A drought period drinking water source water quantity attenuation and water quality deterioration linkage early warning method

Abstract

The application provides a kind of drought period drinking water source place water quantity attenuation and water quality deterioration linkage early warning method, including obtaining the multi-source monitoring data and external environment data of target drinking water source place;According to the multi-source monitoring data, water quantity attenuation characteristic parameter and water quality deterioration characteristic parameter are generated;According to water quantity attenuation characteristic parameter and water quality deterioration characteristic parameter, water quantity-water quality joint distribution function is constructed;According to water quantity-water quality joint distribution function and external environment data, water quantity-water quality joint risk probability in preset period is generated;According to water quantity-water quality joint risk probability, corresponding early warning information is generated and output.The application can realize the linkage early warning of drought period drinking water source place water quantity attenuation and water quality deterioration, improve the comprehensiveness and accuracy of early warning.

Description

Technical Field

[0001] This invention relates to the field of water resource security early warning technology, and more specifically, to a method for linking the decline in water volume and the deterioration of water quality in drinking water sources during drought periods. Background Technology

[0002] Water quantity decline and water quality deterioration in drinking water sources during droughts are two core issues affecting water supply security, and they are often interconnected and occur simultaneously. Water quantity decline is mainly manifested in lower water levels and reduced flow, leading to a decrease in the water body's self-purification capacity and an increase in pollutant concentrations; water quality deterioration is manifested in lower dissolved oxygen and excessive levels of pollutants such as ammonia nitrogen and total phosphorus, further exacerbating the degradation of aquatic ecosystems. In recent years, with the increasing frequency of extreme drought events caused by climate change, drinking water sources are facing increasingly severe dual pressures of water quantity and water quality, placing higher demands on early warning systems for water source safety.

[0003] Currently, early warning technologies for drinking water sources mainly fall into two categories: water quantity early warning technology, which assesses water quantity risk by monitoring hydrological data such as water level and flow rate, combined with historical drought events; and water quality early warning technology, which assesses and warns of water quality using single-factor evaluation or comprehensive index methods by monitoring water quality indicators such as dissolved oxygen, ammonia nitrogen, and total phosphorus. However, existing technologies typically treat water quantity and water quality as independent systems for separate monitoring and early warning, lacking a quantitative characterization of their interrelationship. Specifically, existing water quantity early warning methods do not fully consider the feedback effect of water quality changes on water quantity decline, and existing water quality early warning methods do not incorporate water quantity conditions as a driving factor for water quality deterioration into their analytical framework. Furthermore, existing technologies often rely on a single monitoring data source, failing to effectively integrate multi-source information such as real-time online monitoring data, historical hydrological data, and meteorological forecast data, resulting in insufficient timeliness and predictability of early warning results. In terms of probabilistic risk assessment, existing methods mostly use univariate statistical models to describe the risk probability of water quantity or water quality, making it difficult to characterize the joint risk characteristics of simultaneous water quantity decline and water quality deterioration. Summary of the Invention

[0004] This invention provides a method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, comprising the following steps: acquiring multi-source monitoring data and external environmental data of a target drinking water source; generating water quantity decline characteristic parameters and water quality deterioration characteristic parameters of the target drinking water source based on the multi-source monitoring data; constructing a water quantity-water quality joint distribution function of the target drinking water source based on the water quantity-water quality joint distribution function and the external environmental data; generating a water quantity-water quality joint risk probability of the target drinking water source within a preset time period based on the water quantity-water quality joint risk probability; and generating and outputting early warning information corresponding to the water quantity-water quality joint risk probability based on the water quantity-water quality joint risk probability.

[0005] Furthermore, the acquisition of multi-source monitoring data and external environmental data of the target drinking water source includes: acquiring real-time online monitoring data of the target drinking water source, including water level data, flow rate data, dissolved oxygen data, ammonia nitrogen data, and total phosphorus data; acquiring historical hydrological data of the target drinking water source, including historical daily water level data, historical daily flow rate data, historical daily water quality data, and historical drought event records; and acquiring meteorological forecast data of the target drinking water source, including daily precipitation forecast data and daily temperature forecast data within a preset time period.

[0006] Further, generating water volume decay characteristic parameters and water quality deterioration characteristic parameters of the target drinking water source based on the multi-source monitoring data includes: generating a current inflow sequence of the target drinking water source based on the water level data and the flow rate data; generating a historical average inflow of the target drinking water source based on the historical daily water level data and the historical daily flow rate data in the historical hydrological data; generating a water volume decay rate sequence of the target drinking water source based on the current inflow sequence and the historical average inflow, where each element in the water volume decay rate sequence is a ratio calculated based on the current inflow at the corresponding time in the current inflow sequence and the historical average inflow; and generating a water volume decay rate sequence based on the dissolved oxygen... The data, ammonia nitrogen data, and total phosphorus data are used to generate a current comprehensive water quality index sequence for the target drinking water source; based on the historical daily water quality data in the historical hydrological data, a historical average comprehensive water quality index for the same period of the target drinking water source is generated; based on the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period of the target drinking water source, a water quality deterioration rate sequence is generated, where each element in the water quality deterioration rate sequence is the difference calculated based on the current comprehensive water quality index at the corresponding time in the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period of the target drinking water source; the water volume decay rate sequence is used as the water volume decay characteristic parameter, and the water quality deterioration rate sequence is used as the water quality deterioration characteristic parameter.

[0007] Further, the step of constructing the water quantity-water quality joint distribution function of the target drinking water source based on the water quantity decay characteristic parameters and the water quality deterioration characteristic parameters includes: fitting a water quantity decay marginal distribution function of the target drinking water source based on the water quantity decay rate sequence; fitting a water quality deterioration marginal distribution function of the target drinking water source based on the water quality deterioration rate sequence; calculating the rank correlation coefficient between the water quantity decay rate sequence and the water quality deterioration rate sequence based on the water quantity decay rate sequence and the water quality deterioration rate sequence; and constructing the water quantity-water quality joint distribution function using the Copula function based on the water quantity decay marginal distribution function, the water quality deterioration marginal distribution function, and the rank correlation coefficient.

[0008] Further, the step of constructing the water quantity-water quality joint distribution function using a Copula function based on the water quantity decay marginal distribution function, the water quality deterioration marginal distribution function, and the rank correlation coefficient includes: selecting a target Copula function type from a preset family of Copula functions based on the rank correlation coefficient; constructing an initial joint distribution function using the target Copula function type based on the water quantity decay marginal distribution function and the water quality deterioration marginal distribution function; estimating the parameters in the initial joint distribution function using the maximum likelihood estimation method to obtain estimated parameters; and substituting the estimated parameters into the initial joint distribution function to generate the water quantity-water quality joint distribution function.

[0009] Further, generating the joint risk probability of water quantity and water quality for the target drinking water source within a preset time period based on the joint water quantity-water quality distribution function and the external environmental data includes: generating a predicted drought index sequence for the preset time period based on the precipitation forecast data and the temperature forecast data in the external environmental data; generating a predicted water quantity decay rate sequence and a predicted water quality deterioration rate sequence for the preset time period based on the predicted drought index sequence; substituting each predicted water quantity decay rate in the predicted water quantity decay rate sequence and each predicted water quality deterioration rate in the predicted water quality deterioration rate sequence into the joint water quantity-water quality distribution function to generate a joint probability density sequence for the preset time period; and generating the joint risk probability of water quantity and water quality for the target drinking water source within the preset time period based on the joint probability density sequence.

[0010] Further, the step of generating the predicted drought index sequence for the preset time period based on the precipitation forecast data and the temperature forecast data in the external environmental data includes: generating a daily standardized precipitation index sequence for the preset time period based on the precipitation forecast data; generating a daily standardized temperature index sequence for the preset time period based on the temperature forecast data; and generating the predicted drought index sequence for the preset time period using a preset drought index calculation formula based on the daily standardized precipitation index sequence and the daily standardized temperature index sequence.

[0011] Further, generating the predicted water volume decay rate sequence and the predicted water quality deterioration rate sequence within the preset time period based on the predicted drought index sequence includes: establishing a water volume decay prediction regression model based on the predicted drought index sequence and the water volume decay rate sequence; inputting the predicted drought index sequence into the water volume decay prediction regression model to generate the predicted water volume decay rate sequence; establishing a water quality deterioration prediction regression model based on the predicted drought index sequence and the water quality deterioration rate sequence; and inputting the predicted drought index sequence into the water quality deterioration prediction regression model to generate the predicted water quality deterioration rate sequence.

[0012] Furthermore, the step of generating and outputting early warning information corresponding to the combined risk probability of water quantity and water quality based on the combined risk probability of water quantity and water quality includes: matching the combined risk probability of water quantity and water quality with multiple preset risk probability threshold intervals; retrieving the early warning level text and emergency response measure text corresponding to the matching result from a preset early warning information database based on the matching result; combining the retrieved early warning level text and emergency response measure text to generate the early warning information; and outputting the early warning information to the management terminal.

[0013] Furthermore, the step of matching the combined risk probability of water quantity and water quality with multiple preset risk probability threshold intervals includes: when the combined risk probability of water quantity and water quality is in a first risk probability threshold interval, the matching result is a first risk level; when the combined risk probability of water quantity and water quality is in a second risk probability threshold interval, the matching result is a second risk level; when the combined risk probability of water quantity and water quality is in a third risk probability threshold interval, the matching result is a third risk level; when the combined risk probability of water quantity and water quality is in a fourth risk probability threshold interval, the matching result is a fourth risk level; and when the combined risk probability of water quantity and water quality is in a fifth risk probability threshold interval, the matching result is a fifth risk level.

[0014] The embodiments of the present invention have at least the following beneficial effects: 1. This invention acquires multi-source monitoring data, including real-time online monitoring data, historical hydrological data, and meteorological forecast data, as well as external environmental data, from the target drinking water source. Based on this, it generates water quantity decay characteristic parameters and water quality deterioration characteristic parameters. This solves the problem of single-source early warning data and failure to effectively integrate multi-source information in existing technologies. It achieves comprehensive perception and dynamic tracking of the water quantity and water quality status of drinking water sources, significantly improving the data support capability and information comprehensive utilization level of the early warning system.

[0015] 2. This invention constructs a joint distribution function of water quantity and water quality based on the characteristic parameters of water quantity decline and water quality deterioration. It then uses the Copula function in combination with the marginal distribution function of water quantity decline, the marginal distribution function of water quality deterioration, and the rank correlation coefficient between the two to characterize the interdependent structure between water quantity decline and water quality deterioration. This solves the problem in the prior art of treating water quantity and water quality separately and being unable to quantitatively describe their linkage relationship. It achieves accurate characterization of the joint risk of water quantity and water quality in drinking water sources during drought periods, enabling the early warning results to truly reflect the comprehensive safety situation where water quantity decline and water quality deterioration occur simultaneously.

[0016] 3. This invention generates the probability of combined water quantity and water quality risk within a preset time period based on the combined distribution function of water quantity and water quality and meteorological forecast data in the external environment data. Based on this, it matches the risk level and emergency response measures to output early warnings, which solves the problems of insufficient timeliness and lack of foresight in the existing technology. It realizes the advanced prediction and graded early warning of the combined risks of drinking water sources in the future period, provides timely and effective technical support for management decisions, and significantly enhances the initiative of drinking water source safety management and the scientific nature of emergency response during drought periods. Attached Figure Description

[0017] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of the invention are illustrated in the drawings by way of example and not limitation, wherein: Figure 1 This is a flowchart illustrating a method for early warning of the linkage between water quantity decline and water quality deterioration in drinking water sources during drought periods, provided in an embodiment of the present invention. Detailed Implementation

[0018] The principles and spirit of the invention will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement the invention, and are not intended to limit the scope of the invention in any way. Rather, these embodiments are provided to make the invention more thorough and complete, and to fully convey the scope of the invention to those skilled in the art.

[0019] Those skilled in the art will recognize that embodiments of the present invention can be implemented as a system, apparatus, device, method, or computer program product. Therefore, the present invention can be specifically implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0020] It should be noted that the number of any elements in the accompanying drawings is for illustrative purposes only and not as a limitation, and any naming is for distinction only and has no limiting meaning.

[0021] The following is for reference. Figure 1 , Figure 1 This is a schematic flowchart illustrating a method for linking water quantity decline and water quality deterioration in drinking water sources during drought periods, provided by an embodiment of the present invention. Figure 1 As shown, a method for linking water quantity decline and water quality deterioration in drinking water sources during drought periods includes: S1: Acquire multi-source monitoring data and external environmental data of the target drinking water source.

[0022] Target drinking water sources refer to reservoirs, lakes, or river sections that serve as urban and rural water supply areas, and their water quantity and quality directly affect water supply security. Multi-source monitoring data refers to various monitoring data related to the hydrology and water quality of the target drinking water source obtained from different data sources. External environmental data refers to meteorological forecast information related to the area where the target drinking water source is located. In practice, online water quality monitoring stations and water level / flow monitoring stations deployed at the target drinking water source collect real-time data on water level, flow rate, dissolved oxygen, ammonia nitrogen, and total phosphorus. Water level data is obtained using pressure level gauges or radar level gauges; flow rate data is obtained using acoustic Doppler current profilers or ultrasonic time-of-flight flow meters; dissolved oxygen data is obtained using electrochemical dissolved oxygen sensors; ammonia nitrogen data is obtained using an online ammonia-sensitive electrode method analyzer; and total phosphorus data is obtained using an online ammonium molybdate spectrophotometric analyzer.

[0023] The aforementioned sensors and monitoring equipment collect data at a preset sampling frequency, such as once per hour, and upload the collected data to the data center in real time via the data acquisition and transmission module. Historical hydrological data is obtained from hydrological yearbooks or national hydrological databases of the target drinking water source, including historical daily water level data, historical daily flow data, historical daily water quality data, and historical drought event records for at least the past 30 years. Meteorological forecast data is obtained from the National Meteorological Administration or commercial meteorological service providers, including daily precipitation forecasts and daily temperature forecasts for preset time periods, such as the next 7 days or the next 30 days. All acquired data is organized according to a unified time format and stored in a relational database for subsequent retrieval.

[0024] S2: Based on multi-source monitoring data, generate characteristic parameters of water quantity decline and water quality deterioration of the target drinking water source.

[0025] S21: Generate the current inflow sequence of the target drinking water source based on water level and flow data.

[0026] The current inflow sequence refers to the time series data consisting of the inflow flow values ​​of a target drinking water source at various times within the current time period. In practice, the collected hourly water level data is converted into hourly flow data using a water level-flow relationship curve. This curve is obtained by fitting historical measured water level and flow data using the least squares method. The converted hourly flow data is then arranged in chronological order to form the current inflow sequence. , where t represents the time index. , where n is the number of data points in the current time period. For reservoir-type drinking water sources, inflow refers to the inflow rate; for river-type drinking water sources, inflow refers to the measured flow rate at the monitoring section.

[0027] S22: Based on historical daily water level data and historical daily flow data in historical hydrological data, generate the historical average inflow volume of the target drinking water source for the same period.

[0028] The historical average inflow refers to the average inflow rate of a target drinking water source on the same calendar date as the current date throughout history. In practice, this involves extracting water level and flow data for all historical years that fall on the same month and date as the current date from historical hydrological data. For example, if the current date is July 1, 2026, then the water level and flow data for July 1st of each historical year are extracted. The extracted daily historical water level data is then converted into daily historical flow data using the water level-flow relationship curve described above. Finally, the arithmetic mean of the flow rates for the same date across all historical years is calculated to obtain the historical average inflow. This process can be represented as Where m is the number of historical years with complete records. This represents the daily flow rate for the same calendar date as the current date in the i-th historical year.

[0029] S23: Generate a water volume decay rate sequence for the target drinking water source based on the current water volume sequence and the historical average water volume for the same period.

[0030] A water volume decay rate sequence refers to time series data comprising the degree of decay of the current inflow relative to the historical average inflow for the same period. Each element in the water volume decay rate sequence is the ratio calculated based on the current inflow at the corresponding moment in the current inflow sequence to the historical average inflow for the same period. In specific implementation, for the current inflow sequence... For each time t, calculate the current inflow value and compare it with the historical average inflow value for the same period. The ratio of the two values ​​yields the water decay rate at that moment. The formula for calculating the water decay rate D(t) is as follows:

[0031] in, Let be the water volume decay rate at time t. This represents the historical average water inflow for the same period, expressed in cubic meters per second. Let be the current inflow rate at time t, expressed in cubic meters per second. Arrange the calculation results for all times in chronological order to form a water decay rate sequence. , .

[0032] S24: Based on dissolved oxygen, ammonia nitrogen, and total phosphorus data, generate a current comprehensive water quality index sequence for the target drinking water source.

[0033] The current comprehensive water quality index series refers to the time series data composed of the comprehensive water quality evaluation values ​​of the target drinking water source at various times within the current time period. In practice, for each data collection time t, the dissolved oxygen concentration at that time is obtained. ammonia nitrogen concentration and total phosphorus concentration The current comprehensive water quality index is calculated using the comprehensive water quality index method. This index is obtained by normalizing and weighting multiple water quality indicators. Current comprehensive water quality index. The calculation formula is shown below:

[0034] in, Let t be the current comprehensive water quality index at time t. Let be the dissolved oxygen concentration at time t, expressed in milligrams per liter. The standard limit for dissolved oxygen is 6 mg / L, based on the Class II water standard in the Surface Water Environmental Quality Standard GB 3838-2002. Let be the ammonia nitrogen concentration at time t, in milligrams per liter. The standard limit for ammonia nitrogen is 0.5 mg / L, based on the Class II water standard in the Surface Water Environmental Quality Standard GB 3838-2002. The total phosphorus concentration at time t is expressed in milligrams per liter. The standard limit for total phosphorus is 0.1 mg / L, based on the Class II water standard in the Surface Water Environmental Quality Standard GB 3838-2002. and These are the weighting coefficients for dissolved oxygen, ammonia nitrogen, and total phosphorus, respectively, with a sum of 1. Typical values ​​are 0.3, 0.35, and 0.35, respectively. The calculation results for all time points are arranged in chronological order to form the current comprehensive water quality index sequence. , t=1,2,…,n.

[0035] S25: Generate the historical average comprehensive water quality index for the target drinking water source based on historical daily water quality data from historical hydrological data.

[0036] The historical average comprehensive water quality index refers to the average comprehensive water quality evaluation value of the target drinking water source on the same calendar date in historical years. In practice, daily water quality data for all historical years with the same month and date as the current date are extracted from historical hydrological data, including historical daily dissolved oxygen, ammonia nitrogen, and total phosphorus data. For each historical year and date, the historical comprehensive water quality index is calculated using the same comprehensive water quality index calculation method as in S24. Then, the arithmetic mean of the comprehensive water quality indices for the same date in all historical years is calculated to obtain the historical average comprehensive water quality index. This process can be represented as Where m is the number of historical years with complete records. is the comprehensive water quality index for the same calendar date as the current date in the i-th historical year.

[0037] S26: Generate a water quality deterioration rate sequence for the target drinking water source based on the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period.

[0038] A water quality deterioration rate sequence refers to time-series data comprising the degree of deterioration of current water quality relative to the historical average for the same period. Each element in the water quality deterioration rate sequence is the difference between the current comprehensive water quality index at the corresponding moment in the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period. In specific implementation, for the current comprehensive water quality index sequence... For each time t, calculate the current comprehensive water quality index value and compare it with the historical average comprehensive water quality index for the same period. The difference yields the water quality deterioration rate at that moment. Water quality deterioration rate The calculation formula is shown below:

[0039] in, Let t be the water quality deterioration rate. Let t be the current comprehensive water quality index at time t. This represents the historical average comprehensive water quality index for the same period. The calculation results for all times are arranged chronologically to form a water quality deterioration rate sequence. , .

[0040] S27: Use the water volume decay rate sequence as the water volume decay characteristic parameter and the water quality deterioration rate sequence as the water quality deterioration characteristic parameter.

[0041] After completing the calculations for S23 and S26, the water decay rate sequence D(t) generated by S23 is used as the water decay characteristic parameter, and the water quality deterioration rate sequence generated by S26 is used as the water quality deterioration rate sequence. These parameters serve as characteristic parameters of water quality deterioration and are used in subsequent steps. Both the water volume decline characteristic parameter and the water quality deterioration characteristic parameter are time-related time series data, and they correspond one-to-one in the time dimension.

[0042] S3: Construct a joint distribution function of water quantity and water quality for the target drinking water source based on the characteristic parameters of water quantity decline and water quality deterioration.

[0043] S31: Based on the water decay rate sequence, the marginal distribution function of water decay in the target drinking water source is obtained by fitting.

[0044] The marginal distribution function refers to the probability distribution function of a single random variable. In practical implementation, the water decay rate sequence obtained from S23 is used. Treating the data as a random sample, the maximum likelihood estimation method is used to estimate the parameters of various candidate probability distribution functions, including normal, log-normal, gamma, and Weibull distributions. The likelihood function values ​​for each distribution are calculated, and the distribution with the highest likelihood function value is selected as the optimal marginal distribution for the water decay rate. Taking the gamma distribution as an example, the water decay rate sequence... Obeying shape parameters Scale parameters are The gamma distribution has a probability density function as shown in the following equation:

[0045] in, Let be the probability density value of the water decay rate at the value d, where d is the variable representing the water decay rate. The shape parameter of the gamma distribution is estimated from the water decay rate sequence through maximum likelihood estimation. The scale parameter of the gamma distribution is estimated from the water decay rate sequence through maximum likelihood estimation. The gamma function is defined as follows: Substituting the estimated parameters into the probability density function above, we obtain the marginal distribution function of water volume decline. , where P represents probability.

[0046] S32: Based on the water quality deterioration rate sequence, the marginal distribution function of water quality deterioration of the target drinking water source is obtained by fitting.

[0047] In practical implementation, the water quality deterioration rate sequence obtained from S26 will be used. Treating it as a random sample, the same method as in S31 is used to fit multiple candidate probability distribution functions using maximum likelihood estimation. The distribution type with the largest likelihood function value is selected as the optimal marginal distribution for the water quality deterioration rate. Taking the normal distribution as an example, the water quality deterioration rate sequence... Follow the mean Standard deviation is The probability density function of the normal distribution is shown in the following equation:

[0048] in, Let be the probability density value of the water quality deterioration rate at the value e, where e is the variable representing the water quality deterioration rate. The mean of the value is normally distributed and is estimated from the water quality deterioration rate sequence through maximum likelihood estimation. The standard deviation of the normal distribution is estimated from the water quality deterioration rate sequence through maximum likelihood estimation. Substituting the estimated parameters into the probability density function above yields the marginal distribution function of water quality deterioration. , where P represents probability.

[0049] S33: Calculate the rank correlation coefficient between the water volume decay rate sequence and the water quality deterioration rate sequence based on the water volume decay rate sequence and the water quality deterioration rate sequence.

[0050] The rank correlation coefficient is a nonparametric statistic that measures the degree of monotonic correlation between two variables. In practical implementation, it involves the water decay rate sequence... and water quality deterioration rate sequence Pair them according to time order to form n sets of two-dimensional data points. Kendall's rank correlation coefficient was used to measure the correlation between the two. First, the water decay rate sequence was analyzed. Sort all values ​​in ascending order and assign a rank to each value; for the water quality deterioration rate sequence Sort all values ​​in ascending order and assign a rank to each value. Then compare all data point pairs and count the number of identical pairs. The number of inconsistent pairs The formula for calculating the Kendall rank correlation coefficient τ is shown below:

[0051] in, Kendall's rank correlation coefficient. The number of consistent pairs, i.e., for any two data points i and j, and ,or and Number of data point pairs The number of inconsistent pairs, i.e., for any two data points i and j, and ,or and The number of data point pairs, where n is the total number of data points. (Calculated...) The value reflects the degree of positive or negative correlation between the water volume decline rate and the water quality deterioration rate.

[0052] S34: Based on the marginal distribution function of water quantity decline, the marginal distribution function of water quality deterioration, and the rank correlation coefficient, the Copula function is used to construct the joint distribution function of water quantity and water quality.

[0053] Copula functions are mathematical tools used to connect the marginal distribution functions of multiple random variables to form a joint distribution function. In practice, based on the Kendall rank correlation coefficient τ calculated by S33, a suitable Copula function type is selected from a pre-defined family of Copula functions. Commonly used Copula functions include Gaussian Copula, Clayton Copula, Frank Copula, and Gumbel Copula. The selection principle is: when τ is greater than 0, Gumbel Copula or Clayton Copula are preferred; when τ is less than 0, Frank Copula is preferred; when τ is close to 0, Gaussian Copula is preferred. Taking Gumbel Copula as an example, its two-dimensional Copula function expression is shown below:

[0054] in, Let be the Copula function value, representing the joint probability that the water decay rate does not exceed u and the water quality deterioration rate does not exceed v, where u is the marginal distribution function value of water decay. The value ranges from 0 to 1, where v is the marginal distribution function value of water quality deterioration. The value ranges from 0 to 1. For the relevant parameters of Gumbel Copula, ≥1, Kendall rank correlation coefficient obtained through S33 The calculation is obtained, and the calculation formula is: The water volume decay marginal distribution function obtained from S31. Substituting the water quality deterioration marginal distribution function F_E obtained from S32 into the Copula function above, we obtain the water quantity-water quality joint distribution function. This function can fully describe the joint probability distribution characteristics between the water volume decay rate and the water quality deterioration rate.

[0055] S4: Based on the water quantity-water quality joint distribution function and external environmental data, generate the water quantity-water quality joint risk probability of the target drinking water source within a preset time period.

[0056] S41: Generate a predicted drought index sequence for a preset time period based on precipitation and temperature forecast data from external environmental data.

[0057] The drought index forecast series refers to time series data composed of index values ​​reflecting the future degree of drought, calculated based on meteorological forecast data. In practice, the preset time period is the next 30 days, and the daily precipitation forecast data and daily temperature forecast data are obtained from the National Meteorological Administration, with a time resolution of 1 day.

[0058] First, a daily standardized precipitation index (SPI) sequence for a preset time period is generated based on precipitation forecast data. The Standardized Precipitation Index (SPI) is a drought index obtained by standardizing the precipitation sequence through a probability distribution. Specifically, for each day j within the preset time period, the predicted precipitation value for that day is obtained. The predicted precipitation value for that day. By comparing the historical precipitation data for the same day with long-term precipitation data from the same period in history, and fitting the historical precipitation series using a gamma distribution, the shape parameters are obtained. and scale parameters Then calculate the cumulative probability of the daily precipitation forecast. Cumulative probability The calculation formula is shown below:

[0059] in, This represents the cumulative probability that precipitation will not exceed P, where P is the precipitation variable in millimeters. The shape parameter of the gamma distribution is obtained by maximum likelihood estimation of the historical daily precipitation series for the same period. The scale parameter of the gamma distribution is obtained by maximum likelihood estimation of the historical daily precipitation series for the same period. The function is the gamma function. Then, the cumulative probability is... Perform inverse standard normal distribution transformation to obtain the standardized precipitation index. ,in This is the inverse function of the cumulative distribution function of the standard normal distribution. (The last part, "all days," appears to be a fragment and doesn't translate directly.) Arranged in chronological order, a daily standardized precipitation index sequence is obtained. m is the number of days within the preset time period.

[0060] Next, a daily standardized temperature index (STI) sequence is generated based on the temperature forecast data for a preset time period. The STI is an index obtained by standardizing the temperature sequence. For each day j within the preset time period, the predicted temperature value for that day is obtained. Calculate the historical average temperature for that day. Standard deviation of temperature on the same day in history Standardized temperature index The calculation formula is shown below:

[0061] in, The standardized temperature index for day j. This is the predicted temperature value for day j, in degrees Celsius. This represents the historical average temperature for that day, expressed in degrees Celsius, and is derived from historical hydrological data. This represents the standard deviation of the temperature for the same period in history, expressed in degrees Celsius, and is obtained from historical hydrological data. All days... Arranged chronologically, a daily standardized temperature index sequence is obtained. , .

[0062] Then, based on the daily standardized precipitation index series and the daily standardized temperature index series, a predicted drought index series for a preset time period is generated using a pre-defined drought index calculation formula. The drought index comprehensively reflects the combined effects of precipitation and temperature on drought. Predicted Drought Index The calculation formula is shown below:

[0063] in, The predicted drought index for day j. The standardized precipitation index for day j. The standardized temperature index for day j. This is the weighting coefficient for the standardized precipitation index, typically taken as 0.6. This is the weighting factor for the standardized temperature index, typically taken as 0.4. and The sum of all days is 1. Arranged chronologically, the predicted drought index sequence is obtained. , .

[0064] S42: Based on the predicted drought index sequence, generate the predicted water volume decline rate sequence and the predicted water quality deterioration rate sequence for the preset time period.

[0065] In practical implementation, a water decline prediction regression model is first established based on the predicted drought index sequence and the water decline rate sequence obtained from S23. The predicted drought index from historical data is used as the independent variable, and the corresponding water decline rate is used as the dependent variable. Linear regression or support vector regression methods are employed to establish the mapping relationship between the two. Taking linear regression as an example, the expression for the water decline prediction regression model is shown below:

[0066] in, DI is used to predict the rate of water depletion and is a drought index. The regression coefficients are obtained by fitting historical data using the least squares method. The regression intercept is obtained by fitting historical data using the least squares method. The predicted drought index sequence generated by S41 is then used. Substituting each value into the above water decay prediction regression model generates a predicted water decay rate sequence. , .

[0067] Secondly, based on the predicted drought index sequence and the water quality deterioration rate sequence obtained from S26, a water quality deterioration prediction regression model is established. The predicted drought index from historical data is used as the independent variable, and the corresponding water quality deterioration rate is used as the dependent variable. Linear regression or support vector regression methods are employed to establish the mapping relationship between the two. Taking linear regression as an example, the expression of the water quality deterioration prediction regression model is shown below:

[0068] in, DI is used to predict the rate of water quality deterioration, and it is a predictive drought index. The regression coefficients are obtained by fitting historical data using the least squares method. The regression intercept is obtained by fitting historical data using the least squares method. The predicted drought index sequence generated by S41 is then used. Substituting each value into the above water quality deterioration prediction regression model generates a predicted water quality deterioration rate sequence. .

[0069] S43: Substitute each predicted water volume decay rate in the predicted water volume decay rate sequence and each predicted water quality deterioration rate in the predicted water quality deterioration rate sequence into the water volume-water quality joint distribution function to generate a joint probability density sequence within a preset time period.

[0070] In practical implementation, for each day j within the preset time period, the predicted water volume decay rate generated by S42 will be used. and predicting water quality deterioration rate Substitute the water quantity-water quality joint distribution function constructed by S34 into the inputs respectively. First calculate Corresponding water volume decay marginal distribution function value ,calculate Corresponding water quality deterioration marginal distribution function value Then calculate the joint distribution function value. This value indicates that, under the predicted conditions, the water volume decay rate does not exceed [a certain percentage]. And the water quality deterioration rate does not exceed The joint probability. Arranging the joint distribution function values ​​of all days in chronological order yields the joint probability density sequence. .

[0071] S44: Based on the joint probability density sequence, generate the joint risk probability of water quantity and water quality of the target drinking water source within a preset time period.

[0072] The combined risk probability of water quantity and water quality refers to the comprehensive risk level of a target drinking water source simultaneously experiencing water quantity decline and water quality deterioration within a preset time period. In specific implementation, the combined probability density sequence generated by S43... Statistical analysis was performed. Preset risk thresholds included water degradation rate thresholds. and water quality deterioration rate threshold , The typical value is 0.2. The typical value is 0.3. For each day j, determine... Is it greater than And is E_pred(j) greater than If both conditions are met simultaneously, the day is marked as a risk day. The formula for calculating the combined risk probability R of water quantity and water quality is as follows:

[0073] Where R represents the combined risk probability of water quantity and water quality. The number of days marked as risk within the preset time period is denoted by , and m is the total number of days in the preset time period. The calculated R value is the combined probability of water quantity and water quality risk for the target drinking water source within the preset time period.

[0074] S5: Generate and output early warning information corresponding to the combined risk probability of water quantity and water quality based on the combined risk probability of water quantity and water quality.

[0075] S51: Match the combined risk probability of water quantity and water quality with multiple preset risk probability threshold ranges.

[0076] In practice, five risk probability threshold ranges are preset, each corresponding to one of five risk levels. The first risk probability threshold range is [0, 0.1), corresponding to the first risk level, i.e., low risk; the second risk probability threshold range is [0.1, 0.3), corresponding to the second risk level, i.e., relatively low risk; the third risk probability threshold range is [0.3, 0.5), corresponding to the third risk level, i.e., medium risk; the fourth risk probability threshold range is [0.5, 0.7), corresponding to the fourth risk level, i.e., relatively high risk; and the fifth risk probability threshold range is [0.7, 1.0], corresponding to the fifth risk level, i.e., high risk. The water quantity-water quality joint risk probability R calculated by S44 is compared with the above five intervals in turn: when R is in [0, 0.1), the matching result is the first risk level; when R is in [0.1, 0.3), the matching result is the second risk level; when R is in [0.3, 0.5), the matching result is the third risk level; when R is in [0.5, 0.7), the matching result is the fourth risk level; when R is in [0.7, 1.0], the matching result is the fifth risk level.

[0077] S52: Based on the matching results, retrieve the warning level text and emergency response measure text corresponding to the matching results from the preset warning information database.

[0078] In practice, the early warning information database is pre-stored and contains the early warning level text and emergency response measure text corresponding to each of the five risk levels. The first risk level corresponds to a blue warning, and the corresponding first emergency response measure is to strengthen daily monitoring and pay attention to trends in water quantity and quality. The second risk level corresponds to a yellow warning, and the corresponding second emergency response measure is to initiate routine patrols and increase monitoring frequency to once daily. The third risk level corresponds to an orange warning, and the corresponding third emergency response measure is to initiate emergency consultations, increase monitoring frequency to once every 6 hours, and prepare to activate backup water sources. The fourth risk level corresponds to a red warning, and the corresponding fourth emergency response measure is to initiate emergency water supply scheduling, increase monitoring frequency to once per hour, and activate backup water sources. The fifth risk level corresponds to a brown warning, and the corresponding fifth emergency response measure is to activate the emergency water supply plan, implement unified water resource scheduling, and issue public water conservation notices. Based on the matching results of S51, the corresponding early warning level text and emergency response measure text are retrieved from the early warning information database.

[0079] S53: Combine the retrieved warning level text and emergency response measures text to generate warning information.

[0080] In practice, the warning level text and emergency response measures text retrieved by S52 are combined according to a preset template. The template for generating warning information is: warning level + target drinking water source name + current date + warning result of drought-related drinking water source water volume decline and water quality deterioration + warning level text + recommended measures + emergency response measures text. For example, when the matching result is the third risk level, the generated warning information is an orange warning. The warning result for the drought-related drinking water source water volume decline and water quality deterioration of a certain reservoir on July 1, 2026 is an orange warning. The recommended measures are to initiate an emergency consultation, increase the monitoring frequency to once every 6 hours, and prepare to activate the backup water source.

[0081] S54: Output early warning information to the management terminal.

[0082] In practice, the management terminals include the workstation computers of the water management department, the mobile phones of management personnel, and a large-screen display system. Warning information is output in parallel through multiple methods: pushing warning information to the monitoring platform software interface of the water management department's workstation computers via a Hypertext Transfer Protocol (HTTP) interface; sending warning information to management personnel's mobile phones via an SMS gateway; and pushing warning information to the large-screen display system for scrolling display via a message queue telemetry transmission protocol. All output methods record the sending time and receiving status of the warning information to ensure accurate delivery.

[0083] Furthermore, the storage medium in the embodiments of this application stores program instructions capable of implementing all the above methods. These program instructions can be stored in the storage medium in the form of a software product, including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0084] The above description is merely an explanation of some preferred embodiments of the present invention and the technical principles employed. Those skilled in the art should understand that the scope of the invention as described in the embodiments of the present invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention.

Claims

1. A method for early warning of the linkage between water quantity decline and water quality deterioration in drinking water sources during drought periods, characterized in that, Includes the following steps: Acquire multi-source monitoring data and external environmental data of the target drinking water source; generate water quantity decline characteristic parameters and water quality deterioration characteristic parameters of the target drinking water source based on the multi-source monitoring data; Based on the water quantity decay characteristic parameters and the water quality deterioration characteristic parameters, a joint water quantity-water quality distribution function for the target drinking water source is constructed. Based on the water quantity-water quality joint distribution function and the external environmental data, the water quantity-water quality joint risk probability of the target drinking water source within a preset time period is generated. Based on the combined risk probability of water quantity and water quality, generate and output early warning information corresponding to the combined risk probability of water quantity and water quality.

2. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 1, is characterized in that... The acquisition of multi-source monitoring data and external environmental data of the target drinking water source includes: The real-time online monitoring data of the target drinking water source is obtained, including water level data, flow rate data, dissolved oxygen data, ammonia nitrogen data, and total phosphorus data. Obtain historical hydrological data of the target drinking water source, including historical daily water level data, historical daily flow data, historical daily water quality data, and historical drought event records. Obtain meteorological forecast data for the target drinking water source, including daily precipitation forecast data and daily temperature forecast data for a preset time period.

3. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 2, is characterized in that... The step of generating water quantity decline characteristic parameters and water quality deterioration characteristic parameters for the target drinking water source based on the multi-source monitoring data includes: Based on the water level data and the flow rate data, a current inflow sequence for the target drinking water source is generated; Based on the historical daily water level data and the historical daily flow data in the historical hydrological data, the historical average inflow of the target drinking water source for the same period is generated. Based on the current inflow sequence and the historical average inflow for the same period, a water volume decay rate sequence for the target drinking water source is generated. Each element in the water volume decay rate sequence is a ratio calculated based on the current inflow at the corresponding time in the current inflow sequence and the historical average inflow for the same period. Based on the dissolved oxygen data, ammonia nitrogen data, and total phosphorus data, a current comprehensive water quality index sequence for the target drinking water source is generated; Based on the historical daily water quality data in the historical hydrological data, the historical average comprehensive water quality index of the target drinking water source for the same period is generated. Based on the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period, a water quality deterioration rate sequence for the target drinking water source is generated. Each element in the water quality deterioration rate sequence is the difference between the current comprehensive water quality index at the corresponding time in the current comprehensive water quality index sequence and the historical average comprehensive water quality index for the same period. The water volume decay rate sequence is used as the water volume decay characteristic parameter, and the water quality deterioration rate sequence is used as the water quality deterioration characteristic parameter.

4. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 3, is characterized in that... The step of constructing a joint water quantity-water quality distribution function for the target drinking water source based on the water quantity decay characteristic parameters and the water quality deterioration characteristic parameters includes: Based on the water decay rate sequence, the water decay marginal distribution function of the target drinking water source is obtained by fitting. Based on the water quality deterioration rate sequence, the water quality deterioration marginal distribution function of the target drinking water source is fitted. Calculate the rank correlation coefficient between the water volume decay rate sequence and the water quality deterioration rate sequence based on the water volume decay rate sequence and the water quality deterioration rate sequence; Based on the water quantity decay edge distribution function, the water quality deterioration edge distribution function, and the rank correlation coefficient, the Copula function is used to construct the water quantity-water quality joint distribution function.

5. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 4, is characterized in that... The step of constructing the joint water quantity-water quality distribution function using the Copula function based on the water quantity decline edge distribution function, the water quality deterioration edge distribution function, and the rank correlation coefficient includes: Based on the rank correlation coefficient, a target Copula function type is selected from a preset family of Copula functions; Based on the water volume decay edge distribution function and the water quality deterioration edge distribution function, an initial joint distribution function is constructed using the target Copula function type; The parameters in the initial joint distribution function are estimated using the maximum likelihood estimation method to obtain the estimated parameters; Substitute the estimated parameters into the initial joint distribution function to generate the water quantity-water quality joint distribution function.

6. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 1, is characterized in that... The step of generating the combined water quantity and water quality risk probability of the target drinking water source within a preset time period based on the water quantity-water quality joint distribution function and the external environmental data includes: Based on the precipitation forecast data and temperature forecast data in the external environmental data, a predicted drought index sequence for the preset time period is generated; Based on the predicted drought index sequence, generate a predicted water volume decay rate sequence and a predicted water quality deterioration rate sequence for the preset time period; Substitute each predicted water volume decay rate in the predicted water volume decay rate sequence and each predicted water quality deterioration rate in the predicted water quality deterioration rate sequence into the water volume-water quality joint distribution function to generate the joint probability density sequence within the preset time period. Based on the joint probability density sequence, the water quantity-water quality joint risk probability of the target drinking water source during the preset time period is generated.

7. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 6, is characterized in that... The step of generating a predicted drought index sequence for the preset time period based on the precipitation forecast data and the temperature forecast data in the external environmental data includes: Based on the precipitation forecast data, a daily standardized precipitation index sequence is generated for the preset time period; Based on the temperature forecast data, a daily standardized temperature index sequence for the preset time period is generated; Based on the daily standardized precipitation index sequence and the daily standardized temperature index sequence, the predicted drought index sequence for the preset time period is generated using a preset drought index calculation formula.

8. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 6, is characterized in that... The step of generating a predicted water volume decline rate sequence and a predicted water quality deterioration rate sequence for the preset time period based on the predicted drought index sequence includes: Based on the predicted drought index sequence and the water decay rate sequence, a water decay prediction regression model is established. The predicted drought index sequence is input into the water volume decline prediction regression model to generate the predicted water volume decline rate sequence. Based on the predicted drought index sequence and the water quality deterioration rate sequence, a water quality deterioration prediction regression model is established; The predicted drought index sequence is input into the water quality deterioration prediction regression model to generate the predicted water quality deterioration rate sequence.

9. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 1, is characterized in that... The step of generating and outputting early warning information corresponding to the combined water quantity and water quality risk probability based on the combined water quantity and water quality risk probability includes: The combined risk probability of water quantity and water quality is matched with multiple preset risk probability threshold ranges; Based on the matching results, retrieve the warning level text and emergency response measure text corresponding to the matching results from the preset warning information database; The retrieved warning level text and emergency response measure text are combined to generate the warning information; the warning information is then output to the management terminal.

10. The method for early warning of water quantity decline and water quality deterioration in drinking water sources during drought periods, as described in claim 9, is characterized in that... The step of matching the combined risk probability of water quantity and water quality with multiple preset risk probability threshold intervals includes: When the combined risk probability of water quantity and water quality is within the first risk probability threshold range, the matching result is the first risk level; When the combined risk probability of water quantity and water quality is within the second risk probability threshold range, the matching result is the second risk level; When the combined risk probability of water quantity and water quality is within the third risk probability threshold range, the matching result is the third risk level; When the combined risk probability of water quantity and water quality is within the fourth risk probability threshold range, the matching result is the fourth risk level; When the combined risk probability of water quantity and water quality is within the fifth risk probability threshold range, the matching result is the fifth risk level.

Images