Finally, it outputs a warning level based on the layeral changes, fluorescence enhancement changes, and surface response of the precursor red tide enrichment layer.

[0023] In this embodiment, the system includes a surface support platform, a lidar profile detection module, a multi-channel echo receiving module, a profile inversion module, a background profile update module, a subsurface enrichment layer identification module, a depolarization interference elimination module, a tidal current displacement correction module, a spatial continuity discrimination module, a precursor trend analysis module, a water stratification monitoring module, and an early warning discrimination module. Each module works collaboratively in the following sequence: lidar echo acquisition, profile inversion, background comparison, candidate layer identification, interference elimination, tidal current correction, spatial confirmation, trend analysis, and early warning output.

[0024] The surface platform is used to install a lidar profiling module, a multi-channel echo receiving module, a water stratification monitoring module, and data communication equipment. The surface platform can be a fixed buoy, a shore-based outrigger platform, a shipborne platform, or a near-shore fixed monitoring platform. In this embodiment, a fixed buoy or a shore-based outrigger platform is preferred for continuous monitoring of the same monitoring station.

[0025] The lidar profiling module is mounted on a water-supporting platform and is used to emit pulsed lasers towards the target water body. After the pulsed laser enters the target water body, it interacts with algae particles, inorganic suspended particles, water molecules, and dissolved substances in the water body, forming elastic backscattering signals, chlorophyll fluorescence signals, Raman scattering signals, and polarized echo signals.

[0026] The multi-channel echo receiving module receives elastic backscattered signals, chlorophyll fluorescence signals, Raman scattering signals, and polarized echo signals returned from the target water body, and transmits these signals to the profile inversion module. The profile inversion module performs time zero-point determination, depth conversion, dark level subtraction, and profile registration on each channel signal to generate the current monitoring profile. The current monitoring profile includes a normalized chlorophyll fluorescence profile, a particulate matter backscattering profile, a diffuse attenuation profile, and a depolarized profile.

[0027] The background profile update module calls up the background profile corresponding to the current monitoring conditions. The subsurface enrichment layer identification module compares the current monitoring profile with the background profile to extract candidate subsurface enrichment layers. The depolarization interference elimination module eliminates candidate subsurface enrichment layers caused by inorganic suspended particles or sediment resuspension. The tidal current displacement correction module matches water mass displacement of candidate subsurface enrichment layers in continuous monitoring cycles based on velocity and direction data. The spatial continuity discrimination module determines whether candidate subsurface enrichment layers have spatial continuity and identifies those that meet the criteria as red tide precursor enrichment layers. The precursor trend analysis module calculates the center depth, upper boundary depth, layer thickness, normalized chlorophyll fluorescence enhancement rate, upper boundary upward movement speed, duration, and subsurface to surface fluorescence ratio of the red tide precursor enrichment layer. The water stratification monitoring module collects temperature profiles, salinity profiles, velocity data, and direction data of the target water body. The early warning judgment module is used to output the early warning level based on the results of precursor trend analysis and water stratification monitoring data.

[0028] Through the above data flow, this system does not directly use the chlorophyll signal obtained by lidar inversion as the basis for red tide alarm. Instead, it converts lidar profile data into the identification result of the subsurface continuous water layer. After background profile comparison, depolarization elimination, tidal displacement correction and spatial continuity discrimination, it then determines whether the continuous water layer has the significance of red tide precursor.

[0029] The lidar profiling module includes a pulsed laser, an emitting optics assembly, and an emitting control unit. The pulsed laser emits pulsed laser light towards the target water body. The emitting optics assembly controls the emission direction and spot size of the emitted beam. The emitting control unit records the emission time of each laser pulse and sends the emission time to the synchronous acquisition unit.

[0030] The multi-channel echo receiving module includes a receiving telescope, a beam splitter, an elastic backscattering receiving channel, a chlorophyll fluorescence receiving channel, a Raman scattering receiving channel, a parallel polarization receiving channel, a vertical polarization receiving channel, and a synchronous acquisition unit. The receiving telescope receives the echo light returned from the target water body. The beam splitter is positioned in the output light path of the receiving telescope and directs the echo light into the elastic backscattering receiving channel, chlorophyll fluorescence receiving channel, Raman scattering receiving channel, parallel polarization receiving channel, and vertical polarization receiving channel according to their wavelength and polarization direction.

[0031] The elastic backscattering receiving channel receives the backscattered echo corresponding to the emitted laser wavelength. The chlorophyll fluorescence receiving channel receives the fluorescence echo generated by excited chlorophyll. The Raman scattering receiving channel receives the Raman scattering echo from water molecules. The parallel polarization receiving channel receives the echo component parallel to the polarization direction of the emitted laser. The vertical polarization receiving channel receives the echo component perpendicular to the polarization direction of the emitted laser. The synchronous acquisition unit is connected to each of the above receiving channels and acquires the signals output by each receiving channel using the same sampling clock.

[0032] The center wavelength, bandwidth, photodetector response range, and sampling frequency of each receiving channel are determined by the instrument calibration document during system installation. The instrument calibration document is used to determine the effective signal range, noise standard deviation, channel response coefficient, and depth sampling interval, but is not used as the sole basis for red tide warning.

[0033] The profile inversion module includes a time zero point determination unit, a depth conversion unit, a dark level subtraction unit, and a profile registration unit.

[0034] The zero-time determination unit uses the moment corresponding to the peak value of the air-water interface echo as the zero-time point. The peak value of the air-water interface echo originates from the elastic backscattering receiving channel and corresponds to the location of the strong echo at the air-water interface. By using this peak value as the zero-time point, the influence of the transmission trigger delay and the fixed delay of the receiving circuit on the depth calculation can be eliminated.

[0035] The depth conversion unit converts the signals from each channel into a water depth sampling sequence based on the echo arrival time. For the first... The depth of each depth sampling point is determined according to the following formula: ; In the formula, Indicates the first The depth of each depth sampling point relative to the air-water interface; This represents the speed of light in a vacuum. Indicates the first The echo arrival time corresponding to each sampling point; This indicates the zero point of time corresponding to the peak value of the air-water interface echo. This indicates the angle between the laser propagation direction after entering the water and the vertical direction. This represents the refractive index of the target water body at the laser wavelength. When the laser enters the target water body vertically... Take zero.

[0036] The depth sampling interval is determined by the laser pulse width, sampling frequency, and water refractive index, and is written into the instrument calibration file during system installation and calibration. Subsequent candidate subsurface enrichment layer identification, tidal current displacement matching, and spatial continuity determination all use the aforementioned depth sampling interval as the smallest calculation unit in the depth direction.

[0037] The dark level subtraction unit subtracts the dark level signal acquired by each channel before laser emission. Specifically, before each laser pulse is emitted, a baseline signal without laser echo is acquired from each receiving channel, and the average value of this baseline signal is used as the dark level of the corresponding channel. After subtracting the corresponding dark level from the original echo signal of each channel, it enters the profile registration unit. This processing reduces the fixed deviation caused by the dark current of the photodetector and the bias of the acquisition circuit in the profile inversion.

[0038] The profile registration unit registers the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal, parallel polarization signal, and vertical polarization signal after dark level subtraction according to the same depth sampling sequence to generate the current monitoring profile.

[0039] Normalized chlorophyll fluorescence profiles are generated according to the following formula: ; In the formula, Indicates the first Normalized chlorophyll fluorescence values ​​at each depth sampling point; Indicates the first Chlorophyll fluorescence signal intensity after dark level subtraction at each depth sampling point; Indicates the first The Raman scattering signal intensity after dark level subtraction at each depth sampling point. The Raman scattering signal originates from water molecules themselves and its intensity changes with water body decay. Therefore, normalizing the chlorophyll fluorescence signal using the Raman scattering signal can reduce the influence of water transparency changes on the determination of fluorescence intensity.

[0040] When the Raman scattering signal at a certain depth sampling point is lower than the effective signal lower limit in the instrument calibration file, that depth sampling point is marked as an invalid sampling point. Invalid sampling points are not included in the candidate subsurface enrichment layer identification, background profile statistics, and normalized enhancement calculation. If the number of invalid sampling points in the same candidate interval exceeds half of the total number of sampling points in that candidate interval, that candidate interval will not be included in the subsequent red tide precursor enrichment layer judgment.

[0041] The particulate backscattering profile is generated from the elastic backscattering signal after dark level subtraction and is used to characterize the relative enhancement location of particulate matter in the depth direction in the target water body. The depolarization profile is generated according to the following formula: ; In the formula, Indicates the first Depolarization values ​​at depth sampling points; Indicates the first The vertical polarization receiving channel signal strength at each depth sampling point; Indicates the first The parallel polarization receiving channel signal intensity at each depth sampling point. The depolarization profile is used to help distinguish between fluorescence enhancement caused by algal enrichment and scattering enhancement caused by the resuspension of inorganic suspended particles or sediment.

[0042] The diffuse attenuation profile is generated by the attenuation of Raman scattering signals from adjacent depth sampling points and is used to reflect the optical attenuation changes of the target water body in the depth direction. The diffuse attenuation profile is used to help determine whether the signal within the corresponding depth range is within the effective detection range, and is not used as the sole basis for outputting the red tide warning level.

[0043] The background profile update module includes a historical profile filtering unit, a conditional grouping unit, and a background statistics unit.

[0044] The historical profile screening unit removes monitoring profiles marked as red tide status by on-site verification records from historical monitoring profiles. On-site verification records are derived from artificial water sampling microscopic counting records, chlorophyll a laboratory test records, red tide notification records from marine authorities, or on-site water color anomaly verification records from the same or adjacent monitoring stations. On-site verification records are used to indicate whether historical monitoring profiles can serve as non-red tide background samples and do not directly replace the system's early warning judgment.

[0045] The conditional grouping units grouped the historical monitoring profiles after removal according to monitoring station location, tidal phase, illumination period, and seasonal window. Monitoring station location was used to distinguish differences in water background across different geographical locations. Tidal phase was used to distinguish changes in water mass caused by high tide, low tide, and slack tide. Illumination period was used to distinguish the impact of daytime illumination variations on the vertical distribution of phytoplankton. Seasonal window was used to distinguish differences in water temperature and phytoplankton community background across different seasons.

[0046] The tidal phase is determined based on the tidal curve corresponding to the monitoring station. When the latter tidal level is higher than the former tidal level in two adjacent tidal level records, the corresponding time period is divided into the rising tide phase; when the latter tidal level is lower than the former tidal level, the corresponding time period is divided into the ebb tide phase; the monitoring period of the transition from rising to falling or from falling to rising tidal levels is divided into the slack tide phase.

[0047] The illumination period is determined based on the sunrise, noon, and sunset times of the monitoring stations. Historical monitoring profiles collected between sunrise and noon are classified as the morning illumination period, historical monitoring profiles collected between noon and sunset are classified as the afternoon illumination period, and historical monitoring profiles collected between sunset and the next day's sunrise are classified as the nighttime period.

[0048] Seasonal windows are divided according to the natural quarters: March to May is the spring window, June to August is the summer window, September to November is the autumn window, and December to February of the following year is the winter window. For waters where a long-term red tide monitoring system has been established, seasonal windows can also be divided according to the red tide monitoring seasons published by the competent authority, and these windows should be permanently recorded in the system deployment documents.

[0049] The background statistics unit calculates the background mean and standard deviation for each depth sampling point within the same group, based on the normalized chlorophyll fluorescence profile, particulate matter backscattering profile, and depolarization profile. The background mean and standard deviation are calculated using the following formula: ; ; In the formula, Indicates the type of profile to be analyzed, taking the normalized chlorophyll fluorescence profile, particulate matter backscattering profile, or depolarization profile; Indicates the first in the same group The historical monitoring profile in the first The values ​​of each depth sampling point; This indicates the number of historical monitoring profiles included in the statistics within the same group; Indicates the first The background mean value corresponding to each depth sampling point; Indicates the first The background standard deviation corresponding to each depth sampling point.

[0050] When the background standard deviation corresponding to a certain depth sampling point is zero or less than the noise standard deviation of the corresponding channel in the instrument calibration file, the background profile update module replaces the background standard deviation of that depth sampling point with the noise standard deviation of the corresponding channel. The noise standard deviation is determined by repeated measurement data collected during system installation and calibration under both light-shielded and stable clear water conditions. The replaced background standard deviation is used for outlier extraction and normalized enhancement calculation. This process ensures that calculations using the background standard deviation as the denominator can be performed at all depth sampling points.

[0051] When the number of historical monitoring profiles within a group is insufficient to calculate a stable background profile, the background profile update module supplements the group with historical monitoring profiles from the same monitoring station, adjacent tidal phases, or adjacent seasonal windows, and retains the source of the supplemented group in the system deployment record. The supplemented group is only used for background profile generation and does not change the subsequent processing logic for background comparison under the corresponding monitoring conditions.

[0052] The subsurface enrichment layer identification module includes an abnormal sampling point extraction unit and a continuous water layer merging unit.

[0053] The abnormal sampling point extraction unit compares the current normalized chlorophyll fluorescence profile with the corresponding background profile. For the first... A sampling point at a depth is marked as a chlorophyll fluorescence anomaly sampling point when the following formula is satisfied: ; In the formula, This indicates the current normalized chlorophyll fluorescence profile at the [number]th [year]. The values ​​of each depth sampling point; The background normalized chlorophyll fluorescence profile indicates the first... The background mean of each depth sampling point; The background normalized chlorophyll fluorescence profile indicates the first... The background standard deviation of each depth sampling point.

[0054] The abnormal sampling point extraction unit also compares the current particulate matter backscattering profile with the corresponding background profile. For the first... A depth sampling point is marked as a particulate matter backscattering anomaly sampling point when the following formula is satisfied: ; In the formula, This indicates the current particulate backscattering profile at the [number]th [number]th [number]. The values ​​of each depth sampling point; The background particulate backscattering profile is shown in the first... The background mean of each depth sampling point; The background particulate backscattering profile is shown in the first... The background standard deviation of each depth sampling point.

[0055] In this embodiment, the sum of the background mean and twice the background standard deviation is used as the threshold for extracting outlier sampling points. This threshold is derived from the statistical fluctuation range of historical profiles under the same monitoring conditions in non-red tide states. The background mean represents the normal level of the sampling point at that depth under non-red tide conditions, and the background standard deviation represents the natural fluctuation range of the sampling point at that depth under non-red tide conditions. When the current profile exceeds the sum of the background mean and twice the background standard deviation, it indicates that the sampling point at that depth exceeds the main fluctuation range of the corresponding background profile and is therefore marked as an outlier sampling point.

[0056] The continuous water layer merging unit merges adjacent chlorophyll fluorescence anomaly sampling points into a chlorophyll fluorescence anomaly region, and adjacent particulate matter backscattering anomaly sampling points into a particulate matter backscattering anomaly region. The continuous water layer merging unit then calculates the overlap region in the depth direction between the chlorophyll fluorescence anomaly region and the particulate matter backscattering anomaly region, and identifies the overlap region as a candidate subsurface enrichment layer.

[0057] In this embodiment, the surface water layer is the first two depth sampling layers after the air-water interface echo. The upper boundary of the candidate subsurface enrichment layer must be below the surface water layer. This definition prevents the system from directly treating surface transient floating matter, foam, surface oil film, or water surface echo anomalies caused by wind and waves as subsurface enrichment layers.

[0058] The continuous aquifer merging unit outputs the upper boundary depth, lower boundary depth, center depth, and thickness of the candidate subsurface enrichment layer. The center depth of the candidate subsurface enrichment layer is the arithmetic mean of the upper and lower boundary depths. The thickness of the candidate subsurface enrichment layer is the difference between the lower and upper boundary depths. These geometric quantities are used for subsequent tidal displacement correction, spatial continuity determination, and precursor trend analysis.

[0059] The depolarization interference elimination module includes a normalized enhancement calculation unit and a disturbance layer elimination unit.

[0060] In nearshore waters, sediment resuspension, construction disturbance, and disturbances from ship navigation or tidal currents can increase the number of inorganic suspended particles, resulting in enhanced particulate backscattering profiles. If judgment is based solely on enhanced particulate backscattering and local fluorescence fluctuations, sediment disturbance can easily be mistaken for algal enrichment. This embodiment utilizes depolarization profiles to eliminate interference from candidate subsurface enrichment layers.

[0061] The normalized enhancement calculation unit uses the background standard deviation of the corresponding depth sampling points as the denominator to calculate the normalized fluorescence enhancement of the normalized chlorophyll fluorescence profile, the normalized scattering enhancement of the particulate matter backscattering profile, and the normalized depolarization enhancement of the depolarization profile within the candidate subsurface enrichment layer. For the candidate subsurface enrichment layer... Its standardization enhancement is calculated according to the following formula: ; In the formula, Indicates candidate subsurface enrichment layer The corresponding standard chemical enhancement amount; Indicates the type of profile to be calculated, such as normalized chlorophyll fluorescence profile, particulate matter backscattering profile, or depolarization profile; Indicates candidate subsurface enrichment layer The number of depth sampling points within; Indicates the current monitoring profile at the [number]th [number]. The values ​​of each depth sampling point; This represents the background mean value of sampling points at the same depth. This represents the background standard deviation for sampling points at the same depth. When taking normalized chlorophyll fluorescence profiles, The standardized fluorescence enhancement amount is denoted as .when When taking particulate backscattering profiles, The normalized scattering enhancement is denoted as . .when When removing the polarization profile, To standardize the depolarization enhancement amount, denoted as The disturbance layer exclusion unit will be added to the candidate subsurface enrichment layer when the following formula is satisfied. Marked as an inorganic suspended particle disturbance layer and removed from the candidate results: ; In the formula, Indicates candidate subsurface enrichment layer The standardized depolarization enhancement amount; Indicates candidate subsurface enrichment layer The normalized scattering enhancement; Indicates candidate subsurface enrichment layer The normalized fluorescence enhancement level. This condition indicates that the polarization change and scattering enhancement of the candidate layer are stronger than the fluorescence enhancement, which is consistent with the optical behavior of inorganic suspended particle disturbance. Therefore, it is not further processed as a red tide precursor enrichment layer.

[0062] The water stratification monitoring module is used to collect temperature profiles, salinity profiles, flow velocity data, and flow direction data of the target water body. Temperature and salinity profiles are acquired by temperature-salinity-depth sensors or multi-depth temperature-salinity sensors installed below the water surface support platform. Flow velocity and flow direction data are acquired by acoustic Doppler current meters, mechanical current meters, or electromagnetic current meters.

[0063] Temperature and salinity profiles are used to determine whether stable stratification exists in the target water body. Velocity and direction data are used for tidal current displacement correction. The water stratification monitoring module sends the collected data to the tidal current displacement correction module and the early warning judgment module according to the timestamps consistent with the lidar monitoring cycle. If the sampling times of the temperature and salinity profiles are not completely consistent with those of the lidar profiles, the set of water stratification monitoring data closest to the lidar sampling time is used as the data for the same monitoring cycle.

[0064] When the water stratification monitoring module outputs multi-depth velocity data, the tidal displacement correction module uses the velocity and direction of the depth layer where the center of the candidate subsurface enrichment layer is located. When the candidate subsurface enrichment layer spans multiple velocity measurement depth layers, the arithmetic mean velocity and average direction of each velocity measurement layer within the range from the upper boundary to the lower boundary of the candidate subsurface enrichment layer are used. When the water stratification monitoring module only outputs single-layer velocity data, the single-layer velocity data is used as the water mass displacement calculation data for the corresponding monitoring period, and the installation depth of the velocity sensor is noted in the system deployment record.

[0065] The water stratification monitoring module does not directly determine the red tide status; its output data is used for water mass displacement matching, temperature and salinity stratification judgment, and early warning level verification. This approach avoids using conventional temperature and salinity or flow velocity monitoring as the sole basis for red tide alarms.

[0066] The tidal current displacement correction module includes a water mass displacement calculation unit and an enrichment layer matching unit.

[0067] The water mass displacement calculation unit calculates the horizontal displacement of the water mass within adjacent monitoring cycles based on the time interval, flow velocity data, and flow direction data between adjacent monitoring cycles. For the first... The water mass displacement between the first monitoring cycle and the next monitoring cycle is calculated using the following formula: ; In the formula, Indicates the first The monitoring cycle up to the first The horizontal displacement of water masses between monitoring cycles; This represents the average flow rate between two monitoring periods; This represents the time interval between two monitoring periods. The direction of water mass displacement is determined by the average flow direction for the corresponding time period. The system establishes a horizontal coordinate system with the monitoring station or scanning start point as the origin, and records the profile location of the candidate subsurface enrichment layer in the previous monitoring period as [reference needed]. The horizontal displacement of the water mass within adjacent monitoring periods is recorded as follows: The average flow direction is relative to the coordinate system. The included angle of the axis is denoted as Then, the projection position of the candidate subsurface enrichment layer from the previous monitoring period in the current monitoring period is: ; ; In the formula, and This indicates the projected position of the candidate subsurface enrichment layer from the previous monitoring period in the current monitoring period after correction for water mass displacement. and Indicates the profile location of the candidate subsurface enrichment layer in the previous monitoring period; Indicates the horizontal displacement of the water mass; This represents the average flow direction angle.

[0068] The enrichment layer matching unit projects the candidate subsurface enrichment layer from the previous monitoring period onto the corresponding scanning profile position of the current monitoring period according to the projection position. If the projected candidate subsurface enrichment layer and the candidate subsurface enrichment layer of the current monitoring period satisfy the following conditions: the difference in center depth is no greater than two depth sampling intervals, and the difference in layer thickness is no greater than one-third of the layer thickness of the previous monitoring period, then the two are determined to be candidate subsurface enrichment layers in the same water mass.

[0069] The aforementioned tidal displacement correction is used to distinguish between the horizontal movement of water masses and the vertical movement of enriched layers. Without tidal displacement correction, continuous changes in candidate layer depth at the monitoring station may be caused by different water masses passing through, and cannot be directly identified as the same enriched layer developing towards the surface. In this embodiment, water masses are first matched based on flow velocity and direction, and then the upward movement velocity of the upper boundary of the enriched layer is calculated, so that the upward movement judgment has a clear data basis.

[0070] The spatial continuity discrimination module includes an adjacent profile comparison unit and an enrichment layer confirmation unit.

[0071] The adjacent profile comparison unit compares candidate subsurface enrichment layers in adjacent scanning profiles or adjacent monitoring stations. Adjacent scanning profiles refer to adjacent lidar profiles obtained by the same surface-bearing platform along a preset scanning direction within a monitoring cycle. Adjacent monitoring stations refer to monitoring stations deployed in the same target water area that are adjacent along the shoreline or the main current direction.

[0072] When the candidate subsurface enrichment layers in three consecutive adjacent profiles all satisfy the condition that the center depth difference is no greater than two depth sampling intervals, and the overlap length of adjacent candidate subsurface enrichment layers in the depth direction is no less than one depth sampling interval, the enrichment layer confirmation unit will determine the corresponding candidate subsurface enrichment layer as a red tide precursor enrichment layer.

[0073] Spatial continuity discrimination is used to exclude isolated noise, local floating object projections, or transient particle disturbances in a single profile. Only candidate subsurface enrichment layers that show a continuous distribution in adjacent scan profiles or adjacent monitoring stations are included in the precursor trend analysis module.

[0074] The precursor trend analysis module includes a sliding window statistical unit, a stratigraphic change calculation unit, and a fluorescence change calculation unit.

[0075] The sliding window statistical unit uses four consecutive monitoring periods as a sliding window to extract the center depth, upper boundary depth, layer thickness, normalized mean chlorophyll fluorescence value, and subsurface to surface fluorescence ratio of the red tide precursor enrichment layer in the same water mass.

[0076] For the Red tide precursor enrichment layer in each monitoring cycle Its normalized mean chlorophyll fluorescence value is calculated according to the following formula: ; In the formula, Indicates the first The normalized mean chlorophyll fluorescence value of the red tide precursor enrichment layer during each monitoring period; represents the number of depth sampling points within the red tide precursor enrichment layer; Indicates the first Normalized chlorophyll fluorescence values ​​at each depth sampling point. The mean normalized chlorophyll fluorescence value of the surface water layer is calculated using the following formula: ; In the formula, Indicates the first Normalized mean chlorophyll fluorescence of surface water layer during each monitoring period; and These represent the first and second depth sampling layers after the air-water interface echo, respectively. The fluorescence ratio of the subsurface layer to the surface layer is calculated using the following formula: ; In the formula, Indicates the first The ratio of subsurface to surface fluorescence for each monitoring cycle; Indicates the first Normalized mean chlorophyll fluorescence of the red tide precursor enrichment layer during each monitoring period; Indicates the first Normalized mean chlorophyll fluorescence of the surface water layer during each monitoring period. Thickness of the red tide precursor enrichment layer. The difference between the lower boundary depth and the upper boundary depth in the th monitoring period: ; In the formula, Indicates the first Layer thickness for each monitoring cycle; Indicates the first The lower boundary depth of the red tide precursor enrichment layer during each monitoring cycle; Indicates the first The upper boundary depth of the red tide precursor enrichment layer in each monitoring period. The change in layer thickness within the sliding window is the difference between the layer thickness in the fourth monitoring period and the layer thickness in the first monitoring period.

[0077] The layer change calculation unit calculates the upward movement velocity of the upper boundary based on the slope of the upper boundary depth change over time within the sliding window. The upward movement velocity of the upper boundary is calculated according to the following formula: ; In the formula, This indicates the upward movement velocity of the upper boundary of the red tide precursor enrichment layer within the sliding window; This indicates the upper boundary depth of the first monitoring cycle within the sliding window; This indicates the upper boundary depth of the fourth monitoring cycle within the sliding window; This represents the sampling time for the first monitoring period within the sliding window; represents the sampling time for the fourth monitoring period within the sliding window. If... A positive value indicates a decrease in the depth of the upper boundary, suggesting that the precursory red tide enrichment layer is shifting towards the surface. The fluorescence change calculation unit calculates the normalized chlorophyll fluorescence enhancement rate based on the slope of the change in the normalized mean chlorophyll fluorescence over time within the sliding window. The normalized chlorophyll fluorescence enhancement rate is calculated using the following formula: ; In the formula, This represents the normalized chlorophyll fluorescence enhancement rate within the sliding window; This represents the normalized mean chlorophyll fluorescence value of the red tide precursor enrichment layer during the first monitoring period within the sliding window. This represents the normalized mean chlorophyll fluorescence value of the red tide precursor enrichment layer during the fourth monitoring period within the sliding window. The duration of the red tide precursor enrichment layer is the time difference between the first confirmation of the red tide precursor enrichment layer and the most recent confirmation during a continuous monitoring period. If the red tide precursor enrichment layer is in the fourth monitoring period... The first monitoring cycle was confirmed, and in the [number]th monitoring cycle... If the monitoring cycle is still confirmed, the duration is calculated according to the following formula: ; In the formula, Indicates the duration of the precursor enrichment layer of red tide; This indicates the monitoring time at which the precursor enrichment layer of the red tide was first confirmed; This indicates the monitoring time of the most recent confirmation of the red tide precursor enrichment layer.

[0078] The early warning discrimination module includes a hierarchical discrimination unit. The hierarchical discrimination unit receives the center depth, upper boundary depth, layer thickness, normalized chlorophyll fluorescence enhancement rate, upper boundary upward movement speed, duration, and subsurface to surface fluorescence ratio output by the precursor trend analysis module, and receives temperature profile, salinity profile, flow velocity data, and flow direction data output by the water stratification monitoring module.

[0079] The mean background value of the surface layer is the arithmetic mean of the normalized chlorophyll fluorescence background values ​​corresponding to the two depth sampling layers after the air-water interface echo in the background profile. The standard deviation of the surface layer background is the arithmetic mean of the standard deviations of the normalized chlorophyll fluorescence background values ​​corresponding to the two depth sampling layers after the air-water interface echo in the background profile. The mean normalized chlorophyll fluorescence value of the surface water layer is the arithmetic mean of the normalized chlorophyll fluorescence values ​​corresponding to the two depth sampling layers after the air-water interface echo in the current monitoring profile.

[0080] The classification and discrimination unit outputs the warning level according to the following rules.

[0081] When a red tide precursor enrichment layer exists continuously for four monitoring cycles, and the upper boundary depth of the later monitoring cycle is not always less than the upper boundary depth of the previous monitoring cycle, a subsurface accumulation indication is output. This indicates that a spatially continuous subsurface enrichment layer has formed in the target water body, but this enrichment layer has not yet shown a continuous process of moving towards the surface.

[0082] An early warning for red tide precursor enrichment layer is issued when it exists continuously for four monitoring periods, and in each of these periods, the upper boundary depth of the subsequent monitoring period is less than that of the previous monitoring period, the normalized average chlorophyll fluorescence value of the subsequent monitoring period is greater than that of the previous monitoring period, and the layer thickness of the subsequent monitoring period is not less than that of the previous monitoring period. This indicates that the red tide precursor enrichment layer is continuously moving towards the surface within the same water mass, and the fluorescence intensity and layer thickness are increasing synchronously.

[0083] When the upper boundary depth of the red tide precursor enrichment layer is less than or equal to the sum of the lower boundary depth of the surface water layer and a depth sampling interval, and the normalized mean chlorophyll fluorescence value of the surface water layer is greater than the sum of the corresponding surface background mean and twice the background standard deviation, an early warning of an outbreak in the adjacent surface layer is issued. This situation indicates that the enrichment layer, which was originally located in the subsurface layer, has approached the surface water layer, and the surface water layer has begun to show fluorescence anomalies relative to the background profile.

[0084] Temperature and salinity profiles are used to assist in determining the stratification state of the water body. If a continuous temperature or salinity gradient exists on both sides of the depth range of the red tide precursor enrichment layer within the sliding window, the grading discrimination unit retains the current warning level. If the temperature and salinity profiles show that the target water body is in a strongly mixed state, and the red tide precursor enrichment layer no longer meets the spatial continuity discrimination condition in subsequent monitoring periods, the grading discrimination unit terminates continuous tracking of the enrichment layer. This process is used to avoid mistaking short-term mixing-induced local anomalies as continuously developing red tide precursor enrichment layers.

[0085] After system deployment, the lidar profiling module emits pulsed lasers towards the target water body according to the set monitoring cycle. The multi-channel echo receiving module simultaneously acquires elastic backscattering signals, chlorophyll fluorescence signals, Raman scattering signals, parallel polarization signals, and vertical polarization signals. The profiling inversion module uses the peak value of the air-water interface echo as the zero point of time, converts the signals from each channel to the same depth sampling sequence, and generates the current monitoring profile.

[0086] The background profile update module reads the current monitoring station location, tidal phase, illumination period, and seasonal window, and calls the corresponding background profile parameters. The subsurface enrichment layer identification module compares the current normalized chlorophyll fluorescence profile and particulate matter backscattering profile with the corresponding background profile, extracts the chlorophyll fluorescence anomaly region and the particulate matter backscattering anomaly region, and uses the overlapping area of ​​the two in the depth direction as the candidate subsurface enrichment layer.

[0087] The depolarization interference elimination module calculates the normalized fluorescence enhancement, normalized scattering enhancement, and normalized depolarization enhancement within the candidate subsurface enrichment layer. If a candidate layer exhibits stronger scattering and depolarization enhancements than fluorescence enhancements, it is deleted. The remaining candidate layers are then processed by the tidal current displacement correction module.

[0088] The tidal displacement correction module calculates the horizontal displacement of the water mass based on the flow velocity and direction of adjacent monitoring cycles, and projects the candidate layer from the previous monitoring cycle onto the scanning profile position of the current monitoring cycle. If the projected candidate layer and the current candidate layer satisfy the conditions of center depth difference and layer thickness difference, they are determined to be candidate layers in the same water mass. The spatial continuity discrimination module further determines whether there are continuous candidate layers in adjacent scanning profiles or adjacent monitoring stations. After satisfying the spatial continuity condition, the candidate layer is determined to be a red tide precursor enrichment layer.

[0089] The precursor trend analysis module tracks the red tide precursor enrichment layer over four consecutive monitoring periods, calculating its upper boundary depth, central depth, layer thickness, normalized mean chlorophyll fluorescence, upper boundary upward movement rate, normalized chlorophyll fluorescence enhancement rate, duration, and subsurface to surface fluorescence ratio. The early warning discrimination module outputs subsurface aggregation alerts, early development warnings, or near-surface outbreak warnings according to tiered discrimination rules, and sends the warning results, the depth range of the red tide precursor enrichment layer, monitoring time, and station location information to the shore-based terminal.

[0090] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A red tide early warning system for subsurface water bodies based on lidar, characterized in that, It includes a surface support platform, a lidar detection module, a water condition monitoring module, and a data processing module; The lidar detection module is mounted on a water surface support platform and is used to emit pulsed lasers toward the target water body and receive elastic backscattered signals, chlorophyll fluorescence signals, Raman scattering signals and polarized echo signals returned by the target water body. The water condition monitoring module is used to collect temperature profiles, salinity profiles, flow velocity data, and flow direction data of the target water body; The data processing module is used to generate the current monitoring profile based on the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal and polarization echo signal. The current monitoring profile includes the normalized chlorophyll fluorescence profile, particulate matter backscattering profile and depolarization profile. The data processing module is also used to call the background profile corresponding to the current monitoring conditions, compare the current monitoring profile with the background profile, extract the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region located below the surface, and determine the overlapping area of ​​the chlorophyll fluorescence abnormal region and particulate matter backscattering abnormal region in the depth direction as the candidate subsurface enrichment layer. The data processing module is also used to eliminate inorganic suspended particle disturbances in the candidate subsurface enrichment layer based on the depolarization profile, and to match water mass displacements in the candidate subsurface enrichment layer during the continuous monitoring period based on flow velocity data and flow direction data. The data processing module is also used to identify candidate subsurface enrichment layers that meet the water mass displacement matching relationship as red tide precursor enrichment layers, and output red tide warning levels based on the upper boundary depth, layer thickness, normalized chlorophyll fluorescence mean, duration, and normalized chlorophyll fluorescence changes of the surface water layer of the red tide precursor enrichment layer.

2. The early warning system for red tides in subsurface water based on lidar according to claim 1, characterized in that, The lidar detection module includes a pulsed laser, a transmitting optical component, a receiving telescope, a beam splitting unit, an elastic backscattering receiving channel, a chlorophyll fluorescence receiving channel, a Raman scattering receiving channel, a parallel polarization receiving channel, a vertical polarization receiving channel, and a synchronous acquisition unit. The pulsed laser emits pulsed laser light toward the target water body via a transmitting optical component; The receiving telescope is used to receive the echo light returned by the target water body; The beam splitting unit is set in the output light path of the receiving telescope and introduces the echo light into the elastic backscattering receiving channel, the chlorophyll fluorescence receiving channel, the Raman scattering receiving channel, the parallel polarization receiving channel and the vertical polarization receiving channel, respectively. The synchronous acquisition unit acquires the signals output by each receiving channel using the same sampling clock.

3. The early warning system for red tides in subsurface water based on lidar according to claim 2, characterized in that, The data processing module includes a profile inversion unit; The profile inversion unit is used to take the time corresponding to the peak echo of the gas-water interface as the zero point of the echo time, convert the elastic backscattering signal, chlorophyll fluorescence signal, Raman scattering signal, parallel polarization receiving channel signal and vertical polarization receiving channel signal into the same depth sampling sequence, and deduct the dark level signal collected by each channel before laser emission. The profile inversion unit is also used to register the channel signals after dark level subtraction according to the same depth sampling sequence to form the current monitoring profile.

4. The early warning system for red tides in subsurface water based on lidar according to claim 3, characterized in that, The normalized chlorophyll fluorescence profile is generated by the ratio of the chlorophyll fluorescence signal to the Raman scattering signal at the same depth sampling point. The particulate backscattering profile is generated from the elastic backscattering signal after dark level subtraction; The depolarization profile is generated by the ratio of the vertically polarized receiving channel signal to the parallelly polarized receiving channel signal at the same depth sampling point.

5. The early warning system for red tide in subsurface water based on lidar according to claim 4, characterized in that, The data processing module includes a background profile update unit; The background profile update unit is used to remove monitoring profiles marked as red tide status by on-site verification records from historical monitoring profiles, and to group the removed historical monitoring profiles according to monitoring station location, tidal stage, sunshine period and seasonal window. The background profile update unit is also used to calculate the background mean and background standard deviation corresponding to the depth sampling points for the normalized chlorophyll fluorescence profile, particulate matter backscattering profile and depolarization profile within the same group, and to use the background mean and background standard deviation as background profile parameters under the corresponding monitoring conditions.

6. The early warning system for red tides in subsurface water based on lidar according to claim 5, characterized in that, The data processing module includes a candidate layer identification unit; The candidate layer identification unit is used to mark depth sampling points whose current normalized chlorophyll fluorescence value is greater than the sum of the background mean and twice the background standard deviation of the corresponding depth sampling point as chlorophyll fluorescence abnormal sampling points. The candidate layer identification unit is also used to mark depth sampling points whose current particulate backscattering value is greater than the sum of the background mean and twice the background standard deviation of the corresponding depth sampling point as particulate backscattering anomalous sampling points. The candidate layer identification unit is also used to merge adjacent chlorophyll fluorescence anomalous sampling points into chlorophyll fluorescence anomalous intervals, merge adjacent particulate matter backscattering anomalous sampling points into particulate matter backscattering anomalous intervals, and determine the overlapping area of ​​the chlorophyll fluorescence anomalous interval and the particulate matter backscattering anomalous interval in the depth direction as a candidate subsurface enrichment layer. The upper boundary of the candidate subsurface enrichment layer is located below the surface water layer, which is the first two depth sampling layers after the gas-water interface echo.

7. The early warning system for red tides in subsurface water based on lidar according to claim 6, characterized in that, The data processing module includes a depolarization interference elimination unit; The depolarization interference elimination unit is used to calculate the normalized fluorescence enhancement of the normalized chlorophyll fluorescence profile, the normalized scattering enhancement of the particulate matter backscattering profile, and the normalized depolarization enhancement of the depolarization profile in the candidate subsurface enrichment layer, respectively, using the background standard deviation of the corresponding depth sampling point as the denominator. When the normalized depolarization enhancement in the candidate subsurface enriched layer is greater than the normalized fluorescence enhancement, and the normalized scattering enhancement is greater than the normalized fluorescence enhancement, the depolarization interference elimination unit marks the candidate subsurface enriched layer as an inorganic suspended particle disturbance layer and deletes it from the candidate results.

8. The early warning system for red tide in subsurface water based on lidar according to claim 7, characterized in that, The data processing module includes a current flow displacement correction unit; The tidal current displacement correction unit is used to calculate the horizontal displacement of the water mass within adjacent monitoring cycles based on the time interval, velocity data, and direction data between adjacent monitoring cycles. The current displacement correction unit is also used to project the candidate subsurface enrichment layer of the previous monitoring cycle onto the scanning profile position corresponding to the current monitoring cycle according to the horizontal displacement. If the projected candidate subsurface enrichment layer and the candidate subsurface enrichment layer of the current monitoring period satisfy the condition that the difference in center depth is no greater than two depth sampling intervals and the difference in layer thickness is no greater than one-third of the layer thickness of the previous monitoring period, the tidal displacement correction unit will determine the two as candidate subsurface enrichment layers in the same water mass.

9. A red tide early warning system for subsurface water based on lidar according to claim 8, characterized in that, The data processing module includes a spatial continuity discrimination unit; The spatial continuity discrimination unit is used to compare candidate subsurface enrichment layers in adjacent scanning profiles or adjacent monitoring stations; When the candidate subsurface enrichment layers in three consecutive adjacent profiles all satisfy the condition that the center depth difference is no greater than two depth sampling intervals, and the overlap length of adjacent candidate subsurface enrichment layers in the depth direction is no less than one depth sampling interval, the spatial continuity discrimination unit determines the corresponding candidate subsurface enrichment layer as a red tide precursor enrichment layer.

10. A red tide early warning system for subsurface water based on lidar according to claim 9, characterized in that, The data processing module includes a trend analysis unit and an early warning judgment unit; The trend analysis unit is used to extract the center depth, upper boundary depth, layer thickness, normalized average chlorophyll fluorescence value, and subsurface to surface fluorescence ratio of the red tide precursor enrichment layer in the same water mass, using four consecutive monitoring periods as a sliding window. The trend analysis unit is also used to calculate the upward movement speed of the upper boundary based on the slope of the change of the upper boundary depth within the sliding window over time, to calculate the change in layer thickness based on the slope of the change of layer thickness over time, and to calculate the normalized chlorophyll fluorescence enhancement rate based on the slope of the change of the normalized chlorophyll fluorescence mean over time. The early warning discrimination unit outputs a subsurface aggregation prompt when there are four consecutive monitoring cycles in the red tide precursor enrichment layer, and the upper boundary depth of the later monitoring cycle is not always less than the upper boundary depth of the previous monitoring cycle. When there are four consecutive monitoring periods in the red tide precursor enrichment layer, and the upper boundary depth of the subsequent monitoring period is less than that of the previous monitoring period, the normalized chlorophyll fluorescence mean of the subsequent monitoring period is greater than that of the previous monitoring period, and the layer thickness of the subsequent monitoring period is not less than that of the previous monitoring period, an early development warning is output. When the upper boundary depth of the red tide precursor enrichment layer is less than or equal to the sum of the lower boundary depth of the surface water layer and a depth sampling interval, and the normalized mean chlorophyll fluorescence value of the surface water layer is greater than the sum of the corresponding surface background mean and twice the background standard deviation, an outbreak warning for the adjacent surface layer is output.

Images