Drainage basin sediment set simulation method based on runoff erosion energy

Through the basin sediment ensemble simulation method based on runoff erosion energy, using multiple hydrological models and Bayesian weighted averaging methods, the problem of insufficient accuracy in soil erosion simulation of traditional models is solved, and efficient and accurate simulation of basin sediment is achieved. Support scientific decision-making.

CN118917236APending Publication Date: 2024-11-08XIAN UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410949901.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The traditional runoff and erosion model has shortcomings in describing the physical mechanism of soil erosion, with large differences in calculation accuracy, and the distributed physical model parameters are redundant and data requirements are high, making it difficult to achieve accurate simulation of basin sediment.

Method used

Using the basin sediment ensemble simulation method based on runoff erosion energy, multiple hydrological models are constructed by collecting hydrological data, precipitation, potential evaporation data, water conservancy engineering data and NDVI raster data, and using Bayesian weighted average method to perform mud Sand simulation and accuracy evaluation was performed.

Benefits of technology

The accuracy of basin sediment simulation has been improved, and a solid foundation for physical mechanisms of confluence and sand production processes has been provided, which supports scientific decision-making in basin ecological environment protection and soil and water conservation planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917236A_ABST
    Figure CN118917236A_ABST
Patent Text Reader

Abstract

The invention discloses a drainage basin sediment set simulation method based on runoff erosion energy. The method specifically comprises the following steps: collecting hydrological data, precipitation and potential evapotranspiration data, hydraulic engineering data, NDVI grid data and drainage basin attribute data in a research period of a research drainage basin; calculating to obtain day-by-day and month measured flow data, surface precipitation data, surface potential evapotranspiration data and month-by-month and year measured sediment transport data; a plurality of watershed hydrological models of daily scale and monthly scale are constructed respectively; calculating to obtain monthly and annual watershed water conservancy project measure factor indexes, an NDVI average value and runoff erosion power; establishing monthly-scale and annual-scale energy and sand calculation modules corresponding to the plurality of watershed hydrological models respectively; and performing set simulation on the monthly-scale sediment simulation process and the annual-scale sediment simulation process by using a Bayesian weighted average method, and performing precision evaluation. According to the method, basin sediment set simulation can be efficiently and accurately realized from the aspect of runoff erosion energy, and the reliability of a simulation result is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of ecological environment protection and soil and water loss prevention and control, and relates to a basin sediment collection simulation method based on runoff erosion energy. Background Art

[0002] Climate change and human activities have led to soil erosion in many areas, resulting in the generation of large amounts of sediment, which is constantly transported and deposited in rivers. Intensified climate change, frequent droughts and extreme precipitation further affect the soil erosion process. At the same time, human activities such as rapid changes in land use and large-scale construction of reservoirs and dams have changed the water-sediment process in the basin, generating and silting up a large amount of sediment. The water-sediment process is an important process in the material cycle of the natural environment and has a profound impact on the development of human society. Therefore, it is particularly important to study the causes of changes in the water-sediment process in the basin and to achieve accurate simulation and prediction of sediment.

[0003] The study of the relationship between water and sediment in a watershed is gradually shifting from the traditional relationship between runoff and erosion, rainfall erosivity and erosion, to the analysis of the relationship between erosion energy and erosion, so as to more deeply reveal the essential laws of the soil erosion process. Traditional empirical models and black box models are insufficient in describing the physical mechanism of soil erosion, and the calculation accuracy varies greatly. Distributed physical models have the disadvantages of parameter redundancy and high data requirements, and are not easy to use. Therefore, it is very meaningful to propose a watershed sediment ensemble simulation method based on runoff erosion energy. Summary of the invention

[0004] The purpose of the present invention is to provide a method for simulating a watershed sediment ensemble based on runoff erosion energy, which can efficiently and accurately realize a watershed sediment ensemble simulation from the perspective of runoff erosion energy.

[0005] The technical solution adopted by the present invention is a basin sediment ensemble simulation method based on runoff erosion energy, which specifically includes the following processes: collecting hydrological data, precipitation and potential evapotranspiration data, water conservancy project data, normalized difference vegetation index NDVI raster data and basin attribute data of the study basin corresponding to the study period; calculating daily and monthly measured flow data, surface precipitation data and surface potential evapotranspiration data and monthly and annual measured sediment transport data; constructing a variety of basin hydrological models at daily and monthly scales respectively; calculating the monthly and annual basin water conservancy project measure factor index, NDVI average value and runoff erosion power; establishing monthly and annual scale energy sediment calculation modules corresponding to a variety of basin hydrological models respectively; using the Bayesian weighted average method to perform ensemble simulation of the monthly and annual scale sediment simulation processes respectively, and evaluating the accuracy of the simulation results.

[0006] The present invention is also characterized in that:

[0007] The specific steps include:

[0008] Step 1: Collect daily hydrological data of the basin control hydrological station during the study period, daily precipitation data and potential evapotranspiration data of meteorological stations inside and around the study basin, annual water conservancy project data, monthly and annual NDVI raster data and basin attribute data;

[0009] Step 2: Use the daily hydrological data of the basin control hydrological station during the study period collected in step 1, and after processing, obtain the daily and monthly measured flow data and the monthly and annual measured sediment transport data; according to the daily precipitation data and potential evapotranspiration data of the meteorological stations inside and around the study basin collected in step 1, after processing, obtain the daily and monthly surface precipitation and surface potential evapotranspiration data of the basin;

[0010] Step 3: Based on the daily and monthly measured flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data obtained in step 2, a calibration period and a verification period are divided, and a variety of daily and monthly scale watershed hydrological models are selected to calibrate and verify the models, thereby realizing daily and monthly scale flow process simulation;

[0011] Step 4: Calculate the monthly and annual water conservancy project measure factor index of the basin based on the annual water conservancy project data and watershed attribute data collected in step 1 and the monthly measured flow data obtained in step 2; use the monthly and annual NDVI raster data collected in step 1 to obtain the monthly and annual NDVI average values ​​of the basin after processing; at the same time, calculate the monthly and annual runoff erosion power based on the daily and monthly simulated flow data obtained in step 3;

[0012] Step 5: Based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, and monthly and annual runoff erosion power calculated in step 4, a calibration period and a verification period are divided, and monthly and annual scale energy-sediment calculation modules corresponding to various basin hydrological models are established respectively. Then, the monthly and annual measured sediment transport data calculated in step 2 are used to calibrate and verify the energy-sediment calculation modules respectively, so as to realize the basin sediment process simulation at the monthly and annual scales.

[0013] Step 6. Based on the simulation of the monthly and annual scale basin sediment process in step 5, the Bayesian weighted average method is used to perform ensemble simulation of the monthly and annual scale sediment simulation processes, respectively, to obtain the monthly and annual sediment ensemble simulation processes, and to conduct deterministic and uncertainty accuracy evaluations.

[0014] The specific process of step 2 is:

[0015] Step 2.1: Use the daily hydrological data of the basin control hydrological station during the study period collected in step 1 to directly obtain the daily measured flow data, and calculate the monthly measured flow data, monthly measured sediment transport data, and annual measured sediment transport data:

[0016]

[0017] In the formula, Q obs,i,j is the measured flow rate in the jth month of the i-th year, m 3 / s;Q obs,i,j,k is the measured flow rate on the kth day of the jth month of the i-th year, m 3 / s; D is the number of days in the jth month of the i-th year; ssc i,j,k is the sediment content on the kth day of the jth month of the i-th year, kg / m 3 ; S obs,i,j is the measured sediment discharge in the jth month of the i-th year, kg; S obs,i is the measured sediment transport in the i-th year, kg;

[0018] Step 2.2: Calculate the daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration based on the daily precipitation data and potential evapotranspiration data of the meteorological stations in and around the study basin collected in step 1. The specific formula is as follows:

[0019]

[0020] Where P l,i,j is the precipitation of the lth meteorological station in the jth month of the ith year, mm; P l,i,j,k is the precipitation on the kth day of the jth month of the ith year at the lth weather station, mm; P i,j,k is the basin surface precipitation on the kth day of the jth month of the i-th year, mm; P i,j is the basin surface precipitation in the jth month of the i-th year, mm; PET l,i,j is the potential evapotranspiration of the lth meteorological station in the jth month of the ith year, mm; PET l,i,j,k is the potential evapotranspiration of the lth meteorological station on the kth day of the ith month of the ith year, mm; PET i,j,k is the potential evapotranspiration of the basin area on the kth day of the jth month of the i-th year, mm; PET i,j is the potential evapotranspiration of the basin in the jth month of the i-th year, mm; M is the number of meteorological stations.

[0021] The specific process of step 3 is:

[0022] Step 3.1, based on the daily and monthly measured flow data, daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration data obtained in step 2, divide the flow simulation process into a rate period and a verification period, where the length of the verification period must exceed 50% of the study period;

[0023] Step 3.2, select at least three watershed hydrological models, perform daily-scale flow simulation and monthly-scale flow simulation respectively, input the daily and monthly measured flow data, daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration data of the calibration period into the corresponding watershed hydrological model, and then use the particle swarm optimization algorithm to calibrate the parameters of each model at the daily and monthly scales, and finally use the corresponding data of the validation period to verify the calibrated parameters, where the particle swarm optimization algorithm uses the following formula:

[0024]

[0025] Where: c1 and c2 are learning factors; is the velocity of the hth particle in the dth iteration cycle; is the position of the hth particle in the dth iteration cycle; r1 and r2 are random numbers uniformly distributed in the interval (0,1); w is the inertia weight of the velocity; pBest d is the best position that all particles have passed by until the dth iteration; is the best position that the hth particle has passed by until the dth iteration;

[0026] Among them, the accuracy evaluation index for calibration and verification of basin hydrological model parameters is calculated using the Nash coefficient NSE, as shown below:

[0027]

[0028] Where: MB is the calculation time length; Q obs,md is the measured flow rate on the mdth day or month, m 3 / s;Q sim,md is the simulated flow rate of the mdth day or month, m 3 / s; is the multi-year average annual value of measured flow, m 3 / s; when NSE is higher than 0.5 in both the rate period and the verification period, it indicates that the progress of the corresponding basin hydrological model meets the requirements;

[0029] Step 3.3, use the hydrological model calibrated and verified in step 3.2 to simulate the flow process of the corresponding time scale. The specific formula is as follows:

[0030] Q sim,i,j,k,g =f g (P i,j,k ,PET i,j,k ,α day,g ) (13);

[0031] α day,g ={α day,g,1 , α day,g,2, …, α day,g,nl} (14);

[0032] Q sim,i,j,g =f g (P i,j ,PET i,j ,α month,g ) (15);

[0033] α month,g ={α month,g,1 , α month,g,2 , …, α month,g,nl} (16);

[0034] Where: Q sim,i,j,k,g is the simulated flow of the g-type basin hydrological model on the k-th day of the j-th month of the i-th year, m 3 / s;f g represents the g-th watershed hydrological model; α day,g is the parameter vector representing the daily scale of the g-th basin hydrological model; α day,g,nl is the nlth parameter representing the daily scale of the gth watershed hydrological model; Q sim,i,j,g is the simulated flow of the g-type basin hydrological model in the j-th month of the i-th year, m 3 / s; α month,g is the parameter vector representing the monthly scale of the g-th basin hydrological model; α month,g,nl is the nlth parameter representing the monthly scale of the gth watershed hydrological model.

[0035] Step 4 is as follows:

[0036] Step 4.1, based on the annual water conservancy project data and watershed attribute data collected in step 1 and the monthly measured flow data obtained in step 2, calculate the monthly and annual watershed water conservancy project measure factor index. The calculation formula of the water conservancy project measure factor index is as follows:

[0037]

[0038] Where: MRI jl A is the water conservancy project measure factor index in the jlth year or month; T is the basin area controlled by the basin control hydrological station, km 2 RI jl is the reservoir index of the jlth year or month; V ij,jl is the storage capacity of the ijth reservoir above the basin control hydrological station in the jlth year or month, m 3 MV jl A is the number of reservoirs above the control hydrological station in the basin in the jlth year or month; ij,jl The basin area controlled by the ijth reservoir above the basin control hydrological station in the jlth year or month, km2 ; is the average measured annual runoff of the basin control hydrological station or the average measured monthly runoff in different months, which can be calculated by measuring the monthly runoff, m 3 ;

[0039] Step 4.2, use the monthly and annual NDVI raster data collected in step 1, and process them to obtain the monthly and annual NDVI average values ​​of the basin:

[0040]

[0041] Where: NDVI jl is the average NDVI of the basin in the jlth year or month; NDVI jl,jk is the NDVI value of the jkth grid in the basin in the jlth year or month; MM is the number of NDVI grids in the basin;

[0042] Step 4.3, calculate the monthly and annual runoff erosion power, the formula is as follows:

[0043] E i =Q i ' ,m ·H i (20);

[0044]

[0045] Where: E i is the runoff erosion power in the i-th year, m 4 / (s·km 2 );Q sim,i,j is the simulated flow rate in the jth month of the ith year during the simulated flow process, m 3 / s; Δt is the time resolution length, s; A T is the basin area controlled by the basin control hydrological station, km 2 ;H i is the basin runoff depth in the i-th year, m; Q sim,i,m is the peak flow rate in the ith year during the simulated flow process, m 3 / s;Q i ' ,m is the peak flow modulus of the ith year in the simulated flow process, m 3 / (s·km 2 );

[0046] E j =Q j ' ,m ·H j (twenty three);

[0047]

[0048] Where: E j is the runoff erosion power in the jth month, m 4 / (s·km 2 );Q sim,j,k is the simulated flow rate on the kth day of the jth month in the simulated flow process, m 3 / s;H j is the basin runoff depth in the jth month, m; Q sim,j,m is the peak flow rate in the jth month during the simulated flow process, m 3 / s;Q j ' ,m is the peak flow modulus of the jth month in the simulated flow process, m 3 / (s·km 2 );

[0049] The specific process of step 5 is:

[0050] Step 5.1, based on the monthly and annual measured sediment transport data, monthly and annual watershed water conservancy project measures factor index, monthly and annual NDVI average values, and monthly and annual runoff erosion power calculated in step 4, divide the sediment simulation process into a rate period and a verification period, where the length of the verification period must exceed 50% of the study period, and construct a monthly and annual scale energy sediment calculation module corresponding to the watershed hydrological model in step 3. The specific formula of the energy sediment calculation module is as follows:

[0051]

[0052] Where: S sim,jl,g represents the sediment transport simulated by the energy sediment calculation module corresponding to the g-type basin hydrological model in the jl-th year or month, kg; E jl,g represents the runoff erosion power of the jlth year or month calculated by the simulated flow of the gth watershed hydrological model, m 4 / (s·km 2 );MRI jl Indicates the water conservancy project measures factor index in the jl year or month; NDVI jl represents the average NDVI of the basin in the jlth year or month; a0, b0, c0 and d0 are the parameters of the energy sand calculation module;

[0053] In step 5.2, the monthly and annual measured sediment transport data calculated in step 2 are used to calibrate and verify the energy-sediment calculation module using the particle swarm optimization algorithm. The calibration and verification evaluation index is evaluated using the Naschy coefficient NSE. When the NSE between the measured sediment transport and the simulated sediment transport is greater than 0.5, it indicates that the energy-sediment calculation module meets the accuracy requirements.

[0054] The specific process of step 6 is:

[0055] Step 6.1, based on the simulation of the monthly and annual scale basin sediment process in step 5, use the Bayesian weighted average method to perform ensemble simulation of the monthly and annual scale sediment simulation processes, and obtain the monthly and annual sediment ensemble simulation values, respectively. The specific formula of the Bayesian weighted average method is as follows:

[0056]

[0057] Where: Mod is the monthly or annual sediment load simulation value, kg; S obs Indicates the measured sediment transport value on a monthly or annual basis, kg; p(Mod g |q obs ) represents the measured sediment transport value S for a given month or year obs Under the condition, the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability reflects S obs and Mod g The matching degree between them is the weight ω of the simulation results of the energy and sediment calculation modules of the g basin hydrological models. g ;p g (Mod|Mod g ,S obs ) represents the measured sediment transport value S for a given month or year obs and the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability of the sediment load ensemble simulation value Mod under the condition of ; MS represents the number of model types;

[0058] Step 6.2, the monthly and annual sediment ensemble simulation values ​​obtained in step 6.1 are evaluated for their deterministic and uncertain accuracy, respectively. The deterministic evaluation index uses the Kling efficiency coefficient KGE, and the specific formula is as follows:

[0059]

[0060] Where: r represents the correlation coefficient between the measured sediment discharge value on a monthly or annual basis and the simulated sediment discharge ensemble value; σ sim and σ obs They represent the standard deviation of the monthly or annual sediment load ensemble simulated values ​​and the measured sediment load values ​​respectively; represents the mean of the measured sediment transport values ​​on a monthly or annual basis, kg; represents the mean of the monthly or annual sediment load ensemble simulation values, kg;

[0061] Step 6.3, using the Bayesian weighted average method in step 6.1, obtain 1000 results of the ensemble simulation of sediment in any jlth year or month in the monthly or annual scale basin sediment process simulation by random sampling 1000 times, take the results between the 0.975 and 0.025 quantiles as the 95% uncertainty interval, and evaluate the uncertainty accuracy of the uncertainty interval, including the uncertainty indicators using coverage CR, average bandwidth BR and average offset amplitude DR, a total of three indicators, the specific formula is as follows:

[0062]

[0063] Where: IA represents the number of measured data within the 95% uncertainty interval; CA represents the total number of measured data;

[0064]

[0065] Where: and are the upper and lower bounds of the 95% uncertainty interval for the jth year or month, respectively; T n Indicates the simulation time length;

[0066]

[0067] Where: S obs,jl Indicates the measured sediment transport value in the jlth year or month.

[0068] The beneficial effect of the present invention is that it not only fully considers the physical nature of the erosion process from the perspective of energetics, but also integrates the advantages of multiple models, effectively improves the accuracy of basin sediment simulation, has a solid foundation of the physical mechanism of runoff and sediment production processes, and can effectively support scientific decision-making in basin ecological environment protection and soil and water conservation planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flow chart of the basin sediment ensemble simulation method based on runoff erosion energy of the present invention;

[0070] Figure 2 It is a schematic diagram of the ensemble simulation results of monthly sediment transport in the control basin of Xiangtan Hydrological Station of Xiangjiang River from 1982 to 2018 in Example 3 of the basin sediment ensemble simulation method based on runoff erosion energy of the present invention;

[0071] Figure 3 It is a schematic diagram of the ensemble simulation results of the annual sediment transport in the controlled basin of the Xiangtan Hydrological Station of the Xiangjiang River from 1982 to 2018 in Example 3 of the basin sediment ensemble simulation method based on runoff erosion energy of the present invention. DETAILED DESCRIPTION

[0072] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0073] Example 1

[0074] The method for simulating a watershed sediment collection based on runoff erosion energy of the present invention specifically comprises the following steps:

[0075] Step 1: Determine the study basin, collect daily hydrological data of the basin control hydrological station during the study period, daily precipitation data and potential evapotranspiration data of meteorological stations inside and around the study basin, annual water conservancy project data, monthly and annual NDVI raster data and basin attribute data;

[0076] Step 2: Use the daily hydrological data of the basin control hydrological station during the study period collected in step 1, and after processing, obtain the daily and monthly measured flow data and the monthly and annual measured sediment transport data; according to the daily precipitation data and potential evapotranspiration data of the meteorological stations inside and around the study basin collected in step 1, after processing, obtain the daily and monthly surface precipitation and surface potential evapotranspiration data of the basin;

[0077] Step 2.1: Use the daily hydrological data (mainly measured daily flow and daily sediment content data) of the basin control hydrological station during the study period collected in step 1 to directly obtain the daily measured flow, and calculate the monthly measured flow data, monthly measured sediment transport data, and annual measured sediment transport data:

[0078]

[0079] In the formula, Q obs,i,j is the measured flow rate in the jth month of the i-th year, m 3 / s;Q obs,i,j,k is the measured flow rate on the kth day of the jth month of the i-th year, m 3 / s; D is the number of days in the jth month of the i-th year; ssc i,j,k is the sediment content on the kth day of the jth month of the i-th year, kg / m 3 ; S obs,i,j is the measured sediment discharge in the jth month of the i-th year, kg; S obs,i is the measured sediment transport in the i-th year, kg;

[0080] Step 2.2: Calculate the daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration based on the daily precipitation data and potential evapotranspiration data of the meteorological stations in and around the study basin collected in step 1. The specific formula is as follows:

[0081]

[0082] Where P l,i,jis the precipitation of the lth meteorological station in the jth month of the ith year, mm; P l,i,j,k is the precipitation on the kth day of the jth month of the ith year at the lth weather station, mm; P i,j,k is the basin surface precipitation on the kth day of the jth month of the i-th year, mm; P i,j is the basin surface precipitation in the jth month of the i-th year, mm; PET l,i,j is the potential evapotranspiration of the lth meteorological station in the jth month of the ith year, mm; PET l,i,j,k is the potential evapotranspiration of the lth meteorological station on the kth day of the ith month of the ith year, mm; PET i,j,k is the potential evapotranspiration of the basin area on the kth day of the jth month of the i-th year, mm; PET i,j is the potential evapotranspiration of the basin area in the jth month of the i-th year, mm; M is the number of meteorological stations;

[0083] Step 3: Based on the daily and monthly measured flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data obtained in step 2, a calibration period and a verification period are divided, and a variety of daily and monthly scale watershed hydrological models are selected to calibrate and verify the models, thereby realizing daily and monthly scale flow process simulation;

[0084] Step 3.1, based on the daily and monthly flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data obtained in step 2, divide the flow simulation process into a rate period and a validation period, where the length of the validation period must exceed 50% of the study period, select at least three basin hydrological models, perform daily scale flow simulation and monthly scale flow simulation respectively, input the daily or monthly measured flow data and surface precipitation, and surface potential evapotranspiration data of the rate period into the basin hydrological model of the corresponding time scale, and then use the particle swarm optimization algorithm to realize the parameter calibration of each model at the daily scale and the monthly scale respectively, and input the daily or monthly measured flow data and surface precipitation, and surface potential evapotranspiration data of the validation period into the basin hydrological model of the corresponding time scale for parameter verification, where the basic formula of the particle swarm optimization algorithm is as follows:

[0085]

[0086] Where: c1 and c2 are learning factors that can adjust the convergence speed of the algorithm; is the velocity of the hth particle in the dth iteration cycle; is the position of the hth particle in the dth iteration cycle; r1 and r2 are random numbers uniformly distributed in the interval (0,1); w is the inertia weight of the velocity; pBest is the best position that the hth particle has passed by until the dth iteration; d is the best position that all particles have passed by until the dth iteration;

[0087] Among them, the accuracy evaluation index for calibration and verification of basin hydrological model parameters is calculated using the Nash coefficient NSE, as shown below:

[0088]

[0089] Where: MB is the calculation time length; Q obs,md is the measured flow rate on the mdth day or month, m 3 / s;Q sim,md is the simulated flow rate of the mdth day or month, m 3 / s; is the multi-year average annual value of measured flow, m 3 / s. When NSE is higher than 0.5 in both the rate period and the validation period, the model is considered to meet the progress requirements;

[0090] Step 3.2: Based on the various daily and monthly scale basin hydrological models calibrated and verified in step 3.1, simulate the flow process of the corresponding time scale. The specific formula is as follows:

[0091] Q sim,i,j,k,g =f g (P i,j,k ,PET i,j,k ,α day,g ) (13);

[0092] α day,g ={α day,g,1 , α day,g,2 , …, α day,g,nl} (14);

[0093] Q sim,i,j,g =f g (P i,j ,PET i,j ,α month,g ) (15);

[0094] α month,g ={α month,g,1 , α month,g,2 , …, α month,g,nl} (16);

[0095] Where: Q sim,i,j,k,g is the simulated flow of the g-type basin hydrological model on the k-th day of the j-th month of the i-th year, m 3 / s;f g represents the g-th watershed hydrological model; α day,g is the parameter vector representing the daily scale of the g-th basin hydrological model; α day,g,nl is the nlth parameter representing the daily scale of the gth watershed hydrological model; Q sim,i,j,g is the simulated flow of the g-type basin hydrological model in the j-th month of the i-th year, m3 / s; α month,g is the parameter vector representing the monthly scale of the g-th basin hydrological model; α month,g,nl is the nlth parameter representing the monthly scale of the gth watershed hydrological model;

[0096] Step 4: Calculate the monthly and annual water conservancy project measure factor index of the basin based on the annual water conservancy project data and watershed attribute data collected in step 1 and the monthly measured flow data obtained in step 2; use the monthly and annual NDVI raster data collected in step 1 to obtain the monthly and annual NDVI average values ​​of the basin after processing; at the same time, calculate the monthly and annual runoff erosion power based on the daily and monthly simulated flow data obtained in step 3;

[0097] Step 4.1: Based on the annual water conservancy project data (mainly reservoirs and silt dams) and watershed attribute data (mainly watershed area) collected in step 1 and the monthly measured flow data obtained in step 2, the monthly and annual watershed water conservancy project measures factor index is calculated. The calculation formula of the water conservancy project measures factor index is as follows:

[0098]

[0099] Where: MRI jl A is the water conservancy project measure factor index in the jlth year or month; T is the basin area controlled by the basin control hydrological station, km 2 RI jl is the reservoir index of the jlth year or month; V ij,jl is the storage capacity of the ijth reservoir above the basin control hydrological station in the jlth year or month, m 3 MV jl A is the number of reservoirs above the control hydrological station in the basin in the jlth year or month; ij,jl The basin area controlled by the ijth reservoir above the basin control hydrological station in the jlth year or month, km 2 ; The multi-year average measured annual runoff of the basin control hydrological station or the multi-year average measured monthly runoff of different months can be obtained by summing the time of the monthly measured flow data and taking the average, m 3 . Usually, the reservoir index is calculated mainly using the data of large reservoirs or backbone check dams in the basins controlled by various hydrological stations;

[0100] Step 4.2, use the monthly and annual NDVI raster data collected in step 1, and process them to obtain the monthly and annual NDVI average values ​​of the basin:

[0101]

[0102] Where: NDVIjl is the average NDVI of the basin in the jlth year or month; NDVI jl,jk is the NDVI value of the jkth grid in the basin in the jlth year or month; MM is the number of NDVI grids in the basin;

[0103] Step 4.3, at the same time, based on the daily and monthly flow processes simulated by the hydrological models of multiple watersheds in step 3, the monthly and annual runoff erosion power is calculated. The runoff erosion power index used here has the dimension of energy and can quantify the size of the runoff erosion energy. The calculation formula is as follows:

[0104] E i =Q i ' ,m ·H i (20);

[0105]

[0106]

[0107] Where: E i is the runoff erosion power in the i-th year, m 4 / (s·km 2 );Q sim,i,j is the simulated flow rate in the jth month of the ith year during the simulated flow process, m 3 / s; Δt is the time resolution length, s; A T is the basin area controlled by the basin control hydrological station, km 2 ;H i is the basin runoff depth in the i-th year, m; Q sim,i,m is the peak flow rate in the ith year during the simulated flow process, m 3 / s;Q i ' ,m is the peak flow modulus of the ith year in the simulated flow process, m 3 / (s·km 2 );

[0108] E j =Q j ' ,m ·H j (twenty three);

[0109]

[0110] Where: E j is the runoff erosion power in the jth month, m 4 / (s·km 2 );Q sim,j,k is the simulated flow rate on the kth day of the jth month in the simulated flow process, m3 / s;H j is the basin runoff depth in the jth month, m; Q sim,j,m is the peak flow rate in the jth month during the simulated flow process, m 3 / s;Q j ' ,m is the peak flow modulus of the jth month in the simulated flow process, m 3 / (s·km 2 );

[0111] Step 5: Based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, monthly and annual runoff erosion power, and monthly and annual measured sediment transport data calculated in step 4, a calibration period and a verification period are divided, and monthly and annual scale energy-sediment calculation modules corresponding to various basin hydrological models are established respectively. Then, the monthly and annual measured sediment transport data calculated in step 2 are used to calibrate and verify the energy-sediment calculation modules respectively, so as to realize the simulation of basin sediment processes at the monthly and annual scales.

[0112] Step 5.1, based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, monthly and annual runoff erosion power, and monthly and annual measured sediment transport data calculated in step 4, divide the sediment simulation process into a rate period and a verification period, where the length of the verification period must exceed 50% of the study period, and establish monthly and annual scale energy sediment calculation modules corresponding to various basin hydrological models. The specific formula of the energy sediment calculation module is as follows:

[0113]

[0114] Where: S sim,jl,g represents the sediment transport simulated by the energy sediment calculation module corresponding to the g-type basin hydrological model in the jl-th year or month, kg; E jl,g represents the runoff erosion power of the jlth year or month calculated by the simulated flow of the gth watershed hydrological model, m 4 / (s·km 2 );MRI jl Indicates the water conservancy project measures factor index in the jl year or month; NDVI jl represents the average NDVI of the basin in the jlth year or month; a0, b0, c0 and d0 are the parameters of the energy sand calculation module;

[0115] Step 5.2, using the monthly and annual measured sediment discharge data calculated in step 2, the particle swarm optimization algorithm is used to calibrate and verify the energy-sediment calculation module. The calibration and verification evaluation index is still evaluated by the Nash coefficient NSE. When the NSE between the measured sediment discharge and the simulated sediment discharge is greater than 0.5, it is considered that the energy-sediment calculation module meets the accuracy requirements;

[0116] Step 6. Based on the simulation of the monthly and annual scale basin sediment process in step 5, the Bayesian weighted average method is used to perform ensemble simulation of the monthly and annual scale sediment simulation processes, respectively, to obtain the monthly and annual sediment ensemble simulation processes, and to conduct deterministic and uncertainty accuracy evaluations.

[0117] Example 2

[0118] Based on Example 1, the specific process of step 6 is as follows:

[0119] Step 6.1, based on the simulation of the monthly and annual scale basin sediment process in step 5, use the Bayesian weighted average method to perform ensemble simulation of the monthly and annual scale sediment simulation processes, and obtain the monthly and annual sediment ensemble simulation values, respectively. The specific formula of the Bayesian weighted average method is as follows:

[0120]

[0121] Where: Mod is the monthly or annual sediment load simulation value, kg; S obs Indicates the measured sediment transport value on a monthly or annual basis, kg; p(Mod g |q obs ) represents the measured sediment transport value S for a given month or year obs Under the condition, the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability of S obs and Mod g The matching degree between them is the weight ω of the simulation results of the energy and sediment calculation modules of the g basin hydrological models. g , the larger the weight, the higher the simulation accuracy of the single model; p g (Mod|Mod g ,S obs ) represents the measured sediment transport value S for a given month or year obs and the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability of the sediment load ensemble simulation value Mod under the condition of ; MS represents the number of model types;

[0122] Step 6.2, the monthly and annual sediment ensemble simulation values ​​obtained in step 6.1 are evaluated for deterministic and uncertain accuracy respectively. The deterministic evaluation index uses the Kling efficiency coefficient KGE. The larger the value, the higher the simulation accuracy. The specific formula is as follows:

[0123]

[0124] Where: r represents the correlation coefficient between the measured sediment discharge value on a monthly or annual basis and the simulated sediment discharge ensemble value; σ simand σ obs They represent the standard deviation of the monthly or annual sediment load ensemble simulated values ​​and the measured sediment load values ​​respectively; represents the mean of the measured sediment transport values ​​on a monthly or annual basis, kg; represents the mean of the monthly or annual sediment load ensemble simulation values, kg;

[0125] Step 6.3, using the Bayesian weighted average method in step 6.1, 1000 results of sediment ensemble simulation of any jlth year or month in the monthly or annual scale basin sediment process simulation can be obtained by random sampling 1000 times. The results between the 0.975 and 0.025 quantiles are taken as the 95% uncertainty interval, and the uncertainty accuracy is evaluated respectively. The uncertainty indicators use three indicators: coverage rate CR, average bandwidth BR and average offset amplitude DR. The specific formula is as follows:

[0126]

[0127] Where: IA represents the number of measured data within the 95% uncertainty interval; CA represents the total number of measured data; the larger the CR, the smaller the uncertainty of the simulation results.

[0128]

[0129] Where: and are the upper and lower bounds of the 95% uncertainty interval for the jth year or month, respectively; T n Indicates the length of simulation time; the smaller BR is, the narrower the uncertainty interval of the simulation results is and the higher the simulation accuracy is.

[0130]

[0131] Where: S obs,jl Represents the measured sediment transport value in the jlth year or month; the smaller the DR, the better the symmetry of the uncertainty interval.

[0132] Example 3

[0133] Taking the Xiangtan hydrological station control basin of Xiangjiang River as an example, the technical process is as follows: Figure 1 As shown, the specific implementation steps are as follows:

[0134] Step 1: Collect the daily measured flow and sediment content data of the Xiangjiang River Xiangtan hydrological station from 1982 to 2018, the daily precipitation data and potential evapotranspiration data of 51 meteorological stations in and around the basin, the construction year, storage capacity and controlled basin area data of 17 large reservoirs, the monthly and annual NDVI raster data with a spatial resolution of 5 km and the basin area of ​​81638 km 2 ;

[0135] Step 2: Use the daily measured flow and sediment content data of the Xiangjiang Xiangtan hydrological station from 1982 to 2018 collected in step 1 to directly obtain the daily measured flow from 1982 to 2018, and after processing, obtain the monthly measured flow data and monthly and annual measured sediment transport data from 1982 to 2018; according to step 1, collect the daily precipitation data and potential evapotranspiration data of 51 meteorological stations in and around the control basin of the Xiangjiang Xiangtan hydrological station from 1982 to 2018, and after processing, obtain the daily and monthly surface precipitation and surface potential evapotranspiration data of the basin;

[0136] Step 2.1, using the daily measured flow and sediment content data of the Xiangtan hydrological station of the Xiangjiang River from 1982 to 2018 collected in step 1, directly obtain the daily measured flow from 1982 to 2018, and after processing, obtain the monthly measured flow data and the monthly and annual measured sediment transport data, and use the above formulas (1) to (3) to calculate the monthly measured flow, monthly and annual measured sediment transport;

[0137] In this embodiment, in formulas (1) to (3), i refers to any year from 1982 to 2018, and j = 1 to 12;

[0138] Step 2.2, based on the daily precipitation data and potential evapotranspiration data of 51 meteorological stations in and around the Xiangjiang Xiangtan hydrological station control basin from 1982 to 2018 collected in step 1, the above formula (4) to formula (9) are used to calculate the daily and monthly surface precipitation and surface potential evapotranspiration data of the basin; in this embodiment, in the calculation process of formulas (4) to (9), l=1~51;

[0139] Step 3: Based on the daily and monthly measured flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data of the Xiangjiang Xiangtan hydrological station from 1982 to 2018 obtained in step 2, the daily scale flow simulation rate period and verification period are divided into January 1, 1982 to December 31, 1999 and January 1, 2000 to December 31, 2018, respectively, and the monthly scale flow simulation rate period and verification period are January 1982 to December 1999 and January 2000 to December 2018, respectively. Four daily scale and monthly scale basin hydrological models are constructed, the models are calibrated and verified, and the four models are used to simulate the daily scale and monthly scale flow processes respectively;

[0140] Step 3.1: Based on the daily and monthly measured flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data of the Xiangjiang Xiangtan hydrological station from 1982 to 2018 obtained in step 2, the regular and validation periods of daily-scale flow simulation rate are divided into January 1, 1982 to December 31, 1999 and January 1, 2000 to December 31, 2018, respectively, and the regular and validation periods of monthly-scale flow simulation rate are divided into January 1982 to December 1999 and January 2000 to December 2018, respectively. Then, four daily-scale and monthly-scale basin hydrological models are constructed respectively. The four basin hydrological models are Xin'anjiang model, Hymod model, GR4J model and SCS model. The daily or monthly measured flow data and surface precipitation and surface potential evapotranspiration data of the calibration period are input into the basin hydrological model of the corresponding time scale. The particle swarm optimization algorithm is then used to calibrate the parameters of each model at the daily and monthly scales. The daily or monthly measured flow data and surface precipitation and surface potential evapotranspiration data of the validation period are input into the basin hydrological model of the corresponding time scale for parameter verification. The particle swarm optimization algorithm uses the above formulas (10) to (11) for calculation.

[0141] Among them, the accuracy evaluation index of the calibration and verification of the basin hydrological model parameters is calculated using the Nash coefficient NSE, and the calculation process is calculated using the above formula (12); when NSE is higher than 0.5 in both the calibration period and the verification period, the model progress is considered to meet the requirements; the results show that at the daily scale, the NSE of the four basin hydrological models in the calibration period and the verification period are 0.69, 0.85, 0.83, 0.58 and 0.74, 0.86, 0.86, 0.53 respectively. At the monthly scale, the NSE of the four basin hydrological models in the calibration period and the verification period are 0.92, 0.96, 0.97, 0.62 and 0.94, 0.96, 0.96, 0.65 respectively. The four models meet the accuracy requirements in the daily and monthly scale flow simulation of the Xiangtan Hydrological Station.

[0142] Step 3.2, based on the four daily-scale and monthly-scale basin hydrological models calibrated and verified in step 3.1, simulate the daily and monthly flow processes from 1982 to 2018. The simulation process adopts the above formula (13) to formula (16). In this embodiment, g = 1 to 4.

[0143] Step 4: Based on the data collected in step 1, the construction year, storage capacity and controlled basin area of ​​17 large reservoirs, with a basin area of ​​81,638 km 2The monthly measured flow of the Xiangjiang River Xiangtan Hydrological Station from 1982 to 2018 obtained in step 2 is used to calculate the monthly and annual basin water conservancy project measures factor index; the monthly and annual 5km spatial resolution NDVI raster data of the Xiangjiang River Xiangtan Hydrological Station from 1982 to 2018 collected in step 1 are used to process the average monthly and annual NDVI values ​​of the basin; at the same time, the daily and monthly simulated flow data of the four basin hydrological models obtained in step 3 are simulated to calculate the monthly and annual runoff erosion power;

[0144] Step 4.1: Based on the data collected in step 1 on the construction year, storage capacity and controlled basin area of ​​17 large reservoirs, the basin area is 81,638 km 2 The monthly measured flow of Xiangjiang River Xiangtan hydrological station from 1982 to 2018 obtained in step 2 is used to calculate the monthly and annual basin water conservancy project measures factor index. The water conservancy project measures factor index is shown in the above formulas (17) to (18); where A T =81638km 2 ;

[0145] Step 4.2: Use the monthly and annual 5 km spatial resolution NDVI raster data of the Xiangtan hydrological station of the Xiangjiang River from 1982 to 2018 collected in step 1 to calculate the monthly and annual NDVI average values ​​of the basin using the above formula (19);

[0146] Step 4.3, at the same time, based on the daily and monthly flow processes from 1982 to 2018 simulated by the four watershed hydrological models in step 3, the monthly and annual runoff erosion powers are calculated using the above formulas (20) to (25);

[0147] Step 5: Based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, monthly and annual runoff erosion power, and monthly and annual measured sediment transport data from 1982 to 2018 calculated in step 4, the monthly scale sediment simulation rate regular and verification periods are divided into January 1982 to December 1999 and January 2000 to December 2018, and the annual scale sediment simulation rate regular and verification periods are 1982 to 1999 and 2000 to 2018, respectively. The monthly and annual scale energy sediment calculation modules corresponding to the four basin hydrological models are established respectively. Then, the monthly and annual measured sediment transport data from 1982 to 2018 calculated in step 2 are used to calibrate and verify the energy sediment calculation modules respectively, so as to realize the basin sediment process simulation at the monthly and annual scales.

[0148] Step 5.1, based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, monthly and annual runoff erosion power, and monthly and annual measured sediment transport data from 1982 to 2018 calculated in step 4, the monthly scale sediment simulation rate regular and verification period is divided into January 1982 to December 1999 and January 2000 to December 2018, and the annual scale sediment simulation rate regular and verification period is divided into 1982 to 1999 and 2000 to 2018, respectively. The monthly scale and annual scale energy sediment calculation modules corresponding to the four basin hydrological models are established respectively. The established energy sediment calculation modules are shown in the above formula (26);

[0149] In step 5.2, the measured monthly and annual sediment discharge data from 1982 to 2018 calculated in step 2 are used to calibrate and verify the energy-sediment calculation module using the particle swarm optimization algorithm. The calibration and verification evaluation indicators are still evaluated using the Naschy coefficient NSE. When the NSE between the measured sediment discharge and the simulated sediment discharge is greater than 0.5, the energy-sediment calculation module is considered to meet the accuracy requirements. The results show that at the monthly scale, the NSE of the energy-sediment calculation modules corresponding to the four watershed hydrological models during the calibration period and the verification period are 0.81, 0.84, 0.82, 0.57 and 0.58, 0.64, 0.60, 0.53, respectively. At the annual scale, the NSEs of the energy-sediment calculation modules corresponding to the four watershed hydrological models during the rate period and the verification period were 0.70, 0.73, 0.74, 0.63 and 0.57, 0.52, 0.62, 0.53, respectively. The energy-sediment calculation modules corresponding to the four watershed hydrological models met the accuracy requirements in the monthly and annual scale sediment transport simulations of the Xiangtan Hydrological Station.

[0150] Step 6: Based on the simulation of the monthly and annual scale basin sediment process in step 5, the Bayesian weighted average method is used to perform ensemble simulation of the monthly and annual scale sediment simulation processes, respectively, to obtain the monthly and annual sediment ensemble simulation processes from 1982 to 2018, and to conduct deterministic and uncertain accuracy evaluations;

[0151] Step 6.1: Based on the simulation of the monthly and annual scale basin sediment processes in step 5, the Bayesian weighted average method is used to perform ensemble simulation on the four monthly scale sediment simulation results with the rate period and validation period from January 1982 to December 1999 and from January 2000 to December 2018, and the four annual scale sediment simulation results with the rate period and validation period from 1982 to 1999 and from 2000 to 2018. The monthly and annual sediment ensemble simulation values ​​from 1982 to 2018 are obtained using the above formula (27) according to the Bayesian weighted average method. MS represents the number of model types, and its value is 4.

[0152] Step 6.2, the sediment ensemble simulation values ​​of the monthly and annual rates and the verification period obtained in step 6.1 are evaluated for their deterministic and uncertain accuracy, respectively, where the deterministic evaluation index uses the Kling efficiency coefficient KGE using the above formula (28);

[0153] The results show that at the monthly scale, the KGE of sediment ensemble simulation during the rate period and the validation period are 0.73 and 0.67 respectively; at the annual scale, the KGE of sediment ensemble simulation during the rate period and the validation period are 0.75 and 0.73 respectively. This shows that the ensemble simulation integrates the results of each model and effectively improves the reliability of the simulation results.

[0154] Step 6.3, using the Bayesian weighted average method in step 6.1, 1000 results of the ensemble simulation of sediment in any jl year or month in the monthly or annual scale basin sediment process simulation can be obtained by random sampling 1000 times. The results between the 0.975 and 0.025 quantiles are taken as the 95% uncertainty intervals, and the uncertainty accuracy is evaluated respectively. The monthly and annual sediment ensemble simulation results and 95% uncertainty intervals of Xiangtan Hydrological Station from 1982 to 2018 are shown as follows: Figure 2 and Figure 3 As shown. The uncertainty index uses three indicators, namely, coverage ratio CR, average bandwidth BR and average deviation amplitude DR. The three indicators use the above formulas (29) to (31);

[0155] In formula (29), IA represents the number of measured data within the 95% uncertainty interval; CA represents the total number of measured data; the larger the CR, the smaller the uncertainty of the simulation results. The results show that on a monthly scale, the CR of the sediment ensemble simulation during the rate period and the validation period are 0.977 and 0.991, respectively; on an annual scale, the CR of the sediment ensemble simulation during the rate period and the validation period are 0.944 and 0.999, respectively. This shows that the ensemble simulation effectively reduces the uncertainty of the simulation results.

[0156] In formula (30): and are the upper and lower bounds of the 95% uncertainty interval for the jth year or month, respectively; T n Indicates the length of simulation time; the smaller the BR, the narrower the uncertainty interval of the simulation results and the higher the simulation accuracy. The results show that on a monthly scale, the BR of the sediment ensemble simulation during the rate period and the verification period are 0.02 and 0.01, respectively; on an annual scale, the BR of the sediment ensemble simulation during the rate period and the verification period are 0.063 and 0.062, respectively. This shows that the ensemble simulation effectively reduces the uncertainty of the simulation results.

[0157] In formula (31): S obs,jlrepresents the measured sediment transport value in the jlth year or month; the smaller the DR, the better the symmetry of the uncertainty interval; the results show that on a monthly scale, the DRs of sediment ensemble simulations during the rate period and the validation period are 0.003 and 0.002, respectively; on an annual scale, the DRs of sediment ensemble simulations during the rate period and the validation period are 0.011 and 0.009, respectively. This shows that ensemble simulation effectively reduces the uncertainty of the simulation results.

Claims

1. A basin sediment ensemble simulation method based on runoff erosion energy, characterized by: The specific process includes the following: collecting hydrological data, precipitation and potential evapotranspiration data, water conservancy project data, Normalized Difference Vegetation Index NDVI raster data and watershed attribute data for the corresponding study period of the study basin; calculating the daily and monthly measured flow data, surface precipitation data and surface potential evapotranspiration data, as well as the monthly and annual measured sediment transport data; Multiple basin hydrological models at daily and monthly scales were constructed respectively; the monthly and annual basin water conservancy project measure factor index, NDVI average value and runoff erosion power were calculated; monthly and annual scale energy-sediment calculation modules corresponding to multiple basin hydrological models were established respectively; the Bayesian weighted average method was used to perform ensemble simulation of the monthly and annual scale sediment simulation processes respectively, and the accuracy of the simulation results was evaluated.

2. The method for simulating river basin sediment collection based on runoff erosion energy according to claim 1 is characterized by: The specific steps include: Step 1: Collect daily hydrological data of the basin control hydrological station during the study period, daily precipitation data and potential evapotranspiration data of meteorological stations inside and around the study basin, annual water conservancy project data, monthly and annual NDVI raster data and basin attribute data; Step 2: Use the daily hydrological data of the basin control hydrological station during the study period collected in step 1, and after processing, obtain the daily and monthly measured flow data and the monthly and annual measured sediment transport data; according to the daily precipitation data and potential evapotranspiration data of the meteorological stations inside and around the study basin collected in step 1, after processing, obtain the daily and monthly surface precipitation and surface potential evapotranspiration data of the basin; Step 3: Based on the daily and monthly measured flow data, daily and monthly surface precipitation, and surface potential evapotranspiration data obtained in step 2, a calibration period and a verification period are divided, and a variety of daily and monthly scale watershed hydrological models are selected to calibrate and verify the models, thereby realizing daily and monthly scale flow process simulation; Step 4: Calculate the monthly and annual water conservancy project measure factor index of the basin based on the annual water conservancy project data and watershed attribute data collected in step 1 and the monthly measured flow data obtained in step 2; use the monthly and annual NDVI raster data collected in step 1 to obtain the monthly and annual NDVI average values ​​of the basin after processing; at the same time, calculate the monthly and annual runoff erosion power based on the daily and monthly simulated flow data obtained in step 3; Step 5: Based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, and monthly and annual runoff erosion power calculated in step 4, a calibration period and a verification period are divided, and monthly and annual scale energy-sediment calculation modules corresponding to various basin hydrological models are established respectively. Then, the monthly and annual measured sediment transport data calculated in step 2 are used to calibrate and verify the energy-sediment calculation modules respectively, so as to realize the basin sediment process simulation at the monthly and annual scales. Step 6. Based on the simulation of the monthly and annual scale basin sediment process in step 5, the Bayesian weighted average method is used to perform ensemble simulation of the monthly and annual scale sediment simulation processes, respectively, to obtain the monthly and annual sediment ensemble simulation processes, and to conduct deterministic and uncertainty accuracy evaluations.

3. The method for simulating watershed sediment collection based on runoff erosion energy according to claim 2 is characterized in that: The specific process of step 2 is: Step 2.1, using the daily hydrological data of the basin control hydrological station during the study period collected in step 1, directly obtain the daily measured flow data, calculate the monthly measured flow data, monthly measured sediment transport data and annual measured sediment transport data: In the formula, Q obs,i,j is the measured flow rate in the jth month of the i-th year, m 3 / s;Q obs,i,j,k is the measured flow rate on the kth day of the jth month of the i-th year, m 3 / s; D is the number of days in the jth month of the i-th year; ssc i,j,k is the sediment content on the kth day of the jth month of the i-th year, kg / m 3 ; S obs,i,j is the measured sediment discharge in the jth month of the i-th year, kg; S obs,i is the measured sediment transport in the i-th year, kg; Step 2.2: Calculate the daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration based on the daily precipitation data and potential evapotranspiration data of the meteorological stations in and around the study basin collected in step 1. The specific formula is as follows: Where P l,i,j is the precipitation of the lth meteorological station in the jth month of the ith year, mm; P l,i,j,k is the precipitation on the kth day of the jth month of the ith year at the lth weather station, mm; P i,j,k is the basin surface precipitation on the kth day of the jth month of the i-th year, mm; P i,j is the basin surface precipitation in the jth month of the i-th year, mm; PET l,i,j is the potential evapotranspiration of the lth meteorological station in the jth month of the ith year, mm; PET l,i,j,k is the potential evapotranspiration of the lth meteorological station on the kth day of the ith month of the ith year, mm; PET i,j,k is the potential evapotranspiration of the basin area on the kth day of the jth month of the i-th year, mm; PET i,j is the potential evapotranspiration of the basin in the jth month of the i-th year, mm; M is the number of meteorological stations.

4. The method for simulating watershed sediment collection based on runoff erosion energy according to claim 3 is characterized in that: The specific process of step 3 is as follows: Step 3.1, based on the daily and monthly measured flow, daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration data obtained in step 2, divide the flow simulation process into a rate period and a verification period, where the length of the verification period must exceed 50% of the study period; Step 3.2, select at least three watershed hydrological models, perform daily-scale flow simulation and monthly-scale flow simulation respectively, input the daily and monthly measured flow, daily surface precipitation, monthly surface precipitation, daily surface potential evapotranspiration, and monthly surface potential evapotranspiration data of the calibration period into the corresponding watershed hydrological model, and then use the particle swarm optimization algorithm to calibrate the parameters of each model at the daily and monthly scales, and finally use the corresponding data of the validation period to verify the calibrated parameters, where the particle swarm optimization algorithm uses the following formula: Where: c1 and c2 are learning factors; is the velocity of the hth particle in the dth iteration cycle; is the position of the hth particle in the dth iteration cycle; r1 and r2 are random numbers uniformly distributed in the interval (0,1); w is the inertia weight of the velocity; pBest is the best position that the hth particle has passed by until the dth iteration; d is the best position that all particles have passed by until the dth iteration; Among them, the accuracy evaluation index for calibration and verification of basin hydrological model parameters is calculated using the Nash coefficient NSE, as shown below: Where: MB is the calculation time length; Q obs,md is the measured flow rate on the mdth day or month, m 3 / s;Q sim,md is the simulated flow rate of the mdth day or month, m 3 / s; is the multi-year average annual value of measured flow, m 3 / s; when NSE is higher than 0.5 in both the rate period and the verification period, it indicates that the progress of the corresponding basin hydrological model meets the requirements; Step 3.3, use the hydrological model calibrated and verified in step 3.2 to simulate the flow process of the corresponding time scale. The specific formula is as follows: Q sim,i,j,k,g =f g (P i,j,k ,PET i,j,k ,α day,g ) (13); α day,g ={α day,g,1 ,α day,g,2 ,…,α day,g,nl} (14); Q sim,i,j,g =f g (P i,j ,PET i,j ,α month,g ) (15); α month,g ={α month,g,1 ,α month,g,2 ,…,α month,g,nl} (16); Where: Q sim,i,j,k,g is the simulated flow of the g-type basin hydrological model on the k-th day of the j-th month of the i-th year, m 3 / s;f g represents the g-th watershed hydrological model; α day,g is the parameter vector representing the daily scale of the g-th basin hydrological model; α day,g,nl is the nlth parameter representing the daily scale of the gth watershed hydrological model; Q sim,i,j,g is the simulated flow of the g-type basin hydrological model in the j-th month of the i-th year, m 3 / s; α month,g is the parameter vector representing the monthly scale of the g-th basin hydrological model; α month,g,nl is the nlth parameter representing the monthly scale of the gth watershed hydrological model.

5. The method for simulating river basin sediment collection based on runoff erosion energy according to claim 4 is characterized in that: The step 4 is specifically as follows: Step 4.1, based on the annual water conservancy project data and watershed attribute data collected in step 1 and the monthly measured flow data obtained in step 2, calculate the monthly and annual watershed water conservancy project measure factor index. The calculation formula of the water conservancy project measure factor index is as follows: Where: MRI jl is the water conservancy project measure factor index in the jlth year or month; RI jl is the reservoir index of the jlth year or month; MV jl A is the number of reservoirs above the control hydrological station in the basin in the jlth year or month; ij,jl The basin area controlled by the ijth reservoir above the basin control hydrological station in the jlth year or month, km 2 ; A T is the basin area controlled by the basin control hydrological station, km 2 ; V ij,jl is the storage capacity of the ijth reservoir above the basin control hydrological station in the jlth year or month, m 3 ; is the average measured annual runoff of the basin control hydrological station or the average measured monthly runoff of different months, m 3 ; Step 4.2, use the monthly and annual NDVI raster data collected in step 1, and process them to obtain the monthly and annual NDVI average values ​​of the basin: Where: NDVI jl is the average NDVI of the basin in the jlth year or month; NDVI jl,jk is the NDVI value of the jkth grid in the basin in the jlth year or month; MM is the number of NDVI grids in the basin; Step 4.3, calculate the monthly and annual runoff erosion power, the formula is as follows: E i =Q i ′ ,m ·H i (20); Where: E i is the runoff erosion power in the i-th year, m 4 / (s·km 2 );Q sim,i,j is the simulated flow rate in the jth month of the ith year during the simulated flow process, m 3 / s; Δt is the time resolution length, s; A T is the basin area controlled by the basin control hydrological station, km 2 ; H i is the basin runoff depth in the i-th year, m; Q sim,i,m is the peak flow rate in the ith year during the simulated flow process, m 3 / s;Q i ' ,m is the peak flow modulus of the ith year in the simulated flow process, m 3 / (s·km 2 ); E j =Q j ′ ,m ·H j (23); Where: E j is the runoff erosion power in the jth month, m 4 / (s·km 2 );Q sim,j,k is the simulated flow rate on the kth day of the jth month in the simulated flow process, m 3 / s;H j is the basin runoff depth in the jth month, m; Q sim,j,m is the peak flow rate in the jth month during the simulated flow process, m 3 / s;Q j ' ,m is the peak flow modulus of the jth month in the simulated flow process, m 3 / (s·km 2 ).

6. The method for simulating river basin sediment collection based on runoff erosion energy according to claim 5 is characterized in that: The specific process of step 5 is as follows: Step 5.1, based on the monthly and annual basin water conservancy project measure factor index, monthly and annual NDVI average values, monthly and annual runoff erosion power, and monthly and annual measured sediment transport calculated in step 4, divide the sediment simulation process into a rate period and a verification period, where the length of the verification period must exceed 50% of the study period, and construct a monthly and annual scale energy sediment calculation module corresponding to the basin hydrological model in step 3. The specific formula of the energy sediment calculation module is as follows: Where: S sim,jl,g represents the sediment transport simulated by the energy sediment calculation module corresponding to the g-type basin hydrological model in the jl-th year or month, kg; E jl,g represents the runoff erosion power of the jlth year or month calculated by the simulated flow of the gth watershed hydrological model, m 4 / (s·km 2 );MRI jl Indicates the water conservancy project measures factor index in the jl year or month; NDVI jl represents the average NDVI of the basin in the jlth year or month; a0, b0, c0 and d0 are the parameters of the energy sand calculation module; In step 5.2, the monthly and annual measured sediment transport data calculated in step 2 are used to calibrate and verify the energy-sediment calculation module using the particle swarm optimization algorithm. The calibration and verification evaluation index is evaluated using the Naschy coefficient NSE. When the NSE between the measured sediment transport and the simulated sediment transport is greater than 0.5, it indicates that the energy-sediment calculation module meets the accuracy requirements.

7. The method for simulating watershed sediment collection based on runoff erosion energy according to claim 6 is characterized in that: The specific process of step 6 is as follows: Step 6.1, based on the simulation of the monthly and annual scale basin sediment process in step 5, use the Bayesian weighted average method to perform ensemble simulation of the monthly and annual scale sediment simulation processes, and obtain the monthly and annual sediment ensemble simulation values, respectively. The specific formula of the Bayesian weighted average method is as follows: Where: Mod is the monthly or annual sediment load simulation value, kg; S obs Indicates the measured sediment transport value on a monthly or annual basis, kg; p(Mod g |q obs ) represents the measured sediment transport value S for a given month or year obs Under the condition, the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability reflects S obs and Mod g The matching degree between them is the weight ω of the simulation results of the energy and sediment calculation modules of the g basin hydrological models. g ;p g (Mod|Mod g ,S obs ) represents the measured sediment transport value S for a given month or year obs and the simulation results of the g-th basin hydrological model energy and sediment calculation module Mod g The posterior probability of the ensemble simulated value Mod of sediment discharge under the condition of ; MS represents the number of model types; Step 6.2, the monthly and annual sediment ensemble simulation values ​​obtained in step 6.1 are evaluated for their deterministic and uncertain accuracy, respectively. The deterministic evaluation index uses the Kling efficiency coefficient KGE, and the specific formula is as follows: Where: r represents the correlation coefficient between the measured sediment discharge value on a monthly or annual basis and the simulated sediment discharge ensemble value; σ sim and σ obs They represent the standard deviation of the monthly or annual sediment load ensemble simulated values ​​and the measured sediment load values ​​respectively; represents the mean of the measured sediment transport values ​​on a monthly or annual basis, kg; represents the mean of the monthly or annual sediment load ensemble simulation values, kg; Step 6.3, using the Bayesian weighted average method in step 6.1, obtain 1000 results of the ensemble simulation of sediment in any jlth year or month in the monthly or annual scale basin sediment process simulation by random sampling 1000 times, take the results between the 0.975 and 0.025 quantiles as the 95% uncertainty interval, and evaluate the uncertainty accuracy of the uncertainty interval, including the uncertainty indicators using coverage CR, average bandwidth BR and average offset amplitude DR, a total of three indicators, the specific formula is as follows: Where: IA represents the number of measured data within the 95% uncertainty interval; CA represents the total number of measured data; Where: and are the upper and lower bounds of the 95% uncertainty interval for the jth year or month, respectively; T n Indicates the simulation time length; Where: S obs,jl Indicates the measured sediment transport value in the jlth year or month.