A method for constructing the water level-discharge relationship curve in areas without data by combining intelligent algorithms

Abstract

The present invention provides a method for constructing a water level-discharge relationship curve in data-deficient areas by combining intelligent algorithms. Based on the physical factors affecting the water level-discharge relationship curve and combined with intelligent algorithms, the flood process within the basin is simulated, and then a method for constructing the water level-discharge relationship curve in data-deficient areas by combining intelligent algorithms is proposed. This not only ensures the accuracy and reliability of the calculation results but also solves the problem of constructing the water level-discharge relationship curve in data-deficient areas. Moreover, this method mainly applies the basin digital elevation model and intelligent algorithms, the data source is stable and reliable, the functional relationship between variables in the method is clear, which is conducive to the rapid and automatic estimation of the water level-discharge relationship curve in data-deficient areas, and at the same time ensures the objectivity and rationality of the results, which can further promote the in-depth development of digital hydrology and the rapid improvement of flood control capabilities in mountainous areas.

Description

Technical Field

[0001] The present invention relates to the field of hydrological technology, and in particular to a method for constructing a water level - discharge relationship curve in data - scarce areas by combining intelligent algorithms. Background Art

[0002] In order to further improve the monitoring ability of flood processes in remote mountainous - area basins, it is necessary to adopt a more scientific method for the mutual derivation and conversion of water - level and flow - rate elements at monitoring sections. There is an urgent need for a scientific and reasonable method to construct the water - level - discharge relationship curve in data - scarce areas.

[0003] When constructing the water - level - discharge relationship curve, the first challenge is how to obtain the flow - rate data corresponding to a given water level based on the known water - level monitoring data. In current practical applications, the flow - rate data corresponding to a given water level often needs to rely on manual monitoring during floods to obtain a large number of measured flow - rate data, obtaining multiple sets of "water - level - flow - rate" sample points, and then fitting the functional relationship between the water level and the flow rate. However, the flow - rate measurement under flood conditions is not only costly but also has a high safety - production risk, making it difficult to construct the water - level - discharge relationship curve for river cross - sections in a large number of mountainous basins, restricting the improvement of flood - monitoring capabilities in mountainous basins.

[0004] Aiming at the above deficiencies, how to construct the water - level - discharge relationship curve for river cross - sections in a basin in an economical, effective, scientific, and reasonable way and enhance the flood - monitoring capabilities of river cross - sections in a large number of mountainous basins is exactly the problem that the inventor needs to solve. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for constructing a water - level - discharge relationship curve in data - scarce areas by combining intelligent algorithms in view of the above - mentioned deficiencies of the prior art. The method has the advantages of stable and reliable data sources, high calculation efficiency, objective and reasonable results, etc., which is conducive to the reasonable construction of the water - level - discharge relationship curve in data - scarce areas.

[0006] To achieve the above - mentioned purpose, the present invention adopts the following technical solutions:

[0007] The present invention provides a method for constructing a water - level - discharge relationship curve in data - scarce areas by combining intelligent algorithms, including the following steps:

[0008] S1. Taking the data - scarce hydrological station P s which needs to construct the water - level - discharge relationship curve as the center, searching and judging the data - rich hydrological stations P s with water - level and flow - rate observation data upstream and downstream of the data - scarce hydrological station P c , extracting the data - scarce basin W s with the data - scarce hydrological station P as the outlet s and the data - rich basin W c with the data - rich hydrological station P as the outletc , and extract the watershed without data W s and the basin with data W c The underlying surface characteristics within;

[0009] S2, based on the data-free basin W s and the data basin W c The underlying surface characteristics within the area are used for similarity judgment. For the watershed with data W c , using the basin underlying surface characteristics to estimate the parameter set Par of the hydrological model with data c , and estimate the W for the basin without data s The data-free hydrological model parameter set Par s ;

[0010] S3, based on the hydrological model parameter set Par c As the initial value, combined with the intelligent optimization algorithm, the W suitable for the basin with data is obtained. c The data-based calibration parameter set Par c ′;

[0011] S4. Combined with the watershed without data W s The data-free hydrological model parameter set Par s , there is data for the basin W c The data hydrological model parameter set Par c , and the basin with data W c The data-based calibration parameter set Par c ′, and the data-free basin W is calculated s The data-free calibration parameter set Par s ′;

[0012] S5. Combined with the watershed without data W s The data-free calibration parameter set Par s ′, using rainfall as input to simulate the watershed without data W s Exit, that is, no data hydrological station P s The flow process at the location is based on the combination of the data-free hydrological station P s The observed water level process and the simulated process are fitted to obtain the data-free hydrological station P s Hydrological flow relationship curve.

[0013] Furthermore, in S1, the watershed W without data is extracted s and the basin with data W c The underlying surface characteristics include slope, clay, silt, sand and vegetation cover.

[0014] Further, the S2 is specifically:

[0015] S201. Use principal component analysis to reduce the dimension of the extracted underlying surface characteristics of the basin, streamline the hydrological characteristics of the basin, and obtain the underlying surface characteristic factors that can reflect the hydrological characteristics of the basin.

[0016] S202. Use the underlying surface characteristic factors to calculate the similarity sim of the basin hydrological characteristics:

[0017]

[0018]

[0019] where α c,n is the underlying surface characteristic factor numbered n in the basin W c with data; α s,n is the underlying surface characteristic factor numbered n in the basin W s without data; β n is the weight of the nth underlying surface characteristic factor; sim(α c,n , α s,n ) is the similarity of the underlying surface characteristic factor numbered n between the basin W c with data and the basin W s without data; N is the number of types of underlying surface characteristic factors obtained after principal component analysis.

[0020] S203. For the basin W c with data where the similarity sim of the basin hydrological characteristics is higher than the threshold, estimate the parameter set Par c of the hydrological model with data, and the parameter set Par c of the hydrological model with data includes the tension water storage capacity WM, the free water storage capacity SM, the interflow coefficient KI of subsurface runoff; the groundwater outflow coefficient KG;

[0021] WM = SoilT × (θ f - θ wp );

[0022] SM = SoilT × (θ s - θ f );

[0023]

[0024] KG = 0.7 - KI;

[0025] where SoilT is the soil thickness; θ s is the saturated water content; θ f is the field capacity; θ wp is the wilting coefficient; K is the soil permeability coefficient, and Δt is the length of the calculation period;

[0026] S204. In the gauged basin W c with the initial parameters being the gauged hydrological model parameter set Par c and combining with the SCE-UA algorithm that specifically conforms to the optimization strategy, perform the optimization calibration of the water level model parameters to obtain the gauged calibration parameter set Par c ' applicable to the gauged basin W c ';

[0027] S205. Using the method in S303, estimate the ungauged hydrological model parameter set Par s in the ungauged basin W s .

[0028] Furthermore, the specific implementation of S4 is as follows:

[0029]

[0030]

[0031]

[0032]

[0033]

[0034] where WM s ', SM s ', KI s ', KG s ' are the adjusted parameter values in the ungauged basin W s ; WM s , SM s , KI s , KG s are the parameter values before adjustment in the ungauged basin W s ; WM c ', SM c ', KI c ', KG c ' are the adjusted parameter values in the gauged basin W c ; WM c , SM c , KI c , KG c are the parameter values before adjustment in the gauged basin W c .

[0035] Furthermore, the specific steps of S5 are as follows:

[0036] S501. Combining the ungauged calibration parameter set Par of the ungauged basin W s ​s ′, the rainfall is used as the input to simulate the ungauged basin W s at the outlet of the basin, that is, the flow process at the ungauged hydrological station P s ;

[0037] S502. Combine the water level process observed at the ungauged hydrological station P s with the simulated flow process to fit the water level - discharge relationship curve L of the ungauged hydrological station P s : a

[0038] L a = F(Z, Q);

[0039] where Z is the water level process of the ungauged hydrological station P s ; Q is the flow process of the ungauged hydrological station P simulated in S501 s , and the functional form of the water level - discharge relationship curve L a adopts a cubic equation of one variable.

[0040] The beneficial effects of the present invention are as follows: Based on the physical factors affecting the water level - discharge relationship curve, combined with intelligent algorithms to simulate the flood process in the basin, and then a method for constructing the water level - discharge relationship curve in ungauged areas combined with intelligent algorithms is proposed. This not only ensures the accuracy and reliability of the calculation results, but also solves the problem of constructing the water level - discharge relationship curve in ungauged areas. And this method mainly applies the basin digital elevation model and intelligent algorithms, the data source is stable and reliable, the functional relationship between variables in the method is clear, which is conducive to the rapid and automatic estimation of the water level - discharge relationship curve in ungauged areas, and at the same time ensures the objective rationality of the results, which can further promote the in - depth development of digital hydrology and the rapid improvement of flood control ability in mountainous areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of a method for constructing the water level - discharge relationship curve in ungauged areas combined with intelligent algorithms;

[0042] Figure 2 is a basin digital elevation map;

[0043] Figure 3 is the reference station and the station that needs the water level - discharge relationship curve;

[0044] Figure 4 is a basin slope map;

[0045] Figure 5 is a basin sand grain content distribution map;

[0046] Figure 6 is a basin silt content distribution;

[0047] ​Figure 7 is the clay content distribution in the basin;

[0048] Figure 8 is the SM parameter of the basin;

[0049] Figure 9 is the WM parameter of the basin;

[0050] Figure 10 is the KG parameter of the basin;

[0051] Figure 11 is the KI parameter of the basin;

[0052] Figure 12 are the observed water levels and the scatter points of the simulated discharges;

[0053] Figure 13 is at the site P s where the water level-discharge relationship curve is fitted. Specific Embodiments

[0054] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0055] Please refer to Figure 1 , a method for constructing a water level-discharge relationship curve in data-deficient areas by combining intelligent algorithms, comprising the following steps:

[0056] S1. Centering on the data-deficient hydrological station P s where the water level-discharge relationship curve needs to be constructed, search and determine the data-rich hydrological stations P s upstream and downstream of the data-deficient hydrological station P c with observed water level and discharge data, as shown in Figure 2 ;

[0057] Extract the data-deficient basin W s with the data-deficient hydrological station P s as the outlet and the data-rich basin W c with the data-rich hydrological station P c as the outlet, and extract the underlying surface characteristics within the data-deficient basin W s and the data-rich basin W c ;

[0058] S2. Based on the underlying surface characteristics within the data-deficient basin W s and the data-rich basin W c carry out similarity discrimination. For the data-rich basin W c that passes the similarity discrimination, use the underlying surface characteristics of the basin to estimate the data-rich hydrological model parameter set Par c, and estimate the W for the basin without data s The data-free hydrological model parameter set Par s ;

[0059] S3, based on the hydrological model parameter set Par c As the initial value, combined with the intelligent optimization algorithm, the W suitable for the basin with data is obtained. c The data-based calibration parameter set Par c ′;

[0060] S4. Combined with the watershed without data W s The data-free hydrological model parameter set Par s , there is data for the basin W c The data hydrological model parameter set Par c , and the basin with data W c The data-based calibration parameter set Par c ′, and the data-free basin W is calculated s The data-free calibration parameter set Par s ′;

[0061] S5. Combined with the watershed without data W s The data-free calibration parameter set Par s ′, using rainfall as input to simulate the watershed without data W s Exit, that is, no data hydrological station P s The flow process at the location is based on the combination of the data-free hydrological station P s The observed water level process and the simulated process are fitted to obtain the data-free hydrological station P s Hydrological flow relationship curve.

[0062] In S1, extract the watershed W without data s and the basin with data W c The slope inside, such as Figure 3 , clay particles such as Figure 6 , powder particles such as Figure 5 , sand particles such as Figure 4 , and the underlying surface characteristics of vegetation cover.

[0063] The S2 is specifically:

[0064] S201, using principal component analysis to reduce the dimension of the extracted basin underlying surface characteristics, simplify the basin hydrological characteristics, and obtain underlying surface characteristic factors that can reflect the basin hydrological characteristics;

[0065] S202, using the underlying surface characteristic factors, calculate the basin hydrological characteristic similarity sim:

[0066]

[0067]

[0068] Among them, α c,n is the underlying surface characteristic factor numbered n in the basins with data W c ; α s,n is the underlying surface characteristic factor numbered n in the basins without data W s ; β n is the weight of the nth underlying surface characteristic factor; sim(α c,n , α s,n ) is the similarity of the underlying surface characteristic factor numbered n between the basins with data W c and the basins without data W s ; N is the number of types of underlying surface characteristic factors obtained after principal component analysis;

[0069] S203. For the basins with data W c where the similarity sim of the basin hydrological characteristics is higher than the threshold, estimate the parameter set Par c of the hydrological model with data by using the underlying surface characteristics of the basin. The parameter set Par c of the hydrological model with data includes the tension water storage capacity WM, such as Figure 7 , the free water storage capacity SM, such as Figure 8 , the interflow runoff coefficient KI, such as Figure 9 ; the groundwater outflow coefficient KG, such as Figure 10 ;

[0070] WM = SoilT × (θ f - θ wp );

[0071] SM = SoilT × (θ s - θ f );

[0072]

[0073] KG = 0.7 - KI;

[0074] Among them, SoilT is the soil thickness; θ s is the saturated water content; θf is the field capacity; θ wp is the wilting coefficient; K is the soil permeability coefficient, and Δt is the calculation time step;

[0075] S304. In the basins with data W c , taking the parameter set Par c of the hydrological model with data as the initial parameter, combined with the SCE-UA algorithm that specifically conforms to the optimization strategy, carry out the optimization calibration of the water level model parameters to obtain the parameters applicable to the basins with data W cThe calibrated parameter set Par with data c ′;

[0076] S305. Using the method in S303, estimate the ungauged basin hydrological model parameter set Par within the ungauged basin W s within the ungauged basin W s .

[0077] The specific implementation of S4 is as follows:

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] Among them, WM s ′, SM s ′, KI s ′, KG s ′ are the adjusted parameter values of the ungauged basin W s ; WM s , SM s , KI s , KG s are the parameter values of the ungauged basin W s before adjustment; WM c ′, SM c ′, KI c ′, KG c ′ are the adjusted parameter values of the gauged basin W c ; WM c , SM c , KI c , KG c are the parameter values of the gauged basin W c before adjustment.

[0084] The specific steps of S5 are as follows:

[0085] S501. Combine the ungauged calibrated parameter set Par s ′ of the ungauged basin W, and use rainfall as the input to simulate the outlet of the ungauged basin W s ′, that is, the flow process at the ungauged hydrological station P s at the outlet of the ungauged basin W, such as s ; Figure 11 ;

[0086] S502. Combine the ungauged hydrological station P sThe observed water level process and the simulated flow process are fitted to obtain the water level-flow relationship curve L of the gaugeless hydrological station P s of the water level-flow relationship curve L a as shown in Figure 12 and Figure 13 ;

[0087] L a = F(Z, Q);

[0088] where Z is the water level process of the gaugeless hydrological station P s ; Q is the flow process of the gaugeless hydrological station P obtained by simulation in S501 s , and the functional form of the water level-flow relationship curve L a adopts a cubic equation of one variable.

[0089] Embodiment

[0090] Taking the Daheba River Basin in Shaanxi Province as an example, an intelligent algorithm-based method for constructing the water level-flow relationship curve in gaugeless areas is applied

[0091] at the gaugeless hydrological station P in the Daheba River Basin s , and the function of the fitted water level-flow relationship curve L a is as follows:

[0092] Q = 32.7×Z 3 - 26946×Z 2 + 7401784×Z - 677732309.6;

[0093] The correlation between the observed water level and the simulated flow scatter points reaches 0.875.

[0094] The above embodiments only represent the implementation manners of the present invention, and the description is relatively specific and detailed, but it should not be construed as a limitation to the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be based on the appended claims.

Claims

1. A method for constructing a water level-discharge relationship curve in a data-free area using an intelligent algorithm, characterized in that: The following steps are involved: S1. Centering on the data-free hydrological station Ps for which the water level-flow relationship curve needs to be constructed, search and determine the data-free hydrological station Ps with water level and flow observation data upstream and downstream. c , extract the data-free basin W with the data-free hydrological station Ps as the outlet s and hydrological stations with data P c The watershed W with data for the export c , and extract the watershed without data W s and the basin with data W c The underlying surface characteristics within; S2, based on the data-free basin W s and the data basin W c The underlying surface characteristics within the area are used for similarity judgment. For the watershed with data W c , using the basin underlying surface characteristics to estimate the parameter set Par of the hydrological model with data c , and estimate the W for the basin without data s The data-free hydrological model parameter set Par s ; S3, based on the hydrological model parameter set Par c As the initial value, combined with the intelligent optimization algorithm, the W suitable for the basin with data is obtained. c The data-based calibration parameter set Par c ′; S4. Combined with the watershed without data W s The data-free hydrological model parameter set Par s , there is data for the basin W c The data hydrological model parameter set Par c , and the basin with data W c The data-based calibration parameter set Par c ′, and the data-free basin W is calculated s The data-free calibration parameter set Par s ′; S5. Combined with the watershed without data W s The data-free calibration parameter set Par s ′, using rainfall as input to simulate the watershed without data W s The outlet, that is, the flow process at the data-free hydrological station Ps, is fitted based on the water level process observed at the data-free hydrological station Ps and the flow process obtained by simulation, and the hydrological-discharge relationship curve of the data-free hydrological station Ps is obtained.

2. According to claim 1, a method for constructing a water level-discharge relationship curve in a data-free area in combination with an intelligent algorithm, characterized in that: In S1, extract the watershed W without data s and the basin with data W c The underlying surface characteristics include slope, clay, silt, sand and vegetation cover.

3. According to claim 2, a method for constructing a water level-discharge relationship curve in a data-free area in combination with an intelligent algorithm is characterized in that: The S2 is specifically: S201, using principal component analysis to reduce the dimension of the extracted basin underlying surface characteristics, simplify the basin hydrological characteristics, and obtain underlying surface characteristic factors that can reflect the basin hydrological characteristics; S202, using the underlying surface characteristic factors, calculate the basin hydrological characteristic similarity sim: Among them, αc,n is the basin W with data c The underlying surface characteristic factor with internal number n; α s,n For the basin without data W s The underlying surface characteristic factor with internal number n; β n is the weight of the nth underlying surface characteristic factor; sim(α c,n ,α s,n ) is the basin with data W c and no data basin W s The similarity between the underlying surface characteristic factors numbered n; N is the number of types of underlying surface characteristic factors obtained after principal component analysis; S203, for the basin with data whose hydrological characteristic similarity sim is higher than the threshold value, c , using the basin underlying surface characteristics to estimate the parameter set Par of the hydrological model with data c , the parameter set of the hydrological model with data Par c Including tension water storage capacity WM, free water storage capacity SM, soil outflow coefficient KI; groundwater outflow coefficient KG; WM = SoilT × (θf - θwp); SM = SoilT × (θs - θf); KG = 0.7-KI; Among them, SoilT is soil thickness; θs is saturated water content; θf is field water holding capacity; θwp is wilting coefficient; K is soil permeability coefficient, Δt is calculation period length; S204, in the basin with data W c In the data set, the hydrological model parameter set Par c As the initial parameters, combined with the SCE-UA algorithm that meets the specific optimization strategy, the optimization calibration of the water level model parameters is carried out to obtain the W suitable for the basin with data c The data-based calibration parameter set Par c ′; S205: Using the method in S203, estimate the watershed W without data s The data-free hydrological model parameter set Par s .

4. According to claim 3, a method for constructing a water level-discharge relationship curve in a data-free area in combination with an intelligent algorithm is characterized in that: The specific implementation of S4 is: Among them, WM s ′, SM s ′, KI s ′, KG s ′ is the basin without data W s Adjusted parameter value; WM s , SM s KI s , KG s For the basin without data W s Parameter value before adjustment; WM c ′, SM c ′, KI c ′, KG c ′ is the basin with data W c Adjusted parameter value; WM c , SM c KI c , KG c W is the basin with data c Parameter value before adjustment.

5. According to claim 1, a method for constructing a water level-discharge relationship curve in a data-free area in combination with an intelligent algorithm is characterized in that: The specific steps of S5 are: S501, combined with the watershed without data W s The data-free calibration parameter set Par s ′, using rainfall as input to simulate the watershed without data W s The flow process at the basin outlet, i.e., at the data-free hydrological station Ps; S502, combining the water level process observed at the data-free hydrological station Ps with the simulated flow process to obtain the water level flow relationship curve La of the data-free hydrological station Ps: La=F(Z,Q); Wherein, Z is the water level process of the dataless hydrological station Ps; Q is the flow process of the dataless hydrological station Ps simulated in S501, and the function form of the water level-flow relationship curve La adopts a cubic equation.

Images