A comprehensive evaluation method for river ecological health integrating water chemistry and biological indices
By employing dynamic weight allocation and multi-dimensional coupling analysis, combined with water chemistry and biological indices, the problem of biased evaluation results in traditional river ecological health assessments has been solved, enabling precise quantification and intelligent assessment of river ecological health status.
Patent Information
- Application Number
- CN202511293227.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-11
- Publication Date
- 2025-11-14
- Estimated Expiration
- 2045-09-11
AI Technical Summary
Traditional methods for assessing the ecological health of rivers fail to reflect the dynamic changes in river ecosystems, neglect the influence of hydrological conditions and pollutant concentrations, leading to discrepancies between assessment results and actual conditions. Furthermore, fixed grading standards are difficult to adapt to the geographical backgrounds and ecological baseline differences of different watersheds.
A dynamic weight allocation mechanism is adopted, combined with multi-dimensional coupling analysis of water chemistry and biological indices. Dynamic scenario parameters of rivers are generated through dynamic scenario analysis. The weights are queried using the hydrological period-pollution load coupling matrix, the comprehensive health index of synergistic effect is calculated, and intelligent evaluation is carried out through machine learning model.
It enables precise quantitative assessment of river ecological health status, improves the scientific nature and scenario relevance of the assessment, enhances the sensitivity and accuracy of the assessment model, and avoids the subjectivity and limitations of traditional methods.
Smart Images

Figure CN120782300B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of water environment monitoring and ecological assessment, and in particular to a comprehensive evaluation method for river ecological health that integrates water chemistry and biological indices. Background Technology
[0002] River ecological health refers to a stable state in which a river ecosystem maintains structural integrity, normal function, and the ability to self-recover from external disturbances. Scientific measurement and evaluation of this health is a core technical aspect of water resource protection, water environment management, and water ecological restoration. Traditional evaluations typically rely on chemical composition analysis of water samples collected at specific time points or surveys of biological community structure; these results form the basis for determining the health status of water bodies.
[0003] In existing technologies, the evaluation method for river ecological health typically employs a comprehensive index method. This method first selects a series of representative hydrochemical and biological indicators, such as total phosphorus, total nitrogen, phytoplankton abundance, and benthic animal integrity index. Then, a fixed weight is assigned to each indicator using subjective weighting methods such as expert scoring or analytic hierarchy process (AHP). Finally, the standardized values of each indicator are multiplied by their weights and summed to obtain a comprehensive health index, which is then used to classify the health level according to a pre-set, fixed grading standard.
[0004] However, existing technical solutions have significant drawbacks. Due to their use of fixed weighting systems, these methods fail to reflect the fact that river ecosystems are dynamic and complex systems. In particular, they neglect the significant impact of hydrological conditions, such as high and low water seasons, on pollutant concentrations, migration and transformation patterns, and stress effects on biological communities, leading to potential discrepancies between assessment results and actual ecological conditions under different hydrological scenarios. Furthermore, fixed grading standards are ill-suited to the diverse geographical backgrounds and ecological baselines of different watersheds, limiting the universality and accuracy of the assessments. Summary of the Invention
[0005] To address the aforementioned issues, this invention provides a comprehensive evaluation method for river ecological health that integrates hydrochemical and biological indices. By employing a dynamic weight allocation mechanism and combining multi-dimensional coupling analysis of hydrochemical and biological indices, it is possible to achieve a precise quantitative evaluation of river ecological health status.
[0006] The above objectives can be achieved through the following approach:
[0007] A comprehensive evaluation method for river ecological health integrating hydrochemical and biological indices includes: acquiring hydrochemical index data, biological index data, and hydrological condition data of the river to generate raw monitoring data; performing dynamic scenario analysis on the raw monitoring data to generate river dynamic scenario parameters; indexing and querying the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights; calculating a synergistic effect comprehensive health index based on the raw monitoring data and combining the hydrochemical dynamic weights and biological dynamic weights through coupling analysis; and intelligently evaluating the synergistic effect comprehensive health index to determine the river ecological health level.
[0008] Optionally, the step of performing dynamic scenario analysis on the original monitoring data to generate river dynamic scenario parameters includes: preprocessing the original monitoring data to generate preprocessed hydrochemical data and preprocessed hydrological data; analyzing the preprocessed hydrological data to determine the current hydrological period; evaluating the preprocessed hydrochemical data to determine the pollution load level; and combining the current hydrological period and the pollution load level to generate river dynamic scenario parameters.
[0009] Optionally, the preprocessing of the raw monitoring data includes: obtaining the historical distribution characteristics of the raw monitoring data and generating a data confidence interval; identifying outliers in the raw monitoring data that exceed the data confidence interval and generating a data quality factor; and performing sliding window weighted interpolation correction on the outliers based on the data quality factor to generate preprocessed hydrochemical data and preprocessed hydrological data.
[0010] Optionally, the step of indexing and querying the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights includes: obtaining a hydrological period-pollution load coupling matrix; using the river dynamic scenario parameters as an index to query the hydrological period-pollution load coupling matrix to obtain weight allocation values; and generating hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values.
[0011] Optionally, the step of generating a synergistic comprehensive health index based on the original monitoring data, combined with the hydrochemical dynamic weights and biological dynamic weights, through coupled analysis includes: calculating the initial weights of each indicator using the entropy weight method based on the data dispersion of the original monitoring data; using the initial weights, calculating the closeness of the hydrochemical indicator data and biological index data to their respective ideal states to generate a synergy index; and calculating the synergistic comprehensive health index based on the synergy index and the hydrochemical dynamic weights and biological dynamic weights.
[0012] Optionally, the step of calculating the comprehensive health index of synergy based on the synergy index and the hydrochemical dynamic weight and biological dynamic weight includes: calculating the interaction strength coupling degree using the synergy index; calculating the comprehensive coordination index using the synergy index and the hydrochemical dynamic weight and biological dynamic weight; and calculating the comprehensive health index of synergy by combining the interaction strength coupling degree and the comprehensive coordination index through a coupling formula.
[0013] Optionally, the step of intelligently assessing the synergistic effect comprehensive health index to determine the river's ecological health level includes: acquiring a historical health case database; using the historical health case database to train a machine learning model to generate a dynamic health threshold; and comparing the synergistic effect comprehensive health index with the dynamic health threshold to determine the river's ecological health level.
[0014] Optionally, generating hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values includes: performing structural decomposition on the weight allocation values to generate hydrochemical weight components and biological weight components; and performing data extraction and normalization processing on the hydrochemical weight components and biological weight components to generate hydrochemical dynamic weights and biological dynamic weights.
[0015] Optionally, the step of acquiring river hydrochemical index data, biological index data, and hydrological condition data to generate raw monitoring data includes: performing time series alignment and format standardization on the hydrochemical index data, biological index data, and hydrological condition data to obtain a multi-source fusion dataset; and parsing the multi-source fusion dataset to generate raw monitoring data.
[0016] Based on the same inventive concept, this invention also provides a comprehensive river ecological health evaluation system integrating hydrochemical and biological indices. The system includes: a data acquisition module for acquiring river hydrochemical index data, biological index data, and hydrological condition data to generate raw monitoring data; a dynamic scenario analysis module for performing dynamic scenario analysis on the raw monitoring data to generate river dynamic scenario parameters; a weight allocation module for indexing and querying the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights; a coupling analysis calculation module for generating a synergistic effect comprehensive health index based on the raw monitoring data and combining the hydrochemical dynamic weights and biological dynamic weights through coupling analysis; and an intelligent assessment module for intelligently assessing the synergistic effect comprehensive health index to determine the river ecological health level.
[0017] Compared with the prior art, the present invention has the following advantages:
[0018] This invention introduces dynamic scenario analysis, combining instantaneous hydrological conditions of rivers with pollution load status to generate dynamic scenario parameters. This mechanism enables the evaluation system to identify and quantify the differentiated impacts of pollutants on the ecosystem under different hydrological backgrounds, thereby making the evaluation results more closely reflect the actual response process of the river ecosystem and significantly improving the scientific rigor and scenario relevance of the evaluation.
[0019] This invention innovatively employs a dynamic weight allocation mechanism. By indexing dynamic scenario parameters with a preset coupling matrix, it assigns real-time weights to water chemistry indicators and biological indices. This adaptive weight adjustment completely changes the static model of fixed weights in traditional assessments, ensuring that the assessment focus can be intelligently adjusted according to the main stress factors facing the river, thereby improving the sensitivity and accuracy of the assessment model.
[0020] This invention employs a coupling analysis method to calculate a comprehensive health index based on synergistic effects and introduces a machine learning model for intelligent assessment. This method not only evaluates the health level of each subsystem but also delves deeper into the interactions and coordination between the hydrochemical and biological systems. Furthermore, it avoids the subjectivity and limitations of traditional fixed-threshold grading methods, resulting in an evaluation that is not merely a score but also possesses system-level diagnostic capabilities and adaptive grading abilities.
[0021] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0023] Figure 1 This is a flowchart illustrating a comprehensive evaluation method for river ecological health that integrates water chemistry and biological indices, according to an embodiment of the present invention.
[0024] Figure 2 This is a coupling coordination analysis diagram of an embodiment of the present invention.
[0025] Figure 3 This is a sliding window weighted interpolation correction diagram according to an embodiment of the present invention.
[0026] Figure 4This is a dynamic health threshold diagram according to an embodiment of the present invention.
[0027] Figure 5 This is a dynamic scenario analysis diagram of an embodiment of the present invention.
[0028] Figure 6 This is a schematic diagram of the structure of a comprehensive evaluation system for river ecological health that integrates water chemistry and biological indices, according to an embodiment of the present invention. Detailed Implementation
[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0030] Reference Figure 1 One embodiment of the present invention proposes a comprehensive evaluation method for river ecological health that integrates hydrochemistry and biological indices. It adopts a dynamic weight allocation mechanism and combines multi-dimensional coupling analysis of hydrochemistry and biological indices to achieve accurate quantitative evaluation of river ecological health status.
[0031] The method described in this embodiment specifically includes:
[0032] Acquire water chemical index data, biological index data and hydrological condition data of the river to generate raw monitoring data;
[0033] Dynamic scenario analysis is performed on the raw monitoring data to generate dynamic scenario parameters for the river.
[0034] The river dynamic scenario parameters are indexed and queried to generate hydrochemical dynamic weights and biological dynamic weights;
[0035] Based on the original monitoring data, combined with the water chemical dynamic weight and biological dynamic weight, a synergistic effect comprehensive health index is generated through coupling analysis.
[0036] The comprehensive health index of the synergistic effect is intelligently assessed to determine the ecological health level of the river.
[0037] Specifically, this method first constructs a dynamic river scenario parameter by integrating hydrological and hydrochemical data. This parameter accurately describes the environmental background and stress state of the river at a specific moment. Next, it establishes a mapping mechanism from the dynamic scenario parameter to evaluation weights. Dynamic weights are assigned to the two major evaluation dimensions of hydrochemistry and biology through index lookup, allowing the evaluation model's focus to adaptively adjust according to the real-time river scenario. Finally, it applies coupling analysis theory to comprehensively calculate the dynamically weighted health status of each dimension, generating a comprehensive health index that reflects the synergistic effects of multiple systems. This index is then transformed into a final health level through an intelligent assessment model, thus realizing a complete and logically rigorous dynamic evaluation process. Coupling and coordination analysis, such as... Figure 2 As shown.
[0038] Optionally, the step of performing dynamic scenario analysis on the original monitoring data to generate river dynamic scenario parameters includes:
[0039] The raw monitoring data is preprocessed to generate preprocessed hydrochemical data and preprocessed hydrological data;
[0040] Analyze the preprocessed hydrological data to determine the current hydrological period;
[0041] Evaluate the pretreated water chemistry data to determine the pollution load level;
[0042] By combining the current hydrological period and the pollution load level, dynamic scenario parameters for the river are generated.
[0043] Specifically, the first step is to preprocess the acquired raw monitoring data. The raw monitoring data includes time-series water chemistry data, such as total phosphorus, total nitrogen, and chemical oxygen demand (COD), as well as hydrological condition data, such as flow rate and water level. The purpose of preprocessing is to improve data quality and provide a reliable data foundation for subsequent analysis. This preprocessing step generates preprocessed water chemistry data and preprocessed hydrological data, which are the direct inputs for subsequent analysis. Secondly, based on the preprocessed hydrological data, the current hydrological period is analyzed and determined. The current hydrological period refers to the hydrological state of the river at a specific point in time, which directly affects the dilution, diffusion, and migration capacity of pollutants. Technicians can divide different hydrological characteristic periods based on historical hydrological data, such as long-term average daily flow data. For example, the whole year can be divided into high-water season, normal-water season, low-water season, and dry season. The method for determining the current hydrological period is to compare the real-time monitored preprocessed hydrological data, such as the current flow rate, with pre-set flow rate thresholds for each hydrological period, thereby classifying the current moment into the corresponding hydrological period. Next, based on the pretreated water chemistry data, the pollution load level is assessed and determined. The pollution load level is a quantitative classification of the chemical pollution pressure currently borne by the river. This step requires a comprehensive evaluation of multiple water chemistry indicators. A feasible technical approach is to first calculate a comprehensive pollution index, which can be obtained by weighted summation of standardized values from multiple key water chemistry indicators. The standardization of each indicator can refer to national or local water environmental quality standards. For example, the pollution load index can be defined as follows: :
[0044] ,
[0045] in, The comprehensive pollution index; For the first The weight of each pollutant reflects its impact on the ecosystem; For the first The measured concentrations of these pollutants were derived from pretreated water chemistry data; For the first The corresponding water environmental quality standard limits for each pollutant. Technical personnel need to pre-set the weights of each indicator based on the evaluation objectives and regional water environment characteristics. and standards Calculate the comprehensive pollution index. Subsequently, based on the set grading thresholds, the pollution load is divided into different levels, such as low, medium, and high. Finally, the current hydrological period obtained from the analysis and the pollution load level determined by the assessment are combined to generate the final river dynamic scenario parameters. This combination is a structured label with clear meaning, rather than a simple addition of values. For example, if the current hydrological period is "dry season" and the pollution load level is "high," then the generated river dynamic scenario parameter is "dry season - high pollution load." This parameter constitutes a unique scenario identifier that accurately depicts the dual conditions of hydrological background and chemical stress faced by the river ecosystem at a specific moment, providing a direct and clear index for subsequent dynamic weight allocation.
[0046] Optionally, the preprocessing of the raw monitoring data includes:
[0047] Obtain the historical distribution characteristics of the original monitoring data and generate data confidence intervals;
[0048] Identify outliers in the raw monitoring data that exceed the data confidence interval and generate data quality factors;
[0049] Based on the data quality factor, the outliers are corrected by sliding window weighted interpolation to generate pre-processed hydrochemical data and pre-processed hydrological data.
[0050] Specifically, the historical distribution characteristics of the original monitoring data are first obtained, and confidence intervals are generated based on these characteristics. Technicians first need to collect long-term, continuous historical observation data for each monitoring indicator, such as total phosphorus concentration and flow rate, at the same monitoring section, forming a stable historical dataset. Based on this dataset, the statistical characteristics of the indicator are calculated, primarily the historical mean. and historical standard deviation Data confidence intervals, based on these statistical characteristics, define the reasonable range of fluctuation for a data point. A widely used technique is to use 3D... The criterion is to set the upper and lower limits of the data confidence interval as follows:
[0051] ,
[0052] in, This is the lower limit of the confidence interval. This represents the upper limit of the confidence interval. Any data point falling within this interval is considered a statistically normal value. Then, outliers exceeding the confidence interval in the raw monitoring data are identified, and a data quality factor is generated for each data point. Each data point in the real-time acquired or pending raw monitoring data is compared to the confidence interval of its corresponding indicator. If a data point's value exceeds the upper limit of its interval... or below the lower limit If a data point is not found to be an outlier, it is identified as such. Simultaneously, each original data point is assigned a data quality factor, a discrete identifier representing the data's quality status. For example, a data quality factor of 1 can be set to indicate normal data, and 0 to indicate an outlier. The output of this step is a data sequence labeled with the quality factors. Next, based on the data quality factors, a sliding window weighted interpolation correction is performed on the identified outliers. The sliding window weighted interpolation correction is as follows: Figure 3 As shown, the final result is preprocessed hydrochemical data and preprocessed hydrological data. This step only operates on data points identified as anomalous by the data quality factor; normal data points retain their original values. When a data point... At any moment When identified as an outlier, the system will... Construct a time-shifting sliding window centered on [the element], which contains [the elements]. front and back Normal data at each time point. Then, using the normal data points within the window, a correction value is calculated through a weighted average method. To replace the original outliers Weights are typically inversely proportional to time distance; that is, data closer to the outlier point in time has a greater weight. The formula for calculating the correction value can be expressed as:
[0053] ,
[0054] in, It is the calculated correction value; The time index representing each normal data point within the sliding window; The index within the window is Normal data values; These are the corresponding data points The weight can be calculated using the following formula:
[0055] ,
[0056] This means that the weight is the reciprocal of the absolute value of the time difference. After replacing all outliers with the calculated correction values, the resulting data sequence is the complete and smooth pretreated hydrochemical and hydrological data.
[0057] Optionally, the step of indexing and querying the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights includes:
[0058] Obtain the hydrological period-pollution load coupling matrix;
[0059] The river dynamic scenario parameters are used as indexes to query the hydrological period-pollution load coupling matrix to obtain the weight allocation values.
[0060] Based on the weight allocation values, water chemistry dynamic weights and biological dynamic weights are generated.
[0061] Specifically, the first step is to obtain a pre-constructed hydrological period-pollution load coupling matrix. This matrix is a two-dimensional knowledge base, established by experts in the field of water environment based on extensive ecological mechanism analysis, historical data mining, and practical experience. The rows of this matrix typically correspond to different current hydrological periods, such as high-water season, normal-water season, and low-water season; the columns correspond to different pollution load levels, such as low, medium, and high. The method for constructing the hydrological period-pollution load coupling matrix is as follows:
[0062] The matrix consists of 3 hydrological periods × 3 pollution load levels, with each cell containing a binary pair of "water chemistry-biological weight components". Expert experience was quantified using the Delphi method, and consensus was reached after 3 rounds of feedback iterations (Kendall coefficient ≥ 0.7). A "mechanism-weight" mapping was established based on ecological mechanisms (such as dilution / accumulation effects), and adjusted in conjunction with data validation. Historical data was mined using a random forest model, and the final weights were obtained by integrating model predictions with expert scores. The numerical range and format were clearly defined, and a 3-year update cycle and validation standards (accuracy ≥ 85%) were specified. Each cell in the matrix stores a core weight allocation value. This weight allocation value directly specifies the relative importance of water chemistry indicators and biological indices in the comprehensive health assessment under specific hydrological period and pollution load combinations. For example, in scenarios with low water levels and high pollution loads, the instantaneous stress effect of hydrochemical indicators is extremely significant, thus their weight allocation values will be higher. Conversely, in scenarios with high water levels and low pollution loads, the stability and structural integrity of the biological community better reflect the long-term health of the river, therefore the weight allocation values of the biological index will be correspondingly increased. Secondly, the river dynamic scenario parameters generated in the previous stage are used as indexes for querying within the hydrological period-pollution load coupling matrix. These river dynamic scenario parameters, such as "low water period - high pollution load," are themselves structured coordinates. Technicians use these parameters to directly locate the corresponding row (low water period) and column (high pollution load) in the matrix, thereby accurately retrieving the weight allocation values stored in that cell. This process is similar to consulting a pre-set decision table, characterized by efficiency, intuitiveness, and determinism. Finally, based on the weight allocation values obtained from the query, the final hydrochemical dynamic weights and biological dynamic weights are generated.
[0063] Optionally, the step of generating a synergistic comprehensive health index based on the original monitoring data, combined with the hydrochemical dynamic weights and biological dynamic weights, through coupled analysis includes:
[0064] Based on the data dispersion of the original monitoring data, the initial weights of each indicator are calculated using the entropy weight method;
[0065] Using the initial weights, the degree of closeness between the water chemical index data and the biological index data and their respective ideal states is calculated, and a synergy index is generated.
[0066] Based on the synergy index and the hydrochemical dynamic weight and biological dynamic weight, the comprehensive health index of synergy effect is calculated.
[0067] Specifically, firstly, based on the data dispersion of the original monitoring data, the initial weight of each indicator is objectively calculated using the entropy weight method. This step aims to eliminate the subjectivity of human weighting and allow the data itself to reveal the information contribution of each monitoring indicator in the evaluation system. First, the original monitoring data, including water chemical indicator data and biological index data, needs to be standardized to eliminate the influence of different indicator dimensions and orders of magnitude. Then, for each standardized indicator, its information entropy is calculated. The smaller the information entropy, the greater the data dispersion of the indicator, the more information its data sequence contains, and the higher its weight should be assigned in the evaluation. The initial weight is calculated based on information entropy; specifically, the information redundancy of each indicator is first calculated, and then the redundancy is normalized to obtain the initial weight of the indicator. This process assigns an objective initial weight to all water chemical indicators and biological indices, reflecting the evaluation information carried by their own data variability. Secondly, using the initial weights calculated in the previous step, the closeness of the water chemical indicator data and biological index data to their respective ideal states is calculated, thereby generating a synergy index. The synergy index is a quantitative indicator that measures the coordination and convergence of the components within a subsystem towards an ideal state. Technicians need to assign an ideal value to each indicator, representing its value under the healthiest river conditions. For example, the ideal value for dissolved oxygen could be saturated dissolved oxygen concentration, while the ideal value for pollutants is zero or background concentration. The closeness calculation measures the difference between measured values and ideal values. By weighting and summing the closeness of all indicators within each subsystem using their corresponding initial weights, the synergy index for that subsystem can be obtained. Therefore, this step generates two core synergy indices: one representing the health status of the hydrochemical subsystem, and the other representing the health status of the biological subsystem. The third step, based on the synergy index generated in the previous step and the previously determined hydrochemical and biological dynamic weights, calculates the comprehensive health index of the synergy effect through weighted coupling calculation. This step is the core integration link of the entire evaluation model, organically combining the internal health status of the two subsystems with their relative importance under the current river dynamic scenario.
[0068] Optionally, the calculation of the comprehensive health index based on the synergy index and the hydrochemical dynamic weight and biological dynamic weight includes:
[0069] The interaction strength coupling degree is calculated using the aforementioned synergy index.
[0070] The comprehensive coordination index is calculated using the synergy index and the water chemistry dynamic weight and biological dynamic weight.
[0071] By combining the interaction strength coupling degree and the comprehensive coordination index, the synergistic effect comprehensive health index is calculated using the coupling formula.
[0072] Specifically, the interaction strength and coupling degree between the two subsystems, hydrochemistry and biology, are first calculated using the previously obtained synergy index. Synergy Index and The degree of order or health level within the water chemistry subsystem and the biological subsystem were quantified separately. Interaction strength and coupling degree were also assessed. This aims to measure the degree of interrelation and influence between the development levels of these two subsystems, reflecting whether they develop synchronously or mutually restrict each other. Its calculation is based on the classic formula of the coupling degree model:
[0073] ,
[0074] in, The interaction strength is the coupling degree, which ranges from 0 to 1. It is the interaction strength coupling coefficient. The closer the value is to 1, the closer the developmental levels of the hydrochemical and biological subsystems are, the stronger their interaction and the tighter their coupling; conversely, the closer the value is to 1, the more closely the hydrochemical and biological subsystems are developed, the stronger their interaction and the tighter their coupling. The closer the value is to 0, the greater the gap in development levels between the two subsystems, indicating that they are unrelated or unbalanced. It is the synergy index of the water chemistry subsystem. These are the synergy indices of the biological subsystems; both values are calculated from previous steps. Then, using the synergy index, hydrochemical dynamics weights, and biological dynamics weights, the comprehensive coordination index is calculated. The comprehensive coordination index T is a further quantification of the coordinated development of the two subsystems, based on a comprehensive evaluation of their overall development level. It considers not only the individual health levels of the two subsystems but also their relative importance within the current specific river dynamics scenario. The calculation formula is as follows.
[0075] ,
[0076] in, As a comprehensive coordination index; For water chemistry dynamic weights, These are the biological dynamic weights, which are obtained by querying the hydrological period-pollution load coupling matrix from the river dynamic scenario parameters. and These are the synergy indices for the hydrochemical and biological subsystems, respectively. This index is essentially a weighted average health level of the two subsystems under dynamic weights. Next, combining the interaction strength coupling degree and comprehensive coordination index calculated in the previous two steps, the final synergy effect comprehensive health index is calculated. The calculation formula is as follows:
[0077] ,
[0078] By using square root calculation, the index distribution can be made more uniform, while ensuring that the final result can comprehensively reflect the dual effects of coupling and coordination. The synergistic effect comprehensive health index is a comprehensive reflection of coupling and coordination indices. It reflects the interaction relationships among various elements within the system as well as the overall development level of the system, and is the core indicator ultimately used to evaluate the ecological health status of rivers.
[0079] Optionally, the intelligent assessment of the synergistic effect comprehensive health index to determine the river's ecological health level includes:
[0080] Access the historical health case database;
[0081] A machine learning model is trained using the historical health case database to generate dynamic health thresholds;
[0082] The river's ecological health level is determined by comparing the synergistic effect comprehensive health index with the dynamic health threshold.
[0083] Specifically, the first step is to acquire and build a comprehensive historical health case library.
[0084] The construction of the historical health case database follows these operational procedures: Data sources: Monthly data from 100 monitoring sections within the basin over the past 20 years were collected, including water chemical indicators (pH, dissolved oxygen, etc.), biological indicators (benthic biodiversity, etc.), hydrological data, and pollution discharge records; Labeling basis: A review panel composed of more than 5 experts in the field of ecological environment with over 10 years of experience, the technical guidelines for river ecological health assessment, and on-site survey results (such as vegetation coverage and fish survival status) were used to independently label the health level of each case; Consensus mechanism: The Delphi method was used to reach a consensus on labeling. If the expert opinion disagreement rate was >30%, a third-party expert was introduced for review, ultimately forming a case database containing 10,000+ labeled samples, with the sample ratio for each health level controlled at 1:2:3:2 (Excellent: Good: Average: Poor); Update mechanism: Monitoring data and expert labeling results for the current year are added annually to ensure the timeliness of the case database. This case database is the foundation for training machine learning models. The model uses multi-dimensional feature vectors as input, specifically including: core indices: hydrochemical synergy index (range 0-100), biological synergy index (range 0-100), and synergistic effect comprehensive health index (range 0-100); environmental variables: current hydrological period (categorical variable, mapped to unique heat encoding), watershed pollution load level (categorical variable, divided into 3 levels according to pollutant emissions), seasonal factors (divided by quarter), and climate anomaly index over the past 5 years (calculated based on regional meteorological data). Algorithms that can be used include SVM (Support Vector Machine), random forest, or gradient boosting tree. Hyperparameters (such as kernel function selection and penalty coefficient for SVM) are optimized through grid search, and model performance is evaluated using 5-fold cross-validation, requiring a classification accuracy of ≥85%. Technical personnel need to collect extensive historical monitoring data for specific rivers or similar watersheds. Each case study should contain a complete record consisting of two parts. One part consists of input data used to calculate the comprehensive health index of synergistic effects, namely water chemical index data, biological index data, and hydrological condition data at a specific time point; the other part consists of corresponding river ecological health levels, such as "healthy," "sub-healthy," and "unhealthy," which are authoritatively labeled by domain experts based on multi-source information at the time, including on-site surveys, detailed biological surveys, and socio-economic impacts. The size, quality, and representativeness of this case library directly determine the performance of subsequent models. Then, the constructed historical health case library is used to train one or more machine learning models, with the aim of generating dynamic health thresholds, such as... Figure 4As shown, the "dynamic health threshold" refers to the dynamic decision boundary generated by the model based on the combination of input features. Its "dynamic" nature is reflected in: automatically adjusting weights for different hydrological periods (e.g., higher sensitivity to pollution load during the dry season); correcting the decision threshold by combining seasonal factors and climate anomaly indices (e.g., appropriately relaxing the threshold for organic pollution indicators during the rainy season); and the model output is not a single threshold, but rather directly maps the health level through a multi-dimensional decision boundary in the feature space (e.g., outputting "excellent" when the comprehensive health index of the synergistic effect is >80 and the pollution load is low; outputting "good" when the comprehensive index is 50-80 and the hydrological period is the high-water season). Through this mechanism, the model can adapt to the dynamic changes in the watershed ecosystem, improving the accuracy and adaptability of health level determination. The "dynamic health threshold" here is not a fixed set of values, but a decision model that can dynamically generate grading boundaries based on input conditions. Technicians can choose machine learning algorithms suitable for classification tasks, such as Support Vector Machines (SVM), decision trees, random forests, or neural networks. During training, the model uses raw monitoring data from historical cases or features calculated from them (such as the synergistic effect comprehensive health index, synergy index, etc.) as input features (X), and expert-annotated health levels as target labels (Y). By repeatedly learning these "feature-label" pairs, the model's internal parameters are continuously optimized until it can establish a precise mapping relationship from input data to health levels. The trained machine learning model internalizes a complex, non-linear decision boundary, which is the entity referred to in this invention as the "dynamic health threshold." Next, the synergistic effect comprehensive health index calculated at the current evaluation time, along with other potentially relevant parameters, are input into the trained machine learning model. Upon receiving this new input data, the model uses the decision rules learned during training to perform calculations and inferences. For example, a trained support vector machine model calculates the distance from input data points to each category hyperplane and classifies them to the health level represented by the closest category. Finally, the output of the machine learning model is used as the final river ecological health level. The model's output is directly a clear category, such as "healthy." This process essentially involves an implicit comparison between the newly calculated synergistic effect comprehensive health index and the dynamic health threshold within the model. The term "dynamic" refers to the fact that the model's decisions not only rely on the synergistic effect comprehensive health index but may also take into account other underlying parameters that constitute the index (such as different hydrological periods and pollution load levels), thus enabling the classification standards to adapt to different river scenarios.
[0085] Optionally, generating the hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values includes:
[0086] The weight allocation values are structurally decomposed to generate water chemical weight components and biological weight components;
[0087] The water chemical weight component and the biological weight component are subjected to data extraction and normalization to generate water chemical dynamic weight and biological dynamic weight.
[0088] Specifically, the weight allocation values retrieved from the coupling matrix are first structurally decomposed to generate hydrochemical and biological weight components. Each weight allocation value is a structured data unit stored in a specific cell of the hydrological period-pollution load coupling matrix. This data unit contains a quantitative description of the importance of hydrochemical indicators and biological indices. Technicians need to parse this value according to a predefined data structure. For example, a weight allocation value can be a pair of two elements, denoted as... The task of structural decomposition is to accurately extract the numerical values representing the importance of water chemistry from this complex data structure. and numerical values representing biological importance . This refers to the water chemical weight component. These are the biological weight components. These two components are unstandardized raw weight bias measures. The second step involves data extraction and normalization of the decomposed hydrochemical and biological weight components, ultimately generating directly applicable hydrochemical dynamic weights and biological dynamic weights. Data extraction refers to processing the data obtained in the previous step... and These two values serve as inputs for the calculation. Normalization ensures that the sum of the two final weight values is 1, meeting the basic mathematical requirements of weight coefficients, thus allowing direct use in weighted summation and other operations. The standard method for normalization is to calculate the proportion of each component to the total sum of components. The formula is as follows:
[0089] ,
[0090] ,
[0091] in, It is the final generated dynamic weight of water chemistry; It is the final generated biological dynamic weight; It is the extracted water chemical weight component; These are the extracted biological weight components. These are the two calculated weight values. and Both values are between 0 and 1, and their sum is always equal to 1. This pair of weights accurately reflects the relative importance of the hydrochemical and biological subsystems in the comprehensive evaluation under the current specific river dynamics scenario.
[0092] Optionally, the acquisition of river water chemical index data, biological index data, and hydrological condition data to generate raw monitoring data includes:
[0093] The water chemical index data, biological index data, and hydrological condition data were time-series aligned and format-standardized to obtain a multi-source fusion dataset.
[0094] The multi-source fusion dataset is parsed to generate raw monitoring data.
[0095] Specifically, the first step involves time series alignment and format standardization of three types of data—hydrochemical index data, biological index data, and hydrological condition data—which come from different sources and have varying characteristics, to obtain a unified multi-source fusion dataset. Hydrochemical index data may originate from minute-level high-frequency data from online monitoring stations, biological index data may come from quarterly manual sampling analysis, and hydrological condition data may come from daily average flow records from hydrological stations. The first step is to determine a unified time base and sampling frequency, for example, to a daily scale. For high-frequency data, downsampling is required, such as calculating daily averages; for low-frequency data, interpolation or preservation processing is needed to generate corresponding data points on a unified time axis. The second step is format standardization, which involves converting all data into a unified data structure, such as a two-dimensional data table indexed by time and with each index as a column. Simultaneously, the units and naming conventions of each index need to be standardized to ensure data consistency and comparability. After completing the above operations, a well-structured, time-synchronized multi-source fusion dataset is formed. Next, the multi-source fusion dataset is parsed to generate raw monitoring data that can be directly accessed in subsequent steps. This step involves data organization and encapsulation at the logical level. While the multi-source fusion dataset is structurally unified, in application, the evaluation model needs to clearly identify which data belong to water chemistry indicators, which to biological indices, and which to hydrological conditions. Therefore, the parsing process involves labeling and grouping the columns of the multi-source fusion dataset according to a pre-defined list of indicator classifications. For example, columns such as total nitrogen, total phosphorus, and chemical oxygen demand are labeled as "water chemistry indicator data," columns such as phytoplankton biomass and benthic animal integrity index are labeled as "biological index data," and columns such as flow rate and water level are labeled as "hydrological condition data." After parsing and organization, the final data package is the raw monitoring data. It is a structured dataset with clear internal logic, and each part of the data is easy to index and retrieve. It provides a standardized data input interface for subsequent data preprocessing, dynamic scenario analysis, and other computational steps. Dynamic scenario analysis, such as... Figure 5 As shown.
[0096] Based on the same inventive concept, such as Figure 6As shown, the present invention also provides a comprehensive evaluation system for river ecological health that integrates water chemistry and biological indices, the system comprising:
[0097] The data acquisition module is used to acquire river water chemical index data, biological index data, and hydrological condition data to generate raw monitoring data.
[0098] The dynamic scenario analysis module is used to perform dynamic scenario analysis on the raw monitoring data and generate dynamic scenario parameters for the river.
[0099] The weight allocation module is used to index and query the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights.
[0100] The coupling analysis calculation module is used to generate a synergistic effect comprehensive health index based on the original monitoring data and in combination with the water chemical dynamic weight and biological dynamic weight through coupling analysis calculation.
[0101] The intelligent assessment module is used to intelligently assess the comprehensive health index of the synergistic effect and determine the ecological health level of the river.
[0102] To verify the feasibility of this invention in practice, it was applied to a municipal environmental monitoring center to comprehensively evaluate the ecological health status of the middle and lower reaches of the Qingjiang River within its jurisdiction. The Qingjiang River is a typical urban river, its hydrological conditions significantly affected by seasonal rainfall, and it also carries agricultural non-point source pollution from upstream and urban point source pollution, resulting in a complex and variable ecological health status. Traditional evaluation methods often use fixed weights, making it difficult to accurately reflect the true health status of the river under different hydrological and pollution scenarios. This monitoring center hopes to use this invention to establish an intelligent evaluation system that can dynamically adapt to changes in the river environment.
[0103] In this embodiment, the monitoring center first integrated historical and real-time monitoring data from three monitoring sections set up in the Qingjiang River throughout 2024. The acquired data included water chemical index data (total nitrogen, total phosphorus, permanganate index), biological index data (phytoplankton biomass, benthic animal integrity index), and hydrological condition data (daily average flow). These multi-source data were first aligned to a time series (unified to a daily scale) and standardized in format to generate raw monitoring data.
[0104] When performing dynamic scenario analysis on raw monitoring data, the system first preprocesses the data. For example, on July 15, 2024, a transient spike in flow data was detected, exceeding the historical average by three standard deviations. The system identified this as an outlier and corrected it using a sliding window weighted interpolation method based on its data quality factor, generating smooth and reliable preprocessed hydrological data. Subsequently, based on the preprocessed hydrological data, the system determines the current hydrological period according to the historical flow percentile method (below 25% indicates a dry season, above 75% indicates a wet season). For example, in August 2024, due to concentrated rainfall during the rainy season, the daily average flow was consistently higher than the historical 75th percentile, and it was determined to be a "wet season"; while in December 2024, with scarce rainfall, the flow was below the 25th percentile, and it was determined to be a "dry season". Simultaneously, based on the preprocessed hydrochemical data, the system uses the single-factor pollution index method to assess the pollution load level. During the "high-water season" in August, due to the large runoff and strong dilution effect, the maximum total nitrogen pollution index was 1.3, which was rated as "slight pollution". The combined dynamic scenario parameters for this period were "high-water season - slight pollution". In contrast, during the "low-water season" in December, the runoff was small and pollutants were enriched. The maximum total nitrogen pollution index reached 2.8, which was rated as "severe pollution". The dynamic scenario parameters were "low-water season - severe pollution".
[0105] Next, the system indexes and queries the river's dynamic scenario parameters to generate dynamic weights. Under the "high water season - light pollution" scenario, the stability of the biological community better reflects ecological health. The system retrieves a weight allocation value of (0.4, 0.6) from the hydrological period-pollution load coupling matrix. After normalization, the resulting hydrochemical dynamic weight is 0.4, and the biological dynamic weight is 0.6. However, under the "low water season - heavy pollution" scenario, the acute toxicity risk of hydrochemical indicators becomes the primary concern. The retrieved weight allocation value is (0.8, 0.2), resulting in a hydrochemical dynamic weight of 0.8 and a biological dynamic weight of 0.2.
[0106] In calculating the comprehensive health index of synergistic effect ( When calculating the initial weights of each specific index using the entropy weight method, the system first calculates the initial weights of each index. Then, the synergy index of the water chemistry subsystem is calculated based on these initial weights. Synergy Index of Biological Subsystems Taking the "dry season - severe pollution" scenario in December as an example, the calculated... It is 0.35. The value is 0.55. Based on this, the interaction strength coupling degree is calculated. A score of 0.93 indicates a significant difference in the development levels of the two subsystems. (Comprehensive Coordination Index) Calculated based on dynamic weights =0.8×0.35+0.2×0.55=0.39. Ultimately, the synergistic effect comprehensive health index... = ≈0.60.
[0107] Finally, an intelligent assessment of the synergistic effect comprehensive health index was conducted. The monitoring center constructed a historical health case database using historical monitoring data of the Qingjiang River from the past ten years and corresponding expert assessment conclusions, and trained a support vector machine (SVM) classification model. The synergistic effect comprehensive health index of 0.60 calculated in December, along with the scenario information of "dry season - heavy pollution," was input into the model. The model, combined with its learned dynamic health threshold, ultimately determined the river's ecological health level to be "unhealthy." In contrast, the synergistic effect comprehensive health index calculated under the "high water season - light pollution" scenario in August was 0.75, and the model assessed its level as "sub-healthy."
[0108] Table 1. Calculation data of weights and indices for different dynamic scenarios of the Qingjiang River.
[0109]
[0110] Table 2 Comparison of Evaluation Results between the Evaluation Method of the Invention and Traditional Methods
[0111]
[0112] As can be seen from the data in Tables 1 and 2 above, the evaluation method proposed in this invention exhibits significant advantages. Table 1 clearly demonstrates the core mechanism of this method: the weights of hydrochemical and biological indices can be adaptively adjusted under different hydrological and pollution coupling scenarios. For example, from the wet season to the dry season, the weight of the hydrochemical index dynamically increases from 0.4 to 0.8, which allows the evaluation model to focus on the main stress factors of the current ecosystem. The final calculated synergistic effect comprehensive health index value and health level are also highly correlated with scenario changes.
[0113] Table 2, through comparison with the traditional fixed-weight method, further confirms the effectiveness of the present invention. In May and October, when hydrological conditions and pollution loads are relatively stable, the conclusions obtained by both methods are basically consistent with the expert review levels. However, under the two typical scenarios of the high-water season in August and the low-water season in December, the evaluation results of the traditional method show deviations. Because the traditional method uses fixed weights, it fails to fully consider the scouring effect of large flow rates on benthic organisms during the high-water season and the acute ecotoxicity risks caused by the sharp increase in pollutant concentrations during the low-water season, leading to overly optimistic evaluation results. In contrast, the evaluation level of the present invention completely matches the expert review conclusions, proving that this method, through dynamic scenario analysis and dynamic weight assignment, can more profoundly reveal the true health status of the river ecosystem, resulting in more scientific and accurate evaluation results.
[0114] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.
[0115] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Other embodiments of the invention will be readily apparent to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or conventional techniques in the art not described herein.
Claims
1. A comprehensive evaluation method for river ecological health integrating water chemistry and biological indices, characterized in that, The method includes: Acquire water chemical index data, biological index data and hydrological condition data of the river to generate raw monitoring data; Dynamic scenario analysis is performed on the raw monitoring data to generate river dynamic scenario parameters, including: preprocessing the raw monitoring data to generate preprocessed hydrochemical data and preprocessed hydrological data; analyzing the preprocessed hydrological data to determine the current hydrological period; evaluating the preprocessed hydrochemical data to determine the pollution load level; and combining the current hydrological period and the pollution load level to generate river dynamic scenario parameters. The river dynamic scenario parameters are indexed and queried to generate hydrochemical dynamic weights and biological dynamic weights, including: obtaining the hydrological period-pollution load coupling matrix; using the river dynamic scenario parameters as an index to query the hydrological period-pollution load coupling matrix to obtain weight allocation values; and generating hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values. Based on the original monitoring data, and combining the hydrochemical dynamic weights and biological dynamic weights, a synergistic effect comprehensive health index is generated through coupling analysis. This includes: calculating the initial weights of each indicator using the entropy weight method based on the data dispersion of the original monitoring data; using the initial weights, calculating the closeness of the hydrochemical indicator data and biological index data to their respective ideal states to generate a synergy index; and calculating the synergistic effect comprehensive health index based on the synergistic index and the hydrochemical dynamic weights and biological dynamic weights. This includes: calculating the interaction strength coupling degree using the synergistic index; calculating the comprehensive coordination index using the synergistic index and the hydrochemical dynamic weights and biological dynamic weights; and calculating the synergistic effect comprehensive health index using a coupling formula, combining the interaction strength coupling degree and the comprehensive coordination index. The comprehensive health index of the synergistic effect is intelligently assessed to determine the ecological health level of the river.
2. The method for comprehensive evaluation of river ecological health integrating hydrochemical and biological indices according to claim 1, characterized in that, The preprocessing of the raw monitoring data includes: Obtain the historical distribution characteristics of the original monitoring data and generate data confidence intervals; Identify outliers in the raw monitoring data that exceed the data confidence interval and generate data quality factors; Based on the data quality factor, the outliers are corrected by sliding window weighted interpolation to generate pre-processed hydrochemical data and pre-processed hydrological data.
3. The method for comprehensive evaluation of river ecological health integrating water chemistry and biological indices according to claim 1, characterized in that, The intelligent assessment of the comprehensive health index of the synergistic effect to determine the river's ecological health level includes: Access the historical health case database; A machine learning model is trained using the historical health case database to generate dynamic health thresholds; The river's ecological health level is determined by comparing the synergistic effect comprehensive health index with the dynamic health threshold.
4. The method for comprehensive evaluation of river ecological health integrating water chemistry and biological indices according to claim 1, characterized in that, The generation of hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values includes: The weight allocation values are structurally decomposed to generate water chemical weight components and biological weight components; The water chemical weight component and the biological weight component are subjected to data extraction and normalization to generate water chemical dynamic weight and biological dynamic weight.
5. The method for comprehensive evaluation of river ecological health integrating water chemistry and biological indices according to claim 1, characterized in that, The acquisition of river hydrochemical index data, biological index data, and hydrological condition data to generate raw monitoring data includes: The water chemical index data, biological index data, and hydrological condition data were time-series aligned and format-standardized to obtain a multi-source fusion dataset. The multi-source fusion dataset is parsed to generate raw monitoring data.
6. A comprehensive river ecological health evaluation system integrating hydrochemical and biological indices, applied to the comprehensive river ecological health evaluation method integrating hydrochemical and biological indices as described in any one of claims 1-5, characterized in that, The system includes: The data acquisition module is used to acquire river water chemical index data, biological index data, and hydrological condition data to generate raw monitoring data. The dynamic scenario analysis module is used to perform dynamic scenario analysis on the raw monitoring data and generate river dynamic scenario parameters. This includes: preprocessing the raw monitoring data to generate preprocessed hydrochemical data and preprocessed hydrological data; analyzing the preprocessed hydrological data to determine the current hydrological period; evaluating the preprocessed hydrochemical data to determine the pollution load level; and combining the current hydrological period and the pollution load level to generate river dynamic scenario parameters. The weight allocation module is used to index and query the river dynamic scenario parameters to generate hydrochemical dynamic weights and biological dynamic weights, including: obtaining the hydrological period-pollution load coupling matrix; using the river dynamic scenario parameters as an index to query the hydrological period-pollution load coupling matrix to obtain weight allocation values; and generating hydrochemical dynamic weights and biological dynamic weights based on the weight allocation values. The coupling analysis calculation module is used to generate a synergistic effect comprehensive health index based on the original monitoring data and the hydrochemical dynamic weights and biological dynamic weights through coupling analysis. This includes: calculating the initial weights of each indicator using the entropy weight method based on the data dispersion of the original monitoring data; using the initial weights, calculating the closeness of the hydrochemical indicator data and biological index data to their respective ideal states to generate a synergy index; and calculating the synergistic effect comprehensive health index based on the synergistic index and the hydrochemical dynamic weights and biological dynamic weights. This includes: calculating the interaction strength coupling degree using the synergistic index; calculating the comprehensive coordination index using the synergistic index and the hydrochemical dynamic weights and biological dynamic weights; and calculating the synergistic effect comprehensive health index using a coupling formula by combining the interaction strength coupling degree and the comprehensive coordination index. The intelligent assessment module is used to intelligently assess the comprehensive health index of the synergistic effect and determine the ecological health level of the river.
Citation Information
Patent Citations
River health assessment method based on coordinated development degree
CN106355016A
Multi-index associated river water quality evaluation method and system, equipment and storage medium
CN114037202A
