Method for calculating lake excess flood diversion capacity

Abstract

The application provides a calculation method of lake excess flood diversion capacity, and belongs to the technical field of water conservancy and flood control, and comprises the following steps: dividing the lake into zones, constructing a lake zoning series flood routing model through a river reach water balance equation and a relationship formula of storage capacity and outflow; constructing a runoff and confluence coupling model according to a runoff formation process, a runoff process and a confluence process; calculating the runoff of each zone of the lake through the runoff and confluence coupling model; calculating the storage capacity and the outflow of each zone through the runoff of each zone of the lake and the lake zoning series flood routing model; and evaluating the excess flood diversion capacity of each zone of the lake according to the storage capacity and the outflow of each zone. The application can quantitatively evaluate the lake flood storage capacity, can comprehensively solve the excess flood diversion capacity of each zone of the lake, can optimize the flood control layout of the middle and lower reaches of the Yangtze River, can coordinate the relationship between the strategic development of the lake area and the flood control and safety, and can make the flood control capacity of the lake meet the flood storage and control requirements.

Description

Technical Field

[0001] This invention belongs to the field of water conservancy and flood control technology, specifically relating to a method for calculating excess flood diversion volume of lakes. Background Technology

[0002] Dongting Lake, the mother lake of Hunan and Hubei provinces, is my country's second largest freshwater lake and an important regulating lake for the Yangtze River. It is a vital ecological protection area and the foundation upon which the people living in the lake area depend for their livelihood. Located in the heart of the Changsha-Zhuzhou-Xiangtan urban agglomeration and the Wuhan metropolitan area, the Dongting Lake region has long been known as the "land of fish and rice."

[0003] The unique topography of Dongting Lake creates a complex network of rivers and lakes. The terrain south of Jingzhou is characterized by a high southwest and low northeast, dividing the lake into eastern, southern, and western sections, forming a complex water system. Furthermore, Dongting Lake receives water from three inlets—Songzi, Taiping, and Ouchi—which, after being regulated by the lake, flow into the Yangtze River at Chenglingji, further complicating the river-lake relationship. Since ancient times, there has been a saying that "Dongting is the water of the world," implying both abundant water and the potential for flooding; Dongting Lake is one of the areas in China most prone to flooding.

[0004] Since the 1960s, with the straightening of the Jingjiang River, the damming of the Gezhouba Dam, and especially the construction and operation of the Three Gorges Dam, upstream water and sediment conditions have been adjusted, altering the hydrological conditions of the Dongting Lake area. Given the mismatch between the safe discharge capacity of the middle and lower reaches of the river and the large peak flow from the upstream, as well as the mismatch between the flood control capacity of upstream reservoirs and the excess flood volume, the Dongting Lake area still faces a severe flood control situation during catastrophic floods. Currently, the flood control layout of the middle and lower reaches of the Yangtze River is determined by a large lake evolution hydrological mathematical model based on the water balance equation. This model, based on the flood characteristics of the middle and lower reaches of the Yangtze River, hydrological stations, and the complex river system distribution, determines the excess flood diversion volume near Chenglingji through channel storage and water level-discharge relationship curves.

[0005] As a crucial component of the flood control system in the middle and lower reaches of the Yangtze River, the storage and discharge capacity of Dongting Lake directly impacts the flood control layout of the middle and lower reaches of the Yangtze River. Despite the first and second phases of management, the flood control capacity of the Dongting Lake dikes and embankments still fails to meet the requirements for flood storage and defense. Summary of the Invention

[0006] To address the problem that lakes' flood control capacity does not meet flood storage and defense requirements, this invention provides a method for calculating excess flood diversion capacity of lakes, the method comprising the following steps: The lake is divided into zones, and a lake zone series flood evolution model is constructed by using the river section water balance equation and the relationship between water storage and outflow. A coupled model of runoff generation, runoff, and confluence is constructed based on the runoff generation, runoff, and confluence processes of rainfall. The runoff and precipitation flow rate into each zone of the lake was calculated using a runoff and precipitation coupling model. The storage capacity and outflow of each zone are calculated by using the inflow of each zone into the lake and the lake zone series flood evolution model. The excess flood diversion volume of each lake zone is quantitatively assessed based on the storage capacity and outflow of each zone.

[0007] Preferably, the production-confluence coupling model is as follows: In the formula: S ( t The total water volume of the unit includes surface free water, soil gravity water, and water from ditches and rivers; I ( t ) and ET ( t These represent infiltration and evaporation, respectively, within the unit. R In ( t This refers to rapid (surface or channel) runoff from upstream units; R Out ( t This refers to rapid (surface or channel) runoff flowing downstream; R S,In ( t This represents slow runoff from upstream units; R S,Out ( t () represents slow runoff flowing downstream.

[0008] Preferably, the formula for calculating the water balance equation of the river section is: In the formula, I j for j Inflow during a specific time period I j+1 for j Inbound traffic during the +1 time period Q j for j Outflow rate during a given time period, Q j+1 for j Outflow during the +1 time period For time / day, S j+1 for j The tank storage capacity during the +1 period S j for j Time period The storage capacity of the tank.

[0009] Preferably, the relationship between the water storage capacity and the outflow rate is as follows: In the formula, m The index of the flow-storage relationship curve. k The outflow coefficient is... .

[0010] Preferably, the calculation formula for the flood evolution model is: Preferably, the production-confluence coupling model is divided into multiple units using a DEM, and the production flow rate of each unit is calculated using a water storage capacity curve.

[0011] Preferably, the water storage capacity curve relationship is as follows: In the formula, W m It is the total water storage capacity of the three soil layers above the unit; EX It is to fully store production flow; B It is the index of the water storage capacity curve; W´ mm It is the maximum water storage capacity of the unit; W It is the value at a certain point on the water storage capacity curve; A It is on the curve and W The coordinates corresponding to the value.

[0012] Preferably, the runoff of each unit is divided into fast runoff and slow runoff based on soil permeability, and the calculation formulas for fast runoff and slow runoff are as follows: In the formula: f It is the average soil permeability per unit, in mm / h; EX S It is slow runoff; EX O It is rapid runoff.

[0013] Preferably, after assessing the excess flood diversion capacity of each lake zone based on the channel storage and outflow of each zone, the method further includes using the SCE-UA algorithm to calibrate the parameters of the flood evolution model step by step to verify the flood evolution model. The specific steps are as follows: (1) Algorithm initialization: Assume there is an n-dimensional problem to be optimized, with a total of p (p≥1) complexes participating in CCE evolution, where the number of vertices contained in each complex is denoted as m (m≥n+1). Then the number of sample points to be calculated is: ; (2) Sample point generation: S sample points are generated randomly within the feasible region. Calculate the function value for each sample point. ; (3) Sorting of sample points: Sort the S sample points in ascending order of function values. Arrange them, still denoted as , recorded as ; (4) Complex partitioning: Divide D into p complexes. Each complex contains m vertices, where , ) ,j=1,…,m;k=1,…, p ; (5) Complex evolution: Each complex is evolved according to the complex evolution algorithm CCE; (6) Complex shape mixing: The new point set after evolution is composed of all vertices m of each complex shape. Then, the new set is sorted in ascending order according to the method in step (3), and the CCE algorithm is repeated. The newly generated point set is then denoted as D. (7) Convergence judgment: If the convergence condition is met, stop; otherwise, return to step (4). (8) To avoid an infinite loop, the calculation shall stop when one of the following conditions is met: If the objective function still cannot be improved to the specified precision after multiple iterations, then the points corresponding to the current parameter values ​​reach the flat surface of the feasible region. After several consecutive iterations, the simulation accuracy did not improve and the parameter values ​​could not be changed significantly, so it was considered that the objective function had found the global optimum. Set a maximum number of loop iterations; stop the loop when the maximum number of iterations is reached.

[0014] The method for calculating excess flood diversion in lakes provided by this invention has the following beneficial effects: This invention constructs a lake zonal cascade flood evolution model using the river section water balance equation and the relationship between water storage and outflow. It also constructs a runoff-generating coupling model based on the runoff generation, runoff, and confluence processes of rainfall. This coupling model calculates the runoff generated by rainfall flowing into the lake, and then calculates the excess flood diversion volume of the lake using the runoff generated by each zone of the lake and the lake zonal cascade flood evolution model. By quantitatively assessing the flood diversion and storage capacity of lakes, this invention can comprehensively address the excess flood diversion volume borne by each zone of the lake, optimize the flood control layout of the middle and lower reaches of the Yangtze River, coordinate the strategic development of the lake area with flood control and security, and ensure that the flood control capacity of the lake meets the requirements for flood storage and defense. Attached Figure Description

[0015] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating an embodiment of the present invention; Figure 2 A schematic diagram of CREST unit generation and merging calculation; Figure 3 This is a schematic diagram of the water balance of the CREST unit; Figure 4 A generalized map of the river network in the study area; Figure 5 Analysis results diagram of Luoshan Station; Figure 6 The changes in the correlation between the flow distribution at each of the three outlets of the Jingjiang River and the flow in Zhicheng; Figure 7 Here is the flowchart for the SCE-UA algorithm; Figure 8 Map showing stations and waterways in the Luoshan area; Figure 9 Calibration results of parameters in the Luoshan section in 1996—flow rate and water level; Figure 10 Calibration results of parameters in the Luoshan section in 1998—flow rate and water level; Figure 11 Calibration results of parameters in the Luoshan section in 2017—flow rate and water level; Figure 12 Map showing the catchment area of ​​the Nanzui section; Figure 13 The simulation results of the flood evolution model in the Nanzui section in 1996—discharge rate and water level; Figure 14 The simulation results of the flood evolution model in the Nanzui section in 1998—discharge rate and water level; Figure 15 The simulation results of the flood evolution model in the Nanzui area in 2017—discharge rate and water level; Figure 16 Map showing the catchment area of ​​the Xiaozui section; Figure 17 Results of parameter calibration for the Xiaozui section in 1996—flow rate and water level; Figure 18 Calibration results of parameters for the Xiaozui section in 2017—flow rate and water level; Figure 19 Statistics on the total inflow between Zhicheng and Luoshan in 1954; Figure 20The outflow process at Luoshan Station in 1954 was calculated; Figure 21 Simulated outflow process at Nanzui Station in 1954; Figure 22 Simulated outflow process at Xiaohezui Station in 1954; Figure 23 Statistics on total inflow between Zhicheng and Luoshan in 1998; Figure 24 The outflow process at Luoshan Station in 1998 was calculated. Figure 25 The outflow process at Nanzui Station in 1998 was calculated. Figure 26 The outflow process at Xiaohezui Station in 1998 was calculated; Figure 27 Simulation of the impact of Three Gorges Dam dispatching on Luoshan Station. Detailed Implementation

[0017] Example Taking Dongting Lake as an example, this invention further illustrates a method for calculating excess flood diversion in lakes, as provided by this invention. The specific calculation method includes the following steps: Figure 1 As shown: Step 1: Divide Dongting Lake into zones and construct a zoned flood evolution model of Dongting Lake using the river section water balance equation and the relationship between water storage and outflow.

[0018] The water balance equations between the upper and lower cross sections of the river / lake section are as follows: (1) In the formula, I j for j Inflow during a specific time period I j+1 for j Inbound traffic during the +1 time period Q j for j Outflow rate during a given time period, Q j+1 for j Outflow during the +1 time period For time / day, S j+1 for j The tank storage capacity during the +1 period S j for j The storage capacity of the tank during a given time period.

[0019] Assuming a linear relationship between water storage and outflow, expressed as two functions of water storage and outflow with respect to water level Z, we can skip the intermediate variable Z and directly describe the relationship between flow rate and storage capacity using a linear reservoir. (2) In the formula: k is the outflow coefficient, It needs to be determined based on the analysis of measured data or the optimization of parameters.

[0020] S in formula (2) j+1 Substituting into formula (1) and rearranging, we get: In the formula, except All external quantities are known and can be solved directly.

[0021] For the initial outflow Q 0 can be directly equal to I 0 means that the model calculation should start from a low water level with stable flow, which is in line with the conventions of hydrological calculation.

[0022] The most common relationship between water storage S and outflow q at any given time in a watershed system is: (Q j+1 Δt) m =S j+1 k (4) In the formula: m The index of the flow-storage relationship curve. k It is the outflow coefficient. ,if m A system with a value of 1 is linear, for a river channel. m The range is typically from 0.6 to 1.0.

[0023] S in formula (4) j+1 Substituting into formula (1) and rearranging, we obtain the flood evolution model: In addition to the formula All external quantities are known quantities; outflow under non-real-time forecasting conditions. The solution can be obtained using the bisection method; for real-time forecasting, the upper limit of Q is taken as the historical extreme value.

[0024] Step 2: Construct a runoff-concentration coupling model based on the runoff generation process, runoff process, and confluence process of rainfall, and calculate the runoff volume of rainfall flowing into each zone of Dongting Lake through the runoff-concentration coupling model.

[0025] The Coupled Routing and Excess Storage (CREST) ​​model aims to achieve spatiotemporal simulation and distributed output of water cycle variables using unit-by-unit runoff. Driven by online satellite rainfall products, the model generates unit-by-unit flood warnings through calculations of evapotranspiration, infiltration, and runoff related to soil and vegetation. At lower spatial resolutions (e.g., 30 seconds or less), the unit-by-unit runoff module influences runoff calculations through triple feedback, resulting in more accurate simulations of soil moisture content, unit free water, etc. At higher spatial resolutions (e.g., 150 seconds or larger), the model utilizes features such as water storage capacity curves and linear reservoir outflow to describe the impact of subgrid inhomogeneities on runoff generation and distribution.

[0026] The CREST model uses a DEM (Digital Emission Model) to divide the watershed into numerous regular units. For each unit, rainfall-runoff is calculated using a water storage capacity curve, and fast and slow runoff are distinguished using soil stable infiltration rate. The main architecture is shown below. Figure 2 In the picture, R In It is rapid runoff from upstream units; R Out It is rapid runoff leaving this unit; R S，In It is slow runoff from the upstream unit; R S，Out It is slow runoff leaving this unit; IM It is the proportion of impermeable material; Th It is the slope / channel threshold; f It is the soil stable infiltration rate; EX The unit is fully charged with production flow; E c , E 1, E 2 and E 3 represents evaporation from the canopy and the three soil layers.

[0027] In the CREST model, in addition to the river unit, the slope unit also performs unit-by-unit runoff calculation. That is, through a two-layer planar architecture, after considering the influence of upstream water on the runoff generation and runoff process of the current unit, the calculation results of the current unit are superimposed and output downstream. Based on this, all watershed units are considered as an interconnected whole, and each unit has a reasonable simulated runoff result.

[0028] Convergence Time The inter-grid convergence time is calculated using the following formula: In the formula: t j The water flows from the first j The time required for a grid to flow to its downstream grid;l j It is the first j The distance from a grid center to its downstream grid center; v j It is the average velocity of the water flowing from the j-th grid to its downstream grid; S j It is the average slope from the j-th grid to its downstream grid; Kv This is the flow velocity adjustment coefficient, used to account for the influence of factors such as roughness and hydraulic radius on flow velocity. For a grid with a resolution of 30 seconds, cells with a water catchment threshold of 30 or higher can be treated as channel cells.

[0029] The above formula applies to three different situations: 0. Rapid runoff Kv Primarily determined by land cover; 0. Slow runoff Kv It is mainly determined by the lateral saturated hydraulic conductivity of the soil; 0. Regarding river runoff Kv It is mainly determined by roughness and hydraulic radius.

[0030] Kv It is a parameter with strong physical significance, which can be preset from measured values ​​or deduced in reverse during the parameter calibration process.

[0031] Division of runoff by time period When given a time period dT When it is larger, the first j The runoff from one unit may pass through several downstream units along the steepest slope, in which case: In the formula: t j+i yes j +i times the water flow from each unit needs to reach its downstream unit; n It is a counting indicator, which may be zero in special cases.

[0032] That means j After a period of time dT, the runoff of each unit will move downstream. j+n and j+n+ Between two units, at the instant at the end of the time interval, the first... j The runoff of each unit is superimposed j+n The water volume of the unit is: No. j The remaining runoff of each unit is superimposed on j+n+In Unit 1, considering any downstream unit, the unit water balance is as follows (see the CREST unit water balance diagram). Figure 3 (as shown) In the formula: S ( t The total water volume of the unit includes surface free water, soil gravity water, and water from ditches and rivers; I ( t ) and E ( t These represent infiltration and evaporation, respectively, within the unit. R In ( t This refers to rapid (surface or channel) runoff from upstream units; R Out ( t This refers to rapid (surface or channel) runoff flowing downstream; R S,In ( t This represents slow runoff from upstream units; R S,Out ( t () represents slow runoff flowing downstream.

[0033] evaporation The model needs to consider evaporation from both the canopy and soil, and soil evaporation also needs to take into account the soil's water supply capacity. When the input data only includes precipitation, the model will attempt to automatically acquire evaporation capacity data for the same period (Goal of the Famine Early Warning Systems Network (FEWS NET)).

[0034] When evaporation capacity exceeds precipitation, canopy water interception participates in evaporation first, followed by soil water loss. The effect of soil moisture content on evaporation is calculated using the following formula: In the formula: E canopy This is the water loss from the portion intercepted by the canopy; SAT It is the soil saturation water content; PET It refers to the evaporation capacity of the atmosphere; PWP It refers to the wilting moisture content of the soil; PET The data can be measured values ​​from an evaporation pan (which require evaporation pan correction and land / water surface correction), or relevant formulas can be selected based on meteorological data, such as the Priestley-Taylor formula or the Hargreaves formula.

[0035] abortion Each unit in the two-layer runoff architecture has an independent rainfall runoff generation structure. Given the better adaptability of conceptual rainfall runoff calculations, CREST chooses the full-sumption runoff theory to calculate rainfall runoff. For each unit, the net rainfall is first calculated:

[0036] P Net =P-PET (11) In the formula: P Net This is net rainfall; P It is the amount of rainfall; PET It is potential evaporation capacity.

[0037] Clean Rain P Net First, it is consumed by the canopy cutoff, calculated according to Dickinson's formula: In the formula: CI It is the canopy's retention capacity; d It refers to the degree of canopy closure; LAI It is the leaf area index; P Soil It is rainwater that falls onto the surface of the soil.

[0038] The flow rate calculation uses the Xin'anjiang model, assuming that free water is generated only after the soil pores are filled, and the water volume is: In the formula, W m It is the total water storage capacity of the three soil layers above the unit; EX It is to fully store production flow; B It is the index of the water storage capacity curve; W´ mm It is the maximum water storage capacity of the unit; W It is the value at a certain point on the water storage capacity curve; A It is on the curve and W The coordinates corresponding to the value.

[0039] Production flow rate through soil permeability f Divided into fast and slow runoff: In the formula: f It is the average soil permeability per unit, in mm / h; EX S It is slow runoff; EX O It is rapid runoff.

[0040] Generation and flow coupling At each time period, once the flow rate reaches full capacity, it is first stored in two virtual reservoirs: a fast reservoir and a slow reservoir. The reservoirs use linear outflow. O = K·S (15) In the formula: S This is the virtual reservoir's water storage capacity; O It's about generating traffic; K It is the outflow rate.

[0041] In the CREST model, the virtual reservoir can be replaced by an actual reservoir scheduling module. Furthermore, when the catchment area of ​​a unit exceeds a given threshold, rapid runoff escalates into river flow. For any given unit, its runoff calculation is also affected by the following factors:

[0042] (1) The rapid runoff from several upstream units will be superimposed on the net rainfall of the current unit and participate in the runoff generation calculation: P Soil =P Soil +∑R O,In (t) (16) In the formula: P Soil It is the net rainfall at ground level; R O，In It is the rapid runoff of the upstream unit.

[0043] (2) Slow runoff from several upstream units will increase the soil moisture content of the current unit: I(t)=I(t)+∑R S,In (t) (17) In the formula: I ( t () represents the amount of rainfall infiltrating into the current unit; R S，In It is the slow flow rate of the unit above.

[0044] (3) If the current unit is a river unit, since the evaporation of the river section is water evaporation, the total evaporation calculation for the unit is changed to: In the formula: PET It is potential evapotranspiration capacity; P It is possible precipitation (when calculating evaporation, precipitation is less than PET or zero); PWP It is the moisture content when the plant wilts; E canopy It is canopy evaporation; SAT It is the soil saturation water content; C w It is the ratio of unit water area.

[0045] The aforementioned triple feedback mechanism from flow convergence to flow generation is the most significant difference between CREST and other distributed models. This coupled flow generation and flow simulation can describe in detail the spatial variation of the storage area.

[0046] The parameters with physical mechanisms in the above formulas can be directly determined from measured data, as shown in Table 1. However, due to the deviation of measured data and the difference between laboratory data and production practice, these parameter values ​​should be used with caution.

[0047] Other conceptual parameters, shown in Table 2, require parameter calibration using measured data. Eight of these parameters are considered sensitive, while the other insensitive parameters help improve simulation detail.

[0048] Table 1. Parameters that can be directly determined Table 2 Parameters requiring calibration with measured data For small watersheds, parameter calibration is relatively easy, but for large watersheds, such as the global 0.25-degree distributed simulation which involves more than 3,000 watersheds, the workload is considerable.

[0049] The CREST model recommends the following solution: (1) Parameters are calibrated in batches to improve simulation results in a rolling manner; (2) An automatic parameter optimization algorithm is adopted. The model has an embedded demonstration method called Adaptive Random Search (ARS) for testing. However, it is a local parameter optimization algorithm and has not been optimized for computation speed.

[0050] Hydrological models typically only simulate the flow process at the watershed outlet. Because each cell of the CREST model can output relatively reasonable simulation results, CREST can output many spatial variables for subsequent analysis, as shown in Table 3. Each variable is a set of time-series two-dimensional matrices.

[0051] Table 3 Model Output Step 3: Calculate the channel storage and outflow of each zone by using the inflow into each zone of Dongting Lake and the lake zone series flood evolution model to quantitatively assess the excess flood diversion volume of each zone of Dongting Lake.

[0052] After the flood control capacity of reservoirs and other engineering projects in a river basin is utilized, the flood volume exceeding the safe discharge capacity of rivers and lakes is called excess flood volume. Excess flood volume and its spatiotemporal distribution are important bases for the layout of flood storage and detention areas in a river basin. Flood evolution models can simulate the flood evolution state under the complex river-lake relationship in the middle and lower reaches of the Yangtze River and quantitatively calculate the problem of excess flood volume.

[0053] Calculation range The large lake evolution model calculates the area from Yichang to Luoshan in the middle reaches of the Yangtze River, dividing Dongting Lake into West Dongting Lake, South Dongting Lake, and East Dongting Lake. The boundaries between these three lakes are defined by Nanzui, Xiaohezui, and Lujiao. The main stream of the middle reaches of the Yangtze River has a gentle gradient, with dispersed confluences of tributaries, interconnected rivers and lakes, and a complex river system with intricate river-lake relationships. The main river systems are summarized below:

[0054] The Qingjiang River flows into the Yangtze River between Yichang and Shashi. On the south bank of the Yangtze River, the Songzi River, Hudu River, Ouchi River, and Tiaoxian River (which was blocked by a sluice gate in the winter of 1958) divert floodwaters from the Yangtze River into Dongting Lake. In addition to receiving floodwaters diverted from the Yangtze River, Dongting Lake also stores the Xiang, Zi, Yuan, and Li rivers. After being regulated by the lake, the floodwaters flow into the Yangtze River at Chenglingji. Floodwaters from the rivers and lakes counteract each other near Chenglingji.

[0055] Calculate river segmentation Based on the distribution of Dongting Lake, hydrological observations, and waterway topographic survey data, the calculation interval is divided into four intervals. The total interval is the Zhicheng-Luoshan interval, with the upper boundary defined by the total inflow from Zhicheng, Xiangjiang, Zijiang, Yuanjiang, and Lishui rivers, and the outflow station being Luoshan station. The calculation result for this interval represents the excess flood volume near Chenglingji. The Lujiao interval uses the six stations of Songzi, Taiping, Shimen, Taoyuan, Taojiang, and Xiangtan as the total inflow, and Lujiao as the total outflow, which can be used to simulate the total flood diversion volume of East Dongting Lake. The flood diversion volume of West Dongting Lake is obtained from the Nanzui and Xiaohezui intervals. The Nanzui interval considers the inflow from Songzi, Taiping, and Shimen, while the Xiaohezui interval mainly considers the inflow from Yuanjiang River. The sum of the flood diversion volumes of the two intervals is the flood diversion volume of West Dongting Lake. The required flood diversion volume of South Dongting Lake is obtained from the difference between the total interval and the flood volumes of the eastern and western parts. The Dongting Lake zoning study area is as follows: Figure 4 As shown.

[0056] Calculation method This invention considers the Yichang-Luoshan section of the river and the entire Dongting Lake as a large lake system. The inflow is the sum of the water from the upper boundary and the lake area runoff, specifically the combined inflow from the main stream at Yichang, the Qingjiang Gaobazhou, the four water control stations of Dongting Lake, and the water volume from the interval. The outflow is the flow rate at the Luoshan section. Based on the principle of water balance, the changes in inflow, outflow, and storage volume are combined into a lake water balance system. Flood control calculations are performed using volumetric curves to obtain the water level process at Luoshan station. The model assumes that both storage volume and outflow are functions of water level. Before performing flood control calculations, the water level-discharge relationship at Luoshan station in the foreseeable period must be predicted. The flood evolution model generalizes the complex water level-discharge relationship at Luoshan station into a set of single-valued clusters of lines. During specific forecasting, the corresponding single-valued water level-discharge relationship lines (representing the future trend of the water level-discharge relationship) are selected based on the measured flow data at the time, and combined with the storage capacity curve to form seven calculation curves for flood control calculations, such as... Figure 5 As shown.

[0057] Three-channel flood diversion calculation Due to the three bend reductions in the lower reaches of the Jingjiang River and siltation in the Dongting Lake area, the sediment distribution ratios at the three outlets of the Jingjiang River showed a decreasing trend from 1955 to 1989. Affected by the construction of reservoirs and soil and water conservation projects along the upper reaches of the Yangtze River, the amount of sediment entering the Jingjiang River section showed a decreasing trend from 1990 to 2002, and correspondingly, the sediment distribution at the three outlets also showed a decreasing trend, but the flow rates at the three outlets did not show a significant change. After the Three Gorges Project impounded water in 2003, the flow rates at the three outlets of the Jingjiang River decreased slightly or remained basically unchanged in other years, except for the exceptionally dry years of 2006 and 2011, without any significant trend change. Analysis of the flow rate at Zhicheng Station exceeding 30,000 m³ / s at different stages over more than 60 years showed a positive correlation between the flow rates at the three outlets of the Jingjiang River and the flow rate at Zhicheng Station. During the flood season of 2020, the flow rate at Zhicheng Station increased to 38,000 m³ / s. 3 When the flow rate is above 100 m³ / s, the three-way diversion ratio is generally above 20%. Even during the peak flood season, when the flow rate at Zhicheng Station decreases significantly, the three-way diversion ratio is still significantly higher than that during non-peak flood seasons. For example, during the peak flood season on July 13th, the flow rate at Zhicheng Station was 20,500 m³ / s. 3 At a flow rate of / s, the diversion ratio at the three outlets reached as high as 20.7%, while during non-peak flood periods, the flow rate at Zhicheng Station on June 19 was 20900 m³ / s. 3 At / s, the diversion ratio at the three outlets is only 14.6%. The changes in the correlation between the diversion flow at each of the three outlets of Jingjiang and the flow in Zhicheng are as follows: Figure 6 As shown.

[0058] The flow rates at Songzikou, Taipingkou, and Ouchikou were calculated using the flow splitting ratio method.

[0059] The simplified relationship curve used by the system is as follows: 0) Songzikou: 0)Taipingkou: 0) Lotus pond entrance: Automatic parameter optimization Parameter optimization has always been a challenging problem of common concern to hydrologists both domestically and internationally. The Simple Polygon Evolutionary Algorithm (SCE-UA) is used for automatic parameter optimization in flood evolution models. In the 1990s, Duan et al. in the United States proposed the Simple Polygon Evolutionary Algorithm (SCE-UA), an algorithm capable of effectively solving nonlinearly constrained optimization problems. The SCE-UA algorithm was initially used to study parameter optimization in conceptual rainfall-runoff relationship simulations, primarily based on the characteristics of parameters such as multiple extrema, nonlinearity, interval constraints, and the lack of specific functional expressions. This algorithm combines the advantages of existing algorithms and can find the global optimum. Currently, this algorithm has been widely used in conceptual, semi-distributed, and distributed hydrological models.

[0060] The SCE-UA algorithm is based on the principle of using complex evolutionary search to study natural biological competition. The key to this algorithm is the complex evolutionary algorithm (CCE). When performing CCE calculations, each vertex of a complex is a potential parent of the next sub-complex and may participate in the calculation process of generating the next generation. During the construction of sub-complexes, the application of randomization allows for a more thorough search within the feasible region. The specific steps are as follows (e.g., ...). Figure 7 (as shown)

[0061] After quantitatively assessing the excess flood diversion volume in each zone of Dongting Lake, the process also includes using the SCE-UA algorithm to calibrate the parameters of the flood evolution model step by step to validate the model. The specific steps are as follows: (1) Algorithm initialization: Assume there is an n-dimensional problem to be optimized, with a total of p (p≥1) complexes participating in CCE evolution, where the number of vertices contained in each complex is denoted as m (m≥n+1). Then the number of sample points to be calculated is: ; (2) Sample point generation: S sample points are generated randomly within the feasible region. Calculate the function value for each sample point. ; (3) Sorting of sample points: Sort the S sample points in ascending order of function values. Arrange them, still denoted as , recorded as ; (4) Complex partitioning: Divide D into p complexes. Each complex contains m vertices, where , ) ,j=1,…,m;k=1,…, p ; (5) Complex evolution: Each complex is evolved according to the complex evolution algorithm CCE; (6) Complex shape mixing: The new point set after evolution is composed of all vertices m of each complex shape. Then, the new set is sorted in ascending order according to the method in step (3), and the CCE algorithm is repeated. The newly generated point set is then denoted as D. (7) Convergence judgment: If the convergence condition is met, stop; otherwise, return to step (4). (8) To avoid an infinite loop, the calculation shall stop when one of the following conditions is met: After multiple iterations, the objective function still cannot be improved to the specified precision. It can be assumed that the points corresponding to the current parameter values ​​have reached the flat surface of the feasible region. After several consecutive iterations, the simulation accuracy did not improve and the parameter values ​​could not be changed significantly, so it was considered that the objective function had found the global optimum. Set a maximum number of loop iterations; stop the loop when the maximum number of iterations is reached.

[0062] Partition parameter calibration To simulate flood distribution under different topographic conditions, hydrological data from 1996 (before the Three Gorges Dam was built), 1998 (a typical flood year), and 2017 (after the Three Gorges Dam was built and put into operation) were selected for zonal parameter calibration. Simulation results of the Dongting Lake area experiencing the 1954 and 1998 floods under the three topographic conditions were calculated.

[0063] Zhicheng-Luoshan section This section, abbreviated as (Zhicheng-Luoshan section), considers the main stream of the Yangtze River, the inflow from the four tributaries, and the inflow from within the section itself. Luoshan is the outflow point. The upper boundary is defined by Zhicheng Station, Shimen Station, Taoyuan Station, Taojiang Station, and Xiangtan Station, while the lower boundary is Luoshan Hydrological Station. The catchment area of ​​this section is 62,919 km², and the basin boundaries are as follows: Figure 8 As shown.

[0064] All hydrological data from 1996, 1998, and 2017 were input into the model, and the results after automatic parameter optimization are shown in Table 4. Figure 9 , Figure 10 and Figure 11 The results are as follows:

[0065] Table 4 Key parameters and optimization results for the Luoshan section Nanzui section This section, abbreviated as (Nanzui Section), considers the inflow from the Jingjiang River estuary, the Lishui River, and the section's own inflow. Nanzui is the outflow point. Its upper boundary is defined by Shadaoguan Station, Mituosi Station, and Shimen Station. The lower boundary is Nanzui Station. The catchment area of ​​this section is 11,714 km². The watershed boundaries of this section are as follows: Figure 12 As shown, the calculation results for the Nanzui section represent part of the flood storage capacity of West Dongting Lake.

[0066] All hydrological data from 1996, 1998, and 2017 were input into the model, and the results after automatic parameter optimization are shown in Table 5. Figure 13 , Figure 14 and Figure 15 The results are as follows:

[0067] Table 5 Key parameters and optimization results for the Nanzui section Xiaohezui section This section is referred to as the Xiaohezui section (see the Xiaohezui section catchment area diagram). Figure 16 As shown in the figure, considering the inflow from the Yuan River and the inflow from the inter-regional flow, with Xiaohezui as the outflow, the upper boundary is Taoyuan Station, and the lower boundary is Xiaohezui Station. The catchment area of ​​the inter-regional flow is 6,667 km². The sum of the Xiaohezui and Nanzui inter-regional flows is the total flood storage capacity of West Dongting Lake.

[0068] All hydrological data from 1996 and 2017 were input into the model, and the results after automatic parameter optimization are shown in Table 6. Figure 17 and Figure 18 The results are as follows:

[0069] Table 6 Key parameters and optimization results for the Xiaohezui section Typical flood zone simulation Based on the calibrated parameters (1996, 1998, and 2017), the flood evolution model was used to simulate the extraordinary floods of two typical years, 1954 and 1998, in different time intervals. This invention takes the measured hydrological condition calibration parameters of 1996 as an example to analyze and statistically analyze the flood simulation results.

[0070] Simulation of the 1954 flood Inputting tributary inflow data from 1954 for different intervals, and based on parameters calibrated in 1996, using a runoff generation and confluence model driven by measured precipitation from 80 rain gauge stations, the inflow for each interval was calculated. Figure 19 As shown, the flood evolution model is used to calculate the interval exit, and calculations are performed on the three intervals respectively.

[0071] Taking the parameter calibration results from 1996 as an example and inputting them into the 1954 flood data, the model simulates the outflow process lines of Luoshan, Nanzui, and Xiaohezui, as shown below. Figure 20 , Figure 21 and Figure 22 As shown in Tables 7, 8, and 9, the flood diversion period at Luoshan Station lasted 54 days, from June 30th to August 22nd, with a total flood volume exceeding the guaranteed water level of 47.18 billion m³. The flood diversion period at Nanzui Station lasted 21 days, from June 15th to August 20th, with a total flood volume of 4.38 billion m³. The flood diversion period at Xiaohezui Station lasted 10 days, from June 30th to August 6th, with a total flood volume of 1.92 billion m³.

[0072] Table 7. Calculation results of the lake-regional cascade flood evolution model for the Luoshan section in 1954. Table 8. Calculation results of the lake zone series flood evolution model for the Nanzui section in 1954. Table 9. Calculation results of the lake zone series flood evolution model for the Xiaozui section in 1954. 1998 flood simulation Inputting tributary inflow data from different intervals in 1998, and based on parameters calibrated in 1996, using a runoff generation and confluence model driven by measured precipitation from 80 rain gauge stations, the inflow for each interval was calculated. Figure 23 As shown, the flood evolution model is used to calculate the interval exit, and calculations are performed on the three intervals respectively.

[0073] Taking the parameter calibration results from 1996 as an example and inputting them into the 1998 flood, the outflow process lines of Luoshan, Nanzui, and Xiaohezui are simulated, as follows: Figure 24 , Figure 25 and Figure 26 As shown in Tables 10, 11, and 12. The results show that the flood diversion period at Luoshan Station was from July 16th to September 3rd, lasting 50 days, with a total flood volume exceeding the guaranteed water level of 18 billion m³. The flood diversion period at Nanzui Station was from July 4th to August 20th, lasting 8 days, with a total flood volume of 3.02 billion m³. The flood diversion period at Xiaohezui Station was from June 30th to August 6th, lasting 0 days, with a total flood volume of 0 billion m³.

[0074] Table 10 Calculation results of the 1998 Luoshan Intersection Flood Lake Zoning Series Flood Evolution Model Table 11 Calculation results of the 1998 Nanzui Intersection Flood Lake Zoning Series Flood Evolution Model Table 12 Calculation results of the 1998 Xiaohezui section flood lake zone series flood evolution model Impact of Three Gorges Dam Operation on Excess Flood Volume Near Chenglingji The upper boundary was selected at the Yichang, Shimen, Taoyuan, Taojiang, and Xiangtan hydrological stations, and the lower boundary was at the Luoshan station. Relevant parameters were calibrated using measured data from 1996. Measured data from 1954 and simulated flow rates after the Three Gorges Dam's operation were input, as shown in Table 13. Figure 27 As shown: Table 13 Simulation results of the impact of the Three Gorges Dam dispatch on Luoshan Station Analysis of the flood storage capacity of Dongting Lake A flood evolution model was used to simulate the extraordinary floods of two typical years, 1954 and 1998, by different zones. Taking the calibration parameters of the 1996 flood as an example, the excess flood volume near Chenglingji was 47.18 billion m³. According to the requirement of equal flood diversion between Honghu Lake and Dongting Lake, Honghu Lake on the north side would bear half of the flood diversion volume. Therefore, the flood diversion volume of Dongting Lake in the Zhicheng-Luoshan basin was approximately 23.6 billion m³. Among them, Nanzui + Xiaohezui accounted for 6.3 billion m³, and the flood diversion volume of East + South Dongting Lake was 17.3 billion m³. The proportion of flood diversion volume in the eastern and western parts of the lake area in 1954 was approximately 73.3% and 26.7%, respectively.

[0075] Table 14. Flood evolution model simulation results for the 1954 and 1998 floods by interval. The above-described embodiments are merely preferred embodiments of the present invention, and the scope of protection of the present invention is not limited thereto. Any simple changes or equivalent substitutions of the technical solutions that can be obviously obtained by those skilled in the art within the scope of the technology disclosed in the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for calculating excess flood diversion capacity of a lake, characterized in that, Includes the following steps: The lake is divided into zones, and a lake zone series flood evolution model is constructed by using the river section water balance equation and the relationship between water storage and outflow. A coupled model of runoff generation, runoff, and confluence is constructed based on the runoff generation, runoff, and confluence processes of rainfall. The runoff and precipitation flow rate into each zone of the lake was calculated using a runoff and precipitation coupling model. The storage capacity and outflow of each zone are calculated by using the inflow of each zone into the lake and the lake zone series flood evolution model. The excess flood diversion capacity of each lake zone is assessed based on the channel storage capacity and outflow of each zone. The generation and confluence coupling model is as follows: In the formula: S ( t The total water volume of the unit includes surface free water, soil gravity water, and water from ditches and rivers; I ( t ) and ET ( t These represent infiltration and evaporation, respectively, within the unit. R In ( t This refers to rapid surface or channel runoff from upstream units; R Out ( t This refers to rapid surface or river runoff flowing downstream; R S,In ( t This represents slow runoff from upstream units; R S,Out ( t This refers to slow runoff flowing downstream; The production-confluence coupling model uses a DEM to divide the flow rate into multiple units, and the production flow rate of each unit is calculated using a water storage capacity curve. The formula for the water storage capacity curve is: In the formula, W m It is the total water storage capacity of the three soil layers above the unit; EX It is to fully store production flow; B It is the index of the water storage capacity curve; W´ mm It is the maximum water storage capacity of the unit; W It is the value at a certain point on the water storage capacity curve; A It is on the curve and W The coordinates corresponding to the value; The runoff of each unit is divided into fast runoff and slow runoff based on soil permeability, and the calculation formulas for fast runoff and slow runoff are as follows: In the formula: f It is the average soil permeability per unit, in mm / h; EX S It is slow runoff; EX O It is rapid runoff.

2. The method for calculating excess flood diversion volume of lakes according to claim 1, characterized in that, The formula for calculating the water balance equation of the aforementioned river section is as follows: In the formula, I j for j Inflow during a specific time period I j+1 for j Inbound traffic during the +1 time period Q j for j Outflow during a specific time period Q j+1 for j Outflow during the +1 time period For time / day, S j+1 for j The tank storage capacity during the +1 period S j for j The storage capacity of the tank during a given time period.

3. The method for calculating excess flood diversion volume of lakes according to claim 2, characterized in that, The relationship between the water storage capacity and the outflow rate is as follows: In the formula, m 1 represents the index of the flow-storage relationship curve. k The outflow coefficient is... .

4. The method for calculating excess flood diversion volume of lakes according to claim 3, characterized in that, The calculation formula for the flood evolution model is as follows: 。 5. The method for calculating excess flood diversion volume of lakes according to claim 1, characterized in that, After assessing the excess flood diversion capacity of each lake zone based on the channel storage and outflow of each zone, the process further includes using the SCE-UA algorithm to calibrate the parameters of the flood evolution model step by step to verify the flood evolution model. The specific steps are as follows: (1) Algorithm initialization: Assume there is an n-dimensional problem to be optimized, with a total of p complexes participating in CCE evolution, p≥1. Let m be the number of vertices contained in each complex, m≥n+1. Then the number of sample points is calculated as follows: ; (2) Sample point generation: S sample points are generated randomly within the feasible region. Calculate the function value for each sample point. ; (3) Sorting of sample points: Sort the S sample points in ascending order of function values. Arrange them, still denoted as , recorded as ; (4) Complex partitioning: Divide D into p complexes Each complex contains m vertices, where , ) ,j=1,…,m;k=1,…, p ; (5) Complex evolution: Each complex is evolved according to the complex evolution algorithm CCE; (6) Complex shape mixing: The new point set after evolution is composed of all vertices m of each complex shape. Then, the new set is sorted in ascending order according to the method in step (3), and the CCE algorithm is repeated. The newly generated point set is then denoted as D. (7) Convergence judgment: If the convergence condition is met, stop; otherwise, return to step (4). (8) To avoid an infinite loop, the calculation shall stop when one of the following conditions is met: If the objective function still cannot be improved to the specified precision after multiple iterations, then the points corresponding to the current parameter values ​​reach the flat surface of the feasible region. After several consecutive iterations, the simulation accuracy did not improve and the parameter values ​​could not be changed significantly, so it was considered that the objective function had found the global optimum. Set a maximum number of loop iterations; stop the loop when the maximum number of iterations is reached.

Images