A water ecological environment resilience assessment method and system based on a dynamic baseline

Abstract

This invention discloses a method and system for assessing the resilience of aquatic ecosystems based on dynamic baselines. The method includes: establishing a dynamic baseline; obtaining a time series sequence of the comprehensive water quality index of a target water body over historical periods; extracting the trend component, seasonal component, and residual component from the time series sequence of the comprehensive water quality index; constructing a dynamic baseline and its confidence interval that varies over time; and calculating an aquatic ecosystem resilience index based on an adaptive index, a self-repair index, and a self-purification index. This invention overcomes the shortcomings of existing static assessments and single-dimensional assessments by using a dynamic baseline and a three-dimensional resilience assessment framework, thereby improving the dynamics and comprehensive characterization of aquatic ecosystem resilience and enabling accurate diagnosis and adaptive evaluation of the health status of aquatic ecosystems.

Description

Technical Field

[0001] This invention relates to the field of water environment monitoring and ecological assessment technology, specifically to a method and system for assessing the resilience of the water ecological environment based on dynamic baselines. Background Technology

[0002] Surface water is a dynamic, open, and nonlinear system. Its water quality is not only affected by normal pollution loads but also faces more frequent external shocks such as sudden pollution events and extreme hydrological and climatic conditions. Traditional static evaluation methods, represented by "water quality categories" or "water quality index (WQI)," can only reflect the pollution status at a specific point in time and cannot characterize the system's dynamic adaptation process, restoration capacity, and purification potential under disturbance. This often leads to management that only addresses the symptoms and not the root cause, making it difficult to implement precise regulation on vulnerable links of the ecosystem.

[0003] Ecological resilience theory provides a new paradigm for understanding and assessing a system's ability to cope with change. For example, Chinese patent CN121391041A discloses a quantitative evaluation method and device for water ecological resilience specific to the characteristics of northern rivers. It constructs a three-level evaluation index system through the DPSIR model and uses a combined weighting method combining the analytic hierarchy process (AHP) and entropy weighting to determine the index weights. It comprehensively considers the seasonal fluctuations and freezing-thawing characteristics of northern rivers. However, this method is mainly based on a pre-set index system for evaluation, and its evaluation results reflect more of the stage-specific state characteristics. The characterization of the system's state change process before and after disturbance is still mainly discrete, and it does not model and quantify the system from the complete dynamic chain of "resistance to shock - rapid recovery - long-term stability". Chinese patent CN120316736A discloses a dynamic evaluation system and method for marine ecological resilience based on multi-dimensional indicators. It introduces an LSTM neural network to dynamically adjust the index weights and uses a Bayesian network to analyze the causal relationship between human activities and ecological degradation. However, this method focuses on the marine environment, and its evaluation framework does not design targeted resilience quantification dimensions for the dynamic processes of surface water systems.

[0004] In general, existing technologies suffer from the following shortcomings: First, they rely on static baselines, often using fixed historical averages or water quality standards as the "normal state" baseline, neglecting the dynamic changes of environmental factors such as flow and climate, leading to inaccurate judgments of "abnormality" and "recovery." Second, they are limited in scope, focusing only on single dimensions such as recovery speed or shock resistance magnitude, failing to integrate the complete chain of self-adaptation, self-repair, and self-purification for systematic evaluation. Third, they lack sufficient model mechanistic understanding; most methods rely on traditional empirical formulas, and existing methods largely depend on traditional statistics or empirical formulas, failing to fully integrate high-frequency monitoring data with machine learning techniques to construct dynamic models capable of quantifying the complex coupling relationship between environmental factors and water quality indicators, thus limiting the mechanistic nature and extrapolation applicability of the evaluation results. Therefore, there is an urgent need to develop a new resilience assessment method and technology system that can objectively, dynamically, and multidimensionally diagnose the health and vulnerability of surface water systems and provide support for intelligent and adaptive management. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and system for assessing the resilience of aquatic ecosystems based on dynamic baselines.

[0006] A method for assessing the resilience of aquatic ecosystems based on dynamic baselines includes the following steps: S101. Establish a dynamic baseline; Obtain the time series of the comprehensive water quality index of the target water body in historical periods; The trend component, seasonal component, and residual component of the water quality comprehensive index time series were extracted using the time series decomposition method. The trend component and the seasonal component are aligned and superimposed according to time points to generate a dynamic baseline that changes over time; then the confidence interval of the dynamic baseline is determined based on the standard deviation of the residual components. S102. Calculate the water ecological environment resilience index; Based on the numerical relationship between the observed comprehensive water quality index of the target water body and the dynamic baseline and confidence interval, the adaptive index, self-repair index, and self-purification index are calculated respectively: The adaptive index is the deviation of the observed water quality index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness with which the observed water quality index returns to the confidence interval after the impact event ends; the self-purification index is the proportion of time during which the observed water quality index is within the confidence interval; wherein, the impact event is an event in which the observed water quality index exceeds the confidence interval. The adaptive index, self-repair index, and self-purification index are standardized and then weighted and fused to obtain the aquatic ecological environment resilience index, which is used to indicate the health status and anti-interference ability of the target water body.

[0007] Note: The above method overcomes the shortcomings of traditional static evaluation standards in reflecting the natural fluctuations of water bodies by constructing a dynamic baseline that adaptively adjusts with long-term trends and seasonal patterns, making the identification of abnormal water quality events more scientific and accurate. It innovatively quantifies the resilience of the aquatic ecological environment from three dimensions: "resisting shocks, restoring the original state, and maintaining stability," realizing a leap from single-state evaluation to dynamic capacity diagnosis throughout the entire process. It can provide timely, comprehensive, and targeted decision-making basis for water environment management.

[0008] Furthermore, the water quality comprehensive index time series is a continuous numerical sequence obtained by performing dimensionless processing and weighted summation on multi-index water quality monitoring data arranged at fixed time intervals; The multi-indicator water quality monitoring data includes the concentration values ​​of pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen, and total phosphorus.

[0009] Note: The above method integrates key water quality parameters such as pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen, and total phosphorus, and constructs a single, continuous time series of comprehensive water quality index through dimensionless and weighted fusion. This forms a standardized data foundation that can be directly used for time series analysis, providing a stable, reliable, and physically meaningful input for subsequent dynamic baseline modeling and resilience quantification.

[0010] Furthermore, the time series decomposition method is one of STL decomposition, X-12-ARIMA decomposition, moving average decomposition, or empirical mode decomposition.

[0011] Note: By employing mature and reliable time series decomposition methods such as STL and X-12-ARIMA, long-term trends, seasonal fluctuations, and random residuals can be professionally extracted from the comprehensive water quality index series, thereby constructing a dynamic baseline that can both reflect the inherent evolution of water bodies and be adaptively adjusted.

[0012] Furthermore, the method for determining the confidence interval of the dynamic baseline based on the standard deviation of the residual components includes: Take the standard deviation of the residual components; The standard deviation of the dynamic baseline ± 1.96 times is used as the upper and lower limits of the confidence interval.

[0013] Explanation: By dynamically determining the confidence interval based on the residual standard deviation, the natural fluctuation range of water quality under normal conditions can be objectively quantified, thereby scientifically distinguishing changes caused by predictable factors such as seasons and trends from genuine abnormal shock events, significantly improving the accuracy and reliability of water quality anomaly identification.

[0014] Furthermore, the formula for calculating the adaptive index is as follows: (1) In equation (1), For adaptive exponents, w i For the first i The proportion of days the impact event lasted out of the total number of days. di For the first i The average relative deviation of each impact event.

[0015] Note: The above calculation formula uses the duration (weight) of the impact event. w i ) and severity (deviation) di The system undergoes comprehensive quantification and normalization to scientifically transform the overall deviation impact of the system during disturbances into an index with stable range and strong comparability.

[0016] Furthermore, the formula for calculating the self-healing index is as follows: (2) In equation (2), The self-repair index, The duration of the impact event, This is the time required from the end of the shock event to recovery to within the dynamic baseline confidence interval. To restore the observed comprehensive water quality index values ​​after they have stabilized, This is the dynamic baseline value.

[0017] Note: The above calculation formula integrates the efficiency and effectiveness of the system's recovery from the shock into a single, comparable index by simultaneously quantifying the rate (time ratio) and completeness (proximity to the final state) of the recovery process, thereby scientifically assessing the self-repair capacity of the aquatic ecosystem.

[0018] Furthermore, the formula for calculating the self-purification index is as follows: (3) In equation (3), The self-purification index This refers to the number of time points during which the observed comprehensive water quality index values ​​fall within the dynamic baseline confidence interval across the entire time period. This represents the total number of time points within the entire time period.

[0019] Explanation: The above formula calculates the proportion of time that water quality indicators are within the normal fluctuation range, transforming the system's long-term ability to maintain a stable state into an intuitive and comparable quantitative index, thereby effectively overcoming the shortcomings of traditional methods in lacking a continuous quantitative description of the system's "healthy normal state".

[0020] Furthermore, the formula for calculating the aquatic ecological environment resilience index is as follows: (4) In equation (4), For adaptive exponents, The self-repair index, The self-purification index; , , These are the weighting coefficients, and .

[0021] The present invention also provides a water ecological environment resilience assessment system based on dynamic baselines, for implementing the above-mentioned water ecological environment resilience assessment method based on dynamic baselines, including: a historical data acquisition module, a dynamic baseline construction module, a resilience index calculation module, and a result output module; The historical data acquisition module is used to acquire the time series sequence of the comprehensive water quality index of the target water body in historical periods; The dynamic baseline construction module is used to extract the trend component, seasonal component, and residual component from the time series of the comprehensive water quality index using a time series decomposition method; the trend component and seasonal component are added point-by-point to obtain the dynamic baseline that changes over time; and the confidence interval of the dynamic baseline is determined based on the standard deviation of the residual component. The resilience index calculation module is used to calculate the adaptive index, self-repair index, and self-purification index based on the numerical relationship between the observed comprehensive water quality index of the target water body and the dynamic baseline and confidence interval. The adaptive index is the deviation of the observed comprehensive water quality index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness of the observed comprehensive water quality index returning to the confidence interval after the impact event; and the self-purification index is the proportion of time during which the observed comprehensive water quality index remains within the confidence interval. The impact event is defined as an event where the observed comprehensive water quality index exceeds the confidence interval. The adaptive index, self-repair index, and self-purification index are standardized and then weighted and fused to obtain the aquatic ecological environment resilience index. The results output module is used to classify resilience levels based on the aquatic ecological environment resilience index values ​​and to visualize them in the form of maps, time-series curves, or radar charts.

[0022] Note: The system described above automates and standardizes the entire resilience assessment process based on dynamic baselines. Through modular design, it achieves a complete link from data acquisition, baseline modeling, index calculation to result visualization, ensuring the consistency and efficiency of the assessment process.

[0023] The beneficial effects of this invention are: This invention provides a dynamic baseline that adaptively adjusts to long-term trends and seasonal patterns, fundamentally overcoming misjudgments caused by traditional static baselines and significantly improving the accuracy and consistency of water quality anomaly identification and status assessment. Furthermore, this invention proposes and quantifies for the first time a three-pronged framework for assessing the resilience of the aquatic ecosystem—"adaptive, self-repairing, and self-purifying"—comprehensively depicting the entire process of aquatic ecosystems' resistance to shocks, rapid recovery, and long-term stability, achieving a breakthrough in diagnostic dimensions from a single state to a dynamic process. The method is computationally efficient and highly adaptable, applicable to dynamic assessments at different time scales, and can provide timely and scientific decision support for water environment management and ecological protection. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a time series diagram and dynamic baseline diagram of the comprehensive water quality index in an embodiment of the present invention; Figure 3 This is a radar chart for assessing the resilience of the aquatic ecosystem in an embodiment of the present invention. Figure 4 This is a schematic diagram of the water ecological environment resilience assessment system in an embodiment of the present invention. Detailed Implementation

[0025] To further illustrate the methods and effects of this invention, the technical solution of this invention will be clearly and completely described below in conjunction with experiments.

[0026] The aquatic ecological environment is a dynamic open system composed of aquatic biological communities and their physical and chemical environment. Its health status is directly related to water resource security and ecosystem service functions. The resilience of the aquatic ecological environment refers to the system's ability to maintain structural and functional integrity and quickly recover to a steady state when faced with external disturbances such as pollution impacts and extreme climates.

[0027] However, as can be seen from the background information, existing methods for assessing the resilience of water ecosystems suffer from drawbacks such as static baselines, limited dimensions, and insufficient mechanisms. Specifically, static baselines refer to the use of fixed standards or historical averages as the "normal state," which cannot adapt to dynamic environmental changes. Limited dimensions refer to focusing only on single dimensions such as recovery rate or impact resistance magnitude, lacking a complete assessment of the resistance-recovery-stability chain. Insufficient mechanisms refer to relying on traditional statistical or empirical formulas and failing to fully utilize high-frequency monitoring data and machine learning to construct a dynamically coupled model, ultimately resulting in an inability to accurately describe the complete dynamic response process of surface water systems under disturbance.

[0028] To address this issue, this invention proposes constructing a dynamic baseline using time-series decomposition, quantifying resilience from three independent dimensions: self-adaptation, self-repair, and self-purification, thus solving the problem of existing technologies' difficulty in accurately identifying system vulnerabilities. The specific implementation process is as follows: Example: A method for assessing the resilience of aquatic ecosystems based on dynamic baselines, comprising the following steps: S101. Establish a dynamic baseline; (1) Obtain the time series of the comprehensive water quality index of the target water body in historical periods; Specifically, the water quality comprehensive index time series is a continuous numerical sequence obtained by performing dimensionless processing and weighted summation on multi-indicator water quality monitoring data arranged at fixed time intervals; wherein, the multi-indicator water quality monitoring data includes the concentration values ​​of pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen and total phosphorus. For example, daily water quality monitoring data of a certain river released by the Environmental Monitoring Center from 2021 to 2024 were obtained. The monitoring indicators included water temperature, pH, dissolved oxygen, permanganate index, ammonia nitrogen, and total phosphorus. The raw data underwent quality control and cleaning, including: removing obvious outliers (such as data with pH values ​​exceeding the range of 2-12); handling missing values, using linear interpolation to complete data with no more than 3 consecutive days of missing data, and removing data with long-term missing values; and standardizing the data time resolution by resampling the data into daily data. It should be understood that the handling of outliers and missing values ​​in the above data preprocessing steps can prevent false trends or seasonal distortions in the decomposition results.

[0029] The CCME WQI method was adopted, with the Class II water standard in the Surface Water Environmental Quality Standard (GB 3838-2002) as the target value, and the comprehensive water quality index was calculated daily; the calculation formula is as follows (1-1): (1-1) In equation (1-1), F 1 The proportion of indicators exceeding the standard out of all indicators. F 2 The percentage of measurements exceeding the standard out of all measurements. F 3 The value represents the extent of exceeding the standard (i.e., the maximum ratio of the exceeding concentration to the standard value). The above formula (1-1) converts multi-indicator information into dimensionless values ​​by comprehensively considering the three dimensions of exceeding the standard range, frequency and intensity. The higher the value, the better the water quality.

[0030] Calculations yielded a WQI time series of 1452 days for a certain river from 2021 to 2024, as follows: Figure 2 As shown; from Figure 2It can be observed that the WQI exhibits obvious seasonal fluctuations, with lower values ​​in summer (better water quality) and higher values ​​in winter (poorer water quality), while also showing a long-term trend of slow improvement. This provides a data basis for the subsequent decomposition of trend and seasonal components.

[0031] (2) Using the time series decomposition method, extract the trend component, seasonal component and residual component from the time series of the comprehensive water quality index; The time series decomposition method is one of STL decomposition, X-12-ARIMA decomposition, moving average decomposition, or empirical mode decomposition. Preferably, the choice of different decomposition methods depends on the data characteristics: STL decomposition is suitable for data with robust seasonal cycles and is not sensitive to outliers; X-12-ARIMA is suitable for economic data that needs to handle calendar effects such as leap years and transaction days; moving average decomposition is simple to calculate but has low accuracy; empirical mode decomposition is suitable for non-stationary and nonlinear signals; in this embodiment, the STL decomposition method is used because STL decomposition has a stable effect on extracting long-term trends and annual cycles of water quality data.

[0032] For example, WQI data from the first three years (2021-2023) are used as the training set to construct a dynamic baseline, and data from the fourth year (2024) are used as the validation set to verify the predictive ability of the baseline. The ADF stationarity test is performed on the WQI series of the training set, and the p-value is <0.05, indicating that the series is stationary and no differencing is required. In other embodiments, if the series is not stationary, differencing or transformation processing is required first, otherwise the decomposition result may result in spurious regression.

[0033] The WQI series was decomposed using the Seasonal-Trend Decomposition (STL) method. STL decomposition extracted each component through iterative Locally Weighted Regression (LOESS), and the seasonal cycle was identified using Autocorrelation Function (ACF) analysis. The results showed a significant autocorrelation coefficient with a lag of 365 days (r=0.68, p<0.01), confirming the seasonal cycle as 365 days (annual cycle). The decomposition results included: trend components. Tt Seasonal portion St residual components Rt ; It should be understood that the normal state of aquatic ecosystems is determined by both long-term evolutionary trends and seasonal natural fluctuations. The dynamic baseline formed by the superposition of these two factors can adaptively adjust over time, rather than using a fixed value. For example, a higher baseline value in winter and poorer water quality is normal and should not be misjudged as a pollution shock; conversely, a lower baseline value in summer and better water quality, if the observed value is abnormally high, should be identified as a shock event. Therefore, based on the above decomposition results, the following step (3) is used to obtain a dynamically changing baseline.

[0034] (3) Align and superimpose the trend component and the seasonal component according to the time point to generate a dynamic baseline that changes over time; then determine the confidence interval of the dynamic baseline based on the standard deviation of the residual component. dynamic baseline value Bt It is obtained by superimposing the trend component and the seasonal component, as shown in the following formula: Bt = Tt + St (1-2) In equation (1-2), Tt As a trend component, St For seasonal quantities.

[0035] The method for determining the confidence interval of the dynamic baseline based on the standard deviation of the residual components includes: taking the standard deviation of the residual components; and using the standard deviation of the dynamic baseline ± n times as the upper and lower bounds of the confidence interval. The selection of the above-mentioned value of n directly affects the sensitivity of shock event identification: if n is too small, it will lead to misjudging normal fluctuations as shocks; if n is too large, it will miss real shocks. Usually, n=1.96 is taken to correspond to a 95% confidence level, that is, under the assumption of normal distribution, about 95% of the data points should fall within the confidence interval, and those outside this range are considered to be statistically abnormal. Based on the statistical properties of the residual components, the 95% confidence interval is calculated: (1-3) In equation (1-3), The standard deviation of the residual sequence is calculated to be 3.1505.

[0036] Comparing the validation set (2024) WQI data with the constructed dynamic baseline and confidence intervals, the mean of the validation set residuals was 0.21 and the standard deviation was 3.42, showing no significant difference from the training set residual distribution (KS test, p=0.23), indicating that the dynamic baseline has good predictive ability and stability. The validation steps described above, conducted using the validation set, are necessary to ensure model reliability. If the validation set residual distribution deviates significantly from the training set, it indicates overfitting or structural changes in the baseline construction, requiring readjustment of model parameters or decomposition methods.

[0037] Based on the above dynamic baseline, three independent dimensions can be selected to construct the water ecological resilience, thereby achieving quantitative assessment, as shown in the calculation process of the three indices in S102 below and the water ecological environment resilience index.

[0038] S102. Calculate the water ecological environment resilience index; (1) Based on the numerical relationship between the observed comprehensive water quality index of the target water body and the dynamic baseline and confidence interval, the adaptive index, self-repair index and self-purification index are calculated respectively: The adaptive index is the deviation of the observed water quality index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness with which the observed water quality index returns to the confidence interval after the impact event ends; the self-purification index is the proportion of time during which the observed water quality index is within the confidence interval; wherein, the impact event is an event in which the observed water quality index exceeds the confidence interval. The formula for calculating the adaptive index is as follows: (1) In equation (1), For adaptive exponents, w i For the first i The proportion of days the impact event lasted out of the total number of days. di For the first i The average relative deviation of each impact event; d i The calculation formula is: In the formula, n i For the first i The number of days the event lasted. Ot for t WQI observation at time 10:00 Bt The dynamic baseline value at time t; when Ot < Bt When water quality deteriorates, the deviation is positive. This formula uses relative deviation rather than absolute difference, eliminating the influence of baseline level differences in different seasons and making each impact event comparable. The introduction of the weight wi reflects the management concept that "events with longer durations have a greater impact on system resilience."

[0039] Optionally, the result obtained by applying the softplus function to formula (1) can be nonlinearly transformed to adjust the parameters and make the adaptive exponential distribution more reasonable: ; The formula for calculating the self-healing index is as follows: (2) In equation (2), The self-repair index, The duration of the impact event, This is the time required from the end of the shock event to recovery to within the dynamic baseline confidence interval. and These are the observed values ​​of the comprehensive water quality index after the water quality has stabilized and the dynamic baseline value, respectively.

[0040] The recovery criterion is as follows: WQI observations are considered to be fully recovered if they remain within the confidence interval for three consecutive days. This is the time interval from the end of the impact to the first time the condition is met. Ratio The index reflects the recovery efficiency. A ratio of 1 indicates excellent toughness; if the recovery time exceeds the impact time, the index is less than 1, indicating insufficient toughness. The index is reduced to account for incomplete recovery. If the observed values ​​still deviate from the baseline after recovery, the index will be reduced accordingly.

[0041] The formula for calculating the self-purification index is as follows: (3) In equation (3), The self-purification index This refers to the number of time points during which the observed comprehensive water quality index values ​​fall within the dynamic baseline confidence interval across the entire time period. This represents the total number of time points in the evaluation period (excluding the impact period). The higher the value of this formula, the more stable the system's daily operation and the stronger its self-regulation ability to withstand minor disturbances.

[0042] For example, a water quality shock event is defined as a situation where the observed value of the water quality composite index (actual WQI value) is lower than the lower bound of the dynamic baseline confidence interval. Seven water quality shock events were identified between 2021 and 2024, as shown in Table 1: Table 1. Impact Event Identification Results

[0043] Taking event 1 in Table 1 as an example, w i = n i / 1452, where 1452 represents the total number of days from 2021 to 2024; as can be seen from the table, the impact events are mainly concentrated in 2022-2023, with each event lasting 1-4 days, and an average deviation of about 11%, indicating that although the river occasionally experiences water quality deterioration, the impact intensity is relatively controllable. The calculation process for d1 is as follows: Taking event 1 (April 28, 2022) as an example, on that day... Bt =73.2, Ot =65.1, then d1=(73.2−65.1) / 73.2=0.1107; Using equations (1), (2), and (3) above, we can calculate that... Adaptive index = =0.9010; Optimization using nonlinear transformations (softplus function): After transformation, we obtain the adaptive exponent. : = 0.4263; Self-repair index =0.8900; Of the 1452 days in total, the WQI was within the confidence interval for 1401 days, with a total of 14 days of shock periods. Self-purification index = =0.9743; (2) The adaptive index, self-repair index and self-purification index are standardized and weighted to obtain the water ecological environment resilience index, which is used to indicate the health status and anti-interference ability of the target water body.

[0044] The formula for calculating the aquatic ecological environment resilience index is as follows: (4) In equation (4), For adaptive exponents, The self-repair index, The self-purification index; , , These are the weighting coefficients, and .

[0045] For example, the adaptive index weight α = 0.35, the self-repair index weight β = 0.35, and the self-purification index weight γ = 0.30 are set.

[0046] The aquatic ecological environment resilience index = 0.35 × 0.4263 + 0.35 × 0.8900 + 0.30 × 0.9743 = 75.3; Radar image of the ecological resilience of a certain river. Figure 3The three indices of a certain river were 0.42, 0.89, and 0.97, respectively. According to the resilience level classification standard (high resilience: ≥80 points; medium resilience: 60-79 points; low resilience: 40-59 points; vulnerable: <40 points), the assessment result of the river's aquatic ecological environment resilience level is "medium resilience". The area size of the radar chart reflects the overall resilience level. A shape deviating from an equilateral triangle indicates significant shortcomings. For example, a low adaptive index indicates weak shock resistance, requiring strengthened pollution source control; a low self-repair index indicates impaired recovery mechanisms, requiring attention to the integrity of the ecosystem.

[0047] Therefore, the water ecological environment resilience index obtained by the above methods can be widely applied to scenarios such as differentiated watershed management, ecological compensation mechanism design, emergency pollution early warning response, and climate change adaptation planning. By quantitatively diagnosing the resilience shortcomings of different regions or time periods, it can support precise policy implementation and optimal resource allocation. With the integration of high-frequency monitoring networks and artificial intelligence technology, this index is expected to become a core indicator of smart water management systems, enabling real-time dynamic assessment and predictive management of water ecological health.

[0048] In summary, this invention replaces fixed standards with dynamic baselines, accurately identifies impact events and quantifies the three-dimensional toughness of resistance-recovery-stability, and solves the shortcomings of existing methods, such as static baselines, single dimensions, and insufficient mechanisms. The method is computationally efficient and has moderate data requirements, making it suitable for long-term historical assessments as well as for real-time online monitoring, combining scientific rigor with management practicality.

[0049] Correspondingly, such as Figure 4 As shown, the present invention also provides a water ecological environment resilience assessment system based on dynamic baselines to implement the above-mentioned water ecological environment resilience assessment method based on dynamic baselines, including: a historical data acquisition module 101, a dynamic baseline construction module 102, a resilience index calculation module 103, and a result output module 104. Historical data acquisition module 101 is used to acquire the time series sequence of the comprehensive water quality index of the target water body in historical periods; The dynamic baseline construction module 102 is used to extract the trend component, seasonal component and residual component from the time series of the comprehensive water quality index using a time series decomposition method; add the trend component and seasonal component point by point to obtain the dynamic baseline that changes over time; and then determine the confidence interval of the dynamic baseline based on the standard deviation of the residual component. The resilience index calculation module 103 is used to calculate the adaptive index, self-repair index, and self-purification index based on the numerical relationship between the observed water quality comprehensive index of the target water body and the dynamic baseline and confidence interval. The adaptive index is the deviation of the observed water quality comprehensive index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness of the observed water quality comprehensive index returning to the confidence interval after the impact event; and the self-purification index is the proportion of time during which the observed water quality comprehensive index is within the confidence interval. The impact event is defined as an event in which the observed water quality comprehensive index exceeds the confidence interval. The adaptive index, self-repair index, and self-purification index are standardized and then weighted and fused to obtain the aquatic ecological environment resilience index. The result output module 104 is used to classify the resilience level according to the aquatic ecological environment resilience index value and to visualize it in the form of a map, time series curve or radar chart.

Claims

1. A method for assessing the resilience of aquatic ecosystems based on dynamic baselines, characterized in that, Includes the following steps: S101. Establish a dynamic baseline; Obtain the time series of the comprehensive water quality index of the target water body in historical periods; The trend component, seasonal component, and residual component of the water quality comprehensive index time series were extracted using the time series decomposition method. The trend component and the seasonal component are aligned and superimposed according to time points to generate a dynamic baseline that changes over time. The confidence interval of the dynamic baseline is determined based on the standard deviation of the residual components; S102. Calculate the water ecological environment resilience index; Based on the numerical relationship between the observed comprehensive water quality index of the target water body and the dynamic baseline and confidence interval, the adaptive index, self-repair index, and self-purification index are calculated respectively: The adaptive index is the deviation of the observed water quality index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness with which the observed water quality index returns to the confidence interval after the impact event ends; the self-purification index is the proportion of time during which the observed water quality index is within the confidence interval; wherein, the impact event is an event in which the observed water quality index exceeds the confidence interval. The adaptive index, self-repair index, and self-purification index are standardized and then weighted and fused to obtain the aquatic ecological environment resilience index, which is used to indicate the health status and anti-interference ability of the target water body.

2. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The water quality comprehensive index time series is a continuous numerical sequence obtained by performing dimensionless processing and weighted summation on multi-index water quality monitoring data arranged at fixed time intervals. The multi-indicator water quality monitoring data includes the concentration values ​​of pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen, and total phosphorus.

3. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The time series decomposition method is one of STL decomposition, X-12-ARIMA decomposition, moving average decomposition, or empirical mode decomposition.

4. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The method for determining the confidence interval of the dynamic baseline based on the standard deviation of the residual components includes: Take the standard deviation of the residual components; The standard deviation of the dynamic baseline ± 1.96 times is used as the upper and lower limits of the confidence interval.

5. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The formula for calculating the adaptive index is as follows: （1） In equation (1), For adaptive exponents, w i For the first i The proportion of days the impact event lasted out of the total number of days. di For the first i The average relative deviation of each impact event.

6. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The formula for calculating the self-healing index is as follows: （2） In equation (2), The self-repair index, The duration of the impact event, This is the time required from the end of the shock event to recovery to within the dynamic baseline confidence interval. To restore the observed comprehensive water quality index values ​​after they have stabilized, This is the value of the dynamic baseline.

7. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The formula for calculating the self-purification index is as follows: （3） In equation (3), The self-purification index This refers to the number of time points during which the observed comprehensive water quality index values ​​fall within the confidence interval of the dynamic baseline over the entire time period. This represents the total number of time points within the entire time period.

8. The method for assessing the resilience of aquatic ecosystems based on dynamic baselines as described in claim 1, characterized in that, The formula for calculating the aquatic ecological environment resilience index is as follows: （4） In equation (4), As an index of aquatic ecological environment resilience, For adaptive exponents, The self-repair index, The self-purification index; , , All are weighting coefficients, and .

9. A dynamic baseline-based aquatic ecological environment resilience assessment system, used to implement the dynamic baseline-based aquatic ecological environment resilience assessment method according to any one of claims 1 to 8, characterized in that, include: The system includes a historical data acquisition module, a dynamic baseline construction module, a resilience index calculation module, and a result output module. The historical data acquisition module is used to acquire the time series sequence of the comprehensive water quality index of the target water body in historical periods; The dynamic baseline construction module is used to extract the trend component, seasonal component and residual component from the time series of the comprehensive water quality index using a time series decomposition method. The trend component and the seasonal component are added together point by point to obtain the dynamic baseline that changes over time. The confidence interval of the dynamic baseline is then determined based on the standard deviation of the residual components. The resilience index calculation module is used to calculate the adaptive index, self-repair index, and self-purification index based on the numerical relationship between the observed comprehensive water quality index of the target water body and the dynamic baseline and confidence interval. The adaptive index is the deviation of the observed comprehensive water quality index from the dynamic baseline during the impact event; the self-repair index is the rate and completeness of the observed comprehensive water quality index returning to the confidence interval after the impact event; and the self-purification index is the proportion of time during which the observed comprehensive water quality index remains within the confidence interval. The impact event is defined as an event where the observed comprehensive water quality index exceeds the confidence interval. The adaptive index, self-repair index, and self-purification index are standardized and then weighted and fused to obtain the aquatic ecological environment resilience index. The results output module is used to classify resilience levels based on the aquatic ecological environment resilience index values ​​and to visualize them in the form of maps, time-series curves, or radar charts.

Images