Dynamic early warning method for underground water

Abstract

The invention relates to a dynamic early warning method for underground water. The method comprises monitoring early warning and prediction early warning. The monitoring early warning comprises water level descending type early warning and water level ascending type early warning; the water level decline type early warning comprises overdraft area underground water level early warning and water level and water quantity double-control early warning, the overdraft area underground water level early warning adopts a five-level early warning system, and the water level and water quantity double-control early warning adopts a flat water year index; the water level rising type early warning comprises urban area water level rising type early warning and agricultural irrigation area water level rising type early warning; and a time sequence model, a multiple linear regression model and a neural network model are adopted for prediction and early warning to determine an optimal prediction model of each city. Based on underground water real-time monitoring, double-control index dynamic early warning is realized, and a technical basis is provided for dynamic supervision of drainage basin underground water and continuous utilization of water resources.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic water level control and prediction, and in particular to a groundwater dynamic early warning method. Background Art

[0002] Groundwater monitoring is for groundwater monitoring management departments to monitor groundwater levels, water quality and other data within their jurisdictions, so as to timely grasp dynamic changes and provide long-term protection for groundwater. Sometimes it is difficult to effectively and timely warn during groundwater monitoring, or the warning method cannot effectively reflect the changes in characteristic water points, and it is impossible to timely grasp the changes in water resources, which is not conducive to timely prevention and control, and may affect people's production and life. Summary of the invention

[0003] The present invention aims to solve the deficiencies of the prior art and provide a groundwater dynamic early warning method.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] A groundwater dynamic early warning method, including monitoring early warning and prediction early warning;

[0006] Monitoring and early warning include water level drop early warning and water level rise early warning; water level drop early warning includes groundwater level early warning in over-exploited areas and water level and water volume dual control early warning. Groundwater level early warning in over-exploited areas adopts a five-level early warning system, and water level and water volume dual control early warning adopts normal water year indicators; water level rise early warning includes urban area water level rise early warning and agricultural irrigation area water level rise early warning;

[0007] Forecasting and early warning use time series models, multivariate linear regression models, and neural network models to determine the optimal forecasting model for each city.

[0008] The operation of groundwater level early warning in over-exploited areas is as follows: select the station network, determine the calculation method and benchmark value, compare and analyze the groundwater level depth monitoring data with the benchmark value, dynamically evaluate the monthly average depth change, and adopt a five-level early warning system. The early warning results are assigned five colors: red, orange, yellow, blue, and green. Warnings are issued at three levels: red, orange, and yellow.

[0009] The benchmark values ​​are the maximum, minimum, average and same period last year's values ​​of the series burial depth.

[0010] The warning rules for groundwater level warning in over-exploitation areas are:

[0011] ① If the average water level of the month is less than or equal to the minimum water level of the same period in the previous five years, it is a red warning;

[0012] ② If the average water level of the month is more than 1.0 meter lower than the average water level of the same period in the previous five years and is higher than the minimum water level of the same period in the previous five years, it is an orange warning;

[0013] ③ If the average water level in the current month is more than 1.0 meter lower than the water level in the same period last year, or is less than the average water level of the same period by no more than 1.0 meter, it is a yellow warning;

[0014] ④ If the average water level of the month is greater than or equal to the average water level of the previous five years, and less than the maximum water level of the previous five years, it will be colored blue;

[0015] ⑤ If the average water level of the month is greater than or equal to the maximum water level of the same period in the previous five years, the color is green.

[0016] Water level and water volume dual control early warning:

[0017] The groundwater withdrawal control indicators are comprehensively analyzed by planning analysis and weight decomposition methods, and the groundwater withdrawal control indicators of provincial administrative regions are decomposed into basic units; the annual groundwater withdrawal control is that each province uses the planning analysis method, interpolation method, and trend analysis method to comprehensively analyze the groundwater withdrawal control indicators of each sub-district in its jurisdiction in 2020 and 2030, and determine the groundwater withdrawal control indicators of different levels;

[0018] The objects of regional groundwater level control index determination include shallow groundwater and deep confined water. The regional groundwater level control index determination includes the groundwater level control index of over-exploited areas and the groundwater level control index of non-over-exploited areas. The groundwater level control index determination of over-exploited areas includes two situations: normal water year and different flood and dry water conditions.

[0019] Water level rise warning:

[0020] In urban areas, the impact of rapid groundwater rise on underground projects is taken into consideration; in agricultural irrigation areas, the impact of high groundwater levels on land salinization is taken into consideration, and an early warning method for groundwater level rise is proposed.

[0021] The early warning classification standards for urban areas and agricultural irrigation areas are set at four levels, namely:

[0022] Level 1 warning: The average groundwater depth at the end of the month is ≤ the level 1 warning threshold;

[0023] Level 2 warning: Level 1 warning threshold < average groundwater depth at the end of the month ≤ Level 2 warning threshold;

[0024] Level 3 warning: Level 2 warning threshold < average groundwater depth at the end of the month ≤ Level 3 warning threshold;

[0025] Level 4 warning: Level 3 warning threshold < average groundwater depth at the end of the month ≤ Level 4 warning threshold;

[0026] No warning: Level 4 warning threshold < average groundwater depth at the end of the month.

[0027] The prediction and early warning adopts time series model, multivariate regression model and neural network model to establish a monthly prediction and forecasting scheme for the changes in groundwater level depth in each municipal administrative area, realize the monthly dynamic prediction of groundwater level depth in each municipal administrative area, and provide a basis for the management of groundwater resources in the basin.

[0028] The beneficial effects of the present invention are as follows: based on the analysis of the main controlling factors of groundwater levels in administrative divisions with different hydrogeological conditions, the present invention uses technical methods such as time series, multivariate regression, and neural network models to develop optimal prediction models suitable for different administrative regions, and realize the monthly dynamic prediction of the groundwater depth in over-exploited areas, which strongly supports the dynamic control and prediction of water levels in over-exploited areas; combined with the dual control management of water level and water volume and the requirements for groundwater over-exploitation control, a groundwater level early warning method for stratified and graded groundwater levels in different types of basins is proposed; based on real-time groundwater monitoring, dynamic early warning of dual control indicators is realized, providing a technical basis for dynamic supervision of groundwater in basins and sustainable utilization of water resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the method of the present invention;

[0030] Figure 2 A comparative analysis diagram of the predicted value and the monitored value of the overexploitation zone prediction model in a specific embodiment of the present invention;

[0031] The following is a detailed description of the embodiments of the present invention with reference to the accompanying drawings. DETAILED DESCRIPTION

[0032] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The embodiments are only used to explain the present invention and are not used to limit the scope of the present invention. The present invention is described in more detail by way of example with reference to the accompanying drawings in the following paragraphs. The advantages and features of the present invention will become clearer according to the following description. It should be noted that the accompanying drawings are all in a very simplified form and are not in precise proportions, and are only used to facilitate and clearly assist in explaining the purpose of the embodiments of the present invention.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0034] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0035] A groundwater dynamic early warning method, such as Figure 1 As shown, it includes monitoring and early warning and prediction and early warning.

[0036] Monitoring and early warning include water level drop type warning and water level rise type warning; water level drop type warning includes groundwater level warning in over-exploited areas and water level and water volume dual control warning. The groundwater level warning in over-exploited areas adopts a five-level warning system, and the water level and water volume dual control warning adopts normal water year indicators; water level rise type warning includes water level rise type warning in urban areas and water level rise type warning in agricultural irrigation areas.

[0037] Forecasting and early warning use time series models, multivariate linear regression models, and neural network models to determine the optimal forecasting model for each city.

[0038] The operation of groundwater level warning in over-exploited areas is as follows: select the station network, determine the calculation method and benchmark value. The benchmark value is the maximum, minimum, average value of the series burial depth and the value of the same period last year. According to the groundwater level burial depth monitoring data, a comparative analysis is carried out with the benchmark value, and a dynamic evaluation is conducted on the monthly average burial depth changes. A five-level warning system is adopted, and the warning results are assigned five colors: red, orange, yellow, blue and green. Warnings are issued for three levels: red, orange and yellow.

[0039] The warning rules for groundwater level warning in over-exploitation areas are:

[0040] ① If the average water level of the month is less than or equal to the minimum water level of the same period in the previous five years, it is a red warning;

[0041] ② If the average water level of the month is more than 1.0 meter lower than the average water level of the same period in the previous five years and is higher than the minimum water level of the same period in the previous five years, it is an orange warning;

[0042] ③ If the average water level in the current month is more than 1.0 meter lower than the water level in the same period last year, or is less than the average water level of the same period by no more than 1.0 meter, it is a yellow warning;

[0043] ④ If the average water level of the month is greater than or equal to the average water level of the previous five years, and less than the maximum water level of the previous five years, it will be colored blue;

[0044] ⑤ If the average water level of the month is greater than or equal to the maximum water level of the same period in the previous five years, the color is green.

[0045] Table 1 Groundwater level fluctuation classification levels, divisions, coloring requirements and chromaticity values

[0046] level Compared with the characteristic value of the same period, the water level variation* / m Water level change compared with the same period last year* / m color 1 Monthly average water level ≤ minimum value red 2 Monthly average water level < mean -1 orange color 3 Mean - 1 ≤ Monthly average water level < Mean Average monthly water level < water level of the same period last year - 1 yellow 4 Mean value ≤ monthly average water level < maximum value blue 5 Average monthly water level ≥ maximum green

[0047] Water level and water volume dual control early warning:

[0048] The groundwater withdrawal control indicators are comprehensively analyzed by planning analysis and weight decomposition methods, and the groundwater withdrawal control indicators of provincial administrative regions are decomposed into basic units; the annual groundwater withdrawal control is that each province uses the planning analysis method, interpolation method, and trend analysis method to comprehensively analyze the groundwater withdrawal control indicators of each sub-district in its jurisdiction in 2020 and 2030, and determine the groundwater withdrawal control indicators of different levels;

[0049] The objects of regional groundwater level control index determination include shallow groundwater and deep confined water. The regional groundwater level control index determination includes the groundwater level control index of over-exploited areas and the groundwater level control index of non-over-exploited areas. The groundwater level control index determination of over-exploited areas includes two situations: normal water year and different flood and dry water conditions.

[0050] At present, Tianjin, Hebei, Henan, Shandong and Liaoning provinces in the Haihe River Basin have issued groundwater control indicators, announced groundwater withdrawal control indicators, groundwater level (depth) control indicators and groundwater management indicators, involving a total of 176 county-level administrative regions in the Haihe River Basin. The compliance of groundwater withdrawal control indicators and water level control indicators in county-level administrative regions in the over-exploitation areas of the Haihe River Basin is evaluated. Among them, the groundwater withdrawal control indicator adopts the corresponding value in 2025, and the groundwater withdrawal volume adopts the data of the National Water Use Statistical Survey Direct Reporting Management System, which is analyzed and evaluated on a quarterly basis; the groundwater level indicator value adopts the corresponding value of the 2025 normal water scenario, and the water level adopts the arithmetic mean of the depth data of the groundwater level monitoring wells in the evaluation area, which is analyzed and evaluated on a monthly basis. The warning for areas where control indicators have not been released adopts a five-level warning system.

[0051] Water level rise warning:

[0052] In urban areas, the impact of rapid groundwater rise on underground projects (anti-floating) is considered; in agricultural irrigation areas, the impact of high groundwater levels on land salinization is considered, and an early warning method for groundwater level rise is proposed.

[0053] The early warning classification standards for urban areas and agricultural irrigation areas are set at four levels, namely:

[0054] Level 1 warning: The average groundwater depth at the end of the month is ≤ the level 1 warning threshold;

[0055] Level 2 warning: Level 1 warning threshold < average groundwater depth at the end of the month ≤ Level 2 warning threshold;

[0056] Level 3 warning: Level 2 warning threshold < average groundwater depth at the end of the month ≤ Level 3 warning threshold;

[0057] Level 4 warning: Level 3 warning threshold < average groundwater depth at the end of the month ≤ Level 4 warning threshold;

[0058] No warning: Level 4 warning threshold < average groundwater depth at the end of the month.

[0059] The prediction and early warning adopts the methods of time series, multivariate regression and neural network models to establish a monthly prediction and forecasting plan for the changes in groundwater level depth in various municipal administrative regions, realize the monthly dynamic prediction of groundwater level depth in various municipal administrative regions, and provide a basis for the management of groundwater resources in the basin.

[0060] (1) Time Series Model

[0061] The groundwater level depth series belongs to a non-stationary time series. The autoregressive integrated moving average model (ARIMA) is suitable for the prediction of non-stationary time series. ARIMA models are divided into autoregressive models, moving average models and autoregressive integrated moving average models, namely AR(p), MA(q) and ARIMA(p, d, q) models, where p, d, q represent the autoregressive order, difference order and moving average order of the time series, respectively. If there are seasonal periodic fluctuations in the time series, a seasonal ARIMA model is required, which can be marked as ARIMA(p, d, q)×(P, D, Q), where P, D, Q represent the seasonal autoregressive order, seasonal difference order and seasonal moving average order, respectively.

[0062] The establishment of ARIMA model is divided into the following three steps: ① Model identification. First, a preliminary analysis of the long-term trend of the time series is conducted. If the sequence does not meet the requirements of stationarity, it is necessary to achieve stationarity of the sequence through difference or seasonal difference; ② Determination of the model order. The parameters p and q of the alternative model are determined by autocorrelation and partial correlation analysis charts; p and q are generally determined by trying them one by one from high order to low order. You can also use the auto.arima function in R language to determine multiple alternative models and compare the AIC (Akaike Information Criterion) and BIC (Bayesian Information Criterion) values ​​of different models. The smaller the value, the better the simulation effect of the model; ③ Model diagnosis and prediction. The residual sequence of the model is tested for white noise, that is, the model is tested for Box-Ljung statistics. If the residual sequence is a random sequence, the model is suitable for prediction. After the model is established, the degree of fit of the model is evaluated by MAPE (mean relative error) and RMSE (root mean square error). The lower the value, the better the fit. Finally, the model is used for prediction, and the prediction effect of the model can be evaluated by the relative prediction error.

[0063] (2) Multiple Linear Regression Model

[0064] The regression analysis method is to establish a statistical model to study the closeness of the relationship between variables and the structural state for model prediction.

[0065] When the regression function is a linear function, given a set of independent variable values ​​x 1i , x 2i , …, x pi , that is:

[0066] y i =β0+β1x 1i +β2x 2i +···+β p x pi +ε i (a)

[0067] Among them, β0, β1, β2,···,β p are unknown parameters, often called regression coefficients, β0 is called the regression constant, β1, β2, ···, β p It is called the regression coefficient. When p ≥ 2, it is a multiple linear regression model.

[0068] According to the changing rules and influencing factors of the measured data of groundwater level depth, a multiple regression model was constructed with the groundwater level depth H(t) of the administrative area in the current month as the dependent variable, and the groundwater level depth H(t-1) of the previous month, the monthly precipitation P(t), and the monthly groundwater extraction volume Q(t) as the independent variables. SPSS (Statistical Product Service Solutions) software was used to construct the multiple linear regression model.

[0069] (3) Neural network model

[0070] The radial basis function (RBF) neural network is a local approximation network. Compared with the global approximation network, it has a faster learning convergence speed and can approximate any nonlinear function. It can transform complex nonlinear problems into linear separable problems to analyze the regularity of data, greatly improving its overall simulation prediction efficiency and accuracy compared with the traditional BP neural network.

[0071] The RBF neural network consists of an input layer, a hidden layer, and an output layer. The input sample vector is connected to the hidden layer through a nonlinear radial basis function. The hidden layer maps the low-dimensional vector to a high-dimensional feature space according to a certain weight, thereby analyzing the inherent law of the original sample. Its structure is shown in (a), where x n is the nth input data; Φ P is the radial basis function corresponding to the Pth hidden layer; K P is the training data corresponding to the Pth hidden layer; w P is the weight corresponding to the Pth hidden layer.

[0072] Using the statistical software SPSS, the hidden layer activation function of the RBF neural network is the Softmax function:

[0073]

[0074] Among them, z i is the output value of the ith node, and C is the number of output nodes, that is, the number of categories. The Softmax function can be used to convert the output values ​​of multiple categories into a probability distribution in the range of [0, 1] with a sum of 1.

[0075] The output neuron result is the linear sum of the hidden layer output data and the weights:

[0076]

[0077] Where Y ij is the output value of the jth output sample corresponding to the i-th input sample; Φ ij is the basis function of the jth hidden layer corresponding to the ith input sample; Wj is the weight of the jth hidden layer output.

[0078] This modeling classifies the influence of other factors on the groundwater level depth of the month, except for the monthly precipitation and the monthly artificial mining volume, into the H(t-1) factor, that is, the groundwater level depth of the previous month. In this way, the existing data can be fully utilized and the number of independent variables that need to be input can be reduced. Specific embodiment:

[0080] (1) Groundwater level warning in over-exploitation areas

[0081] According to the requirements of the Ministry of Water Resources for water level forecasting and early warning, the water level changes in the shallow and deep over-exploitation areas in Beijing, Tianjin and Hebei in North China are analyzed on a monthly basis and compared with the benchmark values ​​(2018-2023), using a five-level early warning system. Taking the early warning of the over-exploitation areas in Beijing, Tianjin and Hebei in North China in June 2024 as an example, the results are as follows:

[0082] In June 2024, the average shallow groundwater level in the Beijing-Tianjin-Hebei groundwater over-exploitation area in North China increased by 1.00 meters compared with the same period last year, and the average deep groundwater level decreased by 1.21 meters compared with the same period last year.

[0083] There are 14 administrative districts in the shallow groundwater overexploitation area of ​​Beijing-Tianjin-Hebei in North China, 1 of which is in red alert (greater than the maximum burial depth), namely Xinji City, Hebei Province, and 1 of which is in yellow alert (greater than the average burial depth of the same period sequence by no more than 1.0 meter), namely Zhangjiakou City, Hebei Province; there are 12 administrative districts in the deep groundwater overexploitation area, 1 of which is in orange alert (greater than the average burial depth of the same period sequence by more than 1.0 meter), namely Xinji City, Hebei Province, and 4 of which are in yellow alert, namely Baoding City, Hebei Province (greater than the average burial depth of the same period sequence by no more than 1.0 meter), Handan City (greater than the average burial depth of the same period sequence by no more than 1.0 meter, greater than the burial depth of the same period last year by more than 1.0 meter), Hengshui City (greater than the burial depth of the same period last year by more than 1.0 meter), Xingtai City (greater than the average burial depth of the same period sequence by no more than 1.0 meter, greater than the burial depth of the same period last year by more than 1.0 meter). The shallow and deep warning zoning map of the overexploitation area in Beijing-Tianjin-Hebei is as follows: Figure 2 shown.

[0084] (2) Groundwater “Dual Control” Supervision and Early Warning

[0085] According to the "Notice of the General Office of the Ministry of Water Resources on the Determination of Groundwater Control Indicators" and the control indicators and groundwater monitoring data issued by the basin provinces and cities, by the end of the second quarter of 2024, there will be three counties (cities, districts) in the Haihe River Basin where the actual total groundwater withdrawal exceeds the sequential value of the control indicator at the end of the second quarter (50% of the annual indicator). They are Caofeidian District and Qian'an County of Tangshan City, Hebei Province, and Zhangjiakou City. Their actual total water withdrawal accounts for 64.45%, 51.49% and 52.60% of the annual control indicator.

[0086] In June 2024, there were 23 counties (cities and districts) in the Haihe River Basin where the average monthly water level of shallow groundwater exceeded the control index, namely, Binhai New Area, Jinnan District, Wuqing District, Ninghe District of Tianjin, Zhao County and Jinzhou City of Shijiazhuang City, Hebei Province, Linzhang County, Cheng'an County and Wei County of Handan City, Gaoyang County, Li County and Boye County of Baoding City, Suning County of Cangzhou City, Gu'an County and Dachang Hui Autonomous County of Langfang City, Raoyang County and Anping County of Hengshui City, Xin'an County and Xiong County of Xiong'an New Area, Xinji City, Neihuang County of Anyang City, Henan Province, Hualong District of Puyang City, Huixian City of Xinxiang City, exceeding the control index by 0.34-8.77 meters; deep groundwater level exceeded the control index by 0.34-8.77 meters. There are 33 counties (cities, districts) where the burial depth exceeds the control index, namely Tianjin Binhai New Area, Zhao County of Shijiazhuang City, Hebei Province, Leting County of Tangshan City, Changli County of Qinhuangdao City, Feixiang District, Cheng'an County, Qiu County, Jize County, Guangping County, Wei County, Quzhou County of Handan City, Ningjin County, Julu County, Xinhe County, Guangzong County, Wei County, Qinghe County, Linxi County, Nangong City of Xingtai City, Gaoyang County and Li County of Baoding City, Qing County, Suning County, Nanpi County, Xian County, Botou City of Cangzhou City, Jizhou District, Raoyang County, Jing County, Fucheng County, Shenzhou City of Hengshui City, Xiong County of Xiong'an New Area, and Xinji City, exceeding the control index by 0.27-22.33 meters.

[0087] Table 2 Supervision information on total groundwater withdrawal in the Haihe River Basin, unit: 10,000 m 3

[0088]

[0089] 1.The data comes from the National Water Use Statistics Survey Direct Reporting Management System.

[0090] 2. The total groundwater withdrawal in 2024 for the county-level administrative regions marked in red exceeds the sequential value of the annual control index.

[0091] 3. For counties (cities, districts) where the total groundwater withdrawal is marked in red, indicate “total groundwater withdrawal control index / actual total groundwater withdrawal, percentage”.

[0092] Table 3 Supervision information of groundwater depth in Haihe River Basin, unit: meter

[0093]

[0094] (3) Groundwater level forecast and early warning in the overexploitation areas of Beijing, Tianjin and Hebei

[0095] In accordance with the relevant requirements of the Ministry of Water Resources' "Notice on Carrying out Groundwater Level Forecasting and Early Warning Work in North China" (Shui Wen Bian Zi

[2023] No. 135), a monthly water level change forecast analysis is carried out in North China as a pilot. The over-exploited area water level prediction model is used to predict the changing trend and amplitude of groundwater levels in the over-exploited area of ​​Beijing-Tianjin-Hebei from February to June 2024. The analysis objects include the shallow layer of Beijing, the shallow and deep layers of Tianjin, the shallow and deep layers of Hebei Province, and the shallow and deep layers of the entire over-exploited area of ​​Beijing-Tianjin-Hebei. The comparative analysis of the predicted values ​​and the monitored values ​​of the over-exploited area prediction model is as follows: Figure 2 shown.

[0096] By comparing the predicted values ​​and monitored values ​​of the shallow over-exploitation areas in Beijing-Tianjin-Hebei from February to June 2024, it was found that the average, maximum and minimum values ​​of the errors were 0.21m, 0.37m and 0.02m, respectively, and the changing trend of the predicted water level series from February to June was consistent with the changing trend of the monitored water level. The analysis results of the predicted values ​​and monitored values ​​in the deep over-exploitation areas showed that the average, maximum and minimum values ​​of the errors were 0.33m, 0.65m and 0.14m, respectively, and the changing trend of the predicted water level series from February to June was consistent with the changing trend of the monitored water level. Therefore, according to the changing trend of the predicted and measured value series of water levels in the deep and shallow over-exploitation areas of Beijing-Tianjin-Hebei and the analysis results of the water level amplitude, it can be shown that the model is more accurate in simulating and predicting the changing trend of the deep and shallow water levels in Beijing-Tianjin-Hebei. The prediction results are submitted to the Ministry of Water Resources every month, providing strong support for the over-exploitation management assessment.

[0097] The present invention constructs a layered, zoned, and graded water level early warning system for over-exploited areas in North China and demonstrates its application. In combination with the work requirements of the Ministry of Water Resources for dynamic water level early warning, the average monthly water level depth monitoring data from 2018 to 2023 since the North China groundwater over-exploitation control action was adopted to study and propose monthly early warning thresholds for different administrative divisions in the basin, and construct a layered, zoned, and graded water level early warning system; based on time series, multivariate regression, neural network models and other methods, a monthly water level prediction model suitable for administrative divisions with different geological conditions in over-exploited areas in North China was constructed. By comparing and analyzing the monitoring data from February to June 2024 in the Beijing-Tianjin-Hebei over-exploited area, the predicted water level change trend is consistent with the actual monitoring data. The research results have been applied to the dynamic early warning of water levels in over-exploited areas in North China, which strongly supports the dynamic control and prediction of water levels in over-exploited areas in North China, and provides an advanced demonstration for the national groundwater early warning work.

[0098] The present invention realizes the supervision and early warning of dual control indicators. In combination with the work requirements of the Ministry of Water Resources for the dual control management of groundwater level and water volume, and based on the dual control indicators and groundwater real-time monitoring data analysis issued by various provinces, it took the lead in carrying out dynamic early warning of dual control indicators, providing a technical basis for promoting the rational development and sustainable utilization of groundwater in the Haihe River Basin. According to the early warning analysis, at the end of the second quarter of 2024, there were 3 counties (cities and districts) in the Haihe River Basin where the actual total groundwater withdrawal exceeded the sequential value of the control indicator at the end of the second quarter (50% of the annual indicator), and there were 23 counties (cities and districts) in the Haihe River Basin where the average monthly water level of shallow groundwater exceeded the control indicator; there were 33 counties (cities and districts) where the depth of deep groundwater exceeded the control indicator.

[0099] The present invention is described above by way of example in conjunction with the accompanying drawings. It is obvious that the specific implementation of the present invention is not limited to the above-mentioned method. As long as various improvements are made using the method concept and technical solution of the present invention, or are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A groundwater dynamic early warning method, characterized in that: Including monitoring and early warning and forecasting and early warning; Monitoring and early warning include water level drop early warning and water level rise early warning; water level drop early warning includes groundwater level early warning in over-exploited areas and water level and water volume dual control early warning. Groundwater level early warning in over-exploited areas adopts a five-level early warning system, and water level and water volume dual control early warning adopts normal water year indicators; water level rise early warning includes urban area water level rise early warning and agricultural irrigation area water level rise early warning; Forecasting and early warning use time series models, multivariate linear regression models, and neural network models to determine the optimal forecasting model for each city.

2. A groundwater dynamic early warning method according to claim 1, characterized in that: The operation of groundwater level early warning in over-exploited areas is as follows: select the station network, determine the calculation method and benchmark value, compare and analyze the groundwater level depth monitoring data with the benchmark value, dynamically evaluate the monthly average depth change, and adopt a five-level early warning system. The early warning results are assigned five colors: red, orange, yellow, blue, and green. Warnings are issued at three levels: red, orange, and yellow.

3. A groundwater dynamic early warning method according to claim 2, characterized in that: The benchmark values ​​are the maximum, minimum, average and same period last year's values ​​of the series burial depth.

4. A groundwater dynamic early warning method according to claim 3, characterized in that: The early warning rules for groundwater level warning in over-exploitation areas are: ① If the average water level of the month is less than or equal to the minimum water level of the same period in the previous five years, it is a red warning; ② If the average water level of the month is more than 1.0 meter lower than the average water level of the same period in the previous five years and is higher than the minimum water level of the same period in the previous five years, it is an orange warning; ③ If the average water level in the current month is more than 1.0 meter lower than the water level in the same period last year, or is less than the average water level of the same period by no more than 1.0 meter, it is a yellow warning; ④ If the average water level of the month is greater than or equal to the average water level of the previous five years, and less than the maximum water level of the previous five years, it will be colored blue; ⑤ If the average water level of the month is greater than or equal to the maximum water level of the same period in the previous five years, the color is green.

5. A groundwater dynamic early warning method according to claim 1, characterized in that: Water level and water volume dual control early warning: The groundwater withdrawal control indicators are comprehensively analyzed by planning analysis and weight decomposition methods, and the groundwater withdrawal control indicators of provincial administrative regions are decomposed into basic units; the annual groundwater withdrawal control is that each province uses the planning analysis method, interpolation method, and trend analysis method to comprehensively analyze the groundwater withdrawal control indicators of each sub-district in its jurisdiction in 2020 and 2030, and determine the groundwater withdrawal control indicators of different levels; The objects of regional groundwater level control index determination include shallow groundwater and deep confined water. The regional groundwater level control index determination includes the groundwater level control index of over-exploited areas and the groundwater level control index of non-over-exploited areas. The groundwater level control index determination of over-exploited areas includes two situations: normal water year and different flood and dry water conditions.

6. A groundwater dynamic early warning method according to claim 1, characterized in that: Water level rise warning: In urban areas, the impact of rapid groundwater rise on underground projects is taken into consideration; in agricultural irrigation areas, the impact of high groundwater levels on land salinization is taken into consideration, and an early warning method for groundwater level rise is proposed.

7. A groundwater dynamic early warning method according to claim 6, characterized in that: The early warning classification standards for urban areas and agricultural irrigation areas are set at four levels, namely: Level 1 warning: The average groundwater depth at the end of the month is ≤ the level 1 warning threshold; Level 2 warning: Level 1 warning threshold < average groundwater depth at the end of the month ≤ Level 2 warning threshold; Level 3 warning: Level 2 warning threshold < average groundwater depth at the end of the month ≤ Level 3 warning threshold; Level 4 warning: Level 3 warning threshold < average groundwater depth at the end of the month ≤ Level 4 warning threshold; No warning: Level 4 warning threshold is less than the average groundwater depth at the end of the month.

8. A groundwater dynamic early warning method according to claim 1, characterized in that: The prediction and early warning adopts time series model, multivariate regression model and neural network model to establish a monthly prediction and forecasting scheme for the changes in groundwater level depth in each municipal administrative area, realize the monthly dynamic prediction of groundwater level depth in each municipal administrative area, and provide a basis for the management of groundwater resources in the basin.

Images