Lake process data adaptive monitoring method and lake sensor

Abstract

The application discloses a kind of lake process data self-adapting monitoring method and lake inductor, belong to lake ecological monitoring technical field.Lake process inductor is integrated in-situ monitoring terminal, pre-disposes individual lake process inductor, and continuous generation lake in-situ time series process data;Obtain satellite remote sensing image data to obtain lake global ecological distribution data;Lake global ecological distribution data is carried out error calibration and space-time registration to obtain calibrated lake global ecological distribution data;Fusion same water quality parameter's point analysis result and global spatial distribution information, constructs lake water environment evolution process model and deduces lake ecological development trend.The application is supplemented by intelligent lake in-situ monitoring terminal from source to make up the short board of existing technology lack of lake continuous process data, and provides real and reliable in-situ basic data source for whole process evolution analysis;Meanwhile, based on the more fine double view linkage deduction of the application, effectively improve lake disaster early warning timeliness.

Description

Technical Field

[0001] This invention belongs to the technical field of lake ecological monitoring, and in particular relates to an adaptive monitoring method for lake process data and a lake sensor. Background Technology

[0002] Currently, lake ecological monitoring in the industry mainly relies on three mainstream methods: manual fixed-point sampling and laboratory testing, online in-situ monitoring of single indicators, and large-scale satellite remote sensing surveys. The existing monitoring system suffers from practical technical defects such as poor data continuity, fragmented software and hardware systems, and a lack of connection between air, ground, and water monitoring links. These shortcomings make it difficult to meet the needs of full-process evolution analysis and risk trend prediction of lake ecosystems. The specific technical shortcomings are as follows: First, existing monitoring methods primarily rely on instantaneous, discrete, static indicator collection, failing to acquire continuous ecological process data for lakes. Traditional manual on-site sampling and conventional online monitoring equipment can only perform point-based indicator detection at fixed times and cross-sections, lacking the capability for long-term, high-frequency, in-situ continuous data acquisition; they cannot capture dynamic process characteristics such as small fluctuations in water level, dynamic migration of lake currents, diurnal variations in dissolved oxygen, and in-situ transformation of nitrogen, phosphorus, and nutrients. Such monitoring data contains numerous time-series discontinuities, reflecting only the static state of the lake's water environment at the moment of sampling, and cannot characterize the dynamic evolution of water quality indicators as hydrological parameters change.

[0003] Secondly, the integration of field monitoring terminals is low, and there is no standardized three-dimensional monitoring network covering land, sea, and water. Existing lake field monitoring equipment has limited functionality, with most devices only capable of collecting single-category parameters such as water quality or hydrology. It is impossible to achieve integrated and simultaneous collection of multiple parameters, including water quality physicochemical indicators, topographic elevation, and on-site meteorological data. Various monitoring sensors are scattered, and the layout of the points lacks overall planning. A three-dimensional sensing network that coordinates ground-based in-situ monitoring, water-based terminal sensing, and satellite remote sensing for comprehensive observation has not been established. The monitoring blind spots in lake waters are large, and multi-source monitoring data cannot be interconnected and uniformly scheduled.

[0004] Finally, the existing monitoring and analysis models lack sufficient intelligence, resulting in a significant lag in ecological risk early warning. Current lake water environment assessment standards rely solely on static monitoring data for status quo assessment, lacking adaptive analysis and multi-scale modeling techniques for time-series lake data. They are unable to construct spatiotemporally linked water environment evolution models, hindering the prediction of overall lake ecological development trends. Traditional management models can only address water pollution and ecological anomalies after they occur, failing to enable proactive risk identification and pollution source tracing, and thus failing to support the practical operational needs of precise pollution control and long-term, refined ecological management of lakes. Summary of the Invention

[0005] To address the problems mentioned above in the background art, this invention provides an adaptive monitoring method for lake process data and a lake sensor.

[0006] This invention employs the following technical solution: an adaptive monitoring method for lake process data, comprising the following steps: Pre-deployment within the target lake A series of lake process sensors form an integrated in-situ sensing and monitoring network that combines air, land, and water. Utilizing process sensors in each lake Simultaneous collection of water quality parameters of water bodies Real-time water body data and real-time hydrological data Generate corresponding in-situ time series process data of lakes. ; The numbering of the lake process sensor; According to a predetermined frequency, satellite remote sensing image data of the target lake is acquired. The ecological distribution data of the entire lake area is obtained through a water quality inversion model. Error calibration and spatiotemporal registration are performed on the ecological distribution data of the entire lake area to obtain calibrated ecological distribution data of the entire lake area. Based on in-situ time series data of lakes Conduct water body characteristic analysis to generate corresponding point analysis results; based on calibrated lake-wide ecological distribution data, obtain the overall spatial distribution information for the same time period; By integrating the point analysis results corresponding to the same water quality parameter with the spatial distribution information of the entire region, a model of the evolution process of lake water environment is constructed, and the ecological development trend of lake is deduced.

[0007] In a further embodiment, the water quality parameters At a minimum, this includes: pH value, dissolved oxygen (DO) concentration, turbidity, chlorophyll concentration, total nitrogen concentration, and total phosphorus concentration; correspondingly, real-time water body data. For the first A sensor at any time Collected water quality parameters Real-time monitoring values; The real-time hydrological data It should include at least: latitude and longitude of monitoring points, lake surface elevation, real-time water level, lake flow, and meteorological data.

[0008] In a further embodiment, the in-situ time-series process data of the lake The generation process is as follows: Based on real-time water body data With real-time hydrological data Define the eco-hydrological linkage coupling coefficient To quantitatively characterize the impact of hydrological changes on water quality parameters The real-time driving effect of aquatic ecological indicators; Introducing water quality parameters Rate of water body ecological change Combined with the corresponding eco-hydrological linkage coupling coefficient The water quality parameters are dynamically adjusted using the following formula. sample time interval : In the formula, For adaptation constants; According to the adjusted sampling time interval Sampling was performed to collect real-time water body data. and real-time hydrological data Add a unified hardware timestamp ; Real-time water body data collected each time With real-time hydrological data All are subject to joint verification. For any sampling time; Integrate verified real-time water body data Real-time hydrological data Eco-hydrological linkage coupling coefficient Generate in-situ time series data of lakes .

[0009] In a further embodiment, the process of obtaining the calibrated lake-wide ecological distribution data for a single water quality parameter is as follows: Based on the in-situ time-series data of the lake Construct a point error correction model: ;in, For the first Each monitoring point at time Water quality parameters below The corresponding numerical deviation, For the first Each monitoring point at time The corresponding remote sensing inversion value; Based on deviation samples from multiple monitoring points Water quality parameters are corrected using a pixel-by-pixel correction method. Initial ecological distribution data of the entire lake area: , These are the corresponding corrected remote sensing inversion values; Based on the eco-hydrological linkage coupling coefficient of each monitoring point To constrain and control the eco-hydrological linkage changes in the initial ecological distribution data of the entire lake area after correction; Dynamic sampling interval of in-situ time series data of lakes Using the time base, the corrected remote sensing inversion values Perform time-dimension interpolation and registration; Spatial matching was performed between the geographic coordinates of each monitoring point and the pixel coordinates of the remote sensing image to eliminate spatiotemporal misalignment. After error correction and spatiotemporal joint registration, the calibrated ecological distribution data of the entire lake area was obtained. .

[0010] In a further embodiment, the point analysis results include at least: lake process sensors. Water quality parameters obtained Corresponding rate of water body ecological change ; The spatial distribution information of the entire region corresponds to the water quality parameters. Corrected remote sensing inversion values ​​of the entire target lake .

[0011] In a further embodiment, the deduction method for the lake's ecological development trend is as follows: For a single water quality parameter, a threshold for the rate of change of water body ecology is pre-set. and global concentration benchmark ; the rate of change in aquatic ecosystems and water body ecological change rate threshold Threshold comparisons were performed to determine the rate of change in the lake's ecological state at the monitoring points. Get water quality parameters Corrected remote sensing inversion values ​​of the entire target lake Compare it with the global concentration benchmark value Compare and judge the overall water quality indicators of the lake.

[0012] In a further embodiment, the lake water environment evolution process model is constructed as follows: calibrated lake-wide ecological distribution data In-situ time series data of lakes Perform spatiotemporal registration; Constructing a dual-constraint fusion operator Based on the eco-hydrological linkage coupling coefficient To accommodate the dual constraints, feature-level deep fusion is performed on the two registered data sets, and the fusion mapping relationship is as follows: ; Based on fusion results A dual-view linkage inference architecture is constructed, and the missing data in the time series view is supplemented through a dual-view mutual verification mechanism.

[0013] In a further embodiment, if and If the lake is found to be at risk of large-scale ecological degradation, a corresponding risk alert will be triggered based on the type of water quality parameters. like and If the water quality index rises rapidly only at a local monitoring point, the pre-emptive ecological control and treatment process will be initiated. like The lake's ecological state was determined to be stable, and the routine continuous monitoring mode was maintained.

[0014] A smart lake process sensor is used to realize the adaptive monitoring method of lake process data as described above. The lake process sensor is an integrated in-situ monitoring terminal that is deployed at preset monitoring points in the lake. Includes: a sensing and monitoring unit, a positioning elevation unit, a data transmission unit, and a power supply unit installed on the lake process sensor; Among them, the sensing and monitoring unit is set to collect real-time water data at designated monitoring points in the lake; The elevation unit is used to acquire real-time hydrological data at a specified monitoring point. A data transmission unit is connected to the sensing and monitoring unit and the positioning elevation unit; the data transmission unit is configured to receive real-time water body data and real-time hydrological data, and transmit them to the cloud monitoring platform. The power supply unit is electrically connected to the sensing and monitoring unit, the positioning and elevation unit, and the data transmission unit, respectively.

[0015] In a further embodiment, the smart lake process sensor is adapted to both deep and shallow lakes; the smart lake process sensor can achieve in-situ multi-level stratified monitoring.

[0016] Compared with the prior art, the present invention has the following outstanding advantages: This invention enables in-situ acquisition of continuous dynamic changes in lake eco-hydrology, filling gaps in process monitoring data from the source. Unlike traditional monitoring methods that use fixed-frequency discrete sampling and only acquire static cross-sectional data, this invention quantifies the driving relationship between hydrological factors and water quality indicators through an eco-hydrological linkage coupling coefficient, adaptively and dynamically adjusting the equipment sampling interval. This invention is used to capture subtle, short-term ecological changes such as water level fluctuations, diurnal variations in dissolved oxygen, nitrogen and phosphorus nutrient migration and transformation, and sudden water quality changes induced by lake current disturbances. It comprehensively preserves the entire trajectory of ecological parameters from steady state, gradual change, abrupt change to recovery, overcoming the core deficiency of traditional technologies that cannot capture the intermediate evolutionary processes of lakes and can only monitor the final state.

[0017] This invention's integrated smart lake sensor can simultaneously collect multi-dimensional parameters of water quality, hydrology, and meteorology, avoiding the problem of mismatched parameter changes caused by asynchronous collection times of multiple separate devices; the device supports layered in-situ monitoring, which can accurately obtain the gradient changes of physicochemical indicators in different vertical water layers of the lake; the field-resistant structure ensures that the trend data is not lost or distorted under complex wind and wave conditions, and realizes the long-term stable collection of multi-factor collaborative evolution process data at the lake location scale.

[0018] To address the limitation of satellite remote sensing inversion data, which can only characterize instantaneous spatial distribution and cannot accurately reflect real-world ecological change trends, this invention relies on in-situ location change process data to constrain pixel-by-pixel correction of remote sensing images. A unified eco-hydrological coupling coefficient across the entire region is used to lock in the interconnected hydrological and water quality change patterns of the entire lake, and spatiotemporal registration is completed by combining this with in-situ dynamic sampling time-series benchmarks. This allows the remote sensing data to align with the lake's true natural evolution rhythm, resulting in a standardized dataset that reflects the long-term continuous spatial change process across the entire region, eliminating the problems of process trend distortion and chaotic spatiotemporal evolution logic in traditional remote sensing data.

[0019] This invention abandons the simple data splicing method and achieves feature-level deep fusion of the temporal change process of the original location and the spatial evolution process of the entire remote sensing domain through a dual-constraint fusion operator. It not only retains the fine-grained short-term fluctuation details of single-point monitoring, but also makes up for the shortcomings of single-point monitoring in terms of insufficient spatial coverage and inability to grasp the regional diffusion change process. It generates a rare temporal and spatial dual-dimensional dataset of lake water environment evolution process, which fully covers the gradual change process of local points and the diffusion evolution process of the entire region. Attached Figure Description

[0020] Figure 1 This is a flowchart of the adaptive monitoring method for lake process data in Example 1.

[0021] Figure 2 This is a diagram of the integrated in-situ sensing and monitoring network for air, land, and water in Example 1.

[0022] Figure 3 This is a structural diagram of the lake process sensor in Example 2.

[0023] Figure 3 The components are labeled as follows: Lake process sensor 1, in-situ mounting rod 101, floating component 102, and limiting component 103. Detailed Implementation

[0024] Example 1 like Figure 1 As shown, this invention discloses an adaptive monitoring method for lake process data based on remote sensing, comprising the following steps: Pre-deployment within the target lake Each lake process sensor forms an integrated in-situ sensing and monitoring network encompassing air, land, and water, referencing... Figure 2 .

[0025] Utilizing process sensors in each lake Simultaneous collection of water quality parameters of water bodies Real-time water body data and real-time hydrological data Generate corresponding in-situ time series process data of lakes. ; The numbering of the lake process sensor; According to a predetermined frequency (generally, satellite remote sensing has a long revisit cycle and low data update frequency, and can only output discrete-time global spatial snapshots), satellite remote sensing image data of the target lake is acquired. The lake's global ecological distribution data is obtained through a water quality inversion model, and the lake's global ecological distribution data is then calibrated and spatiotemporally registered to obtain calibrated lake global ecological distribution data. The water quality inversion model described in this embodiment can be an existing band ratio model, multiple linear regression (MLR) model, convolutional neural network (CNN), etc.

[0026] Based on in-situ time-series data of lakes Water body characteristic analysis is conducted to generate corresponding point analysis results; based on calibrated lake-wide ecological distribution data, spatial distribution information of the entire area during the same time period is obtained; therefore, the in-situ time-series process data of the lake described in this embodiment... Relying on the sensor's adaptive dynamic sampling mechanism, the acquisition cycle is much shorter than the revisit cycle of satellite remote sensing images. It can output point temporal change process data at high frequency and continuously, making up for the shortcomings of remote sensing data in terms of sparse sampling in the time dimension and inability to capture the short-term dynamic evolution process of water bodies.

[0027] By integrating the point analysis results corresponding to the same water quality parameter with the spatial distribution information of the entire region, a model of the evolution process of lake water environment is constructed, and the ecological development trend of lake is deduced.

[0028] The water quality parameters described in this embodiment At a minimum, this includes: pH value, dissolved oxygen (DO) concentration, turbidity, chlorophyll concentration, total nitrogen concentration, and total phosphorus concentration; correspondingly, real-time water body data. For the first A sensor at any time Collected water quality parameters Real-time monitoring values; The real-time hydrological data It should include at least: latitude and longitude of monitoring points, lake surface elevation, real-time water level, lake flow, and meteorological data.

[0029] Conventional lake monitoring typically uses fixed-time sampling (e.g., 1 hour / sampling, 4 hours / sampling), which has two drawbacks: First, it misses key changes: Lake algal blooms, sudden water level changes, and sudden increases in nutrients after heavy rainfall are all interconnected eco-hydrological changes that are "non-uniform and non-linear." Second, if the fixed frequency is too slow, it will miss peak changes. Invalid data redundancy: When a lake is in a stable period (e.g., weak algal activity at night and no water level fluctuations), frequent sampling will only generate a large amount of duplicate data, wasting sensor power and increasing the burden on storage and subsequent modeling. Therefore, the process of generating in-situ time-series lake process data described in this embodiment is as follows: Based on real-time water body data With real-time hydrological data Define the eco-hydrological linkage coupling coefficient To quantitatively characterize the impact of hydrological changes on water quality parameters The real-time driving role of aquatic ecological indicators. For example, for nutrient parameters such as chlorophyll, total phosphorus, and total nitrogen, water convection and sediment release caused by hydrological disturbances will lead to rapid fluctuations in these indicators, with a wider range of coupling coefficient values, highlighting the strong driving effect of hydrological elements. For basic physicochemical parameters such as pH, dissolved oxygen, and turbidity, the response of these indicators to changes in hydrological conditions is gradual, with a relatively narrower range of coupling coefficient values, accurately distinguishing the differences in the strength of the influence of hydrological processes on various water quality parameters.

[0030] In this embodiment, the eco-hydrological linkage coupling coefficient The calculation formula is: .

[0031] Introducing water quality parameters Rate of water body ecological change Combined with the corresponding eco-hydrological linkage coupling coefficient The water quality parameters are dynamically adjusted using the following formula. sample time interval : In the formula, As an adaptation constant, it is generally set to a value of .

[0032] Furthermore, the rate of change in aquatic ecosystems The calculation formula is: In other words, for a given water quality parameter, when (The intensity of hydrological impacts on ecology) increases, or When the rate of ecological change increases, It automatically shortens, uses high-frequency data acquisition, and efficiently captures the complete time sequence of key linked changes such as algal blooms and sudden changes in water level.

[0033] when and Synchronous reduction indicates that the lake's hydrological disturbances are weak, water quality indicators change slowly, and the water body is in a stable state. The system automatically lengthens the sampling interval and adopts a sparse sampling mode. Through the above technology, redundant and invalid data in the unchanging stage are greatly reduced, and the power consumption of the equipment caused by continuous sampling of the field monitoring terminal is reduced, which alleviates the pressure of massive time-series data storage and processing on the cloud platform.

[0034] According to the adjusted sampling time interval Sampling was performed to collect real-time water body data. and real-time hydrological data Add a unified hardware timestamp ; Real-time water body data collected each time With real-time hydrological data All are subject to joint verification. For any sampling time; Based on the linkage verification, integrate the real-time water body data that has passed the verification. Real-time hydrological data Eco-hydrological linkage coupling coefficient Generate in-situ time series data of lakes .

[0035] During the sampling process, real-time water body data and real-time hydrological data were synchronized in time and phase. The generated in-situ time-series process data of the lake completely preserved the hydrological-ecological linkage relationship, providing reliable data support for subsequent multi-scale time-series correlation analysis and ecological process modeling.

[0036] In this embodiment, the linkage verification specifically involves: eliminating those that meet the requirements. or Abnormal data points, among which , For monitoring points Historical average water quality parameters ), average historical hydrological data, , All are standard deviations; Simultaneously, regarding the eco-hydrological linkage coupling coefficient Make further judgment: If If it is, it is judged as an abnormal linkage and removed. For monitoring points The historical average of the correlation coefficient, This represents the corresponding standard deviation.

[0037] The physicochemical response mechanisms, hydrological driving sensitivity, and remote sensing spectral inversion characteristics of different water quality parameters all exhibit significant differences. If a unified, full-area calibration is applied, it will cause mutual interference between the correction patterns of various indicators and a mismatch in eco-hydrological linkage constraints, making it impossible to accurately reconstruct the independent spatiotemporal evolution process of a single parameter. Therefore, for a single water quality parameter, the process for obtaining the calibrated lake-wide ecological distribution data is as follows: Based on the in-situ time-series data of the lake Construct a point error correction model: ;in, For the first Each monitoring point at time Water quality parameters below The corresponding numerical deviation, For the first Each monitoring point at time The corresponding remote sensing inversion value; Based on deviation samples from multiple monitoring points Water quality parameters are corrected using a pixel-by-pixel correction method. Initial ecological distribution data of the entire lake area: , These are the corresponding corrected remote sensing inversion values; Based on the eco-hydrological linkage coupling coefficient of each monitoring point To constrain and control the eco-hydrological linkage changes in the initial ecological distribution data of the entire lake area after correction; Dynamic sampling interval of in-situ time series data of lakes Using the time base, the corrected remote sensing inversion values Perform time-dimension interpolation and registration; Spatial matching was performed between the geographic coordinates of each monitoring point and the pixel coordinates of the remote sensing image to eliminate spatiotemporal misalignment. After error correction and spatiotemporal joint registration, the calibrated ecological distribution data of the entire lake area was obtained. .

[0038] In summary, the point analysis results described in this embodiment include at least: lake process sensors. Water quality parameters obtained Corresponding rate of water body ecological change ; The spatial distribution information of the entire region corresponds to the water quality parameters. Corrected remote sensing inversion values ​​of the entire target lake .

[0039] The above methods can analyze the water quality and hydrological time series processes at each monitoring point for a single time period. However, relying solely on discrete point data is insufficient to grasp the spatial evolution patterns of water quality diffusion and nutrient migration across the entire lake area, making it difficult to fully predict the overall ecological evolution trend of the lake. To achieve long-term dynamic assessment of the lake's overall water environment and early identification of potential ecological risks, this embodiment also includes the projection of lake ecological development trends, specifically: For a single water quality parameter, a threshold for the rate of change of water body ecology is pre-set. and global concentration benchmark ; the rate of change in aquatic ecosystems and water body ecological change rate threshold Threshold comparisons were performed to determine the rate of change in the lake's ecological state at the monitoring points. Get water quality parameters Corrected remote sensing inversion values ​​of the entire target lake Compare it with the global concentration benchmark value Compare and judge the overall water quality indicators of the lake.

[0040] Furthermore, if and If this is detected, the lake is deemed to pose a risk of large-scale ecological degradation, and a corresponding risk alert is triggered based on the type of water quality parameters; for example, water quality parameters... When chlorophyll a represents the algal proliferation rate across the entire lake area, and the overall concentration exceeds the standard, a special risk warning for a full-area cyanobacterial bloom will be issued; when water quality parameters... When total phosphorus and total nitrogen are both high, it indicates that eutrophication of the entire lake area is continuing to worsen, and a warning of long-term deterioration of nutrient pollution in the watershed is issued.

[0041] like and If the water quality index at a monitoring point rises rapidly in a localized area, then a preliminary ecological control and treatment process should be initiated. For example, if the rate of change of total phosphorus at the river estuary point increases suddenly, but the concentration obtained from remote sensing inversion of the entire lake area does not exceed the benchmark value, only local preliminary treatments such as interception and control of sewage and pre-dredging of bottom sediment should be carried out in the upstream catchment area of ​​that point, without the need to initiate emergency dispatch measures for the entire area.

[0042] like The system determines that the lake's water ecological state is stable and maintains a regular, continuous monitoring mode. For example, during the low-temperature period in autumn and winter, when the rate of change of various water quality parameters in the lake is slow and the concentration across the entire area is stable and meets the standards, the equipment automatically switches to an adaptive sparse sampling mode and only outputs routine monitoring and analysis reports according to a preset cycle.

[0043] Finally, relying solely on point change rates and overall concentration thresholds can only achieve static classification of a single indicator, failing to fully integrate in-situ temporal changes with overall spatial evolution characteristics, quantify the hydrological-water quality linkage evolution, and intuitively and completely reproduce the entire dynamic evolution of the lake's long-term water environment. To achieve deep coupling of multi-source process data, construct a standardized model capable of fully extrapolating the spatiotemporal evolution of lake ecology, and better quantify the continuous evolution of lake water quality driven by hydrological factors, the lake water environment evolution model described in this embodiment is constructed as follows: calibrated lake-wide ecological distribution data In-situ time series data of lakes Perform spatiotemporal registration; Constructing a dual-constraint fusion operator Based on the eco-hydrological linkage coupling coefficient To accommodate the dual constraints, feature-level deep fusion is performed on the two registered data sets, and the fusion mapping relationship is as follows: ; Based on fusion results A dual-view linkage inference architecture is constructed, and the missing data in the time series view is supplemented through a dual-view mutual verification mechanism.

[0044] In a further embodiment, the dual-view linkage inference architecture is specifically divided into two types of functional views: a temporal process view and a spatial global view. The two types of views are independent of each other but support each other, and together constitute a spatiotemporal integrated representation system of lake ecology and hydrology.

[0045] Among them, the time-series process view is in a fixed spatial position Based on the baseline, integrate in-situ time-series process data of lakes. The calibrated lake-wide ecological distribution data corresponding to this location uses the eco-hydrological coupling coefficient. To constrain this process, a continuous evolution curve of the eco-hydrological parameters at this location over time is generated. This fully depicts the daily and time-period continuous changes of ecological and hydrological indicators at a single location, overcoming the shortcomings of traditional manual discrete sampling, which can only obtain intermittent moment data and cannot present intermediate evolution processes, thus achieving continuous reproduction of the entire temporal evolution of eco-hydrological parameters at a fixed location.

[0046] The spatial global view is taken at any fixed time. Using in-situ measured data from various monitoring points in the lake as the time slice benchmark, and employing these data as the true value calibration benchmark, the calibrated remote sensing data on the ecological distribution of the entire lake area is constrained, extended, and spatially interpolated. While ensuring consistency between local point data and in-situ measured results, the system fully extends the pixel distribution characteristics of the remote sensing data across the entire area, continuously depicting the gradient changes and pattern differences of eco-hydrological parameters with spatial location throughout the entire lake region. This approach overcomes the limitations of traditional single-point monitoring, which can only represent the conditions of local points and cannot reflect the spatial distribution differences of the entire lake, achieving a complete and refined representation of the ecological-hydrological spatial pattern of the entire lake area at a single moment.

[0047] The dual-view mutual verification mechanism is as follows: taking the ecological and hydrological linkage change law as a common constraint benchmark, and using the fixed point time evolution characteristics of the time series process view, the ecological and hydrological parameter distribution deviation of local pixels in the spatial global view is verified and corrected. By synchronously utilizing the continuous spatial distribution characteristics of the spatial global view, interpolation is used to fill in the missing time-series nodes within the discrete sampling interval of the time-series process view.

[0048] Example 2 This embodiment discloses a smart lake process sensor for implementing the adaptive monitoring method of lake process data as described in Embodiment 1. The smart lake process sensor in this embodiment adopts an integrated in-situ monitoring terminal structure, which is deployed at preset monitoring points in the lake. To adapt to different lake depths and achieve standardized in-situ data acquisition across the entire water area, this embodiment uses an installation method compatible with two installation and fixing structures, differentiating between shallow and deep lakes: For monitoring scenarios in shallow lakes, the following methods are adopted: Figure 3 The assembled structure of the pole and floating body shown has been fixed on site. For deep lakes and large reservoirs with large water depths, thick silt at the bottom, or hard rock bottoms where poles cannot be anchored, existing mature single-point / multi-point anchored fixed-point floating installation structures are used to complete the deployment. The monitoring terminals under both deployment schemes support in-situ monitoring of multiple water levels and can simultaneously collect water quality and hydrological time series data from different water depth profiles.

[0049] Furthermore, the smart lake process sensor in this embodiment mainly includes a sensing and monitoring unit, a positioning and elevation unit, a data transmission unit, and a power supply unit. All functional units are integrated and deployed as one unit, which is suitable for long-term in-situ monitoring scenarios of lakes in the field.

[0050] The sensing and monitoring unit is installed on the smart lake process sensor body and is used to collect real-time water physicochemical and ecological monitoring data at preset monitoring points in the lake. The sensing and monitoring unit in this embodiment includes: pH sensor, dissolved oxygen sensor, turbidity sensor, chlorophyll sensor, total nitrogen detection module and total phosphorus detection module.

[0051] The positioning elevation unit is integrated into the lake process sensor to synchronously acquire the geographical location information and real-time hydrological elevation data of the corresponding monitoring point; the positioning elevation unit includes: a GNSS positioning module, an elevation meter and a water level sensor.

[0052] The data transmission unit is communicatively connected to the sensing and monitoring unit and the positioning and elevation unit, respectively. The data transmission unit is used to receive real-time water body data collected by the sensing and monitoring unit and real-time hydrological data obtained by the positioning and elevation unit, and remotely transmit the collected monitoring data to the cloud monitoring platform to realize real-time data transmission and online storage. The power supply unit is electrically connected to the sensing and monitoring unit, the positioning and elevation unit, and the data transmission unit, respectively, to provide a continuous and stable power supply to each functional unit and ensure that the entire in-situ monitoring terminal works uninterruptedly around the clock.

[0053] For ease of understanding, Figure 3 Taking the shallow lake monitoring scenario shown as an example, the whole is mainly composed of an in-situ installation rod 101, a floating body component 102, and a limiting component.

[0054] The in-situ installation rod 1 is made of high-strength, corrosion-resistant fiberglass material. It is hard, resistant to acid and alkali corrosion in lake water, and resistant to underwater aging, making it suitable for long-term underwater deployment in lakes. The in-situ installation rod is a vertical long rod. Its bottom end is deeply buried and fixed in the hard bottom of the lake through a concrete anchor seat, making it firmly installed and not easy to loosen. The top of the rod extends vertically upward and protrudes above the normal water surface of the lake. The overall installation is straight and regular, which can provide stable sliding guidance for the floating components.

[0055] The floating component 102 is integrally molded from high-density waterproof foamed buoyancy material, making it lightweight, highly buoyant, and waterproof and airtight. It can be immersed in lake water for a long time without seeping or deforming. A through-hole is opened in the center of the floating component, and the whole is movably sleeved on the outside of the original mounting rod, which can slide freely along the axis of the rod and float up and down synchronously with the daily rise and fall of the lake water level. A sealed installation cavity is reserved inside the floating component, which integrates the sensing and monitoring unit, positioning elevation unit, data transmission unit and power supply unit into the cavity. The buoyancy of the floating component itself supports the entire set of monitoring functional components, so that the monitoring components are always stably positioned at the preset monitoring water layer position.

[0056] The limiting component 103 adopts a stainless steel wear-resistant retaining ring structure and is fixedly assembled to the upper part of the in-situ mounting rod by fastening. In the actual deployment process, the deployment height of the limiting component is strictly higher than the highest historical water level measured at the monitoring point by 0.5m to 1m, leaving sufficient safety margin. As a rigid blocking structure for the upward sliding of the floating component, the limiting component can strictly limit the maximum upward stroke of the floating component. Even in the case of a surge in water level during the flood season, it can effectively prevent the floating component from sliding upward off the top of the rod, completely avoiding the problem of loss or displacement of the floating component and its internal monitoring components from the in-situ mounting rod. This ensures that the entire in-situ monitoring equipment can be stably deployed and continuously complete data acquisition under various water level environments.

[0057] This embodiment integrates water body sensing and monitoring, positioning and elevation acquisition, remote data transmission and power supply functions into one integrated structure design, replacing the traditional decentralized monitoring equipment deployment mode. It has a compact structure and is easy to install on site. It can be stably deployed at the original location of the lake for a long time, effectively avoiding equipment displacement and data jump caused by wind, waves and water flow disturbance. It can continuously and stably acquire fixed-point in-situ time-series monitoring data of the lake, providing a reliable source data foundation for subsequent construction of lake eco-hydrological process data, and carrying out spatiotemporal fusion and pattern inference.

Claims

1. An adaptive monitoring method for lake process data, characterized in that, Includes the following steps: Pre-deployment within the target lake A series of lake process sensors form an integrated in-situ sensing and monitoring network that combines air, land, and water. Utilizing process sensors in each lake Simultaneous collection of water quality parameters of water bodies Real-time water body data and real-time hydrological data Generate corresponding in-situ time series process data of lakes. ; The numbering of the lake process sensor; According to a predetermined frequency, satellite remote sensing image data of the target lake is acquired. The ecological distribution data of the entire lake area is obtained through a water quality inversion model. Error calibration and spatiotemporal registration are performed on the ecological distribution data of the entire lake area to obtain calibrated ecological distribution data of the entire lake area. Based on in-situ time-series data of lakes Conduct water body characteristic analysis to generate corresponding point analysis results; based on calibrated lake-wide ecological distribution data, obtain the overall spatial distribution information for the same time period; By integrating the point analysis results corresponding to the same water quality parameter with the spatial distribution information of the entire region, a model of the evolution process of lake water environment is constructed, and the ecological development trend of lake is deduced.

2. The adaptive monitoring method for lake process data according to claim 1, characterized in that, The water quality parameters At a minimum, this includes: pH value, dissolved oxygen (DO) concentration, turbidity, chlorophyll concentration, total nitrogen concentration, and total phosphorus concentration; correspondingly, real-time water body data. For the first A sensor at any time Collected water quality parameters Real-time monitoring values; The real-time hydrological data It should include at least: latitude and longitude of monitoring points, lake surface elevation, real-time water level, lake flow and meteorological data.

3. The adaptive monitoring method for lake process data according to claim 1, characterized in that, The in-situ time-series data of the lake The generation process is as follows: Based on real-time water body data With real-time hydrological data Define the eco-hydrological linkage coupling coefficient To quantitatively characterize the impact of hydrological changes on water quality parameters The real-time driving effect of water body ecological indicators; Introducing water quality parameters Rate of water body ecological change Combined with the corresponding eco-hydrological linkage coupling coefficient The water quality parameters are dynamically adjusted using the following formula. sample time interval : In the formula, For adaptation constants; According to the adjusted sampling time interval Sampling was performed to collect real-time water body data. and real-time hydrological data Add a unified hardware timestamp ; Real-time water body data collected each time With real-time hydrological data All are subject to joint verification. For any sampling time; Integrate verified real-time water body data Real-time hydrological data Eco-hydrological linkage coupling coefficient Generate in-situ time series data of lakes .

4. The adaptive monitoring method for lake process data according to claim 1, characterized in that, The process for obtaining the calibrated ecological distribution data of the entire lake area for a single water quality parameter is as follows: Based on the in-situ time-series data of the lake Construct a point error correction model: ;in, For the first Each monitoring point at time Water quality parameters below The corresponding numerical deviation, For the first Each monitoring point at time The corresponding remote sensing inversion value; Based on deviation samples from multiple monitoring points Water quality parameters are corrected using a pixel-by-pixel correction method. Initial ecological distribution data of the entire lake area: , These are the corresponding corrected remote sensing inversion values; Based on the eco-hydrological linkage coupling coefficient of each monitoring point To constrain and control the eco-hydrological linkage changes in the initial ecological distribution data of the entire lake area after correction; Dynamic sampling interval of in-situ time series data of lakes Using the time base, the corrected remote sensing inversion values Perform time-dimension interpolation and registration; Spatial matching was performed between the geographic coordinates of each monitoring point and the pixel coordinates of the remote sensing image to eliminate spatiotemporal misalignment. After error correction and spatiotemporal joint registration, the calibrated ecological distribution data of the entire lake area was obtained. .

5. The adaptive monitoring method for lake process data according to claim 1, characterized in that, The site analysis results include at least: lake process sensors. Water quality parameters obtained Corresponding rate of water body ecological change ; The spatial distribution information of the entire region corresponds to the water quality parameters. Corrected remote sensing inversion values ​​of the entire target lake .

6. The adaptive monitoring method for lake process data based on remote sensing according to claim 1, characterized in that, The deduction method for the ecological development trend of the lake is as follows: For a single water quality parameter, a threshold for the rate of change of water body ecology is pre-set. and global concentration benchmark ; Rate of ecological change in water bodies and water body ecological change rate threshold Threshold comparisons were performed to determine the rate of change in the lake's ecological state at the monitoring points. Get water quality parameters Corrected remote sensing inversion values ​​of the entire target lake Compare it with the global concentration benchmark value Compare and judge the overall water quality indicators of the lake.

7. The adaptive monitoring method for lake process data according to claim 1, characterized in that, The model for the evolution of the lake's water environment is constructed as follows: calibrated lake-wide ecological distribution data In-situ time series data of lakes Perform spatiotemporal registration; Constructing a dual-constraint fusion operator Based on the eco-hydrological linkage coupling coefficient To accommodate the dual constraints, feature-level deep fusion is performed on the two registered data sets, and the fusion mapping relationship is as follows: ； Based on fusion results A dual-view linkage inference architecture is constructed, and the missing data in the time series view is supplemented through a dual-view mutual verification mechanism.

8. The adaptive monitoring method for lake process data according to claim 6, characterized in that, like and If the lake is found to be at risk of large-scale ecological degradation, a corresponding risk alert will be triggered based on the type of water quality parameters. like and If the water quality index rises rapidly only at a local monitoring point, the pre-emptive ecological control and treatment process will be initiated. like The lake's ecological state was determined to be stable, and the routine continuous monitoring mode was maintained.

9. A smart lake process sensor, used to implement the adaptive monitoring method for lake process data as described in any one of claims 1 to 8, characterized in that, The lake process sensor is an integrated in-situ monitoring terminal, which is deployed at the preset monitoring points in the lake. Includes: a sensing and monitoring unit, a positioning elevation unit, a data transmission unit, and a power supply unit installed on the lake process sensor; Among them, the sensing and monitoring unit is set to collect real-time water data at designated monitoring points in the lake; The elevation unit is used to acquire real-time hydrological data at a specified monitoring point. A data transmission unit is connected to the sensing and monitoring unit and the positioning elevation unit; the data transmission unit is configured to receive real-time water body data and real-time hydrological data, and transmit them to the cloud monitoring platform. The power supply unit is electrically connected to the sensing and monitoring unit, the positioning and elevation unit, and the data transmission unit.

10. A smart lake process sensor according to claim 9, characterized in that, The smart lake process sensor is compatible with both deep and shallow lakes. The intelligent lake process sensor can realize in-situ multi-level stratified monitoring of water levels.

Images