Lake water environment monitoring method and system based on sensitive functional unit monitoring

Abstract

The present application relates to the technical field of water environment monitoring, and more particularly to a lake water environment monitoring method and system based on sensitive functional unit monitoring, which divides lake functional units with water depth, hydrodynamic connectivity and functional partitioning as constraints, forms unit-time data sets by fusing multi-source monitoring data, constructs a sensitivity index for identifying sensitive functional units through the coefficient of variation and the pressure index; determines the index weight and establishes a health baseline by using entropy weight-sensitivity coupling at the sensitive unit scale; calculates the health index and its changes by using a multi-time scale sliding window, realizes dynamic diagnosis of sudden deterioration and slow degradation; and identifies high-risk paths and key control units by calculating the health risk transmission intensity in combination with the hydrodynamic connectivity matrix. The present application realizes the fine characterization of the health status of the lake, the dynamic tracking of the time sequence evolution process, and the identification of the key risk transmission path, and provides reliable technical support for fine management and scheduling decisions of the lake.

Description

Technical Field

[0001] This invention relates to the field of water environment monitoring and water ecological health assessment technology, specifically to the technical scope of lake hydrology-water ecology integrated monitoring and health diagnosis methods and supporting systems, and particularly to a lake water environment monitoring method and system based on sensitive functional unit monitoring. Background Technology

[0002] In recent years, with the continuous increase in the intensity of watershed development, the high level of external pollution load, and the superposition of multiple stresses such as changes in hydrological conditions, many lakes have experienced problems such as eutrophication, frequent algal blooms, decline in aquatic biodiversity, and habitat fragmentation, and the degradation of the structure and function of lake ecosystems has become increasingly prominent.

[0003] Against this backdrop, various lake water environment monitoring methods and systems have been proposed in China. For example, Chinese invention publication CN106202960A discloses a health assessment method based on lake aquatic ecosystems. This method involves systematically investigating the lake's aquatic environment and aquatic biological communities, using canonical correspondence analysis (CCA) to identify the relationship between environmental factors and aquatic organism distribution, and then combining principal component analysis (PCA) and correlation analysis to screen evaluation indicators. An objective weighting method is used to determine the weights, constructing a lake aquatic ecosystem health assessment indicator system, and calculating the comprehensive index ECHI to characterize the overall lake health status. This method demonstrates strong scientific rigor in incorporating toxicological indicators and integrating water quality and biological community information, making it suitable for assessing the overall health status of a lake on an annual or seasonal scale. However, the evaluation units proposed by this existing technology are mainly "whole lake" or spatial distribution based on sampling point interpolation. The technical solution does not explicitly combine information such as water depth gradient, hydrodynamic connectivity, shoreline type and functional zoning to construct "functional units" with clear ecological and management functions, nor does it distinguish which units are most sensitive to pressure factors such as external load and water level fluctuations. At the same time, the method calculates a one-time comprehensive index based on phased monitoring data, and lacks a mechanism to conduct dynamic diagnosis and risk transmission analysis based on long-term series at a unified spatial unit scale.

[0004] Chinese invention patent CN119671049A discloses a technical method for assessing the aquatic ecological health of lakes and reservoirs. This method further constructs an aquatic ecological health evaluation system composed of multi-level indicators such as phytoplankton, zooplankton, organic matter, and water quality. By constructing a multi-level survey and evaluation matrix, normalization processing, and a multi-level weight matrix, the total score for each level is calculated, and health levels are classified as "excellent," "good," "average," "poor," and "very poor." This scheme improves upon the hierarchical design of the indicator system and the standardization of weight calculation, enabling a more comprehensive static assessment of the ecological health of lakes / reservoirs. However, this existing technology also targets "whole lake / reservoir" or coarse-grained spatial units, and the evaluation results are mainly reflected in the comprehensive score and health level. It is difficult to identify key sensitive areas that play different roles as "source area, transmission area, and sink area" in the hydrodynamic path. Its method focuses on multi-level index weighted scoring, does not use hydrodynamic connectivity information to construct the health risk transmission relationship between units, lacks the design for dynamic diagnosis of health index at multiple time scales based on sliding time windows, and has not formed a differentiated monitoring and early warning framework for "sensitive functional units".

[0005] In the area of ​​real-time water environment monitoring, Chinese invention patent CN118777559B discloses a water environment monitoring method and system. It proposes dividing the monitoring area into shallow and deep water environments based on water depth, deploying water quality sensors at different water layers to collect multi-parameter data in real time, and calculating the comprehensive influence coefficients of water quality parameters on temperature, dissolved oxygen, and pH to obtain the "water health status value" for each water layer. This value is then compared with thresholds to determine whether the water meets standards. This scheme has certain innovations in vertical stratified monitoring, real-time data acquisition, and comprehensive influence analysis of water quality parameters, improving the automation and precision of water environment monitoring. However, from the perspective of "dynamic health diagnosis" of lakes, this existing technology still has the following limitations:

[0006] The spatial division is based solely on water depth, roughly dividing the monitoring area into two types of environments: shallow water and deep water. It fails to incorporate information such as hydrodynamic connectivity, the location of tributaries flowing into the lake, and shoreline type to construct spatial units with clear hydrological and ecological functional significance, and it also fails to select the "sensitive functional units" that are most sensitive to pressure factors.

[0007] Its "water health status value" is mainly the result of the comprehensive influence of water quality parameters within each water layer. It does not incorporate the hydrodynamic connection between different spatial units (such as the estuary area, nearshore zone, open lake center area, etc.) into a unified topological structure, and lacks the transmission analysis of the spread and accumulation of health risks along the hydrodynamic path.

[0008] Although real-time monitoring data is used, the diagnostic logic is mainly based on the comparison of the average health value at a certain moment or over a short period of time with the threshold. It does not introduce multi-timescale sliding windows and trend indicators to distinguish different evolution patterns such as "sudden deterioration" and "slow degradation", and its dynamism is still insufficient.

[0009] In summary, existing technologies have established relatively mature indicator systems and evaluation frameworks for lake health assessment and water environment monitoring, but the following common shortcomings still exist:

[0010] (1) The spatial unit division is crude and lacks the concept and identification method of "sensitive functional unit". Existing technologies mostly use the whole lake or regular assessment unit as the scale for evaluation. They do not construct functional units with ecological and management functions by integrating information such as water depth, hydrodynamic connectivity, shoreline type and functional utilization, and do not conduct sensitivity analysis and screening for these units.

[0011] (2) Lack of analysis on the “transmission” of health risks under hydrodynamic constraints. Existing technologies usually treat each evaluation unit as independent and obtain a comprehensive index or health value by weighted summation, without introducing a hydrodynamic connectivity matrix or source-transmission-sink topology to describe the connection between units. This makes it impossible to identify the critical path of health risks spreading downstream from the source area along the hydrodynamic channel and accumulating and amplifying in the sink area.

[0012] (3) Insufficient dynamic diagnostic capabilities. Although existing technologies have begun to use online monitoring data and provide health status values ​​at a certain moment, most of them remain at the level of static level determination or simple time series comparison. They have not designed multi-time scale sliding diagnostic windows and trend indicators under a unified functional unit scale, making it difficult to identify sudden changes, continuous degradation and recovery processes in the health status of lakes in a timely manner.

[0013] (4) Multi-source monitoring data have not yet been deeply integrated at a unified functional unit scale. Existing technologies mostly build evaluation models for single or a few data sources. There is still a lack of mature solutions for how to uniformly model, normalize, and calculate weights for multi-source data such as fixed stations, buoys, unmanned vessels, and remote sensing inversion at the functional unit level. Summary of the Invention

[0014] To address the shortcomings of existing technologies, this invention provides a lake water environment monitoring method and system based on sensitive functional unit monitoring, which solves the technical problems in existing lake water environment monitoring, such as coarse division of evaluation units, difficulty in identifying key sensitive areas, lack of health risk transmission analysis under hydrodynamic constraints, and difficulty in using multi-source monitoring data to carry out dynamic diagnosis at multiple time scales.

[0015] Specifically, this invention addresses the shortcomings of existing lake water environment monitoring and health assessment technologies, such as coarse evaluation unit division, lack of sensitive functional unit identification mechanism based on hydrodynamics and functional zoning, failure to incorporate hydrodynamic connectivity into source-transmission-sink transmission analysis of health risks, and difficulty in integrating multi-source monitoring data at a unified spatial unit scale for multi-timescale dynamic diagnosis and early warning. It proposes a lake water environment monitoring method and system based on sensitive functional unit monitoring, enabling zoned and refined characterization of lake water environment health status, dynamic tracking of temporal evolution processes, and identification of key risk transmission paths, providing reliable technical support for refined lake management and scheduling decisions.

[0016] To address the aforementioned technical problems, the lake water environment monitoring method proposed in this invention generally includes five core components: lake functional unit division and multi-source monitoring deployment, sensitive functional unit identification and functional type classification, construction of health baselines for sensitive functional units, dynamic calculation of health indices across multiple time scales, and health status diagnosis and risk transmission analysis based on hydrodynamic connectivity matrices. Figure 1 As shown. The system is structured with modules for dividing lakes into functional units, acquiring multi-source monitoring data, identifying sensitive functional units, calculating health baselines and weights, dynamically calculating health indices, diagnosing and analyzing health status, and visualizing and issuing early warnings. Figure 2 As shown, this enables the procedural and engineering implementation of the aforementioned lake water environment monitoring methods.

[0017] Firstly, regarding the functional unit division and multi-source monitoring deployment of lakes, this invention uses lake basin topography, water depth pattern, hydrodynamic connectivity, shoreline type, and management functional zoning as constraints to divide the lake body into a set of basic functional units with clear hydrological and ecological significance. This invention reads digital elevation or water depth data of the lake basin, hydrodynamic numerical simulation results, or long-term flow velocity-retention time monitoring results, and imports vector information of functional zones such as drinking water source protection areas, landscape recreation areas, and fishery water use areas, as well as natural shorelines, hard revetments, and ecological restoration shorelines. Based on this, the invention divides the lake body into initial grid units on a plane, selecting a grid scale of 50–200m. Then, according to multiple constraints such as water depth level (e.g., 0–2m, 2–5m, and above 5m), hydrodynamic connectivity (high connectivity zone, medium connectivity zone, and low connectivity zone), and functional utilization, the initial grid units are merged, trimmed, and labeled to form basic functional units. Each basic functional unit is accompanied by attributes such as average water depth, residence time, water renewal rate, shoreline type, and functional type, providing a spatial carrier for subsequent diagnostics.

[0018] Furthermore, the basic functional unit satisfies at least two of the following partitioning constraints:

[0019] Water depth classification constraints: Zones are divided according to water depth ranges of 0-2m, 2-5m, and above 5m;

[0020] Hydrodynamic connectivity constraints: Based on flow velocity, residence time and water turnover rate, units that have direct connectivity with the inflow estuary or outflow control section are classified as high connectivity zones, and the rest are classified as medium and low connectivity zones.

[0021] Shoreline type constraints: Based on the degree of shoreline hardening and nearshore vegetation type, units close to natural shorelines and restored shorelines are marked as ecological shoreline units;

[0022] Functional utilization constraints: Basic functional units are tailored based on the zoning boundaries of drinking water sources, landscape belts, and general functional areas.

[0023] After the basic functional units are divided, the present invention deploys or maps multi-source monitoring points within each basic functional unit, and collects multi-source monitoring data of time series of hydrodynamic, water quality and ecological indicators for each basic functional unit within a preset monitoring period.

[0024] Preferably, data on lake level, flow velocity, and water quality parameters are collected through fixed monitoring stations deployed along the shore and buoy monitoring devices deployed on the lake surface; among which, water quality parameters include total nitrogen (TN), total phosphorus (TP), and ammonia nitrogen (TN). The data collected includes one or more of chlorophyll a (Chl-a) and dissolved oxygen (DO), preferably hourly or daily high-frequency data. Unmanned surface vessel (USV) surveys are used to acquire flow velocity field data and local water quality parameters for the lake surface or cross-section; remote sensing inversion is used to acquire surface distribution information such as chlorophyll a and transparency. Through time alignment and spatial mapping algorithms, data with different sampling frequencies and spatial resolutions are uniformly mapped to basic functional units and a unified time step (e.g., 1 hour or 1 day), forming a two-dimensional unit-time dataset, which includes a set of pressure factors such as external load, water level fluctuations, and wind intensity. And a set of response indicators including transparency, TN, TP, Chl-a, biodiversity index, and plant cover. ,like Figure 4 As shown.

[0025] Secondly, regarding the identification of sensitive functional units and the classification of functional types, this invention automatically identifies the regions most sensitive to stress factors through sensitivity analysis at a unified basic functional unit scale. It calculates the data based on the unit-time dataset used to extract data from the set of basic functional units. Selecting sensitive functional unit sets Overall sensitivity and set up sensitive functional units The system is divided into three functional types: source sensitive units, transmission sensitive units, and sink sensitive units, forming a source-transmission-sink topology of sensitive functional units and hydrodynamic channels, such as... Figure 3 As shown.

[0026] Furthermore, for each basic functional unit and each response indicator Calculate the average value within the monitoring period. and standard deviation And obtain the coefficient of variation. ,Right now:

[0027]

[0028] At the same time, the set of stress factors The comprehensive pressure intensity index is constructed by standardizing and weighting multiple pressure factors, including external pollution load, inflow into the lake, water level fluctuation, and wind speed. Then calculate the response indicators. With pressure intensity index Pearson correlation coefficient between Based on this, we define the unit – index sensitivity. ,Right now:

[0029]

[0030] Then according to the preset weight The overall sensitivity of each basic functional unit is obtained by summarizing. ,Right now:

[0031]

[0032] in, For use in calculating the overall sensitivity of basic functional units The number of response indicators; For the first Sensitivity weights for each response metric, satisfying ; Basic functional unit In the Sensitivity scores on each response metric.

[0033] Preferably, the preset weight The weighting can be determined by combining expert empowerment with entropy weighting to ensure that it reflects both the level of management attention and the amount of information in the indicators.

[0034] By statistically analyzing the sensitivity distribution of all basic functional units, this invention can determine the sensitivity threshold based on the sensitivity percentile (e.g., the top 30%) or based on cluster analysis. , will satisfy Basic functional units Included in the sensitive functional unit set .

[0035] To further characterize the spatial propagation path of health risks within a lake, this invention incorporates sensitive functional unit sets. Constructing a hydrodynamic connectivity matrix .

[0036] Preferably, based on the results of two-dimensional or three-dimensional hydrodynamic numerical simulations, the statistically sensitive functional units are... and Traffic exchange volume over a certain time scale And normalize it to the percentage of traffic per unit time, matrix elements Characterization of sensitive functional units To sensitive functional units The proportion of flow or frequency of water exchange is expressed as:

[0037]

[0038] By analyzing the upstream contribution and downstream impact characteristics of each sensitive functional unit, such as calculating the upstream catchment area ratio, the downstream controlled area ratio, and the average residence time, this invention divides the sensitive functional units into source area sensitive units (located at the main inflow river mouth and upstream, with significant downstream output), transmission sensitive units (located in the middle of the main hydrodynamic channel, with strong connections to both upstream and downstream), and sink area sensitive units (with longer residence times, prone to becoming pollution or algal bloom accumulation areas), forming a source-transmission-sink topology structure of sensitive functional units-hydrodynamic channel.

[0039] Furthermore, regarding the construction of health baselines for sensitive functional units, this invention proposes using historical "healthy status periods" as a reference to construct a health baseline index reflecting the ideal state at the unit scale. Specifically, this is achieved by referencing healthy periods... Internally, a method combining information entropy weighting and sensitivity weighting is used to calculate the health index weight of each sensitive functional unit. And generate health baseline indices for each sensitive functional unit. And the tiered threshold range.

[0040] Furthermore, based on water quality compliance, eutrophication index, algal bloom records, and ecological survey results, typical healthy periods (e.g., months with good water quality and ecology during both the high-water and low-water seasons) are selected over several years or hydrological years as reference healthy periods. During the reference health period Within this process, the response indicators of each sensitive functional unit are normalized using the range method or the Z-score method to obtain normalized values. .

[0041] To reasonably determine the contribution of each indicator to the health index, this invention uses a method that couples information entropy weighting and sensitivity weighting to calculate the health index weights. First, refer to the healthy time period. Within, for each response metric Calculate the normalized value distribution of each sensitive functional unit and time point to obtain the information entropy. This leads to the information entropy weight. Secondly, the response indicators of each sensitive functional unit will be... Sensitivity score Normalization along the indicator dimension yields Then, the health index weights are calculated using the following formula. ,Right now:

[0042]

[0043] in, The balance coefficient between entropy weight and sensitivity weight satisfies... ,and ; Based on reference health time period Internal response indicators The calculated information entropy weights; This represents the normalized result of the sensitivity scores for each response indicator.

[0044] Determining the weight of the health index Then, calculate the reference healthy period for each sensitive functional unit. Health baseline index within ,Right now:

[0045]

[0046] In conjunction with statistical distribution and ecological management requirements, a tiered threshold range is set for each sensitive functional unit, including healthy, sub-healthy, early warning, and degraded states.

[0047] Preferably, sensitive functional units can be grouped and a uniform hierarchical threshold can be set to facilitate regional comparison.

[0048] Then, in terms of the dynamic calculation of the health index across multiple time scales, this invention utilizes real-time or near-real-time multi-source monitoring data to continuously refresh the health index and perform time-series diagnosis at the sensitive functional unit scale.

[0049] Furthermore, during the operation phase, the data acquisition and fusion module continuously receives sensor and remote sensing data, and through preprocessing such as temporal resampling, spatial interpolation, and quality control, generates data for each sensitive functional unit at the diagnostic time. index vector According to the reference healthy time period The normalization rules and health index weights used The current health index of the sensitive functional unit is obtained by normalization and weighted summation. ,Right now:

[0050]

[0051] in, For use in calculating health index The number of evaluation indicators; For the first The weights of each evaluation indicator satisfy... ; Sensitive functional unit At any moment The The normalized value of each evaluation indicator.

[0052] To distinguish between sudden deterioration, slow degradation, and recovery processes, this invention preferably uses sliding windows with at least two time scales, such as a daily time scale window. 1 / 2 Scale Window Lunar Scale Window And define the change in health index. for:

[0053]

[0054] in, This represents the average health index within the sliding window.

[0055] By comparing changes in health index at different time scales , and Based on the numerical values ​​and trends of health indicators, this invention can identify sudden deterioration events with sharp fluctuations and significant declines in the short term, as well as slow but continuous declines in the medium to long term, and distinguish recovery phases with significant rebounds in health indicators.

[0056] In terms of dynamic diagnosis of health status and risk transmission analysis, this invention will use the current health index With health baseline index The health index is compared with the grading threshold range to provide a grading diagnosis result for sensitive functional units. Above the health baseline index Upper tolerance boundary, and change in health index When the change is small and there is no significant negative trend, the sensitive functional unit is judged to be in a healthy state; when the health index... At the health baseline index Neighborhood, but the change in health index When there is a slight but persistent decline, the sensitive functional unit is judged to be in a sub-healthy state; when the health index... Significantly below the health baseline index However, it has not yet reached the degradation threshold, and the change in health index on a weekly or monthly scale window is... A sustained decline in the health index triggers a warning; when the health index continues to decline... Below the degradation threshold, and the change in health index across multiple time scales. When there is a significant negative mutation or a continuous downward trend, the sensitive functional unit is determined to be in a degenerate state.

[0057] More importantly, this invention utilizes a hydrodynamic connectivity matrix on this basis. Establish a health risk transmission model to identify key pathways for the downstream spread of anomalies from the source region.

[0058] Furthermore, for any sensitive functional unit with hydrodynamic connections... The present invention defines a health risk transmission model as follows:

[0059]

[0060] in, The intensity of health risk transmission; Hydrodynamic connectivity matrix The elements in the text represent sensitive functional units. To sensitive functional units The proportion of flow or the frequency of water exchange; For sensitive functional units Health baseline index; For sensitive functional units Current health index.

[0061] When health index Below the health baseline index When, it indicates an upstream sensitive functional unit. The health status of the [unit] deteriorates, and this deterioration, weighted by the hydrodynamic connectivity, extends downstream to sensitive functional units. Transmission. This involves assessing the intensity of health risk transmission. Based on statistics over a period of time, this invention can set a threshold. When the intensity of health risk transmission on a certain link Continuously above the threshold At that time, it is identified as a high-risk transmission path, and one or more risk diffusion chains are formed between sensitive units in the source area, key transmission sensitive units, and sensitive units in the sink area, such as... Figure 5 As shown.

[0062] Furthermore, the present invention can further integrate or weight the cumulative risks on the path to measure the comprehensive risk contribution of different paths, thereby providing management departments with quantitative basis for prioritizing intervention paths and key control nodes.

[0063] Based on the above methods, the lake health dynamic diagnosis system proposed in this invention preferably includes three logical layers: a data acquisition and fusion layer, a model calculation and diagnosis layer, and a result display and early warning layer. The data acquisition and fusion layer includes fixed monitoring station terminals, buoy terminals, unmanned surface vessel monitoring terminals, and remote sensing data interfaces, and connects to a central server via a communication network to achieve real-time or near-real-time acquisition of water level, flow velocity, water quality, and remote sensing inversion data. The model calculation and diagnosis layer, implemented on the central server, includes modules for lake functional unit division, sensitive functional unit identification, health baseline and weight calculation, dynamic calculation of health index, and health status diagnosis and transmission analysis. The result display and early warning layer includes a graphical user interface, map services, and early warning push services, used to intuitively present diagnostic results in the form of spatial layers, time series curves, and network paths.

[0064] Specifically, the lake functional unit division module is used to import lake basin topography, water depth, hydrodynamics, and management zoning data, perform grid division, unit merging, and attribute annotation, and maintain the basic functional unit set in the form of a vector layer. The geometric and attribute information; the sensitive functional unit identification module reads the fused unit-time dataset and calculates the comprehensive sensitivity of each basic functional unit. Automatically filter the set of sensitive functional units This data is then highlighted on the map and categorized by function, forming a source-transmission-sink topology of sensitive functional units and hydrodynamic channels. The health baseline and weight calculation module is used to calculate the health index weight within the user-selected reference health period using a coupling method of information entropy weight and sensitivity weight. And generate a health baseline index for each sensitive functional unit. The health index dynamic calculation module periodically reads the latest monitoring data, performs normalization, index calculation, and sliding window analysis to obtain the health index of each sensitive unit. and the changes in health index at different time scales This generates a health level assessment result; based on this, the health status diagnosis and transmission analysis module calls the hydrodynamic connectivity matrix. Calculate the intensity of health risk transmission It identifies high-risk links and risk propagation paths, and outputs the risk transmission topology.

[0065] The results visualization and early warning module, based on a GIS platform, overlays the health levels of sensitive functional units onto a map using different colors or textures. It also provides a timeline control to replay the evolution of the lake's health status. For identified high-risk transmission paths, the system presents the source-transmission-sink chain on the interface in the form of directed edges or a Sankey graph, and users can click to view the health index curves and key influencing indicators of each node. When a sensitive functional unit enters an early warning or degradation state, or when the transmission intensity of a risk transmission path exceeds a threshold, the system pushes early warning information to management personnel via SMS, email, or platform notification.

[0066] Through the above technical solution, this invention organically integrates "sensitive functional unit identification, health baseline construction, multi-timescale health index calculation, risk transmission analysis under hydrodynamic constraints, and visual early warning" at a unified lake functional unit scale. This not only refines the spatial diagnostic units but also enhances the dynamic diagnosis and path identification capabilities, providing stable, interpretable, and operable technical support for the refined management of lake health.

[0067] By employing the above technical solution, the present invention provides a lake water environment monitoring method and system based on the monitoring of sensitive functional units, which has at least the following beneficial effects:

[0068] This invention, while using the conventional lake water environment and water ecology monitoring index system, introduces functional unit division and sensitive functional unit identification based on water depth pattern, hydrodynamic connectivity and functional zoning, so that health diagnosis is transformed from the whole lake or coarse-grained units to a relatively fine characterization of key areas; at the same time, a health baseline is constructed at the sensitive functional unit scale and a multi-time scale sliding window is used to calculate the health index and its changes, which is conducive to identifying the state evolution characteristics such as gradual degradation and short-term anomalies.

[0069] This invention calculates the intensity of health risk transmission through a hydrodynamic connectivity matrix, revealing to some extent the risk transmission relationship between the "source region, transmission region, and sink region," providing a quantitative reference for determining priority intervention units and control nodes. Under the premise of having the necessary monitoring and hydrodynamic data, it can improve the spatial resolution and dynamic identification capability of lake health diagnosis.

[0070] Based on lake hydrodynamics and functional zoning, the lake body is divided into basic functional units with clear hydrological and ecological significance. Key sensitive functional units are identified through sensitivity analysis. Multi-source monitoring data are integrated at a unified functional unit scale to construct a health baseline and dynamic health indices across multiple time scales. A health risk transmission model is established by combining the hydrodynamic connectivity matrix, enabling zoned, dynamic, and transmission path-based health diagnosis and early warning for the lake as a whole and key sensitive functional units. This overcomes the shortcomings of existing technologies in spatial precision, dynamic diagnostic capabilities, and risk transmission identification. Attached Figure Description

[0071] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0072] Figure 1 This is a flowchart of the lake water environment monitoring method in this invention;

[0073] Figure 2 This is a functional structure block diagram of the lake water environment monitoring system in this invention;

[0074] Figure 3 This is a schematic diagram of the spatial distribution of the sensitive functional units in this invention;

[0075] Figure 4 This is a general framework diagram of the lake functional unit division and multi-source monitoring data fusion in this invention;

[0076] Figure 5 This is a schematic diagram of the hydrodynamic connectivity and health risk transmission pathway in this invention;

[0077] Figure 6 This is a distribution diagram of the comprehensive sensitivity and health baseline index of the sensitive functional units in Embodiment 1 of the present invention;

[0078] Figure 7 This is a time series comparison chart of the health index of typical sensitive functional units in Embodiment 2 of the present invention;

[0079] Figure 8 This is a graph showing the changes in health index across multiple time scales in Embodiment 2 of the present invention.

[0080] Figure 9 This is a flowchart of the multi-source monitoring data acquisition and preprocessing process in Embodiment 3 of the present invention;

[0081] Figure 10 This is a diagram illustrating the design pattern of the system's core database in Embodiment 3 of the present invention. Detailed Implementation

[0082] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0083] Example 1: Basic functional unit division and sensitive functional unit identification of a typical lake.

[0084] This embodiment uses a medium-sized shallow lake A as an example to illustrate the basic functional unit division, sensitive functional unit identification, and healthy baseline construction process in the lake water environment monitoring method of the present invention. It should be understood that this embodiment is only used to illustrate the technical solution of the present invention and is not intended to limit the scope of protection of the present invention.

[0085] Lake A is a typical shallow lake in a plain, with a surface area of ​​approximately 30 km². 2 The lake has an average depth of approximately 2.5 meters and a maximum depth of approximately 6.0 meters. It has three tributaries flowing into the lake and one control section at its outlet. Functionally, the lake is divided into three main areas: the upstream inflow zone, the central landscape and recreation zone, and the downstream slow-flowing outflow zone. The area near the upstream inflow is influenced by both watershed non-point source and urban point source pollution. The hydrodynamics in the middle and lower reaches are relatively weak, making them prone to nutrient and algal accumulation. This embodiment uses multi-source monitoring data from 2020–2022 as the basic data source.

[0086] In this embodiment, the water boundary vector file of Lake A, 1:10000 lake basin water depth data, and two-dimensional hydrodynamic numerical simulation results (including multi-year average flow velocity field and residence time field) are first loaded on the GIS platform, and the existing drinking water source protection area, landscape recreation area and general functional area boundaries are superimposed.

[0087] Based on this, the lake area was initially divided into a regular grid of 250m × 250m, with each grid cell having an area of ​​approximately 0.0625 km². 2 After shoreline trimming, a total of 480 initial grids covering the water area were obtained.

[0088] Furthermore, in this embodiment, unit attributes are merged and labeled according to the following constraints: Based on water depth levels, the three depth zones of 0–2m, 2–4m, and 4–6m are marked as shallow water, medium water, and deep water zones, respectively; based on the two-dimensional hydrodynamic simulation results, units with an average flow velocity ≥0.05m / s and a residence time ≤15d are marked as highly connected units, units with an average flow velocity of 0.02–0.05m / s or a residence time of 15–30d are marked as moderately connected units, and the rest are marked as low-connectivity units; units located within drinking water source protection areas are marked as water source functional units, units located within landscape zones are marked as landscape functional units, and the rest are general functional units; units in contact with natural shorelines or ecological restoration shorelines are marked as ecological shoreline units, and units in contact with hard revetments are marked as engineering shoreline units. The merging rule is: adjacent grid units of the same water depth level, the same connectivity level, and the same functional zoning can be merged into one basic functional unit. After merging, a set of 215 basic functional units is obtained. There are approximately 65 upstream lake inflow areas, approximately 90 midstream landscape zones, and approximately 60 downstream slow-flow zones. Each basic functional unit carries the following field information: unit number, geometric boundary, average water depth, average flow velocity, average residence time, shoreline type, and functional type.

[0089] Based on the division of basic functional units, this embodiment deploys multi-source monitoring equipment within lake A and uniformly maps the raw observation data to the basic functional units and a unified time step. Specifically:

[0090] (1) Fixed monitoring stations and buoys:

[0091] Three fixed shore-based monitoring stations were deployed in the upstream lake inlet area, and four multi-parameter buoy stations were deployed in the middle and downstream areas. Monitoring parameters included: water temperature, pH, dissolved oxygen (DO), conductivity, total nitrogen (TN), total phosphorus (TP), and ammonia nitrogen (N). The sampling frequency was once per day, including chlorophyll a (Chl-a), transparency (measured manually by Seychian disk, daily scale value), etc.

[0092] (2) Unmanned surface vessel survey:

[0093] One unmanned surface vessel (USV) was deployed to conduct weekly surveys of Lake A, collecting surface flow velocity, local water quality (TN, TP, Chl-a), and water depth data at over 60 cross-sections along a pre-defined route. This data was used to correct the hydrodynamic model and supplement water quality information in areas with fewer monitoring points.

[0094] (3) Utilization of remote sensing data:

[0095] The surface Chl-a and water transparency of Lake A were retrieved using publicly available satellite data sources (e.g., 10-16 day revisit cycles), with a spatial resolution of approximately 10–30 m. Surface water color products for approximately 60 periods from 2020 to 2022 were obtained through atmospheric correction and a water color retrieval model.

[0096] (4) Data alignment and mapping:

[0097] Various data types are resampled to a daily scale, with the average value from 00:00 to 24:00 representing the daily status. Spatially, observations from fixed stations and buoys are mapped to basic functional units within the coverage area using nearest neighbor or inverse distance weighted interpolation; remote sensing pixels are assigned to corresponding functional units through area overlap; and unmanned surface vessel (USV) survey data is used to correct for biases in the spatial interpolation results. Ultimately, a two-dimensional dataset of functional unit-time is formed, i.e., for each basic functional unit... and each daily time Obtain the set of stress factors and response indicator set .

[0098] In this embodiment, the pressure factor integrates information such as inflow intensity, water level deviation, and wind speed; the response index vector... Variables include TN, TP, Chl-a, transparency SD, and dissolved oxygen. For simplicity, the subsequent sensitivity calculations will only demonstrate four indicators: TN, TP, Chl-a, and SD.

[0099] According to the method of the present invention, for each basic functional unit and each response metric (TN, TP, Chl-a, SD) Calculate the average values ​​during the monitoring period. and standard deviation The coefficient of variation was obtained. ,Right now:

[0100]

[0101] At the same time, a comprehensive pressure intensity index is constructed. The result is obtained by weighted summation after normalization of inflow, daily average water level deviation, and daily average wind speed, and the Pearson correlation coefficient between the time series of response indicators and the pressure intensity series is calculated. Unit – Indicator Sensitivity Defined as:

[0102]

[0103] In this embodiment, for ease of explanation, three representative basic functional units are selected from the 215 basic functional units for calculation: Unit A ( Unit B is located near the upstream inlet of the lake, characterized by shallow water, high connectivity, and general functional functions. ): Located in the central water area of ​​the landscape belt, it is a centrally connected landscape functional unit; Unit C ( ): Located in the downstream slow-flowing zone of the lake, in shallow water, with low connectivity, and near the boundary of the water source protection area.

[0104] Statistical analysis was performed on the above three units based on daily-scale data from 2020 to 2022, and the typical results are as follows (only the statistical results of the mean and standard deviation are listed; the original data can be reproduced from the monitoring records):

[0105] Unit A:

[0106] TN: Therefore , with the comprehensive pressure intensity index correlation coefficient ,but ;

[0107] TP: , Correlation coefficient , ;

[0108] Chl-a: , Correlation coefficient , ;

[0109] SD: , Correlation coefficient , .

[0110] Unit B:

[0111] TN: , , , ;

[0112] TP: , , , ;

[0113] Chl-a: , , , ;

[0114] SD: , , , .

[0115] Unit C:

[0116] TN: , , , ;

[0117] TP: , , , ;

[0118] Chl-a: , , , ;

[0119] SD: , , , .

[0120] In this embodiment, the sensitivity weights of the response indicators are determined based on expert experience and management needs: , , , ,satisfy The overall sensitivity of each unit for:

[0121]

[0122] Substituting the above values, we get:

[0123] Unit A:

[0124]

[0125] Unit B:

[0126]

[0127] Unit C:

[0128]

[0129] Calculate the overall sensitivity for all 215 basic functional units. Subsequently, statistics revealed the overall sensitivity. The 70th percentile is approximately 0.10, and the 80th percentile is approximately 0.13. In this embodiment, a sensitivity threshold is selected. Then it satisfies There are a total of 62 units, which constitute the sensitive functional unit set. Unit A is a sensitive functional unit, while the combined sensitivity of units B and C is... Slightly below the sensitivity threshold Therefore, it is not included in the set of sensitive functional units. .

[0130] After identifying sensitive functional units, this embodiment further constructs a health baseline index at the sensitive functional unit scale. and health index weight First, referring to the water quality records and algal bloom monitoring records of the watershed management department, the six months in 2021 during which the water quality was consistently within acceptable limits (e.g., March–May and October–December), excluding the summer algal bloom period, were selected as the reference healthy period for Lake A. During this period, TN, TP, Chl-a, and SD data for all sensitive functional units were normalized: pollution indicators (TN, TP, Chl-a) were normalized using a lower-than-ideal inverse normalization, i.e.:

[0131]

[0132] Transparency (SD) is considered a metric where higher is always better, so it is standardized, i.e.:

[0133]

[0134] in, For reference to healthy time periods The maximum and minimum values ​​of the corresponding internal indicators across all sensitive functional units.

[0135] For example, during the reference health period Within this range, statistical analysis showed that TN (total nitrate) in sensitive functional units ranged from 0.80–2.00 mg / L, TP from 0.05–0.15 mg / L, Chl-a from 10–50 μg / L, and transparency SD from 0.40–1.50 μm. For unit A mentioned above, during the reference healthy period... Assuming the average values ​​of the indicators within the range are: TN = 1.40 mg / L, TP = 0.09 mg / L, Chl-a = 30 μg / L, and SD = 0.80 m, then the normalized result is:

[0136] TN (lower is better):

[0137]

[0138] TP (lower is better):

[0139]

[0140] Chl-a (lower is better):

[0141]

[0142] SD (higher is better):

[0143]

[0144] Based on this, this embodiment uses a method that couples information entropy weight and sensitivity weight to determine the health index weight. On the one hand, the entropy weight vector is calculated based on the spatial-temporal variability of four indicators within a reference healthy period. For example, the result in this embodiment satisfies ,for:

[0145]

[0146] On the other hand, a sensitivity weight vector is constructed based on the average sensitivity of each indicator in the sensitive functional unit. For example, the raw average sensitivity of TN, TP, Chl-a, and SD can be denoted as:

[0147]

[0148] After normalization, we get:

[0149]

[0150] In this embodiment, a balance coefficient is selected. Then the health index weight for:

[0151]

[0152] Substituting the values, we get:

[0153]

[0154]

[0155]

[0156]

[0157] The above results clearly satisfy .

[0158] Based on this, the sensitivity of functional unit A during the reference health period can be calculated. Health baseline index For ease of explanation, this embodiment uses the above-mentioned average normalized value to approximate the reference healthy period. If the average normalized level of the indicators within the range is:

[0159]

[0160]

[0161] For all sensitive functional units Calculate the health baseline index using the same method. This allows us to obtain the distribution of the health baseline index in sensitive functional units, such as... Figure 6 As shown. Statistical analysis revealed that the health baseline index of the sensitive functional units in this embodiment ranges from approximately 0.45 to 0.75, with a median of approximately 0.55. Based on this distribution and combined with management requirements, health status is categorized into levels; for example, the health index can be... The thresholds are defined as follows: 0.45–0.60 is defined as healthy; 0.30–0.45 is defined as warning; and below 0.30 is defined as degradation. This grading will be used for dynamic diagnosis and early warning in subsequent embodiments. The core of this embodiment is to demonstrate how the present invention achieves sensitive unit identification and health baseline construction at a unified functional unit scale. This embodiment shows that the functional unit division, sensitivity calculation, and health baseline construction method proposed in this invention can be implemented in real lake scenarios based on available monitoring data. The calculation process is clear, the parameter setting logic is clear, and the relevant indicators and values ​​are all within the scientifically reasonable range for water environment. It can be directly reproduced or referenced by peers under similar water body and data conditions.

[0162] Example 2: Lake health dynamic identification and risk transmission analysis based on multi-timescale diagnosis.

[0163] This embodiment focuses on demonstrating how to identify short-term anomalies and slow degradation using multi-timescale health indices at the sensitive functional unit scale, and how to quantitatively characterize the health risk transmission relationship between the "source region—transmission region—sink region" by combining hydrodynamic connectivity matrices. This embodiment is also only used to illustrate the technical solution of the present invention, and not to limit the scope of protection of the present invention.

[0164] In Example 1, based on multi-source monitoring data from 2020–2022, Lake A was divided into 215 basic functional units, and 62 sensitive functional units were identified. This embodiment selects three sensitive functional units with typical locations and strong functional representation, which are denoted as follows: Sensitive Unit S1: Located in the upstream estuary fan-shaped area, shallow water, highly connected, general functional unit, mainly bearing the impact of external pollution loads; Sensitive Unit S2: Located in the main channel in the middle of the lake, medium water, medium connectivity, landscape functional unit, meeting both water quality and landscape requirements; Sensitive Unit S3: Located in the downstream slow flow zone near the water source protection boundary, shallow water, low connectivity, water source function sensitive unit.

[0165] In the construction of the health baseline in Example 1, the health baseline indices of the three sensitive units are as follows: , , .

[0166] This embodiment selects a typical high-flow-rate lake inflow and a 20-day period before and after the summer algal bloom in 2022 as the target diagnostic period, denoted as days 1 to 20. The health index is calculated using the same indicator system and weights as in Example 1: the weights for TN, TP, Chl-a, and SD are as follows. , , , The index normalization method is consistent with the reference healthy period. This invention normalizes the multi-source monitoring data for each day during the diagnostic period and calculates the daily-scale health index of the sensitive unit using the following formula:

[0167]

[0168] For ease of explanation, this embodiment provides the health index sequence (retaining two decimal places) of the three sensitive units over 20 days as shown in Table 1.

[0169] Table 1. Health index of three sensitive units within 20 days of diagnosis.

[0170]

[0171] As shown in Table 1, in the early diagnostic phase (days 1-5), the health indices of the three sensitive units were all slightly higher than their respective baseline indices, indicating that the lakes were in a relatively stable healthy or sub-healthy state. Following a heavy rainfall-high flow inflow process, starting from day 6, the health index of S1 gradually decreased, showing a significant drop on days 9-10, reaching a low of 0.40 on day 10, significantly lower than its baseline of 0.55, corresponding to a warning state. The health index of S2 slowly decreased from day 8 to day 12, reaching a low of 0.51 on day 11, slightly below the baseline but still on the edge of sub-health to warning. The health index of S3 only showed a slight decline, reaching a low of 0.58 on day 11, still close to its baseline of 0.62, indicating a sub-healthy or near-healthy state. Figure 7 As shown.

[0172] like Figure 8 As shown, to quantitatively identify short-term abnormalities, this embodiment sets a 7-day sliding window on a weekly scale, and the calculated change in the health index is as follows:

[0173]

[0174] in, For the first From the 1st to the 1st The 7-day average health index. Taking day 10 as an example, the results for the three sensitive units are as follows (rounded to three decimal places): S1: S2: S3: It can be seen that within the same 7-day window, the health index of the upstream sensitive unit S1 showed a sharp negative bias of approximately 0.10 relative to the previous average, significantly higher than the short-term changes in S2 and S3. This indicates that a sudden deterioration event occurred in S1, while S2 and S3 were less affected and their responses were relatively delayed. This is consistent with the physical process observed during the monitoring period of "short-term high-load impact of tributaries flowing into the lake—slow response in the lake center and downstream areas."

[0175] like Figure 8 As shown, in addition to short-term anomalies, this invention also identifies slow degradation and recovery trends through a longer time window. In this embodiment, for ease of demonstration, a 15-day sliding window is used on a bi-monthly scale, and the calculated change in the health index is as follows:

[0176]

[0177] Taking day 15 as an example, based on the health index sequence of the previous 15 days in Table 1, the 15-day average of the three sensitive units was compared with... The calculation results are as follows:

[0178] S1: S2: S3: .

[0179] It can be seen that although the daily health index of S1 on the 15th day has rebounded from its trough to 0.48, it still has a negative bias of about 0.015 relative to the 15-day window mean, indicating that after the sudden deterioration in the short term, the overall condition is still in a slow recovery phase that is slightly below the recent average level. The 15-day bias of S2 is close to 0, indicating that the central sensitive unit generally maintained slight fluctuations during these 15 days. The bias of S3 is slightly positive, indicating that the downstream sensitive unit is generally maintaining or slightly better than the recent average level. Through this multi-timescale comparison, the present invention can simultaneously reveal the process characteristics from the sharp deterioration at the lake inlet on the 10th day to the incomplete recovery on the 15th day, rather than just the judgment of the health level at a single point in time.

[0180] In Example 1, the hydrodynamic connectivity matrix of lake A was constructed based on the results of two-dimensional hydrodynamic numerical simulation. This example focuses on the connectivity relationships between three sensitive functional units, extracting the following upper triangular hydrodynamic connectivity submatrix (the remaining elements can be considered as 0):

[0181]

[0182] in, This indicates that the amount of water flowing from S1 to S2 per unit time accounts for 60% of the outflow from S1. This indicates that 10% of the water flows directly from S1 to S3. This indicates that the water flowing from S2 to S3 accounts for 50% of the outflow from S2. This connectivity structure reflects the main hydrodynamic path of "upstream inlet → central main channel → downstream slow-flow zone". According to the risk transmission definition of this invention, for any sensitive unit with hydrodynamic connectivity... The intensity of health risk transmission is:

[0183]

[0184] In this embodiment, taking the 10th and 15th days as examples, the risk transmission intensity of S1→S2 and S2→S3 is calculated respectively. According to Table 1:

[0185] Day 10: , The difference is , , , , The difference is , .

[0186] Day 15: , The difference is , , ; , The difference is , .

[0187] In this embodiment, the risk transmission intensity threshold is... A value of 0.05 was used to identify high-risk transmission links. It can be seen that on the 10th day... This indicates that the health risks from S1 have had a significant transmission effect on S2 through the upstream-middle main channel, while , All values ​​were below the threshold, indicating that the risk transmission from S2 to S3 and from S1 to S3 was still at a low to medium level at this time; by the 15th day, as the health index of S1 partially recovered, The value drops to 0.042, which is below the threshold, thus resolving the high-risk transmission state from S1 to S2. and The further reduction only reflects a weaker downstream transmission risk. Combining the changes in the health indices of S2 and S3 in Table 1, it can be seen that during the period from day 8 to day 12, the health index of S2 decreased from around 0.55 to 0.51-0.52, while the health index of S3 only slightly decreased from around 0.60 to 0.58-0.59. This spatial-temporal evolution is consistent with the calculation results of the risk transmission intensity mentioned above. That is, the significant deterioration of the upstream sensitive functional unit S1 is first manifested in its own short-term sharp decline in the health index, and then transmitted to the middle sensitive unit S2 through the main channel, causing S2 to exhibit a moderate degree of delayed response. The impact on the downstream sensitive unit S3 is more mild and delayed.

[0188] This embodiment utilizes real, obtainable monitoring data and hydrodynamic simulation results, selecting three representative sensitive functional units to demonstrate in detail the entire process of calculating the daily-scale health index at the unit scale, identifying short-term anomalies and slow degradation trends through 7-day and 15-day sliding windows, and quantitatively calculating the intensity of health risk transmission in conjunction with the hydrodynamic connectivity matrix. The results show that the multi-timescale health diagnosis and source-transmission-sink risk transmission analysis method proposed in this invention maintains good consistency with the actual lake evolution process, with a clear calculation process, traceable parameters, and good feasibility and reproducibility.

[0189] Example 3: Hardware and software implementation of a lake health dynamic diagnosis system.

[0190] This embodiment, combining measured and simulated data from Lake A, presents a specific engineering implementation scheme for the lake water environment monitoring system of the present invention, illustrating the system's hardware configuration, software modules, data flow, and operational effects. This embodiment is also only used to illustrate the technical solution of the present invention and is not intended to limit the scope of protection of the present invention.

[0191] (1) System overall architecture and hardware configuration.

[0192] The lake water environment monitoring system constructed in this embodiment adopts a three-tier architecture of front-end monitoring equipment + communication network + central server + client visualization terminal. The central side is configured with an application server to run the various computing and diagnostic modules of this invention. Specifically, it is configured with an 8-core CPU, 32GB of memory, a 2TB hard disk drive + a 512GB SSD, and runs on Linux. It deploys a relational database (for storing basic attributes and result data) and a time-series database (for storing continuous monitoring data). The front-end monitoring equipment includes: 3 fixed shoreline water quality-hydrology monitoring stations (corresponding to the 3 stations deployed near the upstream lake inlet in Embodiment 1); 4 multi-parameter buoy monitoring stations (distributed in the central landscape zone and the downstream slow-flow zone); 1 unmanned surface vessel platform (for weekly patrols and hydrodynamic model verification), which connects to the system as a mobile monitoring node; and 1 remote sensing data access interface module (obtaining Chl-a and transparency products from satellite inversion via API).

[0193] Key parameters for fixed monitoring stations and buoy stations are set as follows: water temperature, pH, DO, TN, TP. Eight main indicators were measured, including water temperature, pH, dissolved oxygen (DO), and conductivity. Samples of TN and TP were taken at 15-minute intervals. Chl-a uses a 1-hour sampling period. To simplify data management, this embodiment resamples all indicators to a 15-minute time step before data is stored, meaning each indicator is updated every 15 minutes (for indicators with a 1-hour period, the same value is written at each of the four 15-minute intervals within the same hour). The daily data collection volume can be estimated as follows: Each front-end monitoring node (fixed station or buoy station) collects 8 indicators every 15 minutes, totaling 96 time points in 24 hours. Therefore, the daily record count for a single node is... The system has a total of 7 fixed / buoy nodes, so the total number of daily monitoring data records is approximately: The unmanned surface vessel (USV) conducts weekly surveys, covering approximately 60 monitoring points each time. Each point collects five parameters (flow velocity, TN, TP, Chl-a, and water depth), generating about 60 x 5 = 300 records per survey, significantly less than the data volume of fixed monitoring stations. Remote sensing inversion products are generated on average every 10–16 days, with each scene producing several hundred to over a thousand pixel records for Lake A. Overall, the system's daily data input is in the thousands, within the processing capacity of the server and database.

[0194] (2) Data collection, coding and storage process.

[0195] The front-end monitoring equipment connects to the public network via a 4G / 5G communication module and then to the central server via a VPN secure channel. At the data acquisition end, all devices report monitoring results according to a unified data message format. The data message fields include: device number, observation time, latitude and longitude, parameter number, observed value, unit, quality control mark, etc.

[0196] After the data enters the central server, the specific processing flow for its collection, cleaning, and storage is as follows: Figure 9 As shown. First, the data receiving and encoding submodule parses the message, automatically identifying its functional unit range based on the device number and configuration table; then, the time alignment submodule converts all data to the UTC+8 time base used internally by the system; the quality control submodule marks the observations according to preset rules (such as physical reasonableness range, short-term mutation threshold, sensor self-test status, etc.), marking obviously abnormal data as suspicious or unreliable, but retaining them in the database for later review. The data processed in the above way is written into the time series database, with the following table structure:

[0197] Fields: unit_id (functional unit number), device_id, timestamp, parameter_code, value, qc_flag;

[0198] Where unit_id is the mapped functional unit number, parameter_code corresponds to indicators such as TN, TP, Chl-a, and SD, and qc_flag represents the quality control result.

[0199] To support the data access and correlation analysis needs of subsequent calculation modules, this system has constructed a core database that includes basic information of functional units, monitoring time-series data, health baselines, and diagnostic results. Its overall table structure and relationships are as follows: Figure 10 As shown. In this way, the system can quickly and efficiently retrieve monitoring data at any time according to the dimensions of "functional unit - time interval - indicator", providing a data foundation for subsequent sensitivity calculation, health index calculation and risk transmission analysis.

[0200] (3) Implementation of functional unit management and spatial mapping module.

[0201] In Example 1, Lake A was divided into 215 basic functional units. Each unit's spatial boundaries are stored in the database as vector polygons, and it is associated with attributes such as average water depth, average flow velocity, residence time, and shoreline type. In this example, the functional unit management module implements the following functions:

[0202] Spatial Query and Mapping: The system supports point identification within polygons using built-in spatial indexes (such as R-trees). Whenever new monitoring data (including latitude and longitude) is received, the system automatically calls the spatial query function to match the observation point to a unique functional unit (unit_id). For example, if the coordinates of fixed station ST01 fall within unit u_37, then all observation records from ST01 will be uniformly assigned unit_id = 37 upon data entry.

[0203] Unit attribute maintenance and editing: If the boundaries of functional units need to be adjusted in subsequent engineering practices (such as adding shoreline restoration projects or adjusting the water source protection area), the system allows managers to split or merge unit boundaries through the map interface and automatically update the relevant attribute tables; historical monitoring data that has been entered into the database is still saved with the original unit_id to ensure traceability.

[0204] Sensitive Functional Unit Identification and Grouping: In this embodiment, the 62 sensitive functional units identified in Embodiment 1 are marked as sensitive = 1 in the unit attribute table, and a role field is added for three categories: source area, transmission area, and sink area. Subsequent dynamic diagnosis and risk transmission analysis will, by default, prioritize execution within the sensitive functional unit set.

[0205] Through the above design, this system tightly couples spatial information with monitoring data, ensuring that monitoring data at any point in time can be uniquely mapped to a certain functional unit, thus laying the foundation for diagnosis based on sensitive functional units.

[0206] (4) Systematic implementation of sensitive functional unit identification and health baseline calculation.

[0207] In this embodiment, the sensitive functional unit identification module and the health baseline and weight calculation module run in batch processing on the central server. They are usually automatically triggered at 2:00 AM every day to update or incrementally calculate the data from the past three years, thereby supporting the regular updates of sensitive unit identification and the correction of the health baseline index.

[0208] Sensitivity calculation and unit selection:

[0209] The system selects data from the most recent three years as the time window for sensitivity analysis based on preset configurations. For example, when executing in January 2023, data from 2020 to 2022 is selected; future years can be updated on a rolling basis. The module automatically aggregates data from the time series database by unit_id and parameter_code, and calculates the average value of each indicator for each unit. with standard deviation Then calculate the coefficient of variation. ,Right now:

[0210]

[0211] Simultaneously read the comprehensive pressure intensity index (The response index is calculated by normalizing and weighting indicators such as inflow rate, water level deviation, and wind speed.) With pressure intensity index Pearson correlation coefficient between According to the coefficient of variation Correlation coefficient with Pearson Define Unit – Indicator Sensitivity ,Right now:

[0212]

[0213] Then according to the preset weight The basic functional units are summarized as follows Overall sensitivity And calculate the quantiles of the overall sensitivity of all units. That is:

[0214]

[0215] in, .

[0216] The system by default marks units with a comprehensive sensitivity score not lower than the 70th percentile as sensitive functional units. Administrators can adjust the threshold and re-filter as needed. In the deployment at Lake A, the system's calculation results are consistent with those in Example 1: 62 out of 215 basic functional units meet the criteria. It is automatically marked as a sensitive functional unit.

[0217] Automated process for calculating health baselines and weights:

[0218] The health baseline calculation module first reads the reference health period (e.g., March–May and October–December 2021) set by the user on the interface. Then, it automatically extracts the TN, TP, Chl-a, SD, and other indicator data of sensitive functional units within that period from the database, performs normalization processing according to the formula described in Example 1, and calculates the information entropy weight of each indicator. Simultaneously, the module reads the sensitivity analysis results, normalizes the average sensitivity of the four indicators, and obtains... .

[0219] In this embodiment, the information entropy weight automatically calculated by the system is:

[0220]

[0221] The sensitivity normalization weights are:

[0222]

[0223] In balance coefficient Under this configuration, the system automatically provides a comprehensive weighted health index:

[0224]

[0225] satisfy .

[0226] Subsequently, the system analyzed each sensitive functional unit. Calculate the health baseline index using the following formula within the reference health period and write the results into the health_baseline table.

[0227]

[0228] Taking the sensitive unit S1 in Example 2 as an example, the baseline index automatically calculated by the system is about 0.55, which is consistent with the calculation result, verifying the correctness of the module calculation.

[0229] (5) Implementation of dynamic health diagnosis and risk transmission analysis module.

[0230] In this embodiment, the health index dynamic diagnosis module and the risk transmission analysis module operate in the form of daily scheduling tasks, automatically performing health index calculation and risk transmission analysis on the data of the previous day (or the most recent few days) at 01:00 every day.

[0231] Systematic calculation of health indices across multiple time scales:

[0232] The module reads the previous day's indicator data from the time series database according to sensitive functional units, processes the data for every 15 minutes using the aforementioned normalization formula, and aggregates it according to a set daily scale (e.g., taking the daily average or daily midpoint) to obtain the daily indicator value. Then, it calculates the daily health index using the following formula:

[0233]

[0234] Subsequently, the module retrieves the required historical data from the health_index table according to the configured sliding window length (7d and 15d scales are set in this embodiment) and automatically calculates the change in health index. , And according to threshold rules (e.g.) Determining short-term abnormalities (To determine slow degradation) Generate a health status flag field named status, with values ​​such as healthy, sub-healthy, warning, and degradation.

[0235] In the 20-day health index given in Example 2, the system operation results are consistent with manual calculation: for example, on the 10th day, the 7-day sliding deviation of S1 calculated by the system is about -0.097, which is automatically marked as a short-term abnormality, while the 7-day deviations of S2 and S3 are about -0.037 and -0.017 respectively, which are marked as slight fluctuations.

[0236] Calculation of hydrodynamic connectivity matrix and risk transmission intensity:

[0237] During system initialization, the risk transmission analysis module reads the results of two-dimensional hydrodynamic numerical simulation, performs statistical analysis on the flow exchange between 215 functional units, and generates a complete hydrodynamic connectivity matrix. And store it in the connectivity_matrix table. Elements between two units. In the hydrodynamic connectivity matrix The value in the middle is a constant or a segmented constant based on the season. During the daily diagnosis period, the module calls the health_baseline and health_index tables, and for each pair of connected units... Automatically calculate the intensity of health risk transmission The results are then written to the risk_propagation table.

[0238] Administrators can configure risk propagation thresholds on the interface. (In this embodiment, the value is set to 0.05). The system automatically identifies high-risk transmission links based on the condition R_{ij}(t) > R_{\mathrm{th}} and generates a list of risk paths. Taking the 10th day in Embodiment 2 as an example, the system finds that the health baseline index of S1 is 0.55 and the health index on that day is 0.40, and then automatically calculates the difference. and , Multiplying them together gives:

[0239]

[0240] Simultaneously for S2→S3, the system according to and The calculation yielded:

[0241]

[0242] Therefore, the system automatically marks the "S1→S2" link as a high-risk transmission path and highlights the risk path of S1→S2 with a red arrow in the graphical interface, indicating the significant impact of upstream anomalies on the central main channel.

[0243] (6) Results visualization and early warning information push.

[0244] The results visualization and early warning module in this embodiment is implemented based on Web GIS technology, providing managers with map views, time series views, and risk network views. Figure 3 The interface is as follows: In the map view, 215 functional units are displayed as overlaid polygon layers, with sensitive functional units marked with more prominent boundaries. Each unit is colored according to its current health status (e.g., green for healthy, yellow for sub-healthy, orange for warning, and red for deterioration). Users can use the timeline slider to replay the spatial distribution of any date. In the time series view, after a user clicks on a sensitive functional unit, the interface will display the health index curve of that unit within the selected time period. and 7d and 15d sliding window deviation curves , The system marks the time periods of short-term anomalies and slow degradation. In the risk network view, sensitive functional units are used as nodes, and risk transmission links exceeding the threshold are used as directed edges to construct the network graph. High-risk paths are indicated by thick red arrows, and users can click to view a specific path. Changes over time.

[0245] When the system detects that: a sensitive functional unit is in a warning or degradation state for 3 consecutive days; or a risk transmission link is in a warning or degradation state for more than 2 consecutive days. When the warning module is activated, it will automatically send text messages and emails to preset contacts (such as technical personnel of the lake management agency). The messages will include the abnormal unit (or path) number, location description, health index and risk intensity values, and suggested attention period, so that the management department can take timely on-site investigation and emergency control measures.

[0246] During a year-long trial operation at Lake A, the system constructed in this embodiment successfully identified the sudden drop in the health index of upstream sensitive functional units and the risk transmission path of the S1→S2 main channel during multiple high-flow-rate events and local algal blooms. Compared with manual post-event analysis, the sensitive functional unit ranking, short-term abnormal time windows, and main risk paths provided by the system were largely consistent with the judgments of monitoring personnel based on raw data and on-site records, indicating that the system has good scientific rationality and operability. This embodiment demonstrates that the lake water environment monitoring system proposed in this invention can complete the entire process of data acquisition, functional unit mapping, sensitive functional unit identification, health baseline construction, multi-timescale health diagnosis, and risk transmission analysis in a real lake scenario using existing monitoring conditions. The relevant parameter settings, data scale, and computational complexity are all achievable and reproducible under conventional engineering conditions, demonstrating high engineering application value.

[0247] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0248] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0249] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A lake water environment monitoring method based on sensitive functional unit monitoring, characterized in that, The method includes the following steps: Based on the division constraints, the lake area is divided into several two-dimensional spatial units, forming a system containing multiple basic functional units. Basic functional unit set And collect data from each basic functional unit within the preset monitoring period. Multi-source monitoring data were used to construct a set of stress factors. and response indicator set Unit-time dataset; Calculate the data from the unit-time dataset to extract from the basic functional unit set. Selecting sensitive functional unit sets Overall sensitivity and set up sensitive functional units Functional types are categorized to form a source-transmission-sink topology of sensitive functional units and hydrodynamic channels, thereby selecting a set of sensitive functional units. include: For each basic functional unit and each response indicator Calculate the average value within the monitoring period. and standard deviation And obtain the coefficient of variation. ,Right now: ； At the same time, the set of stress factors The comprehensive pressure intensity index is constructed by standardizing and weighting the multiple pressure factors included. Then calculate the response indicators. With pressure intensity index Pearson correlation coefficient between ; Based on the coefficient of variation Correlation coefficient with Pearson Define Unit – Indicator Sensitivity ,Right now: ； Then according to the preset weight The basic functional units are summarized as follows Overall sensitivity ,Right now: ； in, For use in calculating the overall sensitivity of basic functional units The number of response indicators; For the first Sensitivity weights for each response metric, satisfying ; Basic functional unit In the Sensitivity scores on each response metric; Determine the sensitivity threshold based on sensitivity percentiles or cluster analysis. , will satisfy Basic functional units Included in the sensitive functional unit set ; During the reference health period Internally, a method combining information entropy weighting and sensitivity weighting is used to calculate the health index weight of each sensitive functional unit. And generate health baseline indices for each sensitive functional unit. And the tiered threshold range; Based on the real-time or near-real-time monitoring data of each sensitive functional unit, calculate the current health index of each sensitive functional unit. and the changes in health index at different time scales This is used to dynamically diagnose the health status of each sensitive functional unit; A health risk transmission model based on the source-transmission-sink topology is established to identify key pathways for the downstream spread of anomalies from the source region, enabling graded diagnosis of the health status of the lake as a whole and identification of transmission pathways for key sensitive functional units.

2. The lake water environment monitoring method according to claim 1, characterized in that, In the lake area division, the basic functional units satisfy at least two of the following division constraints: Water depth classification constraints: Zones are divided according to water depth ranges of 0-2m, 2-5m, and above 5m; Hydrodynamic connectivity constraints: Based on flow velocity, residence time and water turnover rate, units that have direct connectivity with the inflow estuary or outflow control section are classified as high connectivity zones, and the rest are classified as medium and low connectivity zones. Shoreline type constraints: Based on the degree of shoreline hardening and nearshore vegetation type, units close to natural shorelines and restored shorelines are marked as ecological shoreline units; Functional utilization constraints: Basic functional units are tailored based on the zoning boundaries, including drinking water sources, landscape belts, and general functional areas.

3. The lake water environment monitoring method according to claim 1, characterized in that, The functional type classification includes: In the collection of sensitive functional units The above is used to construct a hydrodynamic connectivity matrix to characterize the spatial propagation path of health risks in the lake. ; Based on the results of two-dimensional or three-dimensional hydrodynamic numerical simulations, statistically sensitive functional units are analyzed. and Daily flow exchange volume And normalize it to the percentage of traffic per unit time; matrix elements Characterization of sensitive functional units To sensitive functional units The proportion of flow or frequency of water exchange is expressed as: ； By analyzing the upstream contribution and downstream impact characteristics of each sensitive functional unit, the sensitive functional units are divided into: source sensitive units located at the main inflow estuary and upstream and with significant downstream output; transmission sensitive units located in the middle of the main hydrodynamic channel and with strong connections to both upstream and downstream; and sink sensitive units with long residence time and prone to becoming pollution or algal bloom accumulation areas, forming a source-transmission-sink topology structure of sensitive functional units-hydrodynamic channel.

4. The lake water environment monitoring method according to claim 1, characterized in that, The generation of health baseline indices for each sensitive functional unit And the tiered threshold range, including: Based on the water quality compliance status, eutrophication index, algal bloom records, and ecological survey results of sensitive functional units, typical healthy periods over several years or hydrological years are selected as reference healthy periods. ; During the reference health period Within, for each response metric Calculate the normalized value distribution of each sensitive functional unit and time point to obtain the corresponding information entropy. Information entropy weight ; The response indicators of each sensitive functional unit Sensitivity score Normalization along the indicator dimension yields The health index weights of sensitive functional units are calculated using the following formula. ,Right now: ； in, The coefficients for balancing entropy weight and sensitivity weight satisfy the following conditions: ,and ; Based on reference health time period Internal response indicators The calculated information entropy weights; The normalized results are the sensitivity scores of each response indicator; Based on health index weighting Calculate the reference health period for each sensitive functional unit. Health baseline index within ,Right now: ； in, This represents the normalized value obtained by normalizing the various response indicators of each sensitive functional unit using the range method or the Z-score method. Based on statistical distribution, a tiered threshold range is set for each sensitive functional unit, including healthy, sub-healthy, early warning, and deterioration.

5. The lake water environment monitoring method according to claim 4, characterized in that, The health index The calculation formula is: ； in, For use in calculating health index The number of evaluation indicators; For the first The weights of each evaluation indicator satisfy the following: ; Sensitive functional unit At any moment The Normalized values ​​of each evaluation indicator; The different time scales include daily-scale windows. 1 / 2 Scale Window Lunar Scale Window The change in the health index The calculation formula is: ； in, This represents the average health index within the sliding window.

6. The lake water environment monitoring method according to claim 5, characterized in that, The dynamic diagnosis of the health status of each sensitive functional unit includes: Current health index With health baseline index By comparing the grading threshold range, the grading diagnostic results of the sensitive functional units are obtained, namely: When health index Above the health baseline index The upper tolerance boundary, and the change in the health index When the change is small and there is no significant negative trend, the sensitive functional unit is judged to be in a healthy state; When health index At the health baseline index The neighborhood, but the change in the health index If a continuous decline is observed, the sensitive functional unit is determined to be in a sub-healthy state. When health index Below the health baseline index However, it has not yet reached the degradation threshold, and the change in health index on a weekly or monthly scale window is... A sustained decline is considered a warning sign. When health index Below the degradation threshold, and the change in health index across multiple time scales. When a negative mutation or a continuous downward trend is observed, the sensitive functional unit is determined to be in a degenerate state.

7. The lake water environment monitoring method according to claim 1, characterized in that, The establishment of the health risk transmission model includes: For any sensitive functional unit with hydrodynamic connection Based on hydrodynamic connectivity matrix Establish a health risk transmission model, the expression of which is: ； in, The intensity of health risk transmission; Hydrodynamic connectivity matrix The elements in the text represent sensitive functional units. To sensitive functional units The proportion of flow or the frequency of water exchange; For sensitive functional units Health baseline index; For sensitive functional units Current health index; Based on the intensity of health risk transmission To determine the high-risk transmission path, the intensity of health risk transmission along a certain link is... Continuously above the threshold When this occurs, it is identified as a high-risk transmission path, and one or more risk diffusion chains are formed between sensitive units in the source region, sensitive units in key transmission regions, and sensitive units in the sink region.

8. A system for implementing the lake water environment monitoring method according to any one of claims 1-7, characterized in that, include: The lake functional unit division module is used to acquire information on the topography, water depth, hydrodynamics, and functional zoning of the lake area, and to divide the lake area into multiple basic functional units according to water depth level, hydrodynamic connectivity, and shoreline type. ; The multi-source monitoring data acquisition module is used to connect to water level gauges, flow meters, online water quality monitors, buoy sensors, unmanned surface vessel monitoring terminals, and remote sensing data interfaces. It collects and stores multi-source monitoring data from each basic functional unit and constructs a set of pressure factors. and response indicator set Unit-time dataset; The sensitive functional unit identification module calculates the comprehensive sensitivity of each basic functional unit based on multi-source monitoring data, selects a set of sensitive functional units and classifies them by function type, forming a source-transmission-sink topology structure of sensitive functional unit-hydraulic channel; The health baseline and weight calculation module is used to calculate the health index weight of each sensitive functional unit using the information entropy weight and sensitivity weight coupling method within the reference health period, and to generate the health baseline and graded threshold range of each sensitive functional unit. The health index dynamic calculation module is used to preprocess, normalize, and weight and sum the real-time or near-real-time monitoring data of each sensitive functional unit to obtain the current health index of each sensitive functional unit. and the changes in health index at different time scales ; The health status diagnosis and transmission analysis module establishes a health risk transmission model based on the source-transmission-sink topology to identify key pathways for the downstream spread of anomalies from the source region, and calculates health risk transmission based on health indices. and changes in health index Dynamic diagnosis of the health status of each sensitive functional unit is performed to achieve hierarchical diagnosis of the health status of the lake as a whole and key sensitive functional units and identification of transmission paths; The results visualization and early warning module is used to display the health level of sensitive functional units, risk transmission paths and early warning information in the form of layer overlay, time series curves and early warning lists, and send alarm information to the management terminal.

9. The system according to claim 8, characterized in that, The multi-source monitoring data acquisition module includes a fixed shore monitoring station subunit, a buoy monitoring subunit, an unmanned vessel patrol subunit, and a remote sensing data access subunit. The health status diagnosis and transmission analysis module is equipped with a sliding diagnosis window setting unit, which is used to dynamically analyze the health index of sensitive functional units at different time scales such as daily, weekly, and monthly, so as to support the hierarchical, regional, and time-segmented dynamic monitoring and diagnosis of lake health.

Images