Lake ecosystem health evaluation method and system

Abstract

The invention discloses a lake ecosystem health evaluation method and system, and particularly relates to the technical field of data management. A monitoring area is divided through a zoning model of a lake area, distribution density coverage rate characteristics and sampling frequency abnormal fluctuation characteristics are extracted based on distribution density and acquisition frequency data of sampling points, data spatial and temporal distribution uniformity indexes of the sampling points are calculated, and the acquired data are classified according to the uniformity indexes. Establishing an evaluation model and generating a health state report by preferentially utilizing the uniformity data, generating a data quality analysis report aiming at the non-uniformity data and feeding back a sampling strategy optimization suggestion, and performing error accumulation prediction and dynamic sampling adjustment on the incomplete uniformity data; by improving comprehensiveness, uniformity and timeliness of data acquisition, scientificity and reliability of health evaluation of the lake ecosystem are improved, and powerful support is provided for ecological management and restoration.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a lake ecosystem health assessment method and system. Background Art

[0002] Lake ecosystem health assessment refers to a comprehensive analysis and evaluation of the structure, function and sustainability of lake ecosystems, with the aim of determining their current ecological status and overall health. This assessment is based on scientific indicators such as water quality, biodiversity, nutrient levels, ecological balance, etc., to reveal whether the lake can maintain its natural ecological functions and provide stable ecological services. A healthy lake ecosystem usually has clean water, rich biodiversity, a stable food chain and good self-regulation ability.

[0003] Through lake ecosystem health assessment, potential problems affecting the lake environment, such as pollution, eutrophication, biological invasion or human interference, can be discovered. This process not only helps scientific decision-making and resource management, but also provides a basis for ecological restoration and environmental protection, thereby promoting the sustainable use of lakes and the overall stability of regional ecosystems.

[0004] The prior art has the following deficiencies:

[0005] The health status of lake ecosystems changes significantly dynamically, but the frequency and scope of data collection in existing technologies are often insufficient. For example, the layout of sampling points may be biased towards nearshore or accessible areas, ignoring deep water bodies in lakes or hard-to-reach areas, and the monitoring time interval may be too long to capture rapid changes or seasonal effects. In addition, the uneven temporal and spatial distribution of data may lead to biased evaluation results. For example, the water quality in nearshore areas may deteriorate significantly due to the direct impact of pollution sources, while the center of the lake or deep water bodies may remain relatively healthy. If the sampling points are concentrated in the nearshore area, the health risk of the entire lake ecosystem may be overestimated; conversely, ignoring the serious pollution in the nearshore area may underestimate the extent of ecological degradation. Summary of the invention

[0006] The purpose of the present invention is to provide a lake ecosystem health assessment method and system to address the deficiencies in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for evaluating the health of a lake ecosystem, comprising the following steps:

[0008] S1: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted;

[0009] S2: preprocessing the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracting the distribution density coverage characteristics in the preprocessed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data;

[0010] S3: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics, determine the weight assignment of the data spatiotemporal distribution uniformity degree of the sampling point in each time period, and perform weighted average calculation on the data spatiotemporal distribution uniformity degree weight assignment in each time period to obtain the data spatiotemporal distribution uniformity index of each sampling point;

[0011] S4: Compare the obtained spatiotemporal distribution uniformity index of the data of each sampling point with the gradient standard threshold, and divide the lake ecological data collected in each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results;

[0012] S5: For data with uniform spatiotemporal distribution, it is used to build evaluation models and generate health status reports. For data with non-uniform spatiotemporal distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions.

[0013] S6: For the data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in the subsequent time period is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0014] Preferably, in S2, the distribution density coverage rate characteristics in the extracted distribution density data are analyzed to generate a distribution density coverage rate deviation index, and the method for obtaining the distribution density coverage rate deviation index is:

[0015] Get the actual distribution density coverage d in each monitoring area i,实际 , that is, the ratio of the actual number of sampling points to the theoretical number of sampling points in each monitoring area, calculated as: The theoretical coverage rate d is preset according to the ideal sampling density of the monitoring design i,理论 , and determine the importance weight w of each monitoring area i For each monitoring area i, calculate the difference Δd between the actual distribution density coverage and the theoretical distribution density coverage i , the expression is: Δd i =d i,实际 -d i,理论 ; Weight the deviation values ​​and give higher weights to important areas: w i ·(Δd i ) 2 ; Among them, the weight wi The value range of is [0,1]. The weighted deviations of all monitoring areas are summarized and the weighted Euclidean distance D is calculated. The expression is: Where n is the total number of monitoring areas, and the distribution density coverage deviation index is calculated as follows: Where HSK is the distribution density coverage deviation index.

[0016] Preferably, in S2, the extracted sampling frequency abnormal fluctuation characteristics are analyzed to generate a sampling frequency abnormal fluctuation index, and the sampling frequency abnormal fluctuation index is obtained by:

[0017] According to the sampling frequency, the time series data ft is set at each time point. t = 1, 2, ..., N represents the number of sampling points in the time series. The time interval Δt is the sampling interval of the time series. The time series is decomposed by wavelet to decompose the time series ft into components of multiple scales: including: trend component A j , that is, the low-frequency component, reflects the overall trend of the sampling frequency, and the fluctuation component D j , that is, the high-frequency component, indicates rapid fluctuations or abnormal changes in the sampling frequency, using the wavelet decomposition formula: Where j is the number of decomposed scales, and the energy of each component is calculated. The energy is the sum of the squares of the component signals: E(D j ) is high frequency energy, E(A j ) is the low-frequency energy, and the total energy includes the energy of all high-frequency components and low-frequency components E total , the expression is: Calculate the sampling frequency abnormal fluctuation index, the expression is: Where GMH is the sampling frequency abnormal fluctuation index.

[0018] Preferably, in S3, according to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics; determine the weighted assignment of the data spatiotemporal distribution uniformity of the sampling points in each time period, specifically: convert the distribution density coverage deviation index and the sampling frequency abnormal fluctuation index into a comprehensive feature vector, and use the comprehensive feature vector as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the weighted assignment labels of the data spatiotemporal distribution uniformity of the sampling points in each time period as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the data spatiotemporal distribution uniformity of all sampling points in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted assignment of the data spatiotemporal distribution uniformity of the sampling points in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model, and the data spatiotemporal distribution uniformity index of each sampling point is obtained by weighted average calculation of the weighted assignment of the data spatiotemporal distribution uniformity in each time period.

[0019] Preferably, in S4, the obtained data spatiotemporal distribution uniformity index of each sampling point is compared with the gradient standard threshold, specifically: the obtained data spatiotemporal distribution uniformity index of each sampling point is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the data spatiotemporal distribution uniformity index of each sampling point is compared with the first standard threshold and the second standard threshold respectively;

[0020] If the data spatiotemporal distribution uniformity index of the sampling point is greater than the second standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is high. At this time, a high spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected in the sampling point is divided into spatiotemporal distribution uniformity data;

[0021] If the data spatiotemporal distribution uniformity index of the sampling point is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the spatiotemporal distribution uniformity of the data at the sampling point is said to be moderate. At this time, a medium spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into incomplete spatiotemporal distribution uniformity data;

[0022] If the data spatiotemporal distribution uniformity index of the sampling point is less than the first standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is low. At this time, a low spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into spatiotemporal distribution uneven data.

[0023] Preferably, in S6, for the spatiotemporal distribution incomplete uniformity data, the error accumulation of the spatiotemporal distribution incomplete uniformity data in the subsequent time period is predicted, and according to the prediction result, the collection plan of each sampling point is dynamically adjusted, specifically as follows:

[0024] For data with incomplete spatiotemporal distribution uniformity, that is, the data spatiotemporal distribution uniformity index of the generated sampling points is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the data spatiotemporal distribution uniformity index generated in the subsequent time period that is greater than or equal to the first standard threshold and less than or equal to the second standard threshold is collected, and a corresponding data spatiotemporal distribution uniformity index set is established, and the mean and standard deviation of the data spatiotemporal distribution uniformity index set are calculated. After analyzing them, the error accumulation of the data with incomplete spatiotemporal distribution uniformity in the subsequent time period is predicted according to the analysis results, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0025] Preferably, if the mean of the data spatiotemporal distribution uniformity index in the data set is greater than or equal to the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, no warning signal is generated at this time, and the existing sampling scheme is maintained without large-scale adjustment;

[0026] If the mean value of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index mean value, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index standard deviation, a first-level warning signal is generated, the frequency of unstable sampling points is increased, and the sampling point layout plan is readjusted;

[0027] If the mean value of the data spatiotemporal distribution uniformity index is less than the reference threshold value of the data spatiotemporal distribution uniformity index mean value, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index standard deviation, a secondary warning signal is generated, the number of sampling points in the low uniformity area is increased, the performance of the sensor or sampling equipment is checked, and the potential sampling interval problem is repaired;

[0028] If the mean of the data spatiotemporal distribution uniformity index is less than the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, a third-level warning signal is generated and the overall sampling point density is increased; the monitoring plan is redesigned to identify persistent weak areas and give priority to increasing the sampling frequency.

[0029] The present invention also provides a lake ecosystem health assessment system, including a data acquisition module, a data processing module, a data uniformity assessment module, a data classification module, a data utilization and feedback module, and a dynamic sampling optimization module:

[0030] Data collection module: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted;

[0031] Data processing module: pre-processes the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracts the distribution density coverage characteristics in the pre-processed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data;

[0032] Data uniformity assessment module: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics; determine the weight assignment of the data spatiotemporal distribution uniformity of the sampling point in each time period, and perform weighted average calculation on the data spatiotemporal distribution uniformity weight assignment in each time period to obtain the data spatiotemporal distribution uniformity index of each sampling point;

[0033] Data classification module: compare the spatiotemporal distribution uniformity index of the data obtained at each sampling point with the gradient standard threshold, and divide the lake ecological data collected at each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results;

[0034] Data utilization and feedback module: For data with uniform temporal and spatial distribution, it is used to build evaluation models and generate health status reports; for data with non-uniform temporal and spatial distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions;

[0035] Dynamic sampling optimization module: For data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in subsequent time periods is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0036] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0037] 1. The present invention solves the problems of sampling point distribution deviation, insufficient monitoring coverage and difficulty in capturing dynamic changes in time and space in the prior art by combining a partitioning model, a spatiotemporal distribution uniformity index and a dynamic sampling optimization mechanism. By extracting the distribution density coverage and sampling frequency fluctuation characteristics, using a machine learning model to predict the uniformity of the sampling points, classifying and processing the collected data and generating a health status report, the scientificity and accuracy of the lake ecological health assessment are improved. At the same time, in view of the error accumulation of data with incomplete spatiotemporal distribution uniformity, the sampling plan is dynamically adjusted to optimize the sampling point layout and collection frequency, greatly improving the data quality and comprehensiveness of coverage.

[0038] 2. The present invention also introduces an intelligent monitoring solution based on early warning signals, which dynamically identifies weak areas through mean and standard deviation analysis, and adjusts sampling strategies according to different early warning levels, such as increasing sampling density, encrypting sampling frequency, or optimizing equipment performance. It can not only significantly reduce the risk of deviation in lake health assessment, but also efficiently identify ecological risk points, providing strong data support and decision-making basis for the management and restoration of lake ecosystems, and ensuring the sustainable use of lake resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0040] Figure 1 The figure is a flow chart of the method of the present invention.

[0041] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] Example 1, please refer to Figure 1 and Figure 2 As shown, a lake ecosystem health assessment method described in this embodiment includes the following steps:

[0044] S1: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted;

[0045] S2: preprocessing the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracting the distribution density coverage characteristics in the preprocessed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data;

[0046] S3: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics, determine the weight assignment of the data spatiotemporal distribution uniformity degree of the sampling point in each time period, and perform weighted average calculation on the data spatiotemporal distribution uniformity degree weight assignment in each time period to obtain the data spatiotemporal distribution uniformity index of each sampling point;

[0047] S4: Compare the obtained spatiotemporal distribution uniformity index of the data of each sampling point with the gradient standard threshold, and divide the lake ecological data collected in each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results;

[0048] S5: For data with uniform spatiotemporal distribution, it is used to build evaluation models and generate health status reports. For data with non-uniform spatiotemporal distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions.

[0049] S6: For the data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in the subsequent time period is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0050] In S1, the lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are statistically analyzed, specifically:

[0051] Establishing a zoning model includes: dividing the lake according to different ecological functional areas, such as nearshore areas, central lake areas, deep water areas, sediment areas, etc. Consider the special environmental characteristics of the lake, such as the distribution of pollution sources in the basin, river estuaries, high-incidence areas of algal blooms and other key locations. According to the hydrodynamic characteristics, areas with strong fluidity (such as river confluences) and still water areas (such as central lake areas) are monitored separately. Consider physical characteristics such as water temperature stratification (thermocline), nutrient concentration gradients, and depth. Use GIS (geographic information system) technology to grid the lake area. Common methods include regular grid division (such as one grid per square kilometer) or irregular division based on natural boundaries.

[0052] Use high-resolution satellite images to partition the surface features of the lake. Use underwater robots or sonar to obtain the distribution characteristics of the deep water and sediment of the lake to provide support for vertical partitioning. Combine existing watershed land use and hydrological data for auxiliary division.

[0053] Sampling points are arranged in each monitoring area, and the specific number and density are determined according to regional characteristics (such as hydrodynamic complexity or pollution sensitivity). For example: nearshore area: one sampling point is arranged every 200 meters; center of the lake: one sampling point is arranged every 500 meters; deep water body: one vertical sampling point is arranged every 10 meters according to the depth. Calculate the distribution density of sampling points in each monitoring area: distribution density = number of sampling points / area or volume of monitoring area; area is used for surface and nearshore areas, and volume is used for deep water bodies and sediment areas. Output the total number of sampling points and distribution density (points / unit area or points / unit volume) in each area for analysis of monitoring coverage. The time interval for sensors to record data is usually expressed in "times / hour" or "times / day". Record the collection frequency of each sensor during the monitoring period and classify it by region. Mark the sensor locations that are below the set threshold (such as less than 10 sampling times per day) and analyze the reasons for fluctuations in sampling frequency (equipment failure, regional dynamic changes, etc.).

[0054] S2: Preprocess the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extract the distribution density coverage characteristics in the preprocessed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data.

[0055] The distribution density data and sensor acquisition frequency data obtained at the sampling points in each time period are preprocessed to remove noise, repair missing data, and ensure the accuracy of subsequent analysis. Duplicate sampling points are removed to ensure the uniqueness of data in the same area.

[0056] Detect outliers in the distribution density (such as abnormally high or low distribution density in a certain area). You can use the Z-score method to detect outliers: Check whether the sensor has a long data collection interval or a sensor failure (such as no data for a long time). Mark sensors whose collection frequency is lower than a certain threshold (such as less than 5 times a day).

[0057] Use spatial interpolation techniques (such as Kriging interpolation) to complete the missing distribution density data of sampling points to ensure spatial coverage in the area. Use time series prediction algorithms (such as linear interpolation or LSTM models) to complete the time period when sensor data collection is interrupted. In order to unify the magnitude of the eigenvalues, the distribution density data and sampling frequency data are normalized so that the eigenvalue range is limited to [0,1], which is convenient for subsequent analysis and comparison.

[0058] Extracting distribution density coverage features from distribution density data includes: Distribution density coverage features reflect whether the sampling points fully cover the monitoring area in space. The specific extraction process is as follows:

[0059] Analyze the distribution density data of each monitoring area and calculate the actual distribution density coverage. The coverage formula is: distribution density coverage = actual sampling points / theoretical sampling points; distribution density coverage = theoretical sampling points / actual sampling points; the actual number of sampling points is the number of sampling points actually deployed in each time period. The theoretical number of sampling points is the expected number of sampling points calculated based on the area and the designed sampling density (such as 1 sampling point per square kilometer).

[0060] Combined with regional division, a coverage distribution map is generated to mark areas with insufficient coverage (such as coverage below 80%). The mean and standard deviation of coverage are used to indicate the trend of coverage changes over the entire time period, helping to analyze the dynamic stability of coverage.

[0061] Extracting abnormal fluctuation characteristics of sampling frequency from sensor acquisition frequency data includes: abnormal fluctuation characteristics of sampling frequency reflect the change in the time interval of sensor data collection. The specific extraction process is: calculate the standard deviation of the sampling time interval and compare it with the set normal fluctuation threshold. If the standard deviation exceeds the set threshold (such as twice the historical standard deviation), the sensor is marked as abnormal during the time period. Count the frequency of abnormal fluctuations for each sensor during the monitoring period (such as the number of abnormal fluctuations per week). Extract the changing trend of the sampling frequency (increase or decrease) to determine whether the sensor performance is stable. Evaluate the impact of the abnormality on the overall data quality by calculating the amount of missing data during the abnormal fluctuation time period (such as the proportion of abnormal time to the total monitoring time).

[0062] The distribution density coverage rate characteristics in the extracted distribution density data are analyzed to generate the distribution density coverage rate deviation index. The method for obtaining the distribution density coverage rate deviation index is as follows:

[0063] Get the actual distribution density coverage d in each monitoring area i,实际 , that is, the ratio of the actual number of sampling points to the theoretical number of sampling points in each monitoring area, calculated as: The theoretical coverage rate d is preset according to the ideal sampling density of the monitoring design (such as 1 sampling point per square kilometer). i,理论 , and determine the importance weight of each monitoring area to reflect the relative priority of the area in the overall lake monitoring. For example: pollution-sensitive areas (such as nearshore areas, lake estuaries, etc.): give higher weights; ecologically stable areas (such as the center of the lake): give lower weights.

[0064] For each monitoring area i, calculate the difference Δd between the actual distribution density coverage and the theoretical distribution density coverage i , the expression is: Δd i =d i,实际 -d i,理论; Weight the deviation values ​​and give higher weights to important areas: w i ·(Δd i ) 2 ; Among them, the weight w i The value range of is [0,1],

[0065] Summarize the weighted deviations of all monitoring areas and calculate the weighted Euclidean distance D, which is expressed as: Where n is the total number of monitoring areas, and the distribution density coverage deviation index is calculated as follows: Where HSK is the distribution density coverage deviation index.

[0066] The larger the distribution density coverage deviation index is, the more significant the difference between the actual distribution density coverage and the theoretical distribution density coverage is, and the more uneven the distribution of sampling points in the monitoring area is. This usually indicates that the sampling points are too concentrated in certain areas, while other areas may be under-covered, resulting in the monitoring data being unable to fully reflect the ecological status of the entire lake. In this case, it is easy to cause deviations in the monitoring results, especially in lakes with large dynamic changes, and important environmental issues or key ecological processes may be missed.

[0067] On the contrary, the smaller the distribution density coverage deviation index is, the closer the actual distribution density coverage is to the theoretical distribution density coverage, and the spatial distribution of each sampling point is more uniform. This means that the coverage of the monitoring area is more comprehensive, and the collected data can better represent the overall ecological status of the lake, which helps to improve the scientificity and reliability of lake health assessment. At the same time, uniform distribution can reduce the risks caused by data deviation and provide a more accurate basis for subsequent ecological restoration and management.

[0068] After analyzing the extracted sampling frequency abnormal fluctuation characteristics, the sampling frequency abnormal fluctuation index is generated. The method for obtaining the sampling frequency abnormal fluctuation index is as follows:

[0069] Set the time series data ft according to the sampling frequency at each time point, t = 1, 2, ..., N represents the number of sampling points of the time series, and the time interval Δt is the sampling interval of the time series (such as hours, days). Select a wavelet function suitable for data analysis (such as Haar, Daubechies, Meyer). Haar: Suitable for capturing instantaneous changes. Daubechies: Suitable for smooth and continuous signals. Perform wavelet decomposition on the time series and decompose the time series ft into components of multiple scales (levels): including: trend component A j , that is, the low-frequency component, reflects the overall trend of the sampling frequency, and the fluctuation component D j , that is, the high-frequency component, indicates rapid fluctuations or abnormal changes in the sampling frequency, using the wavelet decomposition formula: Where j is the number of decomposed scales, and the energy of each component is calculated. The energy is the sum of the squares of the component signals: E(D j is high frequency energy, E(A j ) is the low-frequency energy, and the total energy includes the energy of all high-frequency components and low-frequency components E total , the expression is: Calculate the sampling frequency abnormal fluctuation index, the expression is: Where GMH is the sampling frequency abnormal fluctuation index.

[0070] The larger the sampling frequency abnormal fluctuation index, the more significant the fluctuation of the sampling frequency and the more uneven the distribution of the sampling data in the time dimension. In this case, there may be some time periods where the sampling is too intensive, while other time periods are insufficient, resulting in key dynamic changes (such as sudden algal blooms or pollution events) may not be captured or evaluated in a timely manner. A high fluctuation index also indicates that the continuity of the data is poor, which will affect the accuracy of lake health assessment and increase the difficulty of understanding the dynamic change process.

[0071] On the contrary, the smaller the sampling frequency abnormal fluctuation index is, the smaller the fluctuation of the sampling frequency is, and the more evenly distributed the sampling data is in the time dimension. This means that the data can evenly cover the monitoring period and capture the seasonal changes and short-term dynamic fluctuations of the lake ecosystem. The data distribution with a low fluctuation index helps to generate more accurate and reliable lake health assessment results and provide comprehensive support for decision makers to ensure timely intervention in ecological problems. The more significant, the higher the degree of abnormal fluctuation.

[0072] S3: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics, determine the weight assignment of the data spatiotemporal distribution uniformity degree of the sampling point in each time period, and perform weighted average calculation on the data spatiotemporal distribution uniformity degree weight assignment in each time period to obtain the data spatiotemporal distribution uniformity index of each sampling point.

[0073] The distribution density coverage deviation index and the sampling frequency abnormal fluctuation index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the weighted assignment labels of the data spatiotemporal distribution uniformity of the sampling points in each time period as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the data spatiotemporal distribution uniformity of all sampling points in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted assignment of the data spatiotemporal distribution uniformity of the sampling points in each time period is determined according to the model output results. Among them, the machine learning model is a polynomial regression model, and the data spatiotemporal distribution uniformity index of each sampling point is obtained by weighted average calculation of the weighted assignment of the data spatiotemporal distribution uniformity in each time period.

[0074] The method for obtaining the weighted value of the uniformity of the spatiotemporal distribution of data at the sampling points in each time period is as follows: from the comprehensive feature vector training data of the trained machine learning model, the corresponding function expression is obtained: LCR=F(HSK, GMH); where F is the output function of the model, HSK is the distribution density coverage deviation index, GMH is the sampling frequency abnormal fluctuation index, and LCR is the weighted value of the uniformity of the spatiotemporal distribution of data at the sampling points in each time period.

[0075] S4: Compare the spatiotemporal distribution uniformity index of the data obtained at each sampling point with the gradient standard threshold, and divide the lake ecological data collected at each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution non-uniformity data according to the comparison results.

[0076] The obtained data spatiotemporal distribution uniformity index of each sampling point is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the data spatiotemporal distribution uniformity index of each sampling point is compared with the first standard threshold and the second standard threshold respectively;

[0077] If the data spatiotemporal distribution uniformity index of the sampling point is greater than the second standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is high. At this time, a high spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected in the sampling point is divided into spatiotemporal distribution uniformity data;

[0078] If the data spatiotemporal distribution uniformity index of the sampling point is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the spatiotemporal distribution uniformity of the data at the sampling point is said to be moderate. At this time, a medium spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into incomplete spatiotemporal distribution uniformity data;

[0079] If the data spatiotemporal distribution uniformity index of the sampling point is less than the first standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is low. At this time, a low spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into spatiotemporal distribution uneven data.

[0080] S5: For data with uniform spatiotemporal distribution, it is used to build evaluation models and generate health status reports first; for data with non-uniform spatiotemporal distribution, a data quality analysis report is generated and suggestions for improving the sampling strategy are provided.

[0081] Spatiotemporal distribution uniformity data are used first in the construction of evaluation models. Spatiotemporal distribution uniformity data have comprehensive coverage, and the distribution of sampling points is highly consistent in space and time, which can provide highly representative data input. When constructing a lake health evaluation model with these data, deviations caused by uneven data distribution can be avoided to ensure that the health status indicators output by the model are more reliable. The core functions of the optimization evaluation model include water quality change trend prediction, eutrophication analysis, and biodiversity dynamics assessment.

[0082] The spatiotemporal distribution uniformity data can cover all areas and time periods of the lake, and the generated report can truly reflect the overall ecological status of the lake. For example, by analyzing the uniformity data, the water quality indicators (dissolved oxygen, pH, nutrient concentration, etc.), pollution source distribution and biological habitat of the lake can be accurately evaluated. Based on the health status report, treatment plans (such as pollution control, ecological restoration, etc.) are proposed to provide a reliable basis for managers. For example, the report can indicate specific areas that need priority treatment and provide adjustment suggestions for long-term monitoring plans.

[0083] The management and feedback of spatiotemporal data include the generation of data quality analysis reports, which include: Data coverage analysis: Identify areas and time periods that are not fully sampled, such as deep water bodies in lakes or insufficient sampling in winter. Fluctuation impact assessment: By analyzing abnormal fluctuations in sampling frequency, evaluate the possibility of missing key dynamic changes (such as pollution peaks and algal blooms) caused by fluctuations. Low-quality data marking: Classify and store spatiotemporal data and mark them to avoid direct impact on the scientific nature of the model. The report can indicate that the density of sampling points in a high-risk pollution area (such as the lake estuary) is insufficient, which may lead to an underestimation of the ecological risk in the area. Through fluctuation analysis, it is identified that the sampling interval of sensors is too long in certain time periods (such as the day and night alternation period), which may miss important ecological data. Regional optimization: Add sampling points in low coverage areas (such as the center of the lake, deep water areas that are difficult to reach) or use drones, underwater robots and other technologies for supplementary monitoring. Time optimization: Encrypt the sampling frequency in time periods with significant dynamic changes (such as the algal bloom period) to capture short-term changes. Replace sensors with large frequency fluctuations, or improve their working time planning to reduce sampling discontinuity. Add backup sensors or real-time monitoring equipment to make up for monitoring gaps caused by equipment failure. Adjust monitoring strategies based on the analysis results of inhomogeneity data. For example, transfer resources from over-covered areas to under-covered areas to improve overall monitoring efficiency.

[0084] S6: For the data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in the subsequent time period is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0085] For data with incomplete spatiotemporal distribution uniformity, that is, the data spatiotemporal distribution uniformity index of the generated sampling points is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the data spatiotemporal distribution uniformity index generated in the subsequent time period that is greater than or equal to the first standard threshold and less than or equal to the second standard threshold is collected, and a corresponding data spatiotemporal distribution uniformity index set is established, and the mean and standard deviation of the data spatiotemporal distribution uniformity index set are calculated. After analyzing them, the error accumulation of the data with incomplete spatiotemporal distribution uniformity in the subsequent time period is predicted according to the analysis results, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0086] If the mean of the data spatiotemporal distribution uniformity index in the data set is greater than or equal to the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, the data mean is high, indicating that the overall distribution uniformity level of the sampling points is good; the data fluctuation is low, indicating that the distribution of the sampling points is stable and no significant abnormalities occur. At this time, no warning signal is generated, the existing sampling plan is maintained, and no large-scale adjustment is required. The sampling frequency is only appropriately optimized in key areas.

[0087] If the mean of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, the data mean is high, indicating that the overall distribution uniformity level of the sampling points is good; the data fluctuation is high, indicating that the uniformity of some sampling points is unstable and there are significant deviations. At this time, a first-level warning signal is generated, the frequency of unstable sampling points is increased, and the spatiotemporal coverage of these areas is ensured; check whether the abnormal fluctuation is caused by equipment problems or environmental characteristics, and readjust the sampling point layout plan.

[0088] If the mean of the data spatiotemporal distribution uniformity index is less than the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, the data mean is low, indicating that the overall distribution uniformity level of the sampling points is poor; the data fluctuation is high, indicating that the distribution of the sampling points is not only uneven, but also has significant local deviations. At this time, a secondary warning signal is generated to increase the number of sampling points in low uniformity areas, especially areas with large deviations from the average level; encrypt sampling during high fluctuation time periods to reduce the impact of dynamic changes on data coverage; check the performance of sensors or sampling equipment, and repair potential sampling interval problems.

[0089] If the mean of the data spatiotemporal distribution uniformity index is less than the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, the data mean is low, indicating that the overall distribution uniformity of the sampling points is insufficient; the data fluctuation is low, indicating that although the distribution is uneven, the deviation is universal, rather than a local anomaly. At this time, a third-level warning signal is generated to increase the overall sampling point density, especially in areas with insufficient coverage; redesign the long-term monitoring plan to ensure that the sampling coverage gradually reaches the uniformity target; identify persistent weak areas based on historical uniformity data and give priority to increasing the sampling frequency.

[0090] It should be noted here that the importance of the first-level warning signal is greater than that of the second-level warning signal, and the importance of the second-level warning signal is greater than that of the third-level warning signal. Relevant personnel can take corresponding handling measures according to different warning signal levels.

[0091] In summary, according to the analysis results, the sampling scheme is dynamically adjusted to reduce the subsequent error accumulation. Specifically, it includes: increasing sampling points: deploying more sampling points in areas or time periods with poor uniformity to improve coverage. Adjusting the sampling frequency: increasing the sampling frequency during periods of high dynamic changes to avoid missing key changes. Optimizing sampling equipment: replacing or calibrating sensors in high-fluctuation areas to improve data collection stability.

[0092] In this embodiment, the lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density of sampling points and the sensor collection frequency data are statistically collected in different time periods; the collected data is preprocessed to extract the distribution density coverage characteristics and the abnormal fluctuation characteristics of the sampling frequency, further determine the weight assignment of the spatiotemporal distribution uniformity of the sampling points, and calculate the data spatiotemporal distribution uniformity index; according to the comparison result of the uniformity index and the gradient standard threshold, the data is divided into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data; spatiotemporal distribution uniformity data is used preferentially to construct an evaluation model and generate a health status report, and a data quality analysis report is generated for the spatiotemporal distribution inhomogeneity data and the sampling strategy is optimized; for the spatiotemporal distribution incomplete uniformity data, the sampling point collection plan is dynamically adjusted by predicting the error accumulation to gradually improve the quality and coverage uniformity of the lake ecological data.

[0093] Embodiment 2, a lake ecosystem health assessment system described in this embodiment includes a data acquisition module, a data processing module, a data uniformity assessment module, a data classification module, a data utilization and feedback module and a dynamic sampling optimization module:

[0094] Data collection module: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted;

[0095] Data processing module: pre-processes the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracts the distribution density coverage characteristics in the pre-processed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data;

[0096] Data uniformity assessment module: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics, the data spatiotemporal distribution uniformity degree weight assignment of the sampling point in each time period is determined, and the data spatiotemporal distribution uniformity index of each sampling point is obtained after weighted average calculation of the data spatiotemporal distribution uniformity degree weight assignment in each time period;

[0097] Data classification module: compare the spatiotemporal distribution uniformity index of the data obtained at each sampling point with the gradient standard threshold, and divide the lake ecological data collected at each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results;

[0098] Data utilization and feedback module: For data with uniform temporal and spatial distribution, it is used to build evaluation models and generate health status reports; for data with non-uniform temporal and spatial distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions;

[0099] Dynamic sampling optimization module: For data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in subsequent time periods is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

[0100] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0101] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0102] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0103] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A lake ecosystem health assessment method, characterized in that: The following steps are involved: S1: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted; S2: preprocessing the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracting the distribution density coverage characteristics in the preprocessed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data; S3: According to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics, determine the weight assignment of the data spatiotemporal distribution uniformity degree of the sampling point in each time period, and perform weighted average calculation on the data spatiotemporal distribution uniformity degree weight assignment in each time period to obtain the data spatiotemporal distribution uniformity index of each sampling point; S4: Compare the obtained spatiotemporal distribution uniformity index of the data of each sampling point with the gradient standard threshold, and divide the lake ecological data collected in each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results; S5: For data with uniform spatiotemporal distribution, it is used to build evaluation models and generate health status reports. For data with non-uniform spatiotemporal distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions. S6: For the data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in the subsequent time period is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

2. A lake ecosystem health assessment method according to claim 1, characterized in that: In S2, the distribution density coverage rate characteristics in the extracted distribution density data are analyzed to generate a distribution density coverage rate deviation index. The distribution density coverage rate deviation index is obtained as follows: Get the actual distribution density coverage d in each monitoring area i,实际 , that is, the ratio of the actual number of sampling points to the theoretical number of sampling points in each monitoring area, calculated as: The theoretical coverage rate d is preset according to the ideal sampling density of the monitoring design i,理论 , and determine the importance weight w of each monitoring area i For each monitoring area i, calculate the difference Δd between the actual distribution density coverage and the theoretical distribution density coverage i , the expression is: Δd i =d i,实际 -d i,理论 ; Weight the deviation values ​​and give higher weights to important areas: w i ·(Δd i ) 2 ; Among them, the weight w i The value range of is [0,1]. The weighted deviations of all monitoring areas are summarized and the weighted Euclidean distance D is calculated. The expression is: Where n is the total number of monitoring areas, and the distribution density coverage deviation index is calculated as follows: Where HSK is the distribution density coverage deviation index.

3. A lake ecosystem health assessment method according to claim 2, characterized in that: In S2, the extracted sampling frequency abnormal fluctuation characteristics are analyzed to generate a sampling frequency abnormal fluctuation index. The sampling frequency abnormal fluctuation index is obtained by: According to the sampling frequency, the time series data ft is set at each time point. t = 1, 2, ..., N represents the number of sampling points in the time series. The time interval Δt is the sampling interval of the time series. The time series is decomposed by wavelet to decompose the time series ft into components of multiple scales: including: trend component A j , that is, the low-frequency component, reflects the overall trend of the sampling frequency, and the fluctuation component D j , that is, the high-frequency component, indicates rapid fluctuations or abnormal changes in the sampling frequency, using the wavelet decomposition formula: Where j is the number of decomposed scales, and the energy of each component is calculated. The energy is the sum of the squares of the component signals: E(D j ) is high frequency energy, E(A j ) is the low-frequency energy, and the total energy includes the energy of all high-frequency components and low-frequency components E total , the expression is: Calculate the sampling frequency abnormal fluctuation index, the expression is: Where GMH is the sampling frequency abnormal fluctuation index.

4. A lake ecosystem health assessment method according to claim 3, characterized in that: In S3, according to the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics; Determine the weighted value of the uniformity of the spatiotemporal distribution of the data of the sampling points in each time period, specifically: convert the distribution density coverage deviation index and the sampling frequency abnormal fluctuation index into a comprehensive feature vector, and use the comprehensive feature vector as the input of the machine learning model. The machine learning model predicts the weighted value of the uniformity of the spatiotemporal distribution of the data of the sampling points in each time period with each group of comprehensive feature vectors as the prediction target, and minimizes the sum of the prediction errors of the weighted value of the spatiotemporal distribution uniformity of all sampling points in each time period as the training target. The machine learning model is trained until the sum of the prediction errors converges and the model training is stopped. The weighted value of the uniformity of the spatiotemporal distribution of the data of the sampling points in each time period is determined according to the output results of the model, wherein the machine learning model is a polynomial regression model, and the weighted average calculation of the weighted value of the spatiotemporal distribution uniformity of the data in each time period is performed to obtain the spatiotemporal distribution uniformity index of the data of each sampling point.

5. A lake ecosystem health assessment method according to claim 4, characterized in that: In S4, the obtained data spatiotemporal distribution uniformity index of each sampling point is compared with the gradient standard threshold, specifically: the obtained data spatiotemporal distribution uniformity index of each sampling point is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the data spatiotemporal distribution uniformity index of each sampling point is compared with the first standard threshold and the second standard threshold respectively; If the data spatiotemporal distribution uniformity index of the sampling point is greater than the second standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is high. At this time, a high spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected in the sampling point is divided into spatiotemporal distribution uniformity data; If the data spatiotemporal distribution uniformity index of the sampling point is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the spatiotemporal distribution uniformity of the data at the sampling point is said to be moderate. At this time, a medium spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into incomplete spatiotemporal distribution uniformity data; If the data spatiotemporal distribution uniformity index of the sampling point is less than the first standard threshold, it is said that the data spatiotemporal distribution uniformity of the sampling point is low. At this time, a low spatiotemporal distribution uniformity signal is generated, and the lake ecological data collected at the sampling point is divided into spatiotemporal distribution uneven data.

6. A lake ecosystem health assessment method according to claim 1, characterized in that: In S6, for the incompletely uniform temporal and spatial distribution data, the error accumulation of the incompletely uniform temporal and spatial distribution data in the subsequent time period is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results, specifically: For data with incomplete spatiotemporal distribution uniformity, that is, the data spatiotemporal distribution uniformity index of the generated sampling points is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, the data spatiotemporal distribution uniformity index generated in the subsequent time period that is greater than or equal to the first standard threshold and less than or equal to the second standard threshold is collected, and a corresponding data spatiotemporal distribution uniformity index set is established, and the mean and standard deviation of the data spatiotemporal distribution uniformity index set are calculated. After analyzing them, the error accumulation of the data with incomplete spatiotemporal distribution uniformity in the subsequent time period is predicted according to the analysis results, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

7. A lake ecosystem health assessment method according to claim 6, characterized in that: If the mean of the data spatiotemporal distribution uniformity index in the data set is greater than or equal to the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, no warning signal will be generated at this time, and the existing sampling plan will be maintained without large-scale adjustment; If the mean value of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index mean value, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index standard deviation, a first-level warning signal is generated, the frequency of unstable sampling points is increased, and the sampling point layout plan is readjusted; If the mean value of the data spatiotemporal distribution uniformity index is less than the reference threshold value of the data spatiotemporal distribution uniformity index mean value, and the standard deviation of the data spatiotemporal distribution uniformity index is greater than or equal to the reference threshold value of the data spatiotemporal distribution uniformity index standard deviation, a secondary warning signal is generated, the number of sampling points in the low uniformity area is increased, the performance of the sensor or sampling equipment is checked, and the potential sampling interval problem is repaired; If the mean of the data spatiotemporal distribution uniformity index is less than the reference threshold of the mean of the data spatiotemporal distribution uniformity index, and the standard deviation of the data spatiotemporal distribution uniformity index is less than the reference threshold of the standard deviation of the data spatiotemporal distribution uniformity index, a third-level warning signal is generated and the overall sampling point density is increased; the monitoring plan is redesigned to identify persistent weak areas and give priority to increasing the sampling frequency.

8. A lake ecosystem health assessment system, used to implement a lake ecosystem health assessment method according to any one of claims 1 to 7, characterized in that: It includes data acquisition module, data processing module, data uniformity assessment module, data classification module, data utilization and feedback module and dynamic sampling optimization module: Data collection module: The lake is divided into multiple monitoring areas through the lake area zoning model, and the distribution density data of sampling points in each monitoring area in several time periods and the sensor collection frequency data are counted; Data processing module: pre-processes the distribution density data and sensor acquisition frequency data obtained at the sampling point in each time period, and extracts the distribution density coverage characteristics in the pre-processed distribution density data and the abnormal fluctuation characteristics of the sampling frequency in the sensor acquisition frequency data; Data uniformity assessment module: based on the extracted distribution density coverage characteristics and sampling frequency abnormal fluctuation characteristics; Determine the weighted value of the uniformity of the data spatiotemporal distribution of the sampling points in each time period, and perform weighted average calculation on the weighted value of the uniformity of the data spatiotemporal distribution in each time period to obtain the uniformity index of the data spatiotemporal distribution of each sampling point; Data classification module: compare the spatiotemporal distribution uniformity index of the data obtained at each sampling point with the gradient standard threshold, and divide the lake ecological data collected at each sampling point into spatiotemporal distribution uniformity data, spatiotemporal distribution incomplete uniformity data and spatiotemporal distribution inhomogeneity data according to the comparison results; Data utilization and feedback module: For data with uniform temporal and spatial distribution, it is used to build evaluation models and generate health status reports; for data with non-uniform temporal and spatial distribution, it generates data quality analysis reports and provides feedback on sampling strategy improvement suggestions; Dynamic sampling optimization module: For data with incomplete spatial and temporal distribution uniformity, the error accumulation of the data with incomplete spatial and temporal distribution uniformity in subsequent time periods is predicted, and the collection plan of each sampling point is dynamically adjusted according to the prediction results.

Images