Base flow simulation method based on multiple linear regression model

Abstract

The present invention discloses a base flow simulation method based on a multiple linear regression model, aiming to improve the simulation accuracy of base flow in a watershed. The method mainly includes: 1) simulating the base flow using multiple widely used base flow separation methods based on the measured outflow sequence data; 2) constructing a new base flow simulation method using a multiple linear model based on the base flow index simulated by each method; and 3) selecting a base flow verification value based on the rainfall sequence data and evaluating the base flow simulation value. The beneficial effects of the present invention are that it can effectively remove the error between the base flow simulation value and the actual value, significantly improve the simulation accuracy of the base flow, build a foundation for future base flow prediction, and improve the reliability of the simulation prediction.

Description

Technical Field

[0001] This application claims the priority of a Chinese patent application filed with the China National Intellectual Property Administration on April 28, 2023, with the application number 202310477320.5 and the invention title "Base Flow Simulation Method Based on Multiple Linear Regression Model", and all of its contents are incorporated herein by reference.

[0002] The present invention relates to the technical field of base flow process simulation, and specifically to a base flow simulation method based on a multiple linear regression model.

Background Art

[0003] Base flow is one of the important components of runoff. The runoff during periods other than rainfall is mainly supplied by base flow, which is very important for river ecosystems and human water use. As the most important replenishment source of runoff during the dry season, base flow is closely related to the basic ecological maintenance, water quality, domestic water supply, etc. of the basin. At the same time, the base flow process is also an important part of the water cycle in the basin and is involved in the water exchange process between surface runoff, unsaturated aquifers, and groundwater. Base flow is mainly affected by related factors such as soil hydrogeological parameter control (i.e., the state of rivers and aquifers), precipitation and soil, land use characteristics, land use intensity, and surface gradient. Understanding the contribution of base flow to runoff is extremely important for the study of runoff generation and confluence characteristics in the basin and the rational allocation of water resources. When basic hydrological data are satisfied, by using multiple types of base flow calculation methods to grasp and analyze the base flow process in the basin, it is helpful to maintain the stability of the water source in the basin and protect the ecological environment. Therefore, how to ensure the accuracy of the separation results of the base flow process through theoretical analysis and numerical simulation is a current hot spot and difficult problem in the hydrological field.

[0004] At present, scholars have provided different base flow separation methods for river basins with different topographic features and climate conditions. Conventionally, there are mainly five types of base flow separation methods, namely numerical simulation method, graph method, analytical method, isotope method and hydrological model method. Each base flow separation method has its own characteristics, but the phenomena of overestimation or underestimation generally exist, and the base flow process path cannot be accurately simulated.

Summary of the Invention

Problems to be Solved by the Invention

[0005] In order to solve the technical defects of conventional base flow simulation, the object of the present invention is to provide a base flow simulation method based on a multiple linear regression model, to solve the phenomenon of overestimation or underestimation existing in each base flow simulation result, to reduce the uncertainty of base flow simulation, and to improve the simulation accuracy of base flow.

Means for Solving the Problems

[0006] In order to achieve the above object, a base flow simulation method based on a multiple linear regression model is A base flow simulation method based on a multiple linear regression model, comprising: Step S1) of obtaining historical outflow data, rainfall daily scale data of each hydrological observation station in the river basin, and the catchment area of the hydrological observation station; For the data obtained in S1), based on the outflow data and catchment area data, use a plurality of base flow separation methods to obtain different total base flow discharge data in step S2); For the data obtained in S2), simulate a plurality of base flow indices BFI by each base flow separation method, and obtain the average value of the base flow index

Number

[0007] Furthermore, the following steps can be further included. Step S5) of obtaining an actual base flow process curve based on the measured rainfall sequence data and historical outflow data obtained in S1). In the selection of the measured rainfall sequence data and historical outflow data in S5), if there is no continuous rainfall record for 15 days before a certain date, the outflow corresponding to that date is approximately equal to the base flow. According to this principle, outflow data is selected to obtain base flow verification data. Step S6) of comparing the base flow process curves obtained in S4) and S5) respectively, and judging the fitting accuracy and degree of coincidence. In S6), the method of judging the fitting accuracy and degree of coincidence is to evaluate each base flow simulation value using the Nash-Sutcliffe Efficiency coefficient (abbreviated as NSE) and the Percent bias (abbreviated as Pbias). The range of NSE is from negative infinity to 1, and the closer it is to 1, the higher the quality of the simulation is judged. The closer Pbias is to 0, the better the performance of the simulation is judged.

[0008] Preferably, the base flow index BFI and the average value of the base flow index in S3)

Number

Number

[0009] Preferably, the construction method of the Multiple Linear Regression Model (abbreviated as MLRM) in S4) is [Number] where In the formula, BF MLRM is the base flow sequence simulated by the MLRM method, and BF n is the base flow sequence simulated by n types of base flow separation methods, and a n is the regression coefficient of BF n .

[0010] Preferably, the base flow separation method in S2) is the HYSEP method, a one-parameter digital filter or a recursive digital filter. [Advantages of the Invention]

[0011] By using the above technical means, the beneficial effects of the present invention are as follows. 1) The present invention constructs a new base flow simulation method and process using a multiple linear regression model based on a plurality of widely used base flow separation methods, reduces the uncertainty of base flow simulation, and improves the simulation accuracy of base flow. 2) The base flow simulation method constructed in the present invention is not affected by the differences in the hydro-meteorological and subsurface conditions of the basin, and can improve the universality of the base flow simulation method. 3) The outflow corresponding to the date without rainfall records for the previous 15 consecutive days defined in the present invention is equal to the base flow, compensating for the lack of measured base flow values in the base flow simulation evaluation.

Brief Description of the Drawings

[0012] To more clearly explain the technical solutions of the embodiments of the present invention or the prior art, the drawings necessary to be used in the embodiments are briefly described below. Obviously, the drawings described below are only a part of the embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative efforts.

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Modes for Carrying Out the Invention

[0014] Hereinafter, with reference to the drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the technical scope of the present invention.

[0015] Hereinafter, with reference to the drawings, the technical solution of the present invention will be described in more detail by way of examples. In order to emphasize the advantages of the present invention, the present invention is specifically implemented by taking three widely used base flow separation methods such as the HYSEP method (HYdrograph SEParation), the one-parameter digital filter (abbreviated as OPDF), and the recursive digital filter (abbreviated as RDF) as examples.

[0016] As shown in FIG. 1, the base flow simulation method based on the multiple linear regression model of the present invention includes the following steps.

[0017] S1) Data collection: Collect the measured daily outflow data, daily rainfall data, and catchment area data of two hydrological observation stations in a certain basin from 1981 to 1995.

[0018] S2) Base flow separation: Use three widely used base flow separation methods such as the HYSEP method (HYdrograph SEParation), the one-parameter digital filter (abbreviated as OPDF), and the recursive digital filter (abbreviated as RDF) to obtain different total base flow discharge data.

[0019] S3) Model construction: First, based on the base flow simulation values obtained by the above three methods, the calculated base flow indices BFI are 0.393, 0.41, 0.418 for the first hydrological observation station and 0.464, 0.483, 0.484 for the second hydrological observation station respectively. Next, calculate the average values of the base flow indices of the two hydrological observation stations, and the results are 0.407 and 0.477 respectively.

[0020] Finally, a Multiple Linear Regression Model (abbreviated as MLRM) is constructed. The construction method is as follows:

Number

Number

[0021] ​S4) Model verification: Using the triple linear regression model constructed in the above steps, simulate the base flow of two hydrological observation stations. Based on the measured rainfall sequence data, it is defined that the outflow corresponding to the date without continuous rainfall records in the previous 15 days is equal to the base flow, and the obtained discontinuous base flow sequence is compared with the simulation value as the base flow verification value. As shown in FIGS. 2 and 3, FIG. 2(a) is the base flow simulation accuracy diagram based on the HYSEP method of the first hydrological observation station, FIG. 2(b) is the base flow simulation accuracy diagram based on the OPDF of the first hydrological observation station, FIG. 2(c) is the base flow simulation accuracy diagram based on the RDF of the first hydrological observation station, FIG. 2(d) is the base flow simulation accuracy diagram based on the MLRM of the first hydrological observation station, FIG. 3(a) is the base flow simulation accuracy diagram based on the HYSEP method of the second hydrological observation station, FIG. 3(b) is the base flow simulation accuracy diagram based on the OPDF of the second hydrological observation station, FIG. 3(c) is the base flow simulation accuracy diagram based on the RDF of the second hydrological observation station, FIG. 3(d) is the base flow simulation accuracy diagram based on the MLRM of the second hydrological observation station. The base flow simulated by the MLRM used in this application is superior to the results of the other three types of base flow separation methods in both the Nash-Sutcliffe efficiency coefficient and the percent bias. In particular, when the MLRM is applied to the first hydrological observation station, the NSE is 0.7 or more and reaches about -13%, and when the MLRM is applied to the second hydrological observation station, the NSE is 0.8 or more and reaches about -5%. The simulation evaluation value of the MLRM is much better than that of the other three simulation methods.

[0022] Figures 4 and 5 illustrate the hydrographs of the outflows and baseflows of two hydrological observation stations within a certain time period. As is evident from the figures, the baseflow process simulated by the MLRM can well capture the variations in the peaks and valleys of the baseflow, and its value is in the middle of the simulation values of the three methods, well compensating for the overestimation or underestimation phenomena existing in each baseflow separation method and significantly reducing the uncertainty of the baseflow simulation.

[0023] The above description is only for implementing the embodiments of the present invention and does not limit the present invention. The number of multiple types of baseflow separation methods in the present invention can be specifically customized based on different research areas. Any modifications, equivalent substitutions, improvements, etc. made within the scope of the claims of the present invention should all be within the protection scope of the present invention.

[0024] Each embodiment in this specification is described progressively. Each embodiment focuses on the differences from other embodiments, and the same or similar parts between each embodiment can be referred to each other.

[0025] In this specification, the principles and embodiments of the present invention are described using specific examples. However, the description of the above embodiments is only for helping to understand the method and its core idea of the present invention. Also, those skilled in the art can change both the specific embodiments and the application scope based on the idea of the present invention. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A base flow simulation method based on a multiple linear regression model, comprising: Step S1) of obtaining historical outflow data, rainfall daily scale data, and catchment area of each hydrological observation station in the basin; Step S2) of obtaining different base flow total flow rate data using a plurality of base flow separation methods based on the outflow data and catchment area data obtained in S1); For the data obtained in S2), simulating a plurality of base flow indices BFI by each base flow separation method, and calculating the average value of the base flow index 【Number 1】 in Step S3); Step S4) of obtaining simulation data using a multiple linear regression model based on the data obtained in S3), constructing a base flow process with the simulation data, and forming a simulation base flow process curve. A base flow simulation method based on a multiple linear regression model, characterized by including the above steps.

2. Step S5) of obtaining an actual base flow process curve based on the measured rainfall sequence data and historical outflow data obtained in S1); The base flow simulation method based on the multiple linear regression model according to Claim 1, further comprising Step S6) of comparing the base flow process curves obtained in S4) and S5) respectively, and determining the fitting accuracy and degree of coincidence.

3. In Step S6), the method for determining the fitting accuracy and degree of coincidence is to evaluate each base flow simulation value using the Nash-Sutcliffe Efficiency coefficient, abbreviated as NSE, and the Percent bias, abbreviated as Pbias. The range of NSE is from negative infinity to 1, and the closer it is to 1, the higher the quality of the simulation is judged. The closer Pbias is to 0, the better the performance of the simulation is judged. A base flow simulation method based on the multiple linear regression model according to Claim 2, characterized by the above.

4. In the selection of the measured rainfall sequence data and historical outflow data in S5), when there is no continuous rainfall record for 15 days before a certain date, the outflow corresponding to the date is approximately equal to the base flow. The outflow data is selected according to this principle to obtain the base flow verification data. The base flow simulation method based on the multiple linear regression model according to claim 2, characterized in that.

5. The base flow index BFI and the average value of the base flow index in S3) 【Number 2】 The calculation formula is 【Number 3】 and where V B is the total baseflow discharge within the time period, V S is the total outflow discharge within the time period, BFI 1 , BFI 2 , …, BFI n are respectively the baseflow index values simulated by n types of methods, n is the number of types of baseflow separation methods, and the numerical values of BFI are in the range of 0 to 1. A baseflow simulation method based on the multiple linear regression model according to any one of claims 1 to 4, characterized in that

6. The method for constructing the multiple linear regression model MLRM in S4) is 【Number 4】 and where BF MLRM is the base flow sequence simulated by the MLRM method, and BF n is the base flow sequence simulated by n types of base flow separation methods, and a n is the regression coefficient of BF n , and BFI 1 , BFI 2 , …, BFI n are respectively the base flow index values simulated by n types of methods. 【Number 5】 is the average value of the base flow index. The base flow simulation method based on the multiple linear regression model according to any one of claims 1 to 4, characterized in that.

7. The base flow separation method in S2) is the HYSEP method, one-parameter digital filter, recursive digital filter, base flow index method, smooth minimum value method, straight line separation method, recession curve method, time step method, graph method, water balance method, hydrological simulation method. The base flow simulation method based on the multiple linear regression model according to any one of claims 1 to 4, characterized in that.

Images